The placebo response in drug development. Part 2: Osteoarthritis

Placebo-controlled clinical trials are the gold standard in drug development, in part to ensure that the efficacy of a new therapy exceeds the placebo response in the indication being studied. The placebo response is a measured improvement in clinical signs and/or symptoms that occurs in patients receiving a sham or “dummy” treatment.  The placebo response is a complex psychological, biological and sociological phenomenon that confounds clinical data analysis, particularly for subjective and patient-reported outcomes. The placebo response is widely known to compromise evaluation of pain endpoints and has been suggested to contribute to as much as ~2/3 of the measured treatment effect in  pain from various etiologies1, contributing to the high rate of Phase II and III clinical trial failure in this indication2. The placebo response is, however, not limited to pain trials; in fact, the understanding of the impact of this phenomenon in a wide variety of therapeutic areas is growing steadily.  This article is part of a series that will examine the impact of the placebo response in drug development in pain and beyond.

Osteoarthritis (OA) is common musculoskeletal disease with increased incidence and prevalence associated with aging. It affects around 10 % of the population and is 10 times more common than rheumatoid arthritis (RA).  OA primarily manifests in knee, hip and hand joints as well as spondylosis. As a degenerative disease, it can evolve asymptomatically over a prolonged period (up to 20 years) before patients complain of pain and stiffness in the affected joint(s) where these symptoms can be worsened with activity. OA is a major cause of disability, loss of productivity and impaired quality of life. Although the disease presents a significant burden to patients, most do not receive the correct therapies or treatments (1). Moreover, development of pharmacological treatments, following long-term administration in randomized clinical trials (RCTs), showed large levels of uncertainty in accurate measurements of efficacy (i.e. pain functional, stiffness and quality of life improvement) compared with placebo (2). The recent FDA draft guidance recommends using classical patient-reported outcome measures that assess pain and function because there is concern about the reliability of predictions of structural parameters (3).  In RCTs, the amelioration of perceived pain by patients due to the placebo response can be so great that some experts have suggested that osteoarthritis patients may benefit from administration of placebo as a treatment option (4).

Compared to other pain conditions, OA presents a significant unmet need for effective treatments. While there is a wide spectrum of current treatments ranging from non-pharmacological (weight loss, ice therapy, orthosis, physiotherapy, electrical stimulation and ultrasound) and pharmacological agents (i.e. paracetamol (acetaminophen), non-steroidal anti-inflammatories [NSAIDs], opioid analgesics, intra-articular injection of hyaluronic acid or corticosteroids) up to surgical intervention (from arthroscopic lavage to joint replacement surgery).  

It is well-recognized that assessments of efficacy of treatments (approved and experimental) are blurred by the placebo effect. In fact, the measurement of pain reduction, functional improvement and stiffness reduction associated with the placebo effect could reach respectively up to 75, 71 and 83 % of the observed treatment efficacy (5). In addition, the range of pain reduction due to placebo administration could varied from 47% after intra-articular corticosteroid injection to 91% after joint lavage (5).  This suggests that understanding the factors influencing the placebo response would improve interpretations of the OA RCT results and improve the development of new treatments. 

The mechanism of placebo response has been studied mainly in chronic pain and Parkinson’s disease, among others. It appears to be predominantly related to expectations, verbal suggestions, previous experiences, and a reward mechanism. These mechanisms could interact at various levels to elicit some pharmacologic mediators, like endogenous opioids, dopamine release and low levels of cholecystokinin, resulting in clinical improvement in patients (6). In addition to general placebo mechanisms described in the literature, it is important to understand some specific characteristics about placebo response in OA patients. Meta-analyses of hip/knee OA characterizing the response to placebo over the time have shown that the level of joint knee pain was significantly reduced by 15mm (visual analog scale (VAS) of 0 to 100 mm) after 2 weeks of placebo administration. The decrease reached a plateau of 20 mm between 6 and 12 weeks of placebo administration. The decrease in hip pain was equivalent to that of knee OA after 2 weeks of placebo administration but did not decrease further thereafter (7). The route of administration also influenced the magnitude of the placebo response, as the effect was significantly greater after intra-articular placebo injection compared to topical placebo administration (effect size, 0.29 with 95% confidence interval between 0.04 to 0.54) (8). Therefore, the characterization of the placebo response is imperative to demonstrate the “true” treatment effect of a therapy.

At Tools4Patient, we have developed a platform solution (Placebell©™) to reduce the impact of the placebo response in clinical trials. This approach utilizes multiple factors on an individual patient basis, including demographics, medical history, baseline disease intensity and certain previously identified important psychological traits. Initial data from peripheral neuropathic pain (PNP) studies suggest that the Placebell©™ approach can reduce data variability related to the placebo response by as much as 30% in some pain indications. This initial model could be used to predict placebo response in a mixed population of PNP and OA patients, suggesting that there are common mechanisms driving placebo response in these two patient populations.  

Furthermore, implementation of Placebell©™ only requires that our proprietary psychological questionnaire be administered to patients at screening or baseline. To request our scientific whitepaper “Predicting the Placebo Response to Reduce Clinical Data Variability and De-Risk Drug Development”, please contact us.


1.        David J Hunter and Sita Bierma-Zeinstra. Osteoarthritis. Lancet 2019; 393: 1745–59

2.        Dario Gregori et al. Association of Pharmacological Treatments With Long-term Pain Control in Patients With Knee Osteoarthritis A Systematic Review and Meta-analysis. JAMA. 2018;320(24):2564-2579. doi:10.1001/jama.2018.19319

3.        Osteoarthritis: Structural Endpoints for the Development of Drugs, Devices, and Biological Products for Treatment Guidance for Industry. DRAFT GUIDANCE. August 2018. link:  FDA Draft Guidance for Industry.

4.        M. Dohertyy and P. Dieppe. The ‘‘placebo’’ response in osteoarthritis and its implications for clinical practice. Osteoarthritis and Cartilage (2009) 17, 1255e1262.

5.        W. Zhang The powerful placebo effect in osteoarthritis. Clin Exp Rheumatol 2019; 37 (Suppl. 120): S118-S123.  

6.        A. Abhishek and M. Doherty. Mechanisms of the placebo response in pain in osteoarthritis.Osteoarthritis and Cartilage 21 (2013) 1229e1235

7.        S. Reiter-Niesert, M. Boers, J. Detert Short-term placebo response in trials of patients with symptomatic osteoarthritis: differences between hip and knee. Osteoarthritis and Cartilage 24 (2016) 1007-1011.

8.        Raveendhara R. Bannuru, et all. Effectiveness and Implications of Alternative Placebo Treatments A Systematic Review and Network Meta-analysis of Osteoarthritis Trials. Ann Intern Med. 2015; 163: 365-372. doi:10.7326/M15-0623