Keith A. Dookeran

Using Genetic Technologies To Reduce, Rather Than Widen, Health Disparities

Evidence shows that both biological and nonbiological factors contribute to health disparities. Genetics, in particular, plays a part in how common diseases manifest themselves. Today, unprecedented advances in genetically based diagnoses and treatments provide opportunities for personalized medicine. However, disadvantaged groups may lack access to these advances, and treatments based on research on non-Hispanic whites might not be generalizable to members of minority groups. Unless genetic technologies become universally accessible, existing disparities could be widened. Addressing this issue will require integrated strategies, including expanding genetic research, improving genetic literacy, and enhancing access to genetic technologies among minority populations in a way that avoids harms such as stigmatization.

Abstract B07: Two-pore domain potassium (K+) channel genes and triple-negative (TN) subtype in The Cancer Genome Atlas (TCGA) breast cancer dataset

Introduction: The family of 2-pore domain K+ (K2P) channel genes has 15 members, are background channels which enable the leak of K+ ions from cells, and are considered to be important for baseline activity of cells at rest including membrane potential, calcium homeostasis and volume regulation. Prior studies support the hypothesis that alterations of expression or function of K2P channels in cancer cells may play a significant role in tumor development and progression. In the vast majority of tumors, the abnormally expressed channel is wild type. The role of K2P channel genes in breast cancer is currently emerging. A recent microarray database study suggests that all but 5 K2P gene family members showed altered expression in breast cancer. Other studies suggest potential clinical utility of K2P channel genes as biomarkers associated with TN subtype. A weighted gene co-expression network analysis study showed that upregulation of KCNK5 was associated with poor outcome for TN related tumors. TN tumors are also known to occur more frequently in women of non-Hispanic (nH) black race and association between black race and unique KCNK4 and KCNK9 breast tumor methylation patterns have been suggested; KCNK9 association with TN subtype has also been observed. However, there is a paucity of studies characterizing K2P genes in clinical breast cancer, and the goal of our study was the systematic evaluation of the relationship between K2P gene DNA methylation/expression and TN breast tumor subtype, in the large publically available TCGA dataset. Methods: TCGA invasive breast cancer data was available for 1040 women of which 767 had Illumina HM450 methylation beta-values and 959 had RSEM mRNA expression z-scores (single values from level 3 data). We evaluated the direction and association of all K2P gene expression/methylation loci and TN subtype using age and race adjusted glm models. Age and race adjusted Cox models were used for analysis of disease free (DFS) and overall survival (OS). CpG methylation loci within 25kb from either end of the genes of interest were included for examination (UCSC genome browser, hg37). A total of 724 CG loci were included using this approach, but after exclusion of probes with null reads, the number of loci available for study was reduced to 608. Methylation glm model results were sorted on p-values (smallest to largest) and the top 10 associated loci were selected for reporting and further analysis. Selected loci were then checked for functionality related to expression using Spearman9s correlation. Bonferroni corrected p-values were used where appropriate. Results: Overexpression of KCNK5, KCNK9, KCNK10 and KCNK12, and underexpression of KCNK6 and KCNK15, were significantly associated with TN subtype (all p Conclusions: TN subtype is associated with specific K2P channel gene over and underexpression patterns, and similar expression patterns observed in blacks is consistent with more frequent TN disease. Our findings also suggest a link between KCNK9 overexpression and worse survival in breast cancer. Both KCNK5 and KCNK9 showed hypomethylation patterns correlated with negative expression, and overall gene overexpression related to TN subtype, and these findings appear to be consistent with prior literature. Citation Format: Keith A. Dookeran, Maria Argos. Two-pore domain potassium (K+) channel genes and triple-negative (TN) subtype in The Cancer Genome Atlas (TCGA) breast cancer dataset. [abstract]. In: Proceedings of the Ninth AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2016 Sep 25-28; Fort Lauderdale, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2017;26(2 Suppl):Abstract nr B07.

Abstract B49: The role of KCNK9 and TP53 on the racial disparity in biologically aggressive breast cancer subtype in The Cancer Genome Atlas (TCGA)

