Contemporary Mitral Valve Surgery for Septuagenarians and Octogenarians

Received: December 17, 2016 Revised: January 12, 2017 Accepted: January 20, 2017 Abstract: Background: The increasing prevalence of degenerative and functional mitral valve disease as the population ages alongside introduction of percutaneous mitral valve interventions mandates revision of outcomes of mitral valve surgery (MVS) in elderly and high risk patients. We compared the characteristics and outcomes of octogenarians and septuagenarians undergoing MVS.


INTRODUCTION
The world's first successful heart valve surgery was performed in 1923 on a 12 year-old girl with rheumatic mitral stenosis, and twenty years later, John Gibbon performed the first successful open heart surgery using cardiopulmonary bypass, often considered the most important advancement in heart valve surgery [1,2].Further refinements of surgical techniques, recommended for treating severe symptomatic valvular heart disease, have since been developed to meet demands of the increasing prevalence of degenerative and functional valve disease for the aging population [3 -5].Mitral valve disease is common at over 10% in those over 75 years of age [6,7].Age however is an established risk factor for cardiac surgery, and part of all conventional mortality risk models to help select favourable surgical candidates [8 -10].Due to the perceived high risk, many of these elderly patients remain untreated compared to their younger cohort [6,7].The decision for intervention has also recently become more complex due to the introduction of percutaneous mitral valve techniques such as the mitraclip [11].Reviewing the outcomes of mitral valve surgery (MVS) in high risk groups such as the elderly is therefore of great importance to guide clinical practice.The aim of this study is to review and compare the characteristics and outcomes of octogenarians and septuagenarians undergoing MVS.

MATERIALS AND METHODS
All patients over the age of 70 years undergoing isolated MVS at Auckland City Hospital during the period 2005-2012 were included, and grouped into 70-79 and >80 years of age for analysis.Study participants were identified retrospectively from the hospital's cardiothoracic surgical unit database.Baseline characteristics, operative variables and post-operative outcomes were retrieved from the hospital's electronic clinical records, with ethics approval attained before the study commencement from the Auckland District Health Board Research Office.
For the variables included in the study, angina class was defined according to the Canadian Cardiovascular Society grading and dyspnea class by the New York Heart Association.Other baseline characteristics followed the definitions of EuroSCORE II parameters [9].EuroSCORE [8], EuroSCORE II [9] and Society of Thoracic Surgeon's (STS) Scores [10] were calculated for all patients.
Mortality data was obtained from the "Births, Deaths and Marriages registers" of New Zealand up till December 2015.Operative mortality was defined as in-hospital deaths or deaths within 30 days of surgery.The five post-operative complications (stroke, renal failure, ventilation >24hrs, mediastinitis and return to theatre) and their composite were defined as per the STS database [10].Survival rates at one, three and five-years were recorded.
Median (lower-upper quartile) and percentages (frequency) are used to present continuous and categorical variables respectively.Univariate analysis was performed using Mann-Whitney U Test and Fisher's Exact Test for continuous and categorical variables respectively.Longitudinal survival was analysed by Kaplan-Meier curves and log-rank (Mantel-Cox) test.Area under the receiver operative characteristics curves (c-statistics) with 95% confidence intervals (95%CI) were used to assess the discriminative ability of risk scores.Variables with P<0.20 in univariate analysis were selected for multivariate analysis to identify independent predictors of various surgical outcomes.Logistic regression and Cox Proportional Hazards Regression were used for cross-sectional and longitudinal outcomes respectively.All tests were two-tailed and P<0.05 considered statistically significant.SPSS(Version 17.0, SPSS Inc., Chicago, IL, USA) and Prism (Version 5, GraphPad Software, San Diego, CA, USA) were used for statistical analyses.

RESULTS
A total of 101 patients who received isolated MVS at Auckland City Hospital from 2005-2012 were included in the study.Out of the 101 patients, 20 patients were part of the octogenarian group at time of surgery, and the remaining 81 in the septuagenarian group.Baseline characteristics are listed in Table 1.There were no significant differences in preoperative characteristics between the two groups apart from age (median age 82 years for octogenarians and 74 years for septuagenarians, P<0.001) and previous stroke (15% in octogenarians and 1.2% septuagenarians, P=0.024).Significantly higher median scores were found in the octogenarians compared to the septuagenarians for two of three risk models: EuroSCORE(11.4% and 7.0% respectively, P=0.01) and STS score (5.6% and 2.8% respectively, P=0.02).

Liao et al.
The operative variables and in-hospital outcomes for each group are demonstrated in Table 2.There were no differences observed for operative variables measured, with approximately half of patients having mitral repair and replacement for both groups.There were also no differences in operative mortality and all complications (stroke, renal failure, ventilation >24 hours, mediastinitis and return to theatre) and their composite between the two groups.Notably none of the patients died in-hospital or within 30-days in the octogenarian group, and 6 (7.4%) died in septuagenarian group although P-value was not statistically significant at 0.596.   1) illustrates the survival during a mean follow-up of 5.1+/-2.6 years for both groups.One, three and five-year survivals were 100.0%,80.0% and 70.0% for octogenarians and 91.4%, 86.4% and 73.8% septuagenarians.There was no statistically significant difference between the two groups (P=0.986).

Survival rate (%)
Table 3 shows predictors of post-operative mortality and complications in multivariate analysis for the study cohort.Cardiopulmonary bypass time and age were independent predictors of operative and long-term mortality respectively; history of congestive heart failure was a predictor of composite morbidity and history of previous cardiac surgery was a predictor of return to theatre.Results of the discriminant analysis for risk scores at predicting outcomes are presented in Table 4. None of the risk scores were able to predict mortality and most morbidities.However, EuroSCORE could predict renal failure (AUC 0.742, 95% CI 0.576-0.906)and mediastinitis (EuroSCORE AUC 0.717, 95% CI 0.598-0.837);EuroSCORE II predicted stroke (AUC 0.722, 95%CI 0.632-0.812)and mediastinitis (AUC 0.818, 95%CI 0.742-0.894);and STS Score predicted composite morbidity (AUC 0.649, 95% CI 0.530-0.769),stroke (AUC 0.742, 95%CI 0.641-0.843)and mediastinitis (AUC 0.687, 95% CI 0.512-0.86).

DISCUSSION
Mitral valve disease is the commonest form of valvular heart disease, and along with our aging population, has led to a growing demand for mitral valve intervention with many patients remaining untreated [6,7].This unmet clinical need is in part due to subgroups deemed at high surgical risk, and warranting the evaluation of surgery in such patients.Our study reported favourable outcomes for isolated mitral valve surgery in carefully selected elderly patients, and these findings alongside other studies in the era of percutaneous valvular interventions have many clinical implications.
Post-operative outcomes between the two age groups were not significantly different statistically, though numerically higher operative mortality in septuagenarians compared to 0% in octogenarians.Baseline characteristics were well matched except higher prevalence of previous stroke in octogenarians.Therefore the main contributor to the higher EuroSCORE and STS Score in octogenarians was predominantly the age difference.This was a similar finding to our previous study of aortic valve replacement in elderly patients [12], and perhaps reflect a more cautious and conservative approach in selecting octogenarians for cardiac surgery compared to younger counterparts [13].However,

Table 3 . Multivariate analysis (for predictors P<0.10).
All ratios are odds ratios, except hazards ratios for predictors of mortality during follow-up.