Predictors of breast cancer triple subtype status, including race

A71

Introduction
 The triple-negative (TN) subtype of breast cancer, tumors not expressing estrogen receptor (ER), progesterone receptor (PR), or Human Epidermal Growth Factor Receptor 2 (HER2), has no viable targeted therapy, leading to poorer survival relative to tumors expressing such receptors. Black race, younger age, and later stage are associated with the TN subtype. What is not known is whether anthropometrics, demographics and reproductive history are associated with triple subtype status and whether these characteristics interact with race. These associations may partially explain the higher prevalence of TN tumors among black women.
 Methods
 Invasive breast tumors from a population-based cohort of 117 black and 362 white Atlanta women aged 20-54, diagnosed between 1990 and 1992, were centrally reviewed and immuno-histochemically analyzed for ER, PR, and HER2; then sub grouped (TN; ER-PR-HER2+; ER+/PR+HER2+; ER/PR+HER2-). Interviews conducted shortly after diagnosis included questions on reproductive and breastfeeding history, physical activity, and body composition throughout life. Anthropometrics (weight, height, waist and hip circumference) were measured at the interview. First, we analyzed whether education, poverty ratio, insurance, age at menarche, age at first birth, recency of birth, and physical activity, body mass index (BMI), and waist to hip ratio were associated with triple subtype status. Next, we investigated if these characteristics were associated with race. Lastly, we stratified the race and triple subtype status estimates by factors that were associated with both triple subtype status and race, to observe if the association between triple subtype and race was similar in all subgroups. Logistic regression was used and age- and stage-adjusted odds ratios (OR) and 95% confidence intervals (CI) were estimated.
 Results 
 Women at ≤ 200% of the poverty level were more likely than women at 200-700% of the poverty level to be diagnosed with TN tumors than with ER/PR+, HER2- tumors [OR = 1.9 (95% CI=1.2, 2.9)]. Women with ≥ 4 births were more likely to have TN tumors than were nulliparous women [2.7 (1.4, 5.1)]. Elevated BMIs at all stages of life were associated with TN tumors [e.g., OR for being obese at interview = 2.8 (1.9, 4.2)]. TN tumors were also more prevalent among the young, alcohol abstainers, and those with less education, or with public or no insurance. Race and TN status estimates did differ significantly among subgroups defined by BMI and poverty ratio. Among those at ≤ 200% of the poverty level, black women were more likely to be diagnosed with TN tumors than white women [OR=4.2 (1.5, 11.2)]. Among women at >700% of the poverty level, black women were not more likely than white women to have TN tumors [OR=1.7 (0.7, 4.2)]. Among under- or normal weight women, black women were more likely to have TN tumors than white women [OR= 4.6 (2.5, 8.4)]. Among obese women, black women were not more likely to have TN tumors than white women [OR=1.5 (0.8, 2.9)].
 Conclusions
 In addition to race, socioeconomics and BMI are associated with TN tumors. Due to the case-only design, no inferences can be made about the etiology of TN disease; rather this study suggests mechanisms affecting survival after diagnosis. The association between race and TN status may be modified by other factors, including poverty and BMI, suggesting these factors may influence the association between race and TN status.