Partial Upper Sternotomy versus full Sternotomy for Mitral Valve Surgery: A Propensity Score Matched Analysis

Objectives: Minimally invasive mitral valve surgery (MIMVS) through anterolateral minithoracotomy (MT) has become the standard therapy for isolated mitral valve disease in experienced centers. Multiple valve disease or other anatomical and certain clinical conditions, however, make this access not suitable for some patients and conventional full sternotomy (FS) is the mostly preferred alternative approach. For those patients, partial upper sternotomy (PS) can be used as a less invasive access. Although FS has been widely investigated, there are not enough insights to the PS approach for mitral valve surgery (MVS). Therefore, we compared the data of both accesses.

Methods: This retrospective analysis includes data on 1,639 patients who underwent either isolated or combined primary MVS at our department from May 2011 to August 2017. Out of these, 663 patients were operated via MT access and were excluded from this analysis. Further 528 patients were excluded mainly due to concomitant coronary artery bypass (CABG) surgery but also because of re-do cases, concomitant aortic surgery or urgent/salvage MVS. Finally, 99 patients who had been judged suitable for PS had either isolated MVS (n = 47, 47.5%) or multivalve surgery (n = 52, 52.5%). In addition, 349 patients with FS for primary MVS were included in our study. To reduce the possibility of selection bias a 1:1 propensity score match making was performed which resulted in 98 pairs. For analysis of postoperative survival Kaplan–Meier curves were calculated.

Results: During a median follow-up time of 1,491 days (478–2,186; PS 1,103 [331–1,806 days], FS 2,180 days [841–3,054]) all-cause mortality was 15.90% (70 of 439 patients). In the propensity score paired model, PS showed statistically significant superior survival compared with FS at 30 days (p = 0.044, hazard ratio [HR] 7.525, 95% confidence interval [CI] 1.06–53.56). Furthermore, 90 and 365 days survival after surgery showed a similar trend, but without reaching statistical significance (p = 0.096, HR 3.898, 95% CI 0.79–19.34; p = 0.077, HR 2.774, 95% CI 0.89–8.61). As secondary end points, number of second pump runs and hospital length of stay were significantly less (p = 0.016, p < 0.001) in PS patients.

Conclusion: The less invasive PS approach for MVS seems to have short- and long-term survival benefits. In patients who are not candidates for MT PS seems a favorable approach although prospective randomized-controlled trials are necessary for confirmation.

No conflict of interest has been declared by the author(s).