In this study, we summarized the characteristics of 425 RA clinical trials registered on ClinicalTrials.gov during the past decade.
Reduced clinical trials, RA drug development dilemma
Over the past decade, there has been a notable decrease in the number of RA clinical trials, with declining from 54 trials in 2013 to 32 trials in 2023, which partly reflected the challenge of RA drug development. Drug development of RA is challenged by the limited understanding of the pathogenetic mechanisms underpinning RA clinical heterogeneity and the specific pathways driving disease in different patients, along with the lack of biomarkers to predict response to RA drugs[12–14]. It helps that certain pioneering drug products for the crowded targets have proven transformative for patients, such as TNF inhibitors, JAK inhibitors, and IL-6/IL-6R agents, however, approximately 40% of patients exhibit no response to individual DMARDs, including bioDMARDs and tsDMARDs, which called ‘ceiling’ treatment response[15]. The declined number of industry-funded RA trials illustrated part of the dilemma in current RA clinical trials. The ‘ceiling’ treatment response, together with the high cost and complexity of RCTs, pharmaceutical companies have hesitated to allocate resources to extensive trials that may only yield response rates comparable to existing drugs, and in some instances they have de-prioritized developing RA drugs[16]. The advancement of novel drugs with diverse targets (BTK, PD-1, GM-CSF et al) or alternative technologies (cell therapy, bispecific antibody, ADC, nano antibody) were aspired to surmount the present impasse.
Excessive industry sponsorship, delayed published results
About three quarters of the RA trials were industry-funded. Previous studies concluded that industry support appeared to be associated with differences in trial design, results, and reporting[17]. For example, Industry sponsorship bias comparator selection, favoring products from the same company, and leads to more favorable efficacy results and conclusions than sponsorship by other sources[18, 19]. Many clinical trials are often either never published or published after a significant delay. We also identified that only half of RA trials yielded results available for at least 2 years after completion during past decade. Although we did not explore the publication rate and time to publication, khan’s study [10] demonstrated that the estimated median time to publication was 38 months in RA field, which is faster than 47 months in oncology[20], but slower than 25 months in neurology[21]. Delays in clinical trial publication hinders an important pathway for accelerating new treatments into clinical practice, policies that enable early preprint release or public posting of completed data analysis should be pursued.
Indistinct RA population, unspecified upper age limit
Among all included study populations, 28% clinical trials didn’t define distinct RA patients, and the enrolled participants were not limited to or didn’t clarify specific disease activity or treatment histories. On the one hand, loose eligibility criteria may create substantial facilitators to patient access to novel therapies, promote trial recruitment and completion, and increase generalizability of trial results. However, on the other hand, synchronizing for disease stage (for example, early, established or late disease) or disease activity (for example, low-to-moderate, moderate-to-high) and taking an appropriate patient treatment history, are already helping to better tailor treatments to the right patient[22]. Moreover, Efficacy and tolerability may differ between DMARD-naïve patients and patients with an inadequate response to one or more DMARDs (biological and synthetic), as well as among patients with different levels of disease activity, the indistinct enrollment might misestimate drug efficacy and safety. Furthermore, certain trials incorporated participants with active RA without providing a clear definition of what constitutes "active RA". The utilization of diverse criteria for defining active RA across studies may introduce potential bias in the results[23].
Meanwhile, we found that 38% of the trials included population with no upper age limit, which leaves the inclusion of elderly participates over 65 years in the study ambiguous. Although there were no upper age limits specified for inclusion in the clinical trial protocol, the inclusion of elderly patients in the actual implementation process could be deemed acceptable; however, disappointingly, this was not the case in reality, elderly patients were often excluded from RA RCTs[9, 24]. It is unfair to exclude older RA patients from clinical trials solely based on age. Firstly, the rise in the number of people aged 65 years and older living with rheumatoid arthritis, as well as the complex management of multimorbidity, polypharmacy and geriatric syndromes, have brought the elderly rheumatoid arthritis population into current research focus[25]. Secondly, older RA patients seem to have a less robust response and higher risks of adverse events to drug therapy compared with younger patients[26]. Ensuring the safety and efficacy of medication for elderly patients necessitates a high regard for the data obtained from elderly participants during the clinical trial process. Yet, the vastly underrepresented of elderly RA patients in clinical trial blocked elderly RA patients have equal access to the cutting edging treatment and receive evidence-based medication.
