Disparities in Early-Stage Breast Cancer and Survival-Reply.

Abstract

We thank Drs. Bazan and Obeng-Gyasi (1) for their interest in our study (2) and we welcome this opportunity for further discourse. In their letter, they suggest that we extend our analysis by examining the study endpoints with the disaggregation of the Hispanic population by race. We agree with them that Hispanic women are a heterogeneous group, and a more granular analyses of racial disparities within Hispanic subgroups may provide additional insights into our finding of worse survival outcomes among Hispanic women with early-stage disease and across subtypes (except triple-negative disease).Although we agree regarding the importance of disaggregating racial and ethnic data in breast cancer research, small sample sizes and limited data availability (on Hispanic origin/subgroups) prohibited a more granular analyses of racial/ethnic disparities within Hispanic race and origin subgroups. For example, Hispanic Black persons comprised only 0.18% of the entire study population, Hispanic Asian women comprised 0.09% (n = 768), Hispanic American Indian/Native Alaskan women, and 0.03% (n = 216) identified as Hispanic Pacific Islander. The limited sample size of Hispanic race groups were a concern given our stratification of models by stage at diagnosis, further limiting the sample size needed to produce statistically reliable estimates. Moreover, because the Surveillance Epidemiology and End Results (SEER) database did not start recording HER2 status until 2010, analyses stratified by molecular subtype are limited to women diagnosed in 2010 and after, further constraining numbers available for analyses. Future studies should examine breast cancer disparities within the Hispanic population by race.Another important point made by Drs. Bazan and Obeng-Gyasi about analyzing Hispanic patients by country of origin, with which we also agree, is regrettably unfeasible as data on country of origin are not available from SEER. Given projected growth of Hispanic populations across the United States, investment in more meaningful data collection to facilitate disaggregation is a critical step in advancing health equity. Indeed, a key strength of the SEER resource, central to its mission, is its oversampling of specific groups in order to facilitate estimation of national trends in incidence, survival, and mortality rates for each group and comparison by group. However, current approaches for such sampling may require reconsideration as aggregating “Hispanics,” “Asians,” and other groups may exert an unintended structural bias that impedes careful evaluation of cancer health disparities instead of exposing it.See the original Letter to the Editor, p. 1867No disclosures were reported.This research and Drs. K.M. Primm and S. Chang are supported by the MD Anderson Cancer Center, and Drs. K.M. Primm and S. Chang are supported by an award from the Cancer Prevention and Research Institute of Texas (CPRIT) for the CPRTP Postdoctoral Fellowship in Cancer Prevention Program (RP 170259).