TY - JOUR
AU - Bartram, C I
AB - Abstract Background This was a prospective study designed to determine the therapeutic impact of magnetic resonance imaging (MRI) in primary fistula in ano, and to assess its effect on outcome. Methods Thirty patients with suspected primary fistula in ano underwent preoperative MRI, and the findings were revealed during surgery following examination under anaesthesia (EUA). Any effect on operative approach was noted. Outcome was assessed at a median of 12 months. Results Two patients had sinuses, one had no sepsis and 27 had fistulas: five superficial, seven intersphincteric, 14 trans-sphincteric and one suprasphincteric. MRI and EUA agreed in 15 patients and MRI findings altered the surgical approach in a further three (10 per cent); two of the latter patients were believed to have a sinus at EUA, which MRI correctly identified as a fistula, allowing definitive treatment. The therapeutic impact of MRI was therefore 10 per cent. Persisting disagreement between MRI and EUA in 12 patients mostly related to minor discrepancies in classification. Only one patient required further unplanned surgery, which was for skin-bridging rather than any new sepsis. Conclusion In experienced hands, MRI has a therapeutic impact of 10 per cent for primary fistula in ano, precipitating surgery that is likely to reduce recurrence in a small, but important, proportion of patients. Introduction Fistula in ano is usually easily treated by surgery1 but may recur in up to 25 per cent2, often from sepsis missed during initial surgery3. It is now generally accepted that magnetic resonance imaging (MRI) accurately identifies sepsis that has been missed during examination under anaesthesia (EUA)4 and predicts relapse better than EUA5. When MRI results are revealed during surgery, the operative approach is altered in a significant number of patients6. Furthermore, MRI-guided surgery reduces further relapse in patients with recurrent disease7. However, it is unknown whether MRI prevents recurrence in patients initially presenting with fistulas believed to be simple, as few studies have specifically assessed the effect of MRI on outcome in this group. While MRI has undoubted benefit in the management of recurrent fistulas, it may also identify poor prognostic features in seemingly simple primary fistulas. This prospective study aimed to determine the therapeutic effect of preoperative MRI in patients presenting with primary fistula in ano and to investigate the consequences of MRI-guided surgery on outcome. Patients and methods Following ethics committee approval, 30 consecutive consenting patients (19 men; median age 37 (range 17–61) years) with primary cryptoglandular fistula in ano were recruited prospectively between November 1998 and August 2001. To avoid selection bias, all patients without a history of previous fistula surgery were eligible, although nine underwent perianal abscess drainage before referral. Nine were tertiary referrals despite having seemingly simple fistulas. In the outpatient department, the assessing clinician indicated whether MRI would have been requested outside the context of a clinical trial. A pre- and post-intervention observational study design was used. All patients underwent supine MRI at 1·0 T (Gyroscan T10-NT™; Philips Medical Systems, Cambridge, UK) using a validated body- or surface-coil protocol described previously7,8. An experienced consultant gastrointestinal radiologist, blinded to clinical detail, supervised scanning, which was performed a median of 4 (range 0–79) days before operation. The radiologist recorded the course of the primary track and any associated extensions using a fistula classification sheet derived from Parks et al.9 and established in previous studies6,10; these notes were sealed in an envelope and taken to the subsequent EUA. Surgery was performed by one of 13 experienced surgeons (six consultants for 23 patients, seven year 6 trainees for seven patients), with the patient in the lithotomy position, aided by probes and/or hydrogen peroxide to help define openings. Laying open and/or coring out of fistulas, or seton placement was undertaken until the surgeon deemed assessment and treatment complete. At this stage the surgeon completed a fistula classification sheet, the ‘initial’ EUA assessment, and stated whether he or she believed that MRI was likely to be helpful. The MRI report