Abstract 6344: HLA and non-HLA associations with childhood and adolescent Hodgkin lymphoma risk

Abstract The genetic etiology for Hodgkin lymphoma (HL) has been investigated with findings reported in both human leukocyte antigens (HLA) and non-HLA genomic regions. However, results of genetic association studies that focus on HL developed among children and adolescent individuals are limited. Given the historically high survival rate, it is reasonable to leverage available genetic data for survivors of childhood/adolescent HL for this investigation. Utilizing existing data from the St. Jude Lifetime Cohort, the Childhood Cancer Survivor Study (CCSS) original cohort and CCSS expansion cohort, we performed a fixed-effects meta-analysis of three genome-wide association studies for a total of 1,286 HL cases, 6,824 other childhood cancer cases and 373 non-cancer controls, all of whom are of European ancestry. Three independent SNPs in the HLA locus (rs28383311, Odds Ratio [OR] = 1.80, P = 2.14×10-21; rs3129198, OR = 1.52, P = 2.05×10-14; rs3129890, OR = 1.51, P = 6.21×10-13) were identified using step-wise conditional logistic regression. HLA SNP rs28383311 had a moderate correlation (r2 = 0.46) with rs2858870 reported previously by Cozen et al. (Blood, 2012) in the HLA-DRA gene region in adolescent/young adults onset HL. HLA SNP rs3129198 had a perfect correlation (r2 = 1.0) with rs2281389 (HLA-DPB1) reported previously by Moutsianas et al. (Blood, 2011) in mostly adults-onset HL. In contrast, SNP rs3129890 had a low correlation (r2 = 0.15) with rs2395185 (HLA-DRA) reported previously by Urayama et al. (JNCI, 2012) in HL cases diagnosed across a wide age range between 16 and 80 years old. Our analysis also identified additional SNPs for non-HLA loci at genome-wide significance level including: 1) rs1432297 (OR = 1.30, P = 2.5×10-8), which is in moderate correlation (r2=0.52) with previously reported rs1432295 (REL); 2) rs13279159 (OR = 1.32, P = 1.7×10-8), which is in high correlation (r2=0.75) with previously reported rs2019960 (PVT1); and 3) rs3824662 (OR = 1.52, P = 3.9×10-10), which had a high correlation (r2 = 0.91) with previously reported rs3781093 (GATA3). In addition, a novel uncommon SNP with a large effect size (rs117361561, OR = 1.95, P = 2.50×10-8, minor allele frequency = 0.02) was identified and mapped to PDGFD, a plausible gene for HL susceptibility. Twelve out 18 (67%) previously reported genome-wide significant non-HLA SNPs were replicated with P < 0.05 based on our meta-analysis. By analyzing early onset HL within pediatric and adolescent age range, we further dissected the associations in HLA locus by replicating three independent SNPs. We also found a novel non-HLA locus and replicated the majority of previous findings in the non-HLA region, suggesting mostly overlapping as well as potentially unique genetic etiology for HL risk across different ages of onset. Citation Format: Cheng Chen, Nan Song, Xiaojun Sun, Qian Dong, Zhenghong Li, Hui Wang, Smita Bhatia, Heather Mulder, John Easton, Jinghui Zhang, Yutaka Yasui, Gregory T. Armstrong, Kirsten K. Ness, Melissa M. Hudson, Leslie L. Robison, Zhaoming Wang. HLA and non-HLA associations with childhood and adolescent Hodgkin lymphoma risk [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 6344.