Right minithoracotomy versus median sternotomy for reoperative mitral valve surgery: a systematic review and meta-analysis of observational studies

Right minithoracotomy versus median sternotomy for reoperative mitral valve surgery: a systematic... Abstract Reoperative mitral valve surgery (MVS) through a median sternotomy (ST-MVS) can be particularly challenging due to dense adhesions and is known to carry a substantial risk of injuries to vascular structures. These injuries occur in 7–9% of cases and are associated with increased mortality rates. A valid alternative that could avoid the risks associated with redo ST-MVS is the right anterolateral minithoracotomy (MT-MVS) approach. The aim of this study was to quantify the effects of MT-MVS compared with those of ST-MVS on morbidity and mortality among patients who underwent prior cardiac surgery through a sternotomy. The MEDLINE and EMBASE databases were searched through 1 November 2017. Data regarding mortality, stroke, reoperation for bleeding and length of hospital stay were extracted and submitted to a meta-analysis using random effects modelling and the I2-test for heterogeneity. Six retrospective observational studies were included, enrolling a total of 777 patients. In a pooled analysis, MT-MVS demonstrated reduced mortality rates compared to a standard sternotomy [odds ratio (OR) 0.41, 95% confidence interval (CI) 0.18–0.96; P = 0.04]. MT-MVS was, moreover, associated with reduced length of hospital stay [difference between the means was −3.81, 95% CI −5.53 to −2.08; P < 0.0001) and reoperation for bleeding (OR 0.32, 95% CI 0.10–0.99; P = 0.0488). The incidence of stroke was similar (OR 1.51, 95% CI 0.65–3.54; P = 0.34), all in the absence of heterogeneity. In conclusion, reoperative minimally invasive MVS through a minithoracotomy is a safe alternative to standard sternotomy, with reduced mortality rates, length of hospital stay and reoperations for bleeding and a comparable risk of stroke. However, because the existing literature provided limited, low-quality evidence, more methodologically rigorous randomized controlled trials are needed. Minimally invasive surgery , Mitral valve surgery , Right minithoracotomy , Reoperation , Median sternotomy INTRODUCTION Redo cardiac surgery has been associated with increased mortality rates compared to primary surgery [1, 2]. Cardiac redo procedures are traditionally performed through a repeat median sternotomy (ST). For the past decade, reoperative mitral valve surgery (MVS) has become more common, representing over 10% of all mitral valve procedures in the USA [3, 4]. However, redo MVS performed through an ST (ST-MVS) can be particularly technically challenging and is known to carry a substantial risk of injuries to patent coronary artery bypass grafts and vascular structures that lie directly substernally and can adhere to the sternum. Resternotomy may furthermore be demanding in patients with (healed) mediastinitis, prior thoracic radiotherapy and dense adhesions or other complications from prior surgery [5–7]. These injuries to cardiac structures occur in 7 to 9% of resternotomies [3, 8, 9] and are reported to be an independent risk factor for in-hospital death [3]. A valid alternative to repeated conventional ST-MVS would be a minimally invasive approach through a right anterolateral minithoracotomy (MT) [5, 10]. An incision of <10 cm is made in the 4th or 5th intercostal space, the goal being to minimize surgical trauma compared to that of a full ST or thoracotomy (20 cm) [11, 12]. MT-MVS can be performed either under direct or video-assisted vision, with the use of long-shafted instruments in both situations. Primary MT-MVS is, besides being associated with less surgical trauma, believed to result in diminished pain, blood loss and need for transfusions, which translates into reduced length of hospital stay (LOHS) [13]. In addition, with MT-MVS, one could avoid the risks associated with resternotomy. Despite these advantages, no general consensus exists on the approach of choice for redo MVS. This consensus should ideally arise from well-designed randomized controlled trials and comprehensive literature reviews that compare redo MT- and ST-MVS among patients with a prior ST. However, to date such data are only available from non-randomized studies, in the form of 2 best evidence topics [14, 15] and 1 small narrative subreview [16]. Therefore, an overview and analysis of the available comparative data, which may aid in determining the optimal approach for redo MVS, are needed. The aim of this report was to review all published observational studies that compare redo MVS through a MT and a conventional median ST approach among patients with prior cardiac surgery through a ST, with mortality as the primary outcome measure. Secondary outcome measures include stroke, reoperation for bleeding, LOHS, wound infection and red blood cell (RBC) transfusions. In order to draw more useful and robust conclusions regarding these outcome metrics, data from individual studies were collected and analysed using meta-analytical techniques. MATERIALS AND METHODS Methods of the analysis, outcome measures and inclusion criteria were specified in advance and documented in a protocol. The review and meta-analysis were conducted using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement [17]. Risk of bias in individual studies and across studies was not assessed, because retrospective observational studies are generally known for possessing a certain level of bias. Eligibility criteria Types of participants Participants aged 18 years and above who presented with mitral valve disease requiring surgery and who underwent at least 1 previous cardiac surgical procedure through a median ST were considered for inclusion. No exclusions were made based on the indication of primary ST (e.g. coronary artery bypass graft, aortic valve replacement). Types of interventions Studies that compared the MT (also denoted as ‘port access’) and the median ST approach for reoperative MVS were considered. Studies in which cardiac procedures were performed concomitantly with the initial mitral valve repair or replacement were also assessed for eligibility. The MT approach was defined as a ≤10 cm right anterolateral incision in the 4th or 5th intercostal space to acquire surgical exposure. No difference was made between video-assisted or direct vision MT; however, studies utilizing robotic telemanipulation such as the Da Vinci robot were excluded because telemanipulation forms a totally different area within minimally invasive MVS. If the report did not define the thoracotomy approach as either mini or full, the authors were contacted for clarification. Redo ST was defined as MVS in which the previous ST incision was re-entered and cardiac exposure was obtained by an oscillating saw to complete the ST. Type of primary outcome measure The primary outcome measure was mortality, reported as in-hospital death, 30-day mortality rate or death as an early postoperative complication. Types of studies Observational studies comparing MT and ST for reoperative MVS after a previous median ST were examined for eligibility. Studies reporting combined data for mitral and aortic valve reoperations were considered only if mitral valve data were presented separately. Studies reporting primary operations were excluded. Literature search Potentially eligible studies were identified by searching the electronic MEDLINE and EMBASE databases through PubMed and Ovid, respectively. No unpublished data were obtained. The search was limited to the English language, human subjects and studies published after 2000, because the latter provide the best evidence for current practice. No publication status restrictions were imposed. The last search was run on 1 November 2017. In addition, a cross-reference and related-articles search was conducted as a check of rigour. The search strategy was first applied to the electronic MEDLINE database and combined the following MeSH and free terms: ‘mitral valve’, ‘mitral valve insufficiency’, ‘mitral valve prolapse’ AND ‘thoracotomy’, ‘minithoracotomy’, ‘port access’ AND ‘ST’ AND ‘reoperation’, ‘reoperative’ and ‘redo’. This search was subsequently adapted for EMBASE. Reports originating from the electronic search were screened for eligibility based on their titles and abstracts. Subsequently, full texts of potentially eligible reports were read and carefully assessed according to the eligibility criteria. Studies meeting these criteria were included for review and, if applicable, for quantitative synthesis (meta-analysis). Study selection was conducted in a non-blinded standardized manner by 2 independent reviewers. Potential inter-reviewer disagreements were resolved by consensus. Data collection For this review, a data extraction sheet was developed and pilot-tested on 3 randomly selected included studies, whereupon the sheet was refined accordingly. Data extraction was performed by 1 review author. The second author validated the correctness of the tabulated data. Potential inter-reviewer disagreements were resolved by consensus. Studies reporting their continuous variables as mean and standard deviation (SD) were extracted without conversion. Variables denoted as median and interquartile range or standard error of the mean were first converted, as described elsewhere [18, 19]. Data were extracted from each included report on (i) general study characteristics: study design, enrolment period, country, setting, in- and exclusion criteria, baseline characteristics and statistical methods; (ii) characteristics of participants: number of patients, mean age (years), gender (male/female), previous surgery and preoperative characteristics such as Society of Thoracic Surgeons (STS) score or EuroSCORE, ejection fraction, New York Heart Association class and several comorbidities; (iii) intervention characteristics: mean time to redo surgery (years), cannulation site, clamping technique, myocardial protection method, concomitant procedures, repair rate in redo surgeries (mitral valve repair versus replacement), conversion to ST (redo MT-MVS only), cardiopulmonary bypass (CPB) and clamping time; (iv) primary outcome measure: death, either reported as 30-day mortality, in-hospital death or early postoperative death; (v) secondary outcome measures: stroke, reoperation for bleeding, LOHS, wound infection and RBC transfusions. If outcome measures were reported in any other way than stated and could not be converted, data were assumed to be not available. In addition, it must be noted that the authors did not discriminate between minor differences in surgical procedures. Statistical analysis Odds ratios (ORs) were used to assess dichotomous outcome measures. The difference in means (MD) was used for continues variables, which were made on the same scale among all studies. If not, the standardized difference was used. Obtained ORs were interpreted as risk ratios. In addition, an OR or MD less than 1 favours MT over ST for MVS. Statistical analyses were performed using Review Manager (RevMan v5.3, Cochrane Collaboration, Oxford, UK). For this analysis, the random effects model with a 95% confidence interval (CI) was used when there was a substantial risk of heterogeneity, originating from the non-randomized nature of the included studies. In addition, because risk profiles and selection criteria differed between centres, the random effects model was favoured. The I2-test for heterogeneity was conducted to assess variability across studies that could not be due to random error alone. High I2-values indicated that the observed variability among studies could not be explained by chance (i.e. a consequence of clinical and/or methodological diversity). Heterogeneity was deemed to be substantial and considerable if I2 >50% and I2 >75%, respectively with P-value <0.10 [18]. No additional analyses were conducted. RESULTS Study selection A total of 6 studies were identified for inclusion in this review and meta-analysis. The MEDLINE and EMBASE database search provided a cumulative number of 250 citations. In addition, 1 citation was obtained via a cross-reference and related-article search. Of these, 84 duplicates were discarded. Another 151 papers were eliminated because their title and abstract clearly did not meet the eligibility criteria. Full texts of the remaining 16 articles were assessed in detail for eligibility. Of these, 10 reports did not meet the criteria as described. Reasons for exclusion were no comparison between MT and median ST (n = 6), results for mitral valve reoperation were not separately reported (n = 3) and the main text for 1 study was in Japanese that could not be translated into English (n = 1). In addition, no unpublished relevant studies were obtained (flow diagram, Fig. 1). Figure 1: View largeDownload slide Study selection procedure shown in a PRISMA flow diagram. Figure 1: View largeDownload slide Study selection procedure shown in a PRISMA flow diagram. Study characteristics Methods All 6 studies finally selected for inclusion were retrospective single-centre observational studies published in English. All studies were conducted in the USA, Canada, Japan or South Korea. The main inclusion criterion encompassed patients who received a MT or median ST for mitral valve disease with at least 1 prior cardiac surgical procedure through a median ST. Overall, studies enrolled patients who were operated on between 1985 and 2011. Participants A total number of 777 participants were included for review. These were subdivided in an MT (n = 237; 31%; mean age 62.8 ± 14.0 years) and an ST (n = 540; 69%; mean age 60.1 ± 14.8 years) group. The number of subjects per individual study ranged from 35 to 287 [20–25] (Table 1). Previous operations as well as preoperative patient characteristics are presented in Table 2. STS scores [25] and EuroSCOREs [23] were both only reported in 1 paper each, whereas others mainly reported ejection fraction, New York Heart Association class and comorbidities such as diabetes mellitus and hypertension. Table 1: Principal data and outcomes of interest from individual studies Study  Country  Study design  Study period  Group (MT/ST)  Participants (total (male/female))  Age (years)  Mortality (events (%))  Stroke (events (%))  LOHS (days)  RFB (events (%))  Wound infection (events (%))  RBC transfusion (patients (%))  Burfeind et al. [20]  USA  Retrospective observational study  1996–2001  MT  60 (33/27)  60 ± 14  0 (0)  4 (7)  NA  1 (2)  NA  NA        1985–1997  ST  155 (42/113)  58 ± 16  21 (14)  7 (5)  NA  8 (5)  NA  NA  Bolotin et al. [21]  USA  Retrospective observational study  January 1996–June 2003  MT  38 (-/-)  68 ± 2  2 (5)  NA  7 ± 6  NA  NA  NA          ST  33 (-/-)  63 ± 2  2 (6)  NA  11 ± 6  NA  NA  NA  Kim et al. [22]  South Korea  Retrospective observational study  September 2007–December 2010  MT  22 (4/18)  