British Journal of Surgery

doi: 10.1002/bjs.10067pmid: N/A

Article PDF first page preview Close This content is only available as a PDF. © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd

Lerut, J; Foss, A

doi: 10.1002/bjs.9937pmid: 26416704

Liver transplantation (LT) for malignant liver diseases represents up to 50 per cent of all indications for this operation. Hepatocellular cancer is the most common of these indications and, when the Milan criteria are applied, 5-year disease-free survival (DFS) rates of 75–85 per cent have been obtained. Recently, these criteria have been widened successfully, based on both tumour morphology and biology combined with different locoregional treatments designed to downstage disease. The success of LT in the treatment of primary liver tumours has led to renewed interest in its role as a treatment for secondary liver tumours (SLTs)1–4. Standard treatment for SLTs includes a combination of chemotherapy and liver resection. Lately, liver resectability has been increased markedly as a result of more effective chemotherapy, parenchyma-sparing operations and/or staged surgical procedures. Further extension of survival can also be achieved by incorporating ablative procedures, arterial embolization, and hormone and radiolabelled chemotherapies into treatment protocols including resection. Despite these advances, curative liver resection remains applicable in only 20 per cent of patients, owing mainly to multifocality and the extent of bilobar disease. Total hepatectomy followed by transplantation for liver-only secondaries is by definition an R0 procedure and is therefore a potentially curative therapy. For metastases from neuroendocrine tumours (NETs), 5-year overall survival and DFS rates ranging from 36 to 92 per cent and from 9 to 78 per cent respectively have been reported. For colorectal metastases, 5-year overall survival rates of 60 per cent have been reached1,2. These results are comparable to those of state-of-the-art oncological treatments and therefore merit further investigation based on refined selection criteria and neoadjuvant and adjuvant therapies. Gastroenteropancreatic NETs are often diagnosed at an advanced stage. Fortunately the metastatic disease often remains confined to the liver, so R0 primary tumour resection and liver resection may represent a curative option. Liver resection is rarely appropriate, however, as detailed intraoperative imaging and thin-slice examination of the resected parenchyma often reveals many more lesions than those identified by routine imaging and pathology5,6. Until recently LT experience for NET was based on small single-centre and heterogeneous multicentre experiences. Recent reports, however, have shed new light on the value of LT1,7. The European Liver Transplant Registry1 and Milan7 series, including 213 and 24 recipients respectively, identified a number of factors as a prerequisite for success: presence of a well differentiated, low-grade tumour with Ki-67 less than 5–10 per cent; primary tumour localization within the portal venous drainage area; tumour burden less than 50 per cent of liver volume; age below 55 years; response to pre-LT treatment with somatostatin analogues and m-Tor inhibitors; and stable or controlled disease for at least 6 months. To exclude extrahepatic disease reliably, sensitive scintigraphy (for example with the octreotide derivative, DOTATOC) should be used. By adhering to these criteria, 5-year overall survival rates were 92 and 79 per cent, and DFS rates were 75 and 57 per cent. Careful post-transplant follow-up, including PET–CT and monitoring of tumour markers, is considered mandatory to enable reintervention for recurrence. As no other treatment has produced similar results for this stage of disease, it is time for the oncological community to integrate LT into the therapeutic NET algorithm3. Colorectal cancer is diagnosed in more than 700 people per million each year worldwide. Many either have liver metastases at presentation or are destined to develop them later. Only 15–20 per cent of those with apparent metastatic disease are candidates for curative surgery following state-of-the-art chemotherapy alone or in conjunction with volume-enhancing radiological or surgical procedures. Successful hepatic resection may result in 5-year overall survival rates of 30–40 per cent8. In a pilot study2 of LT, including patients with colorectal SLTs, 1-, 3- and 5-year overall survival rates of 96, 70 and 60 per cent respectively were obtained. Favourable parameters for survival were: diameter of the largest metastasis less than 55 mm; time interval more than 2 years between colorectal and transplant operations; pre-LT carcinoembryonic antigen level less than 80 ng/ml; and responsive or stable disease under chemotherapy. With aggressive surgical treatment of recurrences, eight of 21 patients were alive 4–8 years after LT and all four with metachronous SLTs were alive 5–8 years after transplantation. These results are markedly superior to those obtained with conventional treatment in similar patients3. It is of interest to note that retrospective analysis of pulmonary micronodules at the time of LT, which later turned out to be lung metastases, did not have a negative impact on overall survival. The results of this pilot LT study should be interpreted with caution owing to the small number of patients, the absence of a control group and a high rate of relapse with pulmonary metastases. It seems that LT may become a valuable treatment for NETs and colorectal SLTs. Despite the development of sequential liver resections and liver volume regenerating procedures, a large number of SLTs will remain unresectable. In some well selected patients, LT may be the answer. Sporadic attempts at LT for uveal malignant melanomas have been made, but with dismal outcome. Few other malignant diseases spread mainly to the liver alone, and transplantation should thus be avoided. It is obvious that the utilization of liver allografts needs careful consideration owing to their limited availability. Using a donor liver in patients with SLTs cannot be viewed as unethical or as a waste of resource in the light of current evidence. It is important to remember in this context that up to 10 per cent of all LTs are retransplants, after which the 5-year overall survival rate reaches around 50 per cent, a figure that seems achievable with LT in selected patients with SLTs. It is predicted that the pool of liver allografts available for liver malignancies will increase as a result of a reduced need for LT and retransplantation in patients with hepatitis C virus (HCV) infection, as a consequence of effective antiviral target therapies in HCV-infected patients with native and transplanted livers9. Novel LT techniques, such as the ‘rapid concept’ combining two-stage hepatectomy, right portal vein transection and regeneration of an auxiliary left lobe liver graft (segments II and III), must be explored. This approach may be applied in split-liver (1 donor liver for 2 recipients) or reduced risk, living donor LT procedures10. Taking recent progress in both medical and surgical liver oncology into account, the time has come for oncologists and liver surgeons to consider LT in the therapeutic algorithm of SLTs. Disclosure The authors declare no conflict of interest. References 1 Le Treut YP , Grégoire E, Klempnauer J, Belghiti J, Jouve E, Lerut J et al. ; for ELITA . Liver transplantation for neuroendocrine tumors in Europe – results and trends in patient selection: a 213-case European Liver Transplant Registry study . Ann Surg 2013 ; 257 : 807 – 815 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Hagness M , Foss A, Line PD, Scholz T, Jørgensen PF, Fosby B et al. Liver transplantation for nonresectable liver metastases from colorectal cancer . Ann Surg 2013 ; 257 : 800 – 806 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Foss A , Lerut JP. Liver transplantation for metastatic liver malignancies . Curr Opin Organ Transplant 2014 ; 19 : 235 – 244 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Hackl C , Schlitt HJ, Kirchner GI, Knoppke B, Loss M. Liver transplantation for malignancy: current treatment strategies and future perspectives . World J Gastroenterol 2014 ; 20 : 5331 – 5344 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Elias D , Lefevre JH, Duvillard P, Goéré D, Dromain C, Dumont F et al. Hepatic metastases from neuroendocrine tumors with a ‘thin slice’ pathological examination: they are many more than you think … Ann Surg 2010 ; 251 : 307 – 310 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Sarmiento JM , Heywood G, Rubin J, Ilstrup DM, Nagorney DM, Que FG. Surgical treatment of neuroendocrine metastases to the liver: a plea for resection to increase survival . J Am Coll Surg 2003 ; 197 : 29 – 37 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Mazzaferro V , Pulvirenti A, Coppa J. Neuroendocrine tumours metastatic to the liver: how to select patients for liver transplantation? J Hepatol 2007 ; 47 : 460 – 466 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Kanas GP , Taylor A, Primrose JN, Langeberg WJ, Kelsh MA, Mowat FS et al. Survival after liver resection in metastatic colorectal cancer: review and meta-analysis of prognostic factors . Clin Epidemiol 2012 ; 4 : 283 – 301 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 9 Afdhal N , Zeuzem S, Kwo P, Chojkier M, Gitlin N, Puoti M et al. ; ION-1 Investigators . Ledipasvir and sofosbuvir for untreated HCV genotype 1 infection . N Engl J Med 2014 ; 370 : 1889 – 1898 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Line PD , Hagness M, Berstad A, Foss A, Dueland S. A novel concept for partial liver transplantation in nonresectable colorectal liver metastases: the RAPID concept . Ann Surg 2015 ; 262 : e5 – e9 . Google Scholar Crossref Search ADS PubMed WorldCat © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd

Cahill, R A; Hompes, R

doi: 10.1002/bjs.9933pmid: 26694990

The latter half of the 20th century saw broad standardization of complex surgical resections for many cancers across all specialties. Total mesorectal excision (TME) of rectal cancer is a typical example of this maturation of intervention beyond simple feasibility1. It proved that correct and fastidious operative understanding and respect of embryological planes could improve disease-specific outcomes and also allow clear communication of critical concepts, representing an advance that could be reproduced across different healthcare systems2. TME brought qualitative assessment of the resection specimen into focus, initially as an indicator and then a driver of surgical quality improvement3. Close interaction with advanced imaging technologies with increasingly accurate delineation of the extent of disease also allowed clearer distinct identification of suboptimal technique4. Through its standardized surgical baseline quality, TME allowed the role of adjuvant and neoadjuvant chemoradiotherapy to be established5. This focus on an oncological package resection also facilitated the logical development of laparoscopic access, providing its proponents with reassurance that the same operation could be performed via reduced abdominal wall access even before the availability of long-term outcome data6. Low rectal access from the abdominal compartment, whether open or laparoscopic, however, remains difficult, most notably in men with increased body mass index, given the constraints of operating within a bony compartment placed at right angles to the peritoneal cavity. This difficulty has led to the emergence of novel technologies, skills sets and concepts (including transanal microsurgery, single-port laparoscopy and natural-orifice transluminal endoscopic surgery) to address the lower two-thirds of the rectum and mesorectum from below, starting the operation by a transanal or transperineal/intersphincteric approach. The past 5 years has seen these operations move from theory to practical reality so that transanal TME (taTME) is now sufficiently mature to be examined critically with a view to wider adoption (Video S1, supporting information). Its development, along the IDEAL (Innovation, Development, Exploration, Assessment, Long-term study) framework7, has gone from initial clinical experiences8,9 through case–control cohort series10 into multicentre trials. It can be taught using standardized nomenclature and an educational curriculum. A key component of this has been an open international registry, run via the LOREC (LOw REctal Cancer) national development programme and the Pelican Cancer Foundation in the UK11, which now has over 400 patients recorded to share experience and truncate learning curve difficulties among the colorectal community12. In concept and early practice, it seems likely that taTME has most potential value for tumours of the mid-rectum (3–4 cm above the anal sphincter), especially when there are concerns about anterior clearance. The exciting prospect, therefore, is that this approach, whether performed by two teams working in tandem or by sequential above and below surgery, will complement and advance existing approaches for patient benefit. The advent of the COLOR III randomized trial13, scheduled to commence later this year specifically to assess its potential impact on reducing circumferential margin positivity compared with that after conventional resection, represents a level of acceptance of both potential and practicality among the colorectal surgical community, reflecting evidence from secondary analysis of the COLOR II study14. If extralevator and intersphinteric approaches can be shown to be valuable, further modifications might identify patient subsets benefiting from a specific approach. Although generally seen as complementary to laparoscopic resection from above, taTME may also assist in the procedure commenced by, or converted to, laparotomy. Proponents of robotic surgery might argue that the control arm for randomization against taTME should be robot-assisted low anterior resection, in which increased freedom of movement overcomes some of the limitations of conventional laparoscopic surgery in the pelvis. Although preliminary results from the ROLARR (Robotic versus Laparoscopic Resection for Rectal Cancer) trial failed to reach statistical significance in its primary endpoints, improved outcomes were seen in certain subgroups of patients with distal rectal cancer (namely men and obese patients). The case for robotic assistance in these challenging patients is therefore also worth advancing. Perhaps a combination of these two innovative techniques may yet prove the most compelling solution for optimal TME dissection in difficult patients with low rectal cancer15. TME itself remains a complex and technically demanding procedure no matter how the pelvis is accessed. There is still a need for further study with taTME, including communication of the technical steps and a clearer understanding of the role that other novel technologies might play in optimizing outcomes, as well as the development of criteria for surgeon, centre and patient selection16. From a purely surgical point of view, the operator must be as certain about the recognition of correct operative planes and critical structures from below as conventionally seen from above, to ensure that new possibilities for mishap and misadventure are minimized. It is here that new technology can help, whether in the form of improved surgical imaging in the planning phase, intraoperative adapted smart insufflation systems (moving away from simple pressure-fixed, intermittently monitored sets towards continuous flow, volume-fixed circuits) or augmented optical imaging (including three-dimensional, stereotactic augmented reality17 and extended electromagnetic field techniques such as near-infrared laparoscopy18). There is also a need to consider methods for operative and concept skill advancement, such as masterclass teaching, mentorship/proctorship and cadaveric dissection training, all of which might improve performance and lead to better outcomes. Integration of novel surgical technologies and operative strategies into stratified or individualized patient care is important and likely to be increasingly so in the future. Genetic profiling of disease will need to be considered alongside selection of the surgical approach that gives the best operative access for the patient's body habitus, general condition and extent of disease. Understanding the benefits of a particular approach under specific circumstances means that it can be built into care pathways that include patient condition optimization via prehabilitation during diagnostic and neoadjuvant therapy windows. Carefully planned studies to evaluate the impact of changes within a care pathway as new capabilities come along should be used to identify real advances, as distinct from items that are merely fashionable. Disclosure The authors declare no conflict of interest. Supporting information Additional supporting information may be found in the online version of this article: Video S1 State-of-the-art lecture on transanal total mesorectal excision by R. Hompes detailing the evolution of the access approach as well as evidence, technique and next steps required. References 1 Heald RJ , Husband EM, Ryall RD. The mesorectum in rectal cancer surgery – the clue to pelvic recurrence? Br J Surg 1982 ; 69 : 613 – 616 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Wibe A , Møller B, Norstein J, Carlsen E, Wiig JN, Heald RJ et al. ; Norwegian Rectal Cancer Group . A national strategic change in treatment policy for rectal cancer – implementation of total mesorectal excision as routine treatment in Norway. A national audit . Dis Colon Rectum 2002 ; 45 : 857 – 866 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Quirke P Training and quality assurance for rectal cancer: 20 years of data is enough . Lancet Oncol 2003 ; 4 : 695 – 702 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Shihab OC , Heald RJ, Rullier E, Brown G, Holm T, Quirke P et al. Defining the surgical planes on MRI improves surgery for cancer of the low rectum . Lancet Oncol 2009 ; 10 : 1207 – 1211 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Nielsen LB , Wille-Jørgensen P. National and international guidelines for rectal cancer . Colorectal Dis 2014 ; 16 : 854 – 865 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Vennix S , Pelzers L, Bouvy N, Beets GL, Pierie JP, Wiggers T et al. Laparoscopic versus open total mesorectal excision for rectal cancer . Cochrane Database Syst Rev 2014 ; ( 4 ) CD005200 . Google Scholar OpenURL Placeholder Text WorldCat 7 McCulloch P , Altman DG, Campbell WB, Flum DR, Glasziou P, Marshall JC et al. ; Balliol Collaboration . No surgical innovation without evaluation: the IDEAL recommendations . Lancet 2009 ; 374 : 1105 – 1112 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Leroy J , Barry BD, Melani A, Mutter D, Marescaux J. No-scar transanal total mesorectal excision: the last step to pure NOTES for colorectal surgery . JAMA Surg 2013 ; 148 : 226 – 230 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Atallah S , Martin-Perez B, Albert M, deBeche-Adams T, Nassif G, Hunter L et al. Transanal minimally invasive surgery for total mesorectal excision (TAMIS-TME): results and experience with the first 20 patients undergoing curative-intent rectal cancer surgery at a single institution . Tech Coloproctol 2014 ; 18 : 473 – 480 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Fernández-Hevia M , Delgado S, Castells A, Tasende M, Momblan D, Díaz del Gobbo G et al. Transanal total mesorectal excision in rectal cancer: short-term outcomes in comparison with laparoscopic surgery . Ann Surg 2015 ; 261 : 221 – 227 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Pelican Cancer Foundation . Transanal Total Mesorectal Excision (TATME) Registry . http://www.lorec.nhs.uk/system/content.asp?pkey=1 [accessed 21 July 2015]. 12 Hompes R , Arnold S, Warusavitarne J. Towards the safe introduction of transanal total mesorectal excision: the role of a clinical registry . Colorectal Dis 2014 ; 16 : 498 – 501 . Google Scholar Crossref Search ADS PubMed WorldCat 13 COLOR III Study Group . COLOR III Trial . http://rectalcancersurgery.eu/color-3-trial/professionals/color-iii/[accessed 22 July 2015]. 14 Bonjer HJ , Deijen CL, Abis GA, Cuesta MA, van der Pas MH, de Lange-de Klerk ES et al. ; COLOR II Study Group . A randomized trial of laparoscopic versus open surgery for rectal cancer . N Engl J Med 2015 ; 372 : 1324 – 1332 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Hompes R Robotics and transanal minimal invasive surgery (TAMIS): the ‘sweet spot’ for robotics in colorectal surgery? Tech Coloproctol 2015 ; 19 : 377 – 378 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Miskovic D , Foster J, Agha A, Delaney CP, Francis N, Hasegawa H et al. Standardization of laparoscopic total mesorectal excision for rectal cancer: a structured international expert consensus . Ann Surg 2015 ; 261 : 716 – 722 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Buchs NC , Hompes R. Stereotactic navigation and augmented reality for transanal total mesorectal excision? Colorectal Dis 2015 ; 17 : 825 – 827 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Cahill RA , Ris F, Mortensen NJ. Near-infrared laparoscopy for real-time intra-operative arterial and lymphatic perfusion imaging . Colorectal Dis 2011 ; 13 ( Suppl 7 ): 12 – 17 . Google Scholar Crossref Search ADS PubMed WorldCat © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd

Day, R W; Fielder, S; Calhoun, J; Kehlet, H; Gottumukkala, V; Aloia, T A

doi: 10.1002/bjs.9918pmid: 26364714

Abstract Background Enhanced recovery (ER) protocols are used widely in surgical practice. As protocols are multidisciplinary with multiple components, it is difficult to compare and contrast reports. The present study examined compliance and transferability to clinical practice among ER publications related to colorectal surgery. Methods PubMed, Embase and Cochrane Central Register databases were searched for current colorectal ER manuscripts. Each publication was assessed for the number of ER elements, whether the element was explained sufficiently so that it could be transferred to clinical practice, and compliance with the ER element. Results Some 50 publications met the reporting criteria for inclusion. A total of 22 ER elements were described. The median number of elements included in each publication was 9, and the median number of included patients was 130. The elements most commonly included in ER pathways were early postoperative diet advancement (49, 98 per cent) and early mobilization (47, 94 per cent). Early diet advancement was sufficiently explained in 43 (86 per cent) of the 50 publications, but only 22 (45 per cent) of 49 listing the variable reported compliance. The explanation for early mobilization was satisfactory in 41 (82 per cent) of the 50 publications, although only 14 (30 per cent) of 47 listing the variable reported compliance. Other ER elements had similar rates of explanation and compliance. The most frequently analysed outcome measures were morbidity (49, 98 per cent), length of stay (47, 94 per cent) and mortality (45, 90 per cent). Conclusion The current standard of reporting is frequently incomplete. To transfer knowledge and facilitate implementation of pathways that demonstrate improvements in perioperative care and recovery, a consistent structured reporting platform is needed. Introduction Although fast-track protocols for surgical care were first described over 20 years ago, it is only in the past 5 years that fast-track and enhanced recovery (ER) protocols have penetrated wide areas of surgical practice1,2. As ER protocols are multidisciplinary and have multiple components, comparison of reports from different centres and across different surgical specialties is not easy. It seems appropriate that reports should describe the various elements of each ER pathway consistently and with sufficient detail that they can be reproduced elsewhere3. Only through such a rigorous approach will clinicians be able to compare outcomes associated with individual programmes. Without complete descriptions, the ability to translate positive results to other centres will be impossible, and the ability to capitalize on the current speed of information transmission will be lost. The introduction of a standard reporting template for clinical trials has resulted in a significant improvement in the ability to interpret and translate data from these trials into clinical practice4,5. Development of a similar structured template for the publication of studies examining outcomes for patients treated with ER protocols is likely to have similar benefits. The purpose of this study was to examine publications related to ER with colorectal surgery, seeking critically to assess the depth and breadth of description in the protocols used. Colorectal surgery was chosen, as it is the area containing the greatest number of publications. Evidence of variability in transparency of describing individual ER elements would strengthen the arguments for a standard reporting data set for all ER studies. This structured reporting would improve the ability to compare outcomes between experiences, translation of protocols from the author's institution to the reader's institution, and provide a structure for quality assessment of future publications. Methods A systematic review of English-language publications was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement6. PubMed, Embase and Cochrane Central Register of Controlled Trials databases were searched for the following specific terms: (‘Fast Track’ OR ‘Enhanced Recovery’ OR ‘ERAS’ OR ‘Accelerated Recovery’ OR ‘Multimodal Rehabilitation’) AND (‘Colon’ OR ‘Rectum’ OR ‘Colorectal’) AND (‘Surgery’ OR ‘Operation’ OR ‘Procedure’). The primary search was carried out by a single researcher, with vetting of the studies done in conjunction with a second researcher. Any discrepancies were resolved by an independent third party. Results from an initial search strategy of the three databases were cross-referenced and duplicates removed to create a single list. Secondary filtering was then performed to eliminate older publications (earlier than 2009) and non-original manuscripts (meta-analyses and reviews). The remaining records were subjected to full-text review and included in the final analysis if they met the following criteria: report comparing outcomes between at least two cohorts (1 of which must have been an ER pathway cohort and 1 of which must have been a conventional or traditional pathway cohort); inclusion of at least 50 patients; and a confirmed focus on colectomy and/or proctectomy. After eliminating publications with significant overlap in patient population and the same primary or senior authorship, a final analysis list was created and assessed in detail. Articles were entered into an electronic database and their content was analysed systematically to determine which elements of ER were stated as being used, whether the stated element was explained, the compliance rate of that element within the study, and the failure rate of the element when applicable. Although all elements were catalogued, particular attention was made to elements with strong evidence for safety and efficacy7–9. Additionally, the International Enhanced Recovery After Surgery (ERAS®) Society Registry variables were reviewed to determine which of the elements in this repository were included in the published literature and could therefore potentially facilitate audits for future publications. To be given credit for naming an ER element, a publication had to list the element as being involved in the authors' ER pathway in one of the following locations: the manuscript text; a figure, chart or table within the main publication, appendices or supplements; or an explicit reference to another publication's pathway. If a reference to an outside publication or external website was quoted, this resource was assessed as an extension of the originally reviewed paper. The total number of ER elements listed in each publication was determined by the number of discrete elements identified in this way. An element did not have to be implemented in exactly the same manner in different publications to count as the same ER element. Once an ER element had been determined as named by the publication, it was further reviewed to determine whether the element had been explained in detail. Although subjective, the criterion for adequate explanation of an ER element was that the description allowed for sufficient understanding to the level that the reader felt confident the element could be implemented based on the description. No judgement was made regarding the adequacy or appropriateness of each intervention, only whether it was described sufficiently. For example, a publication that stated ‘preoperative medications were given’ but did not list the medications was not counted as having explained this ER element sufficiently. Publications were also reviewed to determine whether a corresponding element was described adequately as part of a traditional perioperative pathway, to determine what the ER intervention was changing in the authors' practice. Level of compliance with the ER interventions was then determined by identifying the number of patients who had received the intervention or evidence that the intervention had been implemented universally across the patient population. For example, a publication that listed the use of oral non-narcotic medications in an effort to limit narcotic administration could satisfy compliance by either listing the proportion of patients managed in this manner or by comparing opioid utilization in the ER group versus the traditional group. To assess ER elements that were subject to failure, transparent reporting of failure rates on an element-by-element basis was sought, covering five specific elements. Where nasogastric tube elimination was a component of the ER protocol, was the rate of postoperative nasogastric tube placement disclosed? For those that listed regional/epidural analgesia use, did they indicate the rate of early catheter non-function or haemodynamic lability leading to conversion to intravenous narcotic strategies? Among those that listed early advancement of oral intake, was the rate of postoperative ileus leading to failure of dietary progression stated? Where early and frequent mobilization were evaluated, was failure to progress physically, including adverse events such as falls, reported? Finally, where early urinary catheter removal was included, was the rate of urinary retention requiring catheter replacement stated? Additional information regarding study design, numbers of patients included in each trial, outcomes analysed, whether the number of surgeons or anaesthetists in the practice was listed, implementation process of the ER pathway described, whether discharge criteria were described along with location of the ER pathway, was also recorded and analysed. Quality assessment was performed by using the Downs and Black tool10, as recommended previously for systematic reviews, to address both randomized and non-randomized studies11. Downs and Black scores were grouped into three quality levels: good (20 or more), fair (15–19) and poor (14 or less). Statistical analysis Data regarding ER elements contained in each publication were descriptive in nature and are shown as numbers with percentages. Comparative data across studies are reported as median (range). Results The initial search strategy returned 569 results from PubMed, 1256 from Embase and 161 results from Cochrane reviews. After cross-referencing and removal of duplicates, 1256 publications were identified. Secondary filtering to eliminate older and non-original manuscripts left 527 records that were subject to full-text review. After eliminating 469 articles that did not meet the inclusion criteria and eight further publications with overlap in patient population and the same primary or senior author, 50 publications were identified that met the inclusion criteria; these were assessed in detail (Fig. 1). Fig. 1 Open in new tabDownload slide Flow diagram of articles included in the systematic review Of these, 20 (40 per cent) were randomized trials comparing ER and non-ER protocols12–31, and the remaining 3032–61 described at least one retrospective cohort or used non-random assignment. Some 22 unique ER elements were identified throughout the publications. The median number of implemented ER elements described in these publications was 9 (range 3–17). The median number of patients included was 130 (range 50–1358). Quality assessment scores ranged from 9 to 24 (median 17). There were six studies of poor quality, 29 of fair quality and 15 of good quality. Only 19 publications (38 per cent) described the implementation process of the authors' ER pathway and only 14 specifically identified the number of surgeons participating in the study. For this subset of publications, the median number of surgeons was 3 (range 1–8). Only a single publication listed the number of anaesthetists involved. Most publications (90 per cent) detailed the ER protocol within the methods section of the paper, whereas the remainder referenced another article as containing the ER protocol in use. The most frequently utilized ER elements in the reviewed publications included early postoperative diet advancement in 49 (98 per cent) and early mobilization in 47 (94 per cent) (Fig. 2 and Table 1). How these elements were implemented varied across publications. The method for diet advancement was explained in 43 publications (86 per cent), and compliance with early initiation of diet advancement was reported in 22 (45 per cent) of 49 publications. Similarly, early mobilization was defined in detail in 41 publications (82 per cent), but compliance was reported in only 14 (30 per cent) of 47 publications. Descriptions of diet and mobilization strategies used in the comparative traditional recovery pathways were found in 37 (74 per cent) and 27 (54 per cent) publications respectively. In total, 18 publications disclosed failure rates for postoperative diet advancement and five reported on adverse events related to mobilization. Fig. 2 Open in new tabDownload slide Rate of enhanced recovery (ER) compliance reporting as it relates to ER element naming and explanations. *Indicates element with strong supporting evidence for safety and efficacy8,9. NGT, nasogastric tube; MIS, minimally invasive surgery; PONV, postoperative nausea and vomiting; Fio2, fraction of inspired oxygen Table 1 Enhanced recovery and traditional pathway named elements and explanations with level of evidence and inclusion in the ERAS® Society Registry Factor (22 elements) . Evidence-based8,9 . ERAS® Society Registry variable . Named (n = 50) . ER pathway explained . Traditional pathway explained . Preoperative education Yes* 28 (56) 24 (86) 16 (57) Bowel preparation Yes Yes* 26 (52) 25 (96) 18 (69) Reduced preoperative fasting Yes 29 (58) 27 (93) 20 (69) Carbohydrate loading Yes* 23 (46) 15 (65) 19 (83) Preoperative medications Yes* 5 (10) 2 (40) 4 (80) Withholding sedative medications before surgery Yes* 8 (16) 8 (100) 6 (75) PONV prophylaxis Yes Yes* 8 (16) 5 (63) 6 (75) MIS/incision Yes 8 (16) 8 (100) 6 (75) Intraoperative thermal regulation Yes Yes* 12 (24) 9 (75) 7 (58) Narcotic limitation Yes 24 (48) 18 (75) 9 (38) Increased Fio2 5 (10) 5 (100) 2 (40) NGT management/presence Yes Yes* 37 (74) 32 (86) 22 (59) Epidural analgesia use Yes Yes* 30 (60) 25 (83) 22 (73) Intraoperative fluid restriction Yes Yes* 15 (30) 14 (93) 12 (80) Postoperative fluid restriction Yes Yes 31 (62) 24 (77) 13 (42) Routine laxative or prokinetic use Yes* 16 (32) 16 (100) 12 (75) Postoperative protein supplements Yes* 10 (20) 9 (90) 6 (60) Postoperative carbohydrate supplements Yes* 3 (6) 3 (100) 3 (100) Early postoperative diet Yes Yes* 49 (98) 43 (88) 37 (76) Early mobilization Yes Yes* 47 (94) 41 (87) 27 (57) Urinary catheter management/presence Yes 36 (72) 34 (94) 22 (61) Intra-abdominal drain management/presence Yes Yes* 19 (38) 10 (53) 8 (42) Factor (22 elements) . Evidence-based8,9 . ERAS® Society Registry variable . Named (n = 50) . ER pathway explained . Traditional pathway explained . Preoperative education Yes* 28 (56) 24 (86) 16 (57) Bowel preparation Yes Yes* 26 (52) 25 (96) 18 (69) Reduced preoperative fasting Yes 29 (58) 27 (93) 20 (69) Carbohydrate loading Yes* 23 (46) 15 (65) 19 (83) Preoperative medications Yes* 5 (10) 2 (40) 4 (80) Withholding sedative medications before surgery Yes* 8 (16) 8 (100) 6 (75) PONV prophylaxis Yes Yes* 8 (16) 5 (63) 6 (75) MIS/incision Yes 8 (16) 8 (100) 6 (75) Intraoperative thermal regulation Yes Yes* 12 (24) 9 (75) 7 (58) Narcotic limitation Yes 24 (48) 18 (75) 9 (38) Increased Fio2 5 (10) 5 (100) 2 (40) NGT management/presence Yes Yes* 37 (74) 32 (86) 22 (59) Epidural analgesia use Yes Yes* 30 (60) 25 (83) 22 (73) Intraoperative fluid restriction Yes Yes* 15 (30) 14 (93) 12 (80) Postoperative fluid restriction Yes Yes 31 (62) 24 (77) 13 (42) Routine laxative or prokinetic use Yes* 16 (32) 16 (100) 12 (75) Postoperative protein supplements Yes* 10 (20) 9 (90) 6 (60) Postoperative carbohydrate supplements Yes* 3 (6) 3 (100) 3 (100) Early postoperative diet Yes Yes* 49 (98) 43 (88) 37 (76) Early mobilization Yes Yes* 47 (94) 41 (87) 27 (57) Urinary catheter management/presence Yes 36 (72) 34 (94) 22 (61) Intra-abdominal drain management/presence Yes Yes* 19 (38) 10 (53) 8 (42) Values in parentheses are percentages. ‘Named’ means that the element was listed in the publication; ‘ER pathway explained’ means that the enhanced recovery (ER) element was described with sufficient detail to facilitate transfer to the reader's clinical practice; ‘Traditional pathway explained’ means that the traditional pathway element was described with sufficient detail to facilitate transfer to the reader's clinical practice. * Denotes key field in the Registry. ERAS®, enhanced recovery after surgery; PONV, postoperative nausea and vomiting; MIS, minimally invasive surgery Fio2, fraction of inspired oxygen; NGT, nasogastric tube. Open in new tab Table 1 Enhanced recovery and traditional pathway named elements and explanations with level of evidence and inclusion in the ERAS® Society Registry Factor (22 elements) . Evidence-based8,9 . ERAS® Society Registry variable . Named (n = 50) . ER pathway explained . Traditional pathway explained . Preoperative education Yes* 28 (56) 24 (86) 16 (57) Bowel preparation Yes Yes* 26 (52) 25 (96) 18 (69) Reduced preoperative fasting Yes 29 (58) 27 (93) 20 (69) Carbohydrate loading Yes* 23 (46) 15 (65) 19 (83) Preoperative medications Yes* 5 (10) 2 (40) 4 (80) Withholding sedative medications before surgery Yes* 8 (16) 8 (100) 6 (75) PONV prophylaxis Yes Yes* 8 (16) 5 (63) 6 (75) MIS/incision Yes 8 (16) 8 (100) 6 (75) Intraoperative thermal regulation Yes Yes* 12 (24) 9 (75) 7 (58) Narcotic limitation Yes 24 (48) 18 (75) 9 (38) Increased Fio2 5 (10) 5 (100) 2 (40) NGT management/presence Yes Yes* 37 (74) 32 (86) 22 (59) Epidural analgesia use Yes Yes* 30 (60) 25 (83) 22 (73) Intraoperative fluid restriction Yes Yes* 15 (30) 14 (93) 12 (80) Postoperative fluid restriction Yes Yes 31 (62) 24 (77) 13 (42) Routine laxative or prokinetic use Yes* 16 (32) 16 (100) 12 (75) Postoperative protein supplements Yes* 10 (20) 9 (90) 6 (60) Postoperative carbohydrate supplements Yes* 3 (6) 3 (100) 3 (100) Early postoperative diet Yes Yes* 49 (98) 43 (88) 37 (76) Early mobilization Yes Yes* 47 (94) 41 (87) 27 (57) Urinary catheter management/presence Yes 36 (72) 34 (94) 22 (61) Intra-abdominal drain management/presence Yes Yes* 19 (38) 10 (53) 8 (42) Factor (22 elements) . Evidence-based8,9 . ERAS® Society Registry variable . Named (n = 50) . ER pathway explained . Traditional pathway explained . Preoperative education Yes* 28 (56) 24 (86) 16 (57) Bowel preparation Yes Yes* 26 (52) 25 (96) 18 (69) Reduced preoperative fasting Yes 29 (58) 27 (93) 20 (69) Carbohydrate loading Yes* 23 (46) 15 (65) 19 (83) Preoperative medications Yes* 5 (10) 2 (40) 4 (80) Withholding sedative medications before surgery Yes* 8 (16) 8 (100) 6 (75) PONV prophylaxis Yes Yes* 8 (16) 5 (63) 6 (75) MIS/incision Yes 8 (16) 8 (100) 6 (75) Intraoperative thermal regulation Yes Yes* 12 (24) 9 (75) 7 (58) Narcotic limitation Yes 24 (48) 18 (75) 9 (38) Increased Fio2 5 (10) 5 (100) 2 (40) NGT management/presence Yes Yes* 37 (74) 32 (86) 22 (59) Epidural analgesia use Yes Yes* 30 (60) 25 (83) 22 (73) Intraoperative fluid restriction Yes Yes* 15 (30) 14 (93) 12 (80) Postoperative fluid restriction Yes Yes 31 (62) 24 (77) 13 (42) Routine laxative or prokinetic use Yes* 16 (32) 16 (100) 12 (75) Postoperative protein supplements Yes* 10 (20) 9 (90) 6 (60) Postoperative carbohydrate supplements Yes* 3 (6) 3 (100) 3 (100) Early postoperative diet Yes Yes* 49 (98) 43 (88) 37 (76) Early mobilization Yes Yes* 47 (94) 41 (87) 27 (57) Urinary catheter management/presence Yes 36 (72) 34 (94) 22 (61) Intra-abdominal drain management/presence Yes Yes* 19 (38) 10 (53) 8 (42) Values in parentheses are percentages. ‘Named’ means that the element was listed in the publication; ‘ER pathway explained’ means that the enhanced recovery (ER) element was described with sufficient detail to facilitate transfer to the reader's clinical practice; ‘Traditional pathway explained’ means that the traditional pathway element was described with sufficient detail to facilitate transfer to the reader's clinical practice. * Denotes key field in the Registry. ERAS®, enhanced recovery after surgery; PONV, postoperative nausea and vomiting; MIS, minimally invasive surgery Fio2, fraction of inspired oxygen; NGT, nasogastric tube. Open in new tab After these interventions, the next most commonly described interventions were nasogastric tube management in 37 publications (74 per cent), urinary catheter management in 36 (72 per cent) and epidural analgesia utilization in 30 (60 per cent) (Fig. 2 and Table 1). These elements were explained in detail in 32 (86 per cent), 34 (94 per cent) and 25 (83 per cent) respectively of the publications that named them as elements. The corresponding traditional pathway element or management was explained in 22 (59 per cent), 22 (61 per cent) and 22 (73 per cent) respectively of the publications. Compliance rates for each of these elements were found in eight (22 per cent), nine (25 per cent) and 14 (47 per cent) of the publications. The rate of postoperative nasogastric tube replacement was reported in only ten publications (27 per cent). The rate of urinary retention requiring replacement of a urinary catheter was listed in eight reports (22 per cent), and the rate of epidural failure in nine (30 per cent). Although fewer publications reported or explained in detail the use of minimally invasive surgery or particular types of incision, those that did name this element frequently (7 of 8 papers, 88 per cent) reported compliance in the form of surgical approach and the use of specific incisions. Other elements that had high compliance reporting were intraoperative fluid restriction (9 of 15, 60 per cent), postoperative nausea and vomiting prophylaxis (5 of 8, 63 per cent), postoperative protein supplements (6 of 10, 60 per cent) and postoperative carbohydrate supplements (2 of 3, 67 per cent). Elements with the lowest compliance reporting were: withholding preoperative sedative medications (1 of 8, 13 per cent), treatment with high fraction of inspired oxygen (1 of 5, 20 per cent), premedication administration (1 of 5, 20 per cent) and reduced preoperative fasting period (6 of 29, 21 per cent) (Fig. 2). The outcome measures analysed most frequently in the publications were morbidity in 49 (98 per cent) and length of hospital stay in 47 (94 per cent); only 40 (85 per cent) of these 47publications described utilization of specific discharge criteria. Mortality was reported in 45 publications (90 per cent) and readmissions in 41 (82 per cent). These, as well as other, less frequently reported, outcome measures are shown in Fig. 3. Fig. 3 Open in new tabDownload slide Frequency of reported patient outcome measures assessed as dependent variables in 50 colorectal surgery enhanced recovery publications Discussion This study focused on two areas of reporting regarding ER protocols related to colorectal surgery: transparency of reporting individual components of ER protocols, and compliance and failure rates with the stated components. There was significant variation in the completeness of descriptions of ER protocols. The median number of components described was only 9. Despite heterogeneity between studies, most did report a positive effect of ER on short-term outcomes. This may be attributable to the fact that diet advancement and early mobilization were common to most studies, with a significant impact on short-term endpoints such as length of hospital stay. Authors reported inclusion or exclusion of important components such as fluid management, oral non-narcotic analgesia, and routine laxative or prokinetic use less frequently. When a restrictive fluid management protocol was listed, only rarely did the authors describe their practice, or patient monitoring (such as fluid balance or weight monitoring). Importantly, very few manuscripts that recorded neuraxial analgesia reported epidural/regional level, single injection versus infusion/catheter placement, medications/concentrations used or duration of infusion. The absence of adequate descriptions of ER components significantly impairs the ability to interpret, compare outcomes between studies, or translate practice to other centres. Specific attention to compliance with the most important ER elements also needs to be made clear, as not all of the elements are evidence-based and known to be important in enhancing recovery8,9. Compliance and failure rates with stated components of the protocol also varied, and few studies actually reported the compliance and/or failure rate for each component of the ER protocol. Without these data it is impossible to determine the efficacy of each step in the pathway and its relative effects on the reported outcome measures. Additionally, detailed description of the corollary traditional pathway was frequently lacking. This is of particular importance when there is no apparent difference between outcome measures for ER and traditional groups. Recently, there has been a movement towards standard reporting in medical literature. PRISMA6, Consolidated Standards of Reporting Trials (CONSORT)62 and Template for Intervention Description and Replication (TIDieR)63 recommendations have been proposed for meta-analyses, randomized clinical trials and clinical studies respectively. TIDieR guidelines postdate the studies in this review, but additionally do not take into account all of the necessary features in an ER protocol, such as component compliance, failure rates, control pathway descriptions, graded complications and discharge criteria. As such, it would not be possible to draw meaningful conclusions about the effectiveness of the applied interventions upon outcome measures using the TIDieR instrument alone. The present study included a rigorous literature search and strict inclusion criteria that identified a large number of high-quality studies likely to have high clinical impact. Nevertheless, the non-standard reporting of these elements meant that some of data collection was subjective, although all findings in the analysis have been reported transparently. The finding of significant omissions in reporting ER elements in even the highest-quality studies, including 20 randomized trials, leads to the conclusion that a defined template for reporting these studies would significantly improve the quality of the literature related to ER for colorectal surgery. Based on the deficiencies noted in this analysis a standard reporting template is proposed (Table 2). Although it could prove cumbersome to provide all details within the original publication, the use of appendices or references could be used for complete reporting. Measurement and reporting of compliance for each component has been shown to correlate with patient outcomes66,67. A clear description of discharge criteria, with differentiation of ‘ready for discharge’ compared with actual length of stay with reasons for non-medical extension of inpatient hospitalization should also be seen as essential68. Adherence with these guidelines should make it easier and more reliable to compare practices and outcomes. It may also reverse the observation that compliance with ER pathway elements falls outside the setting of a clinical trial69. Table 2 Recommended reporting elements for studies comparing enhanced recovery with traditional pathways Tabular reporting of all elements included in the examined ER pathway and corresponding elements from the traditional pathway Clearly explain all ER elements with particular attention to reporting specific algorithms and pathways used in clinical management where applicable (such as intravenous fluid rates and criteria for goal-directed fluid therapy, epidural/regional level, single injection versus infusion/catheter placement, medications/concentrations used, duration of infusion, analgesia escalation strategies, drain placement algorithms) Report compliance for all elements named as part of the ER protocol When failure of an ER element is possible, it should be reported and explained, including adverse events that may be related to an ER element When length of stay is used as an outcome measure, discharge criteria or the lack thereof should be reported. If a substitute for length of stay such as ‘readiness for discharge’ is used, there should be a report of actual length of stay and a listing of reasons for non-medical extension of hospitalization When morbidity is used as an outcome measure, efforts should be made to grade complications and stratify them according to severity using a standard system64,65 Tabular reporting of all elements included in the examined ER pathway and corresponding elements from the traditional pathway Clearly explain all ER elements with particular attention to reporting specific algorithms and pathways used in clinical management where applicable (such as intravenous fluid rates and criteria for goal-directed fluid therapy, epidural/regional level, single injection versus infusion/catheter placement, medications/concentrations used, duration of infusion, analgesia escalation strategies, drain placement algorithms) Report compliance for all elements named as part of the ER protocol When failure of an ER element is possible, it should be reported and explained, including adverse events that may be related to an ER element When length of stay is used as an outcome measure, discharge criteria or the lack thereof should be reported. If a substitute for length of stay such as ‘readiness for discharge’ is used, there should be a report of actual length of stay and a listing of reasons for non-medical extension of hospitalization When morbidity is used as an outcome measure, efforts should be made to grade complications and stratify them according to severity using a standard system64,65 ER, enhanced recovery. Open in new tab Table 2 Recommended reporting elements for studies comparing enhanced recovery with traditional pathways Tabular reporting of all elements included in the examined ER pathway and corresponding elements from the traditional pathway Clearly explain all ER elements with particular attention to reporting specific algorithms and pathways used in clinical management where applicable (such as intravenous fluid rates and criteria for goal-directed fluid therapy, epidural/regional level, single injection versus infusion/catheter placement, medications/concentrations used, duration of infusion, analgesia escalation strategies, drain placement algorithms) Report compliance for all elements named as part of the ER protocol When failure of an ER element is possible, it should be reported and explained, including adverse events that may be related to an ER element When length of stay is used as an outcome measure, discharge criteria or the lack thereof should be reported. If a substitute for length of stay such as ‘readiness for discharge’ is used, there should be a report of actual length of stay and a listing of reasons for non-medical extension of hospitalization When morbidity is used as an outcome measure, efforts should be made to grade complications and stratify them according to severity using a standard system64,65 Tabular reporting of all elements included in the examined ER pathway and corresponding elements from the traditional pathway Clearly explain all ER elements with particular attention to reporting specific algorithms and pathways used in clinical management where applicable (such as intravenous fluid rates and criteria for goal-directed fluid therapy, epidural/regional level, single injection versus infusion/catheter placement, medications/concentrations used, duration of infusion, analgesia escalation strategies, drain placement algorithms) Report compliance for all elements named as part of the ER protocol When failure of an ER element is possible, it should be reported and explained, including adverse events that may be related to an ER element When length of stay is used as an outcome measure, discharge criteria or the lack thereof should be reported. If a substitute for length of stay such as ‘readiness for discharge’ is used, there should be a report of actual length of stay and a listing of reasons for non-medical extension of hospitalization When morbidity is used as an outcome measure, efforts should be made to grade complications and stratify them according to severity using a standard system64,65 ER, enhanced recovery. Open in new tab Another important finding in this analysis was the character of the dependent variables assessed in these studies. These were usually crude inpatient indicators of recovery such as morbidity, return of bowel function and length of hospital stay. Patient-reported outcomes and information related to costs of care are lacking. Biological and molecular correlates that might inform future modifications in ER pathways were rarely included. This deficiency has been highlighted before, and it does not appear that reporting of these elements has improved70. ER protocols involve multiple components that will naturally have variable success rates. To transfer knowledge and facilitate implementation of pathways that demonstrate improvements in perioperative care and recovery, a structured reporting platform needs to be agreed and implemented. Acknowledgements V.G. and T.A.A. contributed equally as senior authors of this manuscript. This research was supported in part by the National Institutes of Health (grant number CA016672). Disclosure: The authors declare no conflict of interest. References 1 Kehlet H , Wilmore DW. Fast-track surgery . Br J Surg 2005 ; 92 : 3 – 4 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Kehlet H Fast-track colonic surgery: status and perspectives . Recent Results Cancer Res 2005 ; 165 : 8 – 13 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Basse L , Thorbøl JE, Løssl K, Kehlet H. Colonic surgery with accelerated rehabilitation or conventional care . Dis Colon Rectum 2004 ; 47 : 271 – 277 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Moher D , Hopewell S, Schulz KF, Montori V, Gøtzsche PC, Devereaux PJ et al. ; Consolidated Standards of Reporting Trials Group. CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials . J Clin Epidemiol 2010 ; 63 : e1 – e37 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Brölmann FE , Eskes AM, Sumpio BE, Mayer DO, Moore Z, Agren MS et al. Fundamentals of randomized clinical trials in wound care: reporting standards . Wound Repair Regen 2013 ; 21 : 641 – 647 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Liberati A , Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration . J Clin Epidemiol 2009 ; 62 : e1 – e34 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Slim K , Kehlet H. Commentary: Fast track surgery: the need for improved study design . Colorectal Dis 2012 ; 14 : 1013 – 1014 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Kehlet H Fast-track colorectal surgery . Lancet 2008 ; 371 : 791 – 793 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Gustafsson UO , Scott MJ, Schwenk W, Demartines N, Roulin D, Francis N et al. ; Enhanced Recovery After Surgery (ERAS) Society, for Perioperative Care; European Society for Clinical Nutrition and Metabolism (ESPEN); International Association for Surgical Metabolism and Nutrition (IASMEN). Guidelines for perioperative care in elective colonic surgery: Enhanced Recovery After Surgery (ERAS®) Society recommendations . World J Surg 2013 ; 37 : 259 – 284 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Downs SH , Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions . J Epidemiol Community Health 1998 ; 52 : 377 – 384 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Samoocha D , Bruinvels DJ, Elbers NA, Anema JR, van der Beek AJ. Effectiveness of web-based interventions on patient empowerment: a systematic review and meta-analysis . J Med Internet Res 2010 ; 12 : e23 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Frontera D , Arena L, Corsale I, Francioli N, Mammoliti F, Buccianelli E. Fast track in colo-rectal surgery. Preliminary experience in a rural hospital . G Chir 2014 ; 35 : 293 – 301 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 13 Feng F , Li XH, Shi H, Wu GS, Zhang HW, Liu XN et al. Fast-track surgery combined with laparoscopy could improve postoperative recovery of low-risk rectal cancer patients: a randomized controlled clinical trial . J Dig Dis 2014 ; 15 : 306 – 313 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Jia Y , Jin G, Guo S, Gu B, Jin Z, Gao X et al. Fast-track surgery decreases the incidence of postoperative delirium and other complications in elderly patients with colorectal carcinoma . Langenbecks Arch Surg 2014 ; 399 : 77 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Li K , Li JP, Peng NH, Jiang LL, Hu YJ, Huang MJ. Fast-track improves post-operative nutrition and outcomes of colorectal surgery: a single-center prospective trial in China . Asia Pac J Clin Nutr 2014 ; 23 : 41 – 47 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 16 Mari GM , Costanzi A, Maggioni D, Origi M, Ferrari GC, De Martini P et al. Fast-track versus standard care in laparoscopic high anterior resection: a prospective randomized-controlled trial . Surg Laparosc Endosc Percutan Tech 2014 ; 24 : 118 – 121 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Compagna R , Aprea G, De Rosa D, Gentile M, Cestaro G, Vigliotti G et al. Fast track for elderly patients: is it feasible for colorectal surgery? Int J Surg 2014 ; 12 ( Suppl 2 ): S20 – S22 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Lee SM , Kang SB, Jang JH, Park JS, Hong S, Lee TG et al. Early rehabilitation versus conventional care after laparoscopic rectal surgery: a prospective, randomized, controlled trial . Surg Endosc 2013 ; 27 : 3902 – 3909 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Wang G , Jiang Z, Zhao K, Li G, Liu F, Pan H et al. Immunologic response after laparoscopic colon cancer operation within an enhanced recovery program . J Gastrointest Surg 2012 ; 16 : 1379 – 1388 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Wang Q , Suo J, Jiang J, Wang C, Zhao YQ, Cao X. Effectiveness of fast-track rehabilitation vs conventional care in laparoscopic colorectal resection for elderly patients: a randomized trial . Colorectal Dis 2012 ; 14 : 1009 – 1013 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Ren L , Zhu D, Wei Y, Pan X, Liang L, Xu J et al. Enhanced Recovery After Surgery (ERAS) program attenuates stress and accelerates recovery in patients after radical resection for colorectal cancer: a prospective randomized controlled trial . World J Surg 2012 ; 36 : 407 – 414 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Veenhof AA , Vlug MS, van der Pas MH, Sietses C, van der Peet DL, de Lange-de Klerk ES et al. Surgical stress response and postoperative immune function after laparoscopy or open surgery with fast track or standard perioperative care: a randomized trial . Ann Surg 2012 ; 255 : 216 – 221 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Yang D , He W, Zhang S, Chen H, Zhang C, He Y. Fast-track surgery improves postoperative clinical recovery and immunity after elective surgery for colorectal carcinoma: randomized controlled clinical trial . World J Surg 2012 ; 36 : 1874 – 1880 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Hübner M , Schäfer M, Demartines N, Müller S, Maurer K, Baulig W et al. ; Zurich Fast Track Study Group. Impact of restrictive intravenous fluid replacement and combined epidural analgesia on perioperative volume balance and renal function within a fast track program . J Surg Res 2012 ; 173 : 68 – 74 . Google Scholar Crossref Search ADS PubMed WorldCat 25 García-Botello S , Cánovas de Lucas R, Tornero C, Escamilla B, Espí-Macias A, Esclapez-Valero P et al. [Implementation of a perioperative multimodal rehabilitation protocol in elective colorectal surgery. A prospective randomised controlled study.] Cir Esp 2011 ; 89 : 159 – 166 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Vlug MS , Wind J, Hollmann MW, Ubbink DT, Cense HA, Engel AF et al. Laparoscopy in combination with fast track multimodal management is the best perioperative strategy in patients undergoing colonic surgery: a randomized clinical trial (LAFA-study) . Ann Surg 2011 ; 254 : 868 – 875 . Google Scholar Crossref Search ADS PubMed WorldCat 27 van Bree SH , Vlug MS, Bemelman WA, Hollmann MW, Ubbink DT, Zwinderman AH et al. Faster recovery of gastrointestinal transit after laparoscopy and fast-track care in patients undergoing colonic surgery . Gastroenterology 2011 ; 141 : 872 – 880 .e871–e874. Google Scholar Crossref Search ADS PubMed WorldCat 28 Lee TG , Kang SB, Kim DW, Hong S, Heo SC, Park KJ. Comparison of early mobilization and diet rehabilitation program with conventional care after laparoscopic colon surgery: a prospective randomized controlled trial . Dis Colon Rectum 2011 ; 54 : 21 – 28 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Serclová Z , Dytrych P, Marvan J, Nová K, Hankeová Z, Ryska O et al. Fast-track in open intestinal surgery: prospective randomized study (Clinical Trials Gov Identifier no. NCT00123456) . Clin Nutr 2009 ; 28 : 618 – 624 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Ionescu D , Iancu C, Ion D, Al-Hajjar N, Margarit S, Mocan L et al. Implementing fast-track protocol for colorectal surgery: a prospective randomized clinical trial . World J Surg 2009 ; 33 : 2433 – 2438 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Muller S , Zalunardo MP, Hubner M, Clavien PA, Demartines N; Zurich Fast Track Study Group . A fast-track program reduces complications and length of hospital stay after open colonic surgery . Gastroenterology 2009 ; 136 : 842 – 847 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Ehrlich A , Kellokumpu S, Wagner B, Kautiainen H, Kellokumpu I. Comparison of laparoscopic and open colonic resection within fast-track and traditional perioperative care pathways: clinical outcomes and in-hospital costs . Scand J Surg 2014 ; [Epub ahead of print]. Google Scholar OpenURL Placeholder Text WorldCat 33 Lohsiriwat V . Enhanced recovery after surgery vs conventional care in emergency colorectal surgery . World J Gastroenterol 2014 ; 20 : 13 950 – 13 955 . Google Scholar Crossref Search ADS WorldCat 34 Miller TE , Thacker JK, White WD, Mantyh C, Migaly J, Jin J et al. ; Enhanced Recovery Study Group. Reduced length of hospital stay in colorectal surgery after implementation of an enhanced recovery protocol . Anesth Analg 2014 ; 118 : 1052 – 1061 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Khreiss W , Huebner M, Cima RR, Dozois ER, Chua HK, Pemberton JH et al. Improving conventional recovery with enhanced recovery in minimally invasive surgery for rectal cancer . Dis Colon Rectum 2014 ; 57 : 557 – 563 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Arroyo A , Ramirez JM, Callejo D, Viñas X, Maeso S, Cabezali R et al. ; Spanish Working Group in Fast Track Surgery (GERM) . Influence of size and complexity of the hospitals in an enhanced recovery programme for colorectal resection . Int J Colorectal Dis 2012 ; 27 : 1637 – 1644 . Google Scholar Crossref Search ADS PubMed WorldCat 37 Fierens J , Wolthuis AM, Penninckx F, D'Hoore A. Enhanced recovery after surgery (ERAS) protocol: prospective study of outcome in colorectal surgery . Acta Chir Belg 2012 ; 112 : 355 – 358 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Keane C , Savage S, McFarlane K, Seigne R, Robertson G, Eglinton T. Enhanced recovery after surgery versus conventional care in colonic and rectal surgery . ANZ J Surg 2012 ; 82 : 697 – 703 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Gouvas N , Gogos-Pappas G, Tsimogiannis K, Tsimoyiannis E, Dervenis C, Xynos E. Implementation of fast-track protocols in open and laparoscopic sphincter-preserving rectal cancer surgery: a multicenter, comparative, prospective, non-randomized study . Dig Surg 2012 ; 29 : 301 – 309 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Huibers CJ , de Roos MA, Ong KH. The effect of the introduction of the ERAS protocol in laparoscopic total mesorectal excision for rectal cancer . Int J Colorectal Dis 2012 ; 27 : 751 – 757 . Google Scholar Crossref Search ADS PubMed WorldCat 41 Haverkamp MP , de Roos MA, Ong KH. The ERAS protocol reduces the length of stay after laparoscopic colectomies . Surg Endosc 2012 ; 26 : 361 – 367 . Google Scholar Crossref Search ADS PubMed WorldCat 42 McNicol FJ , Kennedy RH, Phillips RK, Clark SK. Laparoscopic total colectomy and ileorectal anastomosis (IRA), supported by an enhanced recovery programme in cases of familial adenomatous polyposis . Colorectal Dis 2012 ; 14 : 458 – 462 . Google Scholar Crossref Search ADS PubMed WorldCat 43 Walter CJ , Watson JT, Pullan RD, Kenefick NJ, Mitchell SJ, Defriend DJ. Enhanced recovery in major colorectal surgery: safety and efficacy in an unselected surgical population at a UK district general hospital . Surgeon 2011 ; 9 : 259 – 264 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 44 Archibald LH , Ott MJ, Gale CM, Zhang J, Peters MS, Stroud GK. Enhanced recovery after colon surgery in a community hospital system . Dis Colon Rectum 2011 ; 54 : 840 – 845 . Google Scholar Crossref Search ADS PubMed WorldCat 45 Poon JT , Fan JK, Lo OS, Law WL. Enhanced recovery program in laparoscopic colectomy for cancer . Int J Colorectal Dis 2011 ; 26 : 71 – 77 . Google Scholar Crossref Search ADS PubMed WorldCat 46 Christensen HK , Thaysen HV, Rodt SA, Carlsson P, Laurberg S. Short hospital stay and low complication rate are possible with a fully implemented fast-track model after elective colonic surgery . Eur Surg Res 2011 ; 46 : 156 – 161 . Google Scholar Crossref Search ADS PubMed WorldCat 47 Schwarzbach M , Hasenberg T, Linke M, Kienle P, Post S, Ronellenfitsch U. Perioperative quality of care is modulated by process management with clinical pathways for fast-track surgery of the colon . Int J Colorectal Dis 2011 ; 26 : 1567 – 1575 . Google Scholar Crossref Search ADS PubMed WorldCat 48 Teeuwen PH , Bleichrodt RP, de Jong PJ, van Goor H, Bremers AJ. Enhanced recovery after surgery versus conventional perioperative care in rectal surgery . Dis Colon Rectum 2011 ; 54 : 833 – 839 . Google Scholar Crossref Search ADS PubMed WorldCat 49 Kahokehr A , Sammour T, Zargar-Shoshtari K, Srinivasa S, Hill AG. Recovery after open and laparoscopic right hemicolectomy: a comparison . J Surg Res 2010 ; 162 : 11 – 16 . Google Scholar Crossref Search ADS PubMed WorldCat 50 Baird G , Maxson P, Wrobleski D, Luna BS. Fast-track colorectal surgery program reduces hospital length of stay . Clin Nurse Spec 2010 ; 24 : 202 – 208 . Google Scholar Crossref Search ADS PubMed WorldCat 51 Larson DW , Batdorf NJ, Touzios JG, Cima RR, Chua HK, Pemberton JH et al. A fast-track recovery protocol improves outcomes in elective laparoscopic colectomy for diverticulitis . J Am Coll Surg 2010 ; 211 : 485 – 489 . Google Scholar Crossref Search ADS PubMed WorldCat 52 Lloyd GM , Kirby R, Hemingway DM, Keane FB, Miller AS, Neary P. The RAPID protocol enhances patient recovery after both laparoscopic and open colorectal resections . Surg Endosc 2010 ; 24 : 1434 – 1439 . Google Scholar Crossref Search ADS PubMed WorldCat 53 Al Chalabi H , Kavanagh DO, Hassan L, Donnell KO, Nugent E, Andrews E et al. The benefit of an enhanced recovery programme following elective laparoscopic sigmoid colectomy . Int J Colorectal Dis 2010 ; 25 : 761 – 766 . Google Scholar Crossref Search ADS PubMed WorldCat 54 Tsikitis VL , Holubar SD, Dozois EJ, Cima RR, Pemberton JH, Larson DW. Advantages of fast-track recovery after laparoscopic right hemicolectomy for colon cancer . Surg Endosc 2010 ; 24 : 1911 – 1916 . Google Scholar Crossref Search ADS PubMed WorldCat 55 Aboulian A , Hassan Z, Lin MY, Kaji AH, Kumar RR. Successful enhanced recovery program after colorectal surgery in a county institution . Am Surg 2010 ; 76 : 1158 – 1162 . Google Scholar Crossref Search ADS PubMed WorldCat 56 Branagan G , Richardson L, Shetty A, Chave HS. An enhanced recovery programme reduces length of stay after rectal surgery . Int J Colorectal Dis 2010 ; 25 : 1359 – 1362 . Google Scholar Crossref Search ADS PubMed WorldCat 57 de Aguilar-Nascimento JE , Bicudo-Salomão A, Caporossi C, Silva Rde M, Cardoso EA, Santos TP et al. Multimodal approach in colorectal surgery without mechanical bowel cleansing . Rev Col Bras Cir 2009 ; 36 : 204 – 209 . Google Scholar Crossref Search ADS PubMed WorldCat 58 Nygren J , Soop M, Thorell A, Hausel J, Ljungqvist O; ERAS Group. An enhanced recovery protocol improves outcome after colorectal resection already during the first year: a single-center experience in 168 consecutive patients . Dis Colon Rectum 2009 ; 52 : 978 – 985 . Google Scholar Crossref Search ADS PubMed WorldCat 59 Mohn AC , Bernardshaw SV, Ristesund SM, Hovde Hansen PE, Røkke O. Enhanced recovery after colorectal surgery. Results from a prospective observational two-centre study . Scand J Surg 2009 ; 98 : 155 – 159 . Google Scholar Crossref Search ADS PubMed WorldCat 60 Feo CV , Lanzara S, Sortini D, Ragazzi R, De Pinto M, Pansini GC et al. Fast track postoperative management after elective colorectal surgery: a controlled trial . Am Surg 2009 ; 75 : 1247 – 1251 . Google Scholar Crossref Search ADS PubMed WorldCat 61 Wang G , Jiang ZW, Xu J, Gong JF, Bao Y, Xie LF et al. Fast-track rehabilitation program vs conventional care after colorectal resection: a randomized clinical trial . World J Gastroenterol 2011 ; 17 : 671 – 676 . Google Scholar Crossref Search ADS PubMed WorldCat 62 Schulz KF , Altman DG, Moher D; CONSORT Group . CONSORT 2010 Statement: Updated guidelines for reporting parallel group randomised trials . J Clin Epidemiol 2010 ; 63 : 834 – 840 . Google Scholar Crossref Search ADS PubMed WorldCat 63 Hoffmann TC , Glasziou PP, Boutron I, Milne R, Perera R, Moher D et al. Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide . BMJ 2014 ; 348 : g1687 . Google Scholar Crossref Search ADS PubMed WorldCat 64 Strasberg SM , Linehan DC, Hawkins WG. The Accordion severity grading system of surgical complications . Ann Surg 2009 ; 250 : 177 – 186 . Google Scholar Crossref Search ADS PubMed WorldCat 65 Dindo D , Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey . Ann Surg 2004 ; 240 : 205 – 213 . Google Scholar Crossref Search ADS PubMed WorldCat 66 Feroci F , Lenzi E, Baraghini M, Garzi A, Vannucchi A, Cantafio S et al. Fast-track colorectal surgery: protocol adherence influences postoperative outcomes . Int J Colorectal Dis 2013 ; 28 : 103 – 109 . Google Scholar Crossref Search ADS PubMed WorldCat 67 Alcántara-Moral M , Serra-Aracil X, Gil-Egea MJ, Frasson M, Flor-Lorente B, Garcia-Granero E et al. ; E.B.S.Q.-C on behalf of the collaborative Group of Coloproctology Section of the Spanish Association of Surgeons . Observational cross-sectional study of compliance with the fast track protocol in elective surgery for colon cancer in Spain . Int J Colorectal Dis 2014 ; 29 : 477 – 483 . Google Scholar Crossref Search ADS PubMed WorldCat 68 Husted H , Lunn TH, Troelsen A, Gaarn-Larsen L, Kristensen BB, Kehlet H. Why still in hospital after fast-track hip and knee arthroplasty? Acta Orthop 2011 ; 82 : 679 – 684 . Google Scholar Crossref Search ADS PubMed WorldCat 69 Ahmed J , Khan S, Gatt M, Kallam R, MacFie J. Compliance with enhanced recovery programmes in elective colorectal surgery . Br J Surg 2010 ; 97 : 754 – 758 . Google Scholar Crossref Search ADS PubMed WorldCat 70 Neville A , Lee L, Antonescu I, Mayo NE, Vassiliou MC, Fried GM et al. Systematic review of outcomes used to evaluate enhanced recovery after surgery . Br J Surg 2014 ; 101 : 159 – 170 . Google Scholar Crossref Search ADS PubMed WorldCat © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd

