Background/Purpose:

A randomized, double-blind, placebo-controlled phase 1b clinical trial of sprifermin (rhFGF18) in knee OA assessed central medial tibio-femoral compartment cartilage thickness on quantitative MRI (qMRI) as structural primary endpoint and reported no statistically significant response (Lohmander et al. Arthritis Rheumatol 2014;66:1820-31). However, the drug was associated with dose-dependent effects in the lateral tibio-femoral compartment.

Methods:

Sprifermin was administered in 2 cycles of 3 weekly intra-articular injections at 10, 30, and 100 μg. MRIs were assessed semi-quantitatively (sq) by expert radiologists. The post-hoc analysis focused on structural changes in the 100 μg subgroup (n=57) and matching placebo subgroup (n=18) evaluating changes from baseline (BL) up to 12 months. Assessed were cartilage morphology (lesions), bone marrow lesions (BMLs), meniscal damage and extrusion, Hoffa- and effusion-synovitis using the semiquantitative WORMS scoring system. Analyses included multi-dimensional assessment: Delta subregional [DSR] approach (including number of subregions [SRs] showing worsening (>0), no change (0) and improvement (<0), negative values indicating improvement and positive values indicating worsening) and Delta-sum [DSM] approach (including total sum of grades for SRs showing worsening, no change and improvement). Analyses were performed on a whole knee level and separately for medial and lateral tibio-femoral joints (MTFJ, LTFJ), and patello-femoral joint (PFJ). Mann−Whitney−Wilcoxon tests assessed differences between treatment groups. P-values were not adjusted for multiple testing.

Results:

At BL, no significant differences were observed between placebo and treatment groups for all sqMRI parameters, except for whole knee cartilage morphology (29.97 ± [standard deviation] 16.00 for treatment vs 21.11 ± 14.76 vs placebo, p=0.040). For change in cartilage morphology, statistically significant differences were seen from BL to 12 months using DSR approach, with lower values on sprifermin (corresponding to less worsening) vs placebo for the PFJ (0.22 ± 0.55 vs treatment 0.02 ± 0.23, p=0.046), with similar results observed using the DSM approach. Although not statistically significant, less worsening vs placebo was observed in the LTFJ (DSR, 0.11 ± 0.32 vs treatment 0.05 ± 0.48). For change in BMLs, statistically significant changes from 6 to 12 months were observed for the whole knee, with negative values indicating improvement and positive values worsening (placebo 0.44 ± 1.20 vs treatment −0.14 ± 1.24, p=0.042), and borderline for LTFJ (0.22 ± 0.65 vs treatment -0.05 ± 0.80 p=0.064) using the DSR approach. No notable changes from BL were observed for remaining analyzed parameters.

Conclusion:

Cartilage morphology showed less worsening from BL to 12 months and BMLs showed improvement after 100 ug Sprifermin treatment from 6 to 12 months compared to placebo. The potential effect of Sprifermin on subchondral bone (BMLs) is a novel finding that warrants further exploration. This finding may be explained by altered loading as a result of cartilage surface restoration or by direct effects on bone.

ACR Meeting Abstracts - https://acrabstracts.org/abstract/structural-effects-of-sprifermin-in-knee-osteoarthritis-a-post-hoc-analysis-on-cartilage-and-non-cartilaginous-tissue-alterations/