Non‐Hodgkin lymphoma and autoimmunity: Does gender matter?

Abstract

Autoimmune disorders are more frequent in women, whereas most non-Hodgkin lymphomas (NHLs) are common in men; yet, sexspecific autoimmune-lymphoma associations are rarely reported. Detailed data on autoimmune disease were abstracted from medical records of 791 cases (including 316 diffuse large B-cell lymphomas (DLBCLs); 228 follicular lymphomas (FLs); 127 marginal zone lymphomas (MZLs); 64 T-cell lymphomas and 38 mantle cell lymphomas) and 872 controls. The combined prevalence of autoimmune disease was higher among women (15.7% controls; 19.7% cases) than men (6.6% controls; 14.5% cases), but the overall association with NHL was stronger for men (odds ratio 2.4, 95% confidence interval: 1.5-3.8) than women (1.3, 0.9-1.9), the disparity persisting when data for the year immediately preceding lymphoma diagnosis were excluded (men 2.0, 1.3-3.3; women 1.2, 0.8-1.8). For both sexes, the strongest individual associations were for DLBCL, MZL and T-cell lymphomas, with no associations evident for FL. Among women, there were strong links between MZL and both Sjögren's syndrome and idiopathic thrombocytopenia, and among men, between DLBCL and both rheumatoid arthritis and Crohn's disease. The expected association between coeliac disease and T-cell lymphoma was seen in both sexes. Our results add to the accumulating knowledge on this topic and suggest that future studies should analyze data for men and women separately.

Non-Hodgkin lymphoma and gluten-sensitive enteropathy: estimate of risk using meta-analyses

“…This reflects the known association between gluten-sensitive enteropathy and EATCL, for which risks as varied as 6-to over 200-fold have been reported [5,7,8,11,34]. Similarly, risks for T-cell lymphoma as a whole ranged from 6-to over 50-fold [3,8,18,31,32,36]. Diffuse large B-cell [11-17, 19-22, 24, 26, 28, 30, 33, 34]; observed cases of non-Hodgkin lymphoma and total number of 100,000 person-years (P-Yrs) among persons with (Exp) and without (Unexp) gluten-sensitive enteropathy in nested case-control studies [4,18,27,29,35]; or the number of nonHodgkin lymphoma cases and controls who did (Exp) and did not (Unexp) have gluten-sensitive enteropathy in case-control studies [6-10, 23, 25, 31, 32, 36].…”

“…Among the homogeneous studies, five reported risks by sex [23,25,26,31,32]. Women with gluten-sensitive enteropathy were found to have similar risks of NHL to men (pooled risk estimate = 4.87, 95% CI 2.96-8.02, I 2 = 4%, p(heterogeneity) = 0.39; pooled risk estimate = 4.43, 95% CI 2.91-6.74, I 2 = 0%, p(heterogeneity) = 0.98, respectively).…”

“…AgA antigliadin antibodies; DH dermatitis herpetiformis; EMA endomysial antibodies; NSW New South Wales; tTG tissue transglutaminase antibodies. Study-specific risk estimates were pooled using a random effects model lymphoma (DLBCL) was the other subtype commonly found among gluten-sensitive enteropathy cases and an association has been suggested (three risks, pooled risk estimate = 1.97, 95% CI 1.23-3.15, I 2 = 0%, p(heterogeneity) = 0.67) [8,32,36].…”

“…Women with gluten-sensitive enteropathy were found to have similar risks of NHL to men (pooled risk estimate = 4.87, 95% CI 2.96-8.02, I 2 = 4%, p(heterogeneity) = 0.39; pooled risk estimate = 4.43, 95% CI 2.91-6.74, I 2 = 0%, p(heterogeneity) = 0.98, respectively). Where risks were estimated for gluten-sensitive enteropathy diagnosed more than a year before NHL [24,25,27,32,34], the pooled risk was similar (pooled risk estimate = 3.48, 95% CI 2.33-5.20, I 2 = 35%, p(heterogeneity) = 0.19) to when data were not lagged (pooled risk estimate = 4.46, 95% 3.46-5.77, I 2 = 0%, p(heterogeneity) = 0.94). Longer time periods between the diagnoses of gluten-sensitive enteropathy and NHL were considered in one study, where the excess risk decreased with increasing time and persisted five or more years after the detection of gluten-sensitive enteropathy [25].…”

“…Medical records have also been used to identify gluten-sensitive enteropathy. Diagnosis noted within primary care [27,32,35] may have been via serology followed by biopsy, at least for relatively recent cases. Hospital records relating to gluten-sensitive enteropathy [24-26, 29, 34] could relate to an inpatient Pooled risk estimates were calculated using a random effects model.…”