Simillis, C; Thoukididou, S N; Slesser, A A P; Rasheed, S; Tan, E; Tekkis, P P

doi: 10.1002/bjs.9913pmid: 26420725

Abstract Background The aim was to compare the clinical outcomes and effectiveness of surgical treatments for haemorrhoids. Methods Randomized clinical trials were identified by means of a systematic review. A Bayesian network meta-analysis was performed using the Markov chain Monte Carlo method in WinBUGS. Results Ninety-eight trials were included with 7827 participants and 11 surgical treatments for grade III and IV haemorrhoids. Open, closed and radiofrequency haemorrhoidectomies resulted in significantly more postoperative complications than transanal haemorrhoidal dearterialization (THD), LigaSure™ and Harmonic® haemorrhoidectomies. THD had significantly less postoperative bleeding than open and stapled procedures, and resulted in significantly fewer emergency reoperations than open, closed, stapled and LigaSure™ haemorrhoidectomies. Open and closed haemorrhoidectomies resulted in more pain on postoperative day 1 than stapled, THD, LigaSure™ and Harmonic® procedures. After stapled, LigaSure™ and Harmonic® haemorrhoidectomies patients resumed normal daily activities earlier than after open and closed procedures. THD provided the earliest time to first bowel movement. The stapled and THD groups had significantly higher haemorrhoid recurrence rates than the open, closed and LigaSure™ groups. Recurrence of haemorrhoidal symptoms was more common after stapled haemorrhoidectomy than after open and LigaSure™ operations. No significant difference was identified between treatments for anal stenosis, incontinence and perianal skin tags. Conclusion Open and closed haemorrhoidectomies resulted in more postoperative complications and slower recovery, but fewer haemorrhoid recurrences. THD and stapled haemorrhoidectomies were associated with decreased postoperative pain and faster recovery, but higher recurrence rates. The advantages and disadvantages of each surgical treatment should be discussed with the patient before surgery to allow an informed decision to be made. Introduction Haemorrhoids are cushions of specialized submucosal vascular tissue located in the anal canal, and are one of the most common anorectal disorders1. The exact prevalence of symptomatic haemorrhoids is difficult to establish owing to under-reporting by patients. A prospective study2 of patients undergoing screening colonoscopy revealed the presence of haemorrhoids in 38·9 per cent, with 44·7 per cent of these suffering from haemorrhoidal symptoms. Haemorrhoidal symptoms may include bright red bleeding from the rectum, mucous discharge, perianal irritation or pruritus, perianal pain, prolapse of the haemorrhoidal cushions, or protruding mass, soiling and difficulties with hygiene1–4. Treatment options for haemorrhoidal disease range from conservative management with advice on diet, lifestyle changes and application of topical ointments, to interventions that can be performed on an outpatient setting (such as rubber band ligation, infrared coagulation, injection sclerotherapy), as well as surgical treatments3. Based on the degree of prolapse and the classification by Banov and colleagues5, grade III and IV haemorrhoids (prolapsed haemorrhoids requiring manual reduction and non-reducible prolapsed haemorrhoids respectively) are amenable to surgical treatment. Approximately 9000 haemorrhoidectomies (including around 1300 stapled procedures) were performed in England during 2012–20136. Hundreds of studies have been published comparing the surgical treatments available for grade III and IV haemorrhoids, including: open haemorrhoidectomy7–10, closed haemorrhoidectomy11–14, submucosal haemorrhoidectomy15,16, stapled haemorrhoidectomy17–20, transanal haemorrhoidal dearterialization (THD)21–24, LigaSure™ (Valleylab, Boulder, Colorado, USA) haemorrhoidectomy25–28, Harmonic® (Ethicon Endo-Surgery, Cincinnati, Ohio, USA) haemorrhoidectomy29–32, laser haemorrhoidectomy33–36, Starion™ (Starion Instruments, Saratoga, California, USA) haemorrhoidectomy37, radiofrequency haemorrhoidectomy38,39 and bipolar scissors haemorrhoidectomy40. The aim of this study was to perform a systematic review of the literature to identify the surgical treatments available for grade III and IV haemorrhoids, and to carry out a network meta-analysis to compare the clinical outcomes and effectiveness of these treatments. Methods Search strategy A comprehensive literature search using a combination of free-text terms and controlled vocabulary when applicable was undertaken in the following databases: MEDLINE, Embase, Science Citation Index Expanded, and Cochrane Central Register of Controlled Trials (CENTRAL) in the Cochrane Library. The World Health Organization International Clinical Trials Registry Platform search portal and ClinicalTrials.gov were also searched to identify further trials. Details of the search strategy are provided in Table S1 (supporting information). The related articles function in PubMed was used to broaden the search, and all abstracts, studies and citations identified were reviewed. The references of the identified trials were also searched to find additional trials for inclusion. No restrictions were made based on language, publication year or publication status. The latest date for this search was 21 June 2014. Inclusion and exclusion criteria Only randomized clinical trials (RCTs) were considered for this network meta-analysis, and only studies that recruited patients with haemorrhoids clearly defined as grades III and IV were included in the review. Finally, only studies reporting on elective surgical treatments were included; studies reporting on emergency haemorrhoidectomy for painful or thrombosed haemorrhoids were excluded. Outcomes of interest The various surgical interventions aiming to treat grade III and IV haemorrhoids were assessed for the following outcomes: postoperative complications, defined as any deviation from the normal postoperative course41 – all complications reported by the studies were added and included within this outcome; some included complications were analysed and individually – urinary retention, constipation/faecal impaction, postoperative bleeding, anal fissure, wound complications; emergency reoperation owing to early postoperative complications requiring expedited or urgent surgical treatment; duration of surgery in minutes; operative blood loss in millilitres; length of hospital stay in days; postoperative pain measured with a visual analogue scale on days 1, 7 and 14 after surgery; time to first bowel movement in days; time to return to work or normal activities in days; recurrence of haemorrhoids – number of patients with recurrent internal or prolapsed haemorrhoids seen on rectal examination at the clinic, or number of patients complaining of recurrent prolapsed haemorrhoids at follow-up; symptoms reported by patients at follow-up consistent with recurrent haemorrhoidal symptoms – all recurrent symptoms reported were added and included within this outcome; some of the included symptoms were also analysed individually – recurrent haemorrhoidal bleeding, recurrent pruritus, recurrent pain or discomfort related to haemorrhoids; anal stenosis, determined as the proportion of patients complaining of difficulty voiding owing to outlet obstruction or anal stenosis/stricture at follow-up; incontinence, assessed as the proportion of patients experiencing soiling or difficulty with hygiene or incontinence (any grade of incontinence) at follow-up; and perianal skin tag, considered as the proportion of patients complaining of perianal skin tags at follow-up, or perianal skin tags found on rectal examination at the clinic. Data collection The trials for inclusion were identified independently by two review authors by screening the titles and abstracts. The full text was sought for any references identified for potential inclusion by at least one of the authors, and further selection for inclusion was based on the full text. The following data were extracted from each study independently by two review authors: first author, year of publication, language of publication, country, inclusion and exclusion criteria, sample size, participant characteristics (such as age, sex, proportion of participants with grade IV haemorrhoids), study design, including details of the surgical interventions aimed at treating the haemorrhoids, outcomes described above, and risk of bias (see below). Any discrepancies were resolved by discussion; if there was disagreement, the final decision was taken by the senior author. The Cochrane Collaboration's risk of bias tool was used to assess the risk of bias of the included trials based on the following domains: allocation sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessors, incomplete outcome data and selective outcome reporting42. The other source of bias assessed was vested interest bias: whether a trial was conducted by a party with vested interests in the outcome of the trial, such as a drug manufacturer. For each of these risk domains, the studies were categorized as at low, uncertain or high risk of bias. Statistical analysis The systematic review and meta-analysis was conducted according to guidelines from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Group43. For binary data, based on the number of patients developing the adverse event, a binomial model was used for the analysis and the odds ratio (OR) was calculated. For outcomes where some patients may develop multiple adverse events, the total number of adverse events rather than the number of patients was imputed in the analysis, and a Poisson model was used. An arbitrary constant of 1 was added to the denominator and 0·5 to the numerator for trials with zero-event outcomes44. For continuous outcomes the mean difference (MD) was calculated. If the data were likely to be normally distributed, the median was used for the analysis when the mean was not available. If the s.d. was not available from a study, it was calculated from the standard error, P, c.i. or i.q.r., according to guidance given in the Cochrane Handbook for Systematic Reviews of Interventions45. If it was not possible to calculate the s.d. from the standard error, P, c.i. or i.q.r., the s.d. was imputed using the largest s.d. in other trials for that outcome. For each outcome of interest, Stata/IC 11 (StataCorp LP, College Station, Texas, USA) was used to draw a network plot of all the treatments assessed for that specific outcome. Any treatments not connected to the other treatments through the network plot were excluded from the analysis of that outcome. A Bayesian network meta-analysis was conducted using the Markov chain Monte Carlo method in WinBUGS 1.4 (MRC Biostatistics Unit, Cambridge, and Imperial College School of Medicine, London, UK). The treatment contrast (OR for binary outcomes, MD for continuous outcomes) for any two treatments was modelled as a function of comparisons between each individual treatment and an arbitrarily selected reference group46. The chosen reference group was the open haemorrhoidectomy group. The residual deviance and deviance information criterion (DIC) were used for assessing between-study heterogeneity, in accordance with guidance from the National Institute for Health and Care Excellence Decision Support Unit documents44. Three different models were run for each outcome: fixed-effect model, random-effects model and random-effects inconsistency model. The choice of model was based on the model fit. The DIC provides a measure of model fit that penalizes model complexity; therefore, a lower DIC indicated a better model fit44. The simpler model, which was a fixed-effect model, was used if the DIC values were similar between the fixed-effect and random-effects models. The random-effects model, which assumes variation between studies owing to heterogeneity and generates a wider c.i., was used if it resulted in a better model fit as indicated by a DIC lower than that of a fixed-effect model by at least 344. Evidence of inconsistency between direct and indirect comparisons was assessed by examining the geometry of the network diagrams carefully47. In addition, the deviance and DIC statistics of the consistency and inconsistency models were compared and, if the inconsistency model resulted in a better model fit than the consistency model, the results of the network meta-analysis were interpreted with caution47. The probability of ranking of a treatment (that a treatment ranks as the best treatment, second best treatment, third best treatment, etc.) for each outcome of interest was calculated. A probability below 90 per cent of being the best treatment for a particular outcome was not considered by the authors to be high enough to be confidently reported as the best treatment for that outcome of interest48. A network meta-analysis was performed to compare the surgical treatments identified in all included trials. A sensitivity network meta-analysis was undertaken of more recent trials published in or after 2005. The year 2005 was selected because some of the treatment methods compared were developed more recently and were used on patients after 2005. A further sensitivity network meta-analysis was carried out based on larger trials with a greater number of included patients (studies reporting on 80 or more patients). Results Eligible studies A total of 3224 references were identified through electronic searches of CENTRAL (679), MEDLINE (1319), Embase (619) and Science Citation Index Expanded (607). A further 167 references were identified from RCT registers (101 from World Health Organization International Clinical Trials Registry Platform and 66 from ClinicalTrials.gov). One more reference was identified for further assessment by scanning reference lists of the identified RCTs. Some 824 duplicates between databases were excluded. A further 2359 clearly irrelevant references were excluded after screening titles and reading abstracts. Two hundred and nine references were retrieved for further assessment. One hundred and eleven references were excluded after reviewing the studies in full. In total, 98 RCTs met the inclusion criteria7–40,49–112 (Fig. 1). A total of 7827 participants were included in the analysis (Table 1). The risk of bias in the included trials is summarized in Fig. 2, and described for each study in Fig. S1 (supporting information). Fig. 1 Open in new tabDownload slide PRISMA diagram showing selection of articles for review Table 1 Summary of studies included in the analysis Reference . Treatments . Total* . Grade IV . Abo-hashem et al.29 (2010) Harmonic® versus open 80 10 Altomare et al.7 (2008) LigaSure™ versus open 273 119 Ammaturo et al.49 (2012) Stapled versus open 79 n.r. Arslani et al.50 (2012) LigaSure™ versus stapled 98 0 Azolas et al.51 (2010) Closed versus open 48 n.r. Basdanis et al.52 (2005) LigaSure™ versus stapled 95 22 Bassi and Bergami53 (1997) Closed versus open 90 n.r. Bessa54 (2008) LigaSure™ versus open 110 67 Bikhchandani et al.55 (2005) Stapled versus open 84 13 Boccasanta et al.56 (2001) Stapled versus open 80 80 Bouini et al.57 (2012) LigaSure™ versus open 60 n.r. Brown et al.58 (2001) Stapled versus open 30 n.r. Bulut et al.59 (2006) Stapled versus open 30 n.r. Carrabetta et al.60 (2001) Closed versus open 103 n.r. Castellvi et al.61 (2009) LigaSure™ versus open 74 26 Cheetham et al.8 (2003) Stapled versus open 31 25 Chen et al.62 (2007) LigaSure™ versus stapled 86 0 Chung et al.17 (2005) Harmonic® versus stapled 88 0 Chung et al.40 (2002) Harmonic® versus open versus bipolar scissors 86 86 Chung and Wu63 (2003) LigaSure™ versus closed 61 45 Correa-Rovelo et al.11 (2002) Stapled versus closed 84 24 Dell'Abate et al.64 (2005) Stapled versus open 117 n.r. De Nardi et al.21 (2014) THD versus open 47 n.r. Denoya et al.22 (2013) THD versus closed 42 9 Enriquez-Navascues et al.33 (1993) Laser versus open 35 n.r. Fareed et al.12 (2009) LigaSure™ versus closed 80 14 Fazeli et al.65 (2011) LigaSure™ versus open 57 n.r. Festen et al.23 (2009) THD versus stapled 41 4·76 Filingeri et al.15 (2004) Radiofrequency versus submucosal 102 102 Filingeri et al.38 (2004) Radiofrequency versus open 36 36 Filingeri et al.13 (2010) Radiofrequency versus closed 22 22 Franceschilli et al.39 (2011) Radiofrequency versus open 210 17 Franklin et al.66 (2003) LigaSure™ versus open 34 n.r. Ganio et al.67 (2001) Stapled versus open 100 29 Ganio et al.68 (2007) Stapled versus open 100 29 Gao et al.69 (2008) Stapled versus open 100 15 Gençosmanoğlu et al.70 (2002) Closed versus open 80 n.r. Gentile et al.71 (2011) LigaSure™ versus open 52 52 Hasse et al.18 (2004) Stapled versus closed 76 n.r. Helmy19 (2000) Stapled versus open 40 10 Ho et al.72 (2000) Stapled versus open 119 49 Ho and Ho73 (2006) Stapled versus closed 50 n.r. Ho et al.9 (1997) Closed versus open 67 n.r. Hosch et al.16 (1998) Submucosal versus open 34 n.r. Huang et al.74 (2007) Stapled versus closed 596 0 Infantino et al.24 (2012) THD versus stapled 169 0 Jayne et al.75 (2002) LigaSure™ versus open 40 8 Kairaluoma et al.20 (2003) Stapled versus open 60 n.r. Khafagy et al.76 (2009) Stapled versus open versus THD 45 n.r. Khalil et al.77 (2000) Stapled versus closed 40 0 Khanna et al.25 (2010) LigaSure™ versus closed 48 n.r. Kim and Lee78 (2009) Starion™ versus Harmonic® 60 n.r. Kim et al.79 (2013) Stapled versus open 122 n.r. Kraemer et al.26 (2005) LigaSure™ versus stapled 50 2 Krska et al.80 (2003) Stapled versus open 50 n.r. Kwok et al.30 (2005) Harmonic® versus LigaSure™ 49 n.r. Lawes et al.27 (2004)† LigaSure™ versus open 34 n.r. Leventoğlu et al.81 (2008) Stapled versus open versus Harmonic® 60 n.r. Liao et al.82 (2008) Stapled versus open 76 n.r. Mehigan et al.83 (2000) Stapled versus open 40 10 Mik et al.84 (2008) Closed versus open 63 63 Milito et al.28 (2002) LigaSure™ versus open 56 29 Morpurgo et al.85 (2008) Stapled versus closed 20 0 Muzi et al.86 (2007) LigaSure™ versus open 250 81 Nyström et al.87 (2010) Stapled versus open 180 n.r. Ortiz et al.88 (2002) Stapled versus open 55 26 Ortiz et al.89 (2005) Stapled versus open 31 31 Ozer et al.90 (2008) Closed versus open versus Harmonic® 87 n.r. Palazzo et al.91 (2002) LigaSure™ versus open 34 n.r. Palimento et al.92 (2003) Stapled versus open 74 40 Pandini et al.34 (2006) Laser versus open 40 3 Pattana-Arun et al.93 (2006) LigaSure™ versus closed 45 4 Pescatori et al.94 (2000) Closed versus open 55 n.r. Peters et al.95 (2005)‡ LigaSure™ versus open 30 8 Picchio et al.96 (2006)§ Stapled versus open 74 40 Pokharel et al.97 (2009) Closed versus open 56 30 Racalbuto et al.98 (2004) Stapled versus open 100 n.r. Rahmani et al.99 (2012) Closed versus open 100 40 Ramadan et al.10 (2002) Harmonic® versus open 54 n.r. Rowsell et al.100 (2000) Stapled versus open 22 0 Sabanci et al.101 (2007) Stapled versus closed 100 n.r. Sakr102 (2010) LigaSure™ versus open 84 25 Sakr et al.103 (2010) LigaSure™ versus stapled 68 17 Senagore et al.14 (2004) Stapled versus closed 156 n.r. Senagore et al.35 (1993) Laser versus closed 86 n.r. Shelygin et al.31 (2003) Closed versus open versus Harmonic® 70 27 Smyth et al.104 (2003)¶ Stapled versus open 40 10 Stolfi et al.105 (2008) Stapled versus open 171 88 Tan et al.106 (2008) LigaSure™ versus open 43 n.r. Thorbeck and Montes107 (2002) LigaSure™ versus open 112 60 Tsunoda et al.32 (2011) Harmonic® versus LigaSure™ 60 24 Verre et al.108 (2013) THD versus stapled 122 73 Wang et al.36 (2005) Laser versus closed 86 0 Wang et al.109 (2006) LigaSure™ versus closed 84 15 Wang et al.37 (2007) Starion™ versus LigaSure™ 64 17 Wilson et al.110 (2002) Stapled versus open 89 n.r. You et al.111 (2005) Closed versus open 80 18 Zampieri et al.112 (2012) THD versus LigaSure™ 114 61 Reference . Treatments . Total* . Grade IV . Abo-hashem et al.29 (2010) Harmonic® versus open 80 10 Altomare et al.7 (2008) LigaSure™ versus open 273 119 Ammaturo et al.49 (2012) Stapled versus open 79 n.r. Arslani et al.50 (2012) LigaSure™ versus stapled 98 0 Azolas et al.51 (2010) Closed versus open 48 n.r. Basdanis et al.52 (2005) LigaSure™ versus stapled 95 22 Bassi and Bergami53 (1997) Closed versus open 90 n.r. Bessa54 (2008) LigaSure™ versus open 110 67 Bikhchandani et al.55 (2005) Stapled versus open 84 13 Boccasanta et al.56 (2001) Stapled versus open 80 80 Bouini et al.57 (2012) LigaSure™ versus open 60 n.r. Brown et al.58 (2001) Stapled versus open 30 n.r. Bulut et al.59 (2006) Stapled versus open 30 n.r. Carrabetta et al.60 (2001) Closed versus open 103 n.r. Castellvi et al.61 (2009) LigaSure™ versus open 74 26 Cheetham et al.8 (2003) Stapled versus open 31 25 Chen et al.62 (2007) LigaSure™ versus stapled 86 0 Chung et al.17 (2005) Harmonic® versus stapled 88 0 Chung et al.40 (2002) Harmonic® versus open versus bipolar scissors 86 86 Chung and Wu63 (2003) LigaSure™ versus closed 61 45 Correa-Rovelo et al.11 (2002) Stapled versus closed 84 24 Dell'Abate et al.64 (2005) Stapled versus open 117 n.r. De Nardi et al.21 (2014) THD versus open 47 n.r. Denoya et al.22 (2013) THD versus closed 42 9 Enriquez-Navascues et al.33 (1993) Laser versus open 35 n.r. Fareed et al.12 (2009) LigaSure™ versus closed 80 14 Fazeli et al.65 (2011) LigaSure™ versus open 57 n.r. Festen et al.23 (2009) THD versus stapled 41 4·76 Filingeri et al.15 (2004) Radiofrequency versus submucosal 102 102 Filingeri et al.38 (2004) Radiofrequency versus open 36 36 Filingeri et al.13 (2010) Radiofrequency versus closed 22 22 Franceschilli et al.39 (2011) Radiofrequency versus open 210 17 Franklin et al.66 (2003) LigaSure™ versus open 34 n.r. Ganio et al.67 (2001) Stapled versus open 100 29 Ganio et al.68 (2007) Stapled versus open 100 29 Gao et al.69 (2008) Stapled versus open 100 15 Gençosmanoğlu et al.70 (2002) Closed versus open 80 n.r. Gentile et al.71 (2011) LigaSure™ versus open 52 52 Hasse et al.18 (2004) Stapled versus closed 76 n.r. Helmy19 (2000) Stapled versus open 40 10 Ho et al.72 (2000) Stapled versus open 119 49 Ho and Ho73 (2006) Stapled versus closed 50 n.r. Ho et al.9 (1997) Closed versus open 67 n.r. Hosch et al.16 (1998) Submucosal versus open 34 n.r. Huang et al.74 (2007) Stapled versus closed 596 0 Infantino et al.24 (2012) THD versus stapled 169 0 Jayne et al.75 (2002) LigaSure™ versus open 40 8 Kairaluoma et al.20 (2003) Stapled versus open 60 n.r. Khafagy et al.76 (2009) Stapled versus open versus THD 45 n.r. Khalil et al.77 (2000) Stapled versus closed 40 0 Khanna et al.25 (2010) LigaSure™ versus closed 48 n.r. Kim and Lee78 (2009) Starion™ versus Harmonic® 60 n.r. Kim et al.79 (2013) Stapled versus open 122 n.r. Kraemer et al.26 (2005) LigaSure™ versus stapled 50 2 Krska et al.80 (2003) Stapled versus open 50 n.r. Kwok et al.30 (2005) Harmonic® versus LigaSure™ 49 n.r. Lawes et al.27 (2004)† LigaSure™ versus open 34 n.r. Leventoğlu et al.81 (2008) Stapled versus open versus Harmonic® 60 n.r. Liao et al.82 (2008) Stapled versus open 76 n.r. Mehigan et al.83 (2000) Stapled versus open 40 10 Mik et al.84 (2008) Closed versus open 63 63 Milito et al.28 (2002) LigaSure™ versus open 56 29 Morpurgo et al.85 (2008) Stapled versus closed 20 0 Muzi et al.86 (2007) LigaSure™ versus open 250 81 Nyström et al.87 (2010) Stapled versus open 180 n.r. Ortiz et al.88 (2002) Stapled versus open 55 26 Ortiz et al.89 (2005) Stapled versus open 31 31 Ozer et al.90 (2008) Closed versus open versus Harmonic® 87 n.r. Palazzo et al.91 (2002) LigaSure™ versus open 34 n.r. Palimento et al.92 (2003) Stapled versus open 74 40 Pandini et al.34 (2006) Laser versus open 40 3 Pattana-Arun et al.93 (2006) LigaSure™ versus closed 45 4 Pescatori et al.94 (2000) Closed versus open 55 n.r. Peters et al.95 (2005)‡ LigaSure™ versus open 30 8 Picchio et al.96 (2006)§ Stapled versus open 74 40 Pokharel et al.97 (2009) Closed versus open 56 30 Racalbuto et al.98 (2004) Stapled versus open 100 n.r. Rahmani et al.99 (2012) Closed versus open 100 40 Ramadan et al.10 (2002) Harmonic® versus open 54 n.r. Rowsell et al.100 (2000) Stapled versus open 22 0 Sabanci et al.101 (2007) Stapled versus closed 100 n.r. Sakr102 (2010) LigaSure™ versus open 84 25 Sakr et al.103 (2010) LigaSure™ versus stapled 68 17 Senagore et al.14 (2004) Stapled versus closed 156 n.r. Senagore et al.35 (1993) Laser versus closed 86 n.r. Shelygin et al.31 (2003) Closed versus open versus Harmonic® 70 27 Smyth et al.104 (2003)¶ Stapled versus open 40 10 Stolfi et al.105 (2008) Stapled versus open 171 88 Tan et al.106 (2008) LigaSure™ versus open 43 n.r. Thorbeck and Montes107 (2002) LigaSure™ versus open 112 60 Tsunoda et al.32 (2011) Harmonic® versus LigaSure™ 60 24 Verre et al.108 (2013) THD versus stapled 122 73 Wang et al.36 (2005) Laser versus closed 86 0 Wang et al.109 (2006) LigaSure™ versus closed 84 15 Wang et al.37 (2007) Starion™ versus LigaSure™ 64 17 Wilson et al.110 (2002) Stapled versus open 89 n.r. You et al.111 (2005) Closed versus open 80 18 Zampieri et al.112 (2012) THD versus LigaSure™ 114 61 * Number of patients with grade III or IV haemorrhoids. Study reporting long-term outcomes of participants reported previously in † Palazzo et al.91, ‡ Jayne et al.75, § Palimento et al.92 and ¶ Mehigan et al.83. n.r., Not reported; THD, transanal haemorrhoid dearterialization. Open in new tab Table 1 Summary of studies included in the analysis Reference . Treatments . Total* . Grade IV . Abo-hashem et al.29 (2010) Harmonic® versus open 80 10 Altomare et al.7 (2008) LigaSure™ versus open 273 119 Ammaturo et al.49 (2012) Stapled versus open 79 n.r. Arslani et al.50 (2012) LigaSure™ versus stapled 98 0 Azolas et al.51 (2010) Closed versus open 48 n.r. Basdanis et al.52 (2005) LigaSure™ versus stapled 95 22 Bassi and Bergami53 (1997) Closed versus open 90 n.r. Bessa54 (2008) LigaSure™ versus open 110 67 Bikhchandani et al.55 (2005) Stapled versus open 84 13 Boccasanta et al.56 (2001) Stapled versus open 80 80 Bouini et al.57 (2012) LigaSure™ versus open 60 n.r. Brown et al.58 (2001) Stapled versus open 30 n.r. Bulut et al.59 (2006) Stapled versus open 30 n.r. Carrabetta et al.60 (2001) Closed versus open 103 n.r. Castellvi et al.61 (2009) LigaSure™ versus open 74 26 Cheetham et al.8 (2003) Stapled versus open 31 25 Chen et al.62 (2007) LigaSure™ versus stapled 86 0 Chung et al.17 (2005) Harmonic® versus stapled 88 0 Chung et al.40 (2002) Harmonic® versus open versus bipolar scissors 86 86 Chung and Wu63 (2003) LigaSure™ versus closed 61 45 Correa-Rovelo et al.11 (2002) Stapled versus closed 84 24 Dell'Abate et al.64 (2005) Stapled versus open 117 n.r. De Nardi et al.21 (2014) THD versus open 47 n.r. Denoya et al.22 (2013) THD versus closed 42 9 Enriquez-Navascues et al.33 (1993) Laser versus open 35 n.r. Fareed et al.12 (2009) LigaSure™ versus closed 80 14 Fazeli et al.65 (2011) LigaSure™ versus open 57 n.r. Festen et al.23 (2009) THD versus stapled 41 4·76 Filingeri et al.15 (2004) Radiofrequency versus submucosal 102 102 Filingeri et al.38 (2004) Radiofrequency versus open 36 36 Filingeri et al.13 (2010) Radiofrequency versus closed 22 22 Franceschilli et al.39 (2011) Radiofrequency versus open 210 17 Franklin et al.66 (2003) LigaSure™ versus open 34 n.r. Ganio et al.67 (2001) Stapled versus open 100 29 Ganio et al.68 (2007) Stapled versus open 100 29 Gao et al.69 (2008) Stapled versus open 100 15 Gençosmanoğlu et al.70 (2002) Closed versus open 80 n.r. Gentile et al.71 (2011) LigaSure™ versus open 52 52 Hasse et al.18 (2004) Stapled versus closed 76 n.r. Helmy19 (2000) Stapled versus open 40 10 Ho et al.72 (2000) Stapled versus open 119 49 Ho and Ho73 (2006) Stapled versus closed 50 n.r. Ho et al.9 (1997) Closed versus open 67 n.r. Hosch et al.16 (1998) Submucosal versus open 34 n.r. Huang et al.74 (2007) Stapled versus closed 596 0 Infantino et al.24 (2012) THD versus stapled 169 0 Jayne et al.75 (2002) LigaSure™ versus open 40 8 Kairaluoma et al.20 (2003) Stapled versus open 60 n.r. Khafagy et al.76 (2009) Stapled versus open versus THD 45 n.r. Khalil et al.77 (2000) Stapled versus closed 40 0 Khanna et al.25 (2010) LigaSure™ versus closed 48 n.r. Kim and Lee78 (2009) Starion™ versus Harmonic® 60 n.r. Kim et al.79 (2013) Stapled versus open 122 n.r. Kraemer et al.26 (2005) LigaSure™ versus stapled 50 2 Krska et al.80 (2003) Stapled versus open 50 n.r. Kwok et al.30 (2005) Harmonic® versus LigaSure™ 49 n.r. Lawes et al.27 (2004)† LigaSure™ versus open 34 n.r. Leventoğlu et al.81 (2008) Stapled versus open versus Harmonic® 60 n.r. Liao et al.82 (2008) Stapled versus open 76 n.r. Mehigan et al.83 (2000) Stapled versus open 40 10 Mik et al.84 (2008) Closed versus open 63 63 Milito et al.28 (2002) LigaSure™ versus open 56 29 Morpurgo et al.85 (2008) Stapled versus closed 20 0 Muzi et al.86 (2007) LigaSure™ versus open 250 81 Nyström et al.87 (2010) Stapled versus open 180 n.r. Ortiz et al.88 (2002) Stapled versus open 55 26 Ortiz et al.89 (2005) Stapled versus open 31 31 Ozer et al.90 (2008) Closed versus open versus Harmonic® 87 n.r. Palazzo et al.91 (2002) LigaSure™ versus open 34 n.r. Palimento et al.92 (2003) Stapled versus open 74 40 Pandini et al.34 (2006) Laser versus open 40 3 Pattana-Arun et al.93 (2006) LigaSure™ versus closed 45 4 Pescatori et al.94 (2000) Closed versus open 55 n.r. Peters et al.95 (2005)‡ LigaSure™ versus open 30 8 Picchio et al.96 (2006)§ Stapled versus open 74 40 Pokharel et al.97 (2009) Closed versus open 56 30 Racalbuto et al.98 (2004) Stapled versus open 100 n.r. Rahmani et al.99 (2012) Closed versus open 100 40 Ramadan et al.10 (2002) Harmonic® versus open 54 n.r. Rowsell et al.100 (2000) Stapled versus open 22 0 Sabanci et al.101 (2007) Stapled versus closed 100 n.r. Sakr102 (2010) LigaSure™ versus open 84 25 Sakr et al.103 (2010) LigaSure™ versus stapled 68 17 Senagore et al.14 (2004) Stapled versus closed 156 n.r. Senagore et al.35 (1993) Laser versus closed 86 n.r. Shelygin et al.31 (2003) Closed versus open versus Harmonic® 70 27 Smyth et al.104 (2003)¶ Stapled versus open 40 10 Stolfi et al.105 (2008) Stapled versus open 171 88 Tan et al.106 (2008) LigaSure™ versus open 43 n.r. Thorbeck and Montes107 (2002) LigaSure™ versus open 112 60 Tsunoda et al.32 (2011) Harmonic® versus LigaSure™ 60 24 Verre et al.108 (2013) THD versus stapled 122 73 Wang et al.36 (2005) Laser versus closed 86 0 Wang et al.109 (2006) LigaSure™ versus closed 84 15 Wang et al.37 (2007) Starion™ versus LigaSure™ 64 17 Wilson et al.110 (2002) Stapled versus open 89 n.r. You et al.111 (2005) Closed versus open 80 18 Zampieri et al.112 (2012) THD versus LigaSure™ 114 61 Reference . Treatments . Total* . Grade IV . Abo-hashem et al.29 (2010) Harmonic® versus open 80 10 Altomare et al.7 (2008) LigaSure™ versus open 273 119 Ammaturo et al.49 (2012) Stapled versus open 79 n.r. Arslani et al.50 (2012) LigaSure™ versus stapled 98 0 Azolas et al.51 (2010) Closed versus open 48 n.r. Basdanis et al.52 (2005) LigaSure™ versus stapled 95 22 Bassi and Bergami53 (1997) Closed versus open 90 n.r. Bessa54 (2008) LigaSure™ versus open 110 67 Bikhchandani et al.55 (2005) Stapled versus open 84 13 Boccasanta et al.56 (2001) Stapled versus open 80 80 Bouini et al.57 (2012) LigaSure™ versus open 60 n.r. Brown et al.58 (2001) Stapled versus open 30 n.r. Bulut et al.59 (2006) Stapled versus open 30 n.r. Carrabetta et al.60 (2001) Closed versus open 103 n.r. Castellvi et al.61 (2009) LigaSure™ versus open 74 26 Cheetham et al.8 (2003) Stapled versus open 31 25 Chen et al.62 (2007) LigaSure™ versus stapled 86 0 Chung et al.17 (2005) Harmonic® versus stapled 88 0 Chung et al.40 (2002) Harmonic® versus open versus bipolar scissors 86 86 Chung and Wu63 (2003) LigaSure™ versus closed 61 45 Correa-Rovelo et al.11 (2002) Stapled versus closed 84 24 Dell'Abate et al.64 (2005) Stapled versus open 117 n.r. De Nardi et al.21 (2014) THD versus open 47 n.r. Denoya et al.22 (2013) THD versus closed 42 9 Enriquez-Navascues et al.33 (1993) Laser versus open 35 n.r. Fareed et al.12 (2009) LigaSure™ versus closed 80 14 Fazeli et al.65 (2011) LigaSure™ versus open 57 n.r. Festen et al.23 (2009) THD versus stapled 41 4·76 Filingeri et al.15 (2004) Radiofrequency versus submucosal 102 102 Filingeri et al.38 (2004) Radiofrequency versus open 36 36 Filingeri et al.13 (2010) Radiofrequency versus closed 22 22 Franceschilli et al.39 (2011) Radiofrequency versus open 210 17 Franklin et al.66 (2003) LigaSure™ versus open 34 n.r. Ganio et al.67 (2001) Stapled versus open 100 29 Ganio et al.68 (2007) Stapled versus open 100 29 Gao et al.69 (2008) Stapled versus open 100 15 Gençosmanoğlu et al.70 (2002) Closed versus open 80 n.r. Gentile et al.71 (2011) LigaSure™ versus open 52 52 Hasse et al.18 (2004) Stapled versus closed 76 n.r. Helmy19 (2000) Stapled versus open 40 10 Ho et al.72 (2000) Stapled versus open 119 49 Ho and Ho73 (2006) Stapled versus closed 50 n.r. Ho et al.9 (1997) Closed versus open 67 n.r. Hosch et al.16 (1998) Submucosal versus open 34 n.r. Huang et al.74 (2007) Stapled versus closed 596 0 Infantino et al.24 (2012) THD versus stapled 169 0 Jayne et al.75 (2002) LigaSure™ versus open 40 8 Kairaluoma et al.20 (2003) Stapled versus open 60 n.r. Khafagy et al.76 (2009) Stapled versus open versus THD 45 n.r. Khalil et al.77 (2000) Stapled versus closed 40 0 Khanna et al.25 (2010) LigaSure™ versus closed 48 n.r. Kim and Lee78 (2009) Starion™ versus Harmonic® 60 n.r. Kim et al.79 (2013) Stapled versus open 122 n.r. Kraemer et al.26 (2005) LigaSure™ versus stapled 50 2 Krska et al.80 (2003) Stapled versus open 50 n.r. Kwok et al.30 (2005) Harmonic® versus LigaSure™ 49 n.r. Lawes et al.27 (2004)† LigaSure™ versus open 34 n.r. Leventoğlu et al.81 (2008) Stapled versus open versus Harmonic® 60 n.r. Liao et al.82 (2008) Stapled versus open 76 n.r. Mehigan et al.83 (2000) Stapled versus open 40 10 Mik et al.84 (2008) Closed versus open 63 63 Milito et al.28 (2002) LigaSure™ versus open 56 29 Morpurgo et al.85 (2008) Stapled versus closed 20 0 Muzi et al.86 (2007) LigaSure™ versus open 250 81 Nyström et al.87 (2010) Stapled versus open 180 n.r. Ortiz et al.88 (2002) Stapled versus open 55 26 Ortiz et al.89 (2005) Stapled versus open 31 31 Ozer et al.90 (2008) Closed versus open versus Harmonic® 87 n.r. Palazzo et al.91 (2002) LigaSure™ versus open 34 n.r. Palimento et al.92 (2003) Stapled versus open 74 40 Pandini et al.34 (2006) Laser versus open 40 3 Pattana-Arun et al.93 (2006) LigaSure™ versus closed 45 4 Pescatori et al.94 (2000) Closed versus open 55 n.r. Peters et al.95 (2005)‡ LigaSure™ versus open 30 8 Picchio et al.96 (2006)§ Stapled versus open 74 40 Pokharel et al.97 (2009) Closed versus open 56 30 Racalbuto et al.98 (2004) Stapled versus open 100 n.r. Rahmani et al.99 (2012) Closed versus open 100 40 Ramadan et al.10 (2002) Harmonic® versus open 54 n.r. Rowsell et al.100 (2000) Stapled versus open 22 0 Sabanci et al.101 (2007) Stapled versus closed 100 n.r. Sakr102 (2010) LigaSure™ versus open 84 25 Sakr et al.103 (2010) LigaSure™ versus stapled 68 17 Senagore et al.14 (2004) Stapled versus closed 156 n.r. Senagore et al.35 (1993) Laser versus closed 86 n.r. Shelygin et al.31 (2003) Closed versus open versus Harmonic® 70 27 Smyth et al.104 (2003)¶ Stapled versus open 40 10 Stolfi et al.105 (2008) Stapled versus open 171 88 Tan et al.106 (2008) LigaSure™ versus open 43 n.r. Thorbeck and Montes107 (2002) LigaSure™ versus open 112 60 Tsunoda et al.32 (2011) Harmonic® versus LigaSure™ 60 24 Verre et al.108 (2013) THD versus stapled 122 73 Wang et al.36 (2005) Laser versus closed 86 0 Wang et al.109 (2006) LigaSure™ versus closed 84 15 Wang et al.37 (2007) Starion™ versus LigaSure™ 64 17 Wilson et al.110 (2002) Stapled versus open 89 n.r. You et al.111 (2005) Closed versus open 80 18 Zampieri et al.112 (2012) THD versus LigaSure™ 114 61 * Number of patients with grade III or IV haemorrhoids. Study reporting long-term outcomes of participants reported previously in † Palazzo et al.91, ‡ Jayne et al.75, § Palimento et al.92 and ¶ Mehigan et al.83. n.r., Not reported; THD, transanal haemorrhoid dearterialization. Open in new tab Fig. 2 Open in new tabDownload slide Summary of risk of bias across all included studies Surgical treatments compared The following elective surgical treatments were identified for the management of grade III and IV haemorrhoids: open haemorrhoidectomy or Milligan–Morgan haemorrhoidectomy, performed with a scalpel, conventional scissors or diathermy; closed haemorrhoidectomy or Ferguson haemorrhoidectomy, performed with a scalpel, conventional scissors or diathermy; submucosal haemorrhoidectomy or Parks' haemorrhoidectomy, carried out with a scalpel, conventional scissors or diathermy; stapled haemorrhoidectomy or haemorrhoidopexy, procedure for prolapse and haemorrhoids, or Longo procedure/technique; THD or haemorrhoidal artery ligation operation, performed with or without mucopexy, and with or without Doppler guidance; haemorrhoidectomy using a LigaSure™ device; Harmonic® haemorrhoidectomy using a Harmonic® or ultrasonic scalpel; laser haemorrhoidectomy done with a Nd : YAG or carbon dioxide laser; haemorrhoidectomy using the Starion™ system; haemorrhoidectomy performed with a radiofrequency device; and haemorrhoidectomy carried out with bipolar scissors. The results of all pairwise comparisons of the various surgical treatments for the outcomes of interest are shown in Table S2 (supporting information), and statistically significant results only in Table S3 (supporting information). The surgical treatments with the highest probability of ranking from best to worst (1st to 11th) for the outcomes of interest are summarized in Table S4 (supporting information). Fig. 3 shows an example of a network plot; similar plots were created for all outcomes of interest. Fig. 3 Open in new tabDownload slide Network plot for postoperative complications. Similar network plots were produced for each outcome of interest. Circles represent the intervention as a node in the network; lines represent direct comparisons using randomized clinical trials (RCTs); the line thickness indicates the number of RCTs included in each comparison Postoperative complications Seventy-seven trials provided data on 6596 participants and all 11 surgical treatments, for the network meta-analysis on postoperative complications; the network plot is shown in Fig. 3. The fixed-effect model was preferred based on the DIC statistics, and there was no evidence of inconsistency in the networks. The closed and radiofrequency haemorrhoidectomy groups had significantly more postoperative complications than the open, stapled, LigaSure™, Harmonic® and THD groups. Open haemorrhoidectomy had significantly more postoperative complications than the LigaSure™, Harmonic® and THD groups. Furthermore, Harmonic® haemorrhoidectomy resulted in significantly fewer postoperative complications than open, closed, stapled, submucosal and radiofrequency haemorrhoidectomy. The LigaSure™ and THD groups had significantly fewer postoperative complications than the open, closed and radiofrequency haemorrhoidectomy groups. Urinary retention, wound complications, constipation and anal fissure Seventy-one trials (5536 participants; 11 treatments) provided data for the network meta-analysis on urinary retention. Twenty-six trials (2018 participants; 8 treatments) provided data on wound complications (wound infection, discharge, dehiscence, not healing). The random-effects model was preferred for both outcomes based on the DIC statistics, and there was no evidence of inconsistency. The incidence of postoperative urinary retention was significantly higher in the closed haemorrhoidectomy group compared with the LigaSure™ and Harmonic® groups. Furthermore, the closed haemorrhoidectomy group had significantly more wound complications than the open and stapled haemorrhoidectomy groups. There was no significant difference in the other comparisons for these outcomes. Twenty-five trials (2565 participants; 9 treatments) provided data on constipation/faecal impaction. Closed haemorrhoidectomy was complicated by constipation significantly more often than the open and Harmonic® procedures. There was no significant difference in the other comparisons. Postoperative anal fissure was reported by 20 trials (1897 participants; 7 treatments) and there was no significant difference between the surgical treatments for this outcome. The fixed-effect model was preferred for the above two outcomes, and there was no evidence of inconsistency. Postoperative bleeding and emergency reoperation Seventy-one trials provided data for the network meta-analysis on the proportion of patients with postoperative bleeding (6191 participants; 11 treatments). The fixed-effect model was preferred, and there was no evidence of inconsistency in the network. Pairwise comparison of the surgical treatments showed that significantly fewer people had postoperative bleeding after THD compared with open or stapled haemorrhoidectomy. Thirty-nine trials reported on emergency reoperation (3272 participants; 8 treatments). In total, 65 patients (2·0 per cent) needed expedited or urgent reoperation to treat early postoperative complications. In 57 patients (88 per cent) reoperation was needed to stop postoperative bleeding. In the other reoperations, five patients required examination under anaesthesia for significant postoperative pain, one patient in the stapled group required incision and drainage of a submucosal haematoma, another patient in the stapled group required incision of painful thrombosed haemorrhoidal tissue distal to the staple line, and one patient required reoperation to treat perianal sepsis after open haemorrhoidectomy. Network meta-analysis of the trials reporting on reoperation showed that the THD group had significantly fewer reoperations than the open, closed, stapled and LigaSure™ groups. Furthermore, THD had a relatively high probability of being the best treatment for this outcome (P = 0·710) (Table S3, supporting information). Duration of surgery and operative blood loss Seventy-three trials (6140 participants; 11 treatments) were included in the analysis of duration of surgery. The random-effects model was preferred based on the DIC statistics, and there was no evidence of inconsistency in the networks. The open, closed, submucosal and laser haemorrhoidectomy groups had a significantly longer duration of surgery than the stapled, LigaSure™, Harmonic®, THD, Starion™ and radiofrequency haemorrhoidectomy groups. In addition, closed, submucosal and laser haemorrhoidectomies took significantly longer than bipolar scissors haemorrhoidectomy. The closed and laser groups also had a significantly longer operating time than the open haemorrhoidectomy group. Laser haemorrhoidectomy was ranked the worst treatment for this outcome with greater than 90 per cent probability (P = 0·909) (Table S3, supporting information). Twenty trials (2016 participants; 9 treatments) were included in the analysis for operative blood loss. The random-effects model was preferred based on the DIC statistics, and there was evidence of inconsistency between trials because the difference in DIC between the consistency and inconsistency models was significant. The pairwise mean differences of the various group comparisons showed that the open, closed and THD groups had significantly more operative blood loss than the stapled, LigaSure™, Harmonic®, Starion™ and bipolar scissors groups. Length of hospital stay, time to first bowel movement and time to normal activities Forty-six trials (4321 participants; 11 treatments) provided data for the network meta-analysis on length of hospital stay. The random-effects model was preferred based on the DIC statistics, and there was no evidence of inconsistency in the networks. The pairwise comparison of interventions showed the stapled group and the THD group to have a significantly shorter length of hospital stay than the open, closed and Harmonic® groups. Furthermore, stapled haemorrhoidectomy was associated with a significantly shorter length of hospital stay than LigaSure™ haemorrhoidectomy. THD resulted in a significantly shorter hospital stay than laser haemorrhoidectomy. In addition, open and LigaSure™ haemorrhoidectomies had a significantly shorter length of hospital stay compared with closed haemorrhoidectomy. Nineteen trials (1513 participants; 7 treatments) reported on time to first bowel movement. The fixed-effect model was preferred based on the DIC statistics, and there was no evidence of inconsistency in the networks. Pairwise comparisons of the treatments showed the THD and Harmonic® groups to have a significantly shorter time to the first bowel movement compared with the open and closed haemorrhoidectomy groups. THD also had a significantly shorter time to the first bowel movement than the bipolar scissors group. Time to first bowel movement was significantly shorter for LigaSure™ than for closed haemorrhoidectomy. THD ranked the best treatment for this outcome with high probability (P = 0·853) (Table S3, supporting information). Time to return to work or normal activities was reported by 42 trials (4125 participants; 11 treatments). The random-effects model was preferred for this outcome based on the DIC statistics, and there was no evidence of inconsistency in the networks. Patients in the stapled, LigaSure™ and Harmonic® groups needed a significantly shorter time to return to normal activities than those in the open and closed haemorrhoidectomy groups. Time to return to normal activities was shorter after stapled than laser haemorrhoidectomy. Pain Fifty-three trials (4184 participants; 10 treatments) reported on pain on postoperative day 1. The random-effects model was preferred for this outcome, and there was no evidence of inconsistency in the networks. The pairwise mean differences of the treatments showed open and closed haemorrhoidectomies to result in significantly more pain on postoperative day 1 than stapled, LigaSure™, Harmonic®, THD and Starion™ procedures. Pain on postoperative day 7 was reported by 35 trials (2856 participants; 9 treatments). The random-effects model was used, and there was no evidence of inconsistency. The open group had significantly more pain on day 7 compared with the stapled group. There was no significant difference in the other comparisons. Eighteen studies reported on pain on postoperative day 14 (1231 participants; 7 treatments). The open and closed haemorrhoidectomy groups had significantly more pain on postoperative day 14 compared with the stapled and LigaSure™ groups. LigaSure™ haemorrhoidectomy resulted in significantly more pain on day 14 than stapled haemorrhoidectomy. The open haemorrhoidectomy group had significantly more pain compared with the closed haemorrhoidectomy group. Recurrence of haemorrhoids and recurrent haemorrhoidal symptoms Forty-seven trials (3863 participants; 9 treatments) provided data for the network meta-analysis on recurrence of haemorrhoids. The fixed-effect model was preferred based on the DIC statistics, and there was no evidence of inconsistency. Pairwise comparison of the groups showed a significant increase in the proportion of people with recurrence of haemorrhoids in the stapled and THD groups compared with the open, closed and LigaSure™ haemorrhoidectomy groups. THD was associated with a higher recurrence rate than laser and radiofrequency haemorrhoidectomies. THD and stapled haemorrhoidectomies had a higher probability of doing worse than the other treatments, with regard to recurrence of haemorrhoids. THD ranked as worst treatment for this outcome, with a high probability (P = 0·785) (Table S3, supporting information). Recurrence of haemorrhoidal symptoms was reported by 47 trials (3765 participants; 7 treatments). The fixed-effect model was preferred based on the DIC statistics, and there was no evidence of inconsistency. The rate of recurrence of haemorrhoidal symptoms was significantly higher for stapled haemorrhoidectomy than for open and LigaSure™ haemorrhoidectomies. There was no significant difference in the other comparisons. Recurrent bleeding, recurrent pain and recurrent pruritus Recurrent haemorrhoidal bleeding was reported by 33 trials (2651 participants; 6 treatments), and 29 trials (1969 participants; 6 treatments) provided data on recurrent pain or discomfort related to haemorrhoids. There was no evidence of any significant difference between the different interventions for both outcomes. Eighteen trials (1176 participants; 7 treatments) provided data on recurrence of pruritus owing haemorrhoids, and THD was found to have a lower rate than closed haemorrhoidectomy. Anal stenosis, incontinence and skin tag Fifty-one trials (4793 participants; 11 treatments) reported on the proportion of patients complaining of difficulty voiding owing to outlet obstruction or anal stenosis/stricture at follow-up. Fifty-three trials (3856 participants; 9 treatments) reported on the proportion of patients experiencing soiling or difficulty with hygiene or incontinence at follow-up. Perianal skin tags at follow-up were reported by 24 trials (1766 participants; 5 treatments). Pairwise comparisons showed no significant difference between the surgical treatments for these three outcomes. Sensitivity analysis – recent trials Network meta-analysis of more recent reports, published in or after 2005, revealed that the closed haemorrhoidectomy group had significantly more complications than the open, stapled, LigaSure™, Harmonic® and THD groups. Radiofrequency haemorrhoidectomy resulted in significantly more postoperative complications compared with open, stapled, LigaSure™, Harmonic® and THD treatments. Open haemorrhoidectomy resulted in significantly more complications than the LigaSure™ procedure. THD was associated with significantly fewer episodes of postoperative bleeding than open, stapled and radiofrequency haemorrhoidectomies. THD had a relatively high probability of being the best treatment for postoperative bleeding (P = 0·704). THD also had a significantly lower emergency reoperation rate than open haemorrhoidectomy, and a relatively