Eighteen studies were included in the review (n=6,138 participants). All trials performed an ITT analysis.
Efficacy of biological agents in patients with an inadequate response to methotrexate (15 RCTs, n=4,627):
There was no significant difference in response rates at six months between older anti-TNFs and newer ones (p=0.08 compared to certolizumab and p=0.33 compared to golimumab) and so all anti-TNFs were considered as a single group.
All agents showed a significant improvement in ACR50 compared to methotrexate treatment alone (abatacept OR 3.3, 95% CI 2.4 to 4.4; three RCTs, rituximab OR 3.1, 95% CI 2.1 to 4.7; three RCTs and anti-TNF OR 6.0, 95% CI 4.7 to 7.6; 10 RCTs). There was no heterogeneity for either abatacept or rituximab (I2=0%). Heterogeneity was substantial for anti-TNF agents (I2=57%). Anti-TNF agents were associated with a significantly greater likelihood of achieving an ACR50 response at six months compared to all non anti-TNF agents combined (OR 1.55, 95% CI 1.12 to 2.16) and for abatacept (OR 1.81, 95% CI 1.23 to 2.65). There was no significant difference compared to rituximab. Tocilizumab was significantly more effective than abatacept (OR 1.97, 95% CI 1.08 to 3.59). There was no significant between tocilizumab and rituximab and abatacept and rituximab.
Efficacy of biological agents in patients with an inadequate response to anti-TNFs (five RCTs, n=1,511):
There was no significant difference between tocilizumab, rituximab, abatacept and golimumab in the likelihood of response to ACR50.