An updated meta-analysis to understand the variable efficacy of drotrecogin alfa (activated) in severe sepsis and septic shock

Abstract

Background: Significant debate continues over the efficacy of drotrecogin alpha activated (DAA) in sepsis. This updated meta-analysis provides an updated summary effect estimate and explores the reasons for outcome heterogeneity in placebo-controlled randomized clinical trials of DAA on 28-day all-cause mortality in patients with severe sepsis or septic shock.

Methods: Computer searches of MEDLINE, EMBASE, the Cochrane Library, ClinicalTrials.gov, published abstracts from major intensive care meetings and examination of reference lists were used to identify five placebo-controlled randomized clinical trials with 7260 patients. The primary endpoint was 28-day all-cause mortality. Secondary outcomes were 28-day incidence of severe bleeding and intracranial hemorrhage.

Results: DAA was not associated with improved 28-day all-cause mortality in patients with severe sepsis or septic shock (pooled relative risk (RR) of 0.97 [95% CI 0.83-1.14]), and is associated with an increase in serious bleeding. The significant heterogeneity in the pooled RR for 28-day mortality (I2 value of 59.4%, χ2 P-value 0.043) is no longer present with exclusion of the post-study amendment portion of PROWESS (I2 value of 0%, χ2 P-value 0.44 without PROWESS post-amendment). Using meta-regression, the best ranked predictor of outcome heterogeneity was baseline mortality in the placebo arm, which was among the highest in PROWESS.

Conclusion: DAA is not associated with improved survival in patients with severe sepsis or septic shock. Further studies should be done to determine whether changes in supportive therapy for sepsis explain the variable efficacy of DAA in randomized controlled clinical trials observed over time.

Figure 1

Flow diagram of literature search and study selection for inclusion in meta-analysis

Figure 2

Forest plot comparing the effect of DAA vs. placebo on risk ratio (RR) for 28-day all-cause mortality. The pooled RR of 0.97, 95% CI 0.83-1.14 demonstrates that DAA had no effect on 28-day all-cause mortality. CI, confidence interval.

Figure 3

Forest plot comparing the effect of DAA vs. placebo on risk ratio (RR) for severe bleeding. The pooled RR of 1.48, 95% CI 1.10-1.99) demonstrates that there is an association between DAA and severe bleeding events. CI, confidence interval.

Figure 4

Forest plot comparing the effect of DAA vs. placebo on risk ratio (RR) for intracranial hemorrhage. The pooled RR of 1.52, 95% CI 0.78-2.99 demonstrates that there is no statistically significant association between DAA and intracranial hemorrhage. CI, confidence interval.

Figure 5

Forest plot comparing the effect of DAA vs. placebo on risk ratio (RR) for 28-day all-cause mortality. Here, PROWESS is divided into pre- and post- amendment periods. Only PROWESS post-amendment demonstrates a protective effect of DAA on 28-day all-cause mortality, with pooled RR 0.71, 95% CI 0.57-0.87. CI, confidence interval.

Figure 6

Galbraith plot plotting the standardized effect estimate for DAA vs. placebo on 28-day mortality in each study against the standard error of each study. Here the PROWESS post-amendment period falls outside the 95% confidence interval and is a clear outlier.