Reviewers: John Han-Chih Chang, MD
Source: Journal of Clinical Oncology 1999; Volume 17: pages 2365 - 2370
Breast cancer is being "caught" earlier and earlier due to better screening practices. Premenopausal women comprise 1/4th of the breast cancer population. In a recent large meta-analysis, the benefit of chemotherapy (ChT) was found in women less than 50 years of age. For lymph node positive patients the benefit in 10 year overall survival (OS) is reported to be 42% versus 53% in favor of ChT.1 The benefit is of lesser magnitude in lymph node negative patients (71% versus 78% favoring giving ChT).1 However, trade offs for this benefit need to be considered, including premature menopause which is a side effect of ChT. The resultant loss in fertility, and increased risk for heart disease and osteoporosis are considerations.
The data were collected from a larger cohort study of women who had surgery for breast cancer. These women were enrolled at the University of Toronto in Canada to evaluate the prognostic significance of body size and nutrition. The mean age of the selected 183 premenopausal women was 44 years. Most had T1-2 disease (primary tumor no greater than 5 cm), and 1/3rd of the patients were lymph node positive. Nearly 30% received no adjuvant systemic treatment. Twelve percent had tamoxifen alone. The rest had ChT with cyclophosphamide, methotrexate and fluorouracil (CMF) for 6 cycles or epirubicin substituted for methotrexate (CEF).
The results were based on the risk of the patient being menopausal one year after diagnosis due to treatment. A number of factors were evaluated as predictors of menopausal onset. In univariate analysis (evaluation of individual factors), age, weight gain post treatment, tumor stage, nodal stage, and the type of systemic therapy (tamoxifen versus ChT) were all linked to menopausal onset. On multivariate analysis, the possibilities of confounding issues are usually eliminated so that only the true causal factors are found. For example, weight gain is likely a result of being menopausal rather than causing menopause. Another example would be the fact that tumor and nodal stage predicts for who is most likely to receive systemic therapy in this non randomized study. Thus, stage may not predict for menopausal onset but rather ChT utilization. The multivariate analysis demonstrated that the true predictors of menopausal onset were age and the use of systemic therapy (ChT having a greater effect than tamoxifen). The risk of menopause from the use of ChT began to increase at the age of 35 and nearly all patients became menopausal by the age of 45 if ChT was utilized. For tamoxifen, the effects on menopausal onset were much later. The "no systemic treatment" group's curve overlaps the tamoxifen only group's curve until age 40 and does not appear different until age 45.
This study confirms previous findings regarding menopausal risk from systemic therapy. Not addressed was that there was no comparison to adriamycin and cyclophosphamide for 4 cycles, which is also a standard regimen in many other institutions. Their evaluation did compare retrospectively the effect of having an anthracycline (epirubicin) versus none - no difference was found.
This study has tremendous value in counseling patients on menopausal risk from systemic therapy. The older the patient, the higher the risk of menopause with treatment. The benefit of systemic chemotherapy decreases with age.1 Using such data as guidelines, the oncologist can assist patients in making an informed choice regarding their treatment.