Abstract

OBJECTIVES

Several single-centre experiences have reported significant operative mortality and morbidity after mitral valve surgery in redo scenarios (ReMVS). Several preoperative risk factors outlining ‘high-risk’ categories have been reported, but scanty data on the impact of different operative techniques for these major challenging procedures have been analysed to date. The aim of the study is to investigate those intraoperative factors and postoperative events affecting early survival after redo mitral procedures.

METHODS

Operative mortality and major morbidity events from a large multicentre registry enrolling 832 consecutive redo mitral procedures were analysed. Intraoperative technical issues and postoperative complications impacting operative mortality were identified.

RESULTS

ReMVS was associated with significant operative mortality (12.5%), acute myocardial infarction (AMI; 5.9%), stroke (4.9%), acute respiratory insufficiency (14.8%), pneumonia (7.0%), acute renal insufficiency (16.1%) and failure (12.6%), reintervention for bleeding (7.6%), massive transfusion (28.0%), need for permanent pacemaker (10.1%). Injury of a previous patent left internal mammary artery (LIMA) graft [odds ratio (OR) 4.2, 95% confidence interval (CI): 1.6–11.5; P = 0.005], major cardiovascular iatrogenic lesions at re-entry (OR 19.2, 95% CI: 9.2–39.9; P < 0.001), extracellular crystalloid cardioplegia (OR 7.3, 95% CI: 1.4–37.8; P = 0.018), and incremental cardiopulmonary bypass time (OR 1.1, 95% CI: 1.0–1.2; P = 0.001) independently predicted operative mortality, whereas combined antegrade + retrograde cardioplegia (OR 0.2, 95% CI: 0.09–0.4; P = 0.001) was the only protective factor against mortality. Among complications, AMI (OR 4.1, 95% CI: 1.8–9.6; P = 0.001), need for intra-aortic balloon pumping (IABP; OR 3.1, 95% CI: 1.5–6.1; P = 0.001), prolonged intubation >48 h (OR 5.3, 95% CI: 2.9–9.4; P = 0.001) and massive (>6 units) transfusions (OR 4.4, 95% CI: 2.4–8.0; P = 0.001) also predicted operative mortality.

CONCLUSIONS

ReMVS still carries the risk of significant early mortality and major morbidity. Major lesion to cardiovascular structures is the most dreadful iatrogenic complication, and injury of a previous LIMA graft identifies patients at higher risk of operative mortality. Prolonged cross-clamp times, extracellular crystalloid cardioplegia and massive transfusions have profound impact on early outcome, as well as the development of perioperative AMI, eventually requiring IABP and prolonged intubation. The combination of antegrade and retrograde cardioplegia seems to offer a better myocardial protection in these high-risk patients.

INTRODUCTION

Mitral valve surgery after prior cardiac surgery (ReMVS) is considered a demanding procedure with higher risks of adverse events than first-time surgery [ 1 ]. Although a number of institutional series have reported a reduction in operative mortality in ReMVS during the last years [ 2–4 ], a recent study still reported a 12% hospital mortality in these patients [ 5 ]. Several preoperative conditions have been outlined, identifying patients at unusual surgical risk [ 2–5 ], whereas very few studies clarify technical issues and perioperative complications affecting early survival. Furthermore, no data from large multicentre experiences are currently available to thoroughly asses the outcome of ReMVS [ 2–5 ].

Interestingly, on the basis of the incremental operative mortality predicted by current risk scores and already reported in single-centre surgical experiences, ReMVS has been occasionally managed by off-label transcatheter valve implantation, despite the lack for proven evidence-based advantage [ 6 , 7 ]. Indeed, EuroSCORE and STS risk models have been demonstrated to overestimate the risk of death in high-risk mitral surgery, thus promoting, unless put in perspective, the possibility for an overtreatment of redo mitral patients with new emerging but still empirical technologies [ 7 ].

Therefore, it is the aim of the present study to analyse data from a large multicentre European registry (REdo Cardiac Operations Research Database—RECORD) encompassing all-redo heart valve procedures performed in eight European institutions [ 8 ], to assess the early outcome and to identify preoperative and intraoperative technical issues, as well as early postoperative events affecting survival after ReMVS.

MATERIALS AND METHODS

This is an analysis of 832 consecutive adult patients with prior cardiac surgery undergoing mitral valve surgery from January 2005 to December 2014 at eight European centres (Italy, France, Germany and Finland), contributing to the multicentre REdo Cardiac Operation Research Database (RECORD) [ 8 ]. The only inclusion criterion for this study was any type of mitral valve surgery performed in patients with a history of prior cardiac surgery. For the purposes of this study, ‘redo mitral valve surgery (ReMVS)’ refers to those patients undergoing any mitral valve surgery as the index procedure after any major cardiac surgery procedure; ‘number of reintervention’ defines the progressive number of reintervention in each patient; ‘number of mitral intervention’ identifies the progressive number of (re)intervention on the mitral valve. Baseline characteristics were defined according to the EuroSCORE II definition criteria [ 9 ].

The choice to analyse data from 2005—thus excluding earlier patients—was taken to avoid potential biases related to differences in perioperative management and care, as well as to have a picture of ‘current’ ReMVS practice [ 8 ]. Data related to hospitalization for the index procedure were retrieved from institutional databases and hospital charts. Data on hospital course were available for all patients. Institutional Review Board/Ethical Committee approved the study, but individual patient consent was waived due to the retrospective, observational nature of the study.

