PO.PS01.05 · 人群科学

An investigation of etiologic heterogeneity across transcriptomic subtypes of clear cell renal cell carcinoma in two case-control studies

编号 2349 展板 15 时间 4/20 09:00–12:00 区域 Section 36 主讲 Eun Mi Jung, PhD
分会场 Epidemiology: Cancer Incidence, Mortality, Patterns, and Methodology
该海报暂无可访问的完整资料 AACR 官方页面 ↗

作者与单位

Eun Mi Jung1, Kristine Jones2, Diptavo Dutta1, Jonathan N. Hofmann1, Helena Furberg Barnes3, Stephen J. Chanock1, Nat Rothman1, Paul Brennan4, Kai Yu1, Mark P. Purdue1

1Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD,2Cancer Genomics Research Laboratory, Frederick National Laboratory for Cancer Research, Frederick, MD,3Memorial Sloan Kettering Cancer Center, New York, NY,4Section of Genetics, International Agency for Research on Cancer, Lyon, France

摘要 Abstract

The well validated Clearcode34 transcriptomic subtypes of clear cell renal cell carcinoma (ccRCC) are clinically distinct, with ccB tumors having poorer patient survival than ccA tumors, yet few studies have evaluated whether the subtypes possess distinct risk factor profiles. To address this question, we investigated potential etiologic heterogeneity between the ccA and ccB subtypes in two case-control studies conducted in Europe (144 ccA, 106 ccB, 1,476 controls) and the U.S. (75 ccA, 27 ccB, 1,235 controls). We measured expression levels of the 34-gene Clearcode panel to classify ccRCC tumors by subtype. For each study we conducted (1) case-only analyses (ccB vs. ccA) using logistic regression to assess subtype differences in tumor characteristics and control matching factors (e.g., age, sex, race/ethnicity) and (2) case-control analyses using polytomous regression to compute subtype-specific associations with established and suspected RCC risk factors [e.g., body mass index (BMI), smoking, hypertension, polygenic risk score (PRS) derived from genome-wide association studies (European-ancestry subjects)]. Study-specific findings were combined through meta-analysis using random effects models. We also computed associations with additional suspected risk factors that were assessed in the US study only. In the meta-analysis of case-control findings, we found ccA tumors to be more strongly associated with obesity [odds ratio (OR)=3.75, 95% confidence interval (CI)=1.14-12.35] than ccB tumors (OR=1.56, 95% CI=0.96-2.52). Conversely, ccB tumors were more strongly associated with a PRS based on 13 risk variants (90 th vs. 10 th percentile: OR ccA =1.96, 95% CI=1.23-3.11; OR ccB =3.34, 95% CI=1.83-6.11). From the U.S study, we identified that ccB tumors were more common among US Black vs. White patients (case-only OR=6.71, 95% CI=1.22-36.98), more strongly associated with chronic renal failure (OR ccB =22.12, 95% CI=6.22-78.72; no exposed ccA cases), and more highly associated with a recent PRS based on 108 risk variants (90th vs. 10th percentiles: OR ccA =3.26, 95% CI=1.39-7.60; OR ccB =13.12, 95% CI=1.61-107.02). We did not observe any notable differences by subtype for the associations with smoking, hypertension, hysterectomy, oophorectomy, physical activity, and alcohol consumption. Our findings suggest the existence of etiologic heterogeneity between the Clearcode34 subtypes, with ccA tumors more strongly associated with excess weight and ccB tumors more common among Black patients and more strongly associated with chronic renal failure and genetic risk scores. This evidence suggesting risk factor heterogeneity across subtypes highlights the importance of accounting for tumor molecular characteristics in investigations of ccRCC etiology and risk prediction models.
利益披露 Disclosure
E. Jung, None.. K. Jones, None.. D. Dutta, None.. J. N. Hofmann, None.. S. J. Chanock, None.. N. Rothman, None.. P. Brennan, None.. K. Yu, None.. M. P. Purdue, None.

在会议检索中打开