46 ± 15  0 (0)  NA  16 ± 6  0 (0)  NA  NA          ST  13 (5/8)  45 ± 16  1 (8)  NA  20 ± 17  2 (15)  NA  NA  Hiraoka et al. [23]  Japan  Retrospective observational study  January 2006–September 2011  MT  10 (5/5)  68 ± 15  0 (0)  1 (10)  18 ± 8  0 (0)  NA  NA          ST  27 (18/9)  63 ± 15  1 (4)  1 (4)  22 ± 13  2 (7)  NA  NA  Vallabhajosyula et al. [24]  USA  Retrospective observational study  1988–2001  MT  67 (32/35)  64 ± 13  2 (3)  2 (3)  11 ± 15  0 (0)  0 (0)  NA        ST  220 (99/121)  61 ± 15  13 (6)  7 (3)  14 ± 12  3 (1)  1 (0.5)  NA  Losenno et al. [25]  Canada  Retrospective observational study  September 2000–August 2014  MT  40 (28/12)  68 ± 14  2 (5)  2 (5)  8 ± 7  1 (3)  0 (0)  25 (63)          ST  92 (38/54)  62 ± 13  10 (11)  2 (2)  12 ± 10  6 (7)  3 (3)  73 (79)  Study  Country  Study design  Study period  Group (MT/ST)  Participants (total (male/female))  Age (years)  Mortality (events (%))  Stroke (events (%))  LOHS (days)  RFB (events (%))  Wound infection (events (%))  RBC transfusion (patients (%))  Burfeind et al. [20]  USA  Retrospective observational study  1996–2001  MT  60 (33/27)  60 ± 14  0 (0)  4 (7)  NA  1 (2)  NA  NA        1985–1997  ST  155 (42/113)  58 ± 16  21 (14)  7 (5)  NA  8 (5)  NA  NA  Bolotin et al. [21]  USA  Retrospective observational study  January 1996–June 2003  MT  38 (-/-)  68 ± 2  2 (5)  NA  7 ± 6  NA  NA  NA          ST  33 (-/-)  63 ± 2  2 (6)  NA  11 ± 6  NA  NA  NA  Kim et al. [22]  South Korea  Retrospective observational study  September 2007–December 2010  MT  22 (4/18)  46 ± 15  0 (0)  NA  16 ± 6  0 (0)  NA  NA          ST  13 (5/8)  45 ± 16  1 (8)  NA  20 ± 17  2 (15)  NA  NA  Hiraoka et al. [23]  Japan  Retrospective observational study  January 2006–September 2011  MT  10 (5/5)  68 ± 15  0 (0)  1 (10)  18 ± 8  0 (0)  NA  NA          ST  27 (18/9)  63 ± 15  1 (4)  1 (4)  22 ± 13  2 (7)  NA  NA  Vallabhajosyula et al. [24]  USA  Retrospective observational study  1988–2001  MT  67 (32/35)  64 ± 13  2 (3)  2 (3)  11 ± 15  0 (0)  0 (0)  NA        ST  220 (99/121)  61 ± 15  13 (6)  7 (3)  14 ± 12  3 (1)  1 (0.5)  NA  Losenno et al. [25]  Canada  Retrospective observational study  September 2000–August 2014  MT  40 (28/12)  68 ± 14  2 (5)  2 (5)  8 ± 7  1 (3)  0 (0)  25 (63)          ST  92 (38/54)  62 ± 13  10 (11)  2 (2)  12 ± 10  6 (7)  3 (3)  73 (79)  Data are expressed as absolute number, percentages or as mean ± SD. LOHS: length of hospital stay; MT: minithoracotomy; NA: not available; RBC: red blood cell; RFB: reoperation for bleeding; SD: standard deviation; ST: median sternotomy; USA: United States of America. Table 1: Principal data and outcomes of interest from individual studies Study  Country  Study design  Study period  Group (MT/ST)  Participants (total (male/female))  Age (years)  Mortality (events (%))  Stroke (events (%))  LOHS (days)  RFB (events (%))  Wound infection (events (%))  RBC transfusion (patients (%))  Burfeind et al. [20]  USA  Retrospective observational study  1996–2001  MT  60 (33/27)  60 ± 14  0 (0)  4 (7)  NA  1 (2)  NA  NA        1985–1997  ST  155 (42/113)  58 ± 16  21 (14)  7 (5)  NA  8 (5)  NA  NA  Bolotin et al. [21]  USA  Retrospective observational study  January 1996–June 2003  MT  38 (-/-)  68 ± 2  2 (5)  NA  7 ± 6  NA  NA  NA          ST  33 (-/-)  63 ± 2  2 (6)  NA  11 ± 6  NA  NA  NA  Kim et al. [22]  South Korea  Retrospective observational study  September 2007–December 2010  MT  22 (4/18)  46 ± 15  0 (0)  NA  16 ± 6  0 (0)  NA  NA          ST  13 (5/8)  45 ± 16  1 (8)  NA  20 ± 17  2 (15)  NA  NA  Hiraoka et al. [23]  Japan  Retrospective observational study  January 2006–September 2011  MT  10 (5/5)  68 ± 15  0 (0)  1 (10)  18 ± 8  0 (0)  NA  NA          ST  27 (18/9)  63 ± 15  1 (4)  1 (4)  22 ± 13  2 (7)  NA  NA  Vallabhajosyula et al. [24]  USA  Retrospective observational study  1988–2001  MT  67 (32/35)  64 ± 13  2 (3)  2 (3)  11 ± 15  0 (0)  0 (0)  NA        ST  220 (99/121)  61 ± 15  13 (6)  7 (3)  14 ± 12  3 (1)  1 (0.5)  NA  Losenno et al. [25]  Canada  Retrospective observational study  September 2000–August 2014  MT  40 (28/12)  68 ± 14  2 (5)  2 (5)  8 ± 7  1 (3)  0 (0)  25 (63)          ST  92 (38/54)  62 ± 13  10 (11)  2 (2)  12 ± 10  6 (7)  3 (3)  73 (79)  Study  Country  Study design  Study period  Group (MT/ST)  Participants (total (male/female))  Age (years)  Mortality (events (%))  Stroke (events (%))  LOHS (days)  RFB (events (%))  Wound infection (events (%))  RBC transfusion (patients (%))  Burfeind et al. [20]  USA  Retrospective observational study  1996–2001  MT  60 (33/27)  60 ± 14  0 (0)  4 (7)  NA  1 (2)  NA  NA        1985–1997  ST  155 (42/113)  58 ± 16  21 (14)  7 (5)  NA  8 (5)  NA  NA  Bolotin et al. [21]  USA  Retrospective observational study  January 1996–June 2003  MT  38 (-/-)  68 ± 2  2 (5)  NA  7 ± 6  NA  NA  NA          ST  33 (-/-)  63 ± 2  2 (6)  NA  11 ± 6  NA  NA  NA  Kim et al. [22]  South Korea  Retrospective observational study  September 2007–December 2010  MT  22 (4/18)  46 ± 15  0 (0)  NA  16 ± 6  0 (0)  NA  NA          ST  13 (5/8)  45 ± 16  1 (8)  NA  20 ± 17  2 (15)  NA  NA  Hiraoka et al. [23]  Japan  Retrospective observational study  January 2006–September 2011  MT  10 (5/5)  68 ± 15  0 (0)  1 (10)  18 ± 8  0 (0)  NA  NA          ST  27 (18/9)  63 ± 15  1 (4)  1 (4)  22 ± 13  2 (7)  NA  NA  Vallabhajosyula et al. [24]  USA  Retrospective observational study  1988–2001  MT  67 (32/35)  64 ± 13  2 (3)  2 (3)  11 ± 15  0 (0)  0 (0)  NA        ST  220 (99/121)  61 ± 15  13 (6)  7 (3)  14 ± 12  3 (1)  1 (0.5)  NA  Losenno et al. [25]  Canada  Retrospective observational study  September 2000–August 2014  MT  40 (28/12)  68 ± 14  2 (5)  2 (5)  8 ± 7  1 (3)  0 (0)  25 (63)          ST  92 (38/54)  62 ± 13  10 (11)  2 (2)  12 ± 10  6 (7)  3 (3)  73 (79)  Data are expressed as absolute number, percentages or as mean ± SD. LOHS: length of hospital stay; MT: minithoracotomy; NA: not available; RBC: red blood cell; RFB: reoperation for bleeding; SD: standard deviation; ST: median sternotomy; USA: United States of America. Table 2: Supplementary data of interest from individual studies Study  Group  Previous surgery (%)  Mean time to redo surgery (years)  Preoperative patient characteristics  Cannulation site  Clamping technique  Myocardial protection  Clamping time (min)  CPB time (min)  MVP/ MVR (%)  Conversion to sternotomy (number of patients)  Burfeind et al. [20]  MT  MVS (60)  NA  LVEF (45 ± 9%) and mean NYHA (3.4)  FV-FA/AscA  Endoclamp (45%)  VF (55%) and additional percutaneous retrograde cardioplegia (33%)  NA  208 ± 76  NA  NA  ST  MVS (83)  NA  LVEF (54 ± 13%) and mean NYHA (3.5)  IVC and SVC-AscA  NA  VF (5%) and cardioplegia (95%)  NA  157 ± 53  NA    Bolotin et al. [21]  MT  NA  NA  LVEF (46 ± 12%) and mean NYHA (2.7)  FV and JV-FA    Hypothermia (26°C) and VF  NA  160 ± 65  42/58  0  ST  NA  NA  LVEF (55 ± 11%) and mean NYHA (2.6)  NA  NA  NA  NA  157 ± 54  9/91    Kim et al. [22]  MT  MVP (59), MVR (9), MVR + TVP/AVR (9), MVP + TVP (5) and other (18)  12 ± 9  LVEF (61 ± 9%), hypertension (5%) and diabetes mellitus (5%)  FV-FA  Transthoracic clamp  Moderate hypothermia and antegrade cardioplegia  91 ± 27  171 ± 47  77/23  NA  ST  MVP (46), MVP + TVP (8), MVR (8), MVR + TVP/AVR (8), CABG (8), AVR (8) and other (22)  10 ± 9  LVEF (45 ± 16%), hypertension (39%) and diabetes mellitus (23%)  IVC and SVC-AscA  NA  NA  102 ± 57  210 ± 103  100/0    Hiraoka et al. [23]  MT  MVP (50), redo MVR (10), AVR (10) and CABG (30)  NA  LVEF (47 ± 19%), mean NYHA (1.3 ± 0.5) and EuroSCORE (4.8 ± 2.0)  FV and JV-FA/AA    Hypothermia (27–30°C) and VF  90 ± 7 (VF-time)  145 ± 25  0/100  NA  ST  MVP (33), MVR (19), redo MVR (4), AVR (11), CABG (4), Bentall (4) and other (25)  NA  LVEF (64 ± 9), mean NYHA (1.3 ± 0.7) and EuroSCORE (3.8 + 2.4)  FV and SVC-FA  Aortic cross-clamp  Antegrade cardioplegia  84 ± 19  135 ± 28  0/100    Vallabhajosyula et al. [24]  MT  MVP (78), MVR (19) and MVP + MVR (3)  NA  LVEF (54 ± 12%), NYHA ≥2 (63%), hypertension (61%) and diabetes mellitus (13%)  FV and SVC-FA (97%)/AA (3%)  Endoclamp (81%) and Chitwood clamp (7%)  Hypothermia, antegrade cardioplegia (88%) and VF (12%)  104 ± 38  153 ± 42  NA  1  ST  MVP (41), MVR (53) and MVP + MVR (6)  NA  LVEF (54 ± 17%), NYHA ≥2 (69%), hypertension (55%) and diabetes mellitus (17%)  Central aortic cannulation (95%) and FA (5%)  Aortic cross-clamp (100%)  Antegrade and retrograde cardioplegia  130 ± 71  172 ± 83  NA    Losenno et al. [25]  MT  Isolated MVR/MVP (18), MVR/MVP ± other valve ± CABG (13), CABG (50), AVR/repair ± aortic ± CABG (18) and other (18)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (28%)  FV and JV-FA/AA    Hypothermia (28–30°C), VF  123 ± 37 (VF-time)  201 ± 63  20/80  0    ST  Isolated MVR/MVP (62), MVR/MVP ± other valve ± CABG (17), CABG (11), AVR/repair ± aortic ± CABG (5) and other (10)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (15%)  IVC and SVC-AscA  Aortic cross-clamp  Combination of antegrade and retrograde blood cardioplegia (100%)  105 ± 46  180 ± 75  9/91    Study  Group  Previous surgery (%)  Mean time to redo surgery (years)  Preoperative patient characteristics  Cannulation site  Clamping technique  Myocardial protection  Clamping time (min)  CPB time (min)  MVP/ MVR (%)  Conversion to sternotomy (number of patients)  Burfeind et al. [20]  MT  MVS (60)  NA  LVEF (45 ± 9%) and mean NYHA (3.4)  FV-FA/AscA  Endoclamp (45%)  VF (55%) and additional percutaneous retrograde cardioplegia (33%)  NA  208 ± 76  NA  NA  ST  MVS (83)  NA  LVEF (54 ± 13%) and mean NYHA (3.5)  IVC and SVC-AscA  NA  VF (5%) and cardioplegia (95%)  NA  157 ± 53  NA    Bolotin et al. [21]  MT  NA  NA  LVEF (46 ± 12%) and mean NYHA (2.7)  FV and JV-FA    Hypothermia (26°C) and VF  NA  160 ± 65  42/58  0  ST  NA  NA  LVEF (55 ± 11%) and mean NYHA (2.6)  NA  NA  NA  NA  157 ± 54  9/91    Kim et al. [22]  MT  MVP (59), MVR (9), MVR + TVP/AVR (9), MVP + TVP (5) and other (18)  12 ± 9  LVEF (61 ± 9%), hypertension (5%) and diabetes mellitus (5%)  FV-FA  Transthoracic clamp  Moderate hypothermia and antegrade cardioplegia  91 ± 27  171 ± 47  77/23  NA  ST  MVP (46), MVP + TVP (8), MVR (8), MVR + TVP/AVR (8), CABG (8), AVR (8) and other (22)  10 ± 9  LVEF (45 ± 16%), hypertension (39%) and diabetes mellitus (23%)  IVC and SVC-AscA  NA  NA  102 ± 57  210 ± 103  100/0    Hiraoka et al. [23]  MT  MVP (50), redo MVR (10), AVR (10) and CABG (30)  NA  LVEF (47 ± 19%), mean NYHA (1.3 ± 0.5) and EuroSCORE (4.8 ± 2.0)  FV and JV-FA/AA    Hypothermia (27–30°C) and VF  90 ± 7 (VF-time)  145 ± 25  0/100  NA  ST  MVP (33), MVR (19), redo MVR (4), AVR (11), CABG (4), Bentall (4) and other (25)  NA  LVEF (64 ± 9), mean NYHA (1.3 ± 0.7) and EuroSCORE (3.8 + 2.4)  FV and SVC-FA  Aortic cross-clamp  Antegrade cardioplegia  84 ± 19  135 ± 28  0/100    Vallabhajosyula et al. [24]  MT  MVP (78), MVR (19) and MVP + MVR (3)  NA  LVEF (54 ± 12%), NYHA ≥2 (63%), hypertension (61%) and diabetes mellitus (13%)  FV and SVC-FA (97%)/AA (3%)  Endoclamp (81%) and Chitwood clamp (7%)  Hypothermia, antegrade cardioplegia (88%) and VF (12%)  104 ± 38  153 ± 42  NA  1  ST  MVP (41), MVR (53) and MVP + MVR (6)  NA  LVEF (54 ± 17%), NYHA ≥2 (69%), hypertension (55%) and diabetes mellitus (17%)  Central aortic cannulation (95%) and FA (5%)  Aortic cross-clamp (100%)  Antegrade and retrograde cardioplegia  130 ± 71  172 ± 83  NA    Losenno et al. [25]  MT  Isolated MVR/MVP (18), MVR/MVP ± other valve ± CABG (13), CABG (50), AVR/repair ± aortic ± CABG (18) and other (18)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (28%)  FV and JV-FA/AA    Hypothermia (28–30°C), VF  123 ± 37 (VF-time)  201 ± 63  20/80  0    ST  Isolated MVR/MVP (62), MVR/MVP ± other valve ± CABG (17), CABG (11), AVR/repair ± aortic ± CABG (5) and other (10)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (15%)  IVC and SVC-AscA  Aortic cross-clamp  Combination of antegrade and retrograde blood cardioplegia (100%)  105 ± 46  180 ± 75  9/91    Data are expressed as absolute number, percentages or mean ± SD. a Total >100%; some patients received >1 previous surgery. AA: axillary artery; AscA: ascending aorta; AVR: aortic valve replacement; CABG: coronary artery bypass graft; CPB: cardiopulmonary bypass; FA: femoral artery; FV: femoral vein; IVC: inferior vena cava; JV: jugular vein; LVEF: left ventricular ejection fraction; MT: minithoracotomy; MVP: mitral valve plasty; MVR: mitral valve replacement; MVS: mitral valve surgery; NA: not available; NYHA: New York Heart Association; SD: standard deviation; ST: median sternotomy; STS: Society of Thoracic Surgeons; SVC: superior vena cava; TVP: tricuspid valve plasty; VF: ventricular fibrillation. Table 2: Supplementary data of interest from individual studies Study  Group  Previous surgery (%)  Mean time to redo surgery (years)  Preoperative patient characteristics  Cannulation site  Clamping technique  Myocardial protection  Clamping time (min)  CPB time (min)  MVP/ MVR (%)  Conversion to sternotomy (number of patients)  Burfeind et al. [20]  MT  MVS (60)  NA  LVEF (45 ± 9%) and mean NYHA (3.4)  FV-FA/AscA  Endoclamp (45%)  VF (55%) and additional percutaneous retrograde cardioplegia (33%)  NA  208 ± 76  NA  NA  ST  MVS (83)  NA  LVEF (54 ± 13%) and mean NYHA (3.5)  IVC and SVC-AscA  NA  VF (5%) and cardioplegia (95%)  NA  157 ± 53  NA    Bolotin et al. [21]  MT  NA  NA  LVEF (46 ± 12%) and mean NYHA (2.7)  FV and JV-FA    Hypothermia (26°C) and VF  NA  160 ± 65  42/58  0  ST  NA  NA  LVEF (55 ± 11%) and mean NYHA (2.6)  NA  NA  NA  NA  157 ± 54  9/91    Kim et al. [22]  MT  MVP (59), MVR (9), MVR + TVP/AVR (9), MVP + TVP (5) and other (18)  12 ± 9  LVEF (61 ± 9%), hypertension (5%) and diabetes mellitus (5%)  FV-FA  Transthoracic clamp  Moderate hypothermia and antegrade cardioplegia  91 ± 27  171 ± 47  77/23  NA  ST  MVP (46), MVP + TVP (8), MVR (8), MVR + TVP/AVR (8), CABG (8), AVR (8) and other (22)  10 ± 9  LVEF (45 ± 16%), hypertension (39%) and diabetes mellitus (23%)  IVC and SVC-AscA  NA  NA  102 ± 57  210 ± 103  100/0    Hiraoka et al. [23]  MT  MVP (50), redo MVR (10), AVR (10) and CABG (30)  NA  LVEF (47 ± 19%), mean NYHA (1.3 ± 0.5) and EuroSCORE (4.8 ± 2.0)  FV and JV-FA/AA    Hypothermia (27–30°C) and VF  90 ± 7 (VF-time)  145 ± 25  0/100  NA  ST  MVP (33), MVR (19), redo MVR (4), AVR (11), CABG (4), Bentall (4) and other (25)  NA  LVEF (64 ± 9), mean NYHA (1.3 ± 0.7) and EuroSCORE (3.8 + 2.4)  FV and SVC-FA  Aortic cross-clamp  Antegrade cardioplegia  84 ± 19  135 ± 28  0/100    Vallabhajosyula et al. [24]  MT  MVP (78), MVR (19) and MVP + MVR (3)  NA  LVEF (54 ± 12%), NYHA ≥2 (63%), hypertension (61%) and diabetes mellitus (13%)  FV and SVC-FA (97%)/AA (3%)  Endoclamp (81%) and Chitwood clamp (7%)  Hypothermia, antegrade cardioplegia (88%) and VF (12%)  104 ± 38  153 ± 42  NA  1  ST  MVP (41), MVR (53) and MVP + MVR (6)  NA  LVEF (54 ± 17%), NYHA ≥2 (69%), hypertension (55%) and diabetes