Surgery

The choice for a mitral valve repair, re-repair or a mechanical or biological prosthetic replacement, was based on single institutional policies and on patient preference after adequate informed consent. Anaesthesia, surgery, myocardial protection techniques and cardiopulmonary bypass (CPB) strategies, as well as postoperative care were based on individual institution's standardized protocols. Intraoperative details were collected from every patient's intraoperative surgical report. All resternotomies were performed with the aid of an oscillating saw. The use of preoperative computed tomography (CT) scan, institution of CPB before resternotomy and use of thoracotomy approaches varied among experiences, and were analysed as variables potentially impacting operative mortality (see Supplementary Data ).

Outcome end-points

The primary outcome end-point of this study was operative mortality, defined as all-cause mortality during the index hospitalization (including rehabilitation hospital stay, if discharged to a rehabilitation clinic) or during the first 30 postoperative days (if discharged home) [ 8 , 10 ]. Events occurred after index hospital discharge were collected from rehabilitation's clinic charts, outpatient clinics at the individual institutions, phone contacts with referral cardiologists or general practitioners, linking with regional Social Security Death and Events Master files where available, or—in the absence of recent data—by direct phone contact with patients and families [ 8 ]. The secondary outcome end-points were: re-exploration for bleeding, defined as any reoperation during the index hospitalization, due to postoperative bleeding; need for permanent pacemaker implantation; perioperative myocardial infarction (MI), according to the third universal definition of MI [ 8 , 11 ]; low cardiac output syndrome (LCOS), defined as haemodynamic instability for more than 1 h during the stay in the intensive care unit (ICU), with signs of peripheral hypo-perfusion, despite inotropic support and adequate correction of preload, afterload and all electrolyte and blood gas abnormalities [ 8 ]; need for postoperative (since aortic declamping) intra-aortic balloon pumping (IABP); prolonged intubation, defined as mechanical ventilation longer than 48 h; acute respiratory failure (ARF), defined as prolonged intubation and/or respiratory insufficiency after extubation with the need of reintubation or need for non-invasive ventilation lasting more than 48 h [ 8 ]; pneumonia, defined as evidence of bacterial growth in the lung with at least one positive bronchoalveolar fluid lavage culture, together with new alveolar infiltrates at chest roentgenogram, irrespective of the presence of fever or leucocytosis, or as evidence of new alveolar infiltrates with leucocytosis and purulent sputum, confirmed by CT scan and/or by consultation of an independent infectivologist or pneumologist [ 8 ]; stroke, defined as for current guidelines [ 8 , 12 ]: briefly, it was defined as perioperative cerebrovascular accident, whose symptoms lasted longer than 24 h with or without residual disability, confirmed at CT or magnetic resonance imaging; in case of no evidence of stroke at neuro-imaging, the diagnosis of stroke was made by consultant neurologists [ 8 , 12 ]; acute renal insufficiency (ARI), defined as a greater than 50% increase over the preoperative serum creatinine value [ 8 , 13 ]; need for continuous renal replacement therapy; need of transfusions, defined by the transfusion of at least one unit of red packed cells, fresh frozen plasma, platelets [ 8 ]; massive intra- and/or postoperative (i.e. perioperative) transfusions, defined as transfusion of ≥6 units of red blood cells, platelets and/or fresh frozen plasma [ 8 ]; deep sternal wound infection, defined as any bacterial/fungal infection involving the sternum [ 8 ]; length in ICU and length of in-hospital stay. Outcome end-points were adjudicated by two co-authors for each participating centre—in light of current outcome definition criteria—by retrieving data from the hospital charts of each patient enrolled in the registry [ 8 , 11–13 ]. Outcome adjudication was then confirmed by two independent reviewers (Francesco Onorati and Andrea Perrotti ).

Data on the following variables pertaining to the previous cardiac procedures and treatment strategy at index procedure were considered in this analysis: participating centre; number of any cardiac reinterventions ; number of reinterventions on the mitral valve; preoperative CT scanning; instauration of CPB through peripheral vascular access before sternotomy; major cardiovascular iatrogenic injury at re-entry, defined as any severe and/or life-threatening (i.e. requiring re-animation and/or immediate changing of the surgical plan and/or massive transfusions) injury of major vessels or cardiac structures occurring during surgical re-entry [ 8 ]; type of surgical intervention; type of mitral valve surgery at redo (i.e. repair or replacement); type of implanted prosthesis (i.e. biological or mechanical); brand of implanted prosthesis; size of implanted prosthesis; type of mitral ring (if mitral repair was executed, i.e. complete, incomplete, pericardial band); mitral ring size; brand of implanted ring; previous coronary artery bypass grafting (CABG); previous valve surgery; previous aortic surgery; number of previous CABG; number of patent previous grafts; previous left internal mammary artery (LIMA) graft; dissection of a patent previous LIMA graft; clamping of a patent previous LIMA graft; intraoperative injury of a patent previous LIMA graft; route of cardioplegia (i.e. antegrade, retrograde, antegrade + retrograde, no cardioplegia with fibrillating cooled unloaded heart); administration of cardioplegia into previous patent vein grafts; type of medium of cardioplegic solution (blood, intracellular crystalloid, extracellular crystalloid, cold fibrillating heart); temperature of cardioplegia; lowest systemic temperature during CPB; aortic cross-clamping time; CPB time.

