Background:Randomized controlled trials in Systemic Lupus (SLE) have shown disappointing results for decades. Key challenges may include the heterogenous population coupled with high placebo response rates.Objectives:To evaluate trends in SLE study metrics over time and explore associations between primary endpoint failure and response in placebo/standard of care arms.Methods:Data from Phase II or III trials which enrolled ≥ 100 patients with SLE and reported SRI-4 and/or BICLA responses after a minimum of 24 weeks were included in the analysis. Sample size, recruitment rates, regional patient distributions, and results in placebo arms (at 24-36 weeks or 48-52 weeks) were examined according to the start date of each study in order to determine trends over time. Placebo group SRI-4 response rates in studies that met their primary endpoint were compared with those that did not.Results:Twenty-seven (14 phase II and 13 phase III) studies met the search criteria. Eleven of them met their primary endpoints. The study start dates ranged from Dec 2006 to Jan 2017. Mean/median total subject numbers were 461/349. Mean/median placebo subjects’ age at baseline were 39.9/39.2 and SLEDAI: 10.6/10.6. Mean/median placebo SRI-4 responses at Week 24-36 were 47.2%/45.8% and 42.8%/43% at Week 48-52. For BICLA, the rates were 40.3%/37.2% at Week 24-36 and 33.2%/33.5% at Week 48-52.As expected, lower placebo response was found in trials that met primary endpoints vs studies that did not (p=0.005). Total subject numbers and recruitment rates decreased over time while placebo SRI-4 response rates increased overall (Figure). However, there has been a greater range of placebo responses in more recent trials. Similar trends were observed in BICLA responses at Week 24-36 and 48-52, and in a corticosteroid reduction endpoint (percent of patients with reduction in steroid dose by ≥25% and to ≤7.5 mg/day prednisone/equivalent) at Week 48-52. Enrollment of patients from North America decreased while proportions of Eastern Europeans increased over time (Figure).Conclusion:High placebo response rates pose a continuing challenge in SLE studies and are associated with primary endpoint failures.Clinical trial metrics have been changing over time, with declining size and recruitment rates, possibly due to competition from increasing numbers of studies.These trends should be considered while designing and conducting future trials. Attention to site training and data quality may be particularly important to control high placebo rates, especially as trial sizes decrease.Figure.Disclosure of Interests:Ewa Olech Grant/research support from: BMS, Consultant of: Abbvie, Amgen, Remegen, Employee of: IQVIA, Speakers bureau: Abbvie, Amgen, Merck, Pfizer, UCB, Eduard van Rijen Employee of: IQVIA, Faizi Hussain Employee of: IQVIA, Gregory Dennis Employee of: IQVIA, Ali Ashrafzadeh Employee of: IQVIA, Joan T Merrill Grant/research support from: Xencor, Bristol Myers Squibb, Glaxo Smith Kline, Consultant of: Xencor, Abbvie, UCB, Glaxo Smith Kline, EMD Serono, Astellas, Remegen, Celgene/Bristol Myers Squibb, Exagen, Astra Zeneca, Amgen, Jannsen, Servier, ILTOO, Daitchi Sankyo, Lilly, Paid instructor for: Abbvie, Bristol Myers Squibb
Dirichlet-multinomial Model with Varying Response Rates over Time