mellitus (17%)  Central aortic cannulation (95%) and FA (5%)  Aortic cross-clamp (100%)  Antegrade and retrograde cardioplegia  130 ± 71  172 ± 83  NA    Losenno et al. [25]  MT  Isolated MVR/MVP (18), MVR/MVP ± other valve ± CABG (13), CABG (50), AVR/repair ± aortic ± CABG (18) and other (18)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (28%)  FV and JV-FA/AA    Hypothermia (28–30°C), VF  123 ± 37 (VF-time)  201 ± 63  20/80  0    ST  Isolated MVR/MVP (62), MVR/MVP ± other valve ± CABG (17), CABG (11), AVR/repair ± aortic ± CABG (5) and other (10)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (15%)  IVC and SVC-AscA  Aortic cross-clamp  Combination of antegrade and retrograde blood cardioplegia (100%)  105 ± 46  180 ± 75  9/91    Study  Group  Previous surgery (%)  Mean time to redo surgery (years)  Preoperative patient characteristics  Cannulation site  Clamping technique  Myocardial protection  Clamping time (min)  CPB time (min)  MVP/ MVR (%)  Conversion to sternotomy (number of patients)  Burfeind et al. [20]  MT  MVS (60)  NA  LVEF (45 ± 9%) and mean NYHA (3.4)  FV-FA/AscA  Endoclamp (45%)  VF (55%) and additional percutaneous retrograde cardioplegia (33%)  NA  208 ± 76  NA  NA  ST  MVS (83)  NA  LVEF (54 ± 13%) and mean NYHA (3.5)  IVC and SVC-AscA  NA  VF (5%) and cardioplegia (95%)  NA  157 ± 53  NA    Bolotin et al. [21]  MT  NA  NA  LVEF (46 ± 12%) and mean NYHA (2.7)  FV and JV-FA    Hypothermia (26°C) and VF  NA  160 ± 65  42/58  0  ST  NA  NA  LVEF (55 ± 11%) and mean NYHA (2.6)  NA  NA  NA  NA  157 ± 54  9/91    Kim et al. [22]  MT  MVP (59), MVR (9), MVR + TVP/AVR (9), MVP + TVP (5) and other (18)  12 ± 9  LVEF (61 ± 9%), hypertension (5%) and diabetes mellitus (5%)  FV-FA  Transthoracic clamp  Moderate hypothermia and antegrade cardioplegia  91 ± 27  171 ± 47  77/23  NA  ST  MVP (46), MVP + TVP (8), MVR (8), MVR + TVP/AVR (8), CABG (8), AVR (8) and other (22)  10 ± 9  LVEF (45 ± 16%), hypertension (39%) and diabetes mellitus (23%)  IVC and SVC-AscA  NA  NA  102 ± 57  210 ± 103  100/0    Hiraoka et al. [23]  MT  MVP (50), redo MVR (10), AVR (10) and CABG (30)  NA  LVEF (47 ± 19%), mean NYHA (1.3 ± 0.5) and EuroSCORE (4.8 ± 2.0)  FV and JV-FA/AA    Hypothermia (27–30°C) and VF  90 ± 7 (VF-time)  145 ± 25  0/100  NA  ST  MVP (33), MVR (19), redo MVR (4), AVR (11), CABG (4), Bentall (4) and other (25)  NA  LVEF (64 ± 9), mean NYHA (1.3 ± 0.7) and EuroSCORE (3.8 + 2.4)  FV and SVC-FA  Aortic cross-clamp  Antegrade cardioplegia  84 ± 19  135 ± 28  0/100    Vallabhajosyula et al. [24]  MT  MVP (78), MVR (19) and MVP + MVR (3)  NA  LVEF (54 ± 12%), NYHA ≥2 (63%), hypertension (61%) and diabetes mellitus (13%)  FV and SVC-FA (97%)/AA (3%)  Endoclamp (81%) and Chitwood clamp (7%)  Hypothermia, antegrade cardioplegia (88%) and VF (12%)  104 ± 38  153 ± 42  NA  1  ST  MVP (41), MVR (53) and MVP + MVR (6)  NA  LVEF (54 ± 17%), NYHA ≥2 (69%), hypertension (55%) and diabetes mellitus (17%)  Central aortic cannulation (95%) and FA (5%)  Aortic cross-clamp (100%)  Antegrade and retrograde cardioplegia  130 ± 71  172 ± 83  NA    Losenno et al. [25]  MT  Isolated MVR/MVP (18), MVR/MVP ± other valve ± CABG (13), CABG (50), AVR/repair ± aortic ± CABG (18) and other (18)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (28%)  FV and JV-FA/AA    Hypothermia (28–30°C), VF  123 ± 37 (VF-time)  201 ± 63  20/80  0    ST  Isolated MVR/MVP (62), MVR/MVP ± other valve ± CABG (17), CABG (11), AVR/repair ± aortic ± CABG (5) and other (10)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (15%)  IVC and SVC-AscA  Aortic cross-clamp  Combination of antegrade and retrograde blood cardioplegia (100%)  105 ± 46  180 ± 75  9/91    Data are expressed as absolute number, percentages or mean ± SD. a Total >100%; some patients received >1 previous surgery. AA: axillary artery; AscA: ascending aorta; AVR: aortic valve replacement; CABG: coronary artery bypass graft; CPB: cardiopulmonary bypass; FA: femoral artery; FV: femoral vein; IVC: inferior vena cava; JV: jugular vein; LVEF: left ventricular ejection fraction; MT: minithoracotomy; MVP: mitral valve plasty; MVR: mitral valve replacement; MVS: mitral valve surgery; NA: not available; NYHA: New York Heart Association; SD: standard deviation; ST: median sternotomy; STS: Society of Thoracic Surgeons; SVC: superior vena cava; TVP: tricuspid valve plasty; VF: ventricular fibrillation. Intervention Patients underwent surgery through a MT, initiated through a right anterolateral incision in the 4th or 5th intercostal space or a median ST approach, both aimed at repairing or replacing the diseased mitral valve. In some subjects, concomitant surgery was performed. Bolotin et al. [21] and Kim et al. [22] excluded participants who received such concomitant procedures in general whereas others only excluded those receiving any concomitant procedure other than tricuspid repair [20, 23]. Two reports excluded those who received concomitant procedures that were not amenable to a minimally invasive approach [24, 25]. Mean time to redo surgery was reported only by Kim et al. [22], whereas cannulation site, clamping technique, myocardial protection method, repair rate in redo surgeries (MVP/MVR), conversion of redo MT-MVS to ST-MVS, CPB and clamping time were described by at least 3 studies each (Table 2). Outcomes All studies reported the number of deaths, either as 30-day mortality rate, in-hospital deaths or deaths of early postoperative complications. In addition, all secondary outcome measures investigated in this review were assessed at least by 1 study each. Other outcomes investigated by different researchers, but not of interest for this review, were time in the intensive care unit, arrhythmias, chest tube output and several others [20–25]. Synthesis of results See Table 1 for all primary and secondary outcome measures. Death, stroke, reoperation for bleeding and wound infection were described as the number of events and as percentage of the group total, whereas LOHS was denoted as mean and SD. Wound infection and RBC transfusions were excluded from the meta-analysis because these outcomes were only reported in 2 or fewer studies each, and forest plots yield little value and may be seriously biased. Wound infections were reported by Vallabhajosyula et al. [24] and Losenno et al. [25] and occurred in 0.5% (1) and 3.3% (3) of patients who received ST-MVS (220 and 92), respectively. No wound infections were reported for either MT group. RBC transfusion data were reported by Losenno et al. [25], encompassing 132 patients. Overall, 63% (25) and 79% (73) of patients required such blood products in the MT (40) and ST (94) group, respectively. Intraoperative conversion from redo MT-MVS to ST-MVS was described by Bolotin et al. [21], Vallabhajosyula et al. [24] and Losenno et al. [25] (Table 2) and occurred in 0.7% (1) of patients. In addition, data for CPB times were available for all included studies. The mean CPB time was 177 ± 64 and 169 ± 74 min for MT- and ST-MVS, respectively (Table 2). Figures 2–5 present results of each meta-analysis performed, including measures of heterogeneity and 95% CIs. Data comparing deaths following MT- and ST-MVS were available for all 6 studies and were included for quantitative synthesis [20–25]. A total of 2.5% (6) and 8.9% (48) of patients who underwent an MT and an ST mitral valve reoperation (237 and 540; total 777) died, respectively. These percentages reveal a trend towards a reduced number of deaths of patients who had MT-MVS. Analysis confirmed this trend and showed (Fig. 2) that the MT approach was associated with a significantly reduced number of deaths compared to ST-MVS (OR 0.41, 95% CI 0.18–0.96; P = 0.04). In addition, evidence for heterogeneity was absent (I2 = 0%, P = 0.62). Quantitative synthesis for stroke incidence was based on 4 of the analysed studies [20, 23–25]. Overall, stroke occurred in 5.1% (9) and 3.4% (17) of patients who underwent MT- and ST-MVS (177 and 494, total = 671), respectively. Moreover, analysis revealed (Fig. 3) that the MT approach was not significantly associated with a higher occurrence of stroke (OR 1.51, 95% CI 0.65–3.54; P = 0.34) in the absence of heterogeneity (I2 = 0%, P = 0.86). Figure 2: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome death. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 2: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome death. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 3: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome stroke. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 3: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome stroke. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 4: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome reoperation for bleeding. *P = 0.0488. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 4: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome reoperation for bleeding. *P = 0.0488. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 5: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome length of hospital stay. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; SD: standard deviation; ST-MVS: median sternotomy mitral valve surgery. Figure 5: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome length of hospital stay. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; SD: standard deviation; ST-MVS: median sternotomy mitral valve surgery. Reoperation for bleeding data were available for 5 observational studies, encompassing 706 patients [20, 22–25]. These subjects were subdivided into an MT and an ST group (199 and 507, respectively) in which reoperation for bleeding was required in 1.0% (2) and 4.1% (21) of cases, respectively. Initially (Fig. 4), MT-MVS was not significantly associated with a reduction in reoperations due to postoperative bleeding (OR 0.32, 95% CI 0.10–0.99; P = 0.05). However, Review Manager (RevMan v5.3) depicts rounded P-values in its forest plots, so a 4-decimal, comma-separated file was created and exported. This file revealed a P-value of 0.0488 by which MT-MVS was deemed to be significantly associated with a reduced need for reoperation for bleeding. No significant heterogeneity was detected within this comparison (I2 = 0%, P = 0.96). Data regarding LOHS were reported in 5 studies, which included 562 subjects [21–25]. LOHS data were presented on an identical scale among all studies so the difference in means was used. Mean LOHS for MT- and ST-MVS (177 and 385) was, respectively, 10.5 ± 11.0 days and 14.0 ± 11.7 days. A meta-analysis (Fig. 5) demonstrated that MT-MVS was significantly related with a lower LOHS compared to patients receiving MVS through a median ST (MD −3.81, 95% CI −5.53 to −2.08; P < 0.0001). There was no evidence of heterogeneity (I2 = 0%, P = 1.00). DISCUSSION Summary of evidence This systematic review examined whether the MT approach reduces the number of deaths compared to a conventional median ST among patients who received reoperative MVS and underwent prior cardiac surgery through a median ST. Stroke, reoperation for bleeding, LOHS, wound infection and RBC transfusions were added to this comparison as secondary outcome measures. Six non-randomized observational studies that met the eligibility criteria were included for review. The total number of patients was 777, with 237 in the MT group and 540 in the conventional ST group. Forest plots (Figs 2–5) revealed no heterogeneity for all outcome measures subjected to quantitative synthesis. Subsequently, homogeneity across studies was assumed. Detection of publication bias by means of a funnel plot was not performed due to the modest number of studies included for analysis. Quantitative analysis of deaths across studies revealed a homogeneous trend (Fig. 2). MT-MVS was significantly associated with a reduced mortality rate compared to ST-MVS. In addition, even though 3 definitions of mortality were considered as 1 for analysis, no heterogeneity was detected. In this review, the authors assumed death to be a direct consequence of the operative approach. However, all included reports were non-randomized retrospective observational studies, of which 5 [20–24] did not control for confounding. In these studies, baseline patient characteristics may have acted as potential confounders for death and influenced the intervention received at baseline. Apart from potential confounding, the experience of the surgeon may also have affected mortality rates. As reported by Holzhey et al. [26], adverse events, including death, are highly dependent on the surgeon’s experience in minimally invasive MVS. Furthermore, the authors described the influence of a learning curve, which comprises 75–125 procedures. Therefore, reported mortality rates with MT were assumed to be an overestimation compared to those from high volume centres, because all included studies [20–25] reported fewer than 75 minithoracotomies over a period exceeding 1 year at their respective institutions. MT-MVS was associated with a trend (in 3 of 4 studies) towards an increased, although not significant, risk of stroke (Fig. 3). Nevertheless, such embolic events remain a substantial concern when considering reoperative surgery through a MT. The trend in stroke risk may be explained by the observed differences in method for CPB across intervention groups. Retrograde perfusion was most frequently used across all 4 MT groups who reported stroke data [20, 23–25] compared to the ST groups, where 3 of the 4 studies mainly performed antegrade perfusion (AP) via central cannulation of the ascending aorta [20, 24, 25] (Table 2). In general, antegrade ascending aorta perfusion is not used during MT-MVS because the addition of the arterial cannula through the small 3- to 10-cm incision may impede or even prohibit adequate exposure compared to femoral artery exposure, which is easily achieved. Nevertheless, Murzi et al. [27] revealed that retrograde perfusion (compared to AP) was an independent risk factor for stroke (OR 4.28, P = 0.02) in patients who underwent MVS through a MT. Subsequently, AP via the axillary or subclavian artery was preferred over femoral cannulation in patients with peripheral vascular disease for MT-MVS. Axillary or subclavian artery cannulation for MT-MVS was only performed in a minority of cases in 3 studies. Vallabhajosyula et al. [24] performed this type of perfusion in 3% of cases; in contrast, Losenno et al. [23] and Hiraoka et al. [25] only performed axillary AP, depending on individual patient risk factors for atheroembolism and for elderly patients, respectively. Another concern that may contribute to the trend towards increased risk of stroke is inadequate deairing of cardiac cavities before closure. This trend may also be due to the restricted access in minimally invasive MVS, according to Botta et al. [14]. Subsequently, surgeons should utilize CO2 insufflation [28, 29] and venting [30, 31] procedures to mitigate