Statistical analysis

Statistical analysis was performed by the SPSS program for Windows, version 13.0 (SPSS, Inc., Chicago, IL, USA). Continuous variables are presented as the mean and standard deviation, and categorical variables are presented as counts and percentages. Data were checked for normality before statistical analysis. Normally distributed continuous variables were compared using the unpaired t -test, whereas the Mann–Whitney U -test was used for not-normally distributed variables. Categorical variables were analysed using the χ 2 test. Multivariable analysis was performed to identify those preoperative covariates, technical issues and postoperative complications independently predicting operative mortality, employing stepwise logistic regression with backward selection. Only variables with a P < 0.10 at univariable analysis were included into the regression models to avoid overfitting. Hosmer–Lemeshow's test was used to assess the regression models fit. The area under the receiver operating characteristic (ROC) curve was used to represent the discriminatory ability of the regression models. Models were expressed in terms of adjusted odds ratio (OR) and 95% confidence interval (CI). A P -value of less than 0.05 was considered significant.

RESULTS

Baseline and operative characteristics of patients are summarized in Supplementary Data . One hundred and ninety-two patients underwent previous aortic valve replacement (AVR) (23.1%). Although mortality was higher in patients with previous AVR (32/192 patients, 16.7% vs no-previous AVR 72/640, 11.3%; P = 0.047), it was not an independent predictor of operative mortality ( P = 0.164 at multivariable analysis). However, these patients reported longer cross-clamp time (111.7 ± 44.9 vs 92.0 ± 40.2 min, P = 0.001) and CPB time (163.6 ± 67.6 vs 140.2 ± 62.7 min, P = 0.001), higher need for IABP (20.3 vs 13.9%, P = 0.031) and transfusions (82.3 vs 72.4%, P = 0.012) and a higher incidence of stroke (7.8 vs 4.1%, P = 0.035). Other outcome end-points were comparable ( P = not significant).

Native mitral valve at index procedure was reported in 162 patients (19.5%). This baseline covariate reported also longer cross-clamp time (111.9 ± 49.1 min vs previous mitral valve (MV) surgery 92.8 ± 39.5, P = 0.001) and CPB time (159.6 ± 73.7 min vs 142.5 ± 61.7, P = 0.002), together with a higher need for massive transfusions (35.2 vs 26.3%, P = 0.024). However, operative mortality ( P = 0.467) and other morbidity end-points were not different.

The overall rate of MV repair was 11.5% (96 patients). Rate of repair differed between urgent/emergent indication [4.5 vs mitral valve replacement (MVR) 95.5%, P = 0.002] and left ventricular ejection fraction (LVEF >50%: MV repair: 9.6% vs MVR: 90.4%; LVEF 30–50%: 14.2 vs 85.8%; LVEF <30%: 21.1 vs 78.9%; P = 0.013), but it did not differ between the number of reinterventions (repair: 12.1 vs MVR 87.9% in first redo, 9.3 vs 90.7% in second redo, 3.6 vs 96.4% in third redo, 14.3 vs 85.7% in fourth redo; P = 0.470) and type of endocarditis (repair 6.5 vs MVR 93.5% in acute endocarditis; 3.8 vs 96.2% in subacute endocarditis; 14.3 vs 85.7% in healed endocarditis; P = 0.200). However, a clear trend towards a high rate of repair in endocarditis was evident in healed versus acute and subacute aetiology: 14.3 vs 6.5% and 3.8%.

Preoperative CT scan in 194 patients (23.3%), CPB before resternotomy in 25 (3.0%) and alternative thoracotomy approach in 23 patients (2.8%) were carried out. One hundred and four patients (12.5%) died during hospitalization. Perioperative complications were as follows: major cardiovascular injury at re-entry 6.6% (55/832), MI 5.9% (49/832), LCOS 21.3% (177/832), need of IABP 15.4% (128/832), stroke 4.9% (41/832), prolonged intubation 19.4% (161/832), ARF 14.8% (123/832), pneumonia 7.0% (58/832), ARI 16.1% (134/832), acute renal failure requiring renal replacement therapy 12.6% (105/832), need for transfusions 75.4% (627/832), massive transfusions 28.0% (233/832), reoperation for bleeding 7.6% (63/832), need for permanent pacemaker implantation 10.1% (84/832) and deep sternal wound infection 2.6% (22/832).

Determinants of operative mortality

The results of univariable analysis of baseline, operative and postoperative covariates for prediction of operative mortality are reported in the Supplementary Data . Multivariable analysis demonstrated that four preoperative variables were independent predictors of operative mortality: preoperative New York Heart Association (NYHA), acute endocarditis, preoperative LVEF and previous CABG (Table 1 ; area under the ROC curve of the regression model was 0.78; 95% CI: 0.73–0.83, Hosmer–Lemeshow test P = 0.461). Furthermore, intraoperative injury of a patent previous LIMA graft, major cardiovascular injury at re-entry, use of extracellular crystalloid cardioplegia and prolonged CPB time were independent predictors of operative mortality, whereas the combined use of antegrade and retrograde cardioplegia was associated with improved early survival (area under the ROC curve of this regression model was 0.85; 95% CI: 0.81–0.90, Hosmer–Lemeshow test: P = 0.692; Table 2 ). Participating centre ( P = 0.211), preoperative CT scan ( P = 0.420), CPB before resternotomy ( P = 0.490), thoracotomy approach ( P = 0.924), number of previous reinterventions ( P = 0.136), number of mitral reinterventions (Table 1 ), type of mitral surgery at index procedure ( P = 0.512) and the status of the mitral valve at index procedure (Table 1 ) did not affect operative mortality.