high probability of being the best treatment for this outcome too (P = 0·713). Pairwise comparisons of treatments reported in more recent trials revealed that the duration of surgery was significantly longer in the open and closed haemorrhoidectomy groups than in the stapled, LigaSure™, Harmonic®, THD, Starion™ and radiofrequency haemorrhoidectomy groups. Furthermore, the stapled and THD groups had a significantly shorter hospital stay compared with the open and closed haemorrhoidectomy groups. Time to return to normal activities was significantly shorter after stapled than after open and closed haemorrhoidectomies. Sensitivity analysis showed that open and closed haemorrhoidectomies caused significantly more pain on postoperative day 1 than stapled and THD procedures. For postoperative day 7, open haemorrhoidectomy resulted in significantly more pain than stapled haemorrhoidectomy. Moreover, the stapled and THD groups had significantly higher rates of recurrence of haemorrhoids than the open and LigaSure™ groups. In addition, stapled haemorrhoidectomy was associated with significantly greater recurrence of haemorrhoidal symptoms than LigaSure™ haemorrhoidectomy. Sensitivity analysis – larger trials Network meta-analysis of the studies reporting on more patients (80 patients and above) revealed that the closed haemorrhoidectomy group had significantly more complications than the open, stapled, LigaSure™ and THD groups. Open haemorrhoidectomy had significantly more complications than LigaSure™ haemorrhoidectomy. Radiofrequency and submucosal haemorrhoidectomies resulted in significantly more postoperative complications compared with open, stapled, LigaSure™, Harmonic®, THD and laser haemorrhoidectomies. There were fewer episodes of postoperative bleeding after THD than after open, stapled and radiofrequency haemorrhoidectomies. THD was also associated with a significantly lower reoperation rate compared with open and closed haemorrhoidectomy, and had a high probability of being the best treatment for this outcome (P = 0·916). Moreover, the duration of surgery was significantly longer for open and closed haemorrhoidectomies compared with stapled, LigaSure™, THD and radiofrequency procedures. Laser haemorrhoidectomy had a significantly longer operating time than open, stapled, LigaSure™, Harmonic®, THD, radiofrequency and bipolar scissors haemorrhoidectomy. Radiofrequency haemorrhoidectomy was associated with a significantly shorter operating time than open, closed, Harmonic®, laser and submucosal haemorrhoidectomy, and had a relatively high probability of being the best treatment in terms of duration of surgery (P = 0·892). Pairwise comparisons of treatments reported in larger trials revealed that hospital stay was significantly shorter after stapled haemorrhoidectomy compared with open and closed haemorrhoidectomy. Stapled haemorrhoidectomy had a relatively high probability of being the best treatment in terms of hospital stay (P = 0·780). Furthermore, after LigaSure™ and stapled haemorrhoidectomy patients returned significantly more quickly to normal activities than those in the open and closed haemorrhoidectomy groups. Sensitivity analysis also showed that stapled and LigaSure™ haemorrhoidectomies caused significantly less pain on postoperative day 1 than open and closed procedures. On postoperative day 7, stapled haemorrhoidectomy resulted in significantly less pain compared with open haemorrhoidectomy. Finally, the stapled and THD groups had significantly higher haemorrhoid recurrence rates than the open and LigaSure™ groups. Discussion Many standard pairwise meta-analyses4,113–132 have been published previously comparing surgical treatments for haemorrhoids. One disadvantage of these meta-analyses is that they could compare only two surgical treatments directly, rather than all available surgical treatments at once. In addition, most of the meta-analyses published on the surgical treatment of haemorrhoids grouped conventional and excisional haemorrhoidectomies together, rather than comparing them individually. For example, open and closed haemorrhoidectomies were grouped together for comparison with LigaSure™ haemorrhoidectomy114,121–122,125,130. Open, closed and submucosal haemorrhoidectomies were also grouped together (performed using scissors, diathermy, laser, LigaSure™ or Harmonic® scalpel) and were compared with stapled haemorrhoidectomy4,114–115,118,120,126,129,131. Importantly, this network meta-analysis allowed simultaneous comparison of all surgical treatments available for grade III and IV haemorrhoids. A network meta-analysis was ideal for this topic, where multiple interventions have been used and compared for the same disease and outcomes in different head-to-head comparisons. Another advantage of this network meta-analysis over previous standard pairwise meta-analyses was that it combined direct evidence within trials and indirect evidence across trials, facilitating indirect comparisons of multiple interventions that have not been studied in a head-to-head fashion133,134. Therefore, it allowed the relative effectiveness of different surgical treatments to be assessed even when they had not been compared directly in individual RCTs. Because network meta-analyses include evidence from both direct and indirect comparisons, the power may be better than in standard pairwise meta-analyses that include only direct evidence135. Moreover, compared with a standard pairwise meta-analysis, a network meta-analysis may yield more reliable and definitive results, and allows visualization and interpretation of a wider picture of the available evidence, and calculation of treatment rankings with probabilities133,134. Only RCTs reporting on elective haemorrhoidectomy for grade III and IV haemorrhoids were considered for this network meta-analysis. Studies of other design were excluded because of the risk of bias in such trials, and because it was not appropriate to perform network meta-analysis on studies with different designs as this would have made the interpretation more difficult. On one hand, this led to some important non-randomized studies being excluded from the analysis but, on the other hand, the bias of the included studies and the heterogeneity between them were inherently less owing to their study design. The overall quality of the included trials based on the Cochrane Collaboration's risk of bias tool was found to be adequate, except with regard to blinding of participants and personnel (performance bias) and blinding of outcome assessment (detection bias). One limitation of a network meta-analysis can be the inconsistency in the results between direct and indirect comparisons; nevertheless, in the present network meta-analysis there was no evidence of inconsistency in the networks for any of the outcomes investigated. However, it is important to note that failure to detect inconsistency does not imply consistency47. Moreover, the amount of evidence a treatment carries, and the number of comparisons available between treatments, determines the diversity and strength of a network meta-analysis135. Severe imbalance in terms of the amount of evidence available may affect the power and reliability of the network meta-analysis, as inferences may be driven largely from the evidence of few treatments and comparisons135. For example, in the present study, some comparisons were informed by several RCTs by either direct or indirect evidence (for example 29 RCTs compared stapled versus open haemorrhoidectomy), whereas other comparisons were only sparsely informed (only 1 RCT compared THD with open haemorrhoidectomy). Two commonly used excisional procedures worldwide are open (Milligan–Morgan) and closed (Ferguson) haemorrhoidectomies. Postoperative pain and postoperative complications are believed to be the most important disadvantages of these techniques, and this was confirmed by the present study. Open and closed haemorrhoidectomies had significantly more postoperative complications than LigaSure™, Harmonic® and THD procedures, and resulted in significantly more postoperative pain than stapled, THD, LigaSure™ and Harmonic® haemorrhoidectomies. The increased complication rates and higher levels of pain related to open and closed haemorrhoidectomies resulted in a longer hospital stay and a later return to normal activities. Furthermore, open and closed haemorrhoidectomies were associated with greater operative blood loss and a longer operating time compared with the other surgical techniques. Nevertheless, low recurrence rate is perceived to be the most important advantage of open and closed haemorrhoidectomies, and this was confirmed here: open and closed haemorrhoidectomies were found to have a lower recurrence rate than THD and stapled haemorrhoidectomies. Pairwise comparison of open and closed haemorrhoidectomies demonstrated significantly more postoperative complications with closed haemorrhoidectomy, suggesting an advantage of open over closed haemorrhoidectomy. Alternative surgical techniques for excision of haemorrhoids have been developed with the aim of reducing postoperative pain and improving perioperative outcomes, including faster recovery and earlier return to normal daily activities. Standard pairwise meta-analyses114,121–122,124–125,130 comparing conventional haemorrhoidectomy with LigaSure™ haemorrhoidectomy showed the latter to have better outcomes with regard to duration of surgery, operative blood loss, postoperative pain, length of hospital stay and time to return to normal activities. In addition, a previous standard pairwise meta-analysis123 that compared conventional with Harmonic® haemorrhoidectomy showed that the Harmonic® procedure resulted in fewer postoperative complications, less postoperative pain and earlier return to work. Similarly, the present network meta-analysis showed that LigaSure™ and Harmonic® haemorrhoidectomies resulted in fewer postoperative complications, a shorter duration of surgery, less operative blood loss and decreased postoperative pain compared with open and closed haemorrhoidectomies. The decreased complication rate and reduced pain were reflected in the shorter time to the first bowel movement and the quicker return to normal activities after LigaSure™ and Harmonic® haemorrhoidectomies. Stapled haemorrhoidectomy (or haemorrhoidopexy) is another technique developed to decrease postoperative pain, resulting in faster recovery; nevertheless, concerns have been raised regarding its high recurrence rate. Previous standard pairwise meta-analyses4,114–115,117–118,120,126,128–129,131 comparing conventional with stapled haemorrhoidectomy showed the stapled procedure to have better outcomes with regard to operating time, postoperative pain, length of hospital stay and time to return to normal activity. However, stapled haemorrhoidectomy was also reported to have higher rates of skin tags, haemorrhoid recurrence and recurrent prolapse than conventional haemorrhoidectomy4,114–115,117–118,120,129,131. Furthermore, previous pairwise meta-analyses113,119,132 comparing stapled versus LigaSure™ haemorrhoidectomy showed the stapled procedure to have a higher recurrence rate, with no difference in postoperative complications, postoperative pain and length of hospital stay. Similarly, the present network meta-analysis showed that stapled haemorrhoidectomy needed a shorter operating time and resulted in less postoperative pain than open and closed haemorrhoidectomies. As a result of the decreased pain levels, stapled haemorrhoidectomy resulted in a shorter hospital stay and quicker return to normal activities. Nevertheless, the recurrence rate was higher after stapled haemorrhoidectomy than after open, closed and LigaSure™ procedures. Furthermore, stapled haemorrhoidectomy was associated with more postoperative complications compared with Harmonic® haemorrhoidectomy and a higher postoperative bleeding rate than THD. Stapled haemorrhoidectomy is expensive and the cost of this technique should be taken into consideration in the decision process, together with its higher rate of complications and recurrences. The increased cost of the stapling instrument may be largely offset by the shorter hospital stay, decreased operating time and earlier return to work. A previous standard pairwise meta-analysis127 comparing THD with stapled haemorrhoidectomy showed no difference between the two treatments with regard to duration of surgery, postoperative complications and recurrence of haemorrhoids. Nevertheless, the present network meta-analysis demonstrated that fewer people had postoperative bleeding after THD compared with open or stapled haemorrhoidectomy, and this resulted in THD being associated with fewer emergency reoperations than open, closed, stapled and LigaSure™ procedures, with a high probability of being the best treatment for reoperation rate (P = 0·710). In addition, THD was found to have fewer postoperative complications, a shorter operating time and decreased levels of postoperative pain than the other surgical techniques. These resulted in THD having a shorter length of hospital stay and an earlier time to the first bowel movement. On the other hand, THD had a higher recurrence rate than open, closed, LigaSure™, laser and radiofrequency haemorrhoidectomies and, importantly, the highest probability of being the worst treatment for recurrence of haemorrhoids (P = 0·785). The low cost of THD, low complication rate, shorter operating time and decreased levels of postoperative pain, but higher recurrence rate, may suggest THD as a safe, quick and easy initial surgical option. On the contrary, open haemorrhoidectomy, which is associated with increased postoperative complications and greater postoperative pain, could be a better approach for refractory haemorrhoids owing to its low recurrence rate. This network meta-analysis has shown that each surgical treatment for haemorrhoids has pros and cons that should be taken into consideration when deciding which technique to use. The technology available in each surgical department and the cost of surgery also play a role in the decision-making process. Every surgeon should be aware of the individual advantages and disadvantages of each surgical treatment, and should discuss these with the patient. The patient's beliefs and priorities should be taken into consideration; for example, whether their priority is faster recovery and earlier return to normal activities, or a lower risk of recurrence. The patient should be provided with all the information available in this review and be allowed to make a fully informed decision. This will ensure that the best treatment and appropriate personalized care is provided to all patients. Further studies are needed to assess patient expectations before surgery and satisfaction after surgery, in both the short and long term. Further higher-quality RCTs are needed to compare surgical treatments for haemorrhoids, particularly with improved blinding of participants and personnel, and blinding of outcome assessors. Future RCTs should assess and compare the economic cost of the different surgical procedures available for haemorrhoids. They should also investigate whether different surgical approaches should be used for single versus circumferential haemorrhoids. Finally, further studies should be designed to assess bowel habit, possibly using a bowel assessment questionnaire, before and after different surgical treatments for haemorrhoids. Disclosure The authors declare no conflict of interest. Supporting information Additional supporting information may be found in the online version of this article: Appendix S1 Systematic review protocol (Word document) Table S1 Detailed search strategy (Word document) Table S2 Odds ratios and mean differences for the pairwise haemorrhoid treatment comparisons for the outcomes of interest (Word document) Table S3 Statistically significant pairwise odds ratios and mean differences (Word document) Table S4 Surgical treatments with the highest probability of ranking from best to worst (1st to 11th) for the outcomes of interest Fig. S1 Risk of bias for each included study (Word document) References 1 Johanson JF . Nonsurgical treatment of hemorrhoids . J Gastrointest Surg 2002 ; 6 : 290 – 294 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Riss S , Weiser FA, Schwameis K, Riss T, Mittlböck M, Steiner G et al. The prevalence of hemorrhoids in adults . Int J Colorectal Dis 2012 ; 27 : 215 – 220 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Burch J , Epstein D, Baba-Akbari A, Weatherly H, Fox D, Golder S et al. Stapled haemorrhoidectomy (haemorrhoidopexy) for the treatment of haemorrhoids: a systematic review and economic evaluation . Health Technol Assess 2008 ; 12 : iii – iv , ix–x, 1–193. Google Scholar Crossref Search ADS PubMed WorldCat 4 Jayaraman S , Colquhoun PH, Malthaner RA. Stapled versus conventional surgery for hemorrhoids . Cochrane Database Syst Rev 2006 ; ( 4 ) CD005393 . Google Scholar OpenURL Placeholder Text WorldCat 5 Banov L , Knoepp LF, Erdman LH, Alia RT. Management of hemorrhoidal disease . J S C Med Assoc 1985 ; 81 : 398 – 401 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 6 Hospital Episode Statistics (HES) . Hospital Episode Statistics. Admitted Patient Care, England – 2012–13: Procedures and Interventions. Main Procedures and Interventions. 3 Character . http://www.hscic.gov.uk/catalogue/PUB12566/hosp-epis-stat-admi-proc-2012-13-tab.xlsx [accessed 14 July 2015]. 7 Altomare DF , Milito G, Andreoli R, Arcana F, Tricomi N, Salafia C et al. ; Ligasure for Hemorrhoids Study Group. Ligasure™ Precise vs. conventional diathermy for Milligan–Morgan hemorrhoidectomy: a prospective, randomized, multicenter trial . Dis Colon Rectum 2008 ; 51 : 514 – 519 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Cheetham MJ , Cohen CR, Kamm MA, Phillips RK. A randomized, controlled trial of diathermy hemorrhoidectomy vs. stapled hemorrhoidectomy in an intended day-care setting with longer-term follow-up . Dis Colon Rectum 2003 ; 46 : 491 – 497 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Ho YH , Seow-Choen F, Tan M, Leong AF. Randomized controlled trial of open and closed haemorrhoidectomy . Br J Surg 1997 ; 84 : 1729 – 1730 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 10 Ramadan E , Vishne T, Dreznik Z. Harmonic scalpel hemorrhoidectomy: preliminary results of a new alternative method . Tech Coloproctol 2002 : 6 : 89 – 92 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Correa-Rovelo JM , Tellez O, Obregón L, Miranda-Gomez A, Moran S. Stapled rectal mucosectomy vs. closed hemorrhoidectomy: a randomized, clinical trial . Dis Colon Rectum 2002 ; 45 : 1367 – 1374 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Fareed M , El-Awady S, Abd El Monaem H, Aly A. Randomized trial comparing LigaSure™ to closed Ferguson hemorrhoidectomy . Tech Coloproctol 2009 ; 13 : 243 – 246 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Filingeri V , Gravante G, Overton J, Toti L, Iqbal A. Ferguson hemorrhoidectomy with radiofrequency versus classic diathermy . J Invest Surg 2010 ; 23 : 170 – 174 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Senagore AJ , Singer M, Abcarian H, Fleshman J, Corman M, Wexner S et al. ; Procedure for Prolapse and Hemorrhoids (PPH) Multicenter Study Group . A prospective, randomized, controlled multicenter trial comparing stapled hemorrhoidopexy and Ferguson hemorrhoidectomy: perioperative and one-year results . Dis Colon Rectum 2004 ; 47 : 1824 – 1836 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Filingeri V , Gravante G, Baldessari E, Grimaldi M, Casciani CU. Prospective randomized trial of submucosal hemorrhoidectomy with radiofrequency bistoury vs. conventional Parks' operation . Tech Coloproctol 2004 ; 8 : 31 – 36 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Hosch SB , Knoefel WT, Pichlmeier U, Schulze V, Busch C, Gawad KA et al. Surgical treatment of piles – prospective, randomized study of packs vs. Milligan–Morgan hemorrhoidectomy . Dis Colon Rectum 1998 ; 41 : 159 – 164 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Chung CC , Cheung HY, Chan ES, Kwok SY, Li MK. Stapled hemorrhoidopexy vs. Harmonic Scalpel hemorrhoidectomy: a randomized trial . Dis Colon Rectum 2005 ; 48 : 1213 – 1219 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Hasse C , Sitter H, Brune M, Wollenteit I, Lorenz W, Rothmund M. Conventional, closed haemorrhoidectomy versus resection with a circular stapler: a prospective randomized study . Dtsch Med Wochenschr 2004 ; 129 : 1611 – 1617 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Helmy MA . Stapling procedure for hemorrhoids versus conventional haemorrhoidectomy . J Egypt Soc Parasitol 2000 ; 30 : 951 – 958 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 20 Kairaluoma M , Nuorva K, Kellokumpu I. Day-case stapled (circular) vs. diathermy hemorrhoidectomy – a randomized, controlled trial evaluating surgical and functional outcome . Dis Colon Rectum 2003 ; 46 : 93 – 99 . Google Scholar Crossref Search ADS PubMed WorldCat 21 De Nardi P , Capretti G, Corsaro A, Staudacher C. A prospective, randomized trial comparing the short- and long-term results of Doppler-guided transanal hemorrhoid dearterialization with mucopexy versus excision hemorrhoidectomy for grade III hemorrhoids . Dis Colon Rectum 2014 ; 57 : 348 – 353 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Denoya PI , Fakhoury M, Chang K, Fakhoury J, Bergamaschi R. Dearterialization with mucopexy versus haemorrhoidectomy for grade III or IV haemorrhoids: short-term results of a double-blind randomized controlled trial . Colorectal Dis 2013 ; 15 : 1281 – 1288 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Festen S , van Hoogstraten MJ, van Geloven AA, Gerhards MF. Treatment of grade III and IV haemorrhoidal disease with PPH or THD. A randomized trial on postoperative complications and short-term results . Int J Colorectal Dis 2009 ; 24 : 1401 – 1405 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Infantino A , Altomare DF, Bottini C, Bonanno M, Mancini S; THD group of the SICCR (Italian Society of Colorectal Surgery) et al. Prospective randomized multicentre study comparing stapler haemorrhoidopexy with Doppler-guided transanal haemorrhoid dearterialization for third-degree haemorrhoids . Colorectal Dis 2012 ; 14 : 205 – 211 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Khanna R , Khanna S, Bhadani S, Singh S, Khanna AK. Comparison of Ligasure hemorrhoidectomy with conventional Ferguson's hemorrhoidectomy . Indian J Surg 2010 ; 72 : 294 – 297 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Kraemer M , Parulava T, Roblick M, Duschka L, Müller-Lobeck H. Prospective, randomized study: proximate PPH stapler vs. LigaSure™ for hemorrhoidal surgery . Dis Colon Rectum 2005 ; 48 : 1517 – 1522 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Lawes DA , Palazzo FF, Francis DL, Clifton MA. One year follow up of a randomized trial comparing Ligasure with open haemorrhoidectomy . Colorectal Dis 2004 ; 6 : 233 – 235 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Milito G , Gargiani M, Cortese F. Randomised trial comparing LigaSure™ haemorrhoidectomy with the diathermy dissection operation . Tech Coloproctol 2002 ; 6 : 171 – 175 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Abo-hashem AA , Sarhan A, Aly AM. Harmonic Scalpel compared with bipolar electro-cautery hemorrhoidectomy: a randomized controlled trial . Int J Surg 2010 ; 8 : 243 – 247 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Kwok SY , Chung CC, Tsui KK, Li MK. A double-blind, randomized trial comparing Ligasure and Harmonic Scalpel hemorrhoidectomy . Dis Colon Rectum 2005 ; 48 : 344 – 348 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Shelygin Iu A , Blagodarnyi LA, Khmylov LM. [ Choice of hemorrhoidectomy method in chronic hemorrhoid .] Khirurgiia 2003 ; ( 8 ): 39 – 45 . Google Scholar OpenURL Placeholder Text WorldCat 32 Tsunoda A , Sada H, Sugimoto T, Kano N, Kawana M, Sasaki T et al. Randomized controlled trial of bipolar diathermy vs ultrasonic scalpel for closed hemorrhoidectomy . World J Gastrointest Surg 2011 ; 3 : 147 – 152 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Enriquez-Navascues JM , Devesa Múgica JM, Bucheli Proaño P. [ Hemorrhoidectomy: conventional or by Nd : YAG contact laser? A prospective and randomized study .] Rev Esp Enferm Dig 1993 ; 84 : 235 – 239 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 34 Pandini LC , Nahas SC, Nahas CS, Marques CF, Sobrado CW, Kiss DR. Surgical treatment of haemorrhoidal disease with CO2 laser and Milligan–Morgan cold scalpel technique . Colorectal Dis 2006 ; 8 : 592 – 595 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Senagore A , Mazier WP, Luchtefeld MA, MacKeigan JM, Wengert T. Treatment of advanced hemorrhoidal disease: a prospective, randomized comparison of cold scalpel vs. contact Nd : YAG laser . Dis Colon Rectum 1993 ; 36 : 1042 – 1049 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Wang D , Zhong KL, Chen JL, Wang XX, Pan K, Xia LG et al. [ Effect of diode laser coagulation treatment on grade III internal hemorrhoids .] Zhonghua Wei Chang Wai Ke Za Zhi 2005 ; 8 : 325 – 327 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 37 Wang JY , Tsai HL, Chen FM, Chu KS, Chan HM, Huang CJ et al. Prospective, randomized, controlled trial of Starion™ vs. Ligasure™ hemorrhoidectomy for prolapsed hemorrhoids . Dis Colon Rectum 2007 ; 50 : 1146 – 1151 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Filingeri V , Gravante G, Baldessari E, Craboledda P, Bellati F, Casciani CU. A randomised trial comparing submucosal haemorrhoidectomy with radiofrequency bistoury vs. diathermic haemorrhoidectomy . Eur Rev Med Pharmacol Sci 2004 ; 8 : 79 – 85 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 39 Franceschilli L , Stolfi VM, D'Ugo S, Angelucci GP, Lazzaro S, Picone E et al. Radiofrequency versus conventional diathermy Milligan–Morgan hemorrhoidectomy: a prospective, randomized study . Int J Colorectal Dis 2011 ; 26 : 1345 – 1350 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Chung CC , Ha JP, Tai YP, Tsang WW, Li MK. Double-blind, randomized trial comparing Harmonic Scalpel hemorrhoidectomy, bipolar scissors hemorrhoidectomy, and scissors excision: ligation technique . Dis Colon Rectum 2002 ; 45 : 789 – 794 . Google Scholar Crossref Search ADS PubMed WorldCat 41 Dindo D , Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey . Ann Surg 2004 ; 240 : 205 – 213 . Google Scholar Crossref Search ADS PubMed WorldCat 42 Cochrane Bias Methods Group . Assessing Risk of Bias in Included Studies ; 2013 . http://bmg.cochrane.org/assessing-risk-bias-included-studies [accessed 14 July 2015]. 43 Liberati A , Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration . BMJ 2009 ; 339 : b2700 . Google Scholar Crossref Search ADS PubMed WorldCat 44 Dias S , Welton NJ, Sutton AJ, Ades AE. NICE DSU Technical Support Document 2: a Generalised Linear Modelling Framework for Pairwise and Network Meta-analysis of Randomised Controlled Trials ; 2013. http://www.nicedsu.org.uk [accessed 14 July 2015]. 45 Higgins JPT , Green S. Cochrane Handbook for Systematic Reviews of Interventions . Version 5.1.0 [updated March 2011]. www.cochrane-handbook.org [accessed 14 July 2015]. 46 Lu G , Ades AE. Combination of direct and indirect evidence in mixed treatment comparisons . Stat Med 2004 ; 23 : 3105 – 3124 . Google Scholar Crossref Search ADS PubMed WorldCat 47 Dias S , Welton NJ, Sutton AJ, Caldwell DM, Lu G, Ades AE et al. NICE DSU Technical Support Document 4: Inconsistency in Networks of Evidence Based on Randomised Controlled Trials ; 2012. http://www.nicedsu.org.uk [accessed 14 July 2015]. 48 Dias S , Welton NJ, Sutton AJ, Ades AE. NICE DSU Technical Support Document 1: Introduction to Evidence Synthesis for Decision Making ; 2012. http://www.nicedsu.org.uk [accessed 14 July 2015]. 49 Ammaturo C , Tufano A, Spiniello E, Sodano B, Iervolino EM, Brillantino A et al. Stapled haemorrhoidopexy vs. Milligan–Morgan haemorrhoidectomy for grade III haemorrhoids: a randomized clinical trial . G Chir 2012 ; 33 : 346 – 351 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 50 Arslani N , Patrlj L, Rajković Z, Papeš D, Altarac S. A randomized clinical trial comparing Ligasure versus stapled hemorrhoidectomy . Surg Laparosc Endosc Percutan Tech 2012 ; 22 : 58 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat 51 Azolas R , Villalon MR, Danilla S, Hasbun A, Gatica F, Salamanca J. Prospective randomized comparison of open and closed hemorrhoidectomy . Revista Chilena De Cirugia 2010 ; 62 : 382 – 386 . Google Scholar OpenURL Placeholder Text WorldCat 52 Basdanis G , Papadopoulos VN, Michalopoulos A, Apostolidis S, Harlaftis N. Randomized clinical trial of stapled hemorrhoidectomy vs open with Ligasure for prolapsed piles . Surg Endosc 2005 ; 19 : 235 – 239 . Google Scholar Crossref Search ADS PubMed WorldCat 53 Bassi R , Bergami G. [ The surgical treatment of hemorrhoids: diathermocoagulation and traditional technics. A prospective randomized study .] Minerva Chir 1997 ; 52 : 387 – 391 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 54 Bessa SS . Ligasure™ vs. conventional diathermy in excisional hemorrhoidectomy: a prospective, randomized study . Dis Colon Rectum 2008 ; 51 : 940 – 944 . Google Scholar Crossref Search ADS PubMed WorldCat 55 Bikhchandani J , Agarwal PN, Kant R, Malik VK. Randomized controlled trial to compare the early and mid-term results of stapled versus open hemorrhoidectomy . Am J Surg 2005 ; 189 : 56 – 60 . Google Scholar Crossref Search ADS PubMed WorldCat 56 Boccasanta P , Capretti PG, Venturi M, Cioffi U, Simone M, Salamina G et al. Randomised controlled trial between stapled circumferential mucosectomy and conventional circular hemorrhoidectomy in advanced hemorrhoids with external mucosal prolapse . Am J Surg 2001 ; 182 : 64 – 68 . Google Scholar Crossref Search ADS PubMed WorldCat 57 Bouini NR , Pour MA, Hadian HS, Pour RA. Randomised clinical trial comparing ligasure hemorrhoidectomy with conventional hemorrhoidectomy . Journal of Mazandaran University of Medical Sciences 2012 ; 22 : 66 – 73 . Google Scholar OpenURL Placeholder Text WorldCat 58 Brown SR , Ballan K, Ho E, Ho Fams YH, Seow-Choen F. Stapled mucosectomy for acute thrombosed circumferentially prolapsed piles: a prospective randomized comparison with conventional haemorrhoidectomy . Colorectal Dis 2001 ; 3 : 175 – 178 . Google Scholar Crossref Search ADS PubMed WorldCat 59 Bulut A , Evcimen S, Kaya IO, Hoca O. [ Stapled haemorrhoidopexy versus Milligan–Morgan haemorrhoidectomy in treatment of haemorrhoidal disease .] Turkish Journal of Surgery 2006 ; 22 : 67 – 71 . Google Scholar OpenURL Placeholder Text WorldCat 60 Carrabetta S , Nikzat K, Guardini R, Segre D. [ Hemorrhoidectomy. Analysis of comparison between Milligan–Morgan versus Ferguson technique .] Chirurgia 2001 ; 14 : 17 – 19 . Google Scholar OpenURL Placeholder Text WorldCat 61 Castellvi J , Sueiras A, Espinosa J, Vallet J, Gil V, Pi F. Ligasure™ versus diathermy hemorrhoidectomy under spinal anesthesia or pudendal block with ropivacaine: a randomized prospective clinical study with 1-year follow-up . Int J Colorectal Dis 2009 ; 24 : 1011 – 1018 . Google Scholar Crossref Search ADS PubMed WorldCat 62 Chen S , Lai DM, Yang B, Zhang L, Zhou TC, Chen GX. [ Therapeutic comparison between procedure for prolapse and hemorrhoids and Ligasure technique for hemorrhoids .] Zhonghua Wei Chang Wai Ke Za Zhi 2007 ; 10 : 342 – 345 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 63 Chung YC , Wu HJ. Clinical experience of sutureless closed hemorrhoidectomy with LigaSure™ . Dis Colon Rectum 2003 ; 46 : 87 – 92 . Google Scholar Crossref Search ADS PubMed WorldCat 64 Dell'Abate P , Ferrieri G, Del Rio P, Soliani P, Sianesi M. [ Longo hemorrhoidopexy vs Milligan–Morgan hemorrhoidectomy: perspective analysis .] G Chir 2005 ; 26 : 443 – 445 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 65 Fazeli MS , Safari S, Kazemeini A, Larti F, Joneidi E, Rahimi M et al. A prospective study comparing Ligasure and open hemorrhoidectomy . Tehran University Medical Journal 2011 ; 69 : 495 – 501 . Google Scholar OpenURL Placeholder Text WorldCat 66 Franklin EJ , Seetharam S, Lowney J, Horgan PG. Randomized, clinical trial of Ligasure™ vs. conventional diathermy in hemorrhoidectomy . Dis Colon Rectum 2003 ; 46 : 1380 – 1383 . Google Scholar Crossref Search ADS PubMed WorldCat 67 Ganio E , Altomare DF, Gabrielli F, Milito G, Canuti S. Prospective randomized multicentre trial comparing stapled with open haemorrhoidectomy . Br J Surg 2001 ; 88 : 669 – 674 . Google Scholar Crossref Search ADS PubMed WorldCat 68 Ganio E , Altomare DF, Milito G, Gabrielli F, Canuti S. Long-term outcome of a multicentre randomized clinical trial of stapled haemorrhoidopexy versus Milligan–Morgan haemorrhoidectomy . Br J Surg 2007 ; 94 : 1033 – 1037 . Google Scholar Crossref Search ADS PubMed WorldCat 69 Gao RZ , Liang XB, Xu CN, Zhang JY, Wang P, Niu HG. [ Comparison of two-year efficacy between procedure for prolapse and hemorrhoids and Milligan–Morgan hemorrhoidectomy in treatment of III and IV degree internal hemorrhoids .] Zhonghua Wei Chang Wai Ke Za Zhi 2008 ; 11 : 249 – 252 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 70 Gençosmanoğlu R , Sad O, Koç D, Inceoğlu R. Hemorrhoidectomy: open or closed technique? A prospective, randomized clinical trial . Dis Colon Rectum 2002 ; 45 : 70 – 75 . Google Scholar Crossref Search ADS PubMed WorldCat 71 Gentile M , De Rosa M, Pilone V, Mosella F, Forestieri P. Surgical treatment for IV-degree hemorrhoids: LigaSure™ hemorroidectomy vs. conventional diathermy. A prospective, randomized trial . Minerva Chir 2011 ; 66 : 207 – 213 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 72 Ho YH , Cheong WK, Tsang C, Ho J, Eu KW, Tang CL et al. Stapled hemorrhoidectomy – cost and effectiveness. Randomized, controlled trial including incontinence scoring, anorectal manometry, and endoanal ultrasound assessments at up to three months . Dis Colon Rectum 2000 ; 43 : 1666 – 1675 . Google Scholar Crossref Search ADS PubMed WorldCat 73 Ho KS , Ho YH. Prospective randomized trial comparing stapled hemorrhoidopexy versus closed Ferguson hemorrhoidectomy . Tech Coloproctol 2006 ; 10 : 193 – 197 . Google Scholar Crossref Search ADS PubMed WorldCat 74 Huang WS , Chin CC, Yeh CH, Lin PY, Wang JY. Randomized comparison between stapled hemorrhoidopexy and Ferguson hemorrhoidectomy for grade III hemorrhoids in Taiwan: a prospective study . Int J Colorectal Dis 2007 ; 22 : 955 – 961 . Google Scholar Crossref Search ADS PubMed WorldCat 75 Jayne DG , Botterill I, Ambrose NS, Brennan TG, Guillou PJ, O'Riordain DS. Randomized clinical trial of Ligasure™ versus conventional diathermy for day-case haemorrhoidectomy . Br J Surg 2002 ; 89 : 428 – 432 . Google Scholar Crossref Search ADS PubMed WorldCat 76 Khafagy W , Nakeeb A, Fouda E, Omar W, Elhak NG, Farid M et al. Conventional haemorrhoidectomy, stapled haemorrhoidectomy, Doppler guided haemorrhoidectomy artery ligation; post operative pain and anorectal manometric assessment . Hepatogastroenterology 2009 ; 56 : 1010 – 1015 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 77 Khalil KH , O'Bichere A, Sellu D. Randomized clinical trial of sutured versus stapled closed haemorrhoidectomy . Br J Surg 2000 ; 87 : 1352 – 1355 . Google Scholar Crossref Search ADS PubMed WorldCat 78 Kim JH , Lee YP. [ Randomized trial comparing a starion and a harmonic scalpel hemorrhoidectomy .] Journal of the Korean Society of Coloproctology 2009 ; 25 : 8 – 12 . Google Scholar Crossref Search ADS WorldCat 79 Kim JS , Vashist YK, Thieltges S, Zehler O, Gawad KA et al. Stapled hemorrhoidopexy versus Milligan–Morgan hemorrhoidectomy in circumferential third-degree hemorrhoids: long-term results of a randomized controlled trial . J Gastrointest Surg 2013 ; 17 : 1292 – 1298 . Google Scholar Crossref Search ADS PubMed WorldCat 80 Krska Z , Kvasnièka J, Faltýn J, Schmidt D, Sváb J, Kormanová K et al. Surgical treatment of haemorrhoids according to Longo and Milligan Morgan: an evaluation of postoperative tissue response . Colorectal Dis 2003 ; 5 : 573 – 576 . Google Scholar Crossref Search ADS PubMed WorldCat 81 Leventoğlu S , Menteş BB, Akin M, Oğuz M. Haemorrhoidectomy with electrocautery or ultrashears and stapled haemorrhoidopexy . ANZ J Surg 2008 ; 78 : 389 – 393 . Google Scholar Crossref Search ADS PubMed WorldCat 82 Liao XJ , Meng Q, Yang GG, Shen Z, Yang QY, Wu WJ. [ Efficacy of the procedure for prolapse and hemorrhoids combined with external hemorrhoids excision in the treatment of III or IV mixed hemorrhoids .] Zhonghua Wei Chang Wai Ke Za Zhi 2008 ; 11 : 525 – 528 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 83 Mehigan BJ , Monson JR, Hartley JE. Stapling procedure for haemorrhoids versus Milligan–Morgan haemorrhoidectomy: randomised controlled trial . Lancet 2000 ; 355 : 782 – 785 . Google Scholar Crossref Search ADS PubMed WorldCat 84 Mik M , Rzetecki T, Sygut A, Trzcinski R, Dziki A. Open and closed haemorrhoidectomy for fourth degree haemorrhoids – comparative one center study . Acta Chir Iugosl 2008 ; 55 : 119 – 125 . Google Scholar Crossref Search ADS PubMed WorldCat 85 Morpurgo E , Termini B, Tosato SM, Orsini C, Masiero V, Brotto M et al. Anorectal manometric changes after standard and stapled hemorrhoidectomy . Journal of Pelvic Medicine and Surgery 2008 ; 14 : 51 – 55 . Google Scholar Crossref Search ADS WorldCat 86 Muzi MG , Milito G, Nigro C, Cadeddu F, Andreoli F, Amabile D et al. Randomized clinical trial of LigaSure™ and conventional diathermy haemorrhoidectomy . Br J Surg 2007 ; 94 : 937 – 942 . Google Scholar Crossref Search ADS PubMed WorldCat 87 Nyström PO , Qvist N, Raahave D, Lindsey I, Mortensen N; Stapled or Open Pile Procedure (STOPP) trial study group . Randomized clinical trial of symptom control after stapled anopexy or diathermy excision for haemorrhoid prolapse . Br J Surg 2010 ; 97 : 167 – 176 . Google Scholar Crossref Search ADS PubMed WorldCat 88 Ortiz H , Marzo J, Armendariz P. Randomized clinical trial of stapled haemorrhoidopexy versus conventional diathermy haemorrhoidectomy . Br J Surg 2002 ; 89 : 1376 – 1381 . Google Scholar Crossref Search ADS PubMed WorldCat 89 Ortiz H , Marzo J, Armendáriz P, Miguel M. Stapled hemorrhoidopexy vs. diathermy excision for fourth-degree hemorrhoids: a randomized, clinical trial and review of the literature . Dis Colon Rectum 2005 ; 48 : 809 – 815 . Google Scholar Crossref Search ADS PubMed WorldCat 90 Ozer MT , Yigit T, Uzar AI, Mentes O, Harlak A, Kilic S et al. A comparison of different hemorrhoidectomy procedures . Saudi Med J 2008 ; 29 : 1264 – 1269 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 91 Palazzo FF , Francis DL, Clifton MA. Randomized clinical trial of Ligasure versus open haemorrhoidectomy . Br J Surg 2002 ; 89 : 154 – 157 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 92 Palimento D , Picchio M, Attanasio U, Lombardi A, Bambini C, Renda A. Stapled and open hemorrhoidectomy: randomized controlled trial of early results . World J Surg 2003 ; 27 : 203 – 207 . Google Scholar Crossref Search ADS PubMed WorldCat 93 Pattana-Arun J , Sooriprasoet N, Sahakijrungruang C, Tantiphlachiva K, Rojanasakul A. Closed vs Ligasure hemorrhoidectomy: a prospective, randomized clinical trial . J Med Assoc Thai 2006 ; 89 : 453 – 458 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 94 Pescatori M , Favetta U, Amato A. Anorectal function and clinical outcome after open and closed haemorrhoidectomy, with and without internal sphincterotomy. A prospective study . Tech Coloproctol 2000 ; 4 : 17 – 23 . Google Scholar Crossref Search ADS WorldCat 95 Peters CJ , Botterill I, Ambrose NS, Hick D, Casey J, Jayne JD. Ligasure™ vs conventional diathermy haemorrhoidectomy: long-term follow-up of a randomised clinical trial . Colorectal Dis 2005 ; 7 : 350 – 353 . Google Scholar Crossref Search ADS PubMed WorldCat 96 Picchio M , Palimento D, Attanasio U, Renda A. Stapled vs open hemorrhoidectomy: long-term outcome of a randomized controlled trial . Int J Colorectal Dis 2006 ; 21 : 668 – 669 . Google Scholar Crossref Search ADS PubMed WorldCat 97 Pokharel N , Chhetri RK, Malla B, Joshi HN, Shrestha RK. Haemorrhoidectomy: Ferguson's (closed) vs Milligan Morgan's technique (open) . Nepal Med Coll J 2009 ; 11 : 136 – 137 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 98 Racalbuto A , Aliotta I, Corsaro G, Lanteri R, Cataldo A, Licata A. Hemorrhoidal stapler prolapsectomy vs. Milligan–Morgan hemorrhoidectomy: a long-term randomized trial . Int J Colorectal Dis 2004 ; 19 : 239 – 244 . Google Scholar Crossref Search ADS PubMed WorldCat 99 Rahmani N , Sayadi S, Mohammadpur Tahamtan RA, Ali AM, Tayebi P. Comparison with result open vs.closed hemorrhoidectomy . Journal of Mazandaran University of Medical Sciences 2012 ; 21 : 54 – 61 . Google Scholar OpenURL Placeholder Text WorldCat 100 Rowsell M , Bello M, Hemingway DM. Circumferential mucosectomy (stapled haemorrhoidectomy) versus conventional haemorrhoidectomy: randomised controlled trial . Lancet 2000 ; 355 : 779 – 781 . Google Scholar Crossref Search ADS PubMed WorldCat 101 Sabanci U , Ogun I, Candemir G. Stapled haemorrhoidopexy versus Ferguson haemorrhoidectomy: a prospective study with 2-year postoperative follow-up . J Int Med Res 2007 ; 35 : 917 – 921 . Google Scholar Crossref Search ADS PubMed WorldCat 102 Sakr MF . LigaSure™ versus Milligan–Morgan hemorrhoidectomy: a prospective randomized clinical trial . Tech Coloproctol 2010 ; 14 : 13 – 17 . Google Scholar Crossref Search ADS PubMed WorldCat 103 Sakr MF , Moussa MM. LigaSure™ hemorrhoidectomy versus stapled hemorrhoidopexy: a prospective, randomized clinical trial . Dis Colon Rectum 2010 ; 53 : 1161 – 1167 . Google Scholar Crossref Search ADS PubMed WorldCat 104 Smyth EF , Baker RP, Wilken BJ, Hartley JE, White TJ, Monson JR. Stapled versus excision haemorrhoidectomy: long-term follow up of a randomised controlled trial . Lancet 2003 ; 361 : 1437 – 1438 . Google Scholar Crossref Search ADS PubMed WorldCat 105 Stolfi VM , Sileri P, Micossi C, Carbonaro I, Venza M, Gentileschi P et al. Treatment of hemorrhoids in day surgery: stapled hemorrhoidopexy vs Milligan–Morgan hemorrhoidectomy . J Gastrointest Surg 2008 ; 12 : 795 – 801 . Google Scholar Crossref Search ADS PubMed WorldCat 106 Tan KY , Zin T, Sim HL, Poon PL, Cheng A, Mak K. Randomized clinical trial comparing LigaSure™ haemorrhoidectomy with open diathermy haemorrhoidectomy . Tech Coloproctol 2008 ; 12 : 93 – 97 . Google Scholar Crossref Search ADS PubMed WorldCat 107 Thorbeck CV , Montes MF. Haemorrhoidectomy: randomised controlled clinical trial of Ligasure® compared with Milligan–Morgan operation . Eur J Surg 2002 ; 168 : 482 – 484 . Google Scholar Crossref Search ADS PubMed WorldCat 108 Verre L , Rossi R, Gaggelli I, Bella C, Tirone A, Piccolomini A. PPH versus THD: a comparison of two techniques for III and IV degree haemorrhoids Personal experience . Minerva Chir 2013 ; 68 : 543 – 550 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 109 Wang JY , Lu CY, Tsai HL, Chen FM, Huang CJ, Huang YS et al. Randomized controlled trial of LigaSure™ with submucosal dissection versus Ferguson hemorrhoidectomy for prolapsed hemorrhoids . World J Surg 2006 ; 30 : 462 – 466 . Google Scholar Crossref Search ADS PubMed WorldCat 110 Wilson MS , Pope V, Doran HE, Fearn SJ, Brough WA. Objective comparison of stapled anopexy and open hemorrhoidectomy – a randomized, controlled trial . Dis Colon Rectum 2002 ; 45 : 1437 – 1444 . Google Scholar Crossref Search ADS PubMed WorldCat 111 You SY , Kim SH, Chung CS, Lee DK. Open vs. closed hemorrhoidectomy . Dis Colon Rectum 2005 ; 48 : 108 – 113 . Google Scholar Crossref Search ADS PubMed WorldCat 112 Zampieri N , Castellani R, Andreoli R, Geccherle A. Long-term results and quality of life in patients treated with hemorrhoidectomy using two different techniques: Ligasure versus transanal hemorrhoidal dearterialization . Am J Surg 2012 ; 204 : 684 – 688 . Google Scholar Crossref Search ADS PubMed WorldCat 113 Chen HL , Woo XB, Cui J, Chen CQ, Peng JS. Ligasure versus stapled hemorrhoidectomy in the treatment of hemorrhoids: a meta-analysis of randomized control trials . Surg Laparosc Endosc Percutan Tech 2014 ; 24 : 285 – 289 . Google Scholar Crossref Search ADS PubMed WorldCat 114 Chen JS , You JF. Current status of surgical treatment for hemorrhoids – systematic review and meta-analysis . Chang Gung Med J 2010 ; 33 : 488 – 500 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 115 Giordano P , Gravante G, Sorge R, Ovens L, Nastro P. Long-term outcomes of stapled hemorrhoidopexy vs conventional hemorrhoidectomy: a meta-analysis of randomized controlled trials . Arch Surg 2009 ; 144 : 266 – 272 . Google Scholar Crossref Search ADS PubMed WorldCat 116 Ho YH , Buettner PG. Open compared with closed haemorrhoidectomy: meta-analysis of randomized controlled trials . Tech Coloproctol 2007 ; 11 : 135 – 143 . Google Scholar Crossref Search ADS PubMed WorldCat 117 Lan P , Wu X, Zhou X, Wang J, Zhang L. The safety and efficacy of stapled hemorrhoidectomy in the treatment of hemorrhoids: a systematic review and meta-analysis of ten randomized control trials . Int J Colorectal Dis 2006 ; 21 : 172 – 178 . Google Scholar Crossref Search ADS PubMed WorldCat 118 Laughlan K , Jayne DG, Jackson D, Rupprecht F, Ribaric G. Stapled haemorrhoidopexy compared to Milligan–Morgan and Ferguson haemorrhoidectomy: a systematic review . Int J Colorectal Dis 2009 ; 24 : 335 – 344 . Google Scholar Crossref Search ADS PubMed WorldCat 119 Lee KC , Chen HH, Chung KC, Hu WH, Chang CL, Lin SE et al. Meta-analysis of randomized controlled trials comparing outcomes for stapled hemorrhoidopexy versus LigaSure™ hemorrhoidectomy for symptomatic hemorrhoids in adults . Int J Surg 2013 ; 11 : 914 – 918 . Google Scholar Crossref Search ADS PubMed WorldCat 120 Madiba TE , Esterhuizen TM, Thomson SR. Procedure for prolapsed haemorrhoids versus excisional haemorrhoidectomy – a systematic review and meta-analysis . S Afr Med J 2009 ; 99 : 43 – 53 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 121 Mastakov MY , Buettner PG, Ho YH. Updated meta-analysis of randomized controlled trials comparing conventional excisional haemorrhoidectomy with LigaSure™ for haemorrhoids . Tech Coloproctol 2008 ; 12 : 229 – 239 . Google Scholar Crossref Search ADS PubMed WorldCat 122 Milito G , Cadeddu F, Muzi MG, Nigro C, Farinon AM. Haemorrhoidectomy with Ligasure vs conventional excisional techniques: meta-analysis of randomized controlled trials . Colorectal Dis 2010 ; 12 : 85 – 93 . Google Scholar Crossref Search ADS PubMed WorldCat 123 Mushaya CD , Caleo PJ, Bartlett L, Buettner PG, Ho YH. Harmonic scalpel compared with conventional excisional haemorrhoidectomy: a meta-analysis of randomized controlled trials . Tech Coloproctol 2014 ; 18 : 1009 – 1116 . Google Scholar Crossref Search ADS PubMed WorldCat 124 Nienhuijs S , de Hingh I. Conventional versus LigaSure™ hemorrhoidectomy for patients with symptomatic hemorrhoids . Cochrane Database Syst Rev 2009 ; ( 1 ) CD006761 . Google Scholar OpenURL Placeholder Text WorldCat 125 Nienhuijs SW , de Hingh IH. Pain after conventional versus Ligasure haemorrhoidectomy. A meta-analysis . Int J Surg 2010 ; 8 : 269 – 273 . Google Scholar Crossref Search ADS PubMed WorldCat 126 Nisar PJ , Acheson AG, Neal KR, Scholefield JH. Stapled hemorrhoidopexy compared with conventional hemorrhoidectomy: systematic review of randomized, controlled trials . Dis Colon Rectum 2004 ; 47 : 1837 – 1845 . Google Scholar Crossref Search ADS PubMed WorldCat 127 Sajid MS , Parampalli U, Whitehouse P, Sains P, McFall MR, Baig MK. A systematic review comparing transanal haemorrhoidal de-arterialisation to stapled haemorrhoidopexy in the management of haemorrhoidal disease . Tech Coloproctol 2012 ; 16 : 1 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 128 Sgourakis G , Sotiropoulos GC, Dedemadi G, Radtke A, Papanikolaou I, Christofides T et al. Stapled versus Ferguson hemorrhoidectomy: is there any evidence-based information? Int J Colorectal Dis 2008 ; 23 : 825 – 832 . Google Scholar Crossref Search ADS PubMed WorldCat 129 Shao WJ , Li GC, Zhang ZH, Yang BL, Sun GD, Chen YQ. Systematic review and meta-analysis of randomized controlled trials comparing stapled haemorrhoidopexy with conventional haemorrhoidectomy . Br J Surg 2008 ; 95 : 147 – 160 . Google Scholar Crossref Search ADS PubMed WorldCat 130 Tan EK , Cornish J, Darzi AW, Papagrigoriadis S, Tekkis PP. Meta-analysis of short-term outcomes of randomized controlled trials of LigaSure™ vs conventional hemorrhoidectomy . Arch Surg 2007 ; 142 : 1209 – 1218 . Google Scholar Crossref Search ADS PubMed WorldCat 131 Tjandra JJ , Chan MK. Systematic review on the procedure for prolapse and hemorrhoids (stapled hemorrhoidopexy) . Dis Colon Rectum 2007 ; 50 : 878 – 892 . Google Scholar Crossref Search ADS PubMed WorldCat 132 Yang J , Cui PJ, Han HZ, Tong DN. Meta-analysis of stapled hemorrhoidopexy vs LigaSure™ hemorrhoidectomy . World J Gastroenterol 2013 ; 19 : 4799 – 5807 . Google Scholar Crossref Search ADS PubMed WorldCat 133 Jansen JP , Naci H. Is network meta-analysis as valid as standard pairwise meta-analysis? It all depends on the distribution of effect modifiers . BMC Med 2013 ; 11 : 159 . Google Scholar Crossref Search ADS PubMed WorldCat 134 Mills EJ , Ioannidis JP, Thorlund K, Schünemann HJ, Puhan MA, Guyatt GH. How to use an article reporting a multiple treatment comparison meta-analysis . JAMA 2012 ; 308 : 1246 – 1253 . Google Scholar Crossref Search ADS PubMed WorldCat 135 Mills EJ , Thorlund K, Ioannidis JP. Demystifying trial networks and network meta-analysis . BMJ 2013 ; 346 : f2914 . Google Scholar Crossref Search ADS PubMed WorldCat © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd

