Clinical Rheumatology | 2021
Biologic disease-modifying anti-rheumatic drugs and kinase inhibitors: differences in efficacy and safety in rheumatoid arthritis
Abstract
To determine relative efficacy of two molecules, the gold standard is a well-designed head-to-head randomized clinical trial (RCT) comparing the two with a validated and clinically meaningful primary endpoint. An alternate method, which may have inherent bias complicating the interpretation of results, is comparing the efficacy of both molecules against a common comparator in a RCT using a validated primary endpoint. The biases in such comparisons are numerous: the trials compared most likely were conducted in different investigative sites and in different regions of the world with different patients and perhaps significant differences in patient baseline demographics and inclusion–exclusion criteria, amongst other confounders. Despite the potential of bias, such an analysis should provide a sense of relative effectiveness of the two molecules, especially if similar results are seen repeatedly. A third comparison is to ascertain the effectiveness of a molecule in patients who failed to adequately respond to the first molecule. This comparison is also open to similar bias as well as selection bias—why was the first molecule utilized rather than the second? Although it is fashionable to perform a meta-analysis to assess relative efficacy, meta-analyses are also hampered by significant differences in the patient populations studied as well as potentially missing critical information [1, 2]. The use of “real-world” registries for such comparisons may well also give misleading or conflicting results as there can be significant bias on multiple levels, including selection bias and missing clinically relevant data. For these reasons, it is reasonable to examine the relative clinical efficacy of Janus kinase inhibitors (jakinibs, JAKi) compared to biologic disease-modifying anti-rheumatic drugs (bDMARDs) by evaluating well-conducted and powered RCT in rheumatoid arthritis (RA), as there is a wealth of peer-reviewed data available involving head-to-head RCT with a similar comparator and RCT of jakinibs after failure of a bDMARD. The primary endpoint in a head-to-head study in RA should be a validated metric showing a meaningful clinical response or a meaningful depth of response. Several metrics have been shown to effectively discriminate between two molecules. The standard is the American College of Rheumatology response criteria (ACR20/50/70), but one can also investigate change in the Disease Activity Score in 28 joints (DAS28) and the Clinical Disease Activity Index (CDAI). The ACR20 has been shown to be the most suitable metric discriminating an effective medication from placebo [3]. In comparing two effective medications to each other, the most suitable discriminators are deeper responses such as the ACR50/70, CDAI reduction of 85%, or comparison of the reduction of the DAS28 [4]. One can also assess the percent of patients achieving a disease state such as CDAI, Simple Disease Activity Scale (SDAI), or DAS28(ESR) remission or low disease activity [4]. Although there is a discrepancy in the cut-points for LDA and remission using the DAS28(CRP) versus the DAS28(ESR), many studies use the same cut-points, although inappropriately [5]. We will focus on four targeted synthetic disease-modifying anti-rheumatic drugs (tsDMARDs, jakinibs) approved for the treatment of RA: tofacitinib, baricitinib (2 mg), and upadacitinib were approved in the USA; these 3, including baricitinib 4 mg, and filgotinib were approved by the European Medicines Agency and in Japan. There are multiple biologic disease-modifying anti rheumatic drugs (bDMARDs) approved worldwide: tumor necrosis factor inhibitors (TNFi, etanercept, adalimumab, infliximab, certolizumab pegol, and golimumab), Interleukin 6 inhibitors (IL6i) (tocilizumab and sarilumab), the co-stimulatory molecule (abatacept), and a B cell modulator (rituximab). * Roy Fleischmann [email protected]