A Randomized Trial of Early Surgery for Infective Endocarditis

Abstract & Commentary

By Michael H. Crawford, MD, Editor

Sources: Kang DH, et al. Early surgery versus conventional treatment for infective endocarditis. N Engl J Med 2012;366:2466-2473. Gordon SM, Pettersson GB. Native-valve infective endocarditis — When does it require surgery? N Engl J Med 2012;366:2519-2521.

Early surgery for infective endocarditis (IE) has been supported by several observational studies, but the lack of randomized, controlled data has led to conflicting recommendations by major societies. Thus, this group of investigators from Korea conduced a randomized, controlled trial in 76 patients with native left heart valve IE, as defined by the modified Duke criteria, who had severe valve disease and large (> 10 mm) vegetations, and were potential candidates for early surgery. Patients with definite indications for urgent surgery were excluded: heart failure, heart block, annular abscess, fistulas, or fungal infection. Also excluded were patients with major contraindications to surgery (e.g., stroke). All patients were evaluated by transesophageal echocardiography and CT angiography of the brain and abdomen. Patients assigned to surgery were required to have it within 48 hours of randomization. Those in the medical therapy group could have surgery if complications requiring surgery developed. The primary end-point was hospital deaths or clinical embolic events within 6 weeks of randomization. Results: All of the early surgery group underwent surgery within 48 hours, and 30 of the 39 (77%) medical group eventually underwent surgery, most during the initial hospitalization (90%). The primary endpoint occurred in one patient (3%) in the early surgery group, compared to nine (23%) in the conventional treatment group (hazard ratio, 0.10, 95% confidence interval, 0.01-0.82, P = 0.03). One early surgery patient died and almost all of the conventional group patients who had an endpoint had clinical emboli (eight of nine). All the endpoints in the conventional group occurred before any crossover to surgery. The authors concluded that early surgery in patients with large vegetations decreases the risk of systemic embolism.


This is an important study because it is the first randomized trial of surgery in IE. Its focus is on the patients with moderate-to-severe regurgitation and large vegetations (> 10 mm) who are not in heart failure and don’t have any of the other class I indications for surgery such as an abscess, fistula, or heart block. Accordingly, it took almost 5 years for them to enroll 76 patients. However, this is one of the more controversial areas in IE management where observational studies are conflicting. Some, but not all, studies support early surgery for vegetations > 10 mm and if severe valve regurgitation is present. The center of the controversy is the risk of a catastrophic complication that would preclude surgery, such as an embolic stroke vs several days of antibiotics to stabilize the patient and reduce the likelihood of surgical complications or reinfection of the prosthetic valve. The risk of emboli is highest the first week after diagnosis, but drops rapidly after 10-14 days of effective antibiotic therapy. Hence, the focus of this study on early surgery to reduce the early embolic risk vs waiting until surgery was necessary due to complications of IE, with the hope that some could avoid surgery.

The study clearly shows that early surgery prevents clinical thromboembolic events without increasing mortality. Both early and 6-month mortality were not significantly different between the two groups. However, there were so few deaths that the trial was not powered to definitively answer this question. Also, the incidence of recurrent IE was very low and not different between the two groups. As expected, more than three quarters of the conservative treatment group eventually required surgery; almost all of which occurred during the initial hospitalization. So, in some ways, this was a study of surgical timing and it doesn’t appear that waiting gains much. Even the quarter of the conservative group who had not had surgery by 6 months had severe valve regurgitation and it could be argued would eventually need surgery. Among the eight patients with emboli in the conservative group, five had strokes with residual defects.

Since the two most common causes of death in IE are heart failure and thromboemboli, it makes sense to operate early to prevent these complications. Current opinion and guideline recommendations are to operate early for heart failure with IE, despite a paucity of data. It appears that based on this randomized trial, moderate-to-severe valve regurgitation with large vegetations can be confidently added to this list.