Table 1:

Preoperative covariates predictive of operative mortality

Factor OR 95% CI P -values  
Previous CABG 6.7 2.3–20.0 0.001 
Acute endocarditis 5.3 2.5–11.6 0.001 
Preoperative LVEF 2.6 1.7–3.8 0.001 
Preoperative NYHA 1.9 1.3–2.9 0.004 
Mitral valve status at index procedure 4.5 0.4–11.3 0.059 
COPD 3.8 0.5–20.5 0.071 
Urgent/emergent procedure 2.2 0.6–10.5 0.089 
Number of mitral intervention 1.9 0.8–14.3 0.690 
Factor OR 95% CI P -values  
Previous CABG 6.7 2.3–20.0 0.001 
Acute endocarditis 5.3 2.5–11.6 0.001 
Preoperative LVEF 2.6 1.7–3.8 0.001 
Preoperative NYHA 1.9 1.3–2.9 0.004 
Mitral valve status at index procedure 4.5 0.4–11.3 0.059 
COPD 3.8 0.5–20.5 0.071 
Urgent/emergent procedure 2.2 0.6–10.5 0.089 
Number of mitral intervention 1.9 0.8–14.3 0.690 

OR: odds ratio; CI: confidence interval; CABG: coronary artery bypass grafting; COPD: chronic obstructive pulmonary disease; LVEF: left ventricular ejection fraction; NYHA: New York Heart Association.

Table 2:

Surgical predictors of operative mortality after mitral valve surgery in patients with previous cardiac surgery

Factor OR 95% CI P -values  
Major cardiovascular iatrogenic injury at re-entry 19.2 9.2–39.9 <0.001 
Extracellular crystalloid cardioplegia 7.3 1.4–37.8 0.018 
Intraoperative injury of patent LIMA graft 4.2 1.6–11.5 0.005 
CPB duration (per each increasing minute) 1.1 1.0–1.2 0.001 
Combined antegrade and retrograde cardioplegia 0.2 0.09–0.4 0.001 
Lowest systemic temperature during CPB 1.1 0.9–1.2 0.075 
XCT duration (per each increasing minute) 1.0 0.9–1.1 0.100 
Factor OR 95% CI P -values  
Major cardiovascular iatrogenic injury at re-entry 19.2 9.2–39.9 <0.001 
Extracellular crystalloid cardioplegia 7.3 1.4–37.8 0.018 
Intraoperative injury of patent LIMA graft 4.2 1.6–11.5 0.005 
CPB duration (per each increasing minute) 1.1 1.0–1.2 0.001 
Combined antegrade and retrograde cardioplegia 0.2 0.09–0.4 0.001 
Lowest systemic temperature during CPB 1.1 0.9–1.2 0.075 
XCT duration (per each increasing minute) 1.0 0.9–1.1 0.100 

OR: odds ratio; CI: confidence interval; CPB: cardiopulmonary bypass; LIMA: left internal mammary artery; XCT: aortic cross-clamping time.

Among the postoperative adverse events, MI, need for postoperative IABP, prolonged intubation and perioperative massive transfusions were independent predictors of operative mortality (Table 3 ; area under the ROC curve for this regression model was 0.92; 95% CI 0.89–0.95, Hosmer–Lemeshow test: P = 0.594).

Table 3:

Postoperative adverse events associated with operative mortality

Postoperative adverse event OR 95% CI P -values  
Prolonged intubation (>48 h) 5.3 2.9–9.4 0.001 
Perioperative massive transfusions (>6 units of blood products) 4.4 2.4–8.0 0.001 
MI 4.1 1.8–9.6 0.001 
Need for postoperative IABP 3.1 1.5–6.1 0.001 
Low cardiac output syndrome 1.8 0.8–3.5 0.099 
Postoperative adverse event OR 95% CI P -values  
Prolonged intubation (>48 h) 5.3 2.9–9.4 0.001 
Perioperative massive transfusions (>6 units of blood products) 4.4 2.4–8.0 0.001 
MI 4.1 1.8–9.6 0.001 
Need for postoperative IABP 3.1 1.5–6.1 0.001 
Low cardiac output syndrome 1.8 0.8–3.5 0.099 

OR: odds ratio; CI: confidence interval; MI: myocardial infarction; IABP: intra-aortic balloon pump.

DISCUSSION

This is the first multicentre European experience reporting hospital outcome and exploring technical issues responsible for operative mortality in a large population ( n = 832) of patients undergoing redo mitral surgery. A few institutional series of retrospective nature reported on the outcome of patients undergoing mitral valve surgery in the redo setting [ 1–5 , 14 ]. Most of these studies however evaluated heterogenous populations [ 2–4 ], encompassed long study periods during which different anaesthesiological and surgical improvements occurred [ 3 ], included case-mix with different heart valve surgical procedures at redo [ 1 ] or evaluated only mechanical [ 14 ] or biological [ 3 ] failed prostheses. Furthermore, most of these series were of limited size [ 2 , 4 , 5 ], probably due to the sporadic referral for reintervention in high-risk patients.

