In 2006, Costello and colleagues provided the first evidence that OCT can be longitudinally performed to track RNFL thinning after acute MSON. They showed that RNFL loss tended to occur within 3–6 months after MSON onset. Moreover, they reported a threshold of RNFL thickness at 6 months (75 μm), below which RNFL measurements predicted persistent visual dysfunction at 6 months [13].

In 2010, Henderson and colleagues corroborated this strong functional-structural correlation and suggested that RNFL thinning appeared earlier, within just 1–2 months after acute MSON [16]. More importantly, they provided sample size calculations for using RNFL loss after 6 months as a suitable outcome in placebo-controlled trials of acute neuroprotection in optic neuritis [16]. Taking into account the comparison of the affected eye follow-up, adjusted for baseline fellow eye RNFL thickness and with 6-month duration, 80 % power, and 30 and 50 % treatment effects, sample sizes were 97 (95 % CI 54,197) and 35 (95 % CI 20, 71). The study design with no adjustment for fellow eye values would be less powerful since a mean of 159 (95 % CI not shown) and 58 (95 % CI not shown) patients per arm would be needed considering equal settings [16]. In 2013, Kupersmith and colleagues found that RNFL loss at 1 month was predictive of the RNFL thinning at 6 months. These results highlighted the importance of the 1-month time point for predicting the outcome after acute MSON [17].

Considering the aforementioned studies, a number of randomized clinical trials (RCTs) were set up to evaluate molecules with primary or secondary neuroprotective effects in acute MSON by using RNFL thinning as the primary outcome (Table 12.1). Most of these studies are phase II RCT with an average duration of 6 months. These studies require small sample sizes (23–63 participants per arm) for an estimated effect size between 45 and 50 %.

The first RCT that has been published is the VISION PROTECT RCT with erythropoietin (EPO) [34]. This study found that patients treated with EPO as add-on therapy to methylprednisolone had lower RNFL thinning after 4 months than those treated only with methylprednisolone. High-contrast visual acuity (HCVA) after the trial was better for those allocated to the interventional arm compared to controls, although the difference was not statistically significant [34]. They used HCVA instead of low-contrast visual acuity (LCVA), which is more effective at capturing visual dysfunction after acute MSON. As such, the use of HCVA instead of LCVA charts may have underestimated this clinical benefit.

The neuroprotection with phenytoin in MSON study found that the average adjusted affected eye RNFL thickness was 7.15 μm higher in the phenytoin group than the placebo group in intention-to-treat analysis, a 30 % protective treatment effect but with no significant between-group difference in visual outcomes [36].