Article Text
Abstract
Background: In the FORWARD study, sprifermin demonstrated cartilage structure modification over a 2-year (Y) treatment period in patients with knee osteoarthritis (OA) [1]. Some relative treatment effect vs. placebo appeared to be maintained post treatment from Y2 to Y5, but the cartilage analysis following Y2 was performed after the analysis up to Y2 had been completed [2]. These post-treatment measurements were hence done without blinding readers to the relative acquisition order of the MRIs, involving a potential reader bias [2].
Objectives: To determine whether the cartilage accumulated during the 2Y anabolic treatment period was structurally viable and mechanically competent post-treatment (Y2→Y5). To this end, we obtained unbiased estimates of rates of cartilage change in the treatment and placebo groups post-treatment: Y2 and Y5 images were re-read with blinding to acquisition order. Focus was on whether the rate of cartilage loss differs between Sprifermin-treated vs. placebo-treated knees.
Methods: FORWARD was a 5-year, multicenter RCT [1,2]. 549 knee OA patients were randomized 1:1:1:1:1 to three once-weekly i.a. injections of placebo, 30µg sprifermin every 12 (q12M), or 6 months (q6M), or 100µg q12M or q6M (last injection at 18M). After Y2, cartilage segmentation was done on baseline (BL), 6, 12, 18, and 24M (Y2) MRIs by 7 experienced readers, with blinding to acquisition order. After Y5, Y2 and Y5 images (Figure 1) were analyzed together with blinding to relative temporal order, using segmented BL MRIs as a reference. All MRIs of each patient were analyzed by the same reader, also across treatment and post-treatment analysis. Readers were blinded to treatment group throughout. Longitudinal cartilage thickness change during Y2→Y5 was obtained for 494 participants: n=96 treated with placebo; n=99 with 30µg sprifermin q12M, n=99 with 30µg q6M, n=99 with 100µg q12M, and n=101 with 100µg q6M during the prior BL→Y2 treatment period. Femorotibial joint (FTJ) cartilage was the primary analytic endpoint; medial (MFTC) and lateral compartment (LFTC) thickness (Figure 1), and the location-independent thinning score are also reported. A one-way ANOVA was used to test whether there were any statistically significant differences between the rates of change during the post-treatment period between the 5 groups. The (annualized) rate of change of the placebo cohort during the post-treatment period (Y2→Y5) also was compared with that during the treatment-period (BL→Y2, original blinded readings), and compared statistically using a t-test.
Results: The change in FTJ cartilage thickness during the Y2→Y5 post-treatment period was -25.5µm (SD 64.1; 95%CI -32.4,-18.7) across all participants; no statistically significant (or relevant) differences (p=0.80) were observed between any of the treatment groups or placebo. Similar results were found for the medial (p=0.53) and lateral (p=0.78) compartment (Table 1). The location-independent thinning score (-0.88mm; SD 1.18; 95%CI -1.01,-0.76) also did not differ significantly between groups (p=0.91; Table 1). The annualized FTJ change during the post-treatment period (Y2→Y5) was -8.2µm (SD 21.4; 95% CI -13.8, -2.5) and that during the treatment period (BL→Y2) -10.4µm (SD 35.7; 95% CI -18.2, -2.6) in the placebo group, with no significant difference in FTJ, MFTC or LFTC between both periods.
Conclusion: FORWARD is the first study to evaluate the longer-term benefit of a potential anabolic DMOAD on cartilage structure, for 3.5 years after the last treatment [2]. All sprifermin groups lost cartilage throughout the post-treatment period, but not at greater rates than the placebo group during the same or the treatment period. These observations allow us to infer indirectly that the cartilage accumulated during anabolic sprifermin treatment appears structurally viable and mechanically competent post-treatment. This seems fulfilled to the same extent as for natural OA cartilage, as it is not lost at greater rates post-treatment than placebo-group cartilage.
REFERENCES: [1] Hochberg M et al.: JAMA. 2019; 322:1360-1370.
[2] Eckstein F et al.: Ann Rheum Dis. 2021; 80:1062-1069.
Acknowledgements: We would like to thank the FORWARD Study Team, the Clinical and MRI sites running the study, the FORWARD patients, and the Chondrometrics readers.
Disclosure of Interests: Felix Eckstein Shareholder of Chondrometrics GmbH, a company that provides professional image analysis service to researchers in academia and to the pharmaceutical industry., Employee of Chondrometrics GmbH, a company that provides professional image analysis service to researchers in academia and to the pharmaceutical industry., Consulting services to Merck KGA, Tissue Gene, Galapagos, Novartis, 4P Pharma/4 Moving, and TrialSpark/Formation Bio, Merck KGA, Nordik Bioscience, Tissue Gene, Galapagos, Novartis, NIH, FNIH, EU, BMBF, and Paracelsus Medical University, Anna Wisser Chondrometrics GmbH, Susanne Maschek Chondrometrics GmbH, Chondrometrics GmbH, Wolfgang Wirth Chondrometrics GmbH, Chondrometrics GmbH, Christoph Ladel Merck KGaA, Merck KGaA, CHL4special consulting, Merck KGaA (past), CHL4special consulting (present), Regenosine, Curnova, Charité hospital, TrialSpark/Formation Bio, ReumaNederland, Asger R. Bihlet NBCD A/S, NBCD A/S, Chris Knight Formation Bio, Formation Bio, Kenneth Somberg Formation Bio, Formation Bio, Formation Bio, Luping Zhao Formation Bio, Formation Bio
- Imaging
- Clinical Trial
- Disease-modifying Drugs (DMARDs)
- Cartilage
- Randomized controlled trial