Article title: Re-Analyzing Phase III Bremelanotide Trials for “Hypoactive Sexual Desire Disorder” in Women
Authors: Glen I. Spielmans
Journal: The Journal of Sex Research
DOI: http://dx.doi.org/10.1080/00224499.2021.1885601
In the article “Re-analyzing phase III bremelanotide trials for ‘hypoactive sexual desire disorder’ in women” by Glen I Spielmans (Journal of Sex Research. Advance online publication. March 7, 2021), the author mainly relied (erroneously) on versions of the bremelanotide trials’ clinicaltrials.gov entries from 2018. However, the FSDS-DAO #13 was actually originally listed in both trials’ original 2015 clinicaltrials.gov entries as a secondary outcome. Relying on the 2018 entries led to an incorrect conclusion that the FSDS-DAO #13 was originally a coprimary outcome measure in both trials. This led to correcting Table 1, Table 2, the Results (Changed Efficacy Outcomes and Efficacy Outcomes: Coprimary Outcomes sections), and the Discussion: Questionable Research and Measurement Practices.
Second, the author did not cross-check the timing of changing primary outcomes in relation to study completion dates. The author has since found that a) the FSDS-DAO #13 was changed to a coprimary outcome and b) satisfactory sexual events were changed to a secondary outcome after the completion of the phase III studies. This led to correcting Table 1, Table 2, Results: Changed Efficacy Outcomes, and Discussion: Questionable Research and Measurement Practices.
The final two paragraphs of the Results: Changed Efficacy Outcomes section should read as follows (replacing the prior version of the final paragraph): In both studies’ original clinicaltrials.gov entries, the number of sexually satisfying events (SSEs) was listed as one of two coprimary outcomes, with the FSFI-D as the other coprimary outcome. The FSDS-DAO #13 was originally listed as the only secondary outcome, as can be seen if one selects “history of changes” on the trial entries (Clincaltrials.gov, 2018a, Clinicaltrials.gov, 2018b). Study 301 was completed on July 26, 2016, and Study 302 was completed on August 4, 2016 (Kingsberg et al., 2019). In early November 2016, the clinicaltrials.gov entries were updated as follows: a) SSEs were changed to a secondary outcome, b) the FSDS-DAO #13 was shifted to a coprimary outcome, c) all other secondary outcomes were added. In a meeting with the FDA on September 8, 2016, Palatin requested that a) SSEs be changed to a secondary outcome, and b) FSDS-DAO #13 scores replace SSEs as a coprimary outcome (United States Food and Drug Administration, 2019).
According to the FDA, Palatin justified switching outcomes by citing both an October 28, 2014 scientific workshop convened by the FDA (Food and Drug Administration, 2014) and a prior phase II study. During the workshop, outcome measures for HSDD/FSIAD clinical trials were discussed. Among several topics, concerns regarding SSEs as an outcome measure were raised, and the inclusion of desire as a clinical trial outcome was discussed (Food and Drug Administration, 2014). The FDA agreed to the sponsor’s proposed outcome changes. The FDA review states that “It is conceivable that the Applicant was aware from blinded data that the results of SSEs (the original coprimary endpoint) would not meet statistical significance prior to unblinding since neither treatment or [placebo] arm showed any improvement” (United States Food and Drug Administration, 2019, p. 126). Kingsberg et al. (2019) did not report that any outcome changed during the course of the study.
The section heading Efficacy Results: Coprimary Outcomes should read as follows: Efficacy Results: Updated Coprimary Outcomes
The Table 1 row starting with “Change in coprimary measure is unreported” should read as follows, in the second column, “The number of sexually satisfying events (SSEs) was a coprimary measure, but was shifted to a secondary measure after studies were completed. When the number of SSEs was shifted to a secondary outcome, the FSDS-DAO #13 was shifted to a coprimary outcome. Neither change was reported by Kingsberg et al. (2019).”
The Table 2 row starting with FSDS-DAO: Question 13 should read as follows in the second column: “Primary, changed from secondary after studies were completed”. The Table 2 row starting with Number of SSEs should read as follows in the second column: “Secondary, changed from primary after studies were completed”.
The start of the Discussion, Questionable Research and Measurement Practices section should read as follows: “Mean change on the original coprimary (FSFI-D) and updated coprimary (FSDS-DAO #13) measures showed modest benefits over placebo. After study completion, the FDA allowed Palatin to change a coprimary outcome (SSEs) to secondary, while changing a secondary outcome (FSDS-DAO #13) to coprimary. This upgraded the status of an outcome with statistically significant results while downgrading the status of an outcome showing no treatment effect. Palatin cited a phase II study and a relevant FDA workshop from October 28, 2014 to support this change. There may have been solid empirical reasons to change measurement tactics. However, the FDA also noted that Palatin could have seen data indicating a lack of effect on SSEs prior to changing outcomes (United States Food and Drug Administration, 2019). Indeed, if these outcome changes were mainly motivated by a desire for better measurement, it seems they could have been implemented earlier, as opposed to after study completion.”
The original second sentence in the Discussion, Questionable Research and Measurement Practices section should start the second paragraph in this section.