this intracavity air in order to prevent embolic events. In Fig. 5, a homogeneous trend towards reduced LOHS for MT-MVS is seen. This trend was also demonstrated by meta-analysis; reoperation through a MT was significantly associated with a lower LOHS compared to patients who had ST-MVS, in the absence of any heterogeneity. The mean LOHS was 3.81 days shorter for MT-MVS. This more rapid patient recovery may be attributed to the minimally invasive nature of a MT approach, which yields less tissue trauma. This faster recovery was also described by Iribarne et al. [32, 33] for first-time operations. In addition, reoperation due to postoperative bleeding was also expected to be lower as a result of reduced wound surface and less tissue trauma with MT-MVS. Moreover, because MT-MVS utilized retrograde perfusion, no aortic and right atrial cannulations were required, which translated into fewer surgical seams and a reduced risk of postoperative bleeding. This observation was confirmed by analysis (Fig. 4) and consistently translates into diminished blood loss and need for transfusion. However, because transfusion outcomes were reported only in 1 study, quantitative synthesis was not performed. The same applies to wound infections, which were reported in 2 studies. Nevertheless, zero wound infections were observed for MT-MVS compared to 4 for ST-MVS [24, 25]. Among the included studies, only 2 reported use of preoperative computed tomography (CT) scans to address potential procedural difficulties or contraindications. Vallabhajosyula et al. [24] performed a CT scan out of concern for significant disease, whereas Hiraoka et al. [23] even performed an additional magnetic resonance imaging scan systematically. The added value of this preoperative imaging was previously described elsewhere by Heuts et al. [34] and may potentially prevent procedural complications. Apart from the proposed MT approach for redo MVS, several other techniques are currently evolving to circumvent the substantial risks associated with redo ST, such as mitral transcatheter valve-in-valve (MTVIV) and valve-in-ring implantations. Both may be particularly useful for redo surgery patients with high risk or multiple comorbidities after failed mitral bioprosthesis or failed mitral valve repair, respectively. Implantation of these transcatheter, balloon-expandable valves may be performed via a transseptal or transapical approach, whereby the latter is performed through a left anterior MT in the 5th or 6th intercostal space. Despite the fact that a transseptal approach may be less invasive, it is believed to be more technically challenging, by which MTVIV is mainly performed transapically. Nevertheless, the safety and feasibility of transvenous transseptal MTVIV are currently being evaluated by the prospective clinical MITRAL trial (Mitral ImpLantation of TRAnscatheter vaLves) and will potentially be used more frequently in the future [35]. Overall, these percutaneous MTVIV and valve-in-ring implantations have demonstrated excellent haemodynamic performance with low perivalvular regurgitation and transvalvular gradient in feasibility studies [36–39]. In the current European Society of Cardiology/European Association for Cardio-Thoracic Surgery 2017 guidelines for the management of valvular heart disease, these approaches were considered to be a reasonable alternative to redo operations for high-risk patients [40]. In summary, MT-MVS is a safe alternative for ST-MVS, with reduced mortality rates, reduced LOHS, lower incidence of reoperation for bleeding and comparable risk of stroke. Nevertheless, the trend towards increased risk of stroke may remain a substantial concern when considering reoperative MVS through a MT. Limitations Several limitations in this systematic review and meta-analysis should be addressed. The meta-analysis combined data across studies to estimate deaths and secondary outcomes. However, the main limitation of this analysis was the heterogeneity of risk profiles and selection criteria between centres. In addition, 3 different definitions of death were used, which could have led to bias. This systematic review only included retrospective observational studies because no randomized controlled trials have yet been published on this topic. Such observational studies are, however, more prone to confounding. In addition, the number of reports included was small, primarily because of the scarcity of comparative studies. The number of subjects within each MT group was also small and ranged from 10 to 67 compared to relatively larger ST groups, potentially limiting the level of evidence. CONCLUSIONS This meta-analysis is the first performed to test the differences between reoperative minimally invasive MVS through a right MT versus conventional median ST after prior cardiac surgery. The existing literature provided limited data but demonstrated significant differences with regards to mortality, LOHS and reoperation for bleeding, all in favour of MT-MVS. These benefits were evident despite the comparable risk of stroke. Moreover, because the current evidence is of insufficient quality, this review could function as a base for more methodologically rigorous randomized controlled trials. Conflict of interest: none declared. REFERENCES 1 Rankin JS, Hammill BG, Ferguson TB, Glower DD, O'Brien SM, DeLong ER. Determinants of operative mortality in valvular heart surgery. J Thorac Cardiovasc Surg  2006; 131: 547– 57. Google Scholar CrossRef Search ADS PubMed  2 Launcelott S, Ouzounian M, Buth KJ, Légaré J-F. Predicting in-hospital mortality after redo cardiac operations: development of a preoperative scorecard. Ann Thorac Surg  2012; 94: 778– 84. Google Scholar CrossRef Search ADS PubMed  3 Park CB, Suri RM, Burkhart HM, Greason KL, Dearani JA, Schaff HV et al.   Identifying patients at particular risk of injury during repeat sternotomy: analysis of 2555 cardiac reoperations. J Thorac Cardiovasc Surg  2010; 140: 1028– 35. Google Scholar CrossRef Search ADS PubMed  4 Gammie JS, Sheng S, Griffith BP, Peterson ED, Rankin JS, O'Brien SM et al.   Trends in mitral valve surgery in the United States: results from the Society of Thoracic Surgeons Adult Cardiac Database. Ann Thorac Surg  2009; 87: 1431– 9. Google Scholar CrossRef Search ADS PubMed  5 Seeburger J, Borger MA, Falk V, Passage J, Walther T, Doll N et al.   Minimally invasive mitral valve surgery after previous sternotomy: experience in 181 patients. Ann Thorac Surg  2009; 87: 709– 14. Google Scholar CrossRef Search ADS PubMed  6 Botta L, Cannata A, Fratto P, Bruschi G, Trunfio S, Maneggia C et al.   The role of the minimally invasive beating heart technique in reoperative valve surgery. J Card Surg  2012; 27: 24– 8. Google Scholar CrossRef Search ADS PubMed  7 Morales D, Williams E, John R. Is resternotomy in cardiac surgery still a problem? Interact CardioVasc Thorac Surg  2010; 11: 277– 86. Google Scholar CrossRef Search ADS PubMed  8 Roselli EE, Pettersson GB, Blackstone EH, Brizzio ME, Houghtaling PL, Hauck R et al.   Adverse events during reoperative cardiac surgery: frequency, characterization, and rescue. J Thorac Cardiovasc Surg  2008; 135: 316– 23.e6. Google Scholar CrossRef Search ADS PubMed  9 Ellman PI, Smith RL, Girotti ME, Thompson PW, Peeler BB, Kern JA et al.   Cardiac injury during resternotomy does not affect perioperative mortality. J Am Coll Surg  2008; 206: 993– 7. Google Scholar CrossRef Search ADS PubMed  10 Grossi EA, Galloway AC, LaPietra A, Ribakove GH, Ursomanno P, Delianides J et al.   Minimally invasive mitral valve surgery: a 6-year experience with 714 patients. Ann Thorac Surg  2002; 74: 660– 4. Google Scholar CrossRef Search ADS PubMed  11 Glower DD. Surgical approaches to mitral regurgitation. J Am Coll Cardiol  2012; 60: 1315– 22. Google Scholar CrossRef Search ADS PubMed  12 Vleissis AA, Bolling SF. Mini-reoperative mitral valve surgery. J Card Surg  1998; 13: 468– 70. Google Scholar CrossRef Search ADS PubMed  13 Schmitto JD, Mokashi SA, Cohn LH. Minimally-invasive valve surgery. J Am Coll Cardiol  2010; 56: 455– 62. Google Scholar CrossRef Search ADS PubMed  14 Botta L, Cannata A, Bruschi G, Fratto P, Taglieri C, Russo CF et al.   Minimally invasive approach for redo mitral valve surgery. J Thorac Dis  2013; 5(Suppl 6): S686. Google Scholar PubMed  15 Murzi M, Solinas M, Glauber M. Is a minimally invasive approach for re-operative mitral valve surgery superior to standard resternotomy? Interact CardioVasc Thorac Surg  2009; 9: 327– 32. Google Scholar CrossRef Search ADS PubMed  16 Modi P, Hassan A, Chitwood WR. Minimally invasive mitral valve surgery: a systematic review and meta-analysis. Eur J Cardiothorac Surg  2008; 34: 943– 52. Google Scholar CrossRef Search ADS PubMed  17 Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med  2009; 6: e1000097. Google Scholar CrossRef Search ADS PubMed  18 Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions . Hoboken (New Jersey), USA: John Wiley & Sons; 2011. 19 Altman DG, Bland JM. Standard deviations and standard errors. BMJ  2005; 331: 903. Google Scholar CrossRef Search ADS PubMed  20 Burfeind WR, Glower DD, Davis RD, Landolfo KP, Lowe JE, Wolfe WG. Mitral surgery after prior cardiac operation: port-access versus sternotomy or thoracotomy. Ann Thorac Surg  2002; 74: S1323– 5. Google Scholar CrossRef Search ADS PubMed  21 Bolotin G, Kypson AP, Reade CC, Chu VF, Freund WLJr, Nifong LW. Should a video-assisted mini-thoracotomy be the approach of choice for reoperative mitral valve surgery? J Heart Valve Dis  2004; 13: 155– 8; discussion 8. Google Scholar PubMed  22 Kim DC, Chee HK, Song MG, Shin JK, Kim JS, Lee S et al.   Comparative analysis of thoracotomy and sternotomy approaches in cardiac reoperation. Korean J Thorac Cardiovasc Surg  2012; 45: 225. Google Scholar CrossRef Search ADS PubMed  23 Hiraoka A, Kuinose M, Totsugawa T, Chikazawa G, Yoshitaka H. Mitral valve reoperation under ventricular fibrillation through right mini-thoracotomy using three-dimensional videoscope. J Cardiothorac Surg  2013; 8: 81. Google Scholar CrossRef Search ADS PubMed  24 Vallabhajosyula P, Wallen T, Pulsipher A, Pitkin E, Solometo LP, Musthaq S et al.   Minimally invasive port access approach for reoperations on the mitral valve. Ann Thorac Surg  2015; 100: 68– 73. Google Scholar CrossRef Search ADS PubMed  25 Losenno KL, Jones PM, Valdis M, Fox SA, Kiaii B, Chu MW. Higher-risk mitral valve operations after previous sternotomy: endoscopic, minimally invasive approach improves patient outcomes. Can J Surg  2016; 59: 399– 406. Google Scholar CrossRef Search ADS PubMed  26 Holzhey DM, Seeburger J, Misfeld M, Borger MA, Mohr FW. Learning minimally-invasive mitral valve surgery: a cumulative sum sequential probability analysis of 3895 operations from a single high volume center. Circulation  2013; 128: 483– 91. Google Scholar CrossRef Search ADS PubMed  27 Murzi M, Cerillo AG, Miceli A, Bevilacqua S, Kallushi E, Farneti P et al.   Antegrade and retrograde arterial perfusion strategy in minimally invasive mitral-valve surgery: a propensity score analysis on 1280 patients. Eur J Cardiothorac Surg  2013; 43: e167– 72. Google Scholar CrossRef Search ADS PubMed  28 Martens S, Neumann K, Sodemann C, Deschka H, Wimmer-Greinecker G, Moritz A. Carbon dioxide field flooding reduces neurologic impairment after open heart surgery. Ann Thorac Surg  2008; 85: 543– 7. Google Scholar CrossRef Search ADS PubMed  29 Svenarud P, Persson M, Van Der Linden J. Effect of CO2 insufflation on the number and behavior of air microemboli in open-heart surgery. Circulation  2004; 109: 1127– 32. Google Scholar CrossRef Search ADS PubMed  30 Reser D, Walser R, van Hemelrijk M, Holubec T, Weber A, Plass A et al.   Long-term outcomes after minimally invasive aortic valve surgery through right anterior minithoracotomy. Thorac Cardiovasc Surg  2017; 65: 191– 7. Google Scholar CrossRef Search ADS PubMed  31 Sharony R, Grossi EA, Saunders PC, Schwartz CF, Ursomanno P, Ribakove GH et al.   Minimally invasive reoperative isolated valve surgery: early and mid-term results. J Card Surgery  2006; 21: 240– 4. Google Scholar CrossRef Search ADS   32 Iribarne A, Easterwood R, Russo MJ, Chan EY, Smith CR, Argenziano M. Comparative effectiveness of minimally invasive versus traditional sternotomy mitral valve surgery in elderly patients. J Thorac Cardiovasc Surg  2012; 143(4 Suppl): S86– 90. Google Scholar CrossRef Search ADS   33 Iribarne A, Russo MJ, Easterwood R, Hong KN, Yang J, Cheema FH et al.   Minimally invasive versus sternotomy approach for mitral valve surgery: a propensity analysis. Ann Thorac Surg  2010; 90: 1471– 8. Google Scholar CrossRef Search ADS PubMed  34 Heuts S, Maessen JG, Sardari Nia P. Preoperative planning of left-sided valve surgery with 3D computed tomography reconstruction models: sternotomy or a minimally invasive approach? Interact CardioVasc Thorac Surg  2016; 22: 587– 93. Google Scholar CrossRef Search ADS PubMed  35 Guerrero M, Salinger M, Pursnani A, Pearson P, Lampert M, Levisay J et al.   Transseptal transcatheter mitral valve‐in‐valve: a step by step guide from preprocedural planning to postprocedural care. Catheter Cardiovasc Interv  2017; 00: 1– 12. 36 Cheung A, Al-Lawati A. Transcatheter mitral valve-in-valve implantation: current experience and review of literature . Curr Opin Cardiol  2013; 28: 181– 6. Google Scholar CrossRef Search ADS PubMed  37 Cheung A, Webb JG, Barbanti M, Freeman M, Binder RK, Thompson C et al.   5-year experience with transcatheter transapical mitral valve-in-valve implantation for bioprosthetic valve dysfunction. J Am Coll Cardiol  2013; 61: 1759– 66. Google Scholar CrossRef Search ADS PubMed  38 Descoutures F, Himbert D, Maisano F, Casselman F, de Weger A, Bodea O et al.   Transcatheter valve-in-ring implantation after failure of surgical mitral repair. Eur J Cardiothorac Surg  2013; 44: e8– 15. Google Scholar CrossRef Search ADS PubMed  39 Cullen MW, Cabalka AK, Alli OO, Pislaru SV, Sorajja P, Nkomo VT et al.   Transvenous, antegrade melody valve-in-valve implantation for bioprosthetic mitral and tricuspid valve dysfunction: a case series in children and adults. JACC Cardiovasc Interv  2013; 6: 598– 605. Google Scholar CrossRef Search ADS PubMed  40 Falk V, Baumgartner H, Bax JJ, De Bonis M, Hamm C, Holm PJ et al.  ; ESC Scientific Document Group. 2017 ESC/EACTS guidelines for the management of valvular heart disease. Eur J Cardiothorac Surg  2017; 52: 616– 64. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Cardio-Thoracic Surgery Oxford University Press

Right minithoracotomy versus median sternotomy for reoperative mitral valve surgery: a systematic review and meta-analysis of observational studies

Loading next page...
 
/lp/ou_press/right-minithoracotomy-versus-median-sternotomy-for-reoperative-mitral-ow6lYf6k62
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ISSN
1010-7940
eISSN
1873-734X
D.O.I.
