Evaluation of Spin in Clinical Trials of Mesenchymal Stromal Cells for the Treatment of Knee Osteoarthritis: A Systematic Review

Abstract

Background:

The regenerative potential of mesenchymal stromal cells (MSCs) has sparked interest in their use for knee osteoarthritis. Concurrently, there have been investigations on how data in scientific journals are reported and how they may influence readers’ interpretations, or “spin bias.” These studies are at risk for bias, given the limited number of patients and inconsistent blinding or controls. The risk of spin bias also complicates the interpretation, as results may be presented in a way that favors a particular outcome.

Purpose:

To quantify and characterize spin bias in clinical trials of MSCs for knee osteoarthritis.

Study Design:

Systematic review.

Methods:

PubMed and Embase searches were conducted using the terms “mesenchymal stem cells” or “MSCs” and “knee arthritis” or “osteoarthritis” and “therapy” or “treatment” or “regeneration.” Overall, 2 independent reviewers classified spin as high, moderate, or low and as 1 of 3 types: (1) emphasizing statistically significant results, (2) interpreting nonsignificant results as treatment effectiveness, and (3) claiming treatment benefits despite nonsignificant results. Journals were categorized as orthopaedic or nonorthopaedic. Descriptive statistics, the chi-square test, and the Fisher exact test were used to analyze the data, with alpha set at P < .05.

Results:

Among the 54 studies, spin was found in 80.0% of articles, with 14.5% having high, 25.5% moderate, and 40.0% low levels of spin. Type 1 was found in 54.5% of articles, type 2 in 18.2%, and type 3 in 29.1%. Spin was less frequently observed in the Methods section of articles compared with the abstract (17.52; P = .003). Reports on adipose-derived MSCs were associated with a higher frequency and level of spin compared with reports on MSCs from other sources (18.92; P = .026). There was no difference in the frequency of spin between orthopaedic and nonorthopaedic journals (0.48; P = .49) and no association with the impact factor (5.34; P = .07). There was no association between spin and financial disclosures (0.02; P = .577).

Conclusion:

Spin bias was present in most MSC-related trials for knee osteoarthritis, with a higher frequency among those that utilized adipose-derived MSCs. Understanding the prevalence and strategies of spin can mitigate any potential misinterpretations of study outcomes.