BH Mason

AIMS: To assess the effect of prognostic factors on overall survival from node negative breast cancer. METHODS: Information was collected on 1138 node negative breast cancer patients in the Auckland region, diagnosed between 1976 and 1985. Prognostic variables investigated included oestrogen (ER) and progesterone (PR) receptor status, tumour grade, tumour size, body mass index, lactation history and parity. The effects of these variables on overall survival were assessed separately in pre and postmenopausal groups. RESULTS: Over a median follow up time of 10.2 years, improved survival was seen in premenopausal women with PR+ status (p = 0.0007), ER+ status (p = 0.03), positive lactational history (p = 0.03) and low tumour grade (p = 0.04). In postmenopausal women, only ER+ status (p = 0.01) and PR+ status (p = 0.02) were associated with improved survival. Multivariate analysis suggested that positive PR status combined with tumour size provided the best prognostic discrimination in premenopausal women, whereas ER status was the dominant prognostic variable in postmenopausal patients. CONCLUSIONS: For premenopausal node negative women, progesterone receptor status, considered either alone, or together with tumour size, provides the best prognostic prediction of survival. By comparison, oestrogen receptor status is the most important predictor of overall survival in postmenopausal women.

There are conflicting reports of seasonal changes in steroid hormone receptor levels in breast cancer tissue. Estrogen receptor and progesterone (PR) receptor levels from 1132 tumors were thus grouped according to month of initial tumor detection or month of tissue sampling/surgery. There was a significant circannual variation in the mean monthly PR receptor concentration in patients grouped according to month of tissue sampling/surgery with peak PR levels in April (late summer-early autumn) and nadir values in August and September (late winter-early spring). There was no significant cyclic variation in estrogen receptor values. A significant annual variation in tumor PR concentration was also seen when receptor levels from individual tumors were grouped according to month of initial tumor detection, with peak PR levels found in January and February. The time interval between tumor detection and biopsy/surgery was 3.3 +/- 5.3 months (mean +/- SD) which was close to the interval between the peak PR concentration expressed by month of tumor detection compared with month of tissue sampling for receptor assay. There was also a significant seasonal variation in the month of initial tumor detection, with peak detection occurring in December (summer). The close synchrony between month of maximum tumor detection and month of peak PR concentration suggests that seasonal changes in detection of breast cancer may in part relate to seasonal changes in hormone responsiveness within tumor tissue.