Previous studies have identified preoperative risk factors affecting early and late mortality, such as advanced age at surgery [ 3 ], left ventricular dysfunction [ 5 , 15 ], urgent or emergent priority [ 1–3 , 5 ], increased NYHA class [ 2 , 4 , 15 ] and pulmonary hypertension [ 4 ]. In our patient population, we similarly reported preoperative NYHA class and preoperative LVEF, together with active endocarditis and previous CABG, to predict operative mortality. Indeed, most previous studies identified preoperative risk conditions which were usually not modifiable, especially if patients were in labile cardiac compensation because of severe mitral disease. Accordingly, most of these authors advocated an early referral to surgery for these patients as a means to improve outcome [ 1–5 , 15 ]. Overall, most of these factors are well-known risky conditions for every type of cardiac surgery and, indeed, they are largely incorporated in both EuroSCORE and STS scores [ 7 ]. On the contrary, data on intraoperative and/or postoperative truly modifiable factors affecting outcome lack in the current literature.

A number of studies reported excellent results from institutional series showing an improvement in postoperative survival during the recent years [ 1 , 3 ]. Jamieson et al. reported operative mortality as low as 3.4% [ 3 ], Emery et al. ∼9% [ 14 ], Goreishi et al. below 5% [ 4 ], all reflecting differences in baseline characteristics and risk profile of the population. However, a recent study from UK reported a hospital mortality of ∼12% [ 5 ]. Similarly, in our large multicentre experience, we reported an operative mortality of 12.5%, quite well predicted by the EuroSCORE II (mean value of 13%; observed/predicted ratio = 0.9), thus confirming that redo mitral surgery is still a high-risk surgical procedure regardless of surgical centre and baseline characteristics related to ethnicity or individual institutional policies. Furthermore, we also observed a non-negligible rate of major morbidities. These data underline the need for an accurate risk/benefits ratio analysis in light of the currently developing less-invasive transcatheter options for failed mitral repair or dysfunctional bioprosthesis [ 6 , 7 ].

The occurrence of major end-organ complications is of utmost importance since some adverse events (MI, IABP, prolonged intubation and massive transfusions) were independent predictors of operative mortality. These findings claim for further improvements in critical care medicine, and highlight the crucial role of perioperative myocardial protection strategies and the importance of perioperative blood conservation protocols [ 16 ]. The dreadful role of inadequate myocardial protection and excessive bleeding is further highlighted by the identification, in the present study, of crucial technical issues related to the intraoperative management of redo mitral surgery. The most dreadful factor is the development of a major iatrogenic cardiovascular complication at reentry (OR >19) and, in particular, the iatrogenic injury of a patent LIMA graft in patients already subjected to previous CABG, the latter leading to a 4-fold increase in operative mortality. An adequate surgical planning appears therefore mandatory before any resternotomy and requires extreme caution in patients with patent LIMA grafts [ 17 , 18 ]. Furthermore, it has to be underscored that other experiences reported alternative approaches, such as thoracotomy, to overcome the limitations of resternotomy, especially in patients with patent LIMA grafts [ 19 ]. Our data concur with Morales et al. who pointed out the link among iatrogenic injury to major cardiovascular structures at reentry, severe bleeding with massive transfusions and perioperative myocardial damage related to severe systemic hypotension [ 18 ], underscoring the dreadful role of these events when occurring altogether, as also proved by the 20% mortality reported in an STS survey [ 20 ].

It is interesting to note that preoperative CT, considered an important preoperative means for safe surgical planning of repeat cardiac surgery [ 17 , 18 ], was not a protective factor in this study. Similarly, institution of CPB before resternotomy [ 21 ] and alternative thoracotomy approach [ 19 ] were not protective factors towards operative mortality in this experience. One possible explanation for these findings could be the fact that these procedures are usually performed by experienced surgeons regardless of preoperative CT scan, presternotomy CPB or alternative entry sites [ 18 , 19 , 21 ]. In this light can also be interpreted our finding that incremental number of resternotomy did not affect mortality, as already stated by others in recent experiences [ 1 , 15 ]. Another possible explanation can be the crucial concurrent role of other ‘technical issues’—different from the mere availability of preoperative CT scan—in determining operative mortality. Of these, intraoperative myocardial protection plays an eminent role in our experience. Extracellular crystalloid cardioplegia was associated with increased mortality. On the contrary, combined antegrade and retrograde delivery of cardioplegia proved to be the strategy of choice for improving early survival of these high-risk patients. Indeed, a recent meta-analysis of 12 trials on 2866 patients confirmed the superiority of blood cardioplegia over crystalloid cardioplegia in preventing postoperative MI [ 22 ]. Similarly, another meta-analysis of 34 trials demonstrated a significant lower incidence of LCOS and myocardial enzyme leakage with blood cardioplegia compared with crystalloid cardioplegia in different cardiac surgical procedures [ 23 ]. The protective role of combined antegrade and retrograde delivery of cardioplegia observed in this series confirmed the findings of previous studies, demonstrating a beneficial role in terms of a more homogeneous distribution of cardioplegia to the arrested myocardium, a reduced risk of antegrade athero-embolism from patent CABG, the potential for a retrograde dislodgement of already-occurred athero-emboli or air-emboli, as well as the delivery to territories with in situ patent arterial grafts [ 24 ]. Another interesting finding is that we were not able to demonstrate the superiority of clamping of patent LIMA graft compared with a strategy leaving the LIMA graft untouched. On the contrary, inadvertent LIMA graft injury during tissue dissection significantly affected hospital survival as previously observed in aortic valve replacement in the redo setting [ 25 ].

