Background: It is controversial whether the timing of tumor excision relative to the menstrual cycle influences the survival of patients with breast carcinoma.
Methods: Premenopausal patients (n=614) who had surgery for invasive, nonmetastatic breast carcinoma during the period 1978-1988 participated in an epidemiologic survey, reporting their menstrual cycle length and the date of their last menses. We ascertained deaths from any cause before 1993.
Results: Using Cox modeling, we found a nonlinear variation in the relative risk (RR) of death according to the timing of surgery during the menstrual cycle. The curve was best described by a cosine transformation of a 28-day cycle. For patients who had breast carcinoma surgery on the estimated day of ovulation, the risk of death was 0.59 (95% confidence limits [CLI=0.39-0.89, P=0.013) compared with patients who had surgery at the approximate time of menses. We observed this for patients treated in 1978-1981 (RR=0.43, 95% CL=0.23-0.83, P=0.011) and 1982-1983 (0.25, 95% CL=0.10-0.63, P=0.003), but not in 1984-1988 (1.48, 95% CL=0.64-3.4). The difference observed for 1984-1988 was explained by a significant improvement in the mortality rate (P=0.0004) for women whose surgery took place during menses or near to the date predicted for the next menses. No such improvement for women who underwent breast carcinoma surgery around the time of ovulation was observed during the period 1984-1988. These changes were not explained by the performance of lumpectomy or the increasing interval between biopsy and tumor excision.
Conclusions: The shape of the survival curve contradicted the idea that it could be explained by levels of circulating estradiol or progesterone. Because observations that surgery was affected by menstrual timing seem not to have persisted beyond the mid-1980s, this study should not be used to support recommendations that surgeons perform breast carcinoma surgery on any particular day of the menstrual cycle.