medwireNews: The reported efficacy of treatments for rheumatoid arthritis (RA) is consistently higher in phase 2 than phase 3 randomized controlled trials, mainly due to differences in inclusion criteria, say the authors of a systematic review and meta-analysis.
For many agents under investigation for rheumatic and other diseases, “the efficacy results of phase 3 trials have fallen short of the expectations generated by the results of respective preceding phase 2 trials,” and in many cases “regulatory approval has either not been sought or achieved, for example the fostamatinib trials in patients with rheumatoid arthritis,” write the researchers in Nature Medicine.
They explain that RA and psoriatic arthritis trials provide a “unique opportunity” to evaluate the reasons behind this discrepancy because “numerous trials” involving patients with these conditions have been conducted, and “consistent outcome measures have been used over the last 20 years.”
Daniel Aletaha (Medical University of Vienna, Austria) and colleagues analyzed data from 39 RA trials testing 17 agents and involving a total of 10,860 participants, finding that the phase 2 studies “systematically overestimated” ACR response rates compared with subsequent phase 3 results.
Specifically, mixed-model analysis demonstrated that the likelihood of achieving an ACR20 response was 39% higher, on average, in phase 2 than in phase 3 trials, while the odds of achieving ACR50 and ACR70 responses were increased by 34% and 38%, respectively, in phase 2 studies.
“Given the larger sample sizes of phase 3 studies compared to phase 2 studies,” these differences “are likely due to an overestimation of effect by phase 2 rather than an underestimation by phase 3 data,” say the researchers. They note that there was a “similar trend” when analyzing phase 2 and phase 3 psoriatic arthritis trials, but the findings did not reach statistical significance due to the small number of trials (n=12) available for analysis.
They then modeled potential reasons for the difference in ACR20 response rates between phase 2 and 3 RA trials, showing that broader inclusion criteria for active joints in phase 2 trials and the use of 28-joint counts rather than 66- or 68-joint counts were “major predictors” of the discrepancy.
And the disconnect between ACR20 response rates in phase 2 and phase 3 trials “was decreased substantially” after controlling for these variables, say Aletaha et al.
Together, these findings suggest that “[p]hase 2 trials should […] ensure to include patients with sufficiently high disease activity levels, which in the case of arthritis trials, should be at least eight tender and eight swollen joints, ideally more, while utilizing a 66/68-joint count measurement,” write the study authors.
“This will allow a balanced efficacy assessment of novel compounds from early investigative studies and strengthen the directional role of phase 2 trials in major drug development programs,” they conclude.
medwireNews is an independent medical news service provided by Springer Healthcare. © 2020 Springer Healthcare part of the Springer Nature Group