STATIN USE AND PROGNOSIS IN 12,865 NON‐HODGKIN LYMPHOMA PATIENTS TREATED IN THE RITUXIMAB‐ERA

Abstract

Introduction: Statin use has been associated with improved cancer-specific outcomes among patients with several cancer forms, including in a few studies of lymphoma. At the same time, in vitro data suggests that statins may inhibit rituximab binding and thus impair the efficacy of rituximab. Most clinical studies have had limited power to adress this concern as well as to investigate major subtypes separately. We aimed to assess if statin use at non-Hodgkin lymphoma (NHL) diagnosis is associated with lymphoma-specific survival in NHL overall and subtypes. Methods: Using the National Swedish Lymphoma register linked with the population-based Prescribed Drug and Cause-of-Death registers, we identified all persons diagnosed with NHL from January 1, 2007, until Dec 31, 2013. Persons with any dispensing of statins during a 6-month period before diagnosis were considered statin users. We used Cox proportional hazards models to compute risk of lymphoma-specific death among statin users compared to non-users for NHL overall and subtypes, with subgroup analyses of treated and non-treated follicular lymphoma (FL). Models were adjusted for age, sex, year of diagnosis, education level, aa-IPI score and concomitant medication with anticoagulants, diuretics, beta-blockers, ACE inhibitors, calcium blockers and anti-diabetics to capture comorbidity. Results: Among 12,865 NHL patients (4,166 diffuse large B-cell, DLBCL, 1,765 FL, 772 mantle cell, MCL, 772 marginal zone, MZL, 131 Burkitt, 922 T- and NK-cell and 4,337 discordant or other lymphomas), 19% used statins at diagnosis. At end of follow-up (Sept 2013), 2,731 (21%) had died from lymphoma. We found no association between statin use and lymphoma-specific mortality for NHL overall, adjusted HR 0.94 (0.85, 1.04) or its subtypes (DLBCL 1.00 (0.86, 1,16), treated FL1.30 (0.81, 2.10), FL “watch and wait” 1.53 (0.74, 3.13), MCL 0.91 (0.62, 1.32), MZL 0.49 (0.22, 1.09), T- and NK-cell 0.92 (0.65, 1.30), discordant or other lymphomas 0.99 (0.80, 1.23)). However, for Burkitt lymphoma, we found a statistically significant decreased lymphoma-specific mortality, adjusted HR 0.09 (0.01, 0.64). Further dose-response analyses are planned. Conclusions: In this very large population-based unselected lymphoma population, statins were not associated with worse outcomes and thus appear safe to use during lymphoma treatment. Interestingly, Burkitt patients using statins had a statistically significant decreased lymphoma-specific mortality. Keywords: non-Hodgkin lymphoma (NHL); statins.