P4-11-12: Molecular Phenotype of Breast Cancers in a Large Cohort of Young Women According to Time Interval Since Pregnancy.

Abstract

Abstract Background: The increase in breast cancer risk during pregnancy and post partum is well recognized. The cross-over to protective effect does not occur until many years later and varies with age at first birth. Recently, a genomic signature specific to the pregnant compared with the non-pregnant breast has been identified; this signature remains present in the postmenopausal parous breast. Given this, we investigated whether time interval since pregnancy affects the phenotype of breast cancers arising in young women compared with nulliparous women. Methods: We examined molecular phenotype, according to histologic grade and biomarker status, in relation to time since pregnancy in an ongoing prospective cohort study (n=355) of young women (≤40yrs) with breast cancer. Medical records were reviewed for tumor stage and receptor status. Parity was ascertained from questionnaires completed within 9 months of diagnosis. Tumor grade was determined by central pathology review. Using tumor grade and biomarker expression, cancers were categorized as luminal A (ER+ and/or PR+, HER2−, histologic grade 1 or 2); luminal B ( ER+ and/or PR+, HER2+, or ER and/or PR+, HER2− and grade 3); HER2 type (ER-, PR-, HER2+); and triple negative (ER-, PR-, HER2−). Results: The median age of the study population is 37 years (range 17–40). Overall, 80% of women had stage 1 or 2 disease; 67% of cancers were ER positive and 32% showed HER2 overexpression. The distribution of breast cancer molecular phenotypes by time interval since last pregnancy is shown in the table. Distribution of molecular phenotype by interval between last pregnancy and diagnosis In our large cohort of parous young women, we found no differences in the distribution of molecular phenotype according to time interval since pregnancy. However, nulliparous young women were more likely to develop luminal A cancers compared to parous women (40% vs. 29%; unadjusted chi square p-value=0.03) and appeared less likely to develop HER2−type and triple negative cancers (7% vs. 13%, p-value=0.09 and 17% vs. 23%, p-value=0.22 respectively). There were no differences in the distribution of luminal B cancers. Conclusions: The distribution of molecular phenotypes is similar among parous young women regardless of the time interval since parturition. Nulliparous young women appear more likely to develop luminal A cancers compared to parous women. Whether the difference in molecular phenotypes of pregnancy-associated breast cancers vs. cancers arising in nulliparous women is due to the effects of genomic alteration remains to be investigated. Effects of a prior pregnancy appear consistent across a 5-year period, in keeping with the concept of genomic alterations identified in the normal pregnant breast and thereafter. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-11-12.