Retinal thinning differentiates treatment effects in relapsing multiple sclerosis below the clinical threshold.

Objective: To investigate retinal layer thinning as a biomarker of disease-modifying treatment (DMT) effects in relapsing multiple sclerosis (RMS).

Methods: From an ongoing prospective observational study, we included patients with RMS, who (i) had an optical coherence tomography (OCT) scan within 6 to 12 months after DMT start (rebaseline) and ≥1 follow-up OCT ≥12 months after rebaseline and (ii) adhered to DMT during follow-up. Differences between DMT in thinning of peripapillary-retinal-nerve-fiber-layer (pRNFL) and macular ganglion cell-plus-inner plexiform-layer (GCIPL) were analyzed using mixed-effects linear regression. Eyes suffering optic neuritis during follow-up were excluded.

Results: We included 291 RMS patients (mean age 30.8 years [SD 7.9], 72.9% female, median disease duration 9 months [range 6-94], median rebaseline-to-last-follow-up-interval 32 months [12-82]). Mean annualized rates of retinal layer thinning (%/year) in reference to DMF (n = 84, GCIPL 0.28, pRNFL 0.53) were similar under TERI (n = 18, GCIPL 0.34, pRNFL 0.59), GLAT (n = 24, GCIPL 0.32, pRNFL 0.56), and IFNb (n = 13, GCIPL 0.33, pRNFL 0.60) were slightly lower under S1PM (n = 27, GCIPL 0.19, pRNFL 0.42) and CLA (n = 23, GCIPL 0.20, pRNFL 0.42), and were significantly lower under NTZ (n = 47, GCIPL 0.09, pRNFL 0.24; both p < 0.001) and antiCD20 (n = 55, GCIPL 0.10, pRNFL 0.23; both p < 0.001). In patients achieving NEDA-2, observed thinning rates were lower overall, but still significantly lower under NTZ and antiCD20.

Interpretation: Applying a rebaselining concept, retinal layer thinning differentiates DMT effects even in clinically stable patients and, thus, might be a useful biomarker to monitor DMT efficacy on subclinical neuroaxonal degeneration-at least on a group level.