Leaders in Pharmaceutical Business Intelligence (LPBI) Group

Ranieri et al. (May 31 issue)1 report the results of yet another negative trial involving patients with septic shock. I cannot help but wonder whether these negative trials reflect a failure of the study design rather than the study drug. The Prospective Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis and Septic Shock (PROWESS-SHOCK) protocol required patients to begin treatment “within 24 hours after the first dose of a vasopressor.” Did the patients simply receive treatment after the opportunity to affect the sepsis cascade was missed? The authors provide data on the median time from the initiation of antibiotics to initial vasopressor therapy (2.5 hours), but they do not provide information regarding the time from diagnosis, from initiation of vasopressors, or from arrival in the emergency department until study-drug administration. It is hard to close the book on any therapy for critically ill patients until it is known whether it was administered during the time frame when the therapy would be most likely to show benefit. Common sense dictates that early interventions would be more helpful than later interventions.

Judd E. Hollander, M.D.
University of Pennsylvania, Philadelphia, PA
judd.hollander@uphs.upenn.edu

Although disappointing, the results of the PROWESS-SHOCK study in which drotrecogin alfa (activated) (DrotAA) (recombinant human activated protein C) was used for severe sepsis are not surprising. Biologically active molecules such as DrotAA are likely to be effective in targeting specific pathophysiological disturbances in disease states such as severe meningococcal sepsis with disseminated intravascular coagulation or purpura fulminans as the dominant clinical feature.1That no biologically active substance has shown benefit in sepsis may have more to do with studies that involve heterogeneous groups of patients with diverse causes of sepsis rather than a single disease entity. Anecdotal reports of the use of activated protein C in severe meningococcal disease in children and adults2-4 provide support for the view that DrotAA could be therapeutically useful in this condition. Unfortunately, DrotAA was abruptly withdrawn from the market. This means that clinical trials of human recombinant activated protein C in life-threatening meningococcal infection can no longer be conducted.

Vineet Nayyar, F.R.A.C.P., F.C.I.C.M.
Thomas Solano, F.R.A.C.P., F.C.I.C.M.
Westmead Hospital, Westmead, NSW, Australia

Author/Editor Response

We agree with Hollander’s implication that multiple reasons may explain why a study such as PROWESS-SHOCK did not show a benefit. However, PROWESS-SHOCK showed that DrotAA did not benefit patients who met the specific criteria for septic shock that were defined for enrollment in the trial.

It is difficult to define precisely when sepsis begins. We did record the time from the start of treatment with a vasopressor to the start of study-drug infusion1 and found no heterogeneity in the treatment effect on 28-day mortality (see Fig. S1 in the Supplementary Appendix, available with the full text of our article at NEJM.org) and 90-day mortality (Figure 2B of our article), according to quartiles of time of this interval. The mean (±SD) time from the start of a vasopressor to the start of DrotAA was 17.2±5.7 hours, which is similar to the interval from the first organ failure to the start of DrotAA reported in the PROWESS study (17.5±12.8 hours).2 The shortest interval was 4 hours. Thus, we found no evidence of a beneficial effect of DrotAA early during the course of septic shock.

Meningitis was listed as the cause of septic shock in only 24 of 1696 patients enrolled in PROWESS-SHOCK, and Neisseria meningitidis was cultured in samples obtained from only 7 patients. Thus, our study cannot address the question of the efficacy of DrotAA for that specific population. However, as Nayyar and Solano suggest, we did anticipate that if DrotAA was beneficial, the benefit would be greatest among patients presenting with a more coagulopathic phenotype. Thus, there were also predefined subgroups according to the protein C class and baseline Sequential Organ Failure Assessment score for coagulation. There was no heterogeneity in the treatment effect on 28-day mortality (Fig. S1 in the Supplementary Appendix of our article) and 90-day mortality (Figure 2B of our article) in these subgroups. Furthermore, the relatively small Resolution of Organ Failure in Pediatric Patients with Severe Sepsis trial included subgroups of children with septic coagulopathy and purpura fulminans and showed no treatment effect of DrotAA as compared with placebo.3 That all six remaining subgroups in PROWESS-SHOCK did not show a treatment effect should reassure clinicians who no longer have DrotAA available for the treatment of septic shock.