Introduction: Compared to non-Hispanic (nH) white, nH black women are more likely to present with biologically aggressive estrogen and progesterone receptor (ER/PR)-negative (ERN) and triple negative (TN) breast cancer subtypes which are associated with worse prognosis (black:white disparity). These subtypes also exhibit greater genetic instability, with TP53 being the single most recurrently mutated gene. KCNK9 is a maternally imprinted and functionally mono-allelic 2-pore domain potassium channel proto-oncogene located at chromosomal region 8q24.3, encoding TASK-3. KCNK9 overexpression may arise via loss of methylation (LOM) or copy number amplification (CNA). We examined KCNK9 and TP53 as potential mediators of the black:white racial disparity in aggressive breast cancer subtypes. Methods: Using publically available TCGA data on 1040 women with invasive breast cancer, we examined: associations of KCNK9 LOM (HM450 beta values; highest third loss vs. others), amplification (putative CNA from GISTIC; any vs. neutral), and mRNA overexpression (RNA-Seq from RSEM; z-score > +1 SD vs. others) with ERN and TN subtype, and overall survival (OS) and disease-free survival (DFS). General TP53 pathway function was examined using mRNA expression z-scores for six related genes ( TP53 MDM2 MDM4 CDKN2A CDKN2B TP53BP1 ). Prevalence risk ratios (RRs) with 95% confidence intervals (CIs) were estimated via logistic regression with model-based standardization (predictive margins) and hazard ratios (HRs) using Cox proportional hazards models, and all models were adjusted at baseline for age as a continuous variable and stage as categorical variable. In mediation analysis, we estimated the proportionate reduction in the racial disparity in aggressive subtypes after accounting for differences in KCNK9 and TP53 pathway biomarkers by comparison of rescaled coefficients using the method of Karlson, Holm and Breen in adjusted models. Death from any cause was considered an event for all-cause mortality (i.e. OS), and recurrence or progression of disease was considered an event for DFS. Results: KCNK9 LOM, amplification and overexpression were each significantly associated with ERN and TN subtype and black race (all p TP53 pathway markers and race, KCNK9 status significantly predicted outcome for both subtypes suggesting the importance of KCNK9 as a molecular driver of aggressive subtype (ERN: LOM RR, 3.69; 95% CI, 2.28-5.98; p TP53 pathway markers, no KCNK9 markers remained independent predictors of mortality or recurrence; however, TN subtype and CDKN2A overexpression remained significant predictors for both (TN: OS HR, 8.96; 95% CI, 2.65-30.28; p CDKN2A : OS HR, 0.29; 95% CI, 0.09-0.97; p=0.045. DFS HR, 0.21; 95% CI, 0.06-0.76; p=0.017). ERN and TN subtypes were more common among black patients (black:white disparity: ERN RR, 2.16; 95% CI, 1.6-2.8; p Conclusions: KCNK9 appears to be a molecular driver of aggressive breast tumor subtype, and in combination with aberrant TP53 pathway overexpression, largely accounts for observed black:white disparity in more biologically aggressive breast cancer subtypes. Citation Format: Keith A. Dookeran, Jacob K. Kresovich, Maria Argos, Garth H. Rauscher. The role of KCNK9 and TP53 on the racial disparity in biologically aggressive breast cancer subtype in The Cancer Genome Atlas (TCGA). [abstract]. In: Proceedings of the Eighth AACR Conference on The Science of Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; Nov 13-16, 2015; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2016;25(3 Suppl):Abstract nr B49.

Abstract C31: Exploring disparities in sentinel lymph node biopsy within the Breast Cancer Care in Chicago study

Background: Lymph node biopsy is a critical component of a diagnostic work-up for patients with invasive breast cancer and is necessary for evaluating the extent and stage of disease so that appropriate treatment can be recommended. The standard procedure prior to the advent of sentinel lymph node biopsy (SLNB) procedures was axillary lymph node dissection (ALND), which is associated with debilitating side effects such as lymphedema. SLNB diffused into practice during the early 2000s and is associated with fewer side effects; for this reason, SLNB has become the standard of care for women diagnosed with invasive breast cancer without clinically palpable lymph node involvement. Using data from the Breast Cancer Care in Chicago study (BCCC), we investigated racial/ethnic disparities in the receipt of SLNB vs. ALND and whether certain tumor, patient, or facility characteristics might help to explain the likelihood of receiving a SLNB. Methods: A total of 989 female patients newly diagnosed with breast cancer were recruited and interviewed for the study. Eligible patients were female, resided in Chicago, had a first primary in situ or invasive breast cancer; were diagnosed between 2005 and 2008 between the ages of 30 and 79 at diagnosis; and self-identified as either non-Latina (nL) White, nL Black or Latina. Of these, 74% (N=731) consented to medical record abstraction and had complete information on receipt of either ALND or SLNB. The racial/ethnic disparity in receipt of SLNB was estimated as a prevalence difference (PD) using logistic regression with marginal based standardization. Mediation analyses were conducted using the method of rescaled coefficients (Karlson, Holm, and Breen). All models were adjusted for age at diagnosis. Results: Roughly half of patients (57%) received a SLNB; nL White women were more likely to receive a SLNB as compared to nL Black and Hispanic women (67% vs 50% and 50%; p= Conclusion: During the mid to late 2000s when the practice of SLNB as an alternative to ALND was reaching about half of eligible patients in Chicago, ethnic minority patients were less likely than their nL White counterparts to receive SLNB. Facility accreditation explained a substantial portion of this disparity. Better access and opportunities to go to highly accredited facilities may decrease this disparity. Citation Format: Bethliz Irizarry, Keith Dookeran, Garth Rauscher. Exploring disparities in sentinel lymph node biopsy within the Breast Cancer Care in Chicago study. [abstract]. In: Proceedings of the Eighth AACR Conference on The Science of Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; Nov 13-16, 2015; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2016;25(3 Suppl):Abstract nr C31.

