Treatment of optic neuritis with erythropoietin: impact on white matter and optic nerve MRI characteristics.

We previously conducted the TONE study (Treatment of optic neuritis with erythropoietin, NCT01962571), which analysed 103 patients with acute optic neuritis and no prior history of multiple sclerosis after randomization to receive high-dose erythropoietin or placebo as an adjunct to methylprednisolone pulse therapy. All visual system-related outcomes yielded negative results. The goal of the current investigation is to assess whether erythropoietin administration influenced inflammatory optic nerve and cerebral lesions or optic nerve atrophy on MRI. We determined the sensitivity of MRI in detecting optic nerve lesions and assessed the influence of the methylprednisolone treatment delay. In this ancillary analysis of MRI images that were acquired during the TONE trial, we determined the optic nerve cross-sectional area, the extent of optic nerve T2 and gadolinium-enhancing lesions, the number of T2 and gadolinium-enhancing juxtacortical, periventricular and infratentorial white matter lesions, and the fulfilment of radiological dissemination-in-space and dissemination-in-time (2017 McDonald criteria). Lesion changes were analysed by non-parametrical statistical methods (Wilcoxon rank sum test and chi-square test) and the time to the occurrence of any new cerebral lesion was analysed by a Cox proportional hazards model. The impact of the time to methylprednisolone treatment was investigated through stratified analyses with a median split. We obtained baseline MRI from 77/103 patients (75%), six months' follow-up from 40/103 patients (39%), and included 52/103 patients (50%) in the time-to-event analysis. At baseline, gadolinium-enhancing optic nerve lesions were more frequent in the placebo compared to the erythropoietin group. Otherwise, baseline characteristics were balanced. We found no clinically meaningful or statistically significant differences between erythropoietin and placebo recipients in terms of lesion changes or optic atrophy. The sensitivity for detecting any optic nerve lesion was 82% at baseline and 85% at month 6. Outcomes were similar between patients whose methylprednisolone treatment was initiated early (<6 days) versus late (≥6 days) after symptom onset. In conclusion, this study provides class II evidence that high-dose erythropoietin in conjunction with methylprednisolone has no effect on the evolution of optic nerve or cerebral white matter lesions in patients presenting with acute optic neuritis.