10.1093/ejcts/ezy173
Publisher site
See Article on Publisher Site

Abstract

Abstract Reoperative mitral valve surgery (MVS) through a median sternotomy (ST-MVS) can be particularly challenging due to dense adhesions and is known to carry a substantial risk of injuries to vascular structures. These injuries occur in 7–9% of cases and are associated with increased mortality rates. A valid alternative that could avoid the risks associated with redo ST-MVS is the right anterolateral minithoracotomy (MT-MVS) approach. The aim of this study was to quantify the effects of MT-MVS compared with those of ST-MVS on morbidity and mortality among patients who underwent prior cardiac surgery through a sternotomy. The MEDLINE and EMBASE databases were searched through 1 November 2017. Data regarding mortality, stroke, reoperation for bleeding and length of hospital stay were extracted and submitted to a meta-analysis using random effects modelling and the I2-test for heterogeneity. Six retrospective observational studies were included, enrolling a total of 777 patients. In a pooled analysis, MT-MVS demonstrated reduced mortality rates compared to a standard sternotomy [odds ratio (OR) 0.41, 95% confidence interval (CI) 0.18–0.96; P = 0.04]. MT-MVS was, moreover, associated with reduced length of hospital stay [difference between the means was −3.81, 95% CI −5.53 to −2.08; P < 0.0001) and reoperation for bleeding (OR 0.32, 95% CI 0.10–0.99; P = 0.0488). The incidence of stroke was similar (OR 1.51, 95% CI 0.65–3.54; P = 0.34), all in the absence of heterogeneity. In conclusion, reoperative minimally invasive MVS through a minithoracotomy is a safe alternative to standard sternotomy, with reduced mortality rates, length of hospital stay and reoperations for bleeding and a comparable risk of stroke. However, because the existing literature provided limited, low-quality evidence, more methodologically rigorous randomized controlled trials are needed. Minimally invasive surgery , Mitral valve surgery , Right minithoracotomy , Reoperation , Median sternotomy INTRODUCTION Redo cardiac surgery has been associated with increased mortality rates compared to primary surgery [1, 2]. Cardiac redo procedures are traditionally performed through a repeat median sternotomy (ST). For the past decade, reoperative mitral valve surgery (MVS) has become more common, representing over 10% of all mitral valve procedures in the USA [3, 4]. However, redo MVS performed through an ST (ST-MVS) can be particularly technically challenging and is known to carry a substantial risk of injuries to patent coronary artery bypass grafts and vascular structures that lie directly substernally and can adhere to the sternum. Resternotomy may furthermore be demanding in patients with (healed) mediastinitis, prior thoracic radiotherapy and dense adhesions or other complications from prior surgery [5–7]. These injuries to cardiac structures occur in 7 to 9% of resternotomies [3, 8, 9] and are reported to be an independent risk factor for in-hospital death [3]. A valid alternative to repeated conventional ST-MVS would be a minimally invasive approach through a right anterolateral minithoracotomy (MT) [5, 10]. An incision of <10 cm is made in the 4th or 5th intercostal space, the goal being to minimize surgical trauma compared to that of a full ST or thoracotomy (20 cm) [11, 12]. MT-MVS can be performed either under direct or video-assisted vision, with the use of long-shafted instruments in both situations. Primary MT-MVS is, besides being associated with less surgical trauma, believed to result in diminished pain, blood loss and need for transfusions, which translates into reduced length of hospital stay (LOHS) [13]. In addition, with MT-MVS, one could avoid the risks associated with resternotomy. Despite these advantages, no general consensus exists on the approach of choice for redo MVS. This consensus should ideally arise from well-designed randomized controlled trials and comprehensive literature reviews that compare redo MT- and ST-MVS among patients with a prior ST. However, to date such data are only available from non-randomized studies, in the form of 2 best evidence topics [14, 15] and 1 small narrative subreview [16]. Therefore, an overview and analysis of the available comparative data, which may aid in determining the optimal approach for redo MVS, are needed. The aim of this report was to review all published observational studies that compare redo MVS through a MT and a conventional median ST approach among patients with prior cardiac surgery through a ST, with mortality as the primary outcome measure. Secondary outcome measures include stroke, reoperation for bleeding, LOHS, wound infection and red blood cell (RBC) transfusions. In order to draw more useful and robust conclusions regarding these outcome metrics, data from individual studies were collected and analysed using meta-analytical techniques. MATERIALS AND METHODS Methods of the analysis, outcome measures and inclusion criteria were specified in advance and documented in a protocol. The review and meta-analysis were conducted using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement [17]. Risk of bias in individual studies and across studies was not assessed, because retrospective observational studies are generally known for possessing a certain level of bias. Eligibility criteria Types of participants Participants aged 18 years and above who presented with mitral valve disease requiring surgery and who underwent at least 1 previous cardiac surgical procedure through a median ST were considered for inclusion. No exclusions were made based on the indication of primary ST (e.g. coronary artery bypass graft, aortic valve replacement). Types of interventions Studies that compared the MT (also denoted as ‘port access’) and the median ST approach for reoperative MVS were considered. Studies in which cardiac procedures were performed concomitantly with the initial mitral valve repair or replacement were also assessed for eligibility. The MT approach was defined as a ≤10 cm right anterolateral incision in the 4th or 5th intercostal space to acquire surgical exposure. No difference was made between video-assisted or direct vision MT; however, studies utilizing robotic telemanipulation such as the Da Vinci robot were excluded because telemanipulation forms a totally different area within minimally invasive MVS. If the report did not define the thoracotomy approach as either mini or full, the authors were contacted for clarification. Redo ST was defined as MVS in which the previous ST incision was re-entered and cardiac exposure was obtained by an oscillating saw to complete the ST. Type of primary outcome measure The primary outcome measure was mortality, reported as in-hospital death, 30-day mortality rate or death as an early postoperative complication. Types of studies Observational studies comparing MT and ST for reoperative MVS after a previous median ST were examined for eligibility. Studies reporting combined data for mitral and aortic valve reoperations were considered only if mitral valve data were presented separately. Studies reporting primary operations were excluded. Literature search Potentially eligible studies were identified by searching the electronic MEDLINE and EMBASE databases through PubMed and Ovid, respectively. No unpublished data were obtained. The search was limited to the English language, human subjects and studies published after 2000, because the latter provide the best evidence for current practice. No publication status restrictions were imposed. The last search was run on 1 November 2017. In addition, a cross-reference and related-articles search was conducted as a check of rigour. The search strategy was first applied to the electronic MEDLINE database and combined the following MeSH and free terms: ‘mitral valve’, ‘mitral valve insufficiency’, ‘mitral valve prolapse’ AND ‘thoracotomy’, ‘minithoracotomy’, ‘port access’ AND ‘ST’ AND ‘reoperation’, ‘reoperative’ and ‘redo’. This search was subsequently adapted for EMBASE. Reports originating from the electronic search were screened for eligibility based on their titles and abstracts. Subsequently, full texts of potentially eligible reports were read and carefully assessed according to the eligibility criteria. Studies meeting these criteria were included for review and, if applicable, for quantitative synthesis (meta-analysis). Study selection was conducted in a non-blinded standardized manner by 2 independent reviewers. Potential inter-reviewer disagreements were resolved by consensus. Data collection For this review, a data extraction sheet was developed and pilot-tested on 3 randomly selected included studies, whereupon the sheet was refined accordingly. Data extraction was performed by 1 review author. The second author validated the correctness of the tabulated data. Potential inter-reviewer disagreements were resolved by consensus. Studies reporting their continuous variables as mean and standard deviation (SD) were extracted without conversion. Variables denoted as median and interquartile range or standard error of the mean were first converted, as described elsewhere [18, 19]. Data were extracted from each included report on (i) general study characteristics: study design, enrolment period, country, setting, in- and exclusion criteria, baseline characteristics and statistical methods; (ii) characteristics of participants: number of patients, mean age (years), gender (male/female), previous surgery and preoperative characteristics such as Society of Thoracic Surgeons (STS) score or EuroSCORE, ejection fraction, New York Heart Association class and several comorbidities; (iii) intervention characteristics: mean time to redo surgery (years), cannulation site, clamping technique, myocardial protection method, concomitant procedures, repair rate in redo surgeries (mitral valve repair versus replacement), conversion to ST (redo MT-MVS only), cardiopulmonary bypass (CPB) and clamping time; (iv) primary outcome measure: death, either reported as 30-day mortality, in-hospital death or early postoperative death; (v) secondary outcome measures: stroke, reoperation for bleeding, LOHS, wound infection and RBC transfusions. If outcome measures were reported in any other way than stated and could not be converted, data were assumed to be not available. In addition, it must be noted that the authors did not discriminate between minor differences in surgical procedures. Statistical analysis Odds ratios (ORs) were used to assess dichotomous outcome measures. The difference in means (MD) was used for continues variables, which were made on the same scale among all studies. If not, the standardized difference was used. Obtained ORs were interpreted as risk ratios. In addition, an OR or MD less than 1 favours MT over ST for MVS. Statistical analyses were performed using Review Manager (RevMan v5.3, Cochrane Collaboration, Oxford, UK). For this analysis, the random effects model with a 95% confidence interval (CI) was used when there was a substantial risk of heterogeneity, originating from the non-randomized nature of the included studies. In addition, because risk profiles and selection criteria differed between centres, the random effects model was favoured. The I2-test for heterogeneity was conducted to assess variability across studies that could not be due to random error alone. High I2-values indicated that the observed variability among studies could not be explained by chance (i.e. a consequence of clinical and/or methodological diversity). Heterogeneity was deemed to be substantial and considerable if I2 >50% and I2 >75%, respectively with P-value <0.10 [18]. No additional analyses were conducted. RESULTS Study selection A total of 6 studies were identified for inclusion in this review and meta-analysis. The MEDLINE and EMBASE database search provided a cumulative number of 250 citations. In addition, 1 citation was obtained via a cross-reference and related-article search. Of these, 84 duplicates were discarded. Another 151 papers were eliminated because their title and abstract clearly did not meet the eligibility criteria. Full texts of the remaining 16 articles were assessed in detail for eligibility. Of these, 10 reports did not meet the criteria as described. Reasons for exclusion were no comparison between MT and median ST (n = 6), results for mitral valve reoperation were not separately reported (n = 3) and the main text for 1 study was in Japanese that could not be translated into English (n = 1). In addition, no unpublished relevant studies were obtained (flow diagram, Fig. 1). Figure 1: View largeDownload slide Study selection procedure shown in a PRISMA flow diagram. Figure 1: View largeDownload slide Study selection procedure shown in a PRISMA flow diagram. Study characteristics Methods All 6 studies finally selected for inclusion were retrospective single-centre observational studies published in English. All studies were conducted in the USA, Canada, Japan or South Korea. The main inclusion criterion encompassed patients who received a MT or median ST for mitral valve disease with at least 1 prior cardiac surgical procedure through a median ST. Overall, studies enrolled patients who were operated on between 1985 and 2011. Participants A total number of 777 participants were included for review. These were subdivided in an MT (n = 237; 31%; mean age 62.8 ± 14.0 years) and an ST (n = 540; 69%; mean age 60.1 ± 14.8 years) group. The number of subjects per individual study ranged from 35 to 287 [20–25] (Table 1). Previous operations as well as preoperative patient characteristics are presented in Table 2. STS scores [25] and EuroSCOREs [23] were both only reported in 1 paper each, whereas others mainly reported ejection fraction, New York Heart Association class and comorbidities such as diabetes mellitus and hypertension. Table 1: Principal data and outcomes of interest from individual studies Study  Country  Study design  Study period  Group (MT/ST)  Participants (total (male/female))  Age (years)  Mortality (events (%))  Stroke (events (%))  LOHS (days)  RFB (events (%))  Wound infection (events (%))  RBC transfusion (patients (%))  Burfeind et al. [20]  USA  Retrospective observational study  1996–2001  MT  60 (33/27)  60 ± 14  0 (0)  4 (7)  NA  1 (2)  NA  NA        1985–1997  ST  155 (42/113)  58 ± 16  21 (14)  7 (5)  NA  8 (5)  NA  NA  Bolotin et al. [21]  USA  Retrospective observational study  January 1996–June 2003  MT  38 (-/-)  68 ± 2  2 (5)  NA  7 ± 6  NA  NA  NA          ST  33 (-/-)  63 ± 2  2 (6)  NA  11 ± 6  NA  NA  NA  Kim et al. [22]  South Korea  Retrospective observational study  September 2007–December 2010  MT  22 (4/18)  46 ± 15  0 (0)  NA  16 ± 6  0 (0)  NA  NA          ST  13 (5/8)  45 ± 16  1 (8)  NA  20 ± 17  