Abstract GS2-04: Aromatase inhibitors versus tamoxifen in pre-menopausal women with estrogen receptor positive early stage breast cancer treated with ovarian suppression: A patient level meta-analysis of 7,030 women in four randomised trials

Abstract

Abstract Background: For women with early stage hormone receptor positive breast cancer, adjuvant treatment with tamoxifen reduces their 15-year risk of death from breast cancer by about one third. Aromatase inhibitors (AIs) are even more effective than tamoxifen in post-menopausal women but, used alone, are ineffective in pre-menopausal women due to compensatory ovarian oestrogen production. Several trials have assessed whether, if administered with ovarian suppression, AIs may also be more effective than tamoxifen at preventing breast cancer recurrence in premenopausal women but trial results have been inconsistent. Methods: Individual patient data were available from four randomised controlled trials, including 7,030 pre-menopausal women with estrogen receptor positive (ER+) breast cancer. All women received ovarian suppression or ablation and were randomised to receive either an AI or tamoxifen for 3 years (in ABCSG XII) or 5 years in other trials (SOFT, TEXT and HOBOE). Primary outcomes for the meta-analysis were time to invasive breast cancer recurrence (distant, loco-regional, or new contralateral breast primary) and breast cancer mortality. Log-rank analyses were used to estimate first-event rate ratios (RR) and their confidence intervals (CIs). Results: Overall, the annual rate of recurrence averaged 21% lower (RR 0.79, 95% 0.69-0.90, p=0.0005) for women allocated AI compared to tamoxifen. The main benefit from AI was seen in years 0-4 (RR 0.68, 99% CI 0.58-0.80), during the period when treatments differed, with no further benefit, or loss of benefit, in years 5-9 (RR 0.98, 99% 0.73-1.32), and as yet limited follow up data available beyond year 10. The 5-year absolute risk of breast cancer recurrence was 3.2% lower in the AI group than the tamoxifen group (6.9% vs 10.1%, p=0.0005). Although distant recurrence was reduced with AI (RR 0.83, 95% CI 0.71-0.97, p=0.02), there was no difference in breast cancer mortality although longer follow up is needed to assess effects on mortality reliably. The proportional reduction in recurrence during the period when treatments differed did not vary by age, BMI, or by tumour size, tumour grade, histological subtype, or presence and absence of chemotherapy. In contrast to the findings of the meta-analysis of AI vs tamoxifen in postmenopausal women, AI appeared ineffective in N4+ disease (RR=0.49 in N0, RR=0.56 in N1-3, RR=1.03 in N4+; p for trend =0.0009). The only other apparent heterogeneity was between the four trial results (p=0.03), and between HER2-negative and positive disease (p=0.05), which may be chance findings given the borderline statistical significance. There were more bone fractures in women receiving AI (5.0% AI vs 3.8% Tam, p=0.02). Among these younger women, few non-breast cancer deaths occurred: 0.9% AI vs 0.7% Tam (RR 1.30, 95% CI 0.75-2.25), and the incidence of endometrial cancer was low (0.2% AI vs 0.3% tamoxifen, p=0.14). Conclusion: Using an AI rather than tamoxifen, in premenopausal women receiving ovarian suppression, reduces the risk of breast cancer recurrence by about 20%. If the surprising lack of benefit in N4+ disease is a chance finding, then the absolute benefits will be larger for women at a higher absolute risk of recurrence. Citation Format: Rosie Bradley, Jeremy Braybrooke, Richard Gray, Robert K Hills, Zulian Lui, Hongchao Pan, Richard Peto, Jonas Bergh, Sandra M Swain, Prudence Francis, Michael Gnant, Francesco Perrone, Meredith M Regan. Aromatase inhibitors versus tamoxifen in pre-menopausal women with estrogen receptor positive early stage breast cancer treated with ovarian suppression: A patient level meta-analysis of 7,030 women in four randomised trials [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr GS2-04.