Associations between polygenic risk scores for immune dysregulation and Hodgkin lymphoma in childhood cancer survivors

Background: Hodgkin lymphoma (HL) is the most common malignancy diagnosed in adolescents aged 15-19 years. Epidemiologic evidence suggests a link between HL incidence and immune dysregulation, suggesting that genetic variability in immune function may contribute to disease susceptibility. However, the extent of shared genetic overlap between immunologic and oncologic phenotypes has not been fully elucidated. To address this gap, we evaluated associations between HL and polygenic risk scores (PRS) for various immune-related conditions previously implicated in HL risk, to determine whether specific immune predispositions are enriched among cases of HL compared with other non-hematologic malignancies. Methods: We used the St. Jude Survivorship Portal, which aggregates data on ~8000 childhood cancer survivors from the St. Jude Lifetime (SJLIFE) and the Childhood Cancer Survivor Study (CCSS) cohorts. The portal integrates 1,600 phenotypic variables with ~400 million genetic variants and includes PRS available derived from the PGS Catalog. We summarized demographic characteristics using the Portal's summary tools, then applied logistic regression to estimate odds ratios (OR) and 95% confidence intervals (CIs) for HL versus other non-hematologic malignancies per one-unit increase in PRS for five immune-related traits (dermatitis, asthma, Crohn's disease, rheumatoid arthritis, and systemic lupus erythematosus), adjusting for age at cancer diagnosis and sex. For each trait, we selected one PRS from the Portal's precomputed set, prioritizing those with larger discovery samples, broader variant coverage, and ancestry diversity. Analyses were conducted in the combined cohort (3,322 HL and 14,167 other survivors). Results: Among 17,489 survivors of HL and other non-hematologic malignancies, the median age at cancer diagnosis was 7.7 years (IQR 2.8-13.9); 52% were male, and 86% were White, 9% Black, and 5% other. In adjusted logistic regression models, a higher dermatitis PRS was significantly associated with higher odds of HL compared with other non-hematologic malignancies (OR 5.3, 95% CI 1.8-15.4, p=0.002). No significant associations were observed for PRS associated with asthma (OR 1.1, 95% CI 0.7-1.9, p=0.656), Crohn's disease (OR 0.9, 95% CI 0.6-1.3, p=0.604), rheumatoid arthritis (OR 1.0, 95% CI 0.9-1.0, p=0.223), or systemic lupus erythematosus (OR 0.9, 95% CI 0.8-1.1, p=0.455) and HL relative to non-hematologic malignancies. Conclusion: Using the St. Jude Survivorship Portal, we observed that survivors of HL tend to harbor a higher dermatitis PRS compared with survivors of non-hematologic malignancies. Although these findings are exploratory and should be interpreted with caution, they highlight a potential role for immune-related genetic variation in HL etiology and demonstrate a need for further investigation and independent validation.