乌鸦传媒

Patients with prosthetic valve endocarditis (PVE) continue to have a high mortality risk that early surgical intervention does not reduce, results of a large multinational trial showed.

An unadjusted analysis controlling only for treatment selection suggested a large benefit from early surgery. A fully adjusted analysis, however, showed no benefit of surgery versus medical therapy for in-hospital (HR 0.90) or 1-year mortality (HR 1.04), according to Tahaniyat Lalani, MD, of Naval Medical Center Portsmouth in Virginia, and co-authors.

A propensity analysis that examined the probability of surgery showed that patients in the top surgical quintile had significantly lower in-hospital and 1-year mortality, and patients in the fourth surgical quintile had lower 1-year mortality as compared with similar patients treated medically, they reported online in JAMA Internal Medicine.

"Surgical treatment was not associated with a lower mortality at 1 year in the overall PVE cohort after controlling for treatment selection and survivor bias," the authors concluded. "Further studies are needed to define the effect and timing of surgery in patients with PVE who have indications for surgery."

As many as 6% of patients develop PVE within 5 years of valve implantation, and the infection is associated with significant morbidity and mortality. "Approximately one-third of patients with PVE die within 1 year after diagnosis, with mortality strongly associated with other chronic illness, healthcare-associated infection, staphylococcus aureus, and complications of PVE," the authors noted.

Authors of consensus guidelines recommend prompt surgical intervention with debridement and valve replacement for patients with complications. However, current guidelines are based in large part on expert opinion and observational data, Lalani and co-authors noted. Moreover, studies comparing outcomes, including mortality, after surgery and medical therapy for PVE have yielded conflicting results.

Surgery for PVE has not been evaluated in a prospective, randomized clinical trial. Propensity score methodology affords a mechanism to compare results with different therapies by providing an estimated probability that a patient will receive a treatment on the basis of baseline characteristics.

Recent recommendations about the application of propensity-score methods support the method of inverse probability of treatment weighting (IPTW) using the surgical propensity score in regression models for mortality because of the method's superior performance in controlling for selection bias, the authors continued. Survivor bias should be addressed by matching or by evaluation of treatment as a time-dependent covariate.

Lalani and colleagues applied propensity-score methods to a comparison of in-hospital and 1-year mortality in patients with PVE treated surgically or medically. The patients were included in the International Collaboration on Endocarditis-Prospective Cohort Study, a prospective, multinational observational cohort study of infective endocarditis.

Of the 4,166 patients in the study, 1,025 developed PVE. The analysis comprised 719 (70.1%) patients with a prosthetic aortic valve (mechanical valve in 349 cases, bioprosthetic in 353, and repair in 17), and 473 (46.1%) patients with a prosthetic mitral valve or ring (mechanical in 303 cases, bioprosthetic in 86, and repair in 84).

The authors found that 490 (47.8%) patients had early surgery and 535 (52.2%) received only medical therapy during hospitalization. Early surgery was associated with younger age, shorter symptom duration, and transfer from another facility. PVE caused by S. aureus or enterococci was associated with medical therapy, whereas coagulase-negative Staphylococcus was associated with increased use of surgery.

Surgery was associated with an increased frequency of complications, including mitral valve regurgitation (28.8% versus 19.6%), paravalvular complications (43.5% versus 20.2%), and prosthetic valve complications (41.6% versus 24.1%).

An unadjusted Cox proportional hazards model showed that early surgery was associated with lower in-hospital mortality (22.0% versus 26.7%, HR 0.68, P=0.003) and 1-year mortality (27.1% versus 36.6%, HR 0.68, P<0.001). A model that controlled for selection bias showed an even greater advantage for early surgery (in-hospital mortality: HR 0.44, P<0.001; 1-year mortality: HR 0.57, P<0.001).

The advantage of early surgery disappeared in a multivariable model with IPTW, as neither in-hospital nor 1-year mortality differed significantly between patients treated surgically or medically.

Propensity analysis by patient quintiles showed that surgery retained an advantage for in-hospital mortality (P=0.03) and 1-year mortality (P=0.007), and surgery also showed an advantage for patients in the fourth quintile for 1-year mortality (P=0.007).

From the American Heart Association:

Comment