Sixty-four studies (13,982 participants) were included in the review: 30 randomised controlled trials (RCTs), 23 prospective and 11 retrospective observational studies. Thirty-seven studies (60%) scored higher than 70% on quality assessment. Where reported, follow-up duration ranged from six to 196 weeks.
Response: A statistically significantly worse response to treatment (measured using EULAR criteria) was observed at six months in seropositive patients with rheumatoid arthritis taking infliximab or adalimumab (OR 0.03, 95% CI 0.01 to 0.21; three studies; I²=49.6%) and at six to 12 months (OR 0.03, 95% CI 0.00 to 0.30; three studies; I²=71.1%). There were no statistically significant differences in response between patients with rheumatoid arthritis, with and without antibodies, when using the American College of Rheumatology 20 response (three studies; I²=62.8%).
Similarly, no statistically significant differences in response to treatment were observed in patients with spondyloarthropathies or inflammatory bowel disease. Seropositive patients with rheumatoid arthritis, psoriasis or inflammatory bowel disease showed a non-statistically significantly higher rate of loss of response to treatment compared to seronegative patients (OR 3.0, 95% CI 0.99 to 9.09; three studies; I²=31.8%).
Safety: Seropositive patients (regardless of condition) showed statistically significant higher rates of hypersensitivity compared to seronegative patients (OR 3.97, 95% CI 2.36 to 6.67; 16 studies; I²=62.5%). Seropositive patients with rheumatoid arthritis showed statistically significantly higher rates of discontinuation of treatment compared to seronegative patients (OR 3.53, 95% CI 1.60 to 7.82; three studies; I²=34.5%). Concomitant treatment with disease modifying antirheumatic drugs (DMARDs) statistically significantly reduced the risk of becoming seropositive (OR 0.32, 95% CI 0.25 to 0.42; 17 studies; I²=0%).
Results from meta-regression and other secondary outcomes were reported in the review. There was evidence of publication bias for response in patients with rheumatoid arthritis but no evidence of bias for other outcomes.