In conclusion, this multicentre study confirmed that mitral valve surgery in patients with a history of prior cardiac surgery is a demanding procedure associated with significant mortality and morbidity. Avoidance of injury to cardiovascular structure at re-entry, myocardial protection by combined antegrade and retrograde blood cardioplegia and prevention of blood loss requiring massive transfusion seem to represent key issues to improve early survival in these very high-risk patients.

Limitations of the study

A few limitations may affect the results of this study and should be acknowledged. The retrospective nature of our registry is the major limitation of this analysis. Another limitation is the heterogeneity in the surgical strategy and perioperative care among the participating institutions. Finally, because of the retrospective nature of the study data, the precise aetiology of redo MV surgery was not captured in detail although some ‘surrogate’ prognostic factors, directly linked to aetiology (e.g. endocarditis and type of endocarditis, LVEF, urgent/emergent procedures etc.), were collected in the Registry and analysed accordingly. Nevertheless, the above-mentioned limitations stem from the observational and multicentre nature of the registry, which indeed provides a ‘real world’ picture of current redo mitral valve surgery in different European countries. Moreover, this is the first study analysing the potential impact on outcome of a series of surgical details directly linked to the available surgical options.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Conflict of interest: none declared.

REFERENCES

1
Jones
JM
,
O'Kane
H
,
Gladstone
DJ
,
Sarsam
MAI
,
Campalani
G
,
MacGowan
SW
et al
.
Repeat heart valve surgery: risk factors for operative mortality
.
J Thorac Cardiovasc Surg
 
2001
;
122
:
913
8
.
2
Potter
DD
,
Sundt
TM
,
Zehr
KJ
,
Dearani
JA
,
Daly
RC
,
Mullany
CJ
et al
.
Risk of repeat mitral valve replacement for failed mitral valve prostheses
.
Ann Thorac Surg
 
2004
;
78
:
67
72
.
3
Jamieson
WRE
,
Burr
LH
,
Miyagishima
RT
,
Janusz
MT
,
Fradet
GJ
,
Lichtenstein
SV
et al
.
Reoperation for bioprosthetic mitral structural failure: risk assessment
.
Circulation
 
2003
;
108
():
98
102
.
4
Ghoreishi
M
,
Dawood
M
,
Hobbs
G
,
Pasrija
C
,
Riley
P
,
Petrose
L
et al
.
Repeat sternotomy: no longer a risk factor in mitral valve surgical procedures
.
Ann Thorac Surg
 
2013
;
96
:
1358
65
.
5
Vohra
HA
,
Whistance
RN
,
Roubelakis
A
,
Burton
A
,
Barlow
CW
,
Tsang
GMK
et al
.
Outcome after redo-mitral valve replacement in adult patients: a 10-year single-centre experience
.
Interact Cardiovasc Thorac Surg
 
2012
;
14
:
575
9
.
6
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
e15
.
7
Tran
A
,
Ruel
M
,
Chan
V
.
High-risk mitral valve surgery: what is the role of risk calculators?
Curr Opin Cardiol
 
2014
;
29
:
123
6
.
8
Onorati
F
,
Biancari
F
,
De Feo
M
,
Mariscalco
G
,
Messina
A
,
Santarpino
G
et al
.
Mid-term results of aortic valve surgery in redo scenarios in the current practice: results from the multicentre European RECORD (REdo Cardiac Operation Research Database) initiative
.
Eur J Cardiothorac Surg
 
2015
;
47
:
269
80
.
9
Nashef
SA
,
Roques
F
,
Sharpless
LD
,
Nilsson
J
,
Smith
C
,
Goldstone
AR
et al
.
EuroSCORE II
.
Eur J Cardiothorac Surg
 
2012
;
41
:
734
44
.
10
Swinkels
BM
,
Plokker
HW
.
Evaluating operative mortality of cardiac surgery: first define operative mortality
.
Netherlands Heart J
 
2010
;
18
:
344
5
.
11
Thygesen
K
,
Alpert
JS
,
Jaffe
AS
,
Simoons
ML
,
Chaitman
BR
,
White
HD
et al
.
Third universal definition of myocardial infarction
.
Eur Heart J
 
2012
;
33
:
2551
67
.
12
Akins
CW
,
Miller
DC
,
Turina
MI
,
Kouchoukos
NT
,
Blackstone
EH
,
Grunkemeier
GL
et al
.
Guidelines for reporting mortality and morbidity after cardiac valve interventions
.
J Thorac Cardiovasc Surg
 
2008
;
135
:
732
8
.
13
Kappetein
AP
,
Head
SJ
,
Généreux
P
,
Piazza
N
,
van Mieghem
NM
,
Blackstone
EH
et al
.
Updated standardized endpoint definitions for transcatheter aortic valve implantation: the Valve Academic Research Consortium-2 consensus document (VARC-2)
.
Eur J Cardiothorac Surg
 
