Introduction: Drotrecogin alfa (activated; DrotAA) treatment, a 96-hour infusion, reduces 28-day mortality in severe sepsis to approximately 25%. The question remains whether a longer infusion or higher dose could increase rate of survival. The goal of this study was to identify a dependable, sensitive measure with which to monitor disease progression and response in patients during DrotAA treatment.
Methods: Data on severe sepsis patients included in PROWESS (placebo-controlled, double-blind, randomized study of 850 DrotAA and 840 placebo individuals) and ENHANCE (single-arm, open-label study of 2,375 DrotAA patients) studies were analyzed. In these studies, DrotAA (24 mug/kg per hour) or placebo was infused for 96 hours and patients were followed for 28 days. Data on six laboratory measures and five organ dysfunctions were systematically analyzed to identify a potential surrogate end-point for monitoring DrotAA therapy and predicting 28-day mortality at the end of therapy. To allow comparison across variables, sensitivity and specificity analyses identified cut-off values for preferred outcome, and relative risks for being above or below cut-offs were calculated, as was the 'proportion of treatment effect explained' (PTEE) to identify biomarkers that contribute to benefit from DrotAA.
Results: Protein C was the only variable that correlated with outcome across all analyses. Using placebo data, a baseline protein C under 40% was established as a useful predictor of outcome (odds ratio 2.12). Similar odds ratios were associated with cut-off values of other biomarkers, but the treatment benefit associated with DrotAA was significantly greater below the cut-off than above the cut-off only for protein C (relative risk for 28-day mortality 0.66 versus 0.88; P = 0.04). Protein C was the only end-of-infusion biomarker that potentially explained at least 50% of the benefit from DrotAA (PTEE 57.2%). The PTEE was 41% for cardiovascular Sequential Organ Failure Assessment score and for d-dimer. At the end of infusion (day 4), protein C categories (</=40%, 41% to 80%, and > 80%) remained significantly related to mortality, regardless of treatment assignment.
Conclusion: Based on systematic analyses of 11 variables measured in severe sepsis clinical trials, protein C was the only variable consistently correlated with both DrotAA treatment effect and survival. Further study is needed to determine whether longer infusions or higher doses of DrotAA would achieve the goal of normalizing protein C in more patients with severe sepsis.