2 (15)  NA  NA  Hiraoka et al. [23]  Japan  Retrospective observational study  January 2006–September 2011  MT  10 (5/5)  68 ± 15  0 (0)  1 (10)  18 ± 8  0 (0)  NA  NA          ST  27 (18/9)  63 ± 15  1 (4)  1 (4)  22 ± 13  2 (7)  NA  NA  Vallabhajosyula et al. [24]  USA  Retrospective observational study  1988–2001  MT  67 (32/35)  64 ± 13  2 (3)  2 (3)  11 ± 15  0 (0)  0 (0)  NA        ST  220 (99/121)  61 ± 15  13 (6)  7 (3)  14 ± 12  3 (1)  1 (0.5)  NA  Losenno et al. [25]  Canada  Retrospective observational study  September 2000–August 2014  MT  40 (28/12)  68 ± 14  2 (5)  2 (5)  8 ± 7  1 (3)  0 (0)  25 (63)          ST  92 (38/54)  62 ± 13  10 (11)  2 (2)  12 ± 10  6 (7)  3 (3)  73 (79)  Study  Country  Study design  Study period  Group (MT/ST)  Participants (total (male/female))  Age (years)  Mortality (events (%))  Stroke (events (%))  LOHS (days)  RFB (events (%))  Wound infection (events (%))  RBC transfusion (patients (%))  Burfeind et al. [20]  USA  Retrospective observational study  1996–2001  MT  60 (33/27)  60 ± 14  0 (0)  4 (7)  NA  1 (2)  NA  NA        1985–1997  ST  155 (42/113)  58 ± 16  21 (14)  7 (5)  NA  8 (5)  NA  NA  Bolotin et al. [21]  USA  Retrospective observational study  January 1996–June 2003  MT  38 (-/-)  68 ± 2  2 (5)  NA  7 ± 6  NA  NA  NA          ST  33 (-/-)  63 ± 2  2 (6)  NA  11 ± 6  NA  NA  NA  Kim et al. [22]  South Korea  Retrospective observational study  September 2007–December 2010  MT  22 (4/18)  46 ± 15  0 (0)  NA  16 ± 6  0 (0)  NA  NA          ST  13 (5/8)  45 ± 16  1 (8)  NA  20 ± 17  2 (15)  NA  NA  Hiraoka et al. [23]  Japan  Retrospective observational study  January 2006–September 2011  MT  10 (5/5)  68 ± 15  0 (0)  1 (10)  18 ± 8  0 (0)  NA  NA          ST  27 (18/9)  63 ± 15  1 (4)  1 (4)  22 ± 13  2 (7)  NA  NA  Vallabhajosyula et al. [24]  USA  Retrospective observational study  1988–2001  MT  67 (32/35)  64 ± 13  2 (3)  2 (3)  11 ± 15  0 (0)  0 (0)  NA        ST  220 (99/121)  61 ± 15  13 (6)  7 (3)  14 ± 12  3 (1)  1 (0.5)  NA  Losenno et al. [25]  Canada  Retrospective observational study  September 2000–August 2014  MT  40 (28/12)  68 ± 14  2 (5)  2 (5)  8 ± 7  1 (3)  0 (0)  25 (63)          ST  92 (38/54)  62 ± 13  10 (11)  2 (2)  12 ± 10  6 (7)  3 (3)  73 (79)  Data are expressed as absolute number, percentages or as mean ± SD. LOHS: length of hospital stay; MT: minithoracotomy; NA: not available; RBC: red blood cell; RFB: reoperation for bleeding; SD: standard deviation; ST: median sternotomy; USA: United States of America. Table 1: Principal data and outcomes of interest from individual studies Study  Country  Study design  Study period  Group (MT/ST)  Participants (total (male/female))  Age (years)  Mortality (events (%))  Stroke (events (%))  LOHS (days)  RFB (events (%))  Wound infection (events (%))  RBC transfusion (patients (%))  Burfeind et al. [20]  USA  Retrospective observational study  1996–2001  MT  60 (33/27)  60 ± 14  0 (0)  4 (7)  NA  1 (2)  NA  NA        1985–1997  ST  155 (42/113)  58 ± 16  21 (14)  7 (5)  NA  8 (5)  NA  NA  Bolotin et al. [21]  USA  Retrospective observational study  January 1996–June 2003  MT  38 (-/-)  68 ± 2  2 (5)  NA  7 ± 6  NA  NA  NA          ST  33 (-/-)  63 ± 2  2 (6)  NA  11 ± 6  NA  NA  NA  Kim et al. [22]  South Korea  Retrospective observational study  September 2007–December 2010  MT  22 (4/18)  46 ± 15  0 (0)  NA  16 ± 6  0 (0)  NA  NA          ST  13 (5/8)  45 ± 16  1 (8)  NA  20 ± 17  2 (15)  NA  NA  Hiraoka et al. [23]  Japan  Retrospective observational study  January 2006–September 2011  MT  10 (5/5)  68 ± 15  0 (0)  1 (10)  18 ± 8  0 (0)  NA  NA          ST  27 (18/9)  63 ± 15  1 (4)  1 (4)  22 ± 13  2 (7)  NA  NA  Vallabhajosyula et al. [24]  USA  Retrospective observational study  1988–2001  MT  67 (32/35)  64 ± 13  2 (3)  2 (3)  11 ± 15  0 (0)  0 (0)  NA        ST  220 (99/121)  61 ± 15  13 (6)  7 (3)  14 ± 12  3 (1)  1 (0.5)  NA  Losenno et al. [25]  Canada  Retrospective observational study  September 2000–August 2014  MT  40 (28/12)  68 ± 14  2 (5)  2 (5)  8 ± 7  1 (3)  0 (0)  25 (63)          ST  92 (38/54)  62 ± 13  10 (11)  2 (2)  12 ± 10  6 (7)  3 (3)  73 (79)  Study  Country  Study design  Study period  Group (MT/ST)  Participants (total (male/female))  Age (years)  Mortality (events (%))  Stroke (events (%))  LOHS (days)  RFB (events (%))  Wound infection (events (%))  RBC transfusion (patients (%))  Burfeind et al. [20]  USA  Retrospective observational study  1996–2001  MT  60 (33/27)  60 ± 14  0 (0)  4 (7)  NA  1 (2)  NA  NA        1985–1997  ST  155 (42/113)  58 ± 16  21 (14)  7 (5)  NA  8 (5)  NA  NA  Bolotin et al. [21]  USA  Retrospective observational study  January 1996–June 2003  MT  38 (-/-)  68 ± 2  2 (5)  NA  7 ± 6  NA  NA  NA          ST  33 (-/-)  63 ± 2  2 (6)  NA  11 ± 6  NA  NA  NA  Kim et al. [22]  South Korea  Retrospective observational study  September 2007–December 2010  MT  22 (4/18)  46 ± 15  0 (0)  NA  16 ± 6  0 (0)  NA  NA          ST  13 (5/8)  45 ± 16  1 (8)  NA  20 ± 17  2 (15)  NA  NA  Hiraoka et al. [23]  Japan  Retrospective observational study  January 2006–September 2011  MT  10 (5/5)  68 ± 15  0 (0)  1 (10)  18 ± 8  0 (0)  NA  NA          ST  27 (18/9)  63 ± 15  1 (4)  1 (4)  22 ± 13  2 (7)  NA  NA  Vallabhajosyula et al. [24]  USA  Retrospective observational study  1988–2001  MT  67 (32/35)  64 ± 13  2 (3)  2 (3)  11 ± 15  0 (0)  0 (0)  NA        ST  220 (99/121)  61 ± 15  13 (6)  7 (3)  14 ± 12  3 (1)  1 (0.5)  NA  Losenno et al. [25]  Canada  Retrospective observational study  September 2000–August 2014  MT  40 (28/12)  68 ± 14  2 (5)  2 (5)  8 ± 7  1 (3)  0 (0)  25 (63)          ST  92 (38/54)  62 ± 13  10 (11)  2 (2)  12 ± 10  6 (7)  3 (3)  73 (79)  Data are expressed as absolute number, percentages or as mean ± SD. LOHS: length of hospital stay; MT: minithoracotomy; NA: not available; RBC: red blood cell; RFB: reoperation for bleeding; SD: standard deviation; ST: median sternotomy; USA: United States of America. Table 2: Supplementary data of interest from individual studies Study  Group  Previous surgery (%)  Mean time to redo surgery (years)  Preoperative patient characteristics  Cannulation site  Clamping technique  Myocardial protection  Clamping time (min)  CPB time (min)  MVP/ MVR (%)  Conversion to sternotomy (number of patients)  Burfeind et al. [20]  MT  MVS (60)  NA  LVEF (45 ± 9%) and mean NYHA (3.4)  FV-FA/AscA  Endoclamp (45%)  VF (55%) and additional percutaneous retrograde cardioplegia (33%)  NA  208 ± 76  NA  NA  ST  MVS (83)  NA  LVEF (54 ± 13%) and mean NYHA (3.5)  IVC and SVC-AscA  NA  VF (5%) and cardioplegia (95%)  NA  157 ± 53  NA    Bolotin et al. [21]  MT  NA  NA  LVEF (46 ± 12%) and mean NYHA (2.7)  FV and JV-FA    Hypothermia (26°C) and VF  NA  160 ± 65  42/58  0  ST  NA  NA  LVEF (55 ± 11%) and mean NYHA (2.6)  NA  NA  NA  NA  157 ± 54  9/91    Kim et al. [22]  MT  MVP (59), MVR (9), MVR + TVP/AVR (9), MVP + TVP (5) and other (18)  12 ± 9  LVEF (61 ± 9%), hypertension (5%) and diabetes mellitus (5%)  FV-FA  Transthoracic clamp  Moderate hypothermia and antegrade cardioplegia  91 ± 27  171 ± 47  77/23  NA  ST  MVP (46), MVP + TVP (8), MVR (8), MVR + TVP/AVR (8), CABG (8), AVR (8) and other (22)  10 ± 9  LVEF (45 ± 16%), hypertension (39%) and diabetes mellitus (23%)  IVC and SVC-AscA  NA  NA  102 ± 57  210 ± 103  100/0    Hiraoka et al. [23]  MT  MVP (50), redo MVR (10), AVR (10) and CABG (30)  NA  LVEF (47 ± 19%), mean NYHA (1.3 ± 0.5) and EuroSCORE (4.8 ± 2.0)  FV and JV-FA/AA    Hypothermia (27–30°C) and VF  90 ± 7 (VF-time)  145 ± 25  0/100  NA  ST  MVP (33), MVR (19), redo MVR (4), AVR (11), CABG (4), Bentall (4) and other (25)  NA  LVEF (64 ± 9), mean NYHA (1.3 ± 0.7) and EuroSCORE (3.8 + 2.4)  FV and SVC-FA  Aortic cross-clamp  Antegrade cardioplegia  84 ± 19  135 ± 28  0/100    Vallabhajosyula et al. [24]  MT  MVP (78), MVR (19) and MVP + MVR (3)  NA  LVEF (54 ± 12%), NYHA ≥2 (63%), hypertension (61%) and diabetes mellitus (13%)  FV and SVC-FA (97%)/AA (3%)  Endoclamp (81%) and Chitwood clamp (7%)  Hypothermia, antegrade cardioplegia (88%) and VF (12%)  104 ± 38  153 ± 42  NA  1  ST  MVP (41), MVR (53) and MVP + MVR (6)  NA  LVEF (54 ± 17%), NYHA ≥2 (69%), hypertension (55%) and diabetes mellitus (17%)  Central aortic cannulation (95%) and FA (5%)  Aortic cross-clamp (100%)  Antegrade and retrograde cardioplegia  130 ± 71  172 ± 83  NA    Losenno et al. [25]  MT  Isolated MVR/MVP (18), MVR/MVP ± other valve ± CABG (13), CABG (50), AVR/repair ± aortic ± CABG (18) and other (18)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (28%)  FV and JV-FA/AA    Hypothermia (28–30°C), VF  123 ± 37 (VF-time)  201 ± 63  20/80  0    ST  Isolated MVR/MVP (62), MVR/MVP ± other valve ± CABG (17), CABG (11), AVR/repair ± aortic ± CABG (5) and other (10)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (15%)  IVC and SVC-AscA  Aortic cross-clamp  Combination of antegrade and retrograde blood cardioplegia (100%)  105 ± 46  180 ± 75  9/91    Study  Group  Previous surgery (%)  Mean time to redo surgery (years)  Preoperative patient characteristics  Cannulation site  Clamping technique  Myocardial protection  Clamping time (min)  CPB time (min)  MVP/ MVR (%)  Conversion to sternotomy (number of patients)  Burfeind et al. [20]  MT  MVS (60)  NA  LVEF (45 ± 9%) and mean NYHA (3.4)  FV-FA/AscA  Endoclamp (45%)  VF (55%) and additional percutaneous retrograde cardioplegia (33%)  NA  208 ± 76  NA  NA  ST  MVS (83)  NA  LVEF (54 ± 13%) and mean NYHA (3.5)  IVC and SVC-AscA  NA  VF (5%) and cardioplegia (95%)  NA  157 ± 53  NA    Bolotin et al. [21]  MT  NA  NA  LVEF (46 ± 12%) and mean NYHA (2.7)  FV and JV-FA    Hypothermia (26°C) and VF  NA  160 ± 65  42/58  0  ST  NA  NA  LVEF (55 ± 11%) and mean NYHA (2.6)  NA  NA  NA  NA  157 ± 54  9/91    Kim et al. [22]  MT  MVP (59), MVR (9), MVR + TVP/AVR (9), MVP + TVP (5) and other (18)  12 ± 9  LVEF (61 ± 9%), hypertension (5%) and diabetes mellitus (5%)  FV-FA  Transthoracic clamp  Moderate hypothermia and antegrade cardioplegia  91 ± 27  171 ± 47  77/23  NA  ST  MVP (46), MVP + TVP (8), MVR (8), MVR + TVP/AVR (8), CABG (8), AVR (8) and other (22)  10 ± 9  LVEF (45 ± 16%), hypertension (39%) and diabetes mellitus (23%)  IVC and SVC-AscA  NA  NA  102 ± 57  210 ± 103  100/0    Hiraoka et al. [23]  MT  MVP (50), redo MVR (10), AVR (10) and CABG (30)  NA  LVEF (47 ± 19%), mean NYHA (1.3 ± 0.5) and EuroSCORE (4.8 ± 2.0)  FV and JV-FA/AA    Hypothermia (27–30°C) and VF  90 ± 7 (VF-time)  145 ± 25  0/100  NA  ST  MVP (33), MVR (19), redo MVR (4), AVR (11), CABG (4), Bentall (4) and other (25)  NA  LVEF (64 ± 9), mean NYHA (1.3 ± 0.7) and EuroSCORE (3.8 + 2.4)  FV and SVC-FA  Aortic cross-clamp  Antegrade cardioplegia  84 ± 19  135 ± 28  0/100    Vallabhajosyula et al. [24]  MT  MVP (78), MVR (19) and MVP + MVR (3)  NA  LVEF (54 ± 12%), NYHA ≥2 (63%), hypertension (61%) and diabetes mellitus (13%)  FV and SVC-FA (97%)/AA (3%)  Endoclamp (81%) and Chitwood clamp (7%)  Hypothermia, antegrade cardioplegia (88%) and VF (12%)  104 ± 38  153 ± 42  NA  1  ST  MVP (41), MVR (53) and MVP + MVR (6)  NA  LVEF (54 ± 17%), NYHA ≥2 (69%), hypertension (55%) and diabetes mellitus (17%)  Central aortic cannulation (95%) and FA (5%)  Aortic cross-clamp (100%)  Antegrade and retrograde cardioplegia  130 ± 71  172 ± 83  NA    Losenno et al. [25]  MT  Isolated MVR/MVP (18), MVR/MVP ± other valve ± CABG (13), CABG (50), AVR/repair ± aortic ± CABG (18) and other (18)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (28%)  FV and JV-FA/AA    Hypothermia (28–30°C), VF  123 ± 37 (VF-time)  201 ± 63  20/80  0    ST  Isolated MVR/MVP (62), MVR/MVP ± other valve ± CABG (17), CABG (11), AVR/repair ± aortic ± CABG (5) and other (10)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (15%)  IVC and SVC-AscA  Aortic cross-clamp  Combination of antegrade and retrograde blood cardioplegia (100%)  105 ± 46  180 ± 75  9/91    Data are expressed as absolute number, percentages or mean ± SD. a Total >100%; some patients received >1 previous surgery. AA: axillary artery; AscA: ascending aorta; AVR: aortic valve replacement; CABG: coronary artery bypass graft; CPB: cardiopulmonary bypass; FA: femoral artery; FV: femoral vein; IVC: inferior vena cava; JV: jugular vein; LVEF: left ventricular ejection fraction; MT: minithoracotomy; MVP: mitral valve plasty; MVR: mitral valve replacement; MVS: mitral valve surgery; NA: not available; NYHA: New York Heart Association; SD: standard deviation; ST: median sternotomy; STS: Society of Thoracic Surgeons; SVC: superior vena cava; TVP: tricuspid valve plasty; VF: ventricular fibrillation. Table 2: Supplementary data of interest from individual studies Study  Group  Previous surgery (%)  Mean time to redo surgery (years)  Preoperative patient characteristics  Cannulation site  Clamping technique  Myocardial protection  Clamping time (min)  CPB time (min)  MVP/ MVR (%)  Conversion to sternotomy (number of patients)  Burfeind et al. [20]  MT  MVS (60)  NA  LVEF (45 ± 9%) and mean NYHA (3.4)  FV-FA/AscA  Endoclamp (45%)  VF (55%) and additional percutaneous retrograde cardioplegia (33%)  NA  208 ± 76  NA  NA  ST  MVS (83)  NA  LVEF (54 ± 13%) and mean NYHA (3.5)  IVC and SVC-AscA  NA  VF (5%) and cardioplegia (95%)  NA  157 ± 53  NA    Bolotin et al. [21]  MT  NA  NA  LVEF (46 ± 12%) and mean NYHA (2.7)  FV and JV-FA    Hypothermia (26°C) and VF  NA  160 ± 65  42/58  0  ST  NA  NA  LVEF (55 ± 11%) and mean NYHA (2.6)  NA  NA  NA  NA  157 ± 54  9/91    Kim et al. [22]  MT  MVP (59), MVR (9), MVR + TVP/AVR (9), MVP + TVP (5) and other (18)  12 ± 9  LVEF (61 ± 9%), hypertension (5%) and diabetes mellitus (5%)  FV-FA  Transthoracic clamp  Moderate hypothermia and antegrade cardioplegia  91 ± 27  171 ± 47  77/23  NA  ST  MVP (46), MVP + TVP (8), MVR (8), MVR + TVP/AVR (8), CABG (8), AVR (8) and other (22)  10 ± 9  LVEF (45 ± 16%), hypertension (39%) and diabetes mellitus (23%)  IVC and SVC-AscA  NA  NA  102 ± 57  210 ± 103  100/0    Hiraoka et al. [23]  MT  MVP (50), redo MVR (10), AVR (10) and CABG (30)  NA  LVEF (47 ± 19%), mean NYHA (1.3 ± 0.5) and EuroSCORE (4.8 ± 2.0)  FV and JV-FA/AA    Hypothermia (27–30°C) and VF  90 ± 7 (VF-time)  145 ± 25  0/100  NA  ST  MVP (33), MVR (19), redo MVR (4), AVR (11), CABG (4), Bentall (4) and other (25)  NA  LVEF (64 ± 9), mean NYHA (1.3 ± 0.7) and EuroSCORE (3.8 + 2.4)  FV and SVC-FA  Aortic cross-clamp  Antegrade cardioplegia  84 ± 19  135 ± 28  0/100    Vallabhajosyula et al. [24]  MT  MVP (78), MVR (19) and MVP + MVR (3)  NA  LVEF (54 ± 12%), NYHA ≥2 (63%), hypertension (61%) and diabetes mellitus (13%)  FV and SVC-FA (97%)/AA (3%)  Endoclamp (81%) and Chitwood clamp (7%)  Hypothermia, antegrade cardioplegia (88%) and VF (12%)  104 ± 38  153 ± 42  NA  1  ST  MVP (41), MVR (53) and MVP + MVR (6)  NA  LVEF (54 ± 17%), NYHA ≥2 (69%), hypertension (55%) and diabetes mellitus (17%)  Central aortic cannulation (95%) and FA (5%)  Aortic cross-clamp (100%)  Antegrade and retrograde cardioplegia  130 ± 71  172 ± 83  NA    Losenno et al. [25]  MT  Isolated MVR/MVP (18), MVR/MVP ± other valve ± CABG (13), CABG (50), AVR/repair ± aortic ± CABG (18) and other (18)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (28%)  FV and JV-FA/AA    Hypothermia (28–30°C), VF  123 ± 37 (VF-time)  201 ± 63  20/80  0    ST  Isolated MVR/MVP (62), MVR/MVP ± other valve ± CABG (17), CABG (11), AVR/repair ± aortic ± CABG (5) and other (10)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (15%)  IVC and SVC-AscA  Aortic cross-clamp  Combination of antegrade and retrograde blood cardioplegia (100%)  105 ± 46  180 ± 75  9/91    Study  Group  Previous surgery (%)  Mean time to redo surgery (years)  Preoperative patient characteristics  Cannulation site  Clamping technique  Myocardial protection  Clamping time (min)  CPB time (min)  MVP/ MVR (%)  Conversion to sternotomy (number of patients)  Burfeind et al. [20]  MT  MVS (60)  NA  LVEF (45 ± 9%) and mean NYHA (3.4)  FV-FA/AscA  Endoclamp (45%)  VF (55%) and additional percutaneous retrograde cardioplegia (33%)  NA  208 ± 76  NA  NA  ST  MVS (83)  NA  LVEF (54 ± 13%) and mean NYHA (3.5)  IVC and SVC-AscA  NA  VF (5%) and cardioplegia (95%)  NA  157 ± 53  NA    Bolotin et al. [21]  MT  NA  NA  LVEF (46 ± 12%) and mean NYHA (2.7)  FV and JV-FA    Hypothermia (26°C) and VF  NA  160 ± 65  42/58  0  ST  NA  NA  LVEF (55 ± 11%) and mean NYHA (2.6)  NA  NA  NA  NA  157 ± 54  9/91    Kim et al. [22]  MT  MVP (59), MVR (9), MVR + TVP/AVR (9), MVP + TVP (5) and other (18)  12 ± 9  LVEF (61 ± 9%), hypertension (5%) and diabetes mellitus (5%)  FV-FA  Transthoracic clamp  Moderate hypothermia and antegrade cardioplegia  91 ± 27  171 ± 47  77/23  NA  ST  MVP (46), MVP + TVP (8), MVR (8), MVR + TVP/AVR (8), CABG (8), AVR (8) and other (22)  10 ± 9  LVEF (45 ± 16%), hypertension (39%) and diabetes mellitus (23%)  IVC and SVC-AscA  NA  NA  102 ± 57  210 ± 103  100/0    Hiraoka et al. [23]  MT  MVP (50), redo MVR (10), AVR (10) and CABG (30)  NA  LVEF (47 ± 19%), mean NYHA (1.3 ± 0.5) and EuroSCORE (4.8 ± 2.0)  FV and JV-FA/AA    Hypothermia (27–30°C) and VF  90 ± 7 (VF-time)  145 ± 25  0/100  NA  ST  MVP (33), MVR (19), redo MVR (4), AVR (11), CABG (4), Bentall (4) and other (25)  NA  LVEF (64 ± 9), mean NYHA (1.3 ± 0.7) and EuroSCORE (3.8 + 2.4)  FV and SVC-FA  Aortic cross-clamp  Antegrade cardioplegia  84 ± 19  135 ± 28  0/100    Vallabhajosyula et al. [24]  MT  MVP (78), MVR (19) and MVP + MVR (3)  NA  LVEF (54 ± 12%), NYHA ≥2 (63%), hypertension (61%) and diabetes mellitus (13%)  FV and SVC-FA (97%)/AA (3%)  Endoclamp (81%) and Chitwood clamp (7%)  Hypothermia, antegrade cardioplegia (88%) and VF (12%)  104 ± 38  153 ± 42  NA  1  ST  MVP (41), MVR (53) and MVP + MVR (6)  NA  LVEF (54 ± 17%), NYHA ≥2 (69%), hypertension (55%) and diabetes mellitus (17%)  Central aortic cannulation (95%) and FA (5%)  Aortic cross-clamp (100%)  Antegrade and retrograde cardioplegia  130 ± 71  172 ± 83  NA    Losenno et al. [25]  MT  Isolated MVR/MVP (18), MVR/MVP ± other valve ± CABG (13), CABG (50), AVR/repair ± aortic ± CABG (18) and other (18)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (28%)  FV and JV-FA/AA    Hypothermia (28–30°C), VF  123 ± 37 (VF-time)  201 ± 63  20/80  0    ST  Isolated MVR/MVP (62), MVR/MVP ± other valve ± CABG (17), CABG (11), AVR/repair ± aortic ± CABG (5) and other (10)a  NA  Mean NYHA (3.3), STS score (15 ± 11) and diabetes mellitus (15%)  IVC and SVC-AscA  Aortic cross-clamp  Combination of antegrade and retrograde blood cardioplegia (100%)  105 ± 46  180 ± 75  9/91    Data are expressed as absolute number, percentages or mean ± SD. a Total >100%; some patients received >1 previous surgery. AA: axillary artery; AscA: ascending aorta; AVR: aortic valve replacement; CABG: coronary artery bypass graft; CPB: cardiopulmonary bypass; FA: femoral artery; FV: femoral vein; IVC: inferior vena cava; JV: jugular vein; LVEF: left ventricular ejection fraction; MT: minithoracotomy; MVP: mitral valve plasty; MVR: mitral valve replacement; MVS: mitral valve surgery; NA: not available; NYHA: New York Heart Association; SD: standard deviation; ST: median sternotomy; STS: Society of Thoracic Surgeons; SVC: superior vena cava; TVP: tricuspid valve plasty; VF: ventricular fibrillation. Intervention Patients underwent surgery through a MT, initiated through a right anterolateral incision in the 4th or 5th intercostal space or a median ST approach, both aimed at repairing or replacing the diseased mitral valve. In some subjects, concomitant surgery was performed. Bolotin et al. [21] and Kim et al. [22] excluded participants who received such concomitant procedures in general whereas others only excluded those receiving any concomitant procedure other than tricuspid repair [20, 23]. Two reports excluded those who received concomitant procedures that were not amenable to a minimally invasive approach [24, 25]. Mean time to redo surgery was reported only by Kim et al. [22], whereas cannulation site, clamping technique, myocardial protection method, repair rate in redo surgeries (MVP/MVR), conversion of redo MT-MVS to ST-MVS, CPB and clamping time were described by at least 3 studies each (Table 2). Outcomes All studies reported the number of deaths, either as 30-day mortality rate, in-hospital deaths or deaths of early postoperative complications. In addition, all secondary outcome measures investigated in this review were assessed at least by 1 study each. Other outcomes investigated by different researchers, but not of interest for this review, were time in the intensive care unit, arrhythmias, chest tube output and several others [20–25]. Synthesis of results See Table 1 for all primary and secondary outcome measures. Death, stroke, reoperation for bleeding and wound infection were described as the number of events and as percentage of the group total, whereas LOHS was denoted as mean and SD. Wound infection and RBC transfusions were excluded from the meta-analysis because these outcomes were only reported in 2 or fewer studies each, and forest plots yield little value and may be seriously biased. Wound infections were reported by Vallabhajosyula et al. [24] and Losenno et al. [25] and occurred in 0.5% (1) and 3.3% (3) of patients who received ST-MVS (220 and 92), respectively. No wound infections were reported for either MT group. RBC transfusion data were reported by Losenno et al. [25], encompassing 132 patients. Overall, 63% (25) and 79% (73) of patients required such blood products in the MT (40) and ST (94) group, respectively. Intraoperative conversion from redo MT-MVS to ST-MVS was described by Bolotin et al. [21], Vallabhajosyula et al. [24] and Losenno et al. [25] (Table 2) and occurred in 0.7% (1) of patients. In addition, data for CPB times were available for all included studies. The mean CPB time was 177 ± 64 and 169 ± 74 min for MT- and ST-MVS, respectively (Table 2). Figures 2–5 present results of each meta-analysis performed, including measures of heterogeneity and 95% CIs. Data comparing deaths following MT- and ST-MVS were available for all 6 studies and were included for quantitative synthesis [20–25]. A total of 2.5% (6) and 8.9% (48) of patients who underwent an MT and an ST mitral valve reoperation (237 and 540; total 777) died, respectively. These percentages reveal a trend towards a reduced number of deaths of patients who had MT-MVS. Analysis confirmed this trend and showed (Fig. 2) that the MT approach was associated with a significantly reduced number of deaths compared to ST-MVS (OR 0.41, 95% CI 0.18–0.96; P = 0.04). In addition, evidence for heterogeneity was absent (I2 = 0%, P = 0.62). Quantitative synthesis for stroke incidence was based on 4 of the analysed studies [20, 23–25]. Overall, stroke occurred in 5.1% (9) and 3.4% (17) of patients who underwent MT- and ST-MVS (177 and 494, total = 671), respectively. Moreover, analysis revealed (Fig. 3) that the MT approach was not significantly associated with a higher occurrence of stroke (OR 1.51, 95% CI 0.65–3.54; P = 0.34) in the absence of heterogeneity (I2 = 0%, P = 0.86). Figure 2: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome death. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 2: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome death. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 3: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome stroke. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 3: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome stroke. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 4: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome reoperation for bleeding. *P = 0.0488. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 4: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome reoperation for bleeding. *P = 0.0488. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; ST-MVS: median sternotomy mitral valve surgery. Figure 5: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome length of hospital stay. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; SD: standard deviation; ST-MVS: median sternotomy mitral valve surgery. Figure 5: View largeDownload slide Comparison between redo MT-MVS and ST-MVS for the outcome length of hospital stay. CI: confidence interval; MT-MVS: minithoracotomy mitral valve surgery; SD: standard deviation; ST-MVS: median sternotomy mitral valve surgery. Reoperation for bleeding data were available for 5 observational studies, encompassing 706 patients [20, 22–25]. These subjects were subdivided into an MT and an ST group (199 and 507, respectively) in which reoperation for bleeding was required in 1.0% (2) and 4.1% (21) of cases, respectively. Initially (Fig. 4), MT-MVS was not significantly associated with a reduction in reoperations due to postoperative bleeding (OR 0.32, 95% CI 0.10–0.99; P = 0.05). However, Review Manager (RevMan v5.3) depicts rounded P-values in its forest plots, so a 4-decimal, comma-separated file was created and exported. This file revealed a P-value of 0.0488 by which MT-MVS was deemed to be significantly associated with a reduced need for reoperation for bleeding. No significant heterogeneity was detected within this comparison (I2 = 0%, P = 0.96). Data regarding LOHS were reported in 5 studies, which included 562 subjects [21–25]. LOHS data were presented on an identical scale among all studies so the difference in means was used. Mean LOHS for MT- and ST-MVS (177 and 385) was, respectively, 10.5 ± 11.0 days and 14.0 ± 11.7 days. A meta-analysis (Fig. 5) demonstrated that MT-MVS was significantly related with a lower LOHS compared to patients receiving MVS through a median ST (MD −3.81, 95% CI −5.53 to −2.08; P < 0.0001). There was no evidence of heterogeneity (I2 = 0%, P = 1.00). DISCUSSION Summary of evidence This systematic review examined whether the MT approach reduces the number of deaths compared to a conventional median ST among patients who received reoperative MVS and underwent prior cardiac surgery through a median ST. Stroke, reoperation for bleeding, LOHS, wound infection and RBC transfusions were added to this comparison as secondary outcome measures. Six non-randomized observational studies that met the eligibility criteria were included for review. The total number of patients was 777, with 237 in the MT group and 540 in the conventional ST group. Forest plots (Figs 2–5) revealed no heterogeneity for all outcome measures subjected to quantitative synthesis. Subsequently, homogeneity across studies was assumed. Detection of publication bias by means of a funnel plot was not performed due to the modest number of studies included for analysis. Quantitative analysis of deaths across studies revealed a homogeneous trend (Fig. 2). MT-MVS was significantly associated with a reduced mortality rate compared to ST-MVS. In addition, even though 3 definitions of mortality were considered as 1 for analysis, no heterogeneity was detected. In this review, the authors assumed death to be a direct consequence of the operative approach. However, all included reports were non-randomized retrospective observational studies, of which 5 [20–24] did not control for confounding. In these studies, baseline patient characteristics may have acted as potential confounders for death and influenced the intervention received at baseline. Apart from potential confounding, the experience of the surgeon may also have affected mortality rates. As reported by Holzhey et al. [26], adverse events, including death, are highly dependent on the surgeon’s experience in minimally invasive MVS. Furthermore, the authors described the influence of a learning curve, which comprises 75–125 procedures. Therefore, reported mortality rates with MT were assumed to be an overestimation compared to those from high volume centres, because all included studies [20–25] reported fewer than 75 minithoracotomies over a period exceeding 1 year at their respective institutions. MT-MVS was associated with a trend (in 3 of 4 studies) towards an increased, although not significant, risk of stroke (Fig. 3). Nevertheless, such embolic events remain a substantial concern when considering reoperative surgery through a MT. The trend in stroke risk may be explained by the observed differences in method for CPB across intervention groups. Retrograde perfusion was most frequently used across all 4 MT groups who reported stroke data [20, 23–25] compared to the ST groups, where 3 of the 4 studies mainly performed antegrade perfusion (AP) via central cannulation of the ascending aorta [20, 24, 25] (Table 2). In general, antegrade ascending aorta perfusion is not used during MT-MVS because the addition of the arterial cannula through the small 3- to 10-cm incision may impede or even prohibit adequate exposure compared to femoral artery exposure, which is easily achieved. Nevertheless, Murzi et al. [27] revealed that retrograde perfusion (compared to AP) was an independent risk factor for stroke (OR 4.28, P = 0.02) in patients who underwent MVS through a MT. Subsequently, AP via the axillary or subclavian artery was preferred over femoral cannulation in patients with peripheral vascular disease for MT-MVS. Axillary or subclavian artery cannulation for MT-MVS was only performed in a minority of cases in 3 studies. Vallabhajosyula et al. [24] performed this type of perfusion in 3% of cases; in contrast, Losenno et al. [23] and Hiraoka et al. [25] only performed axillary AP, depending on individual patient risk factors for atheroembolism and for elderly patients, respectively. Another concern that may contribute to the trend towards increased risk of stroke is inadequate deairing of cardiac cavities before closure. This trend may also be due to the