Goldstein et al.'s Secondary Analysis of Progesterone Clinical Trial for Traumatic Brain Injury Can Only Reflect the Same Trial Design Flaws: A Response to “Very Early Administration of Progesterone Does Not Improve Neuropsychological Outcomes in Subjects with Moderate to Severe Traumatic Brain Injury”

Dear Editor:

In “Very Early Administration of Progesterone Does Not Improve Neuropsychological Outcomes In Subjects with Moderate to Severe Traumatic Brain Injury,” ¹ Goldstein and coworkers performed “a secondary analysis of neuropsychological outcomes to evaluate whether progesterone is associated with improved recovery of cognitive and motor functioning” after traumatic brain injury (TBI) in the controlled Phase III Progesterone for Traumatic Brain Injury: Experimental Clinical Treatment (ProTECT) trial. The authors reported that “Similar to the ProtecT III trial's results of the primary outcome, the secondary outcomes do not provide evidence of a neuroprotective effect of progesterone.”

Although a more detailed neuropsychological evaluation is certainly better than dependence on a single, blunt evaluation of TBI outcome such as the Glasgow Outcome Scale, we do not understand why, in this case, Goldstein and coworkers would have expected a different outcome based on more neuropsychological testing. It is very commendable that the study authors report working hard to “ensure adequate training of the neuropsychological evaluators,” monitoring “quality of test administration, scoring, and data entry on an ongoing basis” at each of the ≥50 enrolling sites. Nonetheless, we cannot agree with their claim that the “lack of a treatment effect of progesterone was [not] related to study procedural shortcomings.” In fact, the negative findings of the ProTECT III clinical trial, although disappointing, are very much the result of several critical problems in design and execution. If the patients' dosing and duration of treatment were wrong, no amount of sophisticated behavioral testing can be expected to provide better outcomes.

In a recent article published in this journal, ² Howard and colleagues reported that suboptimal dosing and duration of treatment were likely to have been important factors contributing to the negative findings regardless of the outcome measures used. In particular, the ProTECT III trial design did not apply United States Food and Drug Administration (FDA)-recommended allometric scaling procedures in translating effective pre-clinical doses to clinical application. Because of the failure of the trialists to account for the more than fourfold metabolic differences between rodents and humans, ProTECT III patients would have received doses approximately six times higher than those indicated by allometric scaling. In addition, the patients received only 4 days of treatment, whereas 15 days is the human equivalent derived via allometric scaling from the animal data (see Howard and coworkers ² for more details).

A number of previously published pre-clinical studies have repeatedly and specifically shown that too high dosing of progesterone produces a washout (hormetic) effect in animals with severe brain injuries caused by cortical impact or ischemic stroke. ^{3 –6} This is not just the case for progesterone. Loss of protection at higher doses can also occur with drugs developed to block tumor angiogenesis in cancer patients ^7,8 and with the use of superoxide dismutases (SODs) in the treatment of ischemic reperfusion injury to heart, brain, and intestine. ⁹ In a careful review of the problem, Calabrese found that >30 neuroprotective drugs tested in stroke and TBI models had U-shaped dose-response curves “in a broad range of experimental protocols.” ¹⁰

Under these circumstances, and despite the authors' best efforts to extract more behavioral information from the trial data, it is hardly surprising that with a sixfold overdose and too short a duration of treatment, additional neuropsychological testing in the ProtecT III patient samples did not reveal better outcomes. What we know from the recent clinical trial results and from the 100% failure of clinical trials for TBI in general is that what is needed is very careful Phase IIB testing of dose and schedule optimization to determine optimal treatment parameters for any future Phase III trials. Then, the inclusion of the more comprehensive and quantitative assessments described by Goldstein and coworkers will be a major advantage in trial design.