Breast Cancer Molecular Subtypes And Risk Factors In Premenopausal Latin American Women: The PRECAMA Study

Smiths LUEONG, International Agency for Research on Cancer, France
VILLAR S. 1 , DONATO E. 2 , RINALDI S. 1 , CARAYOL M. 1 , GARMENDIA M. 3 , RODRIGUEZ A. 4 , SANCHEZ G. 5 , TORRES G. 6 , PORTER P. 2 , OLIVIER M. 1 , ROMIEU I. 1 , PRECAMA TEAM . 1

1 International Agency for Research on Cancer, Lyon, France
2 Fred Hutchinson Cancer Research Center, Seattle, USA
3 Instituto de Nutrición y Tecnología de los Alimentos, Universidad de Chile, Chile
4 Fundación INCIENSA-Proyecto Epidemiológico Guanacaste, San Jose, Costa Rica
5 Facultad de Medicina, Universidad de Antioquia, Colombia
6 National Institute of Public Health, Mexico

Purpose
Breast cancer (BC) encompasses heterogeneous pathologies that can be further subtyped by genomic alterations. BC genomic subtypes have been associated with different outcomes, but how this molecular heterogeneity relates to risk factors and etiology remains largely unknown. In Latin America (LA), there is a high incidence rate of breast cancer in premenopausal women. The PRECAMA study, a multicenter population-based case-control study that includes tumor sample collections, was developed to characterize the molecular features of premenopausal BC and their link to risk factors.

Methods
Pathological tumor samples collected in Chile, Colombia, Costa Rica and Mexico were analyzed for molecular profiling. Immunohistochemistry (IHC) was performed centrally for ER, PR, HER2, KI67, EGFR, CK56 and p53 protein markers. Targeted deep sequencing was done on genomic DNA extracted from formalin-fixed paraffin-embedded (FFPE) tumor tissues and their paired blood samples to screen for somatic mutations in eight genes frequently mutated in BC (AKT1, CDH1, ERBB2, NOTCH1, PIK3CA, PTEN, RB1, TP53). Statistical differences between BC subtypes defined by their IHC and mutation profiles were analyzed by Chi-2 tests using R software.

Results
The genes analyzed were mutated at frequencies similar to those described in databases covering European and North American populations. As expected, TP53 mutations were more frequent in HER2 and triple-negative IHC subtypes while PIK3CA was more frequent in ER positive tumors. Interestingly, a high proportion of TP53 G:C>T:A mutations was observed in TP53 in PRECAMA cases compared to other series.   

Conclusions
These preliminary results suggest common pathways but possible differences in mutagenic processes giving rise to these tumors. The analysis of a larger number of cases in the near future will allow investigating associations between risk factors and specific molecular features.

Funding sources
Union for International Cancer Control; Pan American Health Organization; and International Agency for Research on Cancer.