Hughes, M J; Harrison, E M; Peel, N J; Stutchfield, B; McNally, S; Beattie, C; Wigmore, S J

doi: 10.1002/bjs.9949pmid: 26447461

Abstract Background Analgesia after liver surgery remains controversial. A previous randomized trial of continuous wound infiltration (CWI) versus thoracic epidural analgesia (TEA) after liver surgery (LIVER trial) showed a faster recovery time in the wound infiltration group but better early postoperative pain scores in the TEA group. High-level evidence is, however, limited and opinion remains divided. The aim was to determine whether there is a difference in functional recovery time between patients having CWI plus abdominal nerve blocks versus TEA after liver resection. Methods A randomized unblinded clinical trial of patients undergoing open liver resection was commenced in December 2012, with follow-up to August 2014. Patients were randomized to receive either wound catheter and nerve block (CWI group) or TEA for 48 h after surgery. The primary outcome measure was functional recovery time. Secondary outcomes were pain scores, complication rates, inflammatory response and central venous pressure (CVP) during transection. Results Of 50 patients randomized initially to each group, 44 received TEA and 49 CWI. Median (i.q.r.) recovery time was 6·5 (5–9·75) and 5·75 (4–7) days in the TEA and CWI groups respectively (P = 0·036). Pain scores were not significantly different between the two groups, and there were no differences in morbidity, inflammatory response or CVP during transection. Conclusion Wound infiltration is associated with a reduced time to recovery after open liver resection compared with epidural analgesia. TEA does not offer an advantage over CWI in terms of attenuation of the inflammatory response or pain scores. Registration number: NCT01747122 (http://www.clinicaltrials.gov). Introduction Following the introduction of enhanced recovery protocols (ERPs) after liver resectional surgery, thoracic epidural analgesia (TEA) has been incorporated as part of multimodal analgesic regimens1–4. However, there is controversy regarding the role of TEA after liver surgery5,6. Perceived disadvantages of TEA include intravenous fluid overload7, coagulopathy following liver resection, epidural haematoma8, a failure rate of up to 25 per cent9 and an increase in blood transfusion requirements10. Alternative techniques are becoming more popular, with continuous wound infiltration (CWI) being used in an increasing number of centres11,12. However, despite evidence suggesting improved recovery times for non-epidural groups, opinion remains divided. Potential reasons for this are the superior analgesic properties of TEA compared with CWI12. TEA is also regarded as being able to attenuate the proinflammatory response to surgery13,14 This appeals to enhanced recovery after surgery principles as proinflammatory cytokine modulation corresponds to an improvement in clinical outcomes after liver resection15. TEA use is also advocated to reduce central venous pressure (CVP) during liver resection in order to diminish blood loss during parenchymal transection16. A previous randomized clinical trial (RCT)12 comparing CWI and TEA conducted by the present authors showed superior early and overall pain scores in the TEA group compared with the CWI group. In the present trial, in an attempt to prevent a delay in establishing effective analgesia in the CWI group, a greater volume of local anaesthetic was delivered as a transversus abdominis and rectus sheath nerve block at the same time as the wound infiltration catheters were being placed. In addition, the intraoperative physiological response was investigated to enable a comprehensive assessment of the influence of analgesic modality on response to surgery, as well as an evaluation of clinical outcomes. Furthermore, postoperative cytokine response and levels of the inflammatory mediator high-mobility group box 1 (HMGB1) were quantified. HMGB1 has been associated with duration of the inflammatory response17, and attenuation of this mediator has been recommended as an area of exploitation in an attempt to improve postoperative outcomes18. The aim of this study was to compare functional recovery time between patients having CWI plus abdominal nerve blocks and those receiving TEA after liver resection in a randomized superiority trial. Methods The LIVER 2 (Effects of Local Infiltration Versus Epidural Following Liver Resection 2) trial was approved by the South East Scotland Research Ethics Committee and the protocol was published before the start (clinicaltrials.gov; NCT01747122). The trial was carried out in accordance with Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guidance19 and reported according to the Consolidated Standards of Reporting Trials (CONSORT) statement20. Recruitment took place from December 2012 and follow-up was completed by August 2014. Patients All patients listed to undergo open liver resection at the Royal Infirmary of Edinburgh, UK, were eligible for inclusion in the trial and approached to discuss participation. After an initial discussion, patients were given written trial information at the preadmission clinic approximately 4 weeks before admission for resection. Written informed consent was obtained on the day of surgery, or sooner if appropriate. Patients were excluded before randomization if they were to undergo a laparoscopic or combined procedure, had a contraindication to epidural or wound catheter, deranged liver function, were aged less than 18 years or did not have capacity to provide informed consent to enter the trial. Perioperative care Patients were admitted on the morning of surgery having fasted for 6 h and taken clear fluids for up to 2 h before surgery. After consenting to the trial, they were randomized to receive either TEA or CWI combined with abdominal nerve block. Sequential, opaque sealed envelopes were filled in advance by an independent adjudicator and opened after consent had been obtained. Randomization was stratified according to extent of resection: either major (3 or more segments) or minor (fewer than 3). Neither participant nor the trial team was blinded to the allocation. No formal blinding of allocation was made during data collection or analysis. Patients randomized to TEA underwent a standard anaesthetic protocol by the same group of anaesthetists experienced in liver resection surgery. In the anaesthetic room before induction of anaesthesia, participants received a thoracic epidural puncture at the level of T8–T9. Some 4 ml 2 per cent lidocaine was administered as a test dose followed by 10 ml levobupivacaine with 100 µg fentanyl to establish epidural block. An infusion of 0·1 per cent levobupivacaine with 2 µg/ml fentanyl was then started. All patients underwent induction of anaesthesia with propofol, atracurium and fentanyl (1–2 µg/kg). Maintenance of anaesthesia was achieved by desflurane or sevoflurane. Intraoperative morphine was administered at the discretion of the anaesthetist. Fentanyl was administered if the patient was known to have had an adverse reaction to morphine. At induction patients routinely received a central venous catheter in the right internal jugular vein, radial arterial cannulation and a urethral catheter. CVP, mean arterial pressure (MAP), end-tidal carbon dioxide, heart rate, oxygen saturation and urine output were monitored throughout the operation. The peritoneal cavity was entered via a right subcostal incision that was extended if required, superiorly in the midline. Liver parenchyma transection was performed with a Cavitron Ultrasonic Aspirator (CUSA®; ValleyLab, Boulder, Colorado, USA) and monopolar diathermy. Patients were managed according to an ERP (Table 1) used in the investigating unit12. Perioperative enhanced recovery principles were adhered to, with routine thromboprophylaxis and perioperative antibiotics administered immediately before incision. Intraoperative warming was performed routinely. As small an incision as required was used. Intraperitoneal drains and nasogastric tubes were not used routinely, and intraoperative fluid management was restricted to 80–100 ml/h. Table 1 Enhanced recovery protocol . Recovery elements . Day before surgery Normal oral nutrition until midnight No preanaesthetic medication Day of surgery Short-acting i.v. anaesthetic agent No nasogastric drainage; if used, remove immediately after surgery Warm i.v. fluids, and upper- and lower-body air-warming device Prophylactic antibiotics Avoidance of excessive i.v. fluids No routine drainage of peritoneal cavity TEA or CWI analgesia Patient mobilizes to chair on evening of surgery Restart free oral intake of water and nutrition POD 1 Arterial and central lines out Patient mobilizes around bed Patient sits out of bed for 6 h Discontinuation of i.v. fluids if patient haemodynamically stable and drinks more than 1 litre of fluid Normal diet Continue morphine PCA ± CWI/TEA 1000 mg paracetamol every 6 h Transfer to general ward if possible POD 2 Morphine PCA (CWI/TEA removed) Continue mobilization Patient to sit out of bed for 8 h 1000 mg paracetamol every 6 h Urinary catheter out Normal diet Oral analgesia Transfer to general ward POD 3 Continue mobilization Normal diet Check discharge criteria POD 4 Check discharge criteria Discharge . Recovery elements . Day before surgery Normal oral nutrition until midnight No preanaesthetic medication Day of surgery Short-acting i.v. anaesthetic agent No nasogastric drainage; if used, remove immediately after surgery Warm i.v. fluids, and upper- and lower-body air-warming device Prophylactic antibiotics Avoidance of excessive i.v. fluids No routine drainage of peritoneal cavity TEA or CWI analgesia Patient mobilizes to chair on evening of surgery Restart free oral intake of water and nutrition POD 1 Arterial and central lines out Patient mobilizes around bed Patient sits out of bed for 6 h Discontinuation of i.v. fluids if patient haemodynamically stable and drinks more than 1 litre of fluid Normal diet Continue morphine PCA ± CWI/TEA 1000 mg paracetamol every 6 h Transfer to general ward if possible POD 2 Morphine PCA (CWI/TEA removed) Continue mobilization Patient to sit out of bed for 8 h 1000 mg paracetamol every 6 h Urinary catheter out Normal diet Oral analgesia Transfer to general ward POD 3 Continue mobilization Normal diet Check discharge criteria POD 4 Check discharge criteria Discharge i.v., Intravenous; TEA, thoracic epidural analgesia; CWI, continuous wound infiltration; POD, postoperative day; PCA, patient-controlled analgesia. Open in new tab Table 1 Enhanced recovery protocol . Recovery elements . Day before surgery Normal oral nutrition until midnight No preanaesthetic medication Day of surgery Short-acting i.v. anaesthetic agent No nasogastric drainage; if used, remove immediately after surgery Warm i.v. fluids, and upper- and lower-body air-warming device Prophylactic antibiotics Avoidance of excessive i.v. fluids No routine drainage of peritoneal cavity TEA or CWI analgesia Patient mobilizes to chair on evening of surgery Restart free oral intake of water and nutrition POD 1 Arterial and central lines out Patient mobilizes around bed Patient sits out of bed for 6 h Discontinuation of i.v. fluids if patient haemodynamically stable and drinks more than 1 litre of fluid Normal diet Continue morphine PCA ± CWI/TEA 1000 mg paracetamol every 6 h Transfer to general ward if possible POD 2 Morphine PCA (CWI/TEA removed) Continue mobilization Patient to sit out of bed for 8 h 1000 mg paracetamol every 6 h Urinary catheter out Normal diet Oral analgesia Transfer to general ward POD 3 Continue mobilization Normal diet Check discharge criteria POD 4 Check discharge criteria Discharge . Recovery elements . Day before surgery Normal oral nutrition until midnight No preanaesthetic medication Day of surgery Short-acting i.v. anaesthetic agent No nasogastric drainage; if used, remove immediately after surgery Warm i.v. fluids, and upper- and lower-body air-warming device Prophylactic antibiotics Avoidance of excessive i.v. fluids No routine drainage of peritoneal cavity TEA or CWI analgesia Patient mobilizes to chair on evening of surgery Restart free oral intake of water and nutrition POD 1 Arterial and central lines out Patient mobilizes around bed Patient sits out of bed for 6 h Discontinuation of i.v. fluids if patient haemodynamically stable and drinks more than 1 litre of fluid Normal diet Continue morphine PCA ± CWI/TEA 1000 mg paracetamol every 6 h Transfer to general ward if possible POD 2 Morphine PCA (CWI/TEA removed) Continue mobilization Patient to sit out of bed for 8 h 1000 mg paracetamol every 6 h Urinary catheter out Normal diet Oral analgesia Transfer to general ward POD 3 Continue mobilization Normal diet Check discharge criteria POD 4 Check discharge criteria Discharge i.v., Intravenous; TEA, thoracic epidural analgesia; CWI, continuous wound infiltration; POD, postoperative day; PCA, patient-controlled analgesia. Open in new tab Patients randomized to receive CWI underwent the same anaesthetic induction and maintenance protocol. At the end of the operation an ON-Q® dual-limb 12·5-cm Painbuster® (B. Braun, Sheffield, UK) was inserted into the musculofascial layers of the abdominal wall by a single surgeon. The lateral limb was placed between the transversus abdominis and internal oblique muscles, and the medial limb was placed in the rectus sheath posterior to the muscle11,12. On closure of the wound the nerve block was performed by instilling a total of 40 ml 0·125 per cent levobupivacaine into the transversus abdominis plane and the rectus sheath, posteriorly via the wound catheter. The wound catheters were subsequently connected to an elastomeric reservoir containing 0·375 per cent levobupivacaine, which was instilled at a fixed rate of 4 ml/h. This was kept in place for 48 h. Postoperative care Patients were transferred from the operating theatre to the recovery ward, where they were extubated and monitored until observations and pain scores were satisfactory according to the unit protocol. They were subsequently transferred to the high-dependency unit (HDU) where they remained until they had achieved discharge criteria and could be transferred to the general ward. Outcomes Functional recovery time The primary outcome was functional recovery time. Patients were determined to be medically fit for discharge when they fulfilled the following criteria4,21: independent mobilization, eating and drinking with no requirement for intravenous fluid infusion in the previous 24 h, adequate pain control on oral analgesics, blood tests normal or returning to normal, and patient willing to go home. Fulfilment of discharge criteria was determined by the senior responsible clinician, independent of the study team. This was assessed twice daily. Time to fulfilment of discharge criteria was deemed a more accurate reflection of recovery time than actual time spent in hospital owing to logistical difficulties of discharging patients on the day they fulfilled recovery criteria. The sample size calculation was based on functional recovery criteria data. Pain scores Pain scores were recorded by nursing staff, independent of the trial and trained in pain score recording. Pain scores were recorded at 2, 6 and 12 h after closure of the incision and on postoperative day (POD) 1, POD 2 and POD 3 at 09.00 hours. Pain was recorded at rest and on movement. The score was reported on a scale from 0 to 10. Opiate consumption was also recorded. Complications Assessment for complications was performed twice daily. The type of complication was defined according to predetermined complication criteria22, and the severity of complication was graded according to the Dindo–Clavien criteria23. The clinical team coordinated the diagnosis and management of complications, and the investigating team observed twice daily for occurrence of a complication. Parameters of recovery Patient recovery parameters were assessed prospectively. Time to sitting, mobilizing independently, drinking, eating, passing flatus and bowel movement were observed prospectively. ERP compliance rates were also recorded prospectively. Inflammatory response to surgery Plasma levels of cytokines – interleukin (IL) 1β, IL-6, IL-8, IL-10 and tumour necrosis factor (TNF) α – and serum concentrations of HMGB1 were assessed before surgery, and at 24 and 72 h after the incision. Whole blood was collected by standard venepuncture into an EDTA tube, transferred immediately in ice to the laboratory and centrifuged at 1000g for 15 min. The separated plasma was then aliquoted into 1·5-ml polypropylene tubes, frozen immediately and stored at −80°C. For HMGB1 analysis whole blood, collected in a serum gel tube, was centrifuged at 1000g for 15 min at 4°C. Serum aliquots were stored at −80°C. Batch analysis was performed once all samples had been obtained. Cytokine levels were measured by enzyme-linked immunosorbent assay (ELISA) using a Milliplex® human cytokine kit (Millipore, St Charles, Missouri, USA) in accordance with the manufacturer's instructions. High-sensitivity analysis was performed to detect levels from a standard curve ranging from 0·13 to 2000 pg/ml. Analysis of wells and microsphere identification and detection was performed using a Luminex® 100 instrument (Luminex, Austin, Texas, USA). HMGB1 levels were obtained by quantitative sandwich ELISA (Shino-Test Corporation, Tokyo, Japan). A standard curve was used to detect HMGB1 levels ranging from 0 to 80 ng/ml. Intraoperative central venous pressure Intraoperative CVP was monitored continuously. Mean CVP during transection was recorded, as were highest and lowest values during surgery. Minimum mean arterial pressure (MAP) was also recorded. Intravenous fluid and blood product administration was recorded prospectively until POD 2. Statistical analysis Based on previous unpublished data within the unit, a mean(s.d.) recovery time of 7(2·5) days was used. A reduction in functional recovery time of 1·5 days was deemed to be clinically significant12. To achieve a significance of 0·05 and a power of 80 per cent, 44 patients were required in each arm of the study. Continuous data were compared by means of the Mann–Whitney U test or t test, depending on the distribution of the data. A paired t test or Wilcoxon signed-rank test was used for analysis of paired data. Categorical data were analysed using the χ2 or Fisher's exact test. Comparisons of multiple measures were performed with two-way ANOVA with repeated measures. GraphPad Prism® version 6.00 for Windows® (GraphPad Software, La Jolla, California, USA) was used for statistical analysis. Results The trial was conducted from December 2012 to June 2014. Follow-up was complete by August 2014. A total of 100 patients were randomized (Fig. 1). Six of the patients initially randomized to TEA did not undergo resection. Forty-four and 49 patients in the TEA and CWI groups respectively were included in the analysis. Two patients in the CWI group did not have blood samples taken on POD 3 (1 refused, 1 died) and so were excluded from the inflammatory response analysis. Two patients (1 in each group) did not have a functioning central venous line and were excluded from the CVP analysis. Fig. 1 Open in new tabDownload slide CONSORT diagram for the trial. TEA, thoracic epidural analgesia; CWI, continuous wound infiltration Baseline characteristics were broadly similar, except for a greater proportion of patients with a histologically normal background liver in the TEA group (Table 2). The duration of anaesthesia was significantly shorter in the CWI group. Intermittent portal triad clamping time did not differ significantly between the groups. There was no significant difference in estimated blood loss (Table 3). Table 2 Baseline characteristics . TEA (n = 44) . CWI (n = 49) . Age (years)* 62·6(11·1) 62·8(12·1) Sex ratio (M : F) 29 : 15 28 : 21 Body mass index (kg/m2)* 27·7(4·2) 28·5(5·0) ASA grade I 3 3 II 32 39 III 9 7 Indication Colorectal liver metastases 29 34 Hepatocellular carcinoma 6 7 Cholangiocarcinoma 0 1 Other malignancy 2 4 Benign condition 7 3 Tumour size (mm)* 47·6(44·8) 45·2(44·0) No. of tumours* 1·4(1·2) 1·7(1·2) Background liver Steatosis 16 20 Steatohepatitis 2 12 Sinusoidal dilatation 10 11 Cirrhosis 3 4 Normal liver 17† 2 Extent of resection Major resection 27 33 Minor resection 17 16 Previous PVE 4 3 Preoperative chemotherapy 23 29 . TEA (n = 44) . CWI (n = 49) . Age (years)* 62·6(11·1) 62·8(12·1) Sex ratio (M : F) 29 : 15 28 : 21 Body mass index (kg/m2)* 27·7(4·2) 28·5(5·0) ASA grade I 3 3 II 32 39 III 9 7 Indication Colorectal liver metastases 29 34 Hepatocellular carcinoma 6 7 Cholangiocarcinoma 0 1 Other malignancy 2 4 Benign condition 7 3 Tumour size (mm)* 47·6(44·8) 45·2(44·0) No. of tumours* 1·4(1·2) 1·7(1·2) Background liver Steatosis 16 20 Steatohepatitis 2 12 Sinusoidal dilatation 10 11 Cirrhosis 3 4 Normal liver 17† 2 Extent of resection Major resection 27 33 Minor resection 17 16 Previous PVE 4 3 Preoperative chemotherapy 23 29 * Values are mean(s.d.). TEA, thoracic epidural analgesia; CWI, continuous wound infiltration; ASA, American Society of Anesthesiologists; PVE, portal vein embolization. † P < 0·001 versus CWI (Fisher's exact test). Open in new tab Table 2 Baseline characteristics . TEA (n = 44) . CWI (n = 49) . Age (years)* 62·6(11·1) 62·8(12·1) Sex ratio (M : F) 29 : 15 28 : 21 Body mass index (kg/m2)* 27·7(4·2) 28·5(5·0) ASA grade I 3 3 II 32 39 III 9 7 Indication Colorectal liver metastases 29 34 Hepatocellular carcinoma 6 7 Cholangiocarcinoma 0 1 Other malignancy 2 4 Benign condition 7 3 Tumour size (mm)* 47·6(44·8) 45·2(44·0) No. of tumours* 1·4(1·2) 1·7(1·2) Background liver Steatosis 16 20 Steatohepatitis 2 12 Sinusoidal dilatation 10 11 Cirrhosis 3 4 Normal liver 17† 2 Extent of resection Major resection 27 33 Minor resection 17 16 Previous PVE 4 3 Preoperative chemotherapy 23 29 . TEA (n = 44) . CWI (n = 49) . Age (years)* 62·6(11·1) 62·8(12·1) Sex ratio (M : F) 29 : 15 28 : 21 Body mass index (kg/m2)* 27·7(4·2) 28·5(5·0) ASA grade I 3 3 II 32 39 III 9 7 Indication Colorectal liver metastases 29 34 Hepatocellular carcinoma 6 7 Cholangiocarcinoma 0 1 Other malignancy 2 4 Benign condition 7 3 Tumour size (mm)* 47·6(44·8) 45·2(44·0) No. of tumours* 1·4(1·2) 1·7(1·2) Background liver Steatosis 16 20 Steatohepatitis 2 12 Sinusoidal dilatation 10 11 Cirrhosis 3 4 Normal liver 17† 2 Extent of resection Major resection 27 33 Minor resection 17 16 Previous PVE 4 3 Preoperative chemotherapy 23 29 * Values are mean(s.d.). TEA, thoracic epidural analgesia; CWI, continuous wound infiltration; ASA, American Society of Anesthesiologists; PVE, portal vein embolization. † P < 0·001 versus CWI (Fisher's exact test). Open in new tab Table 3 Operative details . TEA (n = 44) . CWI (n = 49) . P* . Duration of operation (min) 198 (141–250) 226 (171–306) 0·066 Estimated blood loss (ml) 675 (452–1313) 800 (475–1500) 0·410 Pringle time (min) 12 (0–27) 22 (2–34) 0·077 Transection time (min) 70 (42–100) 75 (48–110) 0·433 Wound size (cm) 31 (28–35) 30 (27–33) 0·255 Duration of anaesthesia (min) 50 (40–56) 35 (30–37) < 0·001 . TEA (n = 44) . CWI (n = 49) . P* . Duration of operation (min) 198 (141–250) 226 (171–306) 0·066 Estimated blood loss (ml) 675 (452–1313) 800 (475–1500) 0·410 Pringle time (min) 12 (0–27) 22 (2–34) 0·077 Transection time (min) 70 (42–100) 75 (48–110) 0·433 Wound size (cm) 31 (28–35) 30 (27–33) 0·255 Duration of anaesthesia (min) 50 (40–56) 35 (30–37) < 0·001 Values are median (i.q.r.). TEA, thoracic epidural analgesia; CWI, continuous wound infiltration. * Mann–Whitney U test. Open in new tab Table 3 Operative details . TEA (n = 44) . CWI (n = 49) . P* . Duration of operation (min) 198 (141–250) 226 (171–306) 0·066 Estimated blood loss (ml) 675 (452–1313) 800 (475–1500) 0·410 Pringle time (min) 12 (0–27) 22 (2–34) 0·077 Transection time (min) 70 (42–100) 75 (48–110) 0·433 Wound size (cm) 31 (28–35) 30 (27–33) 0·255 Duration of anaesthesia (min) 50 (40–56) 35 (30–37) < 0·001 . TEA (n = 44) . CWI (n = 49) . P* . Duration of operation (min) 198 (141–250) 226 (171–306) 0·066 Estimated blood loss (ml) 675 (452–1313) 800 (475–1500) 0·410 Pringle time (min) 12 (0–27) 22 (2–34) 0·077 Transection time (min) 70 (42–100) 75 (48–110) 0·433 Wound size (cm) 31 (28–35) 30 (27–33) 0·255 Duration of anaesthesia (min) 50 (40–56) 35 (30–37) < 0·001 Values are median (i.q.r.). TEA, thoracic epidural analgesia; CWI, continuous wound infiltration. * Mann–Whitney U test. Open in new tab Functional recovery time Functional recovery time was significantly shorter in the CWI group: median (i.q.r.) 5·75 (4–7) days compared with 6·5 (5–9·75) days in the TEA group (P = 0·036). The CWI group also spent less time in the HDU than the TEA group: median (i.q.r.) 42 (24–65) and 46 (42–66) h respectively (P = 0·039). Readmission rates were similar in the two groups: three of 49 and five of 44 respectively (P = 0·470). No differences were observed in achievement of recovery milestones (Table 4) or ERP compliance (Table S1, supporting information). Table 4 Postoperative outcomes . TEA (n = 44) . CWI (n = 49) . P† . Time to recovery (h)* Passage of flatus 51 (23–79) 65 (34–78) 0·281 Bowel movement 90 (55–120) 79 (65–101) 0·473 Sitting 22 (20–36·5) 22 (19–24) 0·524 Mobilization 60 (44–85) 53 (39–72) 0·312 First drink 6 (4–14) 7 (4–12) 0·681 First meal 20 (16–39) 20 (18–24) 0·986 Opiate consumption (mg)* Intraoperative 0 (0–1·9) 24·0 (20·0–37·5) < 0·001 Day of surgery 0 (0–22·6) 24·0 (14·0–51·0) < 0·001 POD 1 0 (0–34·5) 27·0 (16·0–47·0) < 0·001 POD 2 18·5 (10·5–27·0) 12·0 (4·0–20·0) 0·020 POD 3 15·0 (8·0–23·5) 8·0 (3·0–16·0) 0·018 Postoperative outcome Death 0 3 0·244‡ Any complication 31 25 0·056§ Complication type Collection/abscess 3 3 1·000‡ Bleeding 1 2 1·000‡ Wound dehiscence 1 0 0·473‡ Wound infection 3 2 0·665‡ Hypotension 10 3 0·034‡ Lower respiratory tract infection 0 3 0·244‡ Urinary retention 5 2 0·249‡ Ileus 5 4 0·731‡ Acute kidney injury 5 3 0·470‡ Arrhythmia 0 3 0·244‡ Dindo–Clavien classification 0·186§ 0 13 21 1 10 6 II 15 10 IIIa 4 5 IIIb 0 3 IVa 0 0 IVb 2 1 V 0 3 * . TEA (n = 44) . CWI (n = 49) . P† . Time to recovery (h)* Passage of flatus 51 (23–79) 65 (34–78) 0·281 Bowel movement 90 (55–120) 79 (65–101) 0·473 Sitting 22 (20–36·5) 22 (19–24) 0·524 Mobilization 60 (44–85) 53 (39–72) 0·312 First drink 6 (4–14) 7 (4–12) 0·681 First meal 20 (16–39) 20 (18–24) 0·986 Opiate consumption (mg)* Intraoperative 0 (0–1·9) 24·0 (20·0–37·5) < 0·001 Day of surgery 0 (0–22·6) 24·0 (14·0–51·0) < 0·001 POD 1 0 (0–34·5) 27·0 (16·0–47·0) < 0·001 POD 2 18·5 (10·5–27·0) 12·0 (4·0–20·0) 0·020 POD 3 15·0 (8·0–23·5) 8·0 (3·0–16·0) 0·018 Postoperative outcome Death 0 3 0·244‡ Any complication 31 25 0·056§ Complication type Collection/abscess 3 3 1·000‡ Bleeding 1 2 1·000‡ Wound dehiscence 1 0 0·473‡ Wound infection 3 2 0·665‡ Hypotension 10 3 0·034‡ Lower respiratory tract infection 0 3 0·244‡ Urinary retention 5 2 0·249‡ Ileus 5 4 0·731‡ Acute kidney injury 5 3 0·470‡ Arrhythmia 0 3 0·244‡ Dindo–Clavien classification 0·186§ 0 13 21 1 10 6 II 15 10 IIIa 4 5 IIIb 0 3 IVa 0 0 IVb 2 1 V 0 3 * * Values are median (i.q.r.). If a patient had more than one complication, the highest Dindo–Clavien grade was reported. TEA, thoracic epidural analgesia; CWI, continuous wound infiltration; POD, postoperative day. † Mann–Whitney U test, except ‡ Fisher's exact test and § χ2 test. ¶ Dindo–Clavien grade III or more versus less than III. Open in new tab Table 4 Postoperative outcomes . TEA (n = 44) . CWI (n = 49) . P† . Time to recovery (h)* Passage of flatus 51 (23–79) 65 (34–78) 0·281 Bowel movement 90 (55–120) 79 (65–101) 0·473 Sitting 22 (20–36·5) 22 (19–24) 0·524 Mobilization 60 (44–85) 53 (39–72) 0·312 First drink 6 (4–14) 7 (4–12) 0·681 First meal 20 (16–39) 20 (18–24) 0·986 Opiate consumption (mg)* Intraoperative 0 (0–1·9) 24·0 (20·0–37·5) < 0·001 Day of surgery 0 (0–22·6) 24·0 (14·0–51·0) < 0·001 POD 1 0 (0–34·5) 27·0 (16·0–47·0) < 0·001 POD 2 18·5 (10·5–27·0) 12·0 (4·0–20·0) 0·020 POD 3 15·0 (8·0–23·5) 8·0 (3·0–16·0) 0·018 Postoperative outcome Death 0 3 0·244‡ Any complication 31 25 0·056§ Complication type Collection/abscess 3 3 1·000‡ Bleeding 1 2 1·000‡ Wound dehiscence 1 0 0·473‡ Wound infection 3 2 0·665‡ Hypotension 10 3 0·034‡ Lower respiratory tract infection 0 3 0·244‡ Urinary retention 5 2 0·249‡ Ileus 5 4 0·731‡ Acute kidney injury 5 3 0·470‡ Arrhythmia 0 3 0·244‡ Dindo–Clavien classification 0·186§ 0 13 21 1 10 6 II 15 10 IIIa 4 5 IIIb 0 3 IVa 0 0 IVb 2 1 V 0 3 * . TEA (n = 44) . CWI (n = 49) . P† . Time to recovery (h)* Passage of flatus 51 (23–79) 65 (34–78) 0·281 Bowel movement 90 (55–120) 79 (65–101) 0·473 Sitting 22 (20–36·5) 22 (19–24) 0·524 Mobilization 60 (44–85) 53 (39–72) 0·312 First drink 6 (4–14) 7 (4–12) 0·681 First meal 20 (16–39) 20 (18–24) 0·986 Opiate consumption (mg)* Intraoperative 0 (0–1·9) 24·0 (20·0–37·5) < 0·001 Day of surgery 0 (0–22·6) 24·0 (14·0–51·0) < 0·001 POD 1 0 (0–34·5) 27·0 (16·0–47·0) < 0·001 POD 2 18·5 (10·5–27·0) 12·0 (4·0–20·0) 0·020 POD 3 15·0 (8·0–23·5) 8·0 (3·0–16·0) 0·018 Postoperative outcome Death 0 3 0·244‡ Any complication 31 25 0·056§ Complication type Collection/abscess 3 3 1·000‡ Bleeding 1 2 1·000‡ Wound dehiscence 1 0 0·473‡ Wound infection 3 2 0·665‡ Hypotension 10 3 0·034‡ Lower respiratory tract infection 0 3 0·244‡ Urinary retention 5 2 0·249‡ Ileus 5 4 0·731‡ Acute kidney injury 5 3 0·470‡ Arrhythmia 0 3 0·244‡ Dindo–Clavien classification 0·186§ 0 13 21 1 10 6 II 15 10 IIIa 4 5 IIIb 0 3 IVa 0 0 IVb 2 1 V 0 3 * * Values are median (i.q.r.). If a patient had more than one complication, the highest Dindo–Clavien grade was reported. TEA, thoracic epidural analgesia; CWI, continuous wound infiltration; POD, postoperative day. † Mann–Whitney U test, except ‡ Fisher's exact test and § χ2 test. ¶ Dindo–Clavien grade III or more versus less than III. Open in new tab Morbidity Complications occurred in 31 patients (70 per cent) in the epidural group compared with 25 (51 per cent) in the CWI group (P = 0·056) (Table 4). There was no difference in severity of morbidity between groups. Three deaths occurred in the CWI group and none in the TEA group. Two deaths resulted from multiple organ failure following hepatic insufficiency, and one was secondary to multiple organ failure on day 35 as a result of colonic ischaemia. Ten patients in the TEA group developed sustained hypotension related to epidural blockade (P = 0·034). Hypotension was included as morbidity if an infusion of vasopressor was commenced (Dindo–Clavien grade II). Pain scores Pain scores were not significantly different between CWI and TEA groups at rest or on movement (Fig. 2). Opiate consumption was greater in the CWI group up to POD 1. Thereafter patients in the TEA group received a greater amount of opiates (Table 4). Four patients in the CWI group and two in the TEA group were intubated and ventilated at some point during the first 3 days after surgery, and so were not included in this analysis. Fig. 2 Open in new tabDownload slide Mean(s.d.) pain scores a at rest and b on movement after liver resection, measured on a visual analogue scale from 1 to 10. TEA, thoracic epidural analgesia; CWI, continuous wound infiltration. aP = 0·518 and bP = 0·060 (2-way repeated-measures ANOVA) Inflammatory response to surgery Levels of proinflammatory mediators, except IL-1β, were all raised on POD 1 and 3 compared with baseline in both groups (Table S2, supporting information). Levels peaked on POD 1, with no significant difference between groups. Levels then decreased by POD 3 but remained higher than preoperative values. Two-way repeated-measures ANOVA of all three levels for each inflammatory mediator showed no significant difference. Haemodynamic response There were no significant differences in patterns of fluid administration, other than a significantly greater volume of intravenous crystalloid being administered in the TEA group on POD 1: mean (i.q.r.) 1618 (1277–1908) ml versus 1450 (1200–1572) ml in the CWI group (P = 0·024). Four patients in each group received an intraoperative blood transfusion (P = 1·000) (Table S3, supporting information). No difference was observed in CVP change, peak CVP or mean CVP during transection. Minimum MAP was, however, lower in the TEA group than in the CWI group: mean(s.d.) 57·7(5·6) versus 61·2(7·8) mmHg (P = 0·015) (Table S4, supporting information). Discussion This trial has shown in patients undergoing open liver resection that CWI offers an advantage in recovery time compared with epidural analgesia. No difference in resting or dynamic pain scores was observed. The inflammatory response was not attenuated to a greater degree in the epidural group, and no advantage in terms of CVP reduction or blood loss was observed. CWI could be considered as a first-line analgesic modality after liver resection. Epidurals are often advocated as important for ERPs owing to the perception of their role in pain control, postoperative mobilization, and reduction of ileus and postoperative nausea and vomiting1,4,24. However, when all aspects of postoperative care are optimized, an epidural may not actually provide a major advantage, as shown in the present trial. A recent meta-analysis25 looking at recovery after abdominal surgery reported no difference in length of hospital stay when all studies comparing epidural with either systemic opiates or CWI within an ERP were compared. Two previous studies showed epidural analgesia to prolong length of stay within an ERP after liver resection compared with CWI12 or intrathecal morphine injection26. The present randomized comparison of analgesic techniques following liver surgery confirmed that epidurals do not necessarily improve postoperative recovery times, particularly in an ERP setting. TEA was associated with an increased HDU stay, hypotension requiring vasopressor infusion, and increased opiate consumption on POD 2 and 3. These findings may explain why recovery time was prolonged in the epidural group. Epidurals have historically been associated with a reduction in pulmonary and cardiac complications following surgery27,28. However, this evidence base is becoming increasingly outdated owing to the lack of trials including ERPs, comparisons of epidurals and systemic opiates only, and heterogeneity of operation specialty and procedure. A recent meta-analysis25 compared epidurals alone versus all other analgesic modalities (patient-controlled analgesia and CWI) in patients who all underwent ERPs, and showed no difference in complication rates. This is in agreement with the present study, which did not show a difference in overall morbidity rate. This should probably be interpreted as evidence that a well constructed postoperative ERP should mitigate against preventable complications by providing optimal care, and that an epidural is not necessarily required to achieve this. The complication rate here is high, but may reflect the prospective nature of the observations and reporting of all minor complications. Another area of interest is postoperative pain control. This trial showed no difference between TEA and CWI in both dynamic and resting pain scores. In open abdominal surgery epidurals have consistently been shown to provide superior pain scores in comparison with intravenous opiates29. However, a meta-analysis30 comparing epidural analgesia with CWI showed no difference in analgesic benefit following abdominal surgery. The generalizability of these data to liver surgery, however, is not straightforward, as liver resection involves subcostal incisions mandating alternative epidural and wound catheter configurations. The evidence assessing pain relief after hepatic surgery is less extensive; there are no randomized trials comparing TEA with systemic opiates, and heterogeneous comparisons of alternative modalities such as CWI have been presented with no definitive conclusion31. The only previous trial12 comparing TEA and CWI after liver surgery showed notably higher early and overall pain scores in the CWI group. In an effort to mitigate against this, a higher-volume transversus abdominis plane and rectus sheath block was incorporated in the present trial to improve pain scores in the immediate postoperative period. This seemed to have had an effect as both early pain scores, overall pain scores and opiate administration were comparable to those in the TEA group. A further potential argument for the use of epidurals over alternative analgesic methods is the increased attenuation of the inflammatory response by epidural analgesia24. However, a similar rise in inflammatory markers was observed in both groups (except IL-1β) here, implying no advantage in this respect for either type of analgesia. Both TEA and CWI have been reported to attenuate the inflammatory response compared with controls in a variety of abdominal surgical settings13–14,32–33, and a meta-analysis15 has shown an association between attenuation of the inflammatory response and clinical outcome. The present trial has shown that TEA does not provide an advantage in this respect compared with CWI, and it is likely that CWI exhibits its own inflammatory response attenuation via the instillation of local anaesthetic to afferent nerve fibres. The third area of investigation in this trial was the effect of either TEA or CWI on intraoperative liver transection. No difference was observed in terms of CVP, intravenous fluid administration or blood loss during transection between the groups. Large RCTs comparing the effect of different analgesia types on CVP reduction during liver surgery are lacking, and no trial has assessed the effect of epidural analgesia alone on CVP. Epidural analgesia has been included in CVP-lowering protocols along with intravenous fluid restriction, nitrates, furosemide and patient positioning. These trials34–36 have shown significant improvements in blood loss following CVP reduction by these methods. However, the impact and importance of TEA in these multimodal protocols is not clear. Therefore, in light of the present findings, epidural analgesia may not offer an advantage in terms of CVP reduction and should not be considered mandatory to achieve low CVP and minimize bleeding. Intravenous fluid administration was similar in the two groups throughout the trial. TEA is often associated with postoperative fluid overload7, but this was not evident. It is inferred that, provided an appropriate fluid restriction regimen is adhered to, fluid overload can be avoided, regardless of analgesic modality. The results of this trial need to be considered in the context of its limitations. First, it was decided to run the trial as an unblinded study. Sham procedures are not necessarily effective in providing genuine blinding. Moreover, not knowing the analgesic regimen can prevent the application of potentially beneficial care components that either method affords, thus providing an unrealistic and therefore untranslatable picture of the effect of each technique. A further limitation of the trial is the lack of specific assessment of recovery quality in the early postoperative period. However, achievement of ERP milestones was assessed, and similar progression was observed between the groups throughout the trial. Finally, the timing of the assessment of inflammatory mediators requires consideration. The wound catheter was not inserted until the end of the procedure, and it is unlikely that CWI would have affected the inflammatory response immediately after surgery, in contrast to TEA. However, samples taken at 24 and 72 h after incision allowed a fair comparison of the inflammatory response, despite potentially missing the early postoperative peak. Collaborators Collaborators in the LIVER 2 trial were: O. J. Garden, R. W. Parks, J. J. Powell, A. Adair, D. Mole, R. Ravindran, D. McKeown, A. Pollock, B. Cook, E. McDonald, R. Mayes, E. Thompson, E. Wilson and D. Cameron (Royal Infirmary of Edinburgh, Edinburgh, UK). Acknowledgements This trial was funded by the UK Defence Medical Services. Disclosure: The authors declare no conflict of interest. Supporting information Additional supporting information may be found in the online version of this article: Table S1 Compliance with enhanced recovery protocol (Word document) Table S2 Markers of the inflammatory response (Word document) Table S3 Fluid balance and administration (Word document) Table S4 Intraoperative haemodynamic changes (Word document) Open in new tabDownload slide References 1 Jones C , Kelliher L, Dickinson M, Riga A, Worthington T, Scott MJ et al. Randomized clinical trial on enhanced recovery versus standard care following open liver resection . Br J Surg 2013 ; 100 : 1015 – 1024 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Lin DX , Li X, Ye QW, Lin F, Li LL, Zhang QY. Implementation of a fast-track clinical pathway decreases postoperative length of stay and hospital charges for liver resection . Cell Biochem Biophys 2011 ; 61 : 413 – 419 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Schultz NA , Larsen PN, Klarskov B, Plum LM, Frederiksen HJ, Christensen BM et al. Evaluation of a fast-track programme for patients undergoing liver resection . Br J Surg 2013 ; 100 : 138 – 143 . Google Scholar Crossref Search ADS PubMed WorldCat 4 van Dam RM , Hendry PO, Coolsen MM, Bemelmans MH, Lassen K, Revhaug A et al. Initial experience with a multimodal enhanced recovery programme in patients undergoing liver resection . Br J Surg 2008 ; 95 : 969 – 975 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Fazakas J , Tóth S, Füle B, Smudla A, Mándli T, Radnai M et al. Epidural anesthesia? No of course . Transplant Proc 2008 ; 40 : 1216 – 1217 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Tzimas P , Prout J, Papadopoulos G, Mallett SV. Epidural anaesthesia and analgesia for liver resection . Anaesthesia 2013 ; 68 : 628 – 635 . Google Scholar Crossref Search ADS PubMed WorldCat 7 McNally SJ , Revie EJ, Massie LJ, McKeown DW, Parks RW, Garden OJ et al. Factors in perioperative care that determine blood loss in liver surgery . HPB (Oxford) 2012 ; 14 : 236 – 241 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Yuan FS , Ng SY, Ho KY, Lee SY, Chung AY, Poopalalingam R. Abnormal coagulation profile after hepatic resection: the effect of chronic hepatic disease and implications for epidural analgesia . J Clin Anesth 2012 ; 24 : 398 – 403 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Ready LB . Acute pain: lessons learned from 25 000 patients . Reg Anesth Pain Med 1999 ; 24 : 499 – 505 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 10 Page A , Rostad B, Staley CA, Levy JH, Park J, Goodman M et al. Epidural analgesia in hepatic resection . J Am Coll Surg 2008 ; 206 : 1184 – 1192 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Basu S , Tamijmarane A, Bulters D, Wells JK, John TG, Rees M. An alternative method of wound pain control following hepatic resection: a preliminary study . HPB (Oxford) 2004 ; 6 : 186 – 189 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Revie EJ , McKeown DW, Wilson JA, Garden OJ, Wigmore SJ. Randomized clinical trial of local infiltration plus patient-controlled opiate analgesia vs. epidural analgesia following liver resection surgery . HPB (Oxford) 2012 ; 14 : 611 – 618 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Beilin B , Shavit Y, Trabekin E, Mordashev B, Mayburd E, Zeidel A et al. The effects of postoperative pain management on immune response to surgery . Anesth Analg 2003 ; 97 : 822 – 827 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 14 Hadimioglu N , Ulugol H, Akbas H, Coskunfirat N, Ertug Z, Dinckan A. Combination of epidural anesthesia and general anesthesia attenuates stress response to renal transplantation surgery . Transplant Proc 2012 ; 44 : 2949 – 2954 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Orci LA , Toso C, Mentha G, Morel P, Majno PE. Systematic review and meta-analysis of the effect of perioperative steroids on ischaemia–reperfusion injury and surgical stress response in patients undergoing liver resection . Br J Surg 2013 ; 100 : 600 – 609 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Rahbari NN , Koch M, Zimmermann JB, Elbers H, Bruckner T, Contin P et al. Infrahepatic inferior vena cava clamping for reduction of central venous pressure and blood loss during hepatic resection: a randomized controlled trial . Ann Surg 2011 ; 253 : 1102 – 1110 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Takahata R , Ono S, Tsujimoto H, Hiraki S, Kimura A, Kinoshita M et al. Postoperative serum concentrations of high mobility group box chromosomal protein-1 correlates to the duration of SIRS and pulmonary dysfunction following gastrointestinal surgery . J Surg Res 2011 ; 170 : e135 – e140 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Suda K , Takeuchi H, Ishizaka A, Kitagawa Y. High-mobility-group box chromosomal protein 1 as a new target for modulating stress response . Surg Today 2010 ; 40 : 592 – 601 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Chan AW , Tetzlaff JM, Altman DG, Laupacis A, Gøtzsche PC, Krleža-Jerić K et al. SPIRIT 2013 statement: defining standard protocol items for clinical trials . Ann Intern Med 2013 ; 158 : 200 – 207 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Schulz KF , Altman DG, Moher D. CONSORT 2010 statement: updated guidelines for reporting parallel group randomized trials . Ann Intern Med 2010 ; 152 : 726 – 732 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Hendry PO , van Dam RM, Bukkems SF, McKeown DW, Parks RW, Preston T et al. Randomized clinical trial of laxatives and oral nutritional supplements within an enhanced recovery after surgery protocol following liver resection . Br J Surg 2010 ; 97 : 1198 – 1206 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Buzby GP , Knox LS, Crosby LO, Eisenberg JM, Haakenson CM, McNeal GE et al. Study protocol: a randomized clinical trial of total parenteral nutrition in malnourished surgical patients . Am J Clin Nutr 1988 ; 47 ( Suppl ): 366 – 381 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Dindo D , Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey . Ann Surg 2004 ; 240 : 205 – 213 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Gustafsson UO , Scott MJ, Schwenk W, Demartines N, Roulin D, Francis N et al. Guidelines for perioperative care in elective colonic surgery: Enhanced Recovery After Surgery (ERAS®) Society recommendations . Clin Nutr 2012 ; 31 : 783 – 800 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Hughes MJ , Ventham NT, McNally S, Harrison E, Wigmore S. Analgesia after open abdominal surgery in the setting of enhanced recovery surgery: a systematic review and meta-analysis . JAMA Surg 2014 ; 149 : 1224 – 1230 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Koea JB , Young Y, Gunn K. Fast track liver resection: the effect of a comprehensive care package and analgesia with single dose intrathecal morphine with gabapentin or continuous epidural analgesia . HPB Surg 2009 ; 2009 : 271986 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Ballantyne JC , Carr DB, deFerranti S, Suarez T, Lau J, Chalmers TC et al. The comparative effects of postoperative analgesic therapies on pulmonary outcome: cumulative meta-analyses of randomized, controlled trials . Anesth Analg 1998 ; 86 : 598 – 612 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Beattie WS , Badner NH, Choi P. Epidural analgesia reduces postoperative myocardial infarction: a meta-analysis . Anesth Analg 2001 ; 93 : 853 – 858 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Jørgensen H , Wetterslev J, Møiniche S, Dahl JB Epidural local anaesthetics versus opioid-based analgesic regimens on postoperative gastrointestinal paralysis, PONV and pain after abdominal surgery . Cochrane Database Syst Rev 2000 ; ( 4 ) CD001893 . Google Scholar OpenURL Placeholder Text WorldCat 30 Ventham NT , Hughes M, O'Neill S, Johns N, Brady RR, Wigmore SJ. Systematic review and meta-analysis of continuous local anaesthetic wound infiltration versus epidural analgesia for postoperative pain following abdominal surgery . Br J Surg 2013 ; 100 : 1280 – 1289 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Hughes M , McNally S, McKeown D, Wigmore S. The effect of analgesic modality on outcome following open liver surgery: a systematic review of post-operative analgesia . Minerva Anestesiol 2015 ; 81 : 541 – 556 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 32 Liu MH , Tian J, Su YP, Wang T, Xiang Q, Wen L. Cervical sympathetic block regulates early systemic inflammatory response in severe trauma patients . Med Sci Monit 2013 ; 19 : 194 – 201 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 33 Zhou D , Gu FM, Gao Q, Li QL, Zhou J, Miao CH. Effects of anesthetic methods on preserving anti-tumor T-helper polarization following hepatectomy . World J Gastroenterol 2012 ; 18 : 3089 – 3098 . Google Scholar Crossref Search ADS PubMed WorldCat 34 El-Kharboutley WS , El-Wahab MA. The role of adoption of low central venous pressure in hepatic resection with Pringle manoeuvre in reducing blood loss and improving operative outcome . Egyptian J Anaesth 2004 ; 20 : 369 – 376 . Google Scholar OpenURL Placeholder Text WorldCat 35 Liu HZ , Zhou QL, Wang XH, Yang CX, Xu YH. Application of low central venous pressure in liver resection . Zhonghua Gandan Waike Zazhi 2005 ; 11 : 461 – 463 . Google Scholar OpenURL Placeholder Text WorldCat 36 Liu Y , Cai M, Duan S, Peng X, Lai Y, Li Y. Effect of controlled low central venous pressure on renal function in major liver resection. Chinese–German Journal of Clinical Oncology 2008 ; 7 : 7 – 9 . Google Scholar OpenURL Placeholder Text WorldCat Author notes Presented to the Military Health System Research Symposium, Fort Lauderdale, Florida, USA, August 2014 © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd

