Background & aims: Nonsteroidal anti-inflammatory drugs (NSAIDs) might cause hepatic side effects, but the frequency of these laboratory and clinical side effects is uncertain.
Methods: Searches of bibliographic databases MEDLINE and EMBASE and of public archives of the Food and Drug Administration were conducted to identify randomized controlled trials of diclofenac, naproxen, ibuprofen, celecoxib, rofecoxib, valdecoxib, or meloxicam in adults with osteoarthritis or rheumatoid arthritis that provided information on aminotransferase elevations >3 x upper limit of normal, liver-related discontinuations, hepatic serious adverse events, liver-related hospitalizations, or liver-related deaths. The proportion of patients with each of the hepatic toxicity outcomes was calculated separately by using sample size weighted pooling for each NSAID.
Results: Sixty-seven articles from the bibliographic database and 65 studies from the Food and Drug Administration archives met inclusion criteria. Diclofenac (3.55%; 95% confidence interval [CI], 3.12%-4.03%) and rofecoxib (1.80%; 95% CI, 1.52%-2.13%) had higher rates of aminotransferase >3 x upper limit of normal than placebo (0.29; 95% CI, 0.17-0.51) and the other NSAIDs (all < or = 0.43%). The 95% CIs for liver-related discontinuations of all NSAIDs except diclofenac (2.17%; 95% CI, 1.78%-2.64%) overlapped with placebo. Only 1 liver-related hospitalization (among 37,671 patients) and 1 liver-related death (among 51,942 patients) occurred, with naproxen.
Conclusions: Diclofenac and rofecoxib had higher rates of aminotransferase elevations than placebo and other NSAIDs studied. No NSAID studied had increased rates of liver-related serious adverse events, hospitalizations, or deaths.