Placebo Response to Oral Administration in Osteoarthritis Clinical Trials and Its Associated Factors A Model-Based Meta-analysis

IMPORTANCE In osteoarthritis (OA) clinical trials, a placebo is often used as control. Therefore, a thorough understanding of the placebo response is important for guiding drug development in OA. OBJECTIVE To develop an oral placebo response model for OA. DATA SOURCES PubMed, EMBASE, and Cochrane Library databases were searched systematically from January 1, 1991, to July 2, 2022. STUDY SELECTION Randomized double-blind placebo-controlled trials of patients with primary OA were included. The interventions and placebo were administered orally. A total of 3032 trials were identified; of these, 130 (4.3%) met the inclusion criteria. DATA EXTRACTION AND SYNTHESIS Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scores, dosage form of the placebo, sample size, proportion of patients who previously used nonsteroidal anti-inflammatory drugs, publication year, intervention categories, Kellgren-Lawrence grades, proportion of White patients, duration of pain, funding source, and risk of bias were extracted. A model-based meta-analysis was used to evaluate the time course of the placebo response in OA treatment and estimate the influencing factors. For subgroup analyses, a meta-analysis with a random-effects model was used to summarize the typical values of the model parameters and their SEs. MAIN OUTCOMES AND MEASURES The primary end point was the time course of the oral placebo response on the WOMAC pain, stiffness, and function subscale scores. RESULTS The 130 trials selected for analysis included 12 673 participants (mean age, 59.9 years; 68.9% women). The baseline scores of WOMAC pain, stiffness, and function subscales were found to be significantly associated with the placebo response. The placebo response reached 90% of its maximum response between 5 and 7 weeks. The placebo responses on the WOMAC subscales were also associated with the sample size, proportion of patients who had previously used nonsteroidal anti-inflammatory drugs, intervention drugs, and publication year. CONCLUSIONS AND RELEVANCE In this study, an oral placebo response model of OA was developed that may quantitatively describe the placebo response at different baseline levels of symptoms. The findings may provide valuable references for future clinical trial design and decision-making.