2012
;
42
:
S45
60
.
14
Emery
RW
,
Krogh
CC
,
Mc Adams
S
,
Emery
AM
,
Holter
AR
.
Long-term follow-up of patients undergoing reoperative surgery with aortic or mitral valve replacement using a St. Jude Medical prosthesis
.
J Heart Valve Dis
 
2010
;
19
:
473
84
.
15
Akay
TH
,
Gultekin
B
,
Ozkan
SI
,
Aslim
E
,
Uguz
E
,
Sezgin
A
et al
.
Mitral valve replacements in redo patients with previous mitral valve procedures: mid-term results and risk factors for survival
.
J Card Surg
 
2008
;
23
:
415
21
.
16
Ferraris
VA
,
Brown
JR
,
Despotis
GJ
,
Hammon
JW
,
Reece
TB
,
Saha
SP
et al
.
2011 update to the Society of Thoracic Surgeons and the Society of Cardiovascular Anesthesiologists blood conservation clinical practice guidelines
.
Ann Thorac Surg
 
2011
;
91
:
944
82
.
17
Khan
NU
,
Yonan
N
.
Does preoperative computed tomography reduce the risks associated with re-do cardiac surgery?
Interact Cardiovasc Thorac Surg
 
2009
;
9
:
119
23
.
18
Morales
D
,
Williams
E
,
John
R
.
Is resternotomy in cardiac surgery still a problem?
Interact Cardiovasc Thorac Surg
 
2010
;
11
:
277
86
.
19
Moscarelli
M
,
Casula
R
,
Speziale
G
,
Athanasiou
T
.
Can we use minimally invasive mitral valve surgery as a safe alternative to sternotomy in high-risk patients?
Interact Cardiovasc Thorac Surg
 
2016
;
22
:
92
6
.
20
Follis
FM
,
Pett
SB
,
Miller
KB
,
Wong
RS
,
Temes
RT
,
Wernly
JA
.
Catastrophic hemorrhage on sternal reentry: still a dreaded complication?
Ann Thorac Surg
 
1999
;
68
:
2215
9
.
21
Imran Hamid
U
,
Digney
R
,
Soo
L
,
Leung
S
,
Graham
AN
.
Incidence and outcome of re-entry injury in redo cardiac surgery: benefits of preoperative planning
.
Eur J Cardoothorac Surg
 
2015
;
47
:
819
23
.
22
Zeng
J
,
He
W
,
Qu
Z
,
Tang
Y
,
Zhou
Q
,
Zhang
B
.
Cold versus crystalloid cardioplegia for myocardial protection in adult cardiac surgery: a meta-analysis of randomized controlled studies
.
J Cardiothorac Vasc Anesth
 
2014
;
28
:
674
81
.
23
Guru
V
,
Omura
J
,
Alghamdi
AA
,
Weisel
R
,
Fremes
SE
.
Is blood superior to crystalloid cardioplegia? A meta-analysis of randomized clinical trials
.
Circulation
 
2006
;
114
:
I-331
I-338
.
24
La Par
DJ
,
Ailawadi
G
,
Harris
DA
,
Hajzus
VA
,
Lau
CL
,
Kern
JA
et al
.
A protocol-driven approach to cardiac reoperation reduces mortality and cardiac injury at the time of resternotomy
.
Ann Thorac Surg
 
2013
;
96
:
865
70
.
25
Park
CB
,
Suri
RM
,
Burkhart
HM
,
Greason
KL
,
Dearani
GL
,
Schaff
HV
et al
.
What is the optimal myocardial protection strategy at reoperation for aortic valve replacement in the presence of a patent internal thoracic artery?
Eur J Cardiothorac Surg
 
2011
;
39
:
861
5
.

APPENDIX. CONFERENCE DISCUSSION

Dr J. Hlubocky (Prague, Czech Republic) : It is a very risky group of patients, mean age was about 65 years, and that is why it was surprising to me that in the subgroup of patients with replaced valve you had I think 60% of mechanical prosthesis. I would expect a slightly more biological solutions in that group of patients.

My second question, you nicely demonstrated that iatrogenic disasters are an influencing impact on the patients. So do you have any recommendations for us for redo procedures but maybe also for primary procedures to avoid those problems?

Dr Onorati: The answer to the first question is related to the institutional policies of the different centres involved in this registry. I take your point, that this is a relatively young patient population, with a significant high percentage of mechanical valves, but it reflects the fact that some institutions preferentially use mechanical valves and others preferentially use biological valves. It is also quite surprising, in my opinion, the low rate of repair and re-repair but this is possibly related to the fact that some of the patients had such a low surgical threshold for replacement can be foreseen. Therefore, this reflects the need to go out from the theatre in a high risk population of patients, with a good result, at least in terms of mitral valve function.

About the second question, I think that none of us has the correct strategy. In our institution we routinely use preoperative CT scan, which has been demonstrated to reduce iatrogenic complications, but I can say in our institution too, iatrogenic lesions of the cardiac structure sometimes occur. However the use of a preoperative CT scan was a factor analysed in the multivariate analysis, but it was not reported to be associated to a reduction of risks for iatrogenic lesions. It is possible that most of these procedures were conducted by experienced surgeons, therefore experience may overcome the suggestions that opposite to what happens to routine CT scan may add less- experienced surgeons.

