Abstract
Breast carcinoma is a heterogeneous disease affected by patients' ethnicity. Gene expression analysis identified several molecular subtypes, and similar subtyping has now been found to be feasible using immunohistochemistry. This study estimated the distribution of intrinsic breast cancer subtypes using estrogen receptor, progesterone receptor, human epidermal growth factor receptor 2 (Her2/neu), and cytokeratin 5/6 immunostaining in a cohort of 125 Egyptian women diagnosed as having invasive breast carcinoma. Associations with clinicopathologic variables and the prognostic markers Bcl-2 and Cyclin D1 were investigated and statistically analyzed. Population difference in breast cancer subtypes was detected, suggesting etiologic and genetic heterogeneity among demographic groups. As reported worldwide, most tumors were luminal A (39.2%), but basal-like and unclassified subtypes had higher proportions among our cohort (16.8% and 16%, respectively), particularly in premenopausal patients (P = .0001), in contrast to postmenopausal African Americans, premenopausal European Americans, and other populations. Her2-overexpressing subtype was the least common subtype (13.65%) among our patients, although it is more common in Asians. Basal-like and unclassified carcinomas were more frequently grade 3 neoplasms (P = .035). Lobular histology was distributed among luminal A, B and unclassified subtypes (P = .006). The highest frequency of nodal positivity was associated with Her2 overexpressing carcinomas (94.1%, P = .0001). Luminal and unclassified carcinomas more likely expressed Bcl-2 (P = .011) and Cyclin D1 (P = .0001), whereas basal and Her2 subtypes had the lowest expression levels. Immunohistochemistry-based subtyping can be helpful in separating breast carcinoma into subtypes that vary in distribution among different populations. These subtypes have distinct clinicopathologic features and diverse prognostication, which may imply different therapeutic options for each subtype.