Minimum trial curation to reasonably assess long-term efficacy of nonhormonal hot flash therapies

Background: Because hot flashes usually persist for years after menopause, a clinically meaningful hot flash therapy needs to have long-term efficacy; however, it is unclear for how long a therapy needs to be compared with a placebo before long-term efficacy can be reasonably deduced. The Food and Drug Administration (FDA) requires a 12-week treatment period for industry-initiated hot flash trials, whereas most academic-initiated trials have ranged from 4 to 12 weeks. We have focused on reviewing nonhormonal hot flash trials to identify inadequate trial durations as a guide toward deducing adequate trial duration to reasonably assess for long-term efficacy. Methods: An electronic database search of MEDLINE, Web of Science, and PsycINFO was performed from 1966 to May 2009 to identify target studies showing a nonhormonal hot flash therapy to be effective at early time points only to become ineffective at later time points (i.e., showing short-term but not long-term efficacy) in a randomized controlled trial (RCT). The longest early time point of efficacy from the target studies plus 1 additional week would be considered the minimum treatment duration necessary to assess for long-term efficacy. Results: Of 2518 citations, 54 RCTs met our inclusion criteria, from which 3 target studies were identified. These 3 target studies evaluated Bellergal Retard (Sandoz, East Hanover, NJ), soy, and venlafaxine and showed times of 2, 6, and 7 weeks, respectively, when the nonhormonal compound last demonstrated efficacy before subsequently losing efficacy in a single RCT. Conclusions: This analysis supports a hot flash RCT duration of at least 8 weeks to reasonably assess a nonhormonal compound's long-term efficacy.