Differences in risk factors for molecular subtypes of clear cell renal cell carcinoma. Issue 7 (10th June 2021)

• Record Type: Journal Article
• Authors: Purdue, Mark P.; Rhee, Jongeun; Moore, Lee; Gao, Xiaohua; Sun, Xuezheng; Kirk, Erin; Bencko, Vladimir; Janout, Vladimir; Mates, Dana; Zaridze, David; Petruzella, Stacey; Hakimi, Abraham Ari; Linehan, William Marston; Chanock, Stephen J.; Brennan, Paul; Furberg, Helena; Troester, Melissa; Rothman, Nathaniel
• Is Part Of: International journal of cancer. Volume 149:Issue 7(2021)
• Page Start: 1448
• Page End: 1454
• Publication Date: 2021-06-10
• Subjects: case‐control study -- clear cell renal cell carcinoma -- epidemiology -- molecular pathology -- risk factors; Cancer -- Periodicals; Cancer -- Prevention -- Periodicals; 616.994
• DOI: 10.1002/ijc.33701
• Languages: English
• ISSNs: 0020-7136

Abstract

Abstract: The ccA and ccB molecular subtypes of clear cell renal cell carcinoma (ccRCC) have well‐characterized prognostic relevance. However, it is not known whether they possess distinct etiologies. We investigated the relationships between these subtypes and RCC risk factors within a case‐control study conducted in Eastern Europe. We analyzed risk factor data for ccA (n = 144) and ccB (n = 106) cases and 1476 controls through case‐only and case‐control comparisons to assess risk factor differences across subtypes using logistic and polytomous regression models. We also performed a meta‐analysis summarizing case‐only results from our study and three patient cohorts. Patients with ccB tumors had poorer survival than those with ccA tumors and were more likely to be male (case‐only odds ratio [OR] 2.68, 95% confidence interval [CI] 1.43‐5.03). In case‐control analyses, body mass index was significantly associated with ccA tumors (OR 2.45, 95% CI 1.18‐5.10 for ≥35 vs <25 kg/m 2 ) but not with ccB tumors (1.52, 0.56‐4.12), while trichloroethylene was associated with ccB but not ccA (OR 3.09, 95% CI 1.11‐8.65 and 1.25, 0.36‐4.39 respectively for ≥1.58 ppm‐years vs unexposed). A polygenic risk score of genetic variants identified from genome‐wide association studies was associated with both ccA and, in particular, ccB (OR 1.82, 1.11‐2.99 and 2.87, 95% CI 1.64‐5.01 respectively for 90th vs 10th percentile). In a meta‐analysis of case‐only results including three patient cohorts, we still observed the ccB excess for male sex and the ccA excess for obesity. In conclusion, our findings suggest the existence of etiologic heterogeneity across ccRCC molecular subtypes for several risk factors.