Pergamon 0959-8049(95)00301-0 Original Paper zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP Eur@ean_?oumuiofConcnVol. 31A,Nos13/14,P~.2181-2184,1995 Cwvrinht b 1995 Ekvier Science Lrd Printeri i. &a~Britain. AU rights reserved 0959-8049/5’S $9.50+0.00 Cyclical Tumour Variations in Premenopausal Women with Early Breast Cancer R.A. Badwe,l R. Bettelheim,2 R.R. Millis,’ W. Gregory,’ M.A. Richards’ and I.S. Fentiman’ ‘ICRP Clinical Oncology Unit, Guy’s Hospital, London SE1 9RT, U.K.; and 21nternational Breast Cancer Study Group, Berne, Switzerland The hormonal milieu at the time of tumour excision may have a significant impact on survival in premenopausal patients with breast cancer, with those undergoing surgery between days 3 and 12 of the menstrual cycle having a worse prognosis. To investigate possible mechanisms which might explain this finding, histological features of tumours from 363 patients included in two studies from Guy’s Hospital have been reviewed. Axillary nodal involvement occurred in 71/115 (62%) of patients whose primary tumour was excised between days 3 and 12 of the cycle, compared with 116/248 (47%) of patients undergoing surgery at other phases of the cycle (x2 = 7.04, zyxwvut P < 0.01). Vascular invasion was observed in 541115 (47%) of primary tumours removed between days 3 and 12 and 82/248 (33%) of tumours removed at other times (x2 = 6.47, P < 0.02). Multivariate analysis of factors influencing survival indicated that both axillary nodal status and phase of the cycle were highly significant independent predictors of prognosis. Key words: breast cancer, menstrual phase, prognosis, histology EurJ Cancer, Vol. 31A, Nos 13114, pp. 2181-2184,1995 INTRODUCTION FOR PREMENOPAUSAL women with early breast cancer, there is some evidence that the phase of the menstrual cycle at the time of surgery affects survival [l-3]. In our first report [l] involving patients diagnosed between 1975 and 1985, we compared sur- vival for patients who underwent excision of the primary tumour during a phase of predicted unopposed oestrogen secretion (3-12 days after onset of menstruation), with that of patients operated on at other phases of the cycle. After day 12 of the cycle, both oestrogen and progesterone are normally secreted, whereas levels of both hormones are low between 0 and 2 days after the onset of menstruation. In the first study, patients operated on between days 3 and 12 had a 54% 10 year actuarial survival rate, compared with 84% for those operated on at other times (P < 0.001). In a second cohort of patients (treated since 1985), similar results were reported [4]. Although some other groups have been unable to confirm this finding [5-91, a recent meta- analysis of all published studies has shown that overall there is a significant effect of timing of surgery [lo]. The odds reduction on survival for surgery in the luteal phase was 16%. In this study, data from both cohorts of patients have been Correspondence to I.S. Fentiman. Received 15 Dec. 1994; accepted 5 Apr. 1995. combined and a detailed review of histological parameters undertaken. In neither of the previous reports were any signifi- cant differences observed in the distribution of conventional prognostic factors, such as tumour size or type, between those undergoing surgery between days 3 and 12 after the last men- strual period (LMP) and those operated on at other times. However, in both reports, a non-significant excess of patients with positive axillary lymph nodes was observed amongst those undergoing tumorectomy between days 3 and 12. One hypothesis that might explain the findings related to timing of surgery is that tumour handling may lead to spread of malignant cells and the likelihood of such tumour dissemination occurring may be influenced directly or indirectly by the endo- crine environment. Evidence for the link between hormones and prognosis was provided by a recent study which found that node-positive patients with per&operative progesterone levels > 1.5 ng/ml (luteal phase) had a significantly better survival than those with progesterone levels < 1.5 ng/ml [ 1I]. Dissemination can occur via lymphatics or blood vessels [ 121, and so we were interested in investigating the relationship between phase of the menstrual cycle and vascular invasion by tumour cells. PATIENTS AND METHODS Clinicopathological data of 823 premenopausal patients with invasive operable breast cancer treated at Guy’s between 1975 2181