Innmune-Related Risk Factors for Non-Hodgkin Lymphoma in Twins

Jun Wang, Thomas M Mack, Amie Hwang, Bharat N. Nathwani, Ann Hamilton and Wendy Cozen


Abstract 1588

Acquired or inherited immune deficiency is an established risk factor for non-Hodgkin lymphoma (NHL), but does not account for the majority of cases. Recent epidemiologic evidence from the InterLymph Consortium suggests that subclinical, subtle alterations in immune response may also affect risk. Specifically, atopic conditions have been associated with a protective effect and infections at an early age with risk. Unaffected twins of patients are ideal controls because they are partially (fraternal) or completely (identical) matched on genetic factors. We evaluated these factors in twin pairs in whom one member was diagnosed with NHL. At least one member of 197 twin pairs returned detailed questionnaires; in 75 pairs, both the case and unaffected co-twin returned the questionnaire and in 122 pairs only one twin (usually the unaffected co-twin) returned the questionnaire. NHL diagnosis in the case-twin was confirmed by medical record and pathology review. Multivariable conditional logistic regression adjusted for education was used to estimate the effect of the exposures on risk with odds ratios (ORs) and 95% confidence intervals (CI). Eczema in childhood was associated with a 60% decreased risk of NHL (OR= 0.4, 95% CI= 0.2, 0.9). Allergic asthma (OR= 0.2, 95% CI= 0.1, 0.6), hay fever (OR= 0.4, 95% CI= 0.2, 0.8) and allergies to plants (OR= 0.3, 95% CI= 0.1, 0.9) were protective. Compared to no atopic conditions, the twin with ≥ 3 atopic conditions had a 90% decreased risk of NHL (95% CI= 0.0, 0.5). Tonsillectomy and appendectomy were each associated with a significant, ∼ 2-fold increased risk. The twin with more early childhood exposures (more contact with pets, young children, sucking thumb or pacifier) likely to transmit infection was also at an increased risk of NHL. The study confirms the inverse association between allergy and NHL risk in genetically matched cases and controls and provides further evidence that childhood exposures are important mediators of risk.

Disclosures: No relevant conflicts of interest to declare.