A Meta-Analysis of Early versus Delayed Surgery for Valvular Infective Endocarditis Complicated by Embolic Ischemic Stroke

OBJECTIVE: An embolic ischemic stroke occurs in 10% to 40% of patients with valvular infective endocarditis (IE) and confers significant morbidity. The optimal timing of valve surgery in this population is not well defined. METHODS: With the use of PubMed, EMBASE, Ovid, and Cochrane databases, a systematic review identified 14 studies through October 2015 that compared early versus delayed surgery for valvular IE complicated by an ischemic stroke. Early surgery was defined as 3 days or less in one, 7 days or less in eight, and 14 days or less in five studies. Risk ratios (RRs) were calculated by the Mantel-Haenszel method under a fixed- or random-effects model, for the outcomes of perioperative stroke, operative mortality, and 1-year survival. RESULTS: A total of 833 patients (early surgery, 330; delayed surgery, 503) were included. The majority of operations were for aortic and/or mitral valve IE, with prosthetic valve IE present in 0% to 60%. Infection with Staphylococcus aureus ranged from 19% to 66%, and heart failure prevalence at the time of operation was 24% to 66%. Early surgery was associated with an increased risk of operative mortality (RR, 1.72; 95% confidence interval [CI], 1.27-2.34; P = 0.0005), which was significant regardless of surgery within the first 7 days (RR, 2.19; 95% CI, 1.45-3.31; P = 0.0002) or 14 days (RR, 1.72; 95% CI, 1.12-2.64; P = 0.01) after stroke. Surgical timing did not affect the risk of perioperative ischemic or hemorrhagic stroke or 1-year survival. CONCLUSIONS: In patients with valvular IE complicated by ischemic stroke, early surgery is associated with an increased risk of operative mortality, with no observed benefit in 1-year survival.