Objective: To evaluate postoperative outcome of emergency surgery for acute severe mitral regurgitation (ASMR) from a multicentre experience.
Methods: In six centres, 279 patients (mean age 62+/-14 years, 62% female) undergoing emergency surgery for ASMR from December 1986 to March 2007 were analysed and followed up. Aetiology included acute myocardial infarction (AMI) in 126 patients (group 1, 45%), degenerative mitral valve disease in 74 (group 2, 26%), and acute endocarditis (AE) in 79 (group 3, 28%). Preoperatively, all patients were in haemodynamic instability, with 185 patients in cardiogenic shock (66%), 184 (66%) intubated, and 61 (22%) on IABP, respectively. Valve repair was performed in 76 (27%), whereas 203 (73%) underwent valve replacement. Median follow-up (98% complete) was 70.8 months (inter-quartile range 59.8-86.66 months).
Results: Overall 30-day mortality was 22.5% (63/279). Early death was significantly lower in group 2 (p<0.001 and p=0.005 vs group 1 and 3, respectively) whereas no difference was detected between group 1 and 3. At logistic regression analysis AMI, AE, shock, left ventricular dysfunction, and coronary artery disease were predictors of early death. Overall 15-year survival was 67+/-10%. Survival was lower in group 1 (39+/-11%) than in group 2 (75+/-9%) and group 3 (77+/-10%). Cox regression found AMI, and associated coronary artery disease to be predictors of late death. Overall 15-year actuarial and actual freedom from cardiac-related events were 44+/-9% and 28+/-10%, respectively, with the worst outcome in the presence of AE. Associated coronary artery disease, AE, AMI, preoperative atrial fibrillation, and chronic renal failure were independent predictors of cardiac-related events.
Conclusions: Emergency surgery for ASMR remains a surgical challenge for high incidence of early and late cardiac-related events, particularly in patients with associated coronary artery disease and acute endocarditis. Apparently, type of mitral valve surgical approaches (repair or replacement) did not provide any influence on postoperative outcome.