was then revealed. If MRI and ‘initial’ EUA assessments agreed, surgery was considered complete. If there was disagreement, any additional exploration or treatment in the light of MRI findings was undertaken entirely at the surgeon's discretion. The nature of any additional surgery was noted, after which the surgeon completed a further fistula classification sheet for these patients, the ‘review’ EUA assessment. Postoperative care was standardized. Because treatment was by loose seton in some patients, recurrence was defined as the need for further unplanned surgery. Any recurrence was documented and findings at subsequent operations noted. Median follow-up was 12 (range 5–39) months. MRI and EUA assessments were compared with respect to primary and secondary track classification and internal opening location. Categorical frequencies were analysed using Fisher's exact test. Analysis was performed using Arcus Quickstat Biomedical version 1.2 (Research Solutions, Cambridge, UK). Statistical significance was assigned at the 5 per cent level. Results Initial EUA and MRI findings agreed in 15 patients. Fourteen had sepsis, of whom 13 had fistulas (Table 1,Fig. 1); three patients had secondary extensions. MRI findings precipitated additional surgery in three further patients, leading to agreement between MRI and EUA (Table 1, Fig. 2). The therapeutic impact of MRI was therefore 10 per cent. MRI revealed an internal opening in two patients initially thought to have a sinus, allowing definitive treatment, and prevented further exploration in a third in whom a horseshoe suspected clinically was absent on imaging. Fig. 1 Open in new tabDownload slide Axial short T1 inversion recovery sequence magnetic resonance image at mid anal canal level of a 42-year-old man in whom both magnetic resonance imaging and initial examination under anaesthesia had suggested a trans-sphincteric fistula (black arrowheads) with an internal opening at 6 o'clock (white arrow). Fistulotomy was performed and the fistula subsequently healed. IRF, ischiorectal fossa Fig. 2 Open in new tabDownload slide Axial short T1 inversion recovery sequence magnetic resonance image (MRI) at mid anal canal level in a 17-year-old girl. The surgeon, diagnosing a sinus following initial examination under anaesthesia, used the MRI for guidance, which confirmed an anterior internal opening and intersphincteric sepsis (white arrow). A seton was subsequently placed through a trans-sphincteric fistula Table 1 Comparison of findings at examination under anaesthesia and magnetic resonance imaging in 30 patients . Patients in whom MRI and initial EUA agreed . Patients in whom MRI and EUA agreed after further surgery (n = 3) . Patients in whom discrepancy between MRI and EUA persisted (n = 12) . . Findings at MRI and initial EUA . Initial EUA findings . Review EUA findings . MRI findings . EUA findings . Primary track classification  Nil  1 0 0  1  0  Sinus  1 2 0  0  1  Superficial or submucous  4 0 1  1  0  Intersphincteric  5 1 1  7  1  Trans-sphincteric  4 0 1  1  9  Suprasphincteric  0 0 0  1  1  Extrasphincteric  0 0 0  1  0  Total 15 3 3 12 12 Internal opening and level  Nil  2 2 0  1  1  Below dentate line  8 0 1  2  1  At dentate line  4 1 2  7  9  Above dentate line  1 0 0  1  1  Rectal  0 0 0  1  0  Total 15 3 3 12 12 Abscess  Superficial or submucous  1 1 1  1  0  Intersphincteric  1 0 0  0  0  Ischiorectal  1 0 0  1  2  Supralevator  0 0 0  1  1  Total  3 1 1  3  3 Horseshoe  Intersphincteric  0 1 0  1  0  Ischiorectal  0 0 0  0  1  Supralevator  0 0 0  0  0  Total  0 1 0  1  1 . Patients in whom MRI and initial EUA agreed . Patients in whom MRI and EUA agreed after further surgery (n = 3) . Patients in whom discrepancy between MRI and EUA persisted (n = 12) . . Findings at MRI and initial EUA . Initial EUA findings . Review EUA findings . MRI findings . EUA findings . Primary track classification  Nil  1 0 0  1  0  Sinus  1 2 0  0  1  Superficial or submucous  4 0 1  1  0  Intersphincteric  5 1 1  7  1  Trans-sphincteric  4 0 1  1  9  Suprasphincteric  0 0 0  1  1  Extrasphincteric  0 0 0  1  0  Total 15 3 3 12 12 Internal opening and level  Nil  2 2 0  1  1  Below dentate line  8 0 1  2  1  At dentate line  4 1 2  7  9  Above dentate line  1 0 0  1  1  Rectal  0 0 0  1  0  Total 15 3 3 12 12 Abscess  Superficial or submucous  1 1 1  1  0  Intersphincteric  1 0 0  0  0  Ischiorectal  1 0 0  1  