Few head-to-head comparisons, restricted placebo use
We found there were only 36% active-controlled trials and 50% placebo-controlled trials. Few head-to-head trials and most placebo-controlled trials were partly attributed to favoring superiority trials over equivalence trials by international regulatory agencies and much larger sample size for active comparator trials[27, 28]. Therefore, the placebo-controlled clinical trial could be considered appropriate and sometimes is the gold standard for the assessment of a new therapy. However, this is not the case in RA, as numerous medications are already approved for this indication, including promising bioDMARDs and tsDMARDs. Thus, head-to-head trials are promoted in RA field to determine the difference between a new therapy and existing drugs.
Among those placebo-controlled trials, almost half included placebo controls with variants designs, such as add-on or early escape placebo-control, to avoided ethical concerns; and half of trials using placebo without any design were conducted within 12 weeks. From an ethical standpoint, effective therapies for RA and treatment strategies prioritizing early disease control, support for restricting patient exposure to placebos or ineffective therapies for a prolonged period. Hence, the US Food and Drug Administration, Europe Medicines Agency, and the American College of Rheumatology have mandated the restricted use of placebos within 12 weeks or the inclusion of an active comparator in studies longer than 12 weeks as principles for control selection[29–31]. Our study showed that the current use of placebo in RA clinical trials was well constrained. Remarkably, large placebo responses were found in RA clinical trials, approximately 30% patients in placebo arms achieving ACR20 response and higher in RA patients with background therapy, with a rising trend, more importantly, placebo response are higher in patients continued MTX as background therapy[32, 33]. Hence, although both were placebo controls, the placebo response of an add-on placebo control and a placebo control without any design may not be the same. The placebo response is essential to interpreting treatment efficacy, particularly for RA medications with a ceiling to their therapeutic effect, and large placebo responses may explain some failure trials leading to keep effective medication from reaching the clinic. Develop new strategies to minimize the impact of placebo response in clinical trials is an essential issue.
Declining ACR20, ascending DAS28, inadequate and non-recommended treat-to-target assessment
We found that ACR20 and DAS28 were the most commonly used clinical outcomes. The primary concepts for assessing RA disease activity outcomes include state measures, such as the ACR/EULAR remission criteria; response measures, exemplified by the ACR 20/50/70 response metric; or continuous measures, like the DAS28, SDAI, or CDAI[22]. The ACR20 improvement criterium was mandated by the US Food and Drug Administration as primary endpoint in many RA drug approval trials[30]. But we found that clinical trials using ACR20 manifest a declining trajectory in contrast to the ascending trend delineated by DAS28, not complied with FDA guidance recommendations. In recent years, there has been considerable debate surrounding the identification of primary endpoint for RA clinical trials, especially whether ACR20 should still be used. First, although ACR20 remains the most powerful discriminator and as primary endpoint of choice in RA drug approval trials, especially when used at early time points, given the availability of effective RA therapies, higher levels of response are considered clinically important, such as ACR50, ACR70[11, 34]. Second, an ACR20 response is a minimal threshold in the current era, endpoints should be sensitive to change to provide better discriminatory power, such as DAS28, hybrid ACR response, and other continuous variables may be more sensitive to change and provide a more suitable alternative to ACR responder index[35, 36]. Most important, treat-to-target strategy was a recommended strategy for managing RA, but the concept of ACR20 is not consistent with treat-to-target strategy.
Treat-to-target strategy is a strict monitoring of the disease activity and adjustment of management strategies when treatment outcomes are not promising, recommended by the American College of Rheumatology (ACR), European League Against Rheumatism (EULAR), and the Asia Pacific League of Associations for Rheumatology (APLAR)[14, 37, 38]. In 2018, European Medicines Agency have issued guideline to recommend an endpoint reflecting a target disease state (ideally remission, or low disease activity (LDA) at the minimum) should be selected as the primary endpoints for RA clinical trials, which complied with the treat-to-target strategy[31]. The purpose of treatment is achieving either remission or low disease activity, preventing joint damage or its progression, ultimately leading to the amelioration of physical function. SDAI ≤3.3 or CDAI ≤2.8 is considered the remission standard according to ACR/EULAR criteria for remission[39]. Nonetheless, it is frustrated that few trials using treat-to-target assessment as primary clinical endpoint, and most of those assessed by DAS28 (84%). The main limitation of the DAS28 is a broader definition of remission, which is not recommended[40], and too lenient a target may precipitate suboptimal treatment strategies, consequently amplifying the disease burden.
This study has some limitations. First, we only identified trials registered on ClinicalTrials.gov, which may have missed trials not subject to registration requirements and evaluations conducted in other countries. Second, we only analyzed the information obtained from the protocol, and there may be discrepancies between the clinical trial protocol and the practical clinical trial conduct, and the results may be biased.