Determination of serum biomarkers in osteoarthritis patients: a previous interventional imaging study revisited

McAlindon T, Bartnik E, Ried JS, Teichert L, Herrmann M, Flechsenhar K · Journal of Biomedical Research

One marker cleared the paper's own correction threshold and the rest did not. The type II procollagen synthesis marker PIIANP rose 29.9% against 1.2% on placebo over 24 weeks, and that comparison cleared the Bonferroni threshold the authors set for themselves (p=0.0014 against 0.0125). The difference in PIIANP concentration at 24 weeks did not (p=0.0086 against 0.0083), and neither did the aggrecan epitope CS846, whose 24-week gap the abstract reports at p=0.045 while the results section records it as failing the correction threshold of 0.0083. The markers did not track the imaging: 'Consistent correlations of changes in biomarkers PIIANP and CS846 with changes of the dGEMRIC score could not be observed.' Combined-group correlations of 0.34 and 0.36 at 24 weeks turn negative by 48 weeks.

Population

Serum biomarker re-analysis of the same 30 participants from the dGEMRIC trial (NCT00536302)

Intervention

Collagen hydrolysate

Comparator

Placebo

Limitations

Post-hoc re-analysis of a 30-person pilot, so it inherits every limitation of the parent trial. Five of the six authors are employees of Sanofi-Aventis Deutschland GmbH, and the parent trial was sponsored by GELITA using that company's own brand of collagen hydrolysate. With no marker tracking it, the imaging result has no biochemical correlate, and the parent trial's dGEMRIC advantage had itself already disappeared by 48 weeks.