2  Supralevator  0 0 0  1  1  Total  3 1 1  3  3 Horseshoe  Intersphincteric  0 1 0  1  0  Ischiorectal  0 0 0  0  1  Supralevator  0 0 0  0  0  Total  0 1 0  1  1 MRI, magnetic resonance imaging; EUA, examination under anaesthesia. Open in new tab Table 1 Comparison of findings at examination under anaesthesia and magnetic resonance imaging in 30 patients . Patients in whom MRI and initial EUA agreed . Patients in whom MRI and EUA agreed after further surgery (n = 3) . Patients in whom discrepancy between MRI and EUA persisted (n = 12) . . Findings at MRI and initial EUA . Initial EUA findings . Review EUA findings . MRI findings . EUA findings . Primary track classification  Nil  1 0 0  1  0  Sinus  1 2 0  0  1  Superficial or submucous  4 0 1  1  0  Intersphincteric  5 1 1  7  1  Trans-sphincteric  4 0 1  1  9  Suprasphincteric  0 0 0  1  1  Extrasphincteric  0 0 0  1  0  Total 15 3 3 12 12 Internal opening and level  Nil  2 2 0  1  1  Below dentate line  8 0 1  2  1  At dentate line  4 1 2  7  9  Above dentate line  1 0 0  1  1  Rectal  0 0 0  1  0  Total 15 3 3 12 12 Abscess  Superficial or submucous  1 1 1  1  0  Intersphincteric  1 0 0  0  0  Ischiorectal  1 0 0  1  2  Supralevator  0 0 0  1  1  Total  3 1 1  3  3 Horseshoe  Intersphincteric  0 1 0  1  0  Ischiorectal  0 0 0  0  1  Supralevator  0 0 0  0  0  Total  0 1 0  1  1 . Patients in whom MRI and initial EUA agreed . Patients in whom MRI and EUA agreed after further surgery (n = 3) . Patients in whom discrepancy between MRI and EUA persisted (n = 12) . . Findings at MRI and initial EUA . Initial EUA findings . Review EUA findings . MRI findings . EUA findings . Primary track classification  Nil  1 0 0  1  0  Sinus  1 2 0  0  1  Superficial or submucous  4 0 1  1  0  Intersphincteric  5 1 1  7  1  Trans-sphincteric  4 0 1  1  9  Suprasphincteric  0 0 0  1  1  Extrasphincteric  0 0 0  1  0  Total 15 3 3 12 12 Internal opening and level  Nil  2 2 0  1  1  Below dentate line  8 0 1  2  1  At dentate line  4 1 2  7  9  Above dentate line  1 0 0  1  1  Rectal  0 0 0  1  0  Total 15 3 3 12 12 Abscess  Superficial or submucous  1 1 1  1  0  Intersphincteric  1 0 0  0  0  Ischiorectal  1 0 0  1  2  Supralevator  0 0 0  1  1  Total  3 1 1  3  3 Horseshoe  Intersphincteric  0 1 0  1  0  Ischiorectal  0 0 0  0  1  Supralevator  0 0 0  0  0  Total  0 1 0  1  1 MRI, magnetic resonance imaging; EUA, examination under anaesthesia. Open in new tab Disagreement persisted between MRI and EUA assessments in 12 patients (Table 1). In nine, this related to minor differences in fistula classification, for example superficial versus low trans-sphincteric (Fig. 3). In one of these, discrepancy also related to extension location, and in another MRI missed an ischiorectal extension. In a further two patients disagreement related to a major difference in classification; in one MRI suggested an extrasphincteric fistula whereas EUA revealed a trans-sphincteric track opening at the anorectal junction, and in the other MRI suggested a suprasphincteric fistula with a supralevator extension, although the primary track was found to be intersphincteric at EUA. In a further patient, MRI missed a superficial sinus that was found easily at initial EUA. No review EUA was deemed necessary following disclosure of MRI findings in any of these 12 patients; discrepancies were judged as semantic variations in fistula classification that did not affect the operative approach. Fig. 3 Open in new tabDownload slide Axial short T1 inversion recovery sequence magnetic resonance image at the level of the subcutaneous external sphincter (white arrow) of a 47-year-old man who underwent fistulotomy of a trans-sphincteric track (white arrowhead) at examination under anaesthesia (EUA), which had been misclassified as superficial on magnetic resonance imaging. The surgeon did not need to perform a review EUA, and the patient subsequently healed In the outpatient department, surgeons stated that they would have requested MRI routinely in 14 of the 30 patients, including all three in whom MRI subsequently altered surgical approach. There was no significant difference between consultants and trainees with respect to this (five of 14 assessments versus nine of 16 respectively; P = 0·299). After initial EUA, surgeons felt that MRI findings would be helpful in 11 patients, including the three in whom MRI altered management. Again, there was no