The Incidence and Predictors of Lupus-like Reaction in Patients with IBD Treated with Anti-TNF Therapies

The incidence of lupus-like reactions (LLRs) in patients with inflammatory bowel disease (IBD) treated with anti-tumor necrosis factor (ATNF) has not been well defined. We aimed to characterize the features and predictors associated with LLR.We studied a cohort of adult patients with IBD treated with ATNF by a single specialist during 2009. Patients with LLR were characterized and compared with those without LLR for possible predictors.Twenty of 289 patients (6.9%) had LLR (19.9 cases per 1000 patient-years). Female gender and IBD-unclassified were more prevalent in the LLR group (85% versus 54%, P = 0.009; and 15% versus 2.2%, P = 0.018, respectively), with a hazard ratio of 3.89 (95% confidence interval = 1.12-13.55; P = 0.033) and 7.38 (95% confidence interval = 1.93-28.23; P = 0.003), respectively. ATNF duration was shorter in the LLR group (median, 1 year [interquartile range, 0-3] versus 3 years [interquartile range, 1-6.5], P = 0.005). Arthropathy was universal, followed by fatigue and dermatitis (30% each). Antinuclear antibodies were universally positive, and 16 of 20 had anti-double-stranded DNA. ATNF was discontinued in all; 8 patients required corticosteroids and 1 required hydroxychloroquine followed by complete clinical resolution (mean 7.9 ± 5.9 months). Antinuclear antibodies reverted or normalized in 7 of 16 patients (44%). Fourteen patients (70%) were switched to a second ATNF, 2 with concomitant immunomodulators, and 12 as monotherapy. One patient on ATNF monotherapy developed a second LLR and was successfully switched to a third ATNF.LLRs secondary to ATNFs are more frequent than previously reported, more common in women and IBD-unclassified. It is reversible with cessation of the culprit agent and steroids. Switching to an alternative ATNF rarely results in recurrence.