Humes, D J; Walker, A J; Blackwell, J; Hunt, B J; West, J

doi: 10.1002/bjs.9923pmid: 26387670

Abstract Background Guidelines recommend extended thromboprophylaxis following colectomy for malignant disease, but not for non-malignant disease. The aim of this study was to determine absolute and relative rates of venous thromboembolism (VTE) following colectomy by indication, admission type and time after surgery. Methods A cohort study of patients undergoing colectomy in England was undertaken using linked primary (Clinical Practice Research Datalink) and secondary (Hospital Episode Statistics) care data (2001–2011). Crude rates and adjusted hazard ratios (HRs) were calculated for the risk of first VTE following colectomy using Cox regression analysis. Results Some 12 388 patients were identified; 312 (2·5 per cent) developed VTE after surgery, giving a rate of 29·59 (95 per cent c.i. 26·48 to 33·06) per 1000 person-years in the first year after surgery. Overall rates were 2·2-fold higher (adjusted HR 2·23, 95 per cent c.i. 1·76 to 2·50) for emergency compared with elective admissions (39·44 versus 25·78 per 1000 person-years respectively). Rates of VTE were 2·8-fold higher in patients with malignant disease versus those with non-malignant disease (adjusted HR 2·84, 2·04 to 3·94). The rate of VTE was highest in the first month after emergency surgery, and declined from 121·68 per 1000 person-years in the first month to 25·65 per 1000 person-years during the rest of the follow-up interval. Crude rates of VTE were similar for malignant and non-malignant disease (114·76 versus 120·98 per 1000 person-years respectively) during the first month after emergency surgery. Conclusion Patients undergoing emergency colectomy for non-malignant disease have a similar risk of VTE as patients with malignant disease in the first month after surgery. Introduction More than 33 000 colectomies are performed in England each year, and nearly 250 000 in the USA, for non-malignant and malignant disease, with nearly one-quarter occurring as an emergency procedure1. Recent international guidance has recommended that patients undergoing colectomy for malignant disease should receive extended pharmacological thromboprophylaxis for 4 weeks after surgery, based on evidence from four randomized trials including 901 patients, of whom over 80 per cent had malignant disease2–5. A subsequent systematic review6 of these studies reported a reduction in overall venous thromboembolism (VTE) from 14·3 to 6·1 per cent (odds ratio 0·41, 95 per cent c.i. 0·26 to 0·63). However, no such evidence exists for patients undergoing colectomy for non-malignant disease, who account for 40–50 per cent of patients having a colectomy7,8. Hence, no specific guidance has been issued on extended VTE prophylaxis in this group, despite these patients having underlying medical conditions such as ulcerative colitis, Crohn's disease and diverticular disease that are associated with an inherent increased risk of VTE compared with that in the general population, in both the ambulatory and hospital setting9–11. Previous studies describing the risk of VTE following colectomy for non-malignant disease have had limited follow-up (30 days), reducing their ability to quantify the highest-risk period for VTE following surgery7,12. Similarly, studies reporting the risk of VTE following colectomy for malignant disease have used a limited 30-day follow-up13, or have compared the risk of VTE after colectomy with that in patients with advanced metastatic disease, and hence have concluded, probably erroneously, that the risk of VTE is reduced after surgery14. Published studies on VTE following colectomy have failed to quantify the risk in terms of indication for surgery (particularly non-malignant indications), timing of surgery (emergency versus elective), and duration or timing of the highest risk of VTE after operation. The aim of the present study was to determine the magnitude and duration of risk of VTE following colectomy by indication for surgery and timing of surgery using linked electronic healthcare data from England15. Methods The study was approved by the Independent Scientific Advisory Committee approval board (Protocol 11-051R). Data sources The Clinical Practice Research Datalink (CPRD) is a primary care database containing diagnostic and prescription data for approximately 13 million people of the general population in the UK, with 3·4 million active patients contributing data. Diseases are coded within the CPRD using Read codes which have been used by clinicians to record data within primary care since 198516. Hospital Episode Statistics (HES) collects a record for each episode of admitted patient care delivered in England, either by National Health Service (NHS) hospitals or delivered in the independent sector but commissioned by the NHS. It has collected data since 1989, with more than 15 million new records added each year. Records are coded using a combination of ICD-10 codes for diagnosis at discharge along with OPCS-4 codes detailing procedures performed. Increasingly HES data are being used to study surgical diseases, with a recent systematic review17 reporting that approximately two-thirds of studies published using these data concern surgical conditions. Death certificate data from the Office for National Statistics were also used. Anonymized patient identifiers from CPRD and HES were linked by a trusted third party by using the NHS number, date of birth, postcode and sex. Most patients were matched exactly according to NHS number (over 90 per cent of patients are linked in this way), with the remaining patients linked probabilistically on the basis of postcode, date of birth and sex. In the version employed, 53·0 per cent of practices in the CPRD are linked to HES, representing a 3·0 per cent sample of the population of England. These data have been shown previously18 to be similar in terms of age, sex and geographical distribution to data from the UK population. Cohort The cohort was identified using OPCS codes for any colectomy from HES data between 2001 and 2011 (Appendix S1, supporting information). Operations confined to or including the rectum and anal canal were excluded. Person-time at risk commenced on the day after surgery for the overall analysis; therefore VTEs recorded on the same day as the operation were not included in the analysis. Patients were followed up until they developed a VTE event, died, left a participating general practice, or for 1 year after surgery. Patients were excluded if they were: not in a linked general practice or had had a VTE before the admission date for colectomy, as these patients have an inherently increased risk of postoperative VTE. Previous VTE was defined using the same definition as for the entire study. Exposures Patients were identified who had a colorectal cancer diagnosis in the CPRD and HES data (ICD-10 sections C18–20, excluding C18·1 Appendix). The non-malignant diagnoses were confirmed from the ICD-10 discharge codes associated with the admission, including inflammatory bowel disease (IBD) (Appendix S1, supporting information), diverticular disease (Appendix S1) and other. Co-morbidity was determined from the CPRD and classified using the Charlson index before admission for surgery19. Admission type was defined as elective or emergency, based on the type of admission recorded for the surgical procedure. Smoking status was classified as never, ever, current or missing, based on CPRD smoking records before the operation. Body mass index (BMI) was recorded from CPRD records, and patients were classified as having a BMI of less than 30 kg/m2, 30 kg/m2 or more, or missing data. Laparoscopic surgery was defined as any surgery accompanied by a laparoscopic code. Prescriptions for the oral contraceptive pill and hormone replacement therapy (HRT) were identified from the CPRD, and users were classified as current if the prescription was within 3 months of the index operation. Outcome definition VTE diagnosis was determined from medical codes in the CPRD and HES. These were considered to indicate a valid VTE event if supported by either: a prescription for an anticoagulant or other evidence of treatment in an anticoagulation clinic (such as a medical code) between 15 days before and 90 days after the VTE diagnosis, or a date of death within 30 days of the event. Additionally, an underlying cause of death of VTE was included as evidence of a VTE diagnosis. The definition using primary care data alone has been validated previously20 and used in the authors' previous studies of VTE21,22. Statistical analysis The date of diagnosis of VTE was taken to be the episode start date for VTEs occurring within the same hospital admission as the index operation. Absolute rates of VTE (per 1000 person-years) were calculated by dividing the number of individuals with VTE by the person-time at risk for the first year of follow-up after operation. This was done overall and then separately for each exposure of interest. Patients with missing data were analysed as a separate category, and rates and risk were presented separately for them. A Cox proportional hazards model was then created to include all exposures, to estimate hazard ratios (HRs) and 95 per cent c.i. The risk of VTE was then evaluated by indication for surgery and type of admission, adjusting for age and co-morbidity. A female-only analysis was performed in order to account for current use (within 3 months of surgery) of the oral contraceptive pill or HRT, given their known association with an increased risk of VTE23. Age was fitted as an interaction term in the models using likelihood ratio tests to look for the presence of a significant interaction. A further analysis was undertaken of the rates of VTE by calendar month after surgery in order to define the period of greatest risk up to 12 weeks. A sensitivity analysis was performed of all patients from the date of hospital discharge, excluding patients who had an in-hospital VTE, to determine the rates of VTE after discharge. All data management and analyses were performed using Stata® version 12 (StataCorp, College Station, Texas, USA). Results In total, 12 388 patients had a colectomy and 312 (2·5 per cent) developed VTE in the first year after surgery. The proportion of men and women undergoing emergency surgery was similar; however, in the elective setting the proportion of women who had surgery for non-malignant was greater than that for malignant disease (56·4 versus 45·1 per cent; P < 0·001, χ2 test) (Table 1). Patients undergoing surgery for non-malignant disease were younger than those undergoing surgery for malignant disease (median (i.q.r.) 62 (46–74) versus 72 (64–79) years respectively; P = 0·001, Kruskal–Wallis test). Patients undergoing emergency or elective surgery for malignant disease were more likely to be ever or current smokers than those with non-malignant disease (Table 1) (P < 0·001, χ2 test). In total, 12·5 per cent of the patients (1547 of 12 388) had a laparoscopic procedure during the study interval, increasing to 33·7 per cent (302 of 895) in the final year of the study. Table 1 Demographics of colectomy cohort, by admission type and indication for surgery . Emergency (n = 3938) . Elective (n = 8450) . . Non-malignant (n = 2188) . Malignant (n = 1750) . Non-malignant (n = 1891) . Malignant (n = 6559) . Age (years) < 40 389 (17·8) 33 (1·9) 362 (19·1) 57 (0·9) 40–49 218 (10·0) 72 (4·1) 218 (11·5) 212 (3·2) 50–59 310 (14·2) 177 (10·1) 318 (16·8) 767 (11·7) 60–69 409 (18·7) 356 (20·3) 459 (24·3) 1774 (27·0) 70–79 502 (22·9) 579 (33·1) 373 (19·7) 2329 (35·5) 80–89 360 (16·5) 533 (30·5) 161 (8·5) 1420 (21·6) Sex M 985 (45·0) 825 (47·1) 824 (43·6) 3598 (54·9) F 1203 (55·0) 925 (52·9) 1067 (56·4) 2961 (45·1) No. of co-morbidities 0 1120 (51·2) 739 (42·2) 1058 (55·9) 2136 (32·6) 1 466 (21·3) 327 (18·7) 401 (21·2) 880 (13·4) 2 285 (13·0) 323 (18·5) 240 (12·7) 1848 (28·2) ≥ 3 317 (14·5) 361 (20·6) 192 (10·2) 1695 (25·8) Body mass index (kg/m2) < 30 1356 (62·0) 1140 (65·1) 1253 (66·3) 4370 (66·6) ≥ 30 377 (17·2) 300 (17·1) 354 (18·7) 1370 (20·9) Missing 455 (20·8) 310 (17·7) 284 (15·0) 819 (12·5) Smoking status Never 517 (23·6) 274 (15·7) 405 (21·4) 730 (11·1) Ever 887 (40·5) 861 (49·2) 853 (45·1) 3322 (50·7) Current 563 (25·7) 487 (27·8) 523 (27·7) 2209 (33·7) Missing 221 (10·1) 128 (7·3) 110 (5·8) 298 (4·5) . Emergency (n = 3938) . Elective (n = 8450) . . Non-malignant (n = 2188) . Malignant (n = 1750) . Non-malignant (n = 1891) . Malignant (n = 6559) . Age (years) < 40 389 (17·8) 33 (1·9) 362 (19·1) 57 (0·9) 40–49 218 (10·0) 72 (4·1) 218 (11·5) 212 (3·2) 50–59 310 (14·2) 177 (10·1) 318 (16·8) 767 (11·7) 60–69 409 (18·7) 356 (20·3) 459 (24·3) 1774 (27·0) 70–79 502 (22·9) 579 (33·1) 373 (19·7) 2329 (35·5) 80–89 360 (16·5) 533 (30·5) 161 (8·5) 1420 (21·6) Sex M 985 (45·0) 825 (47·1) 824 (43·6) 3598 (54·9) F 1203 (55·0) 925 (52·9) 1067 (56·4) 2961 (45·1) No. of co-morbidities 0 1120 (51·2) 739 (42·2) 1058 (55·9) 2136 (32·6) 1 466 (21·3) 327 (18·7) 401 (21·2) 880 (13·4) 2 285 (13·0) 323 (18·5) 240 (12·7) 1848 (28·2) ≥ 3 317 (14·5) 361 (20·6) 192 (10·2) 1695 (25·8) Body mass index (kg/m2) < 30 1356 (62·0) 1140 (65·1) 1253 (66·3) 4370 (66·6) ≥ 30 377 (17·2) 300 (17·1) 354 (18·7) 1370 (20·9) Missing 455 (20·8) 310 (17·7) 284 (15·0) 819 (12·5) Smoking status Never 517 (23·6) 274 (15·7) 405 (21·4) 730 (11·1) Ever 887 (40·5) 861 (49·2) 853 (45·1) 3322 (50·7) Current 563 (25·7) 487 (27·8) 523 (27·7) 2209 (33·7) Missing 221 (10·1) 128 (7·3) 110 (5·8) 298 (4·5) Values in parentheses are percentages. Open in new tab Table 1 Demographics of colectomy cohort, by admission type and indication for surgery . Emergency (n = 3938) . Elective (n = 8450) . . Non-malignant (n = 2188) . Malignant (n = 1750) . Non-malignant (n = 1891) . Malignant (n = 6559) . Age (years) < 40 389 (17·8) 33 (1·9) 362 (19·1) 57 (0·9) 40–49 218 (10·0) 72 (4·1) 218 (11·5) 212 (3·2) 50–59 310 (14·2) 177 (10·1) 318 (16·8) 767 (11·7) 60–69 409 (18·7) 356 (20·3) 459 (24·3) 1774 (27·0) 70–79 502 (22·9) 579 (33·1) 373 (19·7) 2329 (35·5) 80–89 360 (16·5) 533 (30·5) 161 (8·5) 1420 (21·6) Sex M 985 (45·0) 825 (47·1) 824 (43·6) 3598 (54·9) F 1203 (55·0) 925 (52·9) 1067 (56·4) 2961 (45·1) No. of co-morbidities 0 1120 (51·2) 739 (42·2) 1058 (55·9) 2136 (32·6) 1 466 (21·3) 327 (18·7) 401 (21·2) 880 (13·4) 2 285 (13·0) 323 (18·5) 240 (12·7) 1848 (28·2) ≥ 3 317 (14·5) 361 (20·6) 192 (10·2) 1695 (25·8) Body mass index (kg/m2) < 30 1356 (62·0) 1140 (65·1) 1253 (66·3) 4370 (66·6) ≥ 30 377 (17·2) 300 (17·1) 354 (18·7) 1370 (20·9) Missing 455 (20·8) 310 (17·7) 284 (15·0) 819 (12·5) Smoking status Never 517 (23·6) 274 (15·7) 405 (21·4) 730 (11·1) Ever 887 (40·5) 861 (49·2) 853 (45·1) 3322 (50·7) Current 563 (25·7) 487 (27·8) 523 (27·7) 2209 (33·7) Missing 221 (10·1) 128 (7·3) 110 (5·8) 298 (4·5) . Emergency (n = 3938) . Elective (n = 8450) . . Non-malignant (n = 2188) . Malignant (n = 1750) . Non-malignant (n = 1891) . Malignant (n = 6559) . Age (years) < 40 389 (17·8) 33 (1·9) 362 (19·1) 57 (0·9) 40–49 218 (10·0) 72 (4·1) 218 (11·5) 212 (3·2) 50–59 310 (14·2) 177 (10·1) 318 (16·8) 767 (11·7) 60–69 409 (18·7) 356 (20·3) 459 (24·3) 1774 (27·0) 70–79 502 (22·9) 579 (33·1) 373 (19·7) 2329 (35·5) 80–89 360 (16·5) 533 (30·5) 161 (8·5) 1420 (21·6) Sex M 985 (45·0) 825 (47·1) 824 (43·6) 3598 (54·9) F 1203 (55·0) 925 (52·9) 1067 (56·4) 2961 (45·1) No. of co-morbidities 0 1120 (51·2) 739 (42·2) 1058 (55·9) 2136 (32·6) 1 466 (21·3) 327 (18·7) 401 (21·2) 880 (13·4) 2 285 (13·0) 323 (18·5) 240 (12·7) 1848 (28·2) ≥ 3 317 (14·5) 361 (20·6) 192 (10·2) 1695 (25·8) Body mass index (kg/m2) < 30 1356 (62·0) 1140 (65·1) 1253 (66·3) 4370 (66·6) ≥ 30 377 (17·2) 300 (17·1) 354 (18·7) 1370 (20·9) Missing 455 (20·8) 310 (17·7) 284 (15·0) 819 (12·5) Smoking status Never 517 (23·6) 274 (15·7) 405 (21·4) 730 (11·1) Ever 887 (40·5) 861 (49·2) 853 (45·1) 3322 (50·7) Current 563 (25·7) 487 (27·8) 523 (27·7) 2209 (33·7) Missing 221 (10·1) 128 (7·3) 110 (5·8) 298 (4·5) Values in parentheses are percentages. Open in new tab In the emergency setting, 1750 (44·4 per cent) of 3938 patients had surgery for malignant disease; of those undergoing surgery for non-malignant disease, 502 (22·9 per cent) of 2188 were for IBD and 651 (30·0 per cent) of 2188 were for diverticular disease, with the remainder for other indications. In the elective setting 6559 (77·6 per cent) of 8450 patients underwent surgery for malignant disease; of those having surgery for non-malignant disease, 527 procedures (27·9 per cent) were for IBD and 579 (30·6 per cent) for diverticular disease, with the remainder for other indications. Venous thromboembolism rates overall, by admission type and indication The overall rate of VTE following colectomy for any indication was 48·08 (95 per cent c.i. 41·57 to 55·62) and 29·59 (26·48 to 33·06) per 1000 person-years at 6 months and 1 year respectively. The rate of VTE was highest in the first month after emergency surgery, and declined from 121·68 per 1000 person-years in the first month to 25·65 per 1000 person-years during the rest of the follow-up interval. The overall rate was 2·2-fold higher (adjusted HR 2·23, 95 per cent c.i. 1·76 to 2·50) for emergency compared with elective admissions (crude rate 39·44 versus 25·78 per 1000 person-years respectively) in the year after surgery. For malignant disease, rates of VTE were 2·8-fold higher (adjusted HR 2·84, 2·04 to 3·94) than for non-malignant disease (36·85 versus 14·56 per 1000 person-years respectively). When patients in the non-malignant group without a clear indication for surgery were excluded, there was little difference in the absolute rates (14·56 versus 10·94 per 1000 person-years, including and excluding patients without a clear indication respectively). Laparoscopic surgery was not associated with a reduction in risk of VTE in the first year after surgery (unadjusted HR 0·91, 95 per cent c.i. 0·65 to 1·28). For women, the overall rate of VTE was 2·3-fold higher (adjusted HR 2·23, 1·59 to 3·13) for emergency compared with elective admissions in the year after surgery. When further adjustment was made for current use of either HRT or the oral contraceptive pill this association did not change (adjusted HR 2·24, 1·60 to 3·14); thus, no adjustment for these variables was made in subsequent analyses. Overall rates of venous thromboembolism Elective colectomy Patients who had elective colectomy for non-malignant disease had the lowest crude rate of VTE (6·31 per 1000 person-years) (Table 2). The crude rate of VTE following elective colectomy increased with increasing age, peaking in those aged over 60 years (36·33 per 1000 person-years). The risk of VTE increased with increasing age, although the effects of age were attenuated when accounting for indication, suggesting an interaction between age and indication (P = 0·005, likelihood ratio test). When stratified by indication, the risk of VTE by age increased for patients with a non-malignant indication, peaking in those over 80 years old (crude rate 20·80 per 1000 person-years), but there was no increase in patients with a malignant indication. Table 2 Rates of venous thromboembolism for elective admission, by indication, age, co-morbidity using the Charlson index, body mass index and smoking status . Rate per 1000 person-years . Unadjusted HR . Adjusted HR* . Indication Non-malignant 6·31 (3·49, 11·37) 1·00 (reference) 1·00 (reference) Malignant 31·59 (27·35, 36·49) 4·03 (2·20, 7·41) 4·16 (2·24, 7·72) Age (years) < 40 – – – 40–49 9·79 (3·67, 26·07) 1·00 (reference) 1·00 (reference) 50–59 21·85 (14·39, 33·18) 2·22 (0·77, 6·45) 1·76 (0·61, 5·13) 60–69 36·33 (28·93, 45·63) 3·69 (1·35, 10·08) 2·67 (0·97, 7·37) 70–79 27·91 (21·96, 35·45) 2·02 (1·03, 7·75) 1·98 (0·72, 5·50) ≥ 80 21·37 (14·85, 30·75) 1·44 (0·76, 6·12) 1·56 (0·54, 4·50) No. of co-morbidities 0 25·08 (19·94, 31·55) 1·00 (reference) 1·00 (reference) 1 27·76 (19·63, 39·25) 1·06 (0·70, 1·61) 1·04 (0·68, 1·59) 2 26·48 (20·07, 34·93) 0·96 (0·67, 1·38) 0·81 (0·56, 1·16) 3 24·83 (18·29, 33·73) 0·89 (0·61, 1·30) 0·72 (0·48, 1·07) Body mass index (kg/m2) < 30 21·45 (17·78, 25·88) 1·00 (reference) 1·00 (reference) ≥ 30 41·49 (32·48, 53·01) 1·88 (1·38, 2·56) 1·84 (1·34, 2·51) Missing 23·49 (15·61, 35·34) 1·17 (0·74, 1·83) 1·20 (0·73, 1·95) Smoking status Never 18·84 (12·02, 29·54) 1·00 (reference) 1·00 (reference) Ever 26·72 (21·99, 32·47) 1·36 (0·83, 2·22) 1·19 (0·73, 1·96) Current 28·2 (22·27, 35·70) 1·39 (0·84, 2·31) 1·18 (0·71, 1·98) Missing 19·04 (9·08, 39·93) 1·08 (0·45, 2·56) 0·92 (0·37, 2·32) . Rate per 1000 person-years . Unadjusted HR . Adjusted HR* . Indication Non-malignant 6·31 (3·49, 11·37) 1·00 (reference) 1·00 (reference) Malignant 31·59 (27·35, 36·49) 4·03 (2·20, 7·41) 4·16 (2·24, 7·72) Age (years) < 40 – – – 40–49 9·79 (3·67, 26·07) 1·00 (reference) 1·00 (reference) 50–59 21·85 (14·39, 33·18) 2·22 (0·77, 6·45) 1·76 (0·61, 5·13) 60–69 36·33 (28·93, 45·63) 3·69 (1·35, 10·08) 2·67 (0·97, 7·37) 70–79 27·91 (21·96, 35·45) 2·02 (1·03, 7·75) 1·98 (0·72, 5·50) ≥ 80 21·37 (14·85, 30·75) 1·44 (0·76, 6·12) 1·56 (0·54, 4·50) No. of co-morbidities 0 25·08 (19·94, 31·55) 1·00 (reference) 1·00 (reference) 1 27·76 (19·63, 39·25) 1·06 (0·70, 1·61) 1·04 (0·68, 1·59) 2 26·48 (20·07, 34·93) 0·96 (0·67, 1·38) 0·81 (0·56, 1·16) 3 24·83 (18·29, 33·73) 0·89 (0·61, 1·30) 0·72 (0·48, 1·07) Body mass index (kg/m2) < 30 21·45 (17·78, 25·88) 1·00 (reference) 1·00 (reference) ≥ 30 41·49 (32·48, 53·01) 1·88 (1·38, 2·56) 1·84 (1·34, 2·51) Missing 23·49 (15·61, 35·34) 1·17 (0·74, 1·83) 1·20 (0·73, 1·95) Smoking status Never 18·84 (12·02, 29·54) 1·00 (reference) 1·00 (reference) Ever 26·72 (21·99, 32·47) 1·36 (0·83, 2·22) 1·19 (0·73, 1·96) Current 28·2 (22·27, 35·70) 1·39 (0·84, 2·31) 1·18 (0·71, 1·98) Missing 19·04 (9·08, 39·93) 1·08 (0·45, 2·56) 0·92 (0·37, 2·32) Values in parentheses are 95 per cent c.i. * Adjusted for all other variables in the table. HR, hazard ratio. Open in new tab Table 2 Rates of venous thromboembolism for elective admission, by indication, age, co-morbidity using the Charlson index, body mass index and smoking status . Rate per 1000 person-years . Unadjusted HR . Adjusted HR* . Indication Non-malignant 6·31 (3·49, 11·37) 1·00 (reference) 1·00 (reference) Malignant 31·59 (27·35, 36·49) 4·03 (2·20, 7·41) 4·16 (2·24, 7·72) Age (years) < 40 – – – 40–49 9·79 (3·67, 26·07) 1·00 (reference) 1·00 (reference) 50–59 21·85 (14·39, 33·18) 2·22 (0·77, 6·45) 1·76 (0·61, 5·13) 60–69 36·33 (28·93, 45·63) 3·69 (1·35, 10·08) 2·67 (0·97, 7·37) 70–79 27·91 (21·96, 35·45) 2·02 (1·03, 7·75) 1·98 (0·72, 5·50) ≥ 80 21·37 (14·85, 30·75) 1·44 (0·76, 6·12) 1·56 (0·54, 4·50) No. of co-morbidities 0 25·08 (19·94, 31·55) 1·00 (reference) 1·00 (reference) 1 27·76 (19·63, 39·25) 1·06 (0·70, 1·61) 1·04 (0·68, 1·59) 2 26·48 (20·07, 34·93) 0·96 (0·67, 1·38) 0·81 (0·56, 1·16) 3 24·83 (18·29, 33·73) 0·89 (0·61, 1·30) 0·72 (0·48, 1·07) Body mass index (kg/m2) < 30 21·45 (17·78, 25·88) 1·00 (reference) 1·00 (reference) ≥ 30 41·49 (32·48, 53·01) 1·88 (1·38, 2·56) 1·84 (1·34, 2·51) Missing 23·49 (15·61, 35·34) 1·17 (0·74, 1·83) 1·20 (0·73, 1·95) Smoking status Never 18·84 (12·02, 29·54) 1·00 (reference) 1·00 (reference) Ever 26·72 (21·99, 32·47) 1·36 (0·83, 2·22) 1·19 (0·73, 1·96) Current 28·2 (22·27, 35·70) 1·39 (0·84, 2·31) 1·18 (0·71, 1·98) Missing 19·04 (9·08, 39·93) 1·08 (0·45, 2·56) 0·92 (0·37, 2·32) . Rate per 1000 person-years . Unadjusted HR . Adjusted HR* . Indication Non-malignant 6·31 (3·49, 11·37) 1·00 (reference) 1·00 (reference) Malignant 31·59 (27·35, 36·49) 4·03 (2·20, 7·41) 4·16 (2·24, 7·72) Age (years) < 40 – – – 40–49 9·79 (3·67, 26·07) 1·00 (reference) 1·00 (reference) 50–59 21·85 (14·39, 33·18) 2·22 (0·77, 6·45) 1·76 (0·61, 5·13) 60–69 36·33 (28·93, 45·63) 3·69 (1·35, 10·08) 2·67 (0·97, 7·37) 70–79 27·91 (21·96, 35·45) 2·02 (1·03, 7·75) 1·98 (0·72, 5·50) ≥ 80 21·37 (14·85, 30·75) 1·44 (0·76, 6·12) 1·56 (0·54, 4·50) No. of co-morbidities 0 25·08 (19·94, 31·55) 1·00 (reference) 1·00 (reference) 1 27·76 (19·63, 39·25) 1·06 (0·70, 1·61) 1·04 (0·68, 1·59) 2 26·48 (20·07, 34·93) 0·96 (0·67, 1·38) 0·81 (0·56, 1·16) 3 24·83 (18·29, 33·73) 0·89 (0·61, 1·30) 0·72 (0·48, 1·07) Body mass index (kg/m2) < 30 21·45 (17·78, 25·88) 1·00 (reference) 1·00 (reference) ≥ 30 41·49 (32·48, 53·01) 1·88 (1·38, 2·56) 1·84 (1·34, 2·51) Missing 23·49 (15·61, 35·34) 1·17 (0·74, 1·83) 1·20 (0·73, 1·95) Smoking status Never 18·84 (12·02, 29·54) 1·00 (reference) 1·00 (reference) Ever 26·72 (21·99, 32·47) 1·36 (0·83, 2·22) 1·19 (0·73, 1·96) Current 28·2 (22·27, 35·70) 1·39 (0·84, 2·31) 1·18 (0·71, 1·98) Missing 19·04 (9·08, 39·93) 1·08 (0·45, 2·56) 0·92 (0·37, 2·32) Values in parentheses are 95 per cent c.i. * Adjusted for all other variables in the table. HR, hazard ratio. Open in new tab Emergency colectomy Patients undergoing an emergency colectomy for malignant disease had a 2·4-fold increase in risk of VTE in the year following colectomy compared with those with a non-malignant indication (Table 3). The risk of VTE increased with increasing age, but the effects of age were attenuated when accounting for indication, suggesting an interaction between age and indication (P < 0·001, likelihood ratio test). When stratified by indication, the risk of VTE by age increased for patients with a non-malignant indication, peaking in those over the age of 60 years (crude rate 37·62 per 1000 person-years), but there was no increase in patients with a malignant indication. Table 3 Rates of venous thromboembolism for emergency admission, by indication, age, co-morbidity using the Charlson index, body mass index and smoking status . Rate per 1000 person-years . Unadjusted HR . Adjusted HR* . Indication Non-malignant 23·1 (16·88, 31·62) 1·00 (reference) 1·00 (reference) Malignant 61·44 (49·14, 76·82) 2·59 (1·76, 3·80) 2·42 (1·61, 3·65) Age (years) < 40 16·04 (7·21, 35·71) 1·00 (reference) 1·00 (reference) 40–49 30·44 (15·22, 60·87) 1·91 (0·66, 5·51) 1·48 (0·51, 4·30) 50–59 24·3 (13·07, 45·15) 1·51 (0·55, 4·15) 1·03 (0·37, 2·88) 60–69 54·98 (39·09, 77·34) 3·35 (1·40, 8·00) 2·06 (0·83, 5·09) 70–79 60·19 (45·08, 80·36) 3·61 (1·54, 8·46) 2·14 (0·87, 5·28) ≥ 80 24·6 (14·28, 42·36) 1·44 (0·55, 3·79) 0·83 (0·30, 2·32) No. of co-morbidities 0 42·13 (32·85, 54·04) 1·00 (reference) 1·00 (reference) 1 33·36 (21·52, 51·71) 0·79 (0·48, 1·31) 0·76 (0·45, 1·27) 2 29·17 (16·94, 50·23) 0·68 (0·38, 1·24) 0·52 (0·28, 0·95) ≥ 3 49·45 (32·24, 75·84) 1·12 (0·68, 1·84) 0·85 (0·51, 1·43) Body mass index (kg/m2) < 30 36·84 (29·10, 46·64) 1·00 (reference) 1·00 (reference) ≥ 30 64·71 (46·00, 91·02) 1·75 (1·15, 2·65) 1·70 (1·11, 2·60) Missing 25·07 (14·85, 42·33) 0·67 (0·38, 1·19) 0·83 (0·43, 1·62) Smoking status Never 28·51 (17·72, 45·85) 1·00 (reference) 1·00 (reference) Ever 46·16 (35·92, 59·33) 1·60 (0·94, 2·75) 1·39 (0·80, 2·41) Current 42·89 (30·49, 47·27) 1·49 (0·83, 2·68) 1·24 (0·68, 2·26) Missing 19·67 (8·19, 47·27) 0·68 (0·25, 1·83) 0·83 (0·27, 2·55) . Rate per 1000 person-years . Unadjusted HR . Adjusted HR* . Indication Non-malignant 23·1 (16·88, 31·62) 1·00 (reference) 1·00 (reference) Malignant 61·44 (49·14, 76·82) 2·59 (1·76, 3·80) 2·42 (1·61, 3·65) Age (years) < 40 16·04 (7·21, 35·71) 1·00 (reference) 1·00 (reference) 40–49 30·44 (15·22, 60·87) 1·91 (0·66, 5·51) 1·48 (0·51, 4·30) 50–59 24·3 (13·07, 45·15) 1·51 (0·55, 4·15) 1·03 (0·37, 2·88) 60–69 54·98 (39·09, 77·34) 3·35 (1·40, 8·00) 2·06 (0·83, 5·09) 70–79 60·19 (45·08, 80·36) 3·61 (1·54, 8·46) 2·14 (0·87, 5·28) ≥ 80 24·6 (14·28, 42·36) 1·44 (0·55, 3·79) 0·83 (0·30, 2·32) No. of co-morbidities 0 42·13 (32·85, 54·04) 1·00 (reference) 1·00 (reference) 1 33·36 (21·52, 51·71) 0·79 (0·48, 1·31) 0·76 (0·45, 1·27) 2 29·17 (16·94, 50·23) 0·68 (0·38, 1·24) 0·52 (0·28, 0·95) ≥ 3 49·45 (32·24, 75·84) 1·12 (0·68, 1·84) 0·85 (0·51, 1·43) Body mass index (kg/m2) < 30 36·84 (29·10, 46·64) 1·00 (reference) 1·00 (reference) ≥ 30 64·71 (46·00, 91·02) 1·75 (1·15, 2·65) 1·70 (1·11, 2·60) Missing 25·07 (14·85, 42·33) 0·67 (0·38, 1·19) 0·83 (0·43, 1·62) Smoking status Never 28·51 (17·72, 45·85) 1·00 (reference) 1·00 (reference) Ever 46·16 (35·92, 59·33) 1·60 (0·94, 2·75) 1·39 (0·80, 2·41) Current 42·89 (30·49, 47·27) 1·49 (0·83, 2·68) 1·24 (0·68, 2·26) Missing 19·67 (8·19, 47·27) 0·68 (0·25, 1·83) 0·83 (0·27, 2·55) Values in parentheses are 95 per cent c.i. * Adjusted for all other variables in the table. HR, hazard ratio. Open in new tab Table 3 Rates of venous thromboembolism for emergency admission, by indication, age, co-morbidity using the Charlson index, body mass index and smoking status . Rate per 1000 person-years . Unadjusted HR . Adjusted HR* . Indication Non-malignant 23·1 (16·88, 31·62) 1·00 (reference) 1·00 (reference) Malignant 61·44 (49·14, 76·82) 2·59 (1·76, 3·80) 2·42 (1·61, 3·65) Age (years) < 40 16·04 (7·21, 35·71) 1·00 (reference) 1·00 (reference) 40–49 30·44 (15·22, 60·87) 1·91 (0·66, 5·51) 1·48 (0·51, 4·30) 50–59 24·3 (13·07, 45·15) 1·51 (0·55, 4·15) 1·03 (0·37, 2·88) 60–69 54·98 (39·09, 77·34) 3·35 (1·40, 8·00) 2·06 (0·83, 5·09) 70–79 60·19 (45·08, 80·36) 3·61 (1·54, 8·46) 2·14 (0·87, 5·28) ≥ 80 24·6 (14·28, 42·36) 1·44 (0·55, 3·79) 0·83 (0·30, 2·32) No. of co-morbidities 0 42·13 (32·85, 54·04) 1·00 (reference) 1·00 (reference) 1 33·36 (21·52, 51·71) 0·79 (0·48, 1·31) 0·76 (0·45, 1·27) 2 29·17 (16·94, 50·23) 0·68 (0·38, 1·24) 0·52 (0·28, 0·95) ≥ 3 49·45 (32·24, 75·84) 1·12 (0·68, 1·84) 0·85 (0·51, 1·43) Body mass index (kg/m2) < 30 36·84 (29·10, 46·64) 1·00 (reference) 1·00 (reference) ≥ 30 64·71 (46·00, 91·02) 1·75 (1·15, 2·65) 1·70 (1·11, 2·60) Missing 25·07 (14·85, 42·33) 0·67 (0·38, 1·19) 0·83 (0·43, 1·62) Smoking status Never 28·51 (17·72, 45·85) 1·00 (reference) 1·00 (reference) Ever 46·16 (35·92, 59·33) 1·60 (0·94, 2·75) 1·39 (0·80, 2·41) Current 42·89 (30·49, 47·27) 1·49 (0·83, 2·68) 1·24 (0·68, 2·26) Missing 19·67 (8·19, 47·27) 0·68 (0·25, 1·83) 0·83 (0·27, 2·55) . Rate per 1000 person-years . Unadjusted HR . Adjusted HR* . Indication Non-malignant 23·1 (16·88, 31·62) 1·00 (reference) 1·00 (reference) Malignant 61·44 (49·14, 76·82) 2·59 (1·76, 3·80) 2·42 (1·61, 3·65) Age (years) < 40 16·04 (7·21, 35·71) 1·00 (reference) 1·00 (reference) 40–49 30·44 (15·22, 60·87) 1·91 (0·66, 5·51) 1·48 (0·51, 4·30) 50–59 24·3 (13·07, 45·15) 1·51 (0·55, 4·15) 1·03 (0·37, 2·88) 60–69 54·98 (39·09, 77·34) 3·35 (1·40, 8·00) 2·06 (0·83, 5·09) 70–79 60·19 (45·08, 80·36) 3·61 (1·54, 8·46) 2·14 (0·87, 5·28) ≥ 80 24·6 (14·28, 42·36) 1·44 (0·55, 3·79) 0·83 (0·30, 2·32) No. of co-morbidities 0 42·13 (32·85, 54·04) 1·00 (reference) 1·00 (reference) 1 33·36 (21·52, 51·71) 0·79 (0·48, 1·31) 0·76 (0·45, 1·27) 2 29·17 (16·94, 50·23) 0·68 (0·38, 1·24) 0·52 (0·28, 0·95) ≥ 3 49·45 (32·24, 75·84) 1·12 (0·68, 1·84) 0·85 (0·51, 1·43) Body mass index (kg/m2) < 30 36·84 (29·10, 46·64) 1·00 (reference) 1·00 (reference) ≥ 30 64·71 (46·00, 91·02) 1·75 (1·15, 2·65) 1·70 (1·11, 2·60) Missing 25·07 (14·85, 42·33) 0·67 (0·38, 1·19) 0·83 (0·43, 1·62) Smoking status Never 28·51 (17·72, 45·85) 1·00 (reference) 1·00 (reference) Ever 46·16 (35·92, 59·33) 1·60 (0·94, 2·75) 1·39 (0·80, 2·41) Current 42·89 (30·49, 47·27) 1·49 (0·83, 2·68) 1·24 (0·68, 2·26) Missing 19·67 (8·19, 47·27) 0·68 (0·25, 1·83) 0·83 (0·27, 2·55) Values in parentheses are 95 per cent c.i. * Adjusted for all other variables in the table. HR, hazard ratio. Open in new tab Rates in first month after surgery Elective surgery Patients who had an elective colectomy for malignant disease had a 3·8-fold increased risk of VTE in the first month after surgery versus patients with non-malignant disease (HR 3·81, 95 per cent c.i. 0·90 to 15·99), although this was not significant (Table 4). The absolute rate of VTE in all patients fell after the first postoperative month (Fig. 1). Table 4 Rates of venous thromboembolism in the first month after surgery, by admission type and indication . Rate (per 1000 person-years) . Adjusted HR* . . First month after surgery . Time from first month to 1 year . Elective Non-malignant 12·98 (3·25, 51·89) 5·65 (2·94, 10·86) 1·00 (reference) Malignant 61·83 (43·96, 86·98) 28·56 (24·36, 33·48) 3·81 (0·90, 15·99) Emergency Non-malignant 114·76 (73·20, 179·91) 13·13 (8·47, 20·36) 1·00 (reference) Malignant 120·98 (74·12, 197·48) 54·42 (42·34, 69·94) 1·12 (0·56, 2·27) . Rate (per 1000 person-years) . Adjusted HR* . . First month after surgery . Time from first month to 1 year . Elective Non-malignant 12·98 (3·25, 51·89) 5·65 (2·94, 10·86) 1·00 (reference) Malignant 61·83 (43·96, 86·98) 28·56 (24·36, 33·48) 3·81 (0·90, 15·99) Emergency Non-malignant 114·76 (73·20, 179·91) 13·13 (8·47, 20·36) 1·00 (reference) Malignant 120·98 (74·12, 197·48) 54·42 (42·34, 69·94) 1·12 (0·56, 2·27) Values in parentheses are 95 per cent c.i. * Cox regression analysis adjusted for age and sex. HR, hazard ratio. Open in new tab Table 4 Rates of venous thromboembolism in the first month after surgery, by admission type and indication . Rate (per 1000 person-years) . Adjusted HR* . . First month after surgery . Time from first month to 1 year . Elective Non-malignant 12·98 (3·25, 51·89) 5·65 (2·94, 10·86) 1·00 (reference) Malignant 61·83 (43·96, 86·98) 28·56 (24·36, 33·48) 3·81 (0·90, 15·99) Emergency Non-malignant 114·76 (73·20, 179·91) 13·13 (8·47, 20·36) 1·00 (reference) Malignant 120·98 (74·12, 197·48) 54·42 (42·34, 69·94) 1·12 (0·56, 2·27) . Rate (per 1000 person-years) . Adjusted HR* . . First month after surgery . Time from first month to 1 year . Elective Non-malignant 12·98 (3·25, 51·89) 5·65 (2·94, 10·86) 1·00 (reference) Malignant 61·83 (43·96, 86·98) 28·56 (24·36, 33·48) 3·81 (0·90, 15·99) Emergency Non-malignant 114·76 (73·20, 179·91) 13·13 (8·47, 20·36) 1·00 (reference) Malignant 120·98 (74·12, 197·48) 54·42 (42·34, 69·94) 1·12 (0·56, 2·27) Values in parentheses are 95 per cent c.i. * Cox regression analysis adjusted for age and sex. HR, hazard ratio. Open in new tab Fig. 1 Open in new tabDownload slide Cumulative incidence of venous thromboembolism (VTE) following elective admission by indication for surgery Emergency surgery Patients undergoing emergency colectomy had a similar cumulative incidence of VTE in the first month after surgery regardless of indication, with a higher absolute rate in those with a malignant indication: 120·98 per 1000 person-years versus 114·76 per 1000 person-years in those with non-malignant disease (HR 1·12, 95 per cent c.i. 0·56 to 2·27) (Table 4 and Fig. 2). Fig. 2 Open in new tabDownload slide Cumulative incidence of venous thromboembolism (VTE) following emergency admission by indication for surgery The rate of VTE following a non-malignant emergency colectomy was 2·0-fold greater than that for an elective colectomy with a malignant indication in the first month after surgery (absolute rate 114·76 versus 61·83 per 1000 person-years respectively; HR 2·04, 95 per cent c.i. 1·13 to 3·74). Following the first month after surgery, the absolute rate of VTE in all patients fell during the remainder of the year but remained highest in those undergoing emergency surgery for malignant disease (54·42 per 1000 person-years) (Table 4). Sensitivity analysis after discharge from hospital Elective surgery In the first month after hospital discharge the absolute rates of VTE were similar to those in the first month after surgery (non-malignant: 13·18 versus 12·98 per 1000 person-years; malignant: 68·76 versus 61·83 per 1000 person-years) (Tables 4 and 5). Emergency surgery In the first month after discharge from hospital the absolute rate of VTE was lower following emergency admission for non-malignant disease compared with that in the first month after surgery (78·01 versus 114·76 per 1000 person-years). A reduction was also seen in the month following discharge in patients with a malignant indication: 106·38 per 1000 person-years versus 120·98 per 1000 person-years in the first month after surgery (Tables 4 and 5). Table 5 Rates of venous thromboembolism in the first month after discharge from hospital, by admission type and indication . Rate in first month after hospital discharge (per 1000 person-years) . Elective Non-malignant 13·18 (3·29, 52·71) Malignant 68·76 (49·61, 95·33) Emergency Non-malignant 78·01 (44·31, 137·36) Malignant 106·38 (61·77, 183·21) . Rate in first month after hospital discharge (per 1000 person-years) . Elective Non-malignant 13·18 (3·29, 52·71) Malignant 68·76 (49·61, 95·33) Emergency Non-malignant 78·01 (44·31, 137·36) Malignant 106·38 (61·77, 183·21) Values in parentheses are 95 per cent c.i. Open in new tab Table 5 Rates of venous thromboembolism in the first month after discharge from hospital, by admission type and indication . Rate in first month after hospital discharge (per 1000 person-years) . Elective Non-malignant 13·18 (3·29, 52·71) Malignant 68·76 (49·61, 95·33) Emergency Non-malignant 78·01 (44·31, 137·36) Malignant 106·38 (61·77, 183·21) . Rate in first month after hospital discharge (per 1000 person-years) . Elective Non-malignant 13·18 (3·29, 52·71) Malignant 68·76 (49·61, 95·33) Emergency Non-malignant 78·01 (44·31, 137·36) Malignant 106·38 (61·77, 183·21) Values in parentheses are 95 per cent c.i. Open in new tab Discussion This study found an overall 1-year incidence of VTE following colectomy of 2·5 per cent. The highest overall absolute rates of VTE occurred after emergency surgery for malignant disease, and rates increased with age for non-malignant indications, although not in patients with cancer. Undoubtedly, the period of greatest risk for VTE following colectomy is the first month after surgery: around 0·5 per cent by the end of this period following any colectomy, and 1·0 per cent with an emergency admission. Of particular note is the fact that patients undergoing emergency colectomy for either a malignant or a non-malignant indication had similar high rates of VTE in the early postoperative period. Given that current guidance recommends extending VTE prophylaxis following colectomy for malignancy to 28 days after surgery, randomized clinical trials of the benefits of extended prophylaxis may be warranted in those patients having an emergency colectomy for non-malignant disease. This study used linked data to identify patients undergoing colectomy from population-based data, with identification of operative procedures from secondary care along with defining VTE in a validated manner from primary20 and secondary care, and is thus uniquely placed to quantify VTE risk accurately. The identification of VTE following discharge from hospital relies on clinical suspicion of the general practitioner and subsequent referral for investigation, thereby minimizing the surveillance bias that may occur in patients identified solely in hospital, as has been suggested in other studies24. The indication for surgery in patients with non-malignant disease was not classified specifically in 25·2 per cent of patients (1029 of 4079) and this figure is in keeping with data previously published using stand-alone HES data25. Furthermore, when these patients were excluded from the analysis, the absolute rates of VTE remained broadly similar, suggesting that the risk of VTE in these patients was no greater than that of patients with a clear indication for surgery. Although the authors were unable to identify patients who received thromboprophylaxis at or around the time of surgery during the study interval (2001–2011), only in the last year of the study were there recommendations for prolonged thromboprophylaxis following surgery in the UK. Uptake of this by colorectal surgeons at this time was poor26, so patients would at most have received low molecular weight heparin while an inpatient following surgery15. The sensitivity analysis of patients following discharge demonstrates that patients were still at increased risk of VTE after discharge, regardless of whether they had received thromboprophylaxis in hospital. In addition, rates of thromboprophylaxis at this time were low, with the ENDORSE study27 estimating that only 50 per cent of patients received appropriate thromboprophylaxis, with a lower proportion in those undergoing emergency surgery – in whom the present study found the highest absolute rates of VTE. Nevertheless the possibility cannot be excluded that rates of VTE might have been higher in some groups than was observed in this study, precisely because they received longer prophylaxis. The potential effects of laparoscopic resection on the rates of VTE could not be assessed fully. Laparoscopic resection rates reported here were low, as it was during the study period that laparoscopic resection rates began to increase. The finding of approximately one-third of resections being performed laparoscopically by the end of the study is in keeping with data from the National Bowel Cancer Audit28. Further studies are required to assess the potential benefits of laparoscopic surgery in this group of patients. Rates of VTE following a diagnosis of colorectal cancer were recently the subject of a systematic review and meta-analysis29, which reported an overall incidence of 16 per 1000 person-years based on four studies reporting incidence in average-risk populations. Few studies have, however, described the risk of VTE following surgery for malignant disease, with conflicting results being reported based on the comparison population used. Two population-based studies14,30 reporting on VTE risk following a diagnosis of colorectal cancer found a decreased risk of VTE after surgery; however, the comparison group comprised patients who did not have surgery because of advanced disease or significant co-morbidity, who have the greatest risk of VTE. Alcalay and colleagues14 reported a 6-month rate of VTE after surgery of 40 per 1000 patient-years, which is similar to the present 6-month rate of 48·08 (95 per cent c.i. 41·57 to 55·62) per 1000 person-years. Their study, however, did not report rates by emergency or elective resection. A study13 using data from the American College of Surgeons National Surgery Quality Improvement Program (NSQIP®) database reported an overall 30-day VTE rate of 2·0 per cent (446 of 21 943) following colectomy for colorectal cancer, but gave limited information on timing after surgery as follow-up was limited to 30 days, and did not report rates by emergency and elective resections. A further study31 using NSQIP® data from 2005 to 2008 reported an overall VTE rate of 2·5 per cent following colectomy, with a postdischarge rate of 0·7 per cent, which is in keeping with the 0·5 per cent rate found in the present study for all patients following colectomy. Fleming and co-workers31 found no difference in VTE rates between non-malignant and malignant disease (both 0·7 per cent), although they did not report rates according to whether surgery was performed as an emergency or electively. The present study, however, has shown a clear increased risk for malignant disease, of 2·4-fold in the first year after emergency surgery and 4·2-fold following elective surgery. Few studies have reported the risk of VTE after surgery for non-malignant disease. A study12 using data from the NSQIP® database reported a 30-day risk of VTE following surgery for IBD of 2·3 per cent (242 of 10 431). This is higher than the 30-day rate for non-malignant disease in the present study; however, it is not clear whether Wallaert et al.12 excluded patients who had had a VTE previously, as in the present study. Their study did, however, find that emergency surgery resulted in a 1·8-fold (odds ratio 1·8, 95 per cent c.i. 1·2 to 2·6) increase in risk of VTE compared with elective surgery. Sweetland and colleagues32 reported an increased VTE risk in women undergoing gastrointestinal surgery (relative risk 56·0, 95 per cent c.i. 39·4 to 80·4) and cancer surgery (relative risk 91·6, 73·9 to 113·4) compared with the risk in women who did not have surgery, with the greatest risk in the first 6 weeks after operation. These authors were, however, unable to describe the risk in detail according to the timing of surgery (emergency or elective), or indication for surgery in terms of non-malignant and malignant disease. Current guidance from the National Institute for Health and Care Excellence15 and the American College of Chest Physicians33 suggests that patients undergoing colectomy for malignant disease should receive 28 days (4 weeks) of extended pharmacological prophylaxis after surgery. This guidance is supported by the present findings, which indicate an increased risk of VTE in patients undergoing surgery for malignant disease in both the elective and emergency setting that extends beyond the inpatient period. However, the present results also indicate that patients having emergency surgery for non-malignant disease have a similar increased risk of VTE during the 30 days after surgery to that of patients with malignant disease operated on as an emergency. Indeed, the magnitude of the risk in patients undergoing emergency non-malignant colectomy was greater than that of patients undergoing colectomy for malignant disease in the elective setting, who currently receive extended prophylaxis. Given the findings of the ENDORSE study27, every effort should be made to assess risk in these patients before operation and ensure that appropriate thromboprophylaxis is prescribed during their hospital stay. Interventional studies assessing the benefit and risks of extended prophylaxis in patients undergoing emergency colectomy for non-malignant disease may be warranted. Acknowledgements This work was funded by a National Institute for Health Research postdoctoral fellowship awarded to D.J.H. J.W. is funded by a University of Nottingham/Nottingham University Hospitals NHS Trust Senior Clinical Research Fellowship. The funders had no role in the design of the study, the collection, analysis and interpretation of data, the writing of the article, or the decision to submit it for publication. Disclosure: The authors declare no conflict of interest. Supporting information Additional supporting information may be found in the online version of this article: Appendix S1 OPCS and ICD codes used to identify colectomy, inflammatory bowel disease and diverticular disease (Word document) Open in new tabDownload slide References 1 Health and Social Care Information Centre . Hospital Episode Statistics . http://www.hscic.gov.uk/hes [accessed 28 July 2015]. 2 Lausen I , Jensen R, Jorgensen LN, Rasmussen MS, Lyng KM, Andersen M et al. Incidence and prevention of deep venous thrombosis occurring late after general surgery: randomised controlled study of prolonged thromboprophylaxis . Eur J Surg 1998 ; 164 : 657 – 663 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Bergqvist D , Agnelli G, Cohen AT, Eldor A, Nilsson PE, Le Moigne-Amrani A et al. ; ENOXACAN II Investigators . Duration of prophylaxis against venous thromboembolism with enoxaparin after surgery for cancer . N Engl J Med 2002 ; 346 : 975 – 980 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Rasmussen MS , Jorgensen LN, Wille-Jørgensen P, Nielsen JD, Horn A, Mohn AC et al. ; FAME Investigators. Prolonged prophylaxis with dalteparin to prevent late thromboembolic complications in patients undergoing major abdominal surgery: a multicenter randomized open-label study . J Thromb Haemost 2006 ; 4 : 2384 – 2390 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Jørgensen LN , Lausen I, Rasmussen MS, Wille-Jørgensen P, Berqvist D. Prolonged thromboprophylaxis with low-molecular weight heparin following major general surgery: an individual patient data meta-analysis . Blood 2002 ; 100 : Abstract 1952. Google Scholar OpenURL Placeholder Text WorldCat 6 Rasmussen MS , Jørgensen LN, Wille-Jørgensen P. Prolonged thromboprophylaxis with low molecular weight heparin for abdominal or pelvic surgery . Cochrane Database Syst Rev 2009 ; ( 1 ) CD004318 . Google Scholar OpenURL Placeholder Text WorldCat 7 Shapiro R , Vogel JD, Kiran RP. Risk of postoperative venous thromboembolism after laparoscopic and open colorectal surgery: an additional benefit of the minimally invasive approach? Dis Colon Rectum 2011 ; 54 : 1496 – 1502 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Byrne BE , Mamidanna R, Vincent CA, Faiz O. Population-based cohort study comparing 30- and 90-day institutional mortality rates after colorectal surgery . Br J Surg 2013 ; 100 : 1810 – 1817 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Walker AJ , West J, Card TR, Humes DJ, Grainge MJ. Variation in the risk of venous thromboembolism in people with colorectal cancer: a population-based cohort study from England . J Thromb Haemost 2014 ; 12 : 641 – 649 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Strate LL , Erichsen R, Horváth-Puhó E, Pedersen L, Baron JA, Sørensen HT. Diverticular disease is associated with increased risk of subsequent arterial and venous thromboembolic events . Clin Gastroenterol Hepatol 2014 ; 12 : 1695 – 1701 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Van Assche G , Vermeire S, Rutgeerts P. Management of acute severe ulcerative colitis . Gut 2011 ; 60 : 130 – 133 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Wallaert JB , De Martino RR, Marsicovetere PS, Goodney PP, Finlayson SRG, Murray JJ et al. Venous thromboembolism after surgery for inflammatory bowel disease: are there modifiable risk factors? Data from ACS NSQIP . Dis Colon Rectum 2012 ; 55 : 1138 – 1144 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Davenport DL , Vargas HD, Kasten MW, Xenos ES. Timing and perioperative risk factors for in-hospital and post-discharge venous thromboembolism after colorectal cancer resection . Clin Appl Thromb Hemost 2012 ; 18 : 569 – 575 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Alcalay A , Wun T, Khatri V, Chew HK, Harvey D, Zhou H et al. Venous thromboembolism in patients with colorectal cancer: incidence and effect on survival . J Clin Oncol 2006 ; 24 : 1112 – 1118 . Google Scholar Crossref Search ADS PubMed WorldCat 15 National Institute for Health and Care Excellence (NICE) . Venous Thromboembolism in Adults Admitted to Hospital – Reducing the Risk [CG92] . NICE : London , 2010 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 16 Benson T The history of the Read Codes: the inaugural James Read Memorial Lecture 2011 . Inform Prim Care 2011 ; 19 : 173 – 182 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 17 Sinha S , Peach G, Poloniecki JD, Thompson MM, Holt PJ. Studies using English administrative data (Hospital Episode Statistics) to assess health-care outcomes – systematic review and recommendations for reporting . Eur J Public Health 2013 ; 23 : 86 – 92 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Crooks C . Epidemiology of Upper Gastrointestinal Bleeding: Studying its Causes and Outcomes using Case Control Studies and Survival Analyses (PhD thesis). University of Nottingham : Nottingham , 2013 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 19 Charlson ME , Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation . J Chronic Dis 1987 ; 40 : 373 – 383 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Lawrenson R , Todd JC, Leydon GM, Williams TJ, Farmer RD. Validation of the diagnosis of venous thromboembolism in general practice database studies . Br J Clin Pharmacol 2000 ; 49 : 591 – 596 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Abdul Sultan A , Tata LJ, Grainge MJ, West J. The incidence of first venous thromboembolism in and around pregnancy using linked primary and secondary care data: a population based cohort study from England and comparative meta-analysis . PloS One 2013 ; 8 : e70310 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Walker AJ , Card TR, West J, Crooks C, Grainge MJ. Incidence of venous thromboembolism in patients with cancer – a cohort study using linked United Kingdom databases . Eur J Cancer 2013 ; 49 : 1404 – 1413 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Vinogradova Y , Coupland C, Hippisley-Cox J. Use of combined oral contraceptives and risk of venous thromboembolism: nested case–control studies using the QResearch and CPRD databases . BMJ 2015 ; 350 : h2135 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Bilimoria KY , Chung J, Ju MH, Haut ER, Bentrem DJ, Ko CY et al. Evaluation of surveillance bias and the validity of the venous thromboembolism quality measure . JAMA 2013 ; 310 : 1482 – 1489 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Faiz O , Warusavitarne J, Bottle A, Tekkis PP, Clark SK, Darzi AW et al. Nonelective excisional colorectal surgery in English National Health Service Trusts: a study of outcomes from Hospital Episode Statistics data between 1996 and 2007 . JAMA 2010 ; 210 : 390 – 401 . Google Scholar OpenURL Placeholder Text WorldCat 26 Srinivasaiah N , Arsalani-Zadeh R, Monson JR. Thrombo-prophylaxis in colorectal surgery: a National Questionnaire Survey of the members of the Association of Coloproctology of Great Britain and Ireland . Colorectal Dis 2012 ; 14 : e390 – e393 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Kakkar AK , Cohen AT, Tapson VF, Bergmann J-F, Goldhaber SZ, Deslandes B et al. ; ENDORSE Investigators . Venous thromboembolism risk and prophylaxis in the acute care hospital setting (ENDORSE Survey): findings in surgical patients . Ann Surg 2010 ; 251 : 330 – 338 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Health and Social Care Information Centre . National Bowel Cancer Audit Annual Report 2013 . http://www.hscic.gov.uk/catalogue/PUB11105/nati-clin-audi-supp-prog-bowe-canc-2013-rep1.pdf [accessed 28 July 2015]. 29 Horsted F , West J, Grainge MJ. Risk of venous thromboembolism in patients with cancer: a systematic review and meta-analysis . PLoS Med 2012 ; 9 : e1001275 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Choi S , Lee KW, Bang SM, Kim S, Lee JO, Kim YJ et al. Different characteristics and prognostic impact of deep-vein thrombosis/pulmonary embolism and intraabdominal venous thrombosis in colorectal cancer patients . Thromb Haemost 2011 ; 106 : 1084 – 1094 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 31 Fleming FJ , Kim MJ, Salloum RM, Young KC, Monson JR. How much do we need to worry about venous thromboembolism after hospital discharge? A study of colorectal surgery patients using the National Surgical Quality Improvement Program database . Dis Colon Rectum 2010 ; 53 : 1355 – 1360 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Sweetland S , Green J, Liu B, Berrington de González A, Canonico M, Reeves G et al. Million Women Study collaborators. Duration and magnitude of the postoperative risk of venous thromboembolism in middle aged women: prospective cohort study . BMJ 2009 ; 339 : b4583 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Gould MK , Garcia DA, Wren SM, Karanicolas PJ, Arcelus JI, Heit JA et al. ; American College of Chest Physicians. Prevention of VTE in nonorthopedic surgical patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines . Chest 2012 ; 141 ( Suppl ): e227S – e277S . Google Scholar Crossref Search ADS PubMed WorldCat Author notes Presented to the Tripartite Colorectal Meeting, Birmingham, UK, June 2014; published in abstract form as Colorectal Dis 2014; 16(Suppl 2): 15 © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd

Waaijer, L; van Diest, P J; Verkooijen, H M; Dijkstra, N -E; van der Pol, C C; Borel Rinkes, I H M; Witkamp, A J

doi: 10.1002/bjs.9950pmid: 26447629

Abstract Background Surgery is the intervention of choice for definitive diagnosis and treatment in women with pathological nipple discharge (PND). Ductoscopy has been reported to improve diagnosis, but as an interventional procedure it may also reduce the need for surgery. This study evaluated interventional ductoscopy in patients with PND. Methods A prospective study on ductoscopy was conducted in consecutive patients with PND, but without a suspected malignancy on routine diagnostic evaluation. Intraductal lesions were removed by ductoscopic extraction. Surgery was undertaken if there were suspicious ductoscopic findings or at the patient's request. Therapeutic efficacy was determined by cannulation success, detection and removal rates, symptom resolution and avoided surgery. Results Ductoscope introduction was successful in 71 (87 per cent) of 82 patients, with abnormalities visualized in 53 (65 per cent); these were mostly polypoid lesions (29 patients). The lesion was removed in 27 of 34 attempted ductoscopic extractions. Twenty-six (32 per cent) of the 82 patients underwent surgery, whereas surgery was avoided in 56 (68 per cent). After a median follow-up of 17 (range 3–45) months, 40 patients (49 per cent) no longer experienced symptoms of PND, 13 of 34 patients experienced an insufficient therapeutic effect after attempted ductoscopic extraction, and the outcome was unknown in two (2 per cent). Malignancy was diagnosed in four patients (5 per cent); two had been missed at ductoscopy and two at initial surgery after ductoscopy. Conclusion Interventional ductoscopy is technically feasible and may help to avoid surgery in the majority of patients. As endoscopic removal of intraductal lesions is not always possible and malignancy can be the underlying cause of PND, ductoscopic instruments should be further optimized to allow definitive histological diagnosis. Introduction Pathological nipple discharge (PND) is defined as spontaneous, unilateral haemorrhagic or serous discharge during a non-lactational period and is the cause of approximately 5 per cent of surgical referrals to a breast clinic1. As PND can be associated with breast cancer, numerous women with such discharge undergo exploratory surgery, such as major duct excision or microdochectomy, to rule out malignancy and to treat symptoms. However, the vast majority of these women have surgery for benign disease, as malignancy is diagnosed in only 3–6 per cent of patients with PND but without abnormalities on breast imaging2–5. There are reports5–8 of a higher risk of an underlying breast cancer, up to 23 per cent, but these studies included patients undergoing surgery or those with a suspected malignancy on breast imaging. As the underlying cause of PND in the majority of patients (57–69 per cent) is a benign papilloma, most women undergo surgery with a concomitant risk of complications and effects on cosmesis, breast-feeding and sensitivity of the nipple, for a benign diagnosis2,9–10. Furthermore, lesions are frequently found at some distance from the nipple orifice9,11. As a consequence, a blind surgical excision may lead to unsuccessful removal of peripheral lesions5. Cytology of nipple aspiration fluid has been proposed as a preoperative diagnostic tool, but predictive values are disappointing, with positive predictive values of 50–55 per cent and negative predictive values of 76–76·5 per cent12,13. Ductoscopy, a minimally invasive intraductal approach that enables direct access to the ductal system via the nipple14–16, has proved a useful tool in the diagnostic evaluation of patients with nipple discharge17. Moncrief and colleagues9 identified 97·3 per cent of pathologically diagnosed papillomas at endoscopy. With improving breast imaging quality and the addition of a third working channel next to channels for optics and irrigation, the first interventional ductoscopy was developed. To improve diagnostic accuracy, intraductal biopsy tools18–20 and a ‘basket’ intervention device were introduced to acquire tissue and remove the lesion under direct vision21–23. Bender and co-workers22 showed ductoscopy to be a promising tool, by demonstrating that 21 of 22 solitary papillomas could be removed successfully by endoscopy22. There have been no studies of the overall and long-term therapeutic efficacy of this attractive minimally invasive approach for patients with PND. The aim of the present study was, therefore, to evaluate the therapeutic efficacy of interventional ductoscopy in patients with PND but no suspicious imaging findings. The endpoints were therapeutic efficacy, determined by cannulation success, lesion detection and removal rates, symptom resolution and surgery avoided. Methods This prospective cohort study included consecutive patients who presented with PND to the University Medical Centre Utrecht (UMCU), The Netherlands, between May 2010 and May 2014. PND was defined as spontaneous, single-duct nipple discharge during a non-lactational period, persisting for more than 3 months. The appearance of the discharge was serous, haemorrhagic or serohaemorrhagic. Informed consent was obtained from all patients according to the Law on Medical Treatment Agreement, after comprehensive information on content, expected course, potential complications/risks and prognosis of this treatment had been provided. The study protocol is outlined in Fig. 1. Before ductoscopy, standard diagnostic evaluation was performed in all patients, including a complete history and physical examination, and recent imaging (mammography, ultrasonography and/or MRI and/or core needle biopsy (CNB) within 3 months) if indicated24. No patient was thought to have a malignancy after diagnostic evaluation. In 50 patients cytological evaluation of the nipple fluid had been performed by the referring hospital. Fig. 1 Open in new tabDownload slide Study protocol Patients with a final assessment category of 3 or less according to the American College of Radiology Breast Imaging Reporting and Data System (BIRADS), or histological proof of a benign cause, underwent ductoscopy. Breast imaging and/or pathology findings obtained at another unit were reviewed by radiologists and pathologists at UMCU. Ductoscopy system The ductoscopic procedure was conducted in an outpatient setting at UMCU. Endoscopy was performed using a 6000-pixel 0·55-mm optic (LaDuScope T-flex; Polydiagnost, Pfaffenhofen, Germany) inserted into a Polyshaft® (1·15 mm outer diameter, PD-DS-1015; Polydiagnost). This provided a separate irrigation channel for saline infusion, and a working channel for ductoscopic extraction of papillary polypoid lesions, for which a grasping endobasket (3-wire basket, diameter 380 µm, PD-TI-2104; Polydiagnost) was used. The ductoscope had a working length of 80 mm, a 0° angle direct view and a field vision of 70°, and was gas-sterilized. Cannulation Before ductoscopy, the nipple and areola of the breast were disinfected with 70 per cent ethanol. After periareolar infiltration of 1 per cent lidocaine, a salivary duct probe (size 0000 to 1; Karl Storz, Tuttlingen, Germany) and an obturator (Polydiagnost) were used for dilatation of the productive lactiferous duct orifice in the nipple. Upward traction of the nipple was used for straightening of the subareolar ducts to facilitate cannulation. The port through which the ductoscope was introduced (SoLex nipple expander®; Polydiagnost) was placed into the duct orifice. Cannulation was considered successful once this had been achieved. Inspection of ductal system After initial intraductal infusion of 4 ml lidocaine, sterile saline was used for ductal distension. Endoscopic exploration was performed starting in the fluid-producing duct. The major lactiferous ducts and segmental branches were visualized in an orderly fashion until the scope could not be advanced any further. Images of ductal abnormalities were recorded. All ductoscopic procedures were performed by a single surgeon, who had experience of more than ten procedures before the start of this study25. A modified version of previously proposed macroscopic classifications was used26–28. Lesions were classified as polypoid, epithelial or miscellaneous, and any of these was considered a positive finding at ductoscopy. Polypoid lesions were defined as localized expansive lesions, smooth-surfaced without haemorrhage or atypical vessels. Epithelial lesions were defined as superficial spreading lesions, accompanied by no obvious elevations and mostly without haemorrhage or atypical vessels. A subtype was the suspicious epithelial lesion, defined as a fissured-surface lesion that was red, friable or displaying atypical vessels. The group of miscellaneous lesions included intraluminal debris and fibrotic tissue, defined as a sudden duct ending (blind-ended duct). When no intraductal abnormalities were found, ductoscopy was considered negative. Histopathological diagnoses were divided into: malignancies, including (in situ) ductal and lobular carcinoma; papilloma, consisting of intraductal papilloma or papillomatosis; and miscellaneous, including benign epithelial lesions or fibrotic tissue, ductal hyperplasia, apocrine metaplasia, fibroadenomas, mastopathy, cysts, ductectasia or periductal fibrosis. An accepted positive result for pathology was the presence of any of these lesions in the ductoscopically or surgically acquired tissue specimen. Interventional procedure Ductoscopic extraction was attempted in patients with intraductal polypoid lesions. For this purpose, the basket was moved forward distally to the lesion, opened, rotated, and moved back to excise and remove the lesion. Removal of (part of) the intraductal lesion was considered as a successful (partial) ductoscopic extraction. Initial diameter and remnant diameter after basket extraction were estimated. Removal was considered complete when less than 10 per cent, and partial when 10 per cent or more, of the lesion remained. Pain assessment For patients' evaluation of the procedure, in the last 18 consecutive patients a paper questionnaire was used to assess pain intensity on a visual analogue scale (VAS) from 0 (no pain) to 10 (maximum pain)29 for lidocaine infiltration anaesthesia, cannulation and endoscopic examination of the ducts. Surgery All suspicious epithelial lesions were treated surgically. Surgery was offered to all remaining patients (either targeted excision when visible on imaging or microdochectomy and/or major duct excision) and left up to the patient's discretion. Complications were ranked according to the Clavien–Dindo classification30. Follow-up The effect of treatment on symptoms was evaluated in all patients during outpatient clinic visits or by telephone consultations after 2–4 weeks, with subsequent follow-up in patients with ongoing symptoms or surgery. In May 2014, all patients were approached to acquire information on symptoms, whether they had undergone surgery elsewhere and the occurrence of new lesions. Results Ductoscopy was performed in 82 women with a mean age of 47 (range 20–72) years (Table 1). The duration of nipple discharge ranged from 0·5 to 108 (mean 18·2) months. For logistical reasons, one patient who had experienced symptoms for less than 3 months underwent ductoscopy. The colour of the discharge was described as haemorrhagic by 68 patients (83 per cent). Nipple abnormalities were reported by five patients, including reversible nipple retraction (4) and skin lesions (1). Follow-up was available for 78 patients (95 per cent). Of the remaining patients, three had died and one was lost to follow-up. All deaths were non-breast-related; these patients had not undergone surgery, nor was there a diagnosis or indication of breast cancer before their death. Table 1 Patient characteristics and baseline breast imaging data . No. of patients* (n = 82) . Age (years)† 47(11) (20–72) Duration of symptoms (months)† 18·2(26·0) (0·5–108) Duration of follow-up (months)‡ 17 (3–45) Side of nipple discharge Unilateral 75 (91) Left 39 Right 36 Bilateral 7 (9) Type of nipple discharge (Sero)sanguineous 68 (83) Other 14 (17) Nipple retraction/skin lesions Yes 5 (6) No 77 (94) Previous surgery on ipsilateral breast Yes 12 (15) No 70 (85) BIRADS category 1 31 (38) 2 36 (44) 3 13 (16) 4 2 (2) Mammographic abnormalities§ Present 11 (13) Not present 69 (84) No mammogram¶ 2 (2) Ultrasound abnormalities# Present 25 (30) Not present 55 (67) No ultrasound performed** 2 (2) MRI abnormalities†† Present 5 (6) Not present 5 (6) No MRI 72 (88) Cytology performed Yes 50 (61) No 32 (39) Core needle biopsy Yes 15 (18) Yes, showing no abnormalities 2 (2) No biopsy 65 (79) Follow-up Lost to follow-up 1 (1) Died 3 (4) One-step surgery‡‡ 4 (5) . No. of patients* (n = 82) . Age (years)† 47(11) (20–72) Duration of symptoms (months)† 18·2(26·0) (0·5–108) Duration of follow-up (months)‡ 17 (3–45) Side of nipple discharge Unilateral 75 (91) Left 39 Right 36 Bilateral 7 (9) Type of nipple discharge (Sero)sanguineous 68 (83) Other 14 (17) Nipple retraction/skin lesions Yes 5 (6) No 77 (94) Previous surgery on ipsilateral breast Yes 12 (15) No 70 (85) BIRADS category 1 31 (38) 2 36 (44) 3 13 (16) 4 2 (2) Mammographic abnormalities§ Present 11 (13) Not present 69 (84) No mammogram¶ 2 (2) Ultrasound abnormalities# Present 25 (30) Not present 55 (67) No ultrasound performed** 2 (2) MRI abnormalities†† Present 5 (6) Not present 5 (6) No MRI 72 (88) Cytology performed Yes 50 (61) No 32 (39) Core needle biopsy Yes 15 (18) Yes, showing no abnormalities 2 (2) No biopsy 65 (79) Follow-up Lost to follow-up 1 (1) Died 3 (4) One-step surgery‡‡ 4 (5) * With percentages in parentheses unless indicated otherwise; † values are mean(s.d.) (range) and ‡ median (range). § A mass, indeterminate/suspicious calcifications, or architectural distortion. ¶ Based on the patient's age (less than 30 years), mammography was replaced by MRI. # A mass or intraductal mass(es). ** Patients with an American College of Radiology Breast Imaging Reporting and Data System (BIRADS) category of 1 on mammography. For logistical reasons ultrasonography was not performed. †† A mass or suspicious enhancement pattern. ‡‡ The ductoscopic therapeutic efficacy was unknown in patients who had one-step surgery. Open in new tab Table 1 Patient characteristics and baseline breast imaging data . No. of patients* (n = 82) . Age (years)† 47(11) (20–72) Duration of symptoms (months)† 18·2(26·0) (0·5–108) Duration of follow-up (months)‡ 17 (3–45) Side of nipple discharge Unilateral 75 (91) Left 39 Right 36 Bilateral 7 (9) Type of nipple discharge (Sero)sanguineous 68 (83) Other 14 (17) Nipple retraction/skin lesions Yes 5 (6) No 77 (94) Previous surgery on ipsilateral breast Yes 12 (15) No 70 (85) BIRADS category 1 31 (38) 2 36 (44) 3 13 (16) 4 2 (2) Mammographic abnormalities§ Present 11 (13) Not present 69 (84) No mammogram¶ 2 (2) Ultrasound abnormalities# Present 25 (30) Not present 55 (67) No ultrasound performed** 2 (2) MRI abnormalities†† Present 5 (6) Not present 5 (6) No MRI 72 (88) Cytology performed Yes 50 (61) No 32 (39) Core needle biopsy Yes 15 (18) Yes, showing no abnormalities 2 (2) No biopsy 65 (79) Follow-up Lost to follow-up 1 (1) Died 3 (4) One-step surgery‡‡ 4 (5) . No. of patients* (n = 82) . Age (years)† 47(11) (20–72) Duration of symptoms (months)† 18·2(26·0) (0·5–108) Duration of follow-up (months)‡ 17 (3–45) Side of nipple discharge Unilateral 75 (91) Left 39 Right 36 Bilateral 7 (9) Type of nipple discharge (Sero)sanguineous 68 (83) Other 14 (17) Nipple retraction/skin lesions Yes 5 (6) No 77 (94) Previous surgery on ipsilateral breast Yes 12 (15) No 70 (85) BIRADS category 1 31 (38) 2 36 (44) 3 13 (16) 4 2 (2) Mammographic abnormalities§ Present 11 (13) Not present 69 (84) No mammogram¶ 2 (2) Ultrasound abnormalities# Present 25 (30) Not present 55 (67) No ultrasound performed** 2 (2) MRI abnormalities†† Present 5 (6) Not present 5 (6) No MRI 72 (88) Cytology performed Yes 50 (61) No 32 (39) Core needle biopsy Yes 15 (18) Yes, showing no abnormalities 2 (2) No biopsy 65 (79) Follow-up Lost to follow-up 1 (1) Died 3 (4) One-step surgery‡‡ 4 (5) * With percentages in parentheses unless indicated otherwise; † values are mean(s.d.) (range) and ‡ median (range). § A mass, indeterminate/suspicious calcifications, or architectural distortion. ¶ Based on the patient's age (less than 30 years), mammography was replaced by MRI. # A mass or intraductal mass(es). ** Patients with an American College of Radiology Breast Imaging Reporting and Data System (BIRADS) category of 1 on mammography. For logistical reasons ultrasonography was not performed. †† A mass or suspicious enhancement pattern. ‡‡ The ductoscopic therapeutic efficacy was unknown in patients who had one-step surgery. Open in new tab Cannulation and inspection of the ductal system Ductoscopic results are summarized in Fig. 2. The ductal system was cannulated successfully in 71 (87 per cent) of the 82 patients. In 11 patients (13 per cent) ductoscopy failed during the dilatation part of the procedure, because the duct orifices were too narrow (7) or unidentifiable (1), or owing to nipple retraction (3). Fig. 2 Open in new tabDownload slide Results of ductoscopic extraction (DE) and outcomes of patients with pathological nipple discharge (PND). *Including one papilloma with atypical ductal hyperplasia. †Ductal carcinoma in situ diagnosed after 45 months Ductoscopy was positive in 53 patients. The abnormalities found were polypoid lesions in 29 patients, epithelial lesions in 16 and miscellaneous in eight. False routes (duct perforation) occurred in 12 patients (15 per cent), and this prevented cannulation in three. Despite the false route, a lesion was subsequently removed by ductoscopic extraction in four patients. Eight of 12 patients with false routes had persistent symptoms and four were symptom-free, of whom three had undergone a ductoscopic lesion extraction. Ductoscopic extraction Owing to the size or type of lesions (such as a very small or obstructing large polypoid lesion, epithelial abnormalities and intraductal debris not suitable for basket extraction), ductoscopic extraction was not attempted in 19 (36 per cent) of 53 patients in whom intraductal abnormalities were visualized. Only the 34 patients in whom ductoscopic extraction was attempted were included in the subsequent analyses of the interventional procedure. In 27 of 34 attempts, the observed intraductal lesion was removed. Lesion removal was complete in 21 and partial in six. Seven of 34 attempts failed, owing to either problematic basket passage (6) or lesion dimensions being too small to be grasped with the basket (1). A sufficient amount of tissue for adequate histological assessment was collected in 13 of the 34 attempts. All of these were diagnosed as intraductal papilloma. The reason for failure of histological assessment was insufficient tissue quantity and intraductal loss of the removed tissue. Pain assessment Mild pain was reported for infiltration of lidocaine and for cannulation, with median (range) VAS scores of 3 (0–8) and 2 (0–8) respectively. Pain was reported to be moderate during ductoscopy, with a median VAS score of 3·5 (1–7). Therapeutic efficacy After a median follow-up of 17 (range 3–45) months after ductoscopy, 40 patients (49 per cent of all those included in the study) no longer experienced symptoms of PND without having surgery. Thirty-six patients (44 per cent) had persistent PND symptoms after ductoscopy. Six patients (7 per cent) had unknown therapeutic efficacy, including four who had surgery immediately after ductoscopy, one who died before therapeutic efficacy had been assessed, and one who was lost to follow-up after negative ductoscopy. Among the 27 patients in whom a lesion had been removed successfully, symptom resolution was reported by 19 patients. In 12 of the 40 symptom-free patients, only ductal flushing with saline had been used; the endoscopic appearance was epithelial lesions (7), debris (3), a small papilloma (1) and a blind-ended duct (1). In eight symptom-free patients ductoscopy had been negative and in one patient cannulation was unsuccessful. Surgery Surgery was performed in 26 (32 per cent) of 82 patients, after a median interval of 3 (range 0–8) months. Four patients underwent one-step surgery immediately after negative ductoscopy (2) or unsuccessful or partial ductoscopic extraction (2). Other indications for surgery were ductoscopic findings of suspicious epithelial lesions (3) or patient's preference, owing to symptoms (10) or for reassurance (9). Surgery involved either major duct excision (17), microdochectomy (2), a combination of both (3) or excision biopsy (4). Surgery was performed without imaging guidance, except for the excision biopsies. Here, ultrasound-guided localization (3) or ductoscopy-assisted wire placement (1) was used, as described previously31. Correlation between ductoscopic and histological findings A total of 36 patients (44 per cent) had a diagnosis based on tissue acquired from either ductoscopic extraction (13) and/or surgery (26) (Table 2). Of the 29 polypoid lesions, histology was acquired in 21, showing papilloma in 20 and benign epithelial changes in one patient. A total of 23 papillomas were diagnosed histologically, of which 20 had been visualized as a polypoid lesion by ductoscopy. Where histology was acquired from both ductoscopic extraction and surgical excision (2 patients) the results were in concordance, revealing intraductal papilloma. In one patient with inconclusive histology of ductoscopic extraction, surgical excision revealed papilloma. Table 2 Ductoscopic appearance of lesions in relation to histological diagnosis in 36 patients . Histopathology . Total . . Papilloma(tosis) . (In situ) malignancy . Benign lesion/fibrosis . Normal breast tissue . No material obtained . Ductoscopy-positive Polypoid lesion 20* 0 1 0 8 29 Epithelial lesion 1 2 2 0 11 16 Suspicious 2 1 3 Miscellaneous 0 0 1 0 7 8 Fibrotic duct 1† 1 2 Intraductal debris 6 6 Ductoscopy-negative 1 0 2 2 13 18 Unsuccessful cannulation 1 1 1 1 7 11 Total 23 3 7 3 46 82 . Histopathology . Total . . Papilloma(tosis) . (In situ) malignancy . Benign lesion/fibrosis . Normal breast tissue . No material obtained . Ductoscopy-positive Polypoid lesion 20* 0 1 0 8 29 Epithelial lesion 1 2 2 0 11 16 Suspicious 2 1 3 Miscellaneous 0 0 1 0 7 8 Fibrotic duct 1† 1 2 Intraductal debris 6 6 Ductoscopy-negative 1 0 2 2 13 18 Unsuccessful cannulation 1 1 1 1 7 11 Total 23 3 7 3 46 82 * One polypoid lesion was diagnosed histopathologically as intraductal papilloma with a focus of atypical ductal hyperplasia. † One malignancy was diagnosed at breast reduction surgery 45 months after ductoscopy, but histology of the major duct excision 1 month after ductoscopy showed fibrotic tissue. Open in new tab Table 2 Ductoscopic appearance of lesions in relation to histological diagnosis in 36 patients . Histopathology . Total . . Papilloma(tosis) . (In situ) malignancy . Benign lesion/fibrosis . Normal breast tissue . No material obtained . Ductoscopy-positive Polypoid lesion 20* 0 1 0 8 29 Epithelial lesion 1 2 2 0 11 16 Suspicious 2 1 3 Miscellaneous 0 0 1 0 7 8 Fibrotic duct 1† 1 2 Intraductal debris 6 6 Ductoscopy-negative 1 0 2 2 13 18 Unsuccessful cannulation 1 1 1 1 7 11 Total 23 3 7 3 46 82 . Histopathology . Total . . Papilloma(tosis) . (In situ) malignancy . Benign lesion/fibrosis . Normal breast tissue . No material obtained . Ductoscopy-positive Polypoid lesion 20* 0 1 0 8 29 Epithelial lesion 1 2 2 0 11 16 Suspicious 2 1 3 Miscellaneous 0 0 1 0 7 8 Fibrotic duct 1† 1 2 Intraductal debris 6 6 Ductoscopy-negative 1 0 2 2 13 18 Unsuccessful cannulation 1 1 1 1 7 11 Total 23 3 7 3 46 82 * One polypoid lesion was diagnosed histopathologically as intraductal papilloma with a focus of atypical ductal hyperplasia. † One malignancy was diagnosed at breast reduction surgery 45 months after ductoscopy, but histology of the major duct excision 1 month after ductoscopy showed fibrotic tissue. Open in new tab Atypical ductal hyperplasia (ADH) was found in one of 23 papillomas diagnosed. Ductoscopy was performed in the operating room because a 10-mm diameter lesion had been measured on ultrasound examination. A large polypoid lesion obstructing the duct was visualized endoscopically. Ductoscopic extraction with the basket was not possible and the patient underwent surgery. Histopathology confirmed the preoperative CNB diagnosis of intraductal papilloma, but also revealed a focus of ADH. The surgical specimen of two suspicious epithelial lesions on ductoscopy contained malignancy (Table 2). After a median follow-up of 17 (range 3–45) months a total of four patients (5 per cent) had been diagnosed with invasive breast cancer (3) or ductal carcinoma in situ (DCIS) (1). Of these, lesions in two patients had been identified correctly as suspicious by ductoscopic evaluation. Malignancies were identified surgically at a median of 5 (3–45) months after initial ductoscopy, either following a suspicious finding at ductoscopy (2) or at follow-up (2). One malignancy was missed at both ductoscopy and initial surgery; ductoscopy showed a white–yellow duct obstruction indicative of fibrotic tissue, which was confirmed at microdochectomy. However, extensive DCIS was found at breast reduction 45 months after initial ductoscopy and microdochectomy. It is unknown whether this malignancy was already present at the time of ductoscopy and initial surgery. Nevertheless, it was regarded as ‘missed malignancy’ in the present study. Complications Three women (4 per cent) developed mastitis in the examined breast for which antibiotic treatment (3) and/or surgical incision (1) was initiated. Mastitis was diagnosed in two patients and treated by the general practitioner, without cell culture. At the patient's explicit request, ductoscopy was performed in one woman during lactation in the presence of a galactocele. Four days after ductoscopy the woman presented to UMCU with mastitis and an abscess; surgical incision was performed and oral antibiotics were given based on the Staphylococcus aureus cultured. Among the 26 patients who underwent surgery, iatrogenic injury to the nipple resulting in an areola scar (1), necrosis of the nipple (1), postoperative mastitis (1) or wound infection (1) occurred, resulting in a postoperative complication rate of 15 per cent. Although complications occurred more frequently after surgery, complications with a Clavien–Dindo classification as high as grade IIIa were observed after ductoscopy alone (1 patient). Discussion Ductoscopy has emerged as a minimally invasive approach to identify intraductal neoplasia, but more recently intraductal interventional techniques have been introduced. These may aid in non-surgical treatment and improve selection for surgery, leading to possible improvements in quality of life while reducing costs and complications. The results of the present study, based on a prospective cohort of 82 patients, indicate that interventional ductoscopy is technically feasible for the diagnosis and treatment of PND but no suspicious findings on imaging, avoiding surgery in the majority of patients. The 87 per cent cannulation rate in the present study is in accordance with previously reported rates of 81–100 per cent in women with PND22,32–35. With intraductal abnormalities visualized in 53 (75 per cent) of 71 successfully cannulated breasts, the positive ductoscopy rate was higher in the present study than reported previously: 41–65 per cent in studies of 54–1093 successfully cannulated breasts22,27,33. A possible explanation could be the use of more advanced imaging technology. The focus of the present study was the therapeutic effect of ductoscopy and the prevention of unnecessary surgery for patients with PND with a low suspicion of malignancy (BIRADS category of 3 or less, or benign histology). The rate of successful lesion removal of 79 per cent (27 of 34), with symptom resolution in 70 per cent (19 of 27), is in accordance with the results of Bender and colleagues22; they reported an 85 per cent success rate in 26 solitary papillomas, which led to symptom resolution in 95 per cent of patients. Kamali and co-workers34, on the other hand, reported successful ductoscopic extraction in only 47 per cent of 47 solitary papillomas, with symptom resolution in 29 per cent; however, their study also included patients with significant indications of malignancy and/or palpable masses confirmed by mammography. Nevertheless, further improvement of the therapeutic efficacy of current interventional techniques is needed. One potential method is transductal laser ablation; this is currently being tested in a feasibility trial at UMCU, which is recruiting patients who have been diagnosed with an intraductal papilloma and persistent PND after partial ductoscopic extraction. After a median follow-up of 17 months, 40 of 82 patients no longer experienced PND symptoms, despite not having surgery. However, ductoscopic lesion extraction had been performed in only 19 of these patients; this could indicate a self-limiting disease or a therapeutic effect of ductoscopic saline flushing alone. In 12 of the 19 patients in whom ductoscopic extraction of the intraductal, mostly epithelial, abnormality was not attempted, symptom resolution occurred within 3 months. This strengthens the suggestion of a therapeutic effect of ductal saline lavage. As repeat ductoscopies without signs of intraductal fibrosis were performed, it is not likely that symptom resolution would be due to postductoscopic fibrosis. Symptom persistence might be explained by the intraductal abnormality remaining in situ despite lavage, or the lesion causing symptoms not being visualized (for example an intraductal papilloma). After ductoscopy, 56 patients (68 per cent) did not require further surgery, as they were either symptom-free or received sufficient reassurance. However, for definitive reassurance, ductoscopy should distinguish adequately between the different types of lesion. Here, histology of two of the three macroscopically suspicious findings showed malignancy, whereas 20 of the 21 benign-appearing polypoid lesions were confirmed histologically as papillomas. Although histology was lacking for 56 per cent of patients, the results imply a correlation between endoscopic appearance of intraductal lesions and final histological diagnosis, as proposed previously9,26. In previous studies9,33–34,36 there was a significant correlation between the ductoscopic identification (73–97·3 per cent) and histological diagnosis of papillomas, with a positive predictive value of 73–92 per cent. Comparing ductoscopy for the diagnosis of PND with standard investigation methods, Kamali and colleagues34 reported a 72 per cent sensitivity for papillomas on ultrasonography, 62·9 per cent at mammography, 81·4 per cent for galactography and 86·6 per cent with ductoscopy. To increase the diagnostic yield, Shen and co-workers37 combined ductoscopy in 259 women with symptoms of PND, regardless of malignancy suspicion and findings on breast imaging, with cytological analysis of ductal washings. The ductal washing technique yielded an average of 5000 cells per washed duct, compared with 50 in nipple aspirate fluid alone. A papillary lesion was visualized in 36 per cent of patients, with an overall 83 per cent positive predictive value of ductoscopy itself; this improved to 86 per cent when combined with cytological analysis of ductal washings. The diagnostic value of ductoscopy for malignancy was studied in 1048 patients27; the reported sensitivity of ductoscopy for detection of malignant lesions was 94·2 per cent, but the specificity was low. Others38 have shown that there are no visual characteristics that can accurately predict a final diagnosis of malignancy. Therefore, a device to facilitate adequate intraductal biopsy and subsequent diagnosis seems indispensable for future ductoscopic diagnosis of malignancy, especially considering that all malignant findings visualized by ductoscopy were epithelial lesions not suitable for basket extraction. The current basket device is useful for tissue collection from polypoid lesions; removal was successful in over three-quarters of attempts in the present study. However, to improve tissue collection for pathological assessment, a better tissue grip with the basket is needed. In the present cohort, three lesions were diagnosed as a solid mass measuring more than 10 mm on ultrasonography, and ductoscopy visualized an obstructing intraductal polypoid lesion for which basket extraction was unsuccessful. One solid mass, measuring 10 mm in diameter, had been diagnosed as intraductal papilloma on initial ultrasound-guided CNB, but was upstaged to papilloma with ADH in the excision specimen. Upstaging of pure intraductal papillomas on CNB owing to atypia or malignancy on excision has been reported in up to 16·4 per cent39,40, depending partly on the type and size of biopsy used41. Previous authors39,42 therefore recommended excision of all papillomas diagnosed on CNB. These studies, however, included lesions initially diagnosed on CNB, inherent to BIRADS 3 or above. Cheng et al.43 and Chang and co-workers44 reported that upgrading to malignancy was significantly more common with increased size on ultrasonography for lesions larger than 1·2 and 1·5 cm respectively. Irrespective of the possible risk of underlying atypia or malignancy, the present results show that transductal basket extraction of lesions larger than 10 mm on ultrasonography is not feasible. Therefore, surgical excision is recommended for treatment of PND caused by papillomas larger than 10 mm. Although the inclusion criteria in the present study focused on avoiding malignancies, a total of four patients (5 per cent) were still diagnosed with invasive breast cancer or DCIS. This is in accordance with rates of 3–6 per cent in breast clinic-referred women with PND2–4, and higher than the 0–2 per cent reported among patients with PND but a negative physical examination, mammogram and ultrasound examination4,45. Whether initial surgery would have diagnosed these malignancies is questionable. In two patients, subsequent surgery found only benign disease, and malignancy was diagnosed at follow-up. Malignancies have often been shown to be localized in the periphery of the breast46. In addition, in a study of galactograms11 it was found that 70 per cent patients with carcinomas presenting with nipple discharge had lesions further than 2 cm from the nipple. Conventional surgery by major duct excision or microdochectomy is therefore likely to miss these lesions because only the proximal ducts, usually closer than 3 cm, are removed. This is underlined by the higher incidence of multiple lesions identified by ductoscopy, suggestive of underestimation of the true causes of bleeding from deeper lesions that would not be identified by classical blind retroareolar resection47. Although the present results indicate the need for awareness in patients with PND, with a systematic approach the risk of underlying carcinoma can probably be defined further, whereas benign pathology can be diagnosed and treated without surgery. In current practice, either surgical excision or follow-up is advised. Long-term annual follow-up visits and imaging are associated with high costs and increased patient anxiety, so patients frequently opt for surgical excision. Recognizing the limitations of the present study, a ductoscopic approach is suggested for patients with PND, BIRADS category 3 or lower, and/or benign histology, except for those with irreversible nipple retraction, skin abnormalities or with lesions larger than 10 mm on ultrasound examination. Ductoscopic extraction can be used for polypoid lesions. Surgery is indicated in the event of suspicious epithelial findings or unsuccessful ductoscopic extraction without definitive benign histology. Disclosure The authors declare no conflict of interest. Snapshot quiz 15/12 Open in new tabDownload slide References 1 Dixon J , Mansel R. ABC of breast diseases. Symptoms assessment and guidelines for referral . BMJ 1994 ; 309 : 722 – 726 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Vargas HI , Vargas MP, Eldrageely K, Gonzalez KD, Khalkhali I. Outcomes of clinical and surgical assessment of women with pathological nipple discharge . Am Surg 2006 ; 72 : 124 – 128 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Seltzer MH . Breast complaints, biopsies, and cancer correlated with age in 10 000 consecutive new surgical referrals . Breast J 2004 ; 10 : 111 – 117 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Gray RJ , Pockaj BA, Karstaedt PJ. Navigating murky waters: a modern treatment algorithm for nipple discharge . Am J Surg 2007 ; 194 : 850 – 854 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Dillon MF , Mohd Nazri SR, Nasir S, McDermott EW, Evoy D, Crotty TB et al. The role of major duct excision and microdochectomy in the detection of breast carcinoma . BMC Cancer 2006 ; 6 : 164 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Morrogh M , Morris EA, Liberman L, Borgen PI, King TA. The predictive value of ductography and magnetic resonance imaging in the management of nipple discharge . Ann Surg Oncol 2007 ; 14 : 3369 – 3377 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Kocdor MA , Sevinc AI, Canda T, Balci P, Saydam S, Cavdaroglu O et al. Pathologic nipple discharge in patients with radiologically invisible mass: review of 28 consecutive sub-areolar explorations . Breast J 2009 ; 15 : 230 – 235 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Cabioglu N , Hunt KK, Singletary SE, Stephens TW, Marcy S, Meric F et al. Surgical decision making and factors determining a diagnosis of breast carcinoma in women presenting with nipple discharge . J Am Coll Surg 2003 ; 196 : 354 – 364 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Moncrief RM , Nayar R, Diaz LK, Staradub VL, Morrow M, Khan SA. A comparison of ductoscopy-guided and conventional surgical excision in women with spontaneous nipple discharge . Ann Surg 2005 ; 241 : 575 – 581 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Dietz JR , Crowe JP, Grundfest S, Arrigain S, Kim JA. Directed duct excision by using mammary ductoscopy in patients with pathologic nipple discharge . Surgery 2002 ; 132 : 582 – 587 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Hou MF , Huang TJ, Liu GC. The diagnostic value of galactography in patients with nipple discharge . Clin Imaging 2001 ; 25 : 75 – 81 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Simmons R , Adamovich T, Brennan M, Christos P, Schultz M, Eisen C et al. Nonsurgical evaluation of pathologic nipple discharge . Ann Surg Oncol 2003 ; 10 : 113 – 116 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Morrogh M , Park A, Elkin EB, King TA. Lessons learned from 416 cases of nipple discharge of the breast . Am J Surg 2010 ; 200 : 73 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Okazaki A , Okazaki M, Asaishi K, Satoh H, Watanabe Y, Mikami T et al. Fiberoptic ductoscopy of the breast: a new diagnostic procedure for nipple discharge . Jpn J Clin Oncol 1991 ; 21 : 188 – 193 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 15 Berna JD , Garcia-Medina V, Kuni CC. Ductoscopy: a new technique for ductal exploration . Eur J Radiol 1991 ; 12 : 127 – 129 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Makita M , Sakamoto G, Akiyama F, Namba K, Sugano H, Kasumi F et al. Duct endoscopy and endoscopic biopsy in the evaluation of nipple discharge . Breast Cancer Res Treat 1991 ; 18 : 179 – 187 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Kapenhas-Valdes E , Feldman SM, Cohen JM, Boolbol SK. Mammary ductoscopy for evaluation of nipple discharge . Ann Surg Oncol 2008 ; 15 : 2720 – 2727 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Matsunaga T , Kawakami Y, Namba K, Fujii M. Intraductal biopsy for diagnosis and treatment of intraductal lesions of the breast . Cancer 2004 ; 101 : 2164 – 2169 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Hünerbein M , Raubach M, Gebauer B, Schneider W, Schlag PM. Ductoscopy and intraductal vacuum assisted biopsy in women with pathologic nipple discharge . Breast Cancer Res Treat 2006 ; 99 : 301 – 307 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Ling H , Liu G, Lu J, Love S, Zhang J, Xu X et al. Fiberoptic ductoscopy-guided intraductal biopsy improve the diagnosis of nipple discharge . Breast J 2009 ; 15 : 168 – 175 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Balci FL , Feldman SM. Interventional ductoscopy for pathological nipple discharge . Ann Surg Oncol 2013 ; 20 : 3352 – 3354 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Bender O , Balci FL, Yüney E, Akbulut H. Scarless endoscopic papillomectomy of the breast . Onkologie 2009 ; 32 : 94 – 98 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 23 Kamali S , Bender O, Aydin MT, Yuney E, Kamali G. Ductoscopy in the evaluation and management of nipple discharge . Ann Surg Oncol 2010 ; 17 : 778 – 783 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Oncoline . Dutch Guideline Breast Cancer 2012 . http://www.oncoline.nl/breastcancer [accessed 20 July 2014 ]. 25 Zagouri F , Sergentanis TN, Giannakopoulou G, Panopoulou E, Chrysikos D, Bletsa G et al. Breast ductal endoscopy: how many procedures qualify? BMC Res Notes 2009 ; 2 : 115 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Makita M , Akiyama F, Gomi N, Ikenaga M, Yoshimoto M, Kasumi F et al. Endoscopic classification of intraductal lesions and histological diagnosis . Breast Cancer 2002 ; 9 : 220 – 225 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Liu GY , Lu JS, Shen KW, Wu J, Chen C-M, Hu Z et al. Fiberoptic ductoscopy combined with cytology testing in the patients of spontaneous nipple discharge . Breast Cancer Res Treat 2008 ; 108 : 271 – 277 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Rose C , Bojahr B, Grunwald S, Frese H, Jäger B, Ohlinger R. Ductoscopy-based descriptors of intraductal lesions and their histopathologic correlates . Onkologie 2010 ; 33 : 307 – 312 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Huskisson E Measurement of pain . Lancet 1974 ; 79 : 1127 – 1131 . Google Scholar OpenURL Placeholder Text WorldCat 30 Clavien PA , Barkun J, de Oliveira ML, Vauthey JN, Dindo D, Schulick RD et al. The Clavien–Dindo classification of surgical complications: five-year experience . Ann Surg 2009 ; 250 : 187 – 196 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Hahn M , Fehm T, Solomayer EF, Siegmann KC, Hengstmann AS, Wallwiener D et al. Selective microdochectomy after ductoscopic wire marking in women with pathological nipple discharge . BMC Cancer 2009 ; 9 : 151 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Grunwald S , Heyer H, Paepke S, Schwesinger G, Schimming A, Hahn M et al. Diagnostic value of ductoscopy in the diagnosis of nipple discharge and intraductal proliferations in comparison to standard methods . Onkologie 2007 ; 30 : 243 – 248 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 33 Denewer A , El-Etribi K, Nada N, El-Metwally M. The role and limitations of mammary ductoscope in management of pathologic nipple discharge . Breast J 2008 ; 14 : 442 – 449 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Kamali S , Bender O, Kamali GH, Aydin MT, Karatepe O, Yuney E et al. Diagnostic and therapeutic value of ductoscopy in nipple discharge and intraductal proliferations compared with standard methods . Breast Cancer 2014 ; 21 : 154 – 161 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Fisher CS , Margenthaler JA. A look into the ductoscope: its role in pathologic nipple discharge . Ann Surg Oncol 2011 ; 18 : 3187 – 3191 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Vaughan A , Crowe JP, Brainard J, Dawson A, Kim J, Dietz JR. Mammary ductoscopy and ductal washings for the evaluation of patients with pathologic nipple discharge . Breast J 2009 ; 15 : 254 – 260 . Google Scholar Crossref Search ADS PubMed WorldCat 37 Shen KW , Wu J, Lu JS, Han QX, Shen ZZ, Nguyen M et al. Fiberoptic ductoscopy for patients with nipple discharge . Cancer 2000 ; 89 : 1512 – 1519 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Louie LD , Crowe JP, Dawson AE, Lee KB, Baynes DL, Dowdy T et al. Identification of breast cancer in patients with pathologic nipple discharge: does ductoscopy predict malignancy? Am J Surg 2006 ; 192 : 530 – 533 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Jaffer S , Nagi C, Bleiweiss IJ. Excision is indicated for intraductal papilloma of the breast diagnosed on core needle biopsy . Cancer 2009 ; 115 : 2837 – 2843 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Mercado C , Hamele-bena D, Oken SM, Singer CI, Cangiarella J. Papillary lesions of the breast at percutaneous core-needle biopsy . Radiology 2006 ; 238 : 801 – 808 . Google Scholar Crossref Search ADS PubMed WorldCat 41 Chang JM , Han W, Moon WK, Cho N, Noh DY, Park IA et al. Papillary lesions initially diagnosed at ultrasound-guided vacuum-assisted breast biopsy: rate of malignancy based on subsequent surgical excision . Ann Surg Oncol 2011 ; 18 : 2506 – 2514 . Google Scholar Crossref Search ADS PubMed WorldCat 42 Valdes EK , Feldman SM, Boolbol SK. Papillary lesions: a review of the literature . Ann Surg Oncol 2007 ; 14 : 1009 – 1013 . Google Scholar Crossref Search ADS PubMed WorldCat 43 Cheng TY , Chen CM, Lee MY, Lin KJ, Hung CF, Yang PS et al. Risk factors associated with conversion from nonmalignant to malignant diagnosis after surgical excision of breast papillary lesions . Ann Surg Oncol 2009 ; 16 : 3375 – 3379 . Google Scholar Crossref Search ADS PubMed WorldCat 44 Chang JM , Moon WK, Cho N, Han W, Noh DY, Park IA et al. Risk of carcinoma after subsequent excision of benign papilloma initially diagnosed with an ultrasound (US)-guided 14-gauge core needle biopsy: a prospective observational study . Eur Radiol 2010 ; 20 : 1093 – 1100 . Google Scholar Crossref Search ADS PubMed WorldCat 45 Sabel MS , Helvie MA, Breslin T, Curry A, Diehl KM, Cimmino VM et al. Is duct excision still necessary for all cases of suspicious nipple discharge? Breast J 2012 ; 18 : 157 – 162 . Google Scholar Crossref Search ADS PubMed WorldCat 46 Shen KW , Wu J, Lu JS, Han QX, Shen ZZ, Nguyen M et al. Fiberoptic ductoscopy for breast cancer patients with nipple discharge . Surg Endosc 2001 ; 15 : 1340 – 1345 . Google Scholar Crossref Search ADS PubMed WorldCat 47 Dooley WC . Routine operative breast endoscopy for bloody nipple discharge . Ann Surg Oncol 2002 ; 9 : 920 – 923 . Google Scholar Crossref Search ADS PubMed WorldCat © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd

Vos, E L; Voogd, A C; Verhoef, C; Siesling, S; Obdeijn, I M; Koppert, L B

doi: 10.1002/bjs.9947pmid: 26445887

Abstract Background Although evidence for the benefits of preoperative MRI in breast cancer is lacking, use of MRI is increasing and characterized by large interhospital variation. The aim of the study was to evaluate MRI use and surgical outcomes retrospectively. Methods Women with invasive breast cancer (pT1–3) or ductal carcinoma in situ (DCIS), diagnosed in 2011–2013, were selected from the Netherlands Cancer Registry and subdivided into the following groups: invasive cancer, high-grade DCIS, non-palpable cancer, age 40 years or less, and invasive lobular cancer. Associations between preoperative MRI use and initial mastectomy, resection margin after breast-conserving surgery (BCS), re-excision after BCS, and final mastectomy were analysed. Results In total, 5514 women were included in the study; 1637 (34·1 per cent) of 4801 women with invasive cancer and 150 (21·0 per cent) of 713 with DCIS had preoperative MRI. Positive resection margins were found in 18·1 per cent women who had MRI and in 15·1 per cent of those who did not (adjusted odds ratio (OR) 1·20, 95 per cent c.i. 1·00 to 1·45), with no differences in subgroups. Re-excision rates were 9·8 per cent in the MRI group and 7·2 per cent in the no-MRI group (adjusted OR 1·33, 1·04 to 1·70), with no differences in subgroups. In the MRI group, 38·8 per cent of patients ultimately underwent mastectomy, compared with 24·2 per cent in the no-MRI group (adjusted OR 2·13, 1·87 to 2·41). This difference was not found for patients aged 40 years or less, or for those diagnosed with lobular cancer. Conclusion No subgroup was identified in which preoperative MRI influenced the risk of margin involvement or re-excision rate after BCS. MRI was significantly associated with more extensive surgery, except in patients aged 40 years or less and those with invasive lobular cancer. These results suggest that use of preoperative MRI should be more targeted, and that general, widespread use be discouraged. Introduction The primary goal of breast-conserving surgery (BCS) for breast cancer is to obtain complete tumour excision. If the excision is incomplete, re-excision may be necessary, which will increase healthcare costs, the burden to the patient and the risk of a poor cosmetic result. Obtaining a high complete tumour excision rate is important to the patient, as well as to the many healthcare stakeholders that use the re-excision rate as a quality indicator of breast cancer care. A meta-analysis1 has shown that MRI can detect mammographically and clinically occult disease in the ipsilateral breast in around 16 per cent of patients with invasive cancer or ductal carcinoma in situ (DCIS). MRI has also been shown2 to be more accurate than mammography or ultrasonography in determining tumour size and delineating tumour margins. As a consequence, preoperative MRI is believed to improve the surgical planning and likelihood of complete tumour excision at the first attempt. Especially in DCIS with high nuclear grade, MRI complementary to mammography could help improve the ability to diagnose the extent of the DCIS3,4. The American College of Radiology guidelines suggest that contrast-enhanced MRI of the breast may be useful to determine both the extent of disease and the presence of multifocality and multicentricity in patients with invasive carcinoma and DCIS. However, currently there is no convincing evidence that preoperative MRI does improve surgical outcomes, such as the rates of positive margins, re-excision or breast conservation, in the average patient with breast cancer5–7. An exception to this is the subgroup of patients with lobular cancer, for whom a significantly reduced rate of re-excision has repeatedly been shown following preoperative MRI5–6,8–9. This may support the targeted use of preoperative MRI in particular subgroups of patients with breast cancer, as recommended by the European Society of Breast Cancer Specialists (EUSOMA)10. Although multiple studies have suggested that MRI is not beneficial in patients with breast cancer in general, and that subgroups must be identified, studies investigating the benefit of MRI in specific subgroups are lacking. Owing to lack of evidence of the benefits of preoperative MRI regarding short- and long-term treatment outcomes, the value of MRI is heavily debated. Unlike most European countries, guidelines are available in the Netherlands regarding the use of preoperative MRI in women for whom BCS is being considered. In 2011, preoperative MRI was recommended in patients with a (non)-invasive tumour with poorly defined margins, located in dense breast tissue, or with an extensive intraductal component11. Since 2012, MRI has been advised in patients with: an invasive tumour and a discrepancy in size between physical examination, mammography and ultrasound imaging; invasive lobular breast cancer; uncertainty regarding the extent of high-grade DCIS; or suspected (micro)invasive breast cancer in DCIS12. Despite these guidelines, extremely wide interhospital variation in the use of preoperative MRI still exists in the Netherlands, with a range of 0–85 per cent for women with invasive cancer13. Most studies on the potential benefits of preoperative MRI have described single-centre hospital cohorts, and no European population-based studies have been published. In the present study, the population-based Netherlands Cancer Registry was used to determine the association between preoperative MRI and initial mastectomy rate, surgical margin status after BCS, re-excision rate after BCS, and final mastectomy rate in subgroups of patients with invasive cancer, high-grade DCIS, non-palpable cancer, age 40 years or less, and lobular type of invasive cancer. The aim was to identify subgroups in which the preoperative use of MRI might result in improved surgical outcomes after BCS. Methods The population-based Netherlands Cancer Registry (Eindhoven region) registers all new cancer diagnoses in an area of south-east Netherlands that has 2·4 million inhabitants, and ten large community and teaching (but no academic) hospitals. The registry collects data based on notifications from the automated pathology archive (PALGA) according to international guidelines by specially trained personnel, and meets high-quality standards with completeness exceeding 95 per cent14,15. It provides detailed information on patient demographics, tumour characteristics and treatment. Since 2011, the registry has included data commissioned by the National Breast Cancer Audit, which includes information on preoperative MRI and the date it was performed13. In each hospital, dynamic contrast-enhanced MRI was performed according to local protocol, using various MRI scanners. Breast radiologists read the images using the Breast Imaging Reporting and Data System (BI-RADS). All patients, including their imaging findings, were discussed in preoperative multidisciplinary meetings. Patients diagnosed with a new invasive breast cancer or DCIS between 1 January 2011 and 1 January 2014 were selected. Exclusion criteria were: male sex, neoadjuvant chemotherapy or hormone therapy, clinical or pathological tumour stage T4, distant metastasis at presentation, unknown pathological tumour stage or T0, and unknown type of surgery or unknown surgical margin status. Clinical and pathological TNM stage was according to the seventh edition of the TNM staging system16. Margin status after the initial surgical procedure was registered in detail as negative, focally positive, or more than focally positive for both the non-invasive and invasive component of the tumour separately. Statistical analysis Patients with concurrent contralateral cancer or contralateral new primary cancer later in time were eligible for inclusion. The contralateral breast cancer was analysed as a new patient. The study population was divided into a no-MRI group and an MRI group according to preoperative use of MRI. Differences in patient characteristics between the two groups were tested using the Mann–Whitney U test for continuous variables and χ2 test for categorical variables. The association between MRI and time from diagnosis to surgery (in days) was determined with the Mann–Whitney U test. Subsequently, the total study population was allocated to none, one or more of the following subgroups: invasive cancer, purely high-grade DCIS (defined as Bloom and Richardson grade 2 or 3), non-palpable invasive cancer, young patients (40 years or less) at time of diagnosis, and lobular type of invasive cancer. A negative margin was defined as ‘no ink on tumour’, a focally positive margin as ‘tumour at the resection margin over a length of less than 4 mm’, and a more than focally positive margin as ‘tumour at the resection margin over a length of 4 mm or more’. Univariable and multivariable binary logistic regression analysis was used to test the association between MRI and the following outcomes: initial mastectomy rate (versus initial BCS), positive margin rate after BCS (versus negative margin after BCS), re-excision rate after BCS (versus no re-excision after BCS) and final mastectomy rate (versus final BCS). The multivariable model was performed by the enter method and included all variables displayed in Table 1 that were associated with the outcome of interest in univariable analysis with a P value of less than 0·100. Both age and tumour size were included as continuous variables in the univariable and multivariable regression analyses. To study the association between surgical margin status after BCS and MRI in more detail, χ2 analysis was also performed. Statistical tests were two-sided, and P < 0·050 was considered statistically significant. SPSS® version 20 (IBM, Armonk, New York, USA) was used for all statistical analyses. Table 1 Patient and tumour characteristics of the 4801 patients with invasive breast cancer of the total study population of 55144 . MRI (n = 1637) . No MRI (n = 3164) . P§ . Age (years) < 0·001 ≤ 40 102 (6·2) 76 (2·4) 41–59 767 (46·9) 1096 (34·6) ≥ 60 768 (46·9) 1992 (63·0) Palpability 0·120 No 676 (41·3) 1347 (42·6) Yes 920 (56·2) 1764 (55·8) Unknown 41 (2·5) 53 (1·7) Histology < 0·001 Ductal 1105 (67·5) 2712 (85·7) Lobular 449 (27·4) 231 (7·3) Other 83 (5·1) 221 (7·0) DCIS adjacent to tumour 0·006 No 790 (48·3) 1682 (53·2) Yes 846 (51·7) 1480 (46·8) Unknown 1 (0·1) 2 (0·1) Pathological tumour size (mm)* 16 (11–23) 15 (9–21) < 0·001¶ Tumour category < 0·001 T1 1126 (68·8) 2397 (75·8) T2 474 (29·0) 731 (23·1) T3 37 (2·3) 36 (1·1) Differentiation grade < 0·001 1 483 (29·5) 1076 (34·0) 2 752 (45·9) 1259 (39·8) 3 355 (21·7) 754 (23·8) Unknown 47 (2·9) 75 (2·4) Oestrogen receptor status 0·001 Positive 1438 (87·8) 2657 (84·0) Negative 184 (11·2) 481 (15·2) Unknown 15 (0·9) 26 (0·8) Progesterone receptor status 0·007 Positive 1200 (73·3) 2185 (69·1) Negative 422 (25·8) 952 (30·1) Unknown 15 (0·9) 27 (0·9) Her2/Neu receptor status 0·154 Negative 1417 (86·6) 2755 (87·1) Positive 189 (11·5) 327 (10·3) Unknown 31 (1·9) 82 (2·6) Node category 0·004 N0 1155 (70·6) 2292 (72·4) N1 348 (21·3) 627 (19·8) N2 71 (4·3) 102 (3·2) N3 36 (2·2) 49 (1·5) Unknown 27 (1·6) 94 (3·0) . MRI (n = 1637) . No MRI (n = 3164) . P§ . Age (years) < 0·001 ≤ 40 102 (6·2) 76 (2·4) 41–59 767 (46·9) 1096 (34·6) ≥ 60 768 (46·9) 1992 (63·0) Palpability 0·120 No 676 (41·3) 1347 (42·6) Yes 920 (56·2) 1764 (55·8) Unknown 41 (2·5) 53 (1·7) Histology < 0·001 Ductal 1105 (67·5) 2712 (85·7) Lobular 449 (27·4) 231 (7·3) Other 83 (5·1) 221 (7·0) DCIS adjacent to tumour 0·006 No 790 (48·3) 1682 (53·2) Yes 846 (51·7) 1480 (46·8) Unknown 1 (0·1) 2 (0·1) Pathological tumour size (mm)* 16 (11–23) 15 (9–21) < 0·001¶ Tumour category < 0·001 T1 1126 (68·8) 2397 (75·8) T2 474 (29·0) 731 (23·1) T3 37 (2·3) 36 (1·1) Differentiation grade < 0·001 1 483 (29·5) 1076 (34·0) 2 752 (45·9) 1259 (39·8) 3 355 (21·7) 754 (23·8) Unknown 47 (2·9) 75 (2·4) Oestrogen receptor status 0·001 Positive 1438 (87·8) 2657 (84·0) Negative 184 (11·2) 481 (15·2) Unknown 15 (0·9) 26 (0·8) Progesterone receptor status 0·007 Positive 1200 (73·3) 2185 (69·1) Negative 422 (25·8) 952 (30·1) Unknown 15 (0·9) 27 (0·9) Her2/Neu receptor status 0·154 Negative 1417 (86·6) 2755 (87·1) Positive 189 (11·5) 327 (10·3) Unknown 31 (1·9) 82 (2·6) Node category 0·004 N0 1155 (70·6) 2292 (72·4) N1 348 (21·3) 627 (19·8) N2 71 (4·3) 102 (3·2) N3 36 (2·2) 49 (1·5) Unknown 27 (1·6) 94 (3·0) Values in parentheses are percentages unless indicated otherwise; * values are median (i.q.r.). † Details of patients with ductal carcinoma in situ (DCIS) alone are described in the main text. ‡ Tumour size was not known in five patients in the MRI group and eight in the no-MRI group. § χ2 test, except ¶ Mann–Whitney U test. Open in new tab Table 1 Patient and tumour characteristics of the 4801 patients with invasive breast cancer of the total study population of 55144 . MRI (n = 1637) . No MRI (n = 3164) . P§ . Age (years) < 0·001 ≤ 40 102 (6·2) 76 (2·4) 41–59 767 (46·9) 1096 (34·6) ≥ 60 768 (46·9) 1992 (63·0) Palpability 0·120 No 676 (41·3) 1347 (42·6) Yes 920 (56·2) 1764 (55·8) Unknown 41 (2·5) 53 (1·7) Histology < 0·001 Ductal 1105 (67·5) 2712 (85·7) Lobular 449 (27·4) 231 (7·3) Other 83 (5·1) 221 (7·0) DCIS adjacent to tumour 0·006 No 790 (48·3) 1682 (53·2) Yes 846 (51·7) 1480 (46·8) Unknown 1 (0·1) 2 (0·1) Pathological tumour size (mm)* 16 (11–23) 15 (9–21) < 0·001¶ Tumour category < 0·001 T1 1126 (68·8) 2397 (75·8) T2 474 (29·0) 731 (23·1) T3 37 (2·3) 36 (1·1) Differentiation grade < 0·001 1 483 (29·5) 1076 (34·0) 2 752 (45·9) 1259 (39·8) 3 355 (21·7) 754 (23·8) Unknown 47 (2·9) 75 (2·4) Oestrogen receptor status 0·001 Positive 1438 (87·8) 2657 (84·0) Negative 184 (11·2) 481 (15·2) Unknown 15 (0·9) 26 (0·8) Progesterone receptor status 0·007 Positive 1200 (73·3) 2185 (69·1) Negative 422 (25·8) 952 (30·1) Unknown 15 (0·9) 27 (0·9) Her2/Neu receptor status 0·154 Negative 1417 (86·6) 2755 (87·1) Positive 189 (11·5) 327 (10·3) Unknown 31 (1·9) 82 (2·6) Node category 0·004 N0 1155 (70·6) 2292 (72·4) N1 348 (21·3) 627 (19·8) N2 71 (4·3) 102 (3·2) N3 36 (2·2) 49 (1·5) Unknown 27 (1·6) 94 (3·0) . MRI (n = 1637) . No MRI (n = 3164) . P§ . Age (years) < 0·001 ≤ 40 102 (6·2) 76 (2·4) 41–59 767 (46·9) 1096 (34·6) ≥ 60 768 (46·9) 1992 (63·0) Palpability 0·120 No 676 (41·3) 1347 (42·6) Yes 920 (56·2) 1764 (55·8) Unknown 41 (2·5) 53 (1·7) Histology < 0·001 Ductal 1105 (67·5) 2712 (85·7) Lobular 449 (27·4) 231 (7·3) Other 83 (5·1) 221 (7·0) DCIS adjacent to tumour 0·006 No 790 (48·3) 1682 (53·2) Yes 846 (51·7) 1480 (46·8) Unknown 1 (0·1) 2 (0·1) Pathological tumour size (mm)* 16 (11–23) 15 (9–21) < 0·001¶ Tumour category < 0·001 T1 1126 (68·8) 2397 (75·8) T2 474 (29·0) 731 (23·1) T3 37 (2·3) 36 (1·1) Differentiation grade < 0·001 1 483 (29·5) 1076 (34·0) 2 752 (45·9) 1259 (39·8) 3 355 (21·7) 754 (23·8) Unknown 47 (2·9) 75 (2·4) Oestrogen receptor status 0·001 Positive 1438 (87·8) 2657 (84·0) Negative 184 (11·2) 481 (15·2) Unknown 15 (0·9) 26 (0·8) Progesterone receptor status 0·007 Positive 1200 (73·3) 2185 (69·1) Negative 422 (25·8) 952 (30·1) Unknown 15 (0·9) 27 (0·9) Her2/Neu receptor status 0·154 Negative 1417 (86·6) 2755 (87·1) Positive 189 (11·5) 327 (10·3) Unknown 31 (1·9) 82 (2·6) Node category 0·004 N0 1155 (70·6) 2292 (72·4) N1 348 (21·3) 627 (19·8) N2 71 (4·3) 102 (3·2) N3 36 (2·2) 49 (1·5) Unknown 27 (1·6) 94 (3·0) Values in parentheses are percentages unless indicated otherwise; * values are median (i.q.r.). † Details of patients with ductal carcinoma in situ (DCIS) alone are described in the main text. ‡ Tumour size was not known in five patients in the MRI group and eight in the no-MRI group. § χ2 test, except ¶ Mann–Whitney U test. Open in new tab Results In 2011–2013, a total of 6685 patients with a new diagnosis of invasive breast cancer or DCIS were registered in the Netherlands Cancer Registry (Eindhoven region). After applying the exclusion criteria, 5514 patients were eligible. Invasive cancer was diagnosed in 4801 women (87·1 per cent), of whom 1637 (34·1 per cent) had preoperative MRI; their characteristics are summarized in Table 1. Pure DCIS was diagnosed in 713 patients (12·9 per cent), of whom 150 (21·0 per cent) had preoperative MRI. Of the patients with DCIS in the MRI and no-MRI group, 12 (8·0 per cent) and seven (1·2 per cent) respectively were aged 40 years or less, and 60 (40·0 per cent) and 298 (52·9 per cent) were aged 60 years or over (P < 0·001). Furthermore, of the patients with DCIS in the MRI and no-MRI group, 12 (8·0 per cent) and 82 (14·6 per cent) respectively had differentiation grade 1 disease, 55 (36·7 per cent) and 217 (38·5 per cent) had differentiation grade 2, and 81 (54·0 per cent) and 261 (46·4 per cent) had differentiation grade 3; the grade was unknown in two (1·3 per cent) and three (0·5 per cent) patients respectively (P = 0·089). In the total study population, the incidence of contralateral invasive breast cancer or DCIS diagnosed within 3 months after diagnosis of the first invasive breast cancer or DCIS was 58 (3·2 per cent) in the MRI group and 45 (1·2 per cent) in the no-MRI group. The time between diagnosis and surgery in the total study population was longer for the MRI group: median (i.q.r.) 34 (24–45) days versus 22 (15–30) days in the no-MRI group (P < 0·001). The median time from MRI to surgery was 24 (15–36) days. Preoperative MRI and initial mastectomy rate In the total study population, 1480 patients (26·8 per cent) were initially treated with mastectomy, 651 (36·4 per cent) of the 1787 patients in the MRI group and 829 (22·2 per cent) of the 3727 in the no-MRI group (both unadjusted and adjusted P < 0·001) (Table 2). Likewise, significantly higher initial mastectomy rates were seen in MRI versus no-MRI groups in the subgroups of patients with invasive cancer (35·9 versus 23·1 per cent respectively), high-grade DCIS (43·4 versus 18·2 per cent) and non-palpable cancer (28·8 versus 11·8 per cent) (all unadjusted and adjusted P < 0·001). In contrast, initial mastectomy rates were not significantly different between MRI and no-MRI groups in patients aged 40 years or less (40·4 versus 42 per cent), or in patients with lobular type of cancer (41·2 versus 40·7 per cent) (all unadjusted and adjusted P > 0·200) (Table 2). Table 2 Univariable and multivariable logistic regression analysis for initial mastectomy rate, positive margin rate after breast-conserving surgery, re-excision rate after BCS and final mastectomy rate, according to preoperative MRI use . Total* . MRI* . No MRI* . Unadjusted OR† . Unadjusted P . Adjusted OR† . Adjusted P . Total study population‡ Initial mastectomy 1480 of 5514 (26·8) 651 of 1787 (36·4) 829 of 3727 (22·2) 2·00 (1·77, 2·27) < 0·001 2·18 (1·92, 2·48) < 0·001 Positive margin 645 of 4034 (16·0) 206 of 1136 (18·1) 439 of 2898 (15·1) 1·24 (1·03, 1·49) 0·020 1·20 (1·00, 1·45) 0·052 Re-excision 321 of 4034 (8·0) 111 of 1136 (9·8) 210 of 2898 (7·2) 1·39 (1·09, 1·76) 0·008 1·33 (1·04, 1·70) 0·026 Final mastectomy 1595 of 5514 (28·9) 693 of 1787 (38·8) 902 of 3727 (24·2) 1·99 (1·76, 2·24) < 0·001 2·13 (1·87, 2·41) < 0·001 Invasive cancer Initial mastectomy§ 1318 of 4801 (27·5) 588 of 1637 (35·9) 730 of 3164 (23·1) 1·87 (1·64, 2·13) < 0·001 1·80 (1·54, 2·09) < 0·001 Positive margin¶ 548 of 3483 (15·7) 188 of 1049 (17·9) 360 of 2434 (14·8) 1·26 (1·04, 1·53) 0·020 0·98 (0·79, 1·22) 0·882 Re-excision# 239 of 3483 (6·9) 95 of 1049 (9·1) 144 of 2434 (5·9) 1·58 (1·21, 2·08) 0·001 1·27 (0·94, 1·72) 0·125 Final mastectomy§ 1406 of 4801 (29·3) 623 of 1637 (38·1) 783 of 3164 (24·7) 1·87 (1·64, 2·12) < 0·001 1·74 (1·50, 2·03) < 0·001 High-grade DCIS Initial mastectomy** 146 of 614 (23·8) 59 of 136 (43·4) 87 of 478 (18·2) 3·44 (2·28, 5·20) <0·001 3·18 (2·09, 4·82) < 0·001 Positive margin‡ 90 of 468 (19·2) 18 of 77 (23) 72 of 391 (18·4) 1·35 (0·75, 2·43) 0·314 1·28 (0·70, 2·32) 0·426 Re-excision‡ 75 of 468 (16·0) 16 of 77 (21) 59 of 391 (15·1) 1·48 (0·80, 2·73) 0·216 1·38 (0·73, 2·59) 0·320 Final mastectomy** 171 of 614 (27·9) 66 of 136 (48·5) 105 of 478 (22·0) 3·35 (2·25, 5·00) < 0·001 3·11 (2·07, 4·66) < 0·001 Non-palpable invasive cancer Initial mastectomy†† 354 of 2023 (17·5) 195 of 676 (28·8) 159 of 1347 (11·8) 3·03 (2·40, 3·83) < 0·001 2·68 (2·05, 3·50) < 0·001 Positive margin‡‡ 229 of 1669 (13·7) 74 of 481 (15·4) 155 of 1188 (13·0) 1·21 (0·90, 1·64) 0·209 0·92 (0·66, 1·29) 0·645 Re-excision§§ 97 of 1669 (5·8) 38 of 481 (7·9) 59 of 1188 (5·0) 1·64 (1·08, 2·50) 0·021 1·33 (0·83, 2·13) 0·234 Final mastectomy†† 380 of 2023 (18·8) 207 of 676 (30·6) 173 of 1347 (12·8) 3·00 (2·38, 3·76) < 0·001 2·58 (1·99, 3·47) < 0·001 Age ≤ 40 years Initial mastectomy‡ 81 of 197 (41·1) 46 of 114 (40·4) 35 of 83 (42) 0·93 (0·52, 1·65) 0·798 0·68 (0·37, 1·27) 0·226 Positive margin¶¶ 24 of 116 (20·7) 16 of 68 (24) 8 of 48 (17) 1·54 (0·60, 3·95) 0·371 1·43 (0·55, 3·76) 0·463 Re-excision## 13 of 116 (11·2) 9 of 68 (13) 4 of 48 (8) 1·68 (0·49, 5·80) 0·414 – – Final mastectomy‡ 90 of 197 (45·7) 52 of 114 (45·6) 38 of 83 (46) 0·99 (0·56, 1·75) 0·981 0·75 (0·41, 1·39) 0·358 Lobular invasive cancer Initial mastectomy*** 279 of 680 (41·0) 185 of 449 (41·2) 94 of 231 (40·7) 1·02 (0·74, 1·41) 0·898 1·00 (0·68, 1·45) 0·977 Positive margin††† 102 of 401 (25·4) 65 of 264 (24·6) 37 of 137 (27·0) 0·88 (0·55, 1·41) 0·603 0·80 (0·47, 1·38) 0·419 Re-excision# 45 of 401 (11·2) 29 of 264 (11·0) 16 of 137 (11·7) 0·93 (0·49, 1·79) 0·835 0·97 (0·44, 2·12) 0·933 Final mastectomy*** 301 of 680 (44·3) 198 of 449 (44·1) 103 of 231 (44·6) 0·98 (0·71, 1·35) 0·903 0·95 (0·65, 1·39) 0·791 . Total* . MRI* . No MRI* . Unadjusted OR† . Unadjusted P . Adjusted OR† . Adjusted P . Total study population‡ Initial mastectomy 1480 of 5514 (26·8) 651 of 1787 (36·4) 829 of 3727 (22·2) 2·00 (1·77, 2·27) < 0·001 2·18 (1·92, 2·48) < 0·001 Positive margin 645 of 4034 (16·0) 206 of 1136 (18·1) 439 of 2898 (15·1) 1·24 (1·03, 1·49) 0·020 1·20 (1·00, 1·45) 0·052 Re-excision 321 of 4034 (8·0) 111 of 1136 (9·8) 210 of 2898 (7·2) 1·39 (1·09, 1·76) 0·008 1·33 (1·04, 1·70) 0·026 Final mastectomy 1595 of 5514 (28·9) 693 of 1787 (38·8) 902 of 3727 (24·2) 1·99 (1·76, 2·24) < 0·001 2·13 (1·87, 2·41) < 0·001 Invasive cancer Initial mastectomy§ 1318 of 4801 (27·5) 588 of 1637 (35·9) 730 of 3164 (23·1) 1·87 (1·64, 2·13) < 0·001 1·80 (1·54, 2·09) < 0·001 Positive margin¶ 548 of 3483 (15·7) 188 of 1049 (17·9) 360 of 2434 (14·8) 1·26 (1·04, 1·53) 0·020 0·98 (0·79, 1·22) 0·882 Re-excision# 239 of 3483 (6·9) 95 of 1049 (9·1) 144 of 2434 (5·9) 1·58 (1·21, 2·08) 0·001 1·27 (0·94, 1·72) 0·125 Final mastectomy§ 1406 of 4801 (29·3) 623 of 1637 (38·1) 783 of 3164 (24·7) 1·87 (1·64, 2·12) < 0·001 1·74 (1·50, 2·03) < 0·001 High-grade DCIS Initial mastectomy** 146 of 614 (23·8) 59 of 136 (43·4) 87 of 478 (18·2) 3·44 (2·28, 5·20) <0·001 3·18 (2·09, 4·82) < 0·001 Positive margin‡ 90 of 468 (19·2) 18 of 77 (23) 72 of 391 (18·4) 1·35 (0·75, 2·43) 0·314 1·28 (0·70, 2·32) 0·426 Re-excision‡ 75 of 468 (16·0) 16 of 77 (21) 59 of 391 (15·1) 1·48 (0·80, 2·73) 0·216 1·38 (0·73, 2·59) 0·320 Final mastectomy** 171 of 614 (27·9) 66 of 136 (48·5) 105 of 478 (22·0) 3·35 (2·25, 5·00) < 0·001 3·11 (2·07, 4·66) < 0·001 Non-palpable invasive cancer Initial mastectomy†† 354 of 2023 (17·5) 195 of 676 (28·8) 159 of 1347 (11·8) 3·03 (2·40, 3·83) < 0·001 2·68 (2·05, 3·50) < 0·001 Positive margin‡‡ 229 of 1669 (13·7) 74 of 481 (15·4) 155 of 1188 (13·0) 1·21 (0·90, 1·64) 0·209 0·92 (0·66, 1·29) 0·645 Re-excision§§ 97 of 1669 (5·8) 38 of 481 (7·9) 59 of 1188 (5·0) 1·64 (1·08, 2·50) 0·021 1·33 (0·83, 2·13) 0·234 Final mastectomy†† 380 of 2023 (18·8) 207 of 676 (30·6) 173 of 1347 (12·8) 3·00 (2·38, 3·76) < 0·001 2·58 (1·99, 3·47) < 0·001 Age ≤ 40 years Initial mastectomy‡ 81 of 197 (41·1) 46 of 114 (40·4) 35 of 83 (42) 0·93 (0·52, 1·65) 0·798 0·68 (0·37, 1·27) 0·226 Positive margin¶¶ 24 of 116 (20·7) 16 of 68 (24) 8 of 48 (17) 1·54 (0·60, 3·95) 0·371 1·43 (0·55, 3·76) 0·463 Re-excision## 13 of 116 (11·2) 9 of 68 (13) 4 of 48 (8) 1·68 (0·49, 5·80) 0·414 – – Final mastectomy‡ 90 of 197 (45·7) 52 of 114 (45·6) 38 of 83 (46) 0·99 (0·56, 1·75) 0·981 0·75 (0·41, 1·39) 0·358 Lobular invasive cancer Initial mastectomy*** 279 of 680 (41·0) 185 of 449 (41·2) 94 of 231 (40·7) 1·02 (0·74, 1·41) 0·898 1·00 (0·68, 1·45) 0·977 Positive margin††† 102 of 401 (25·4) 65 of 264 (24·6) 37 of 137 (27·0) 0·88 (0·55, 1·41) 0·603 0·80 (0·47, 1·38) 0·419 Re-excision# 45 of 401 (11·2) 29 of 264 (11·0) 16 of 137 (11·7) 0·93 (0·49, 1·79) 0·835 0·97 (0·44, 2·12) 0·933 Final mastectomy*** 301 of 680 (44·3) 198 of 449 (44·1) 103 of 231 (44·6) 0·98 (0·71, 1·35) 0·903 0·95 (0·65, 1·39) 0·791 Values in parentheses are * percentages and † 95 per cent c.i. Adjustment for variables associated with (initial and final) mastectomy, positive resection margin, and re-excision with P < 0·100 in univariable analysis: ‡ age and differentiation grade, § age, palpability, histology, tumour size, differentiation grade, oestrogen receptor status, progesterone receptor status, Her2/Neu receptor status and regional lymph node status, ¶ age, palpability, histology, presence of ductal carcinoma in situ (DCIS) component, tumour size, differentiation grade, oestrogen receptor status and regional lymph node status, # age, palpability, histology, presence of DCIS component, tumour size, differentiation grade and regional lymph node status, ** age, †† age, histology, presence of DCIS component, tumour size, differentiation grade, oestrogen receptor status, progesterone receptor status, Her2/Neu receptor status and regional lymph node status, ‡‡ histology, presence of DCIS component, tumour size, differentiation grade and regional lymph node status, §§ histology, presence of DCIS component, tumour size, differentiation grade, progesterone receptor status, Her2/Neu receptor status and regional lymph node status, ¶¶ differentiation grade, ## none, *** palpability, tumour size, differentiation grade and regional lymph node status, ††† presence of DCIS component, tumour size, Her2/Neu receptor status and regional lymph node status. OR, odds ratio; BCS, breast-conserving surgery. Open in new tab Table 2 Univariable and multivariable logistic regression analysis for initial mastectomy rate, positive margin rate after breast-conserving surgery, re-excision rate after BCS and final mastectomy rate, according to preoperative MRI use . Total* . MRI* . No MRI* . Unadjusted OR† . Unadjusted P . Adjusted OR† . Adjusted P . Total study population‡ Initial mastectomy 1480 of 5514 (26·8) 651 of 1787 (36·4) 829 of 3727 (22·2) 2·00 (1·77, 2·27) < 0·001 2·18 (1·92, 2·48) < 0·001 Positive margin 645 of 4034 (16·0) 206 of 1136 (18·1) 439 of 2898 (15·1) 1·24 (1·03, 1·49) 0·020 1·20 (1·00, 1·45) 0·052 Re-excision 321 of 4034 (8·0) 111 of 1136 (9·8) 210 of 2898 (7·2) 1·39 (1·09, 1·76) 0·008 1·33 (1·04, 1·70) 0·026 Final mastectomy 1595 of 5514 (28·9) 693 of 1787 (38·8) 902 of 3727 (24·2) 1·99 (1·76, 2·24) < 0·001 2·13 (1·87, 2·41) < 0·001 Invasive cancer Initial mastectomy§ 1318 of 4801 (27·5) 588 of 1637 (35·9) 730 of 3164 (23·1) 1·87 (1·64, 2·13) < 0·001 1·80 (1·54, 2·09) < 0·001 Positive margin¶ 548 of 3483 (15·7) 188 of 1049 (17·9) 360 of 2434 (14·8) 1·26 (1·04, 1·53) 0·020 0·98 (0·79, 1·22) 0·882 Re-excision# 239 of 3483 (6·9) 95 of 1049 (9·1) 144 of 2434 (5·9) 1·58 (1·21, 2·08) 0·001 1·27 (0·94, 1·72) 0·125 Final mastectomy§ 1406 of 4801 (29·3) 623 of 1637 (38·1) 783 of 3164 (24·7) 1·87 (1·64, 2·12) < 0·001 1·74 (1·50, 2·03) < 0·001 High-grade DCIS Initial mastectomy** 146 of 614 (23·8) 59 of 136 (43·4) 87 of 478 (18·2) 3·44 (2·28, 5·20) <0·001 3·18 (2·09, 4·82) < 0·001 Positive margin‡ 90 of 468 (19·2) 18 of 77 (23) 72 of 391 (18·4) 1·35 (0·75, 2·43) 0·314 1·28 (0·70, 2·32) 0·426 Re-excision‡ 75 of 468 (16·0) 16 of 77 (21) 59 of 391 (15·1) 1·48 (0·80, 2·73) 0·216 1·38 (0·73, 2·59) 0·320 Final mastectomy** 171 of 614 (27·9) 66 of 136 (48·5) 105 of 478 (22·0) 3·35 (2·25, 5·00) < 0·001 3·11 (2·07, 4·66) < 0·001 Non-palpable invasive cancer Initial mastectomy†† 354 of 2023 (17·5) 195 of 676 (28·8) 159 of 1347 (11·8) 3·03 (2·40, 3·83) < 0·001 2·68 (2·05, 3·50) < 0·001 Positive margin‡‡ 229 of 1669 (13·7) 74 of 481 (15·4) 155 of 1188 (13·0) 1·21 (0·90, 1·64) 0·209 0·92 (0·66, 1·29) 0·645 Re-excision§§ 97 of 1669 (5·8) 38 of 481 (7·9) 59 of 1188 (5·0) 1·64 (1·08, 2·50) 0·021 1·33 (0·83, 2·13) 0·234 Final mastectomy†† 380 of 2023 (18·8) 207 of 676 (30·6) 173 of 1347 (12·8) 3·00 (2·38, 3·76) < 0·001 2·58 (1·99, 3·47) < 0·001 Age ≤ 40 years Initial mastectomy‡ 81 of 197 (41·1) 46 of 114 (40·4) 35 of 83 (42) 0·93 (0·52, 1·65) 0·798 0·68 (0·37, 1·27) 0·226 Positive margin¶¶ 24 of 116 (20·7) 16 of 68 (24) 8 of 48 (17) 1·54 (0·60, 3·95) 0·371 1·43 (0·55, 3·76) 0·463 Re-excision## 13 of 116 (11·2) 9 of 68 (13) 4 of 48 (8) 1·68 (0·49, 5·80) 0·414 – – Final mastectomy‡ 90 of 197 (45·7) 52 of 114 (45·6) 38 of 83 (46) 0·99 (0·56, 1·75) 0·981 0·75 (0·41, 1·39) 0·358 Lobular invasive cancer Initial mastectomy*** 279 of 680 (41·0) 185 of 449 (41·2) 94 of 231 (40·7) 1·02 (0·74, 1·41) 0·898 1·00 (0·68, 1·45) 0·977 Positive margin††† 102 of 401 (25·4) 65 of 264 (24·6) 37 of 137 (27·0) 0·88 (0·55, 1·41) 0·603 0·80 (0·47, 1·38) 0·419 Re-excision# 45 of 401 (11·2) 29 of 264 (11·0) 16 of 137 (11·7) 0·93 (0·49, 1·79) 0·835 0·97 (0·44, 2·12) 0·933 Final mastectomy*** 301 of 680 (44·3) 198 of 449 (44·1) 103 of 231 (44·6) 0·98 (0·71, 1·35) 0·903 0·95 (0·65, 1·39) 0·791 . Total* . MRI* . No MRI* . Unadjusted OR† . Unadjusted P . Adjusted OR† . Adjusted P . Total study population‡ Initial mastectomy 1480 of 5514 (26·8) 651 of 1787 (36·4) 829 of 3727 (22·2) 2·00 (1·77, 2·27) < 0·001 2·18 (1·92, 2·48) < 0·001 Positive margin 645 of 4034 (16·0) 206 of 1136 (18·1) 439 of 2898 (15·1) 1·24 (1·03, 1·49) 0·020 1·20 (1·00, 1·45) 0·052 Re-excision 321 of 4034 (8·0) 111 of 1136 (9·8) 210 of 2898 (7·2) 1·39 (1·09, 1·76) 0·008 1·33 (1·04, 1·70) 0·026 Final mastectomy 1595 of 5514 (28·9) 693 of 1787 (38·8) 902 of 3727 (24·2) 1·99 (1·76, 2·24) < 0·001 2·13 (1·87, 2·41) < 0·001 Invasive cancer Initial mastectomy§ 1318 of 4801 (27·5) 588 of 1637 (35·9) 730 of 3164 (23·1) 1·87 (1·64, 2·13) < 0·001 1·80 (1·54, 2·09) < 0·001 Positive margin¶ 548 of 3483 (15·7) 188 of 1049 (17·9) 360 of 2434 (14·8) 1·26 (1·04, 1·53) 0·020 0·98 (0·79, 1·22) 0·882 Re-excision# 239 of 3483 (6·9) 95 of 1049 (9·1) 144 of 2434 (5·9) 1·58 (1·21, 2·08) 0·001 1·27 (0·94, 1·72) 0·125 Final mastectomy§ 1406 of 4801 (29·3) 623 of 1637 (38·1) 783 of 3164 (24·7) 1·87 (1·64, 2·12) < 0·001 1·74 (1·50, 2·03) < 0·001 High-grade DCIS Initial mastectomy** 146 of 614 (23·8) 59 of 136 (43·4) 87 of 478 (18·2) 3·44 (2·28, 5·20) <0·001 3·18 (2·09, 4·82) < 0·001 Positive margin‡ 90 of 468 (19·2) 18 of 77 (23) 72 of 391 (18·4) 1·35 (0·75, 2·43) 0·314 1·28 (0·70, 2·32) 0·426 Re-excision‡ 75 of 468 (16·0) 16 of 77 (21) 59 of 391 (15·1) 1·48 (0·80, 2·73) 0·216 1·38 (0·73, 2·59) 0·320 Final mastectomy** 171 of 614 (27·9) 66 of 136 (48·5) 105 of 478 (22·0) 3·35 (2·25, 5·00) < 0·001 3·11 (2·07, 4·66) < 0·001 Non-palpable invasive cancer Initial mastectomy†† 354 of 2023 (17·5) 195 of 676 (28·8) 159 of 1347 (11·8) 3·03 (2·40, 3·83) < 0·001 2·68 (2·05, 3·50) < 0·001 Positive margin‡‡ 229 of 1669 (13·7) 74 of 481 (15·4) 155 of 1188 (13·0) 1·21 (0·90, 1·64) 0·209 0·92 (0·66, 1·29) 0·645 Re-excision§§ 97 of 1669 (5·8) 38 of 481 (7·9) 59 of 1188 (5·0) 1·64 (1·08, 2·50) 0·021 1·33 (0·83, 2·13) 0·234 Final mastectomy†† 380 of 2023 (18·8) 207 of 676 (30·6) 173 of 1347 (12·8) 3·00 (2·38, 3·76) < 0·001 2·58 (1·99, 3·47) < 0·001 Age ≤ 40 years Initial mastectomy‡ 81 of 197 (41·1) 46 of 114 (40·4) 35 of 83 (42) 0·93 (0·52, 1·65) 0·798 0·68 (0·37, 1·27) 0·226 Positive margin¶¶ 24 of 116 (20·7) 16 of 68 (24) 8 of 48 (17) 1·54 (0·60, 3·95) 0·371 1·43 (0·55, 3·76) 0·463 Re-excision## 13 of 116 (11·2) 9 of 68 (13) 4 of 48 (8) 1·68 (0·49, 5·80) 0·414 – – Final mastectomy‡ 90 of 197 (45·7) 52 of 114 (45·6) 38 of 83 (46) 0·99 (0·56, 1·75) 0·981 0·75 (0·41, 1·39) 0·358 Lobular invasive cancer Initial mastectomy*** 279 of 680 (41·0) 185 of 449 (41·2) 94 of 231 (40·7) 1·02 (0·74, 1·41) 0·898 1·00 (0·68, 1·45) 0·977 Positive margin††† 102 of 401 (25·4) 65 of 264 (24·6) 37 of 137 (27·0) 0·88 (0·55, 1·41) 0·603 0·80 (0·47, 1·38) 0·419 Re-excision# 45 of 401 (11·2) 29 of 264 (11·0) 16 of 137 (11·7) 0·93 (0·49, 1·79) 0·835 0·97 (0·44, 2·12) 0·933 Final mastectomy*** 301 of 680 (44·3) 198 of 449 (44·1) 103 of 231 (44·6) 0·98 (0·71, 1·35) 0·903 0·95 (0·65, 1·39) 0·791 Values in parentheses are * percentages and † 95 per cent c.i. Adjustment for variables associated with (initial and final) mastectomy, positive resection margin, and re-excision with P < 0·100 in univariable analysis: ‡ age and differentiation grade, § age, palpability, histology, tumour size, differentiation grade, oestrogen receptor status, progesterone receptor status, Her2/Neu receptor status and regional lymph node status, ¶ age, palpability, histology, presence of ductal carcinoma in situ (DCIS) component, tumour size, differentiation grade, oestrogen receptor status and regional lymph node status, # age, palpability, histology, presence of DCIS component, tumour size, differentiation grade and regional lymph node status, ** age, †† age, histology, presence of DCIS component, tumour size, differentiation grade, oestrogen receptor status, progesterone receptor status, Her2/Neu receptor status and regional lymph node status, ‡‡ histology, presence of DCIS component, tumour size, differentiation grade and regional lymph node status, §§ histology, presence of DCIS component, tumour size, differentiation grade, progesterone receptor status, Her2/Neu receptor status and regional lymph node status, ¶¶ differentiation grade, ## none, *** palpability, tumour size, differentiation grade and regional lymph node status, ††† presence of DCIS component, tumour size, Her2/Neu receptor status and regional lymph node status. OR, odds ratio; BCS, breast-conserving surgery. Open in new tab Preoperative MRI and margin status BCS was performed in 4034 (73·2 per cent) of the total study population. In the MRI group, a focally positive margin was found more frequently than a more than focally positive margin (P = 0·048) (Table 3). However, MRI was not significantly associated with a positive margin after adjustment for possible confounders (odds ratio (OR) 1·20, 95 per cent c.i. 1·00 to 1·45; P = 0·052) (Table 2). In all subgroups, no differences in negative, focally positive, or more than focally positive margin rates were seen between MRI and no-MRI groups in univariable analysis (all P > 0·050) (Table 3). In addition, in multivariable analysis MRI was not associated with a positive surgical margin (all P > 0·050) (Table 2). In patients with lobular type of cancer, preoperative MRI was more frequently associated with a negative margin (75·4 per cent versus 73·0 per cent in the no-MRI group), and more rarely with a more than focally positive margin (5·7 and 8·8 per cent respectively) (Table 3). However, this difference was not significant in univariable analysis (P = 0·507) (Table 3) or multivariable analysis (OR 0·80, 95 per cent c.i. 0·47 to 1·38; P = 0·419) (Table 2). Table 3 Surgical resection margin after breast-conserving surgery according to preoperative use of MRI Surgical resection margin . MRI . No MRI . P* . Total study population n = 1136 n = 2898 0·048 Negative 930 (81·9) 2459 (84·9) Focally positive 147 (12·9) 326 (11·2) More than focally positive 59 (5·2) 113 (3·9) Invasive cancer n = 1049 n = 2434 0·062 Negative 861 (82·1) 2074 (85·2) Focally positive 135 (12·9) 264 (10·8) More than focally positive 53 (5·1) 96 (3·9) High-grade DCIS n = 77 n = 391 0·396 Negative 59 (77) 319 (81·6) Focally positive 12 (16) 55 (14·1) More than focally positive 6 (8) 17 (4·3) Non-palpable invasive cancer n = 481 n = 1188 0·312 Negative 407 (84·6) 1033 (87·0) Focally positive 52 (10·8) 117 (9·8) More than focally positive 22 (4·6) 38 (3·2) Age ≤ 40 years n = 68 n = 48 0·638 Negative 52 (76) 40 (83) Focally positive 11 (16) 6 (13) More than focally positive 5 (7) 2 (4) Lobular invasive cancer n = 264 n = 137 0·507 Negative 199 (75·4) 100 (73·0) Focally positive 50 (18·9) 25 (18·2) More than focally positive 15 (5·7) 12 (8·8) Surgical resection margin . MRI . No MRI . P* . Total study population n = 1136 n = 2898 0·048 Negative 930 (81·9) 2459 (84·9) Focally positive 147 (12·9) 326 (11·2) More than focally positive 59 (5·2) 113 (3·9) Invasive cancer n = 1049 n = 2434 0·062 Negative 861 (82·1) 2074 (85·2) Focally positive 135 (12·9) 264 (10·8) More than focally positive 53 (5·1) 96 (3·9) High-grade DCIS n = 77 n = 391 0·396 Negative 59 (77) 319 (81·6) Focally positive 12 (16) 55 (14·1) More than focally positive 6 (8) 17 (4·3) Non-palpable invasive cancer n = 481 n = 1188 0·312 Negative 407 (84·6) 1033 (87·0) Focally positive 52 (10·8) 117 (9·8) More than focally positive 22 (4·6) 38 (3·2) Age ≤ 40 years n = 68 n = 48 0·638 Negative 52 (76) 40 (83) Focally positive 11 (16) 6 (13) More than focally positive 5 (7) 2 (4) Lobular invasive cancer n = 264 n = 137 0·507 Negative 199 (75·4) 100 (73·0) Focally positive 50 (18·9) 25 (18·2) More than focally positive 15 (5·7) 12 (8·8) Values in parentheses are percentages. DCIS, ductal carcinoma in situ. * χ2 test. Open in new tab Table 3 Surgical resection margin after breast-conserving surgery according to preoperative use of MRI Surgical resection margin . MRI . No MRI . P* . Total study population n = 1136 n = 2898 0·048 Negative 930 (81·9) 2459 (84·9) Focally positive 147 (12·9) 326 (11·2) More than focally positive 59 (5·2) 113 (3·9) Invasive cancer n = 1049 n = 2434 0·062 Negative 861 (82·1) 2074 (85·2) Focally positive 135 (12·9) 264 (10·8) More than focally positive 53 (5·1) 96 (3·9) High-grade DCIS n = 77 n = 391 0·396 Negative 59 (77) 319 (81·6) Focally positive 12 (16) 55 (14·1) More than focally positive 6 (8) 17 (4·3) Non-palpable invasive cancer n = 481 n = 1188 0·312 Negative 407 (84·6) 1033 (87·0) Focally positive 52 (10·8) 117 (9·8) More than focally positive 22 (4·6) 38 (3·2) Age ≤ 40 years n = 68 n = 48 0·638 Negative 52 (76) 40 (83) Focally positive 11 (16) 6 (13) More than focally positive 5 (7) 2 (4) Lobular invasive cancer n = 264 n = 137 0·507 Negative 199 (75·4) 100 (73·0) Focally positive 50 (18·9) 25 (18·2) More than focally positive 15 (5·7) 12 (8·8) Surgical resection margin . MRI . No MRI . P* . Total study population n = 1136 n = 2898 0·048 Negative 930 (81·9) 2459 (84·9) Focally positive 147 (12·9) 326 (11·2) More than focally positive 59 (5·2) 113 (3·9) Invasive cancer n = 1049 n = 2434 0·062 Negative 861 (82·1) 2074 (85·2) Focally positive 135 (12·9) 264 (10·8) More than focally positive 53 (5·1) 96 (3·9) High-grade DCIS n = 77 n = 391 0·396 Negative 59 (77) 319 (81·6) Focally positive 12 (16) 55 (14·1) More than focally positive 6 (8) 17 (4·3) Non-palpable invasive cancer n = 481 n = 1188 0·312 Negative 407 (84·6) 1033 (87·0) Focally positive 52 (10·8) 117 (9·8) More than focally positive 22 (4·6) 38 (3·2) Age ≤ 40 years n = 68 n = 48 0·638 Negative 52 (76) 40 (83) Focally positive 11 (16) 6 (13) More than focally positive 5 (7) 2 (4) Lobular invasive cancer n = 264 n = 137 0·507 Negative 199 (75·4) 100 (73·0) Focally positive 50 (18·9) 25 (18·2) More than focally positive 15 (5·7) 12 (8·8) Values in parentheses are percentages. DCIS, ductal carcinoma in situ. * χ2 test. Open in new tab Preoperative MRI and re-excision rate Re-excision after BCS was performed in 111 (9·8 per cent) and 210 (7·2 per cent) patients in the MRI and no-MRI group respectively (unadjusted OR 1·39, 95 per cent c.i. 1·09 to 1·76; P = 0·008), and remained significantly different after adjustment for age and differentiation grade (OR 1·33, 1·04 to 1·70; P = 0·026) (Table 2). In the subgroup with invasive cancer, re-excision was needed in 9·1 and 5·9 per cent of patients in the MRI and no-MRI group respectively (unadjusted P = 0·001), and in those with non-palpable cancer re-excision was required in 7·9 and 5·0 per cent respectively (unadjusted P = 0·021). However, these differences were not significant after multivariable analysis (OR 1·27, 0·94 to 1·72, P = 0·125, and OR 1·33, 0·83 to 2·13, P = 0·234, respectively). The number of re-excisions was not significantly different between the MRI and no-MRI group in patients with high-grade DCIS (21 versus 15·1 per cent respectively), patients aged 40 years or less (13 versus 8 per cent), and patients with lobular type of cancer (11·0 versus 11·7 per cent) (all unadjusted and adjusted P > 0·200). Preoperative MRI and final mastectomy rate Including re-excisions, 1595 (28·9 per cent) of the 5514 patients in the total study population finally had a mastectomy: 693 (38·8 per cent) of the 1787 patients in the MRI group and 902 (24·2 per cent) of 3727 in the no-MRI group (both unadjusted and adjusted P < 0·001) (Table 2). In addition, significantly higher final mastectomy rates were seen in patients with MRI than without MRI in the subgroups of invasive cancer (38·1 versus 24·7 per cent respectively), high-grade DCIS (48·5 versus 22·0 per cent) and non-palpable cancer (30·6 versus 12·8 per cent) (all unadjusted and adjusted P < 0·001). However, final mastectomy rates were similar in MRI and no-MRI groups in patients aged 40 years or less (45·6 versus 46 per cent respectively) and in those with lobular type of cancer (44·1 versus 44·6 per cent) (all unadjusted and adjusted P > 0·300). Discussion The hypothesis of this retrospective study in patients with invasive or non-invasive breast cancer was that there are subgroups in which the preoperative use of MRI will result in improved surgical outcomes after BCS. In the total study population, patients who underwent MRI had a small, but significantly higher, positive margin rate (18·1 per cent versus 15·1 per cent in those who did not have MRI). However, this does not imply a clinically relevant difference, and in multivariable analysis the difference was no longer significant (P = 0·052). Preoperative MRI was associated with neither improved resection margins nor lower re-excision rates after BCS in any of the subgroups studied. However, MRI was associated with higher initial and final mastectomy rates, except in the subgroup of younger patients (aged 40 years or less) and in patients with lobular type of invasive breast cancer. This is a large, multicentre, population-based analysis of the effect of preoperative MRI on surgical outcomes in Europe. Multivariable analyses were performed, adjusting for multiple patient, tumour and treatment characteristics associated with each outcome separately, which is one of the strengths of the study. Another strength is the large number of patients, which made it possible to focus on patient subgroups, which so far have been studied insufficiently. The association between preoperative MRI and resection margins after BCS has not been studied in detail before. Contradictory results have been found in multiple small, single-centre studies that have focused on the surgical outcomes of BCS in patients with invasive breast cancer17–22. Similar to the findings in the present study, in the only other population-based study, by Fortune-Greeley and colleagues6, preoperative MRI was not associated with improved surgical outcomes. Only small, single-centre studies with contradictory results have reported on preoperative MRI in patients with pure DCIS23–27. In line with the present findings, the only population-based study, performed by Wang and co-workers7, showed no association between preoperative MRI and positive resection margin and re-excision rates after BCS. To date, only the MONET trial28 has focused on patients with non-palpable invasive cancer. In contrast to the present findings, the MONET trial (based on 149 patients) found that the addition of preoperative MRI to routine clinical care was associated with an increased re-excision rate. It is clear from the present findings, and has also been shown frequently in the literature1,5, that there is a strong association between MRI and mastectomy in all subgroups, except for patients aged 40 years or less and those with invasive lobular breast cancer. A weakness of the study is its retrospective design, and thus the inherent lack of information, such as the presence of multifocality or multicentricity, the indication for performing MRI, the result of the MRI and whether it changed the surgical plan. Recommendations in the Dutch breast cancer guidelines regarding the preoperative use of MRI, described above, can, however, shed some light on the decision-making process underlying the results of this study. The retrospective, non-randomized design of the study also explains the differences in patient characteristics, such as patients undergoing MRI having larger tumours and being more likely to have a DCIS component adjacent to the invasive tumour (Table 1). It is known that these factors increase the risk of incomplete excision, and this could have been the reason for performing MRI and thus be a source of selection bias. Even though factors associated with mastectomy were adjusted for in the multivariable analysis, residual confounding may be present owing to factors not taken into consideration. It must also be noted that the study did not include data from university hospitals, and it is therefore likely that MRI was used for surgical planning rather than screening purposes. The median (i.q.r.) time between MRI and surgery was 24 (15–36) days, supporting the assumption that MRI was used for surgical planning. In addition, the lack of information regarding the incidence of additional occult disease in the ipsilateral breast (estimated to be 16 per cent) detected by MRI prevents firm conclusions from being drawn1. The detection rate for contralateral breast cancer by preoperative MRI has been estimated to be 4·1 per cent29. In the present study, the incidence of contralateral breast cancer diagnosed within 3 months after diagnosis of the first tumour was 3·2 per cent in patients who had MRI and 1·2 per cent in those who did not. This difference might be larger in patients with lobular cancer and older patients, owing to the higher absolute risk of contralateral breast cancer in these subgroups30. Whether preoperative MRI reduces the risk of local and distant recurrence is still a matter of debate31,32. Because the cancer registry has included information on preoperative MRI only since 2011, long-term outcomes could not be studied. It can be expected, however, that residual disease in the breast results in positive resection margins that will be treated by re-excision, regardless of the preoperative MRI. It has been shown previously that overall survival is similar in women who have a re-excision and those who do not33. Moreover, long-term prognosis has proved to be similar in women having BCS and those undergoing mastectomy34, and preoperative MRI is therefore unlikely to influence prognosis. Thus, short-term surgical outcomes remain important endpoints for studying the benefits of MRI. Overall, MRI was used with a relatively high frequency of 32·4 per cent, compared with rates in the above-mentioned population-based studies. In these studies6,7, which used the US SEER–Medicare-linked database, 6·6–12·2 per cent of patients underwent MRI. The greater use of MRI in the Netherlands could be explained by the more recent time interval, compared with the periods covered in previous studies. It could also be explained by the fact that preoperative staging by MRI has been advised for invasive lobular cancer by the Dutch breast cancer guideline since 201212, and by the EUSOMA working group since 201010. There is also a growing body of evidence that the targeted use of MRI in this subgroup improves surgical planning5–6,8–9. In the present study, MRI was used in 449 (66·0 per cent) of the 680 patients with invasive lobular breast cancer, and all surgical outcomes were similar for patients with and without MRI. At least these results indicate that MRI is unlikely to have a negative effect in this subgroup. The third explanation for the more widespread use of MRI in the Netherlands, in comparison with values reported by the studies based on the SEER–Medicare database, could be the younger age of the population. The SEER–Medicare database contains data only for patients aged 65 years or more, whereas the present study included women of all ages, making it the first population-based study to include patients aged less than 65 years. Interestingly, the subgroup analysis of patients aged 40 years or less showed that MRI was significantly associated with neither more extensive surgery (initial and final mastectomy rates) nor the other surgical outcomes studied. However, considering the small number of patients in this subgroup (197), there is a risk of a type II error, and thus the results need to be interpreted with caution. The finding is merely an observation of statistical association and does not provide evidence for a causative relationship. The present study has shown in a population-based retrospective cohort that preoperative MRI does not result in improved surgical outcomes after BCS, but instead leads to more extensive surgery in patients with breast cancer in general, a finding in line with previous studies. An exception could be in patients aged 40 years or less, and those with invasive lobular breast cancer. Furthermore, MRI may cause a delay in treatment. Large prospective studies are needed urgently to define patient subgroups in which preoperative MRI is of value in short- and long-term outcomes. Acknowledgements The authors thank the registration teams of the Comprehensive Cancer Centre Netherlands and Comprehensive Cancer Centre South for the collection of data for the Netherlands Cancer Registry, and the scientific staff of the Comprehensive Cancer Centre Netherlands. Disclosure: The authors declare no conflict of interest. Snapshot quiz 15/12 Open in new tabDownload slide References 1 Houssami N , Ciatto S, Macaskill P, Lord SJ, Warren RM, Dixon JM et al. Accuracy and surgical impact of magnetic resonance imaging in breast cancer staging: systematic review and meta-analysis in detection of multifocal and multicentric cancer . J Clin Oncol 2008 ; 26 : 3248 – 3258 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Van Goethem M , Tjalma W, Schelfout K, Verslegers I, Biltjes I, Parizel P. Magnetic resonance imaging in breast cancer . Eur J Surg Oncol 2006 ; 32 : 901 – 910 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Kuhl CK , Schrading S, Bieling HB, Wardelmann E, Leutner CC, Koenig R et al. MRI for diagnosis of pure ductal carcinoma in situ: a prospective observational study . Lancet 2007 ; 370 : 485 – 492 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Schouten van der Velden AP , Boetes C, Bult P, Wobbes T. The value of magnetic resonance imaging in diagnosis and size assessment of in situ and small invasive breast carcinoma . Am J Surg 2006 ; 192 : 172 – 178 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Houssami N , Turner R, Morrow M. Preoperative magnetic resonance imaging in breast cancer: meta-analysis of surgical outcomes . Ann Surg 2013 ; 257 : 249 – 255 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Fortune-Greeley AK , Wheeler SB, Meyer AM, Reeder-Hayes KE, Biddle AK, Muss HB et al. Preoperative breast MRI and surgical outcomes in elderly women with invasive ductal and lobular carcinoma: a population-based study . Breast Cancer Res Treat 2014 ; 143 : 203 – 212 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Wang SY , Kuntz KM, Tuttle TM, Jacobs DR Jr, Kane RL, Virnig BA. The association of preoperative breast magnetic resonance imaging and multiple breast surgeries among older women with early stage breast cancer . Breast Cancer Res Treat 2013 ; 138 : 137 – 147 . Google Scholar Crossref Search ADS PubMed WorldCat 8 McGhan LJ , Wasif N, Gray RJ, Giurescu ME, Pizzitola VJ, Lorans R et al. Use of preoperative magnetic resonance imaging for invasive lobular cancer: good, better, but maybe not the best? Ann Surg Oncol 2010 ; 17 ( Suppl 3 ): 255 – 262 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Mann RM , Loo CE, Wobbes T, Bult P, Barentsz JO, Gilhuijs KG et al. The impact of preoperative breast MRI on the re-excision rate in invasive lobular carcinoma of the breast . Breast Cancer Res Treat 2010 ; 119 : 415 – 422 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Sardanelli F , Boetes C, Borisch B, Decker T, Federico M, Gilbert FJ et al. Magnetic resonance imaging of the breast: recommendations from the EUSOMA working group . Eur J Cancer 2010 ; 46 : 1296 – 1316 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Nationaal Borstkanker Overleg Nederland (NABON) . Breast Cancer Guideline 2008 . http://www.oncoline.nl/mammacarcinoom [accessed 31 August 2015 ]. 12 Nationaal Borstkanker Overleg Nederland (NABON) . Breast Cancer Guideline 2012 . http://www.oncoline.nl/mammacarcinoom [accessed 31 August 2015 ]. 13 Dutch Institute for Clinical Auditing (DICA) . NABON Breast Cancer Audit (NBCA) . http://nbca.clinicalaudit.nl/ [accessed 31 August 2015 ]. 14 Parkin DM , Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002 . CA Cancer J Clin 2005 ; 55 : 74 – 108 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Schouten LJ , Hoppener P, van den Brandt PA, Knottnerus JA, Jager JJ. Completeness of cancer registration in Limburg, The Netherlands . Int J Epidemiol 1993 ; 22 : 369 – 376 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Sobin LH , Gospodarowicz MK, Wittekind C. TNM Classification of Malignant Tumours (7th edn). Wiley–Blackwell : Chichester , 2009 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 17 Fancellu A , Soro D, Castiglia P, Marras V, Melis M, Cottu P et al. Usefulness of magnetic resonance in patients with invasive cancer eligible for breast conservation: a comparative study . Clin Breast Cancer 2014 ; 14 : 114 – 121 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Miller BT , Abbott AM, Tuttle TM. The influence of preoperative MRI on breast cancer treatment . Ann Surg Oncol 2012 ; 19 : 536 – 540 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Shin HC , Han W, Moon HG, Yom CK, Ahn SK, You JM et al. Limited value and utility of breast MRI in patients undergoing breast-conserving cancer surgery . Ann Surg Oncol 2012 ; 19 : 2572 – 2579 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Hwang N , Schiller DE, Crystal P, Maki E, McCready DR. Magnetic resonance imaging in the planning of initial lumpectomy for invasive breast carcinoma: its effect on ipsilateral breast tumor recurrence after breast-conservation therapy . Ann Surg Oncol 2009 ; 16 : 3000 – 3009 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Bleicher RJ , Ciocca RM, Egleston BL, Sesa L, Evers K, Sigurdson ER et al. Association of routine pretreatment magnetic resonance imaging with time to surgery, mastectomy rate, and margin status . J Am Coll Surg 2009 ; 209 : 180 – 187 ; quiz 294–185. Google Scholar Crossref Search ADS PubMed WorldCat 22 Pengel KE , Loo CE, Teertstra HJ, Muller SH, Wesseling J, Peterse JL et al. The impact of preoperative MRI on breast-conserving surgery of invasive cancer: a comparative cohort study . Breast Cancer Res Treat 2009 ; 116 : 161 – 169 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Davis KL , Barth RJ, Gui J, Dann E, Eisenberg B, Rosenkranz K. Use of MRI in preoperative planning for women with newly diagnosed DCIS: risk or benefit? Ann Surg Oncol 2012 ; 19 : 3270 – 3274 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Itakura K , Lessing J, Sakata T, Heinzerling A, Vriens E, Wisner D et al. The impact of preoperative magnetic resonance imaging on surgical treatment and outcomes for ductal carcinoma in situ . Clin Breast Cancer 2011 ; 11 : 33 – 38 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Allen LR , Lago-Toro CE, Hughes JH, Careaga E, Brown AT, Chernick M et al. Is there a role for MRI in the preoperative assessment of patients with DCIS? Ann Surg Oncol 2010 ; 17 : 2395 – 2400 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Pilewskie M , Kennedy C, Shappell C, Helenowski I, Scholtens D, Hansen N et al. Effect of MRI on the management of ductal carcinoma in situ of the breast . Ann Surg Oncol 2013 ; 20 : 1522 – 1529 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Kropcho LC , Steen ST, Chung AP, Sim MS, Kirsch DL, Giuliano AE. Preoperative breast MRI in the surgical treatment of ductal carcinoma in situ . Breast J 2012 ; 18 : 151 – 156 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Peters NH , van Esser S, van den Bosch MA, Storm RK, Plaisier PW, van Dalen T et al. Preoperative MRI and surgical management in patients with nonpalpable breast cancer: the MONET – randomised controlled trial . Eur J Cancer 2011 ; 47 : 879 – 886 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Brennan ME , Houssami N, Lord S, Macaskill P, Irwig L, Dixon JM et al. Magnetic resonance imaging screening of the contralateral breast in women with newly diagnosed breast cancer: systematic review and meta-analysis of incremental cancer detection and impact on surgical management . J Clin Oncol 2009 ; 27 : 5640 – 5649 . Google Scholar Crossref Search ADS PubMed WorldCat 30 de Glas NA , Engels CC, Bastiaannet E, van de Water W, Siesling S, de Craen AJ et al. Contralateral breast cancer risk in relation to tumor morphology and age – in which patients is preoperative MRI justified? Breast Cancer Res Treat 2015 ; 150 : 191 – 198 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Houssami N , Turner R, Macaskill P, Turnbull LW, McCready DR, Tuttle TM et al. An individual person data meta-analysis of preoperative magnetic resonance imaging and breast cancer recurrence . J Clin Oncol 2014 ; 32 : 392 – 401 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Houssami N , Ciatto S, Martinelli F, Bonardi R, Duffy SW. Early detection of second breast cancers improves prognosis in breast cancer survivors . Ann Oncol 2009 ; 20 : 1505 – 1510 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Vos EL , Jager A, Verhoef C, Voogd AC, Koppert LB. Overall survival in patients with a re-excision following breast conserving surgery compared to those without in a large population-based cohort . Eur J Cancer 2015 ; 51 : 282 – 291 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Litière S , Werutsky G, Fentiman IS, Rutgers E, Christiaens MR, Van Limbergen E et al. Breast conserving therapy versus mastectomy for stage I–II breast cancer: 20 year follow-up of the EORTC 10801 phase 3 randomised trial . Lancet Oncol 2012 ; 13 : 412 – 419 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes Presented to the San Antonio Breast Cancer Symposium, San Antonio, Texas, USA, December 2014 © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd

Beek, M A; Verheuvel, N C; Luiten, E J T; Klompenhouwer, E G; Rutten, H J T; Roumen, R M H; Gobardhan, P D; Voogd, A C

doi: 10.1002/bjs.9955pmid: 26694991

Abstract Background Axillary lymph node dissection (ALND) in patients with breast cancer provides prognostic information. For many years, positive nodes were the most important indication for adjuvant systemic therapy. It was also believed that regional control could not be achieved without axillary clearance in a positive axilla. However, during the past 20 years the treatment and staging of the axilla has undergone many changes. This large population-based study was conducted in the south-east of the Netherlands to evaluate the changing patterns of care regarding the axilla, including the introduction of sentinel lymph node biopsy (SLNB) in the late 1990s, implementation of the results of the American College of Surgeons Oncology Group Z0011 study, and the initial effects of the European Organization for Research and Treatment of Cancer AMAROS study. Methods Data from the population-based Eindhoven Cancer Registry of all women diagnosed with invasive breast cancer in the south of the Netherlands between January 1993 and July 2014 were used. Results The proportion of 34 037 women staged by SLNB without completion ALND increased from 0 per cent in 1993–1994 to 69·0 per cent in 2013–2014. In the same period the proportion undergoing ALND decreased from 88·8 to 18·7 per cent. Among women with one to three positive lymph nodes, the proportion undergoing SLNB alone increased from 10·6 per cent in 2011–2012 to 37·6 per cent in 2013–2014. Conclusion This population-based study demonstrated the radical transformation in management of the axilla since the introduction of SLNB and following the recent publication of trials on management of the axilla with a low metastatic burden. Introduction Axillary lymph node dissection (ALND) in patients with breast cancer has been performed for disease control, tumour staging, and to aid in decision-making regarding the use of adjuvant systemic treatment and regional nodal irradiation1,2. An important reason for critical re-evaluation of the need for ALND was the postoperative morbidity associated with the procedure3–5. In the late 1990s this led to the introduction of sentinel lymph node biopsy (SLNB)6–9, which has become the standard of care in clinically node-negative early-stage breast cancer10–13. SLNB was considered to be reliable for staging of the axilla and for the decision to omit ALND in patients with a negative sentinel lymph node (SLN). The results of the American College of Surgeons Oncology Group (ACOSOG) Z0011 trial, published in 2011, have cast doubt on the need to perform completion ALND in patients with SLN metastases14,15. This trial showed that omitting ALND does not result in inferior overall and disease-free survival rates in patients with cT1–2 N0 breast cancer, with one or two SLN metastases treated with breast-conserving therapy as well as adjuvant systemic therapy. Recently, the results of the European Organization for Research and Treatment of Cancer (EORTC) AMAROS (After Mapping of the Axilla: Radiotherapy or Surgery) trial16, published in 2014, showed that both completion ALND and axillary radiotherapy after a positive SLN procedure provided excellent and comparable axillary control for patients with T1–2 primary breast cancer and no palpable lymphadenopathy. The aim of the present study, using data from a large population-based cancer registry, was to evaluate patterns of care in treatment of the axilla with regard to the introduction of SLNB in 1998, and the impact of the ACOSOG Z0011 and EORTC AMAROS trials on management of the axilla in daily clinical practice. Methods Data were retrieved from the Eindhoven Cancer Registry, which is part of the Dutch Cancer Registry and records data on all patients newly diagnosed with cancer in the south-east region of the Netherlands, an area with approximately 2·4 million inhabitants. Data were derived from ten hospitals: six large non-university teaching and four community hospitals. Data on patient and tumour characteristics, and treatment were collected based on the pathology reports and medical records. The patients were staged according to the TNM system of the International Union Against Cancer (UICC)17. Data were revised continuously to ensure completeness and accuracy. To perform a truly population-based analysis, all women diagnosed with primary invasive breast cancer in the period from January 1993 to July 2014 were included in the analyses. Patterns of care for the axilla were studied for the group as a whole, and also according to age group and pathological axillary node status (pN category). The following age groups were defined: less than 50 years, 50–69 years, 70–79 years or 80 years or above. The following subgroups were defined according to pN category: pN0, patients without lymph node metastases; pN0(i+), patients with isolated tumour cells (metastases smaller than 0·2 mm); pN1mi, patients with micrometastases (metastases of 0·2–2·0 mm in size); pN1a, patients with metastases in one to three axillary lymph nodes, including at least one larger than 2 mm; pN > 1a, patients with metastases in more than three axillary lymph nodes or macrometastases in either supraclavicular and/or internal mammary lymph nodes. Patients with pN1a disease were further subdivided with regard to type of breast surgery: mastectomy or breast-conserving surgery. Isolated tumour cells have been documented by the Dutch Cancer Registry only since 2003, following publication of the revised TNM system in 200217. Statistical analysis The χ2 test was used to determine whether observed trends in axillary management were statistically significant. Statistical analyses were performed using SPSS® version 21·0 (IBM, Armonk, New York, USA). P values less than 0·05 were considered significant. Results Between January 1993 and July 2014, a total of 34 037 women were treated for primary invasive breast cancer in the south-east region of the Netherlands. Baseline characteristics are shown in Table 1. Overall, 51·8 per cent of the women had breast-conserving surgery, 39·4 per cent had a mastectomy, and 8·8 per cent did not undergo any definitive breast surgery or the type of surgery was not documented properly. Table 1 Characteristics at diagnosis of patients treated for primary breast cancer in south-east Netherlands, 1993–2014 . No. of patients (n = 34 037) . Age (years) < 50 8217 (24·1) 50–69 16 514 (48·5) ≥70 9306 (27·3) Period of diagnosis 1993–1994 2455 (7·2) 1995–1996 2386 (7·0) 1997–1998 2640 (7·8) 1999–2000 2821 (8·3) 2001–2002 3095 (9·1) 2003–2004 3382 (9·9) 2005–2006 3303 (9·7) 2007–2008 3568 (10·5) 2009–2010 3640 (10·7) 2011–2012 3883 (11·4) 2013–2014 2864 (8·4) Tumour category pT1 19 111 (56·1) pT2 11 199 (32·9) pT3 1200 (3·5) pT4 1724 (5·1) Not known 803 (2·4) Node category pN0 18 181 (53·4) pN0(i+) 522 (1·5) pNmi 1326 (3·9) pN1a 7074 (20·8) pN >1a 2763 (8·1) Not known 4171 (12·3) Histology Ductal 26 763 (78·6) Lobular mixed 5159 (15·2) Other 1378 (4·0) Unknown 737 (2·2) Type of surgery Breast-conserving surgery 17 636 (51·8) Mastectomy 13 416 (39·4) None/unknown 2985 (8·8) Type of axillary staging None 3685 (10·8) ALND only 13 421 (39·4) SLNB only 12 351 (36·3) SLNB + ALND 4396 (12·9) Not known 184 (0·5) . No. of patients (n = 34 037) . Age (years) < 50 8217 (24·1) 50–69 16 514 (48·5) ≥70 9306 (27·3) Period of diagnosis 1993–1994 2455 (7·2) 1995–1996 2386 (7·0) 1997–1998 2640 (7·8) 1999–2000 2821 (8·3) 2001–2002 3095 (9·1) 2003–2004 3382 (9·9) 2005–2006 3303 (9·7) 2007–2008 3568 (10·5) 2009–2010 3640 (10·7) 2011–2012 3883 (11·4) 2013–2014 2864 (8·4) Tumour category pT1 19 111 (56·1) pT2 11 199 (32·9) pT3 1200 (3·5) pT4 1724 (5·1) Not known 803 (2·4) Node category pN0 18 181 (53·4) pN0(i+) 522 (1·5) pNmi 1326 (3·9) pN1a 7074 (20·8) pN >1a 2763 (8·1) Not known 4171 (12·3) Histology Ductal 26 763 (78·6) Lobular mixed 5159 (15·2) Other 1378 (4·0) Unknown 737 (2·2) Type of surgery Breast-conserving surgery 17 636 (51·8) Mastectomy 13 416 (39·4) None/unknown 2985 (8·8) Type of axillary staging None 3685 (10·8) ALND only 13 421 (39·4) SLNB only 12 351 (36·3) SLNB + ALND 4396 (12·9) Not known 184 (0·5) Values in parentheses are percentages. pN0(i+), isolated tumour cells; pNmi, micrometastases; ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy. Open in new tab Table 1 Characteristics at diagnosis of patients treated for primary breast cancer in south-east Netherlands, 1993–2014 . No. of patients (n = 34 037) . Age (years) < 50 8217 (24·1) 50–69 16 514 (48·5) ≥70 9306 (27·3) Period of diagnosis 1993–1994 2455 (7·2) 1995–1996 2386 (7·0) 1997–1998 2640 (7·8) 1999–2000 2821 (8·3) 2001–2002 3095 (9·1) 2003–2004 3382 (9·9) 2005–2006 3303 (9·7) 2007–2008 3568 (10·5) 2009–2010 3640 (10·7) 2011–2012 3883 (11·4) 2013–2014 2864 (8·4) Tumour category pT1 19 111 (56·1) pT2 11 199 (32·9) pT3 1200 (3·5) pT4 1724 (5·1) Not known 803 (2·4) Node category pN0 18 181 (53·4) pN0(i+) 522 (1·5) pNmi 1326 (3·9) pN1a 7074 (20·8) pN >1a 2763 (8·1) Not known 4171 (12·3) Histology Ductal 26 763 (78·6) Lobular mixed 5159 (15·2) Other 1378 (4·0) Unknown 737 (2·2) Type of surgery Breast-conserving surgery 17 636 (51·8) Mastectomy 13 416 (39·4) None/unknown 2985 (8·8) Type of axillary staging None 3685 (10·8) ALND only 13 421 (39·4) SLNB only 12 351 (36·3) SLNB + ALND 4396 (12·9) Not known 184 (0·5) . No. of patients (n = 34 037) . Age (years) < 50 8217 (24·1) 50–69 16 514 (48·5) ≥70 9306 (27·3) Period of diagnosis 1993–1994 2455 (7·2) 1995–1996 2386 (7·0) 1997–1998 2640 (7·8) 1999–2000 2821 (8·3) 2001–2002 3095 (9·1) 2003–2004 3382 (9·9) 2005–2006 3303 (9·7) 2007–2008 3568 (10·5) 2009–2010 3640 (10·7) 2011–2012 3883 (11·4) 2013–2014 2864 (8·4) Tumour category pT1 19 111 (56·1) pT2 11 199 (32·9) pT3 1200 (3·5) pT4 1724 (5·1) Not known 803 (2·4) Node category pN0 18 181 (53·4) pN0(i+) 522 (1·5) pNmi 1326 (3·9) pN1a 7074 (20·8) pN >1a 2763 (8·1) Not known 4171 (12·3) Histology Ductal 26 763 (78·6) Lobular mixed 5159 (15·2) Other 1378 (4·0) Unknown 737 (2·2) Type of surgery Breast-conserving surgery 17 636 (51·8) Mastectomy 13 416 (39·4) None/unknown 2985 (8·8) Type of axillary staging None 3685 (10·8) ALND only 13 421 (39·4) SLNB only 12 351 (36·3) SLNB + ALND 4396 (12·9) Not known 184 (0·5) Values in parentheses are percentages. pN0(i+), isolated tumour cells; pNmi, micrometastases; ALND, axillary lymph node dissection; SLNB, sentinel lymph node biopsy. Open in new tab In 1997–1998, SLNB was used for the first time in patients treated in the Eindhoven region (Fig. 1). In this period, a total of 2621 women were treated for primary invasive breast cancer, of whom 80 (3·1 per cent) had staging of the axilla with SLNB alone and 229 (8·7 per cent) with SLNB followed by completion ALND. In 2013–2014, a total of 2843 women were treated for primary invasive breast cancer, of whom 1961 (69·0 per cent) had axillary staging with SLNB alone and 192 (6·8 per cent) with SLNB plus completion ALND. The proportion of women without axillary staging (no ALND or SLNB) ranged from 11·2 per cent in 1993–1994 to 12·4 per cent in 2013–2014 (Fig. 1). Fig. 1 Open in new tabDownload slide Trend in axillary staging and management for all patients, 1993–2014. Values in parentheses are numbers of patients. SLNB, sentinel lymph node biopsy; ALND, axillary lymph node dissection Trends in axillary staging according to age group Among women aged 50 years or less, the proportion staged with SLNB alone increased from 17·7 per cent in 1999–2000 to 64·3 per cent in 2013–2014 (Fig. S1a, supporting information). A similar trend was observed in the age groups of 50–69 and 70–79 years (Fig. S1b,c, supporting information). In contrast to the other three groups, the proportion of women aged 80 years or above who did not undergo axillary staging increased from 45·3 per cent in 1993–1994 to 56·8 per cent in 2013–2014 (Fig. S1d, supporting information). In this age group, the proportion undergoing ALND decreased from 54·7 per cent in 1993–1994 to 9·8 per cent in 2013–2014. Trends in axillary staging according to pathological axillary node status Among women with pN0 disease, the proportion staged with ALND decreased from 100 per cent in 1993–1994 to 6·1 per cent in 2013–2014 (Fig. 2a). Fig. 2 Open in new tabDownload slide Trend in axillary staging and management, 1993–2014, in patients with: a no axillary metastasis (pN0), b isolated tumour cells (metastases smaller than 0·2 mm, pN0(i+)), c micrometastases (metastases between 0·2 and 2·0 mm, pN1mi), d metastases in one to three axillary lymph nodes, including at least one larger than 2 mm (pN1a), e more than three axillary lymph nodes or macrometastases in either supraclavicular and/or internal mammary lymph nodes (pN > 1a). Values in parentheses are numbers of patients. SLNB, sentinel lymph node biopsy; ALND, axillary lymph node dissection In 2003–2004, 47·4 per cent of the women diagnosed with isolated tumour cells by SLNB underwent completion ALND; by 2013–2014 this proportion had decreased to 1·7 per cent (Fig. 2b). The use of SLNB alone in women with micrometastases increased from 34·9 per cent in 1999–2000 to 86·2 per cent in 2013–2014. Until 2012, most patients with micrometastases demonstrated by SLNB also underwent completion ALND; by 2013–2014 this proportion had decreased to 6·9 per cent (Fig. 2c). Between 1999 and 2011, about 5 per cent of women with pN1a disease were staged by SLNB alone, compared with 37·6 per cent of those diagnosed in 2013–2014 (Fig. 2d). When analysed according to type of surgery, in 2013–2014 the proportion of women undergoing breast-conserving surgery who had SLNB alone was 49·6 per cent, compared with 22·4 per cent of women having a mastectomy (Fig. 3a,b). Fig. 3 Open in new tabDownload slide Trend in axillary staging and management, 1993–2014, in patients with one to three axillary lymph node metastases (pN1a) undergoing a breast-conserving surgery or b mastectomy. Values in parentheses are numbers of patients. SLNB, sentinel lymph node biopsy; ALND, axillary lymph node dissection Among women with pN > 1a disease, the proportion who had SLNB alone was 5·4 per cent in 2013–2014 versus 3 per cent or less before that period (Fig. 2e). All time trends in axillary staging described above were statistically significant (P < 0·001). Of 110 women with one to three positive axillary lymph nodes who had breast-conserving surgery and SLNB without completion ALND in 2013–2014, 108 (98·2 per cent) received adjuvant systemic treatment. In addition, of the 39 women with one to three positive axillary lymph nodes who had a mastectomy and SLNB without completion ALND in the same period, 32 (82 per cent) underwent postmastectomy radiotherapy. Discussion In this large population-based study, the patterns of axillary staging based on a large and unselected patient population are presented. A major shift to less invasive axillary lymph node surgery was found over an interval of almost two decades. SLNB was first introduced in Dutch national breast cancer treatment guidelines in 199918,19. Currently, SLNB is used in more than 70 per cent of all women with breast cancer. As a consequence of the implementation of SLNB, the proportion of women undergoing primary ALND has decreased. This replacement of ALND by SLNB has been described previously in several studies18,20–21. Before publication of the results of the ACOSOG Z0011 trial, the finding of one or more positive SLNs was a clear indication for completion ALND. The situation was always less distinct when the SLN contained micrometastases or isolated tumour cells and no macrometastatic disease22,23. The debate on the need for completion ALND in these patients is reflected by the treatment patterns observed in the present study, and has been described in more detail previously23. In 2009, the St Gallen Consensus Panel recommended that completion ALND should not be performed in patients with micrometastases or isolated tumour cells in the SLN24,25. According to the present data, in 2013–2014 more than 85 per cent of the women with SLN micrometastases were staged by SLNB alone, and for women with SLN isolated tumour cells this was almost 100 per cent. Although the ACOSOG Z0011 trial has been criticized for several reasons, including a heavy selection of patients, too many participating centres with low accrual rates, and only half of the initially planned 1900 patients randomized, publication of its results in 2011 can be considered as a practice-changing event in the management of the axilla. The findings from this trial questioned the need for completion ALND in patients with SLNB metastases14. By also including the periods 2011–2012 and 2013–2014 in the present study, a comparison of the situation before and after the publication of the ACOSOG Z0011 trial can be made. It is clear that axillary staging radically changed for patients with pN1a disease, even though the results of the Z0011 trial had not been integrated in the Dutch breast cancer treatment guidelines. In 2013–2014, 37·6 per cent of these women had axillary staging by SLNB alone, compared with about 5 per cent in 2009–2010. This difference was even more pronounced when the type of surgery was taken into account. Before publication of the ACOSOG Z0011 trial, 7·9 per cent of patients treated with breast-conserving surgery (one of the main inclusion criteria of the ACOSOG Z0011 trial) were staged by SLNB alone, compared with 49·6 per cent in 2013–2014. The clinical impact of the ACOSOG Z0011 trial has been evaluated previously in several other studies. Wright and colleagues26 performed a retrospective population-based study to identify the impact of the ACOSOG Z0011 trial on breast cancer surgery in a community hospital setting. In this study significant changes in practice over time in the post-ACOSOG Z0011 trial era, including reduced rates of completion ALND, were observed. The rate of completion ALND in patients meeting ACOSOG Z0011 criteria decreased from 90 to 29 per cent. A limitation of this population-based study, compared with the present study, was its relatively small number of patients (n = 1125). A single-centre study by Caudle and co-workers27 found a decrease in the rate of completion ALND from 85 per cent in the years before publication of the ACOSOG Z0011 results to 24 per cent thereafter in patients with a positive SLN who met the ACOSOG Z0011 eligibility criteria. Caretta-Weyer et al.28 suggested that 38 per cent of their patients could be spared a completion ALND, whereas Dengel and colleagues29 showed that completion ALND could be avoided in 84 per cent of patients having breast-conserving surgery. Two other single-institution studies30,31 have also reported rates of approximately 45 per cent. In another single-centre study, Robinson and co-workers32, using a prospectively collected database of patients with cT1–2 N0 breast cancer, found that before publication of the ACOSOG Z0011 trial 84 per cent of all patients with macrometastases discovered during SLNB underwent completion ALND, compared with 63 per cent afterwards. Gainer and colleagues33 and Massimino et al.34 used questionnaire surveys to evaluate the impact of the ACOSOG Z0011 trial on surgical practice. Gainer and co-workers asked 2759 members of the American Society of Breast Surgeons to complete a questionnaire regarding the practice type and duration, familiarity with ACOSOG Z0011, and mode of receiving the trial results. The majority (97 per cent) of respondents indicated familiarity with the ACOSOG Z0011 trial, and many had incorporated it into their practice. Massimino and colleagues reported a 63 per cent decrease in completion ALND for patients in the Pacific north-west, similar to the patients in the ACOSOG Z0011 trial. Recently, the EORTC AMAROS trial16 and the International Breast Cancer Study Group (IBCSG) trial35 challenged the need for completion ALND in patients with SLN metastases. The AMAROS trial compared completion ALND with axillary radiotherapy in patients with a positive SLN, T1–2 primary breast cancer and no palpable lymphadenopathy, and showed that excellent axillary control can be achieved with radiotherapy. Before publication of the results of the AMAROS trial in 2014, and probably also as a result of early reporting of the results at the annual meeting of the American Society of Clinical Oncology (ASCO) in 2013, it could be observed from the present registry data that ALND was being omitted in an increasing number of patients with one to three positive lymph nodes undergoing mastectomy. In 2013–2014, 22·4 per cent of these patients were staged by SLNB alone, versus about 6 per cent before publication. Furthermore, the present study shows that elderly women with breast cancer (aged 80 years or above) underwent axillary staging less frequently than younger women, as has also been reported by others36,37. The present study, using a large population-based registry, has demonstrated the radical transformation in management of the axilla since the introduction of SLNB and following the publication of recent studies. Acknowledgements The authors thank the registration team of the Eindhoven Cancer Registry and the Comprehensive Cancer Centre the Netherlands (IKNL) for data collection. Disclosure: The authors declare no conflict of interest. Supporting information Additional supporting information may be found in the online version of this article: Fig. S1 Trend in axillary staging and management, 1993–2014, in patients aged: a less than 50 years, b 50–69 years, c 70–79 years, and d 80 years or more (Word document) References 1 van de Vijver MJ , He YD, van't Veer LJ, Dai H, Hart AA, Voskuil DW et al. A gene-expression signature as a predictor of survival in breast cancer . N Engl J Med 2002 ; 347 : 1999 – 2009 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Goldhirsch A , Glick JH, Gelber RD, Senn HJ. Meeting highlights: International Consensus Panel on the Treatment of Primary Breast Cancer . J Natl Cancer Inst 1998 ; 90 : 1601 – 1608 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Liljegren G , Holmberg L. Arm morbidity after sector resection and axillary dissection with or without postoperative radiotherapy in breast cancer stage I. Results from a randomised trial. Uppsala–Orebro Breast Cancer Study Group . Eur J Cancer 1997 ; 33 : 193 – 199 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Ivens D , Hoe AL, Podd TJ, Hamilton CR, Taylor I, Royle GT. Assessment of morbidity from complete axillary dissection . Br J Cancer 1992 ; 66 : 136 – 138 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Noguchi M , Miwa K, Michigishi T, Yokoyama K, Nishijima H, Takanaka T et al. The role of axillary lymph node dissection in breast cancer management . Breast Cancer 1997 ; 4 : 143 – 153 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Krag DN , Weaver DL, Alex JC, Fairbank JT. Surgical resection and radiolocalization of the sentinel lymph node in breast cancer using a gamma probe . Surg Oncol 1993 ; 2 : 335 – 359 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Ververs JM , Roumen RM, Vingerhoets AJ, Vreugdenhil G, Coebergh JW, Crommelin MA et al. Risk, severity and predictors of physical and psychological morbidity after axillary lymph node dissection for breast cancer . Eur J Cancer 2001 ; 37 : 991 – 999 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Schijven M , Rutten H, Roumen R. Implementation of the sentinel node biopsy: a survey among surgeons in the Netherlands . Eur J Surg Oncol 2000 ; 26 : 431 – 432 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Schijven MP , Vingerhoets AJ, Rutten HJ, Nieuwenhuijzen GA, Roumen RM, van Bussel ME et al. Comparison of morbidity between axillary lymph node dissection and sentinel node biopsy . Eur J Surg Oncol 2003 ; 29 : 341 – 350 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Lyman GH , Giuliano AE, Somerfield MR, Benson AB, Bodurka DC, Burstein HJ et al. American Society of Clinical Oncology guideline recommendations for sentinel lymph node biopsy in early-stage breast cancer . J Clin Oncol 2005 ; 23 : 7703 – 7720 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Edge SB , Niland JC, Bookman MA, Theriault RL, Ottesen R, Lepisto E et al. Emergence of sentinel node biopsy in breast cancer as standard-of-care in academic comprehensive cancer centers . J Natl Cancer Inst 2003 ; 95 : 1514 – 1521 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Veronesi U , Paganelli G, Viale G, Luini A, Zurrida S, Galimberti V et al. A randomized comparison of sentinel-node biopsy with routine axillary dissection in breast cancer . N Engl J Med 2003 ; 349 : 546 – 553 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Posther KE , McCall LM, Blumencranz PW, Burak WE, Beitsch PD, Hansen NM et al. Sentinel node skills verification and surgeon performance: data from a multicenter clinical trial for early-stage breast cancer . Ann Surg 2005 ; 242 : 593 – 599 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Giuliano AE , Hunt KK, Ballman KV, Beitsch PD, Whitworth PW, Blumencranz PW et al. Axillary dissection vs no axillary dissection in women with invasive breast cancer and sentinel node metastasis: a randomized clinical trial . JAMA 2011 ; 305 : 569 – 575 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Giuliano AE , McCall L, Beitsch P, Whitworth PW, Blumencranz P, Leitch AM et al. Locoregional recurrence after sentinel lymph node dissection with or without axillary dissection in patients with sentinel lymph node metastases: the American College of Surgeons Oncology Group Z0011 randomized trial . Ann Surg 2010 ; 252 : 426 – 432 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Donker M , van Tienhoven G, Straver ME, Meijnen P, van de Velde CJ, Mansel RE et al. Radiotherapy or surgery of the axilla after a positive sentinel node in breast cancer (EORTC 10981–22023 AMAROS): a randomised, multicentre, open-label, phase 3 non-inferiority trial . Lancet Oncol 2014 ; 15 : 1303 – 1310 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Sobin LH , Wittekind CH (eds). TNM Classification of Malignant Tumors (6th edn). Wiley-Liss , New York , 2002 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 18 Roumen RM , Pijpers HJ, Thunnissen FB, Ruers TJ. [ Summary of the guideline ‘Sentinel node biopsy in breast cancer.’ Dutch Work Group ‘Sentinel Node Biopsy for Breast Cancer’ .] Ned Tijdschr Geneeskd 2000 ; 144 : 1864 – 1867 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 19 Ho VK , van der Heiden-van der Loo M, Rutgers EJ, van Diest PJ, Hobbelink MG, Tjan-Heijnen VC et al. Implementation of sentinel node biopsy in breast cancer patients in the Netherlands . Eur J Cancer 2008 ; 44 : 683 – 691 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Tvedskov TF , Jensen MB, Balslev E, Ejlertsen B, Kroman N. Stage migration after introduction of sentinel lymph node dissection in breast cancer treatment in Denmark: a nationwide study . Eur J Cancer 2011 ; 47 : 872 – 878 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Maaskant AJ , van de Poll-Franse LV, Voogd AC, Coebergh JW, Tutein Nolthenius-Puylaert MC, Nieuwenhuijzen GA. Stage migration due to introduction of the sentinel node procedure: a population-based study . Breast Cancer Res Treat 2009 ; 113 : 173 – 179 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Gatzemeier W , Mann GB. Which sentinel lymph-node (SLN) positive breast cancer patient needs an axillary lymph-node dissection (ALND) – ACOSOG Z0011 results and beyond . Breast 2013 ; 22 : 211 – 216 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Maaskant-Braat AJ , Voogd AC, van de Poll-Franse LV, Coebergh JW, Nieuwenhuijzen GA. Axillary and systemic treatment of patients with breast cancer and micrometastatic disease or isolated tumor cells in the sentinel lymph node . Breast 2012 ; 21 : 524 – 548 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Singletary SE , Allred C, Ashley P, Bassett LW, Berry D, Bland KI et al. Revision of the American Joint Committee on Cancer staging system for breast cancer . J Clin Oncol 2002 ; 20 : 3628 – 3636 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Jörger M , Senn H-J, Thürlimann B. St. Gallen 2009 recommendation on the treatment of early breast cancer: consensus and controversy . Magazine of European Medical Oncology 2009 ; 2 : 229 – 231 . Google Scholar Crossref Search ADS WorldCat 26 Wright GP , Mater ME, Sobel HL, Knoll GM, Oostendorp LD, Melnik MK et al. Measuring the impact of the American College of Surgeons Oncology Group Z0011 trial on breast cancer surgery in a community health system . Am J Surg 2015 ; 209 : 240 – 245 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Caudle AS , Hunt KK, Tucker SL, Hoffman K, Gainer SM, Lucci A et al. American College of Surgeons Oncology Group (ACOSOG) Z0011: impact on surgeon practice patterns . Ann Surg Oncol 2012 ; 19 : 3144 – 3151 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Caretta-Weyer H , Greenberg CG, Wilke LG, Weiss J, LoConte NK, Decker M et al. Impact of the American College of Surgeons Oncology Group (ACOSOG) Z0011 trial on clinical management of the axilla in older breast cancer patients: a SEER–Medicare analysis . Ann Surg Oncol 2013 ; 20 : 4145 – 4152 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Dengel LT , Van Zee KJ, King TA, Stempel M, Cody HS, El-Tamer M et al. Axillary dissection can be avoided in the majority of clinically node-negative patients undergoing breast-conserving therapy . Ann Surg Oncol 2014 ; 21 : 22 – 27 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Guth U , Myrick ME, Viehl CT, Schmid SM, Obermann EC, Weber WP. The post ACOSOG Z0011 era: does our new understanding of breast cancer really change clinical practice? Eur J Surg Oncol 2012 ; 38 : 645 – 650 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Yi M , Kuerer HM, Mittendorf EA, Hwang RF, Caudle AS, Bedrosian I et al. Impact of the American College of Surgeons Oncology Group Z0011 criteria applied to a contemporary patient population . J Am Coll Surg 2013 ; 216 : 105 – 113 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Robinson KA , Pockaj BA, Wasif N, Kaufman K, Gray RJ. Have the American College of Surgeons Oncology Group Z0011 trial results influenced the number of lymph nodes removed during sentinel lymph node dissection? Am J Surg 2014 ; 208 : 1060 – 1064 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Gainer SM , Hunt KK, Beitsch P, Caudle AS, Mittendorf EA, Lucci A. Changing behavior in clinical practice in response to the ACOSOG Z0011 trial: a survey of the American Society of Breast Surgeons . Ann Surg Oncol 2012 ; 19 : 3152 – 3158 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Massimino KP , Hessman CJ, Ellis MC, Naik AM, Vetto JT. Impact of American College of Surgeons Oncology Group Z0011 and National Surgical Adjuvant Breast and Bowel Project B-32 trial results on surgeon practice in the Pacific Northwest . Am J Surg 2012 ; 203 : 618 – 622 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Galimberti V , Cole BF, Zurrida S, Viale G, Luini A, Veronesi P et al. Axillary dissection versus no axillary dissection in patients with sentinel-node micrometastases (IBCSG 23–01): a phase 3 randomised controlled trial . Lancet Oncol 2013 ; 14 : 297 – 305 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Wanebo HJ , Cole B, Chung M, Vezeridis M, Schepps B, Fulton J et al. Is surgical management compromised in elderly patients with breast cancer? Ann Surg 1997 ; 225 : 579 – 586 . Google Scholar Crossref Search ADS PubMed WorldCat 37 Mustacchi G , Cazzaniga ME, Pronzato P, De Matteis A, Di Costanzo F, Floriani I et al. Breast cancer in elderly women: a different reality? Results from the NORA study . Ann Oncol 2007 ; 18 : 991 – 996 . Google Scholar Crossref Search ADS PubMed WorldCat © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) © 2015 BJS Society Ltd Published by John Wiley & Sons Ltd