difference between consultants or trainees (eight of 23 versus three of seven respectively; P = 1·000). Overall, surgeons found imaging useful in 14 patients, including the three in whom MRI had therapeutic impact, and there was no difference between consultants or trainees (ten of 23 versus four of seven respectively; P = 0·675). Only one patient, from the group of three in whom MRI had precipitated further surgery, had further unplanned surgery, but for skin bridging rather than any new sepsis. Discussion Most fistulas in ano are easily managed by surgery alone11, although MRI is useful in complex cases. Lunniss et al.4 found that MRI could classify cryptoglandular fistula and also predict relapse with greater accuracy than EUA. A later study, which independently categorized fistulas into simple or complex based on MRI and EUA assessments, found that MRI better predicted relapse5. This has resulted in recommendations to employ preoperative MRI in all patients with complex fistulas12, an approach that can significantly reduce further recurrence7. However, the role of MRI in the management of seemingly simple fistulas is less clear. A study of 27 patients, in which MRI findings were revealed during operation, included 25 with primary fistulas, and MRI precipitated additional surgery in only one patient13. However, the authors admitted that their radiologist was initially unfamiliar with fistula MRI and concluded that the considerable learning curve associated with the technique may have prejudiced their results. The present study aimed to determine the therapeutic effect of MRI in patients with primary fistula in ano using experienced radiologists, and to document the cause of any relapse, using a well validated and previously applied6,7 pre- and post-intervention design that measures the therapeutic effect of imaging14. Fistula recurrence is usually due to sepsis missed at EUA3. Extensions occur in approximately 10–15 per cent of patients15, and are more prevalent in recurrent or Crohn's fistulas. This study has confirmed their relative scarcity in consecutive patients with primary fistula. Nevertheless, imaging influenced surgery in 10 per cent, a figure not too distant from the 21 per cent established in a group of 71 subjects with recurrent disease7. However, in the latter study relapse occurred in over 50 per cent of patients in whom there was persistent disagreement between MRI and EUA7. Most disagreement related to the identification of sepsis by MRI, unlike the present study in which disagreement was predominantly related to semantic variations in classification. The two patients in this study believed to have a sinus at EUA, but who were found to have a fistula following disclosure of MRI results, would probably have suffered a recurrence in the absence of imaging7. Indeed, there was no relapse due to new sepsis in the present study, and further unplanned surgery was required only to divide skin bridging in one patient; longer-term follow-up might verify this. Whether MRI should be performed in all patients with primary fistula in ano in return for a therapeutic impact of 10 per cent is a matter for local debate. MRI is a target for cost control8. However, the scanning costs for this study totalled £2100, substantially less than the estimated cost of £2844 to treat a single recurrence, notwithstanding the personal cost to the patient in terms of suffering and lost employment. A cheaper alternative may be anal endosonography which, while initially disappointing in detecting fistula in ano16, has shown greater promise in studies using higher-frequency transducers17. Further research is required in this area. Although the therapeutic impact of MRI was 10 per cent, in 11 patients the surgeons perceived a need for imaging before the results were revealed and felt that it was useful in retrospect in 14 patients. This was most probably related to increased diagnostic confidence, which may occur without any formal change in operative approach or fistula classification. MRI was not always correct in the present study. However, disagreement was usually clinically insignificant, mostly reflecting imprecise localization of the height of the internal opening, which occurs because the dentate line is not discernable on MRI, even when using endoanal coils with improved spatial resolution18. Acknowledgements The authors are grateful for a Kodak bursary