Purpose and design: This is a critical review of the growing body of data, 32 retrospective studies of the outcomes of 9,665 women published since 1989, relevant to the possibility that the timing of primary breast cancer resection within the menstrual cycle impacts breast cancer recurrence and/or spread and patient survival. This article evaluates and contrasts the adequacy of information and data analysis presented in each publication. The overall purpose of this exercise is to rigorously determine the relative strength of the hypothesis that the menstrual cycle timing of operation impacts outcome and, thereby, to determine whether or not a specific change in the practice of breast surgical oncology can be recommended.
Results: The single most completely reported and thoroughly analyzed series, involving 1,175 young women, indicates that surgical resection timing is likely to be relevant to outcome. Seven additional high-quality studies involving 2,864 women have been most completely reported. While two of these eight find no impact, six (75%) of these studies find that breast cancer outcome is affected by operative timing. Nine of the remaining 24 less-complete studies report a statistically significant effect of operative timing. Among these 15 studies of the fates of more than 5,000 women, the opportune menstrual cycle phase almost invariably includes the putative luteal phase. A large number of retrospective studies of widely varying quality find no outcome difference as a function of resection timing. The adequacy of design of each of four ongoing prospective studies is found lacking.
Conclusions: Although it is likely that the menstrual cycle phase of operation is relevant to outcome, the nature of the available data cannot allow a clear recommendation of precisely when to operate. It is, therefore, concluded that current retrospective data are inadequate to recommend an immediate change in practice. Prospective studies of this potentially important question are essential. The prospective trials initiated to date will not be able to definitively answer this question because of inadequate chronobiological design. The minimal requirements for adequate prospective study are delineated.