Antibiotic use and risk of non-Hodgkin lymphomas.

Clinical case reports have suggested that specific bacterial infections are associated with certain non-Hodgkin lymphoma (NHL) subtypes. Epidemiological case-control studies have been conducted using antibiotics as a proxy for bacterial infections, but with inconclusive results. The aim of this study was, in a cohort design, based on the unique nationwide Danish registers, to investigate the association between use of antibiotics and the risk of NHL subtypes. On the basis of the Civil Registration System, we established a cohort of the entire adult (≥ 15 years) Danish population. Information on use of antibiotics came from the Danish Drug Prescription Registry and lymphoma diagnosis from the Danish Cancer Registry. Associations were assessed by adjusted rate ratios (RRs). In total, 13,602 patients were diagnosed with one of the NHL subtypes during 51.6 million person-years of follow-up (1995-2008). We observed positive associations between use of antibiotics and plasma cell myeloma [RR = 1.11, 95% confidence intervals (CIs) = 1.00-1.24], chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) (RR = 1.32, 95% CI = 1.20-1.45), mantle cell lymphoma (MCL) (RR = 1.40, 95% CI = 1.04-1.88) and anaplastic large T-cell lymphoma (ALCL) (RR = 1.83, 95% CI = 1.00-3.36). Among these, the increased risk of CLL/SLL, MCL and ALCL, respectively, did not vary by years since use, and only the risk of CLL/SLL risk differed by number of prescriptions. While causality could not be established in our study, an intriguing positive long-term association between antibiotic use and CLL/SLL risk was observed. To what extent these findings indicate a role for bacteria in lymphoma pathogenesis requires further investigation.