via the Royal College of Radiologists, to the surgeons of St Mark's Hospital, who allowed their patients to enter this study, and to Carole Jenkins, Sister Annie Driscoll and Caroline Francis for their contribution. References 1 Thompson H . The orthodox conception of fistula-in-ano and its treatment . Proc R Soc Med 1962 ; 55 : 754 – 756 . Google Scholar OpenURL Placeholder Text WorldCat 2 Lilius HG . Fistula-in-ano, an investigation of human foetal anal ducts and intramuscular glands and a clinical study of 150 patients . Acta Chir Scand Suppl 1968 ; 383 : 7 – 88 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 3 Seow-Choen F , Phillips RKS. Insights gained from the management of problematical anal fistulae at St Mark's Hospital, 1984–1988 . Br J Surg 1991 ; 78 : 539 – 541 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Lunniss PJ , Armstrong P, Barker PG, Reznek RH, Phillips RKS. Magnetic resonance imaging of anal fistulae . Lancet 1992 ; 340 : 394 – 396 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Spencer JA , Chapple K, Wilson D, Ward J, Windsor AC, Ambrose NS. Outcome after surgery for perianal fistula: predictive value of MR imaging . AJR Am J Roentgenol 1998 ; 171 : 403 – 406 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Beets-Tan RG , Beets GL, van der Hoop AG, Kessels AG, Vliegen RF, Baeten CG et al. Preoperative MR imaging of anal fistulas: does it really help the surgeon? Radiology 2001 ; 218 : 75 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Buchanan G , Halligan S, Williams A, Cohen CRG, Tarroni D, Phillips RKS et al. The effect of MRI on clinical outcome of recurrent fistula-in-ano . Lancet 2002 ; 360 : 1661 – 1662 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Halligan S , Bartram CI. MR imaging of fistula in ano: are endoanal coils the gold standard? AJR Am J Roentgenol 1998 ; 171 : 407 – 412 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Parks AG , Gordon PH, Hardcastle JD. A classification of fistula-in-ano . Br J Surg 1976 ; 63 : 1 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Lunniss PJ , Barker PG, Sultan AH, Armstrong P, Reznek RH, Bartram CI et al. Magnetic resonance imaging of fistula-in-ano . Dis Colon Rectum 1994 ; 37 : 708 – 718 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Shouler PJ , Grimley RP, Keighley MR, Alexander-Williams J. Fistula-in-ano is usually simple to manage surgically . Int J Colorectal Dis 1986 ; 1 : 113 – 115 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Zbar AP , de Souza NM, Puni R, Kmiot WA. Comparison of endoanal magnetic resonance imaging with surgical findings in perirectal sepsis . Br J Surg 1998 ; 85 : 111 – 114 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Scholefield JH , Berry DP, Armitage NC, Wastie ML. Magnetic resonance imaging in the management of fistula in ano . Int J Colorectal Dis 1997 ; 12 : 276 – 279 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Dixon AK . Evidence-based diagnostic radiology . Lancet 1997 ; 350 : 509 – 512 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Marks CG , Ritchie JK. Anal fistulas at St Mark's Hospital . Br J Surg 1977 ; 64 : 84 – 91 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Choen S , Burnett S, Bartram CI, Nicholls RJ. Comparison between anal endosonography and digital examination in the evaluation of anal fistulae . Br J Surg 1991 ; 78 : 445 – 447 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Schwartz DA , Wiersema MJ, Dudiak KM, Fletcher JG, Clain JE, Tremaine WJ et al. A comparison of endoscopic ultrasound, magnetic resonance imaging, and exam under anesthesia for evaluation of Crohn's perianal fistulas . Gastroenterology 2001 ; 121 : 1064 – 1072 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Stoker J , Hussain SM, van Kempen D, Elevelt AJ, Lameris JS. Endoanal coil in MR imaging of anal fistulas . AJR Am J Roentgenol 1996 ; 166 : 360 – 362 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2003 British Journal of Surgery 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) Copyright © 2003 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.
TI - Magnetic resonance imaging for primary fistula in ano
JO - British Journal of Surgery
DO - 10.1002/bjs.4125
DA - 2003-07-04
UR - https://www.deepdyve.com/lp/oxford-university-press/magnetic-resonance-imaging-for-primary-fistula-in-ano-YKAgF117bF
SP - 877
EP - 881
VL - 90
IS - 7
DP - DeepDyve
ER -