We evaluated a non-invasive method for predicting the ovarian function of premenopausal breast cancer patients with amenorrhoea after chemotherapy. A total of 34 patients had ovarian function, whereas 56 had no ovarian function. Logistic regression analysis indicated that age (P = 0.034; hazards ratio [HR], 0.29; confidence interval [CI], 0.091-0.910), follicle-stimulating hormone (P = 0.032; HR 0.97; CI 0.944-0.997) and oestradiol (P = 0.047; HR 1.01; CI 1.000-1.015) were independent influencing factors that determine ovarian function. The ovarian function score (OFS) (P < 0.001; HR 48.00; CI 10.174-226.452) was obtained through a comprehensive analysis of these three variables, and it could more effectively predict ovarian function. According to receiver operating characteristic curve analysis, the OFS had the highest values compared with the other three variables (sensitivity, 94.6 %; specificity, 79.3 %). The OFS is simple and easy to use; thus, it is expected to become a new method for determining drug-induced amenorrhoea in women with breast cancer. Ovarian function likely still exists if the OFS is ≤1.
Keywords: Amenorrhoea; Breast cancer; Ovarian function; Ovariectomy; Sex hormone.