Thrombolytic therapy or surgery for valve prosthesis thrombosis: systematic review and meta-analysis.

BACKGROUND Prosthetic heart valve thrombosis is a life-threatening condition for which treatment has not been evaluated by randomized trials. OBJECTIVES To compare surgery with thrombolytic therapy for the treatment of prosthetic valve thrombosis. METHODS A comprehensive systematic review was carried out by independent researchers using PubMed, Web of Knowledge, HINARI, LILACS and EMBASE, including papers indexed up to 23 October 2013 with at least 10 patients, evaluating one or both treatment options. Outcomes of interest were death, stroke, bleeding, embolic events and success. Study quality was assessed by the Newcastle-Ottawa Scale (ranges from 0 to 9). Comprehensive Meta Analysis Software was used to pool study results, for publication bias analysis and for meta-regression. RESULTS Forty-eight studies were included (2302 patients). No randomized study was identified, and all were observational in design. Study quality varied from 4 to 6 stars using the Newcastle-Ottawa Scale. Mortality for surgery was 18.1% (95%CI, 14.6-22.1; I(2) , 62% [42-75]), while mortality for thrombolytic therapy was 6.6% (95% CI, 4.8-8.9; I(2) , 0% [0-13]). Surgical mortality appeared to increase with NYHA-class, according to meta-regression. Other results related to surgery and thrombolytic therapy, respectively, were: embolic events, 4.6% (2.9-7.3) and 12.8% (10.8-15.2); stroke, 4.3%,(2.7-6.6%) and 5.6%,(4.3-7.4%); success rate, 81.9% (77.2-85.8) and 80.7% (75.6-85.0); bleeding, 4.6% (2.9-7.1) and 6.8% (5.4-8.6); and death or stroke, 19.0% (14.8-24.2) and 11.4% (8.7-14.7). CONCLUSION Mortality in patients treated by thrombolytic therapy for valve prosthesis thrombosis is significantly lower than in patients treated surgically. As we cannot yet ascertain whether this difference is due to the treatment alone, more studies are now necessary to further clarify these findings.