Abstract B51: Exploring the role of reproductive factors and DNA methylation in ethnic disparities in breast cancer tumor aggressiveness

Purpose: Non-Hispanic (nH) Black and Hispanic (or minority) breast cancer patients tend to be diagnosed with more aggressive forms of breast cancer compared to their nH White counterparts. Prior research as well as analyses from the current study has identified hormonal and reproductive factors associated with breast cancer aggressiveness subtypes. We explored the potential role of hormonal and reproductive factors, and the potential contribution of DNA methylation, in explaining the racial/ethnic disparity in tumor aggressiveness in a population based study of breast cancer disparities. Methods: The breast Cancer Care in Chicago (BCCC) study included 989 recently diagnosed nH White, nH Black and Hispanic patients with first primary breast cancer. Analyses include a subset of 286 patients with available tumor immunohistochemistry (IHC) data on estrogen and progesterone receptor (ER/PR), HER2, p53 and Ki67 status. A tumor aggressiveness score (TAS) with a high internal reliability coefficient (Chronbach9s alpha=0.76) was created from tumor grade and IHC data on ER, PR, HER2, p53 and Ki67. Values were standardized to have a mean of 0 and standard deviation of 1, then summed together and re-standardized to create the score. Pyrosequencing assays for DNA methylation were conducted on a set of DNA sequences identified based on prior literature, ENCODE data for DNA methylation, and transcription for normal vs. cancer cell lines from the UCSC genome browser. Of the 286 patients with tumor aggressiveness data, 214 had available methylation data for BRCA1, GSTM2, EGFR, RASSF1, Sat2 and TFF1 genes. Multivariable linear regression models were estimated with standardized aggressiveness score as the dependent variable and using nested models to conduct likelihood ratio tests for both forward (type 1) and backwards (type 3) analyses. A method of rescaled-coefficients was then used to estimate an average controlled direct effect representing the ethnic disparity in breast cancer tumor aggressiveness and the extent to which the disparity might be explained by patient reproductive factors and tumor DNA methylation. Because stage at diagnosis is downstream of, and strongly influenced by, tumor aggressiveness, it was excluded from our analyses. Results: Factors significantly associated with having a higher TAS (p Conclusions: Our findings suggest that DNA methylation of specific genes may influence breast cancer tumor aggressiveness and may help to explain the preponderance of aggressive subtypes diagnosed in ethnic minority women. DNA methylation may represent a promising avenue for biomarker development for early detection for biologically aggressive tumor types in vulnerable populations. These findings require replication and validation in other studies. Citation Format: Keith A. Dookeran, Abeer M. Mahmoud, Matthew Poulin, Liying Yan, Melanie Ehrlich, Jacob K. Kresovich, Virgilia Macias, Andre Kajdacsy-Balla, Elizabeth Wiley, Garth H. Rauscher. Exploring the role of reproductive factors and DNA methylation in ethnic disparities in breast cancer tumor aggressiveness. [abstract]. In: Proceedings of the Seventh AACR Conference on The Science of Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; Nov 9-12, 2014; San Antonio, TX. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2015;24(10 Suppl):Abstract nr B51.

Abstract B34: Race/ethnicity and disparities in mastectomy use in the Breast Cancer Care in Chicago Study