Dr T. Hanke (Luebeck, Germany) : Do you have any thoughts on the use of femoral cannulation for this group of patients?

Dr Onorati: Yes, we also considered this factor in the analysis, and again, to my surprise this was not associated with any reduction of iatrogenic complications. I personally always prepare femoral vessels, but it was not a protective factor in this experience.

Dr F. Van Praet (Aalst, Belgium): About the access, it is true. Out of the 2,850 endoscopic mitral valves we have done, 350 were redos and I think the endoscopic approach of redo mitral valve surgery is a major advantage. The logistic Euro SCORE of our series is around 20; it is very high risk. Some of them or not score able because they come from everywhere, are not accepted everywhere, and we give the patients a chance. So in fact you cannot score. I know that the logistic Euro SCORE tends to over score. I sometimes have the feeling I cannot score them.

The mortality is 6.5% in this series. Obviously we don't have entry problems except for pleural adhesions, which you cannot always predict. The majority of the adhesions you can free them easily, and I am talking about a very small incision, no rib spreading; you can hardly put your fingers in. Of course, in a very old lady when you ruin the lung, then the advantage is gone.

The issue of thromboembolic phenomena, in the 2,850 cases, redos included, early learning curve included, and the thromboembolic rate is 1.3%, which is very low. In the redos it is a little bit higher; it is below 2%. So those are really not arguments against the technique.

I personally think that the way to go and if you want to avoid technical issues by re-entry, I was amazed by the odds ratio you showed over there, a mini thoracotomy is, in my opinion, the way to go.

Dr Onorati : First of all, this is, I would say, a real world series where minimally invasive mitral valve surgery is not a huge practice. We have some patients with thoracotomy access to the mitral valve. Obviously you did not face with major cardiovascular lesions. However this technique was also entered in the multivariate analysis and was not a protective factor.

Also the last thing, 46% of this population were combined procedures, so also aortic, tricuspid etc. Maybe a very skilled surgeon can face some of these problems by minimally invasive surgery, but I am not sure that all these patients can be treated with a minimally invasive approach. But I take your point.

Dr C. Huber (Bern, Switzerland) : So the message here might be that these redo mitrals should go into specialized centres as well, because 24.4% is quite a high ratio as a risk increase.

Dr Van Praet: Of course, I forgot to say, that to do those redo procedures you need experience with the Endo -Clamp, the Intra- Clude balloon, which is under a lot of discussion. You need a lot of experience with the Endo-Clamp or the Intra-Clude balloon, the intra-aortic occlusion balloon, in order to master those difficult redo cases.

Dr Onorati: I am sure that this kind of access might reduce major cardiovascular lesions to 0%, but I think the majority of our colleagues do these kinds of procedures by redo sternotomy worldwide.

Dr M. Antunes (Coimbra, Portugal): One of the main technical problems is, dealing with a patent IMA, and we have found that, especially when we do aortic redos after coronary surgery we do not worry about the inferior mesenteric artery; we leave it perfusing the myocardium normally. We will just perfuse cardioplegia in the aortic root or the vein grafts that we might have to redo. Somehow, it seems as if there is an equilibrium, provided that you repeat the cardioplegia a little bit more frequently and that, if possible, you occlude the left main orifice. We have had excellent results using this technical modification.

Dr Onorati: I agree with you and I think that one of the explanations for the protective role of ante-grade plus retrograde cardioplegia may be related to the fact that it can perfuse also a patent arterial graft pushing blood into the arrested myocardium, therefore, it is possible, the association with retrograde delivery counterbalance the left anterior descending artery graft perfused and protect better the myocardium. Indeed I don't have the definite answer.

Dr Huber: So stay away from the re-operative field as much as you can. I do the same. I would like to make a little vote in the room. In this situation, who is controlling the mammary, hands up, who is trying to clamp the mammary in any way, hands up?

(Show of hands).

Dr Huber: So there are nine people. And who is not touching the mammary whenever possible?

(Show of hands).

Dr Huber: It is the majority.

Dr Livesey: Also is anybody using crystalloid cardioplegia?

(Show of hands).

Dr T. Hanke (Luebeck, Germany): One question. With this rather high complication rate and a mortality rate of almost 13%, wouldn't it be better in those patients to advocate for a trans-apical approach, because 13%, we have to discuss with the cardiologists now these patients and remembering all these trials. When you present those complication rates, they are rather unwilling, at least in Germany, to give you the patients for a redo procedure.

Dr Onorati: This can be an option. We also go inside the database, and we are now doing another study aimed at analysing only those patients potentially candidates to transcatheter aortic valve replacement at the time of index procedure in order to see any differences and maybe rule out tips and tricks. But this study is not already completed, so I cannot tell you the results.

But one thing I can say, on that study, is that these two populations of patients, seem to be quite different in terms of risk-profile which seems to be better in candidates to traditional surgery than those potentially candidates to transcatheter aortic valve replacement Moreover not all redo mitral patients can be treated by transcatheter aortic valve replacement.

Author notes

Presented at the 29th Annual Meeting of the European Association for Cardio-Thoracic Surgery, Amsterdam, Netherlands, 3–7 October 2015.