New OA Trial Bank Aims to Make 1+1 = >2

May 7, 2015

(OARSI2015) Osteoarthritis specialists hope researchers will help them unite individual patient data from different trials to conquer difficult questions about the disease.

Corticosteroid injections have a demonstrable benefit for knee osteoarthritis (OA), but only in the short term, and mostly for patients with severe symptoms. This statement can be made with some confidence, based on a systematic review of individual-level results from 620 subjects in seven separate randomized trials.1

Announced last week Osteoarthritis Research Society International (OARSI) annual meeting in Seattle, these results for the first pilot study from the OA Trial Bankdemonstrate what the fledgling organization intends to accomplish: Merging individual-subject data from numerous clinical trials to resolve important problems in osteoarthritis research.

Interventions in osteoarthritis have notoriously moderate impact. This is due in part to small trial sizes and patient heterogeneity, said Sita Bierma-Zienstra MD, a member of the OA Trial Bank steering group and professor of osteoarthritis at University Medical Center in Rotterdam.  By merging individual-level data from trials, the organization hopes to be able to analyze subgroups of patients and gain better insights into the effects of treatments. 

The group needs international buy-in from researchers to optimize the odds of success.

The pilot project began by paring 68 published randomized trials of glucocorticoids for knee or hip OA down to the seven that ultimately contributed data. (Fully 30 of these were duplicate publications; 36 had to be excluded for some reason, and data were no longer available for 9. Only one team didn't want to share the data.)

The effort allowed them to amass records for 620 subjects, of whom 222 got placebo, 169 got lavage only, and 142 had glucocorticoid injections. All data were transferred to Erasmus University in the Netherlands for analysis. The primary outcome was pain severity 4 weeks post-injection. Secondary outcomes were pain at 12 weeks and 12 months.

The effect size was significant (0.56 vs placebo), but only in the short term and mostly for patients with severe pain (>70 on a 0-100 scale). The team could not draw firm conclusions about effects in the subgroups defined by inflammation.

The steering committee was pleased to announce a call from the World Health Organization for public disclosure of trial data, made public a few days before the presentation at OARSI. But speakers said that a lack of long-term funding is a threat to the Trial Bank's future. Seed grants have come from OARSI, the European League Against Rheumatism (EULAR), and the Dutch arthritis foundation Reumafonds.

A review of glucosamine sulfate is under way, and analyses of exercise therapy and topical treatments for OA are in preparation. Topics for the future include nonsteroidal anti-inflammatory drugs, hyaluronic acid injections, arthroscopic interventions, braces and insoles, manual therapy, and nutraceuticals.

"I had a ... grant to do this for RA," commented David Felson MD during the discussion session. "You haven't seen many publications because it's incredibly difficult to pull off." For one thing, he said, the amount of programming involved is "enormous."

He added that gaining collaboration from commercial sponsors of trials is a challenge because they fear that independent analysis might show their product is not as effective as they claim.

A professor of medicine and epidemiology at the Boston University School of Medicine and coauthor on many OA trials, Felson also offered workarounds for this problem and a suggestion: Take time to analyze trial data for measures that best detect change in OA status.

References:

1.  van Middelkoop M1, Dziedzic KS, Doherty M, et al. Individual patient data meta-analysis of trials investigating the effectiveness of intra-articularglucocorticoid injections in patients with knee or hip osteoarthritis: an OA Trial Bank protocol for a systematic review. Syst Rev. 2013;2:54. doi:10.1186/2046-4053-2-54.