Purpose: To examine population-based racial/ethnic disparities in mastectomy use among women with breast cancer, and explore potential predictors of this disparity. Methods: Participants were 989 women aged 30-79 years, in a cross-sectional, population-based (NCI-funded, Breast Cancer Care in Chicago) study of newly diagnosed female breast cancer (primary in-situ/invasive) in Chicago, Illinois, from 2005-2008. Participants submitted to an in-person interview of social, health care and attitudinal factors. Medical records were abstracted for tumor characteristics and diagnostic/treatment variables. Multivariable logistic regression models with model-based-standardization were used to estimate risk differences (RDs) with bias-corrected bootstrapped 95% confidence Intervals (CIs). Models incorporated variables pertaining to various domains (socioeconomic disadvantage, cultural beliefs, health care access, and tumor aggression/progression) in order to estimate disparities in mastectomy use, and the extent to which various domains might account for the disparity. Results: There were 397 non-Hispanic (nH) White, 411 nH Black and 181 Hispanic participants. Prevalence proportions for mastectomy use were 40% overall, 34% among nH Whites, 44% among nH Blacks, and 45% among Hispanics. Factors significantly associated with increased mastectomy use overall (p Conclusions: In this population-based study, compared to nH Whites, significant disparities in mastectomy use for treatment of primary breast cancer were observed in nH Black and Hispanic women. These findings could substantially be explained by differences in tumor aggression/progression. Differences in beliefs, sociodemographics, and access variables may be contributing to disparities in mastectomy either directly or through their influence on stage at diagnosis. Our findings suggest that the best approach to reducing disparities in mastectomy would be to intervene on factors that could reduce disparities in stage at diagnosis. Citation Format: Keith Anthony Dookeran, Abigail Silva, Garth Rauscher. Race/ethnicity and disparities in mastectomy use in the Breast Cancer Care in Chicago Study. [abstract]. In: Proceedings of the Sixth AACR Conference: The Science of Cancer Health Disparities; Dec 6–9, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2014;23(11 Suppl):Abstract nr B34. doi:10.1158/1538-7755.DISP13-B34

Abstract 859: Race-based survival and prognosis among lower SES women with ER/PR negative breast cancer

Proceedings: AACR 101st Annual Meeting 2010‐‐ Apr 17‐21, 2010; Washington, DC Background: Heterogeneity exists in survival and prognosis among women of different race with ER/PR negative (-) breast cancer. Minority women often have lower SES, which may be a confounding factor. We evaluated survival and prognosis in a cohort of lower SES non-Hispanic White (NHW), African-American (AA), and Hispanic (HIS) women, in an attempt to disaggregate the effects of race and SES in ER/PR- breast cancer. Methods: Chi-square test was used to examine relationship significance [odds ratios (OR), 95% confidence intervals (CIs)]; survival function estimates were generated using Kaplan-Meier (KM) method and compared using log-rank test; proportional hazards regression models [hazard ratios (HR), 95% CIs] were used to select and evaluate factors prognostic for all-cause mortality, in 213 consecutive [30 NHW, 135 AA & 48 HIS] women treated at an urban hospital [44 months median follow-up] with ER/PR- disease. Results: HIS women were younger than NHW [OR, 0.36; 95% CI, 0.14-0.94; p =0.0368] and AA [OR, 0.50; 95% CI, 0.26-0.96; p =0386]. Compared to NHW and HIS, AA women had more comorbid disease [ORs: 3.53; 95% CI, 1.43-8.66; p =0.0053; & 2.12; 95% CI, 1.04-4.33; p =0.0392], and worse poverty status level (PL) [ORs: 5.43; 95% CI, 2.17-13.69; p =0.0001; & 2.79; 95% CI 1.17-6.65; p =0.0192]. No significant differences were noted between groups for stage at diagnosis, grade, p53 or HER2 status, and chemotherapy use. Baseline prognostic factors were: age [HR, 0.99/yr; 95% CI, 0.98-1.02; p =0.822]; stage [(II-IV/I) HR, 2.45; 95% CI, 1.85-3.24; p /≤ census mean) HR, 2.43; 95% CI, 1.12-5.28; p =0.025]; and chemotherapy [(+/-) HR, 0.51; 95% CI, 0.29-0.89; p =0.017]. Race was not associated with greater hazard mortality [(Other/AA) HR, 0.83; 95% CI, 0.60-1.14; p =0.255], and unadjusted 5-yr survival for NHW, AA and HIS women was 60.9%, 52.4%, and 64.4%. 5-yr survival by race was also not different for women aged <50 yrs (p =0.3287) or ≥50 yrs (p =0.6217). Multivariable models indicated that only stage [HR, 2.45; 95% CI, 1.61-3.74; p <0.001] and chemotherapy [HR, 0.31; 95% CI, 0.10-0.95; p =0.041] remained significant prognostic factors when considered together with the other above-mentioned factors. Further, models for Triple Negative [i.e. ER-, PR-, & HER2- (TN)] phenotype (without HER2 covariate) showed similar results: stage [HR, 2.32; 95% CI, 1.45-3.70; p <0.001]; chemotherapy [HR, 0.23; 95% CI, 0.07-0.75; p =0.015]. Conclusion: Survival is not significantly different among lower SES women with ER/PR- breast cancer of different race. Stage and chemotherapy use, but not race, remained independent prognostic factors in Cox models for ER/PR- and TN disease. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 859.