Sepsis Drug Withdrawn, but Meta-Analysis Shows Benefits

Nancy A. Melville

July 17, 2012

July 17, 2012 — A meta-analysis of the sepsis drug drotrecogin alfa (activated) in real-life usage over the course of 10 years shows efficacy and safety rates similar to those reported in the randomized controlled trial that originally led to the drug's approval. The newly reported analysis runs counter to a second randomized trial that led to the drug's withdrawal.

In a study published online July 17 in the Lancet Infectious Diseases, Andre C. Kalil, MD, from the Department of Internal Medicine, University of Nebraska Medical Center, Omaha, and Steven P. LaRosa, MD, from Beverly Hospital in Massachusetts, report that a review of 9 controlled trials and 16 single-group studies, together involving 47,223 patients, and 20 safety analyses involving another 8245 patients showed a significant reduction in hospital mortality with drotrecogin alfa (activated). The analysis also showed increased rates of serious bleeding with the drug.

The new findings are consistent with the randomized, double-blind, placebo-controlled Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis (PROWESS) study that led to the drug's approval in the United States in 2001 and in Europe in 2002.

However, amid concerns that the drug was not effective in patients with less severe sepsis and children, researchers conducted a second clinical study, called PROWESS-SHOCK. The results of this trial did not favor the drug and resulted in Eli Lilly withdrawing the drug in October 2011.

FDA Issues Safety Announcement

The US Food and Drug Administration, meanwhile, issued a Safety Announcement indicating that patients being treated with drotrecogin alfa (activated) should be taken off the drug and that no new patients should be prescribed.

Drotrecogin alfa (activated), or recombinant human activated protein C, is an anticoagulating, anti-inflammatory, and profibrinolytic molecule. Its main mechanism of action is believed to be related to endothelial protein C receptor–activated protein C–protease activated receptor 1 signaling, which provides a cytoprotective effect on the endothelium, according to the authors.

The newly reported 10-year analysis indicates that patients treated with drotrecogin alfa (activated) had an 18% reduction in hospital mortality compared with control patients (relative risk [RR], 0.822; 95% confidence interval [CI], 0.779 - 0.867; P < .0001; I ², 40%).

The reduction was in line with the rate indicated in PROWESS (RR, 0.851; 95% CI, 0.740 - 0.979), but smaller than that of patients in PROWESS who had high disease severity (RR, 0.708; 95% CI, 0.590 - 0.849).

Significant mortality reduction with lower heterogeneity was also seen with propensity-adjusted studies (RR, 0.844; 95% CI, 0.800 - 0.891; P < .0001; I ², 18%). The addition of PROWESS-SHOCK results did not alter the findings, the authors note.

Improved benefits with drotrecogin alfa (activated) were also seen with metaregression in relation to increasing control mortality (P = .01) and more severe disease (P = .04).

Single group studies showed hospital mortality rates for patients treated with drotrecogin alfa (activated) to be 41% (95% CI, 35% - 48%), which is higher than the 31% indicated in the PROWESS study (95% CI, 27% - 36%; P < .0001).

Patients receiving drotrecogin alfa (activated), meanwhile, had a serious bleeding rate of 5.6% (95% CI, 4.5% - 6.9%), which was greater than the 3.5% (95% CI, 2.5% - 5.0%) shown in PROWESS (P = .003), but similar to that reported in PROWESS high disease severity (P = .073).

"The risk of severe bleeding with drotrecogin alfa (activated) was greater in clinical practice than it was in the registration trial and seems to be because of off-label use rather than...treatment of patients with increased disease severity," the authors conclude. "Our effectiveness findings accord with PROWESS but not with the PROWESS-SHOCK trial."

In a related commentary, Jean-Louis Vincent, MD, from the Free University of Brussels, Belgium, said the study is important in offering a real-world perspective on drotrecogin alfa (activated).

"The Death Rate From Sepsis Is Unacceptable"

"[C]ompared with the original PROWESS population, drotrecogin alfa (activated) has been used in more severely ill patients and is associated with a somewhat higher risk of bleeding, but does seem to be effective," he writes. "Although these data might not represent the highest levels of evidence, many of us (including me) maintain that we can learn a lot from such real-life studies."

"The death rate from sepsis is unacceptable and new drugs are needed urgently," he writes.

Dr. Vincent suggests that the dismissal of drotrecogin alfa (activated) based on the earlier research may have been a case of "throwing the baby out with the bathwater."

"By withdrawing it, have some patients been denied an effective treatment? Should the focus not instead have been on establishing which patients were most likely to benefit from its use?" he asks.

"Researchers need to focus on the key lessons learned from drotrecogin alfa (activated), including the notion that targeting of 'sepsis' itself is an inadequate approach to the development of therapies to improve outcomes in this heterogeneous group of patients," he adds.

Clifford S. Deutschman, MD, president of the Society for Critical Care Medicine, agreed that the need to better manage sepsis is urgent.

"This is an awful disorder that kills about a quarter of a million people in the US every year, and that's probably underestimated," he told Medscape Medical News.

"It is responsible for more deaths than the 4 most common types of cancer put together in the US and probably kills millions worldwide, yet it's unrecognized and research is woefully underfunded," said Dr. Deutschman, a professor of anesthesiology, critical care, and surgery at the University of Pennsylvania in Philadelphia.

He expressed doubt, however, about the results of the meta-analysis changing anything in terms of the future of drotrecogin alfa (activated).

"The problem with any meta-analyses is they require a level of statistical manipulation that makes the results difficult to interpret. It's not straightforward like a randomized controlled trial, where the answer is yes or no," he noted.

"It looks to me like the authors did everything they could to make this as fair and unbiased as possible, and they did a good job of selecting which studies to use," he said.

"There may indeed be people buried in these studies who do benefit from the drug, but the problem is figuring out who they are and doing so cost-effectively, and we're unfortunately reliant on companies that have to turn out a profit in order to justify to shareholders what they're doing, so that really makes it difficult," Dr. Deutschman concluded.

Two authors have each received research grant support from Eisai, AstraZeneca, and Agennix, and one author has consulted for Artisan Pharma. Dr. Vincent has been an investigator in trials sponsored by Eli Lilly, including the PROWESS study. Dr. Deutschman has disclosed no relevant financial relationships.

Lancet Infect Dis. Published online July 17, 2012. Article abstract, Commentary extract