Neurology Progesterone for TBI Flops in Big Trials Published: Dec 10, 2014 | Updated: Dec 11, 2014 By John Gever, Managing Editor, MedPage Today save|AA Progesterone treatment failed to help patients with severe acute traumatic brain injury (TBI) in two large, independent trials -- in stark contrast to the positive results seen in two earlier trials. In one of the new multicenter, placebo-controlled trials, which was stopped prematurely for futility, there was no difference in the likelihood of favorable neurological and functional outcomes at 6 months in 882 randomized patients (relative benefit of progesterone 0.95, 95% CI 0.85-1.06, P=0.35), and rates of phlebitis and thrombophlebitis were three times as common with progesterone versus placebo, according to David W. Wright, MD, of Emory University in Atlanta, and colleagues. These results were replicated closely in a separate multinational trial involving 1,195 patients: 50.4% of those assigned to progesterone achieved favorable 6-month outcomes compared with 50.5% of the placebo group, reported Brett E. Skolnick, PhD, at North Shore University Hospital in Manhasset, N.Y., and colleagues. Both studies were published online in the New England Journal of Medicine. In an accompanying editorial, Lee Schwamm, MD, of Massachusetts General Hospital in Boston, argued that the negative results should not have been a complete surprise. He characterized the effect sizes seen in the earlier, smaller trials as relatively modest and, despite statistically significant P-values, potentially the result of chance. Schwamm said the phase III failures may be indicative of a fundamental bias in the current practice of scientific inquiry, in which investigators move too eagerly from limited phase II studies into larger and longer trials. Slower-than-expected enrollment, overly optimistic effect sizes, better-than-expected performance in the placebo groups, and sample-size estimates that were based on one of multiple efficacy outcomes in small safety trials are commonly observed patterns in failed trials, he wrote, and most of these were present in the progesterone studies. At least to an extent, Skolnick and colleagues agreed, writing that the failures should stimulate a rethinking of procedures for drug development and testing in TBI . They acknowledged that TBI is complex and often poorly characterized in the clinic, objective biomarkers are largely lacking, and clinical outcome measures are insensitive. For their part, Wright and colleagues simply threw up their hands. They argued that they had taken what they believed were appropriate steps to ensure that the trial would find a benefit if one existed. Although reducing the heterogeneity in injury type could make trials easier to design for success, it would come at the cost of more homogenous pathological findings and decreased generalizability of the results, they noted. Schwamm recommended that the trialist community should take more care at the phase II stage, performing more extensive and sequential studies before embarking on large pivotal trials. He said this would make phase II more expensive and time-consuming, but with the benefit of fewer costly failures at the phase III stage. Background and Study Details Impetus for the trials came initially from bench studies and animal studies suggesting that progesterone has neuroprotective functions that could aid recovery from serious head injuries. It originates in the nervous system and a host of lab studies have shown that it inhibits cerebral edema and neuronal damage and death, and promotes functional recovery. That led to clinical studies, with a U.S.-based trial reported in 2006 and a Chinese study published in 2008 both finding that patients with severe TBI, treated within hours of injury with injections or infusions of progesterone, had better outcomes than those receiving placebo. In the U.S. trial, the salient benefit was in 30-day neurologic outcome; the Chinese trial found benefits in 6-month mortality and Glasgow Outcome Scale score, a categorical index of functional ability. Those results led Donald Stein, PhD, an author of the 2006 U.S. paper, and a colleague in 2009 to recommend that progesterone be used routinely in severe TBI cases. However, as Schwamm noted, both trials had multiple outcome measures, and not all of them showed an advantage for progesterone. For the phase III trials, both groups used 6-month Glasgow scores for their primary outcomes. Also in both trials, patients were accepted into the trial if they had severe TBI (Glasgow Coma Scale scores of 8 or less in the study by Skolnicks group 4-12 in the study by Wrights group). Treatment was initiated within hours of injury, with progesterone or placebo given by infusion over 4 to 5 days. There was no suggestion of benefit in either study. In addition to no difference in Glasgow Outcome Scale between groups at 6 months, mortality rates were similar, and there were no specific subgroups of patients defined by injury severity or type of ancillary treatment such as surgery who appeared to do significantly better than others. Neither investigator group nor Schwamm suggested that it would be worth testing progesterone again in severe TBI with different treatment parameters or patient characteristics. The study by Skolnicks group was funded by BHR Pharma, a division of Besins Healthcare. Skolnick disclosed serving as a vice president of Hillhurst Biopharmaceuticals, which is developing a different potential treatment for TBI. Co-authors disclosed relevant relationships with BHR Pharma/Besins Healthcare. The study by Wrights group was funded by the NIH. Wright disclosed holding patents on progesterone for TBI, licensed to BHR Pharma. Co-authors disclosed no relevant relationships with industry. Schwamm disclosed no relevant relationships with industry.
Posted on: Thu, 11 Dec 2014 22:06:34 +0000