Design and conduct of clinical trials in patients with osteoarthritis : Recommendations from a task force of the Osteoarthritic Research Society

H. Summary H.1. symptom modifying drugs The primary outcome variable is a specific aspect of joint pain, although a ‘signal' symptom or some measure of function may also be studied. Trials of drugs with a rapid onset of effect can be as short as 1–4 weeks but may be as long as 12 weeks. Longer trials (up to 2 years) may be needed to evaluate longer-term toxicity, determine optimal long-term dosing regimens, or establish long-term benefit. Supplemental escape analgesia should be minimized, monitored and discontinued prior to evaluation of efficacy. Some agents that provide symptom relief may not provide benefit until weeks after initiation of therapy. Under these circumstances, trials will vary from 3–12 months in length. If the agent is administered in courses, episodic readministration of the drug may be needed in long-term trials. Longer trials (up to 2 years) may be required to exclude toxicity or establish long-term benefit. H.2. structure modifying drugs These drugs are intended to prevent, retard, stabilize or reverse development of the morphologic changes of OA. Although this has been called ‘chondroprotection', the term is misleading and should be avoided, because all structures of the joint are involved in OA, not articular cartilage alone. The benefits of disease modifying therapy may not be apparent until years after the onset of treatment. The selection of high-risk groups may shorten the time of investigation. Improvement in symptoms (i.e., joint pain) is not a requisite for the efficacy of a drug in this category. In these studies, it may be necessary to permit concomitant use of drugs for relief of symptoms (NSAIDs, analgesics). The confounding effects of glucocorticoids and NSAIDs in these trials is not yet understood and very restricted use of IA depocorticosteroids is recommended. Demonstration of structure modification will require the use of direct measures of joint anatomy, such as radiography, particularly measurement of the radiographic joint space. As stated above, the plain radiograph is presently the most reproducible and readily available method for assessment of disease modification. Studies are needed to validate surrogate markers of disease activity, since they may help shorten Phase 2 structure modifying drug trials. As an alternative to radiography, some trials may utilize arthroscopy. As we approach the beginning of the twenty-first century, concepts of clinical trials of OA drugs are changing. Methodology and techniques for the evaluation of new agents for OA have been refined dramatically over the last decade. We look forward to the future with excitement as we anticipate the development of new agents that may alter the symptoms and course of OA. The above recommendations are intended to help us ascertain which of these new agents are effective.