Effectiveness and safety of drotrecogin alfa (activated) for severe sepsis: a meta-analysis and metaregression.

BACKGROUND Drotrecogin alfa (activated) was approved for use in severe sepsis in 2001 on the basis of the Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis (PROWESS) trial, but controversies about its effectiveness remain. We aimed to assess effectiveness and safety of use of this drug in the past 10 years and compare them with the original PROWESS results. METHODS We searched PubMed, Embase, Ovid, Cochrane Library, Evidence-Based Medicine, and the American College of Physicians Journal Club databases for experimental and analytical studies of drotrecogin alfa (activated) in adults with severe sepsis until Jan 31, 2012. We calculated adjusted risk ratios for effectiveness and safety outcomes with random-effects models. We did a metaregression to assess the effect of severity of illness on the risk of death and the risk of bleeding associated with drotrecogin alfa (activated). FINDINGS We included nine controlled trials (41,401 patients) and 16 single-group studies (5822 patients) in effectiveness analyses and 20 studies (8245 patients) in safety analyses. Hospital mortality was reduced by 18% with drotrecogin alfa (activated) compared with controls (relative risk 0·822, 95% CI 0·779-0·867; p<0·0001; I(2)=40%). This mortality reduction was much the same as was noted in PROWESS (0·851, 0·740-0·979), but smaller than that of patients in PROWESS with high disease severity (0·708, 0·590-0·849). Propensity-adjusted studies also showed a significant mortality reduction with lower heterogeneity (0·844, 0·800-0·891; p<0·0001, I(2)=18%). These findings were not changed by the addition of PROWESS-SHOCK results. Metaregression showed greater benefits of drotrecogin alfa (activated) with increasing control mortality (p=0·01) and more severe disease (p=0·04). Hospital mortality for single-group studies of drotrecogin alfa (activated) was 41% (95% CI 35-48), and was higher than that noted in PROWESS at 31% (27-36; p<0·0001). The serious bleeding rate with drotrecogin alfa (activated) was 5·6% (4·5-6·9), which was higher than the 3·5% (2·5-5·0) noted in PROWESS (p=0·003), but similar to that reported in PROWESS high disease severity (p=0·073). INTERPRETATION Real-life use of drotrecrogin alfa (activated) was associated with significant reduction in hospital mortality and increased rates of bleeding in patients with severe sepsis. Our effectiveness findings were in line with the PROWESS trial but not with the PROWESS-SHOCK trial. FUNDING None.