restricted access in minimally invasive MVS, according to Botta et al. [14]. Subsequently, surgeons should utilize CO2 insufflation [28, 29] and venting [30, 31] procedures to mitigate this intracavity air in order to prevent embolic events. In Fig. 5, a homogeneous trend towards reduced LOHS for MT-MVS is seen. This trend was also demonstrated by meta-analysis; reoperation through a MT was significantly associated with a lower LOHS compared to patients who had ST-MVS, in the absence of any heterogeneity. The mean LOHS was 3.81 days shorter for MT-MVS. This more rapid patient recovery may be attributed to the minimally invasive nature of a MT approach, which yields less tissue trauma. This faster recovery was also described by Iribarne et al. [32, 33] for first-time operations. In addition, reoperation due to postoperative bleeding was also expected to be lower as a result of reduced wound surface and less tissue trauma with MT-MVS. Moreover, because MT-MVS utilized retrograde perfusion, no aortic and right atrial cannulations were required, which translated into fewer surgical seams and a reduced risk of postoperative bleeding. This observation was confirmed by analysis (Fig. 4) and consistently translates into diminished blood loss and need for transfusion. However, because transfusion outcomes were reported only in 1 study, quantitative synthesis was not performed. The same applies to wound infections, which were reported in 2 studies. Nevertheless, zero wound infections were observed for MT-MVS compared to 4 for ST-MVS [24, 25]. Among the included studies, only 2 reported use of preoperative computed tomography (CT) scans to address potential procedural difficulties or contraindications. Vallabhajosyula et al. [24] performed a CT scan out of concern for significant disease, whereas Hiraoka et al. [23] even performed an additional magnetic resonance imaging scan systematically. The added value of this preoperative imaging was previously described elsewhere by Heuts et al. [34] and may potentially prevent procedural complications. Apart from the proposed MT approach for redo MVS, several other techniques are currently evolving to circumvent the substantial risks associated with redo ST, such as mitral transcatheter valve-in-valve (MTVIV) and valve-in-ring implantations. Both may be particularly useful for redo surgery patients with high risk or multiple comorbidities after failed mitral bioprosthesis or failed mitral valve repair, respectively. Implantation of these transcatheter, balloon-expandable valves may be performed via a transseptal or transapical approach, whereby the latter is performed through a left anterior MT in the 5th or 6th intercostal space. Despite the fact that a transseptal approach may be less invasive, it is believed to be more technically challenging, by which MTVIV is mainly performed transapically. Nevertheless, the safety and feasibility of transvenous transseptal MTVIV are currently being evaluated by the prospective clinical MITRAL trial (Mitral ImpLantation of TRAnscatheter vaLves) and will potentially be used more frequently in the future [35]. Overall, these percutaneous MTVIV and valve-in-ring implantations have demonstrated excellent haemodynamic performance with low perivalvular regurgitation and transvalvular gradient in feasibility studies [36–39]. In the current European Society of Cardiology/European Association for Cardio-Thoracic Surgery 2017 guidelines for the management of valvular heart disease, these approaches were considered to be a reasonable alternative to redo operations for high-risk patients [40]. In summary, MT-MVS is a safe alternative for ST-MVS, with reduced mortality rates, reduced LOHS, lower incidence of reoperation for bleeding and comparable risk of stroke. Nevertheless, the trend towards increased risk of stroke may remain a substantial concern when considering reoperative MVS through a MT. Limitations Several limitations in this systematic review and meta-analysis should be addressed. The meta-analysis combined data across studies to estimate deaths and secondary outcomes. However, the main limitation of this analysis was the heterogeneity of risk profiles and selection criteria between centres. In addition, 3 different definitions of death were used, which could have led to bias. This systematic review only included retrospective observational studies because no randomized controlled trials have yet been published on this topic. Such observational studies are, however, more prone to confounding. In addition, the number of reports included was small, primarily because of the scarcity of comparative studies. The number of subjects within each MT group was also small and ranged from 10 to 67 compared to relatively larger ST groups, potentially limiting the level of evidence. CONCLUSIONS This meta-analysis is the first performed to test the differences between reoperative minimally invasive MVS through a right MT versus conventional median ST after prior cardiac surgery. The existing literature provided limited data but demonstrated significant differences with regards to mortality, LOHS and reoperation for bleeding, all in favour of MT-MVS. These benefits were evident despite the comparable risk of stroke. Moreover, because the current evidence is of insufficient quality, this review could function as a base for more methodologically rigorous randomized controlled trials. Conflict of interest: none declared. REFERENCES 1 Rankin JS, Hammill BG, Ferguson TB, Glower DD, O'Brien SM, DeLong ER. Determinants of operative mortality in valvular heart surgery. J Thorac Cardiovasc Surg  2006; 131: 547– 57. Google Scholar CrossRef Search ADS PubMed  2 Launcelott S, Ouzounian M, Buth KJ, Légaré J-F. Predicting in-hospital mortality after redo cardiac operations: development of a preoperative scorecard. Ann Thorac Surg  2012; 94: 778– 84. Google Scholar CrossRef Search ADS PubMed  3 Park CB, Suri RM, Burkhart HM, Greason KL, Dearani JA, Schaff HV et al.   Identifying patients at particular risk of injury during repeat sternotomy: analysis of 2555 cardiac reoperations. J Thorac Cardiovasc Surg  2010; 140: 1028– 35. Google Scholar CrossRef Search ADS PubMed  4 Gammie JS, Sheng S, Griffith BP, Peterson ED, Rankin JS, O'Brien SM et al.   Trends in mitral valve surgery in the United States: results from the Society of Thoracic Surgeons Adult Cardiac Database. Ann Thorac Surg  2009; 87: 1431– 9. Google Scholar CrossRef Search ADS PubMed  5 Seeburger J, Borger MA, Falk V, Passage J, Walther T, Doll N et al.   Minimally invasive mitral valve surgery after previous sternotomy: experience in 181 patients. Ann Thorac Surg  2009; 87: 709– 14. Google Scholar CrossRef Search ADS PubMed  6 Botta L, Cannata A, Fratto P, Bruschi G, Trunfio S, Maneggia C et al.   The role of the minimally invasive beating heart technique in reoperative valve surgery. J Card Surg  2012; 27: 24– 8. Google Scholar CrossRef Search ADS PubMed  7 Morales D, Williams E, John R. Is resternotomy in cardiac surgery still a problem? Interact CardioVasc Thorac Surg  2010; 11: 277– 86. Google Scholar CrossRef Search ADS PubMed  8 Roselli EE, Pettersson GB, Blackstone EH, Brizzio ME, Houghtaling PL, Hauck R et al.   Adverse events during reoperative cardiac surgery: frequency, characterization, and rescue. J Thorac Cardiovasc Surg  2008; 135: 316– 23.e6. Google Scholar CrossRef Search ADS PubMed  9 Ellman PI, Smith RL, Girotti ME, Thompson PW, Peeler BB, Kern JA et al.   Cardiac injury during resternotomy does not affect perioperative mortality. J Am Coll Surg  2008; 206: 993– 7. Google Scholar CrossRef Search ADS PubMed  10 Grossi EA, Galloway AC, LaPietra A, Ribakove GH, Ursomanno P, Delianides J et al.   Minimally invasive mitral valve surgery: a 6-year experience with 714 patients. Ann Thorac Surg  2002; 74: 660– 4. Google Scholar CrossRef Search ADS PubMed  11 Glower DD. Surgical approaches to mitral regurgitation. J Am Coll Cardiol  2012; 60: 1315– 22. Google Scholar CrossRef Search ADS PubMed  12 Vleissis AA, Bolling SF. Mini-reoperative mitral valve surgery. J Card Surg  1998; 13: 468– 70. Google Scholar CrossRef Search ADS PubMed  13 Schmitto JD, Mokashi SA, Cohn LH. Minimally-invasive valve surgery. J Am Coll Cardiol  2010; 56: 455– 62. Google Scholar CrossRef Search ADS PubMed  14 Botta L, Cannata A, Bruschi G, Fratto P, Taglieri C, Russo CF et al.   Minimally invasive approach for redo mitral valve surgery. J Thorac Dis  2013; 5(Suppl 6): S686. Google Scholar PubMed  15 Murzi M, Solinas M, Glauber M. Is a minimally invasive approach for re-operative mitral valve surgery superior to standard resternotomy? Interact CardioVasc Thorac Surg  2009; 9: 327– 32. Google Scholar CrossRef Search ADS PubMed  16 Modi P, Hassan A, Chitwood WR. Minimally invasive mitral valve surgery: a systematic review and meta-analysis. Eur J Cardiothorac Surg  2008; 34: 943– 52. Google Scholar CrossRef Search ADS PubMed  17 Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med  2009; 6: e1000097. Google Scholar CrossRef Search ADS PubMed  18 Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions . Hoboken (New Jersey), USA: John Wiley & Sons; 2011. 19 Altman DG, Bland JM. Standard deviations and standard errors. BMJ  2005; 331: 903. Google Scholar CrossRef Search ADS PubMed  20 Burfeind WR, Glower DD, Davis RD, Landolfo KP, Lowe JE, Wolfe WG. Mitral surgery after prior cardiac operation: port-access versus sternotomy or thoracotomy. Ann Thorac Surg  2002; 74: S1323– 5. Google Scholar CrossRef Search ADS PubMed  21 Bolotin G, Kypson AP, Reade CC, Chu VF, Freund WLJr, Nifong LW. Should a video-assisted mini-thoracotomy be the approach of choice for reoperative mitral valve surgery? J Heart Valve Dis  2004; 13: 155– 8; discussion 8. Google Scholar PubMed  22 Kim DC, Chee HK, Song MG, Shin JK, Kim JS, Lee S et al.   Comparative analysis of thoracotomy and sternotomy approaches in cardiac reoperation. Korean J Thorac Cardiovasc Surg  2012; 45: 225. Google Scholar CrossRef Search ADS PubMed  23 Hiraoka A, Kuinose M, Totsugawa T, Chikazawa G, Yoshitaka H. Mitral valve reoperation under ventricular fibrillation through right mini-thoracotomy using three-dimensional videoscope. J Cardiothorac Surg  2013; 8: 81. Google Scholar CrossRef Search ADS PubMed  24 Vallabhajosyula P, Wallen T, Pulsipher A, Pitkin E, Solometo LP, Musthaq S et al.   Minimally invasive port access approach for reoperations on the mitral valve. Ann Thorac Surg  2015; 100: 68– 73. Google Scholar CrossRef Search ADS PubMed  25 Losenno KL, Jones PM, Valdis M, Fox SA, Kiaii B, Chu MW. Higher-risk mitral valve operations after previous sternotomy: endoscopic, minimally invasive approach improves patient outcomes. Can J Surg  2016; 59: 399– 406. Google Scholar CrossRef Search ADS PubMed  26 Holzhey DM, Seeburger J, Misfeld M, Borger MA, Mohr FW. Learning minimally-invasive mitral valve surgery: a cumulative sum sequential probability analysis of 3895 operations from a single high volume center. Circulation  2013; 128: 483– 91. Google Scholar CrossRef Search ADS PubMed  27 Murzi M, Cerillo AG, Miceli A, Bevilacqua S, Kallushi E, Farneti P et al.   Antegrade and retrograde arterial perfusion strategy in minimally invasive mitral-valve surgery: a propensity score analysis on 1280 patients. Eur J Cardiothorac Surg  2013; 43: e167– 72. Google Scholar CrossRef Search ADS PubMed  28 Martens S, Neumann K, Sodemann C, Deschka H, Wimmer-Greinecker G, Moritz A. Carbon dioxide field flooding reduces neurologic impairment after open heart surgery. Ann Thorac Surg  2008; 85: 543– 7. Google Scholar CrossRef Search ADS PubMed  29 Svenarud P, Persson M, Van Der Linden J. Effect of CO2 insufflation on the number and behavior of air microemboli in open-heart surgery. Circulation  2004; 109: 1127– 32. Google Scholar CrossRef Search ADS PubMed  30 Reser D, Walser R, van Hemelrijk M, Holubec T, Weber A, Plass A et al.   Long-term outcomes after minimally invasive aortic valve surgery through right anterior minithoracotomy. Thorac Cardiovasc Surg  2017; 65: 191– 7. Google Scholar CrossRef Search ADS PubMed  31 Sharony R, Grossi EA, Saunders PC, Schwartz CF, Ursomanno P, Ribakove GH et al.   Minimally invasive reoperative isolated valve surgery: early and mid-term results. J Card Surgery  2006; 21: 240– 4. Google Scholar CrossRef Search ADS   32 Iribarne A, Easterwood R, Russo MJ, Chan EY, Smith CR, Argenziano M. Comparative effectiveness of minimally invasive versus traditional sternotomy mitral valve surgery in elderly patients. J Thorac Cardiovasc Surg  2012; 143(4 Suppl): S86– 90. Google Scholar CrossRef Search ADS   33 Iribarne A, Russo MJ, Easterwood R, Hong KN, Yang J, Cheema FH et al.   Minimally invasive versus sternotomy approach for mitral valve surgery: a propensity analysis. Ann Thorac Surg  2010; 90: 1471– 8. Google Scholar CrossRef Search ADS PubMed  34 Heuts S, Maessen JG, Sardari Nia P. Preoperative planning of left-sided valve surgery with 3D computed tomography reconstruction models: sternotomy or a minimally invasive approach? Interact CardioVasc Thorac Surg  2016; 22: 587– 93. Google Scholar CrossRef Search ADS PubMed  35 Guerrero M, Salinger M, Pursnani A, Pearson P, Lampert M, Levisay J et al.   Transseptal transcatheter mitral valve‐in‐valve: a step by step guide from preprocedural planning to postprocedural care. Catheter Cardiovasc Interv  2017; 00: 1– 12. 36 Cheung A, Al-Lawati A. Transcatheter mitral valve-in-valve implantation: current experience and review of literature . Curr Opin Cardiol  2013; 28: 181– 6. Google Scholar CrossRef Search ADS PubMed  37 Cheung A, Webb JG, Barbanti M, Freeman M, Binder RK, Thompson C et al.   5-year experience with transcatheter transapical mitral valve-in-valve implantation for bioprosthetic valve dysfunction. J Am Coll Cardiol  2013; 61: 1759– 66. Google Scholar CrossRef Search ADS PubMed  38 Descoutures F, Himbert D, Maisano F, Casselman F, de Weger A, Bodea O et al.   Transcatheter valve-in-ring implantation after failure of surgical mitral repair. Eur J Cardiothorac Surg  2013; 44: e8– 15. Google Scholar CrossRef Search ADS PubMed  39 Cullen MW, Cabalka AK, Alli OO, Pislaru SV, Sorajja P, Nkomo VT et al.   Transvenous, antegrade melody valve-in-valve implantation for bioprosthetic mitral and tricuspid valve dysfunction: a case series in children and adults. JACC Cardiovasc Interv  2013; 6: 598– 605. Google Scholar CrossRef Search ADS PubMed  40 Falk V, Baumgartner H, Bax JJ, De Bonis M, Hamm C, Holm PJ et al.  ; ESC Scientific Document Group. 2017 ESC/EACTS guidelines for the management of valvular heart disease. Eur J Cardiothorac Surg  2017; 52: 616– 64. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

European Journal of Cardio-Thoracic SurgeryOxford University Press

Published: Apr 23, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off