IntroductionSmall increases in mean serum creatinine (SCr) were observed in studies of rheumatoid arthritis patients during tofacitinib treatment. These SCr changes were investigated and potential mechanisms explored.MethodsSCr values and renal adverse event data were pooled from five Phase 3 and two long-term extension (LTE) studies. Dose-response relationships and association with inflammation (C-reactive protein (CRP)) were explored using Phase 2 data and confirmed with Phase 3 data.ResultsIn Phase 3, least squares mean SCr differences from placebo at Month 3 were 0.02 and 0.04 mg/dl for tofacitinib 5 and 10 mg twice daily (BID) (P <0.05), respectively. During Months 0 to 3, confirmed SCr ≥33% increases over baseline were reported in 17 (1.4%; 5 mg BID) and 23 (1.9%; 10 mg BID) patients. Generally, elevations plateaued and remained within normal limits throughout Phase 3 and LTE studies. Exposure-response modeling demonstrated small, reversible effects of tofacitinib on mean SCr, and significant (P <0.05) effects of CRP on model parameters. Phase 3 data confirmed that patients with higher baseline CRP or greater CRP decreases following tofacitinib treatment had the largest increases in SCr. Across Phase 3 and LTE studies, 22 tofacitinib-treated patients had clinical acute renal failure (ARF), predominantly in the setting of concurrent serious illness.ConclusionsTofacitinib treatment was associated with small, reversible mean increases in SCr that plateaued early. The mechanism behind these SCr changes remains unknown, but may involve effects of tofacitinib on inflammation. ARF occurred infrequently, was associated with concurrent serious illness, and was unrelated to prior SCr increases.Electronic supplementary materialThe online version of this article (doi:10.1186/ar4673) contains supplementary material, which is available to authorized users.