Abstract 758: The clinical utility of BRCA1/2 pathogenic variants in breast cancer patient prognosis

Abstract

Abstract Loss-of-function variants in the BRCA1 and BRCA2 genes cause hereditary breast cancer with high penetrance. The pathogenic BRCA1/2 variants are associated with high-grade tumors and early age at diagnosis, which both are associated with poor patient survival. However, it is unclear, whether the variants have prognostic value beyond the diagnosis age and tumor pathology. We present preliminary results from analyses of combined data from the CIMBA (Consortium of Investigators of Modifiers of BRCA1/2) and BCAC (Breast Cancer Association Consortium), including complete follow-up and pathology data on ER, PR, Grade, T, and N from 827 BRCA1 and 632 BRCA2 variant carriers, and 16624 non-carriers. The data was divided into derivation and validation sets as follows. The variant carriers from the CIMBA (nBRCA1=690, nBRCA2=458) and about 90% of non-carriers from the BCAC (n=15053) were used in building two age-specific survival models, one for BRCA1 and ER-negative breast cancer, and another for BRCA2 and ER-positive breast cancer. All BRCA1/2 carriers (nBRCA1=137, nBRCA2=174) and the remaining non-carriers from the BCAC (n=1571) were used for model validation. The analyses were performed with Cox regression, as in the R library survival, for overall 10-year survival for ER-negative breast cancer, and for overall 15-year survival for ER-positive breast cancer. In the ER-negative breast cancer survival model, the prognosis of the pathogenic BRCA1 variant carriers got worse with increasing age at diagnosis. In contrast, the age-dependent survival of the non-carriers with ER-negative breast cancer followed a U-shaped curve, so that the survival was worst in the very young patients, improving in patients diagnosed between 35 and 50 years of age, and declining again in the older patients. This age-specific trend was replicated in the validation data, with Gönen & Heller unbiased concordance statistic K=0.71, in comparison to K=0.69 for a model without the BRCA1 status. In distinct age groups of the derivation data, the pathology-adjusted hazard ratio for BRCA1 status was 0.39 (95% confidence interval 0.16-0.98) in patients diagnosed at or before the age of 35 years, 1.64 (95% CI 1.09-2.47) in patients diagnosed between 35 and 50 years, and 0.92 (95% CI 0.52-1.60) in older patients. The survival of BRCA2 variant carriers with ER-positive breast cancer remained rather constant over age, whereas for non-carriers the prognosis of middle-aged patients was better than the prognosis of younger or older patients. Therefore, the survival difference between BRCA2 carriers and non-carriers was marked in the age group 45-55 years at diagnosis, with HR 1.67 (95% CI 1.07-2.60) associated with BRCA2. This finding was replicated in the validation data (HR 2.53, 95% CI 1.23-5.22). Based on our findings, information on germline BRCA1 and BRCA2 pathogenic variants could be used to improve the prognostic estimation of breast cancer patients. Citation Format: Taru A. Muranen, Anna Morra, Georgia Chenevix-Trench, Antonis C. Antoniou, Douglas F. Easton, Marjanka K. Schmidt, Carl Blomqvist, Heli Nevanlinna, on behalf of CIMBA and BCAC. The clinical utility of BRCA1/2 pathogenic variants in breast cancer patient prognosis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 758.