Using the cross-sectional baseline data set of an ongoing longitudinal trial on visual parameters in MS/CIS patients we here addressed the relations between SLCLA, retinal integrity, VEP, health-related QoL, and disability. Our main findings are i) SLCLA determined at 2.5% contrast level is predicted by RNFL thickness and P100 latency; ii) QoL determined by SF36 PCS reflects predominantly ambulation and, to a lesser degree, upper extremity function but not visual function as assessed by SLCLA; iii) in contrast, visual dysfunction in MS patients, determined by SLCLA, is captured by the “general vision” and “near activities” subscores of the NEIVFQ39; and iv) implementation of SLCLA as additional component might enhance the performance of the MSFC as it seems to capture visual dysfunction as QoL-relevant aspect of disability not covered by the conventional MSFC.
In line with recent studies on retinal morphology in relation to visual disability in MS [25, 26] our data showing a significant association between SLCLA and retinal axonal integrity independent of a previous ON support the hypothesis that SLCLA reflects diffuse neurodegenerative processes in the retina, which could either be due to primary degeneration, secondary to subclinical inflammation, or both. In addition to morphological aspects, SLCLA testing also seems to capture physiologic function of the visual system as indicated by the significant association with P100 latency. The ability of SLCLA testing to capture visual deficits that have a major impact on the QoL of MS patients but are not adequately captured by standard visual assessments has been addressed in another study by the same group . When using the 25-item version of the NEIVFQ plus the 10-item neuro-ophthalmic supplement in a cohort of 167 patients with RR or progressive MS, the authors observed significant correlations between both HCVA and LCVA with both total score and most of the subscores, with “near activities” “role difficulties” and “driving” subscores showing the strongest correlations. Notably, associations between SLCLA and QoL were reported to remain significant when controlling for HCVA. Interestingly, that study also revealed significant correlations between SLCLA and the PCS of SF36 in a sub-cohort of 115 patients. In contrast, in our cohort, LCVA was reflected only by the subscores “general vision” and “near activities”, while HCVA was not related to QoL. Moreover, in our study, SLCLA did not predict QoL as determined by SF36, a difference that cannot be explained by different sample sizes. Our observation that only particular NEIVFQ subscores were correlated with SLCLA raises the question which aspects of visual demands in daily life are covered by the respective subscores. “General vision” can be considered a more common aspect that is important in daily life. Thus, already mild reduction of LCVA might easily interfere with the “general vision” aspect. The item “near activities” reflects all actions at near distance and is thus affected by tasks requiring accurate execution or ability to distinguish small or pale objects, e.g. reading, writing, or painting. Thus, correlation of this aspect with impairment of LCVA seems to be likely. In contrast, as ocular pain for example is not a typical feature in MS outside acute ON one would not expect interference with the “eye pain” subscore of the NEIVFQ. While in the study by Mowry et al.  a different measure of visual performance, a different version of NEIVFQ, and a larger sample size was used, which impedes a direct comparison with our data and may account for some of the observed discrepancies, the bottom line information of both studies are that in MS reduction of SLCLA negatively impacts QoL and that the NEIVFQ is a promising tool to capture vision-related QoL. Our data strengthen the roles of LCVA testing and NEIVFQ for evaluation of visual disability and its impact on QoL in MS.
Against this background SLCLA testing is a promising candidate to complement the current three-dimensional MSFC3 by a fourth visual component in order to compensate for its obvious limitation that it does not cover visual dysfunction. Earlier retrospective studies already suggested that such a four-dimensional MSFC4 would indeed capture relevant visual aspects of disability not covered by EDSS or standard MSFC . A recent review on disability outcome parameters in MS deemed formal inclusion of SLCLA test in the MSFC appropriate, but demanded prospective testing to confirm whether such an extension improves MSFC performance overall . Here, we evaluated the inclusion of SLCLA in the MSFC in a prospective approach. QoL was chosen as dependent variable as this parameter is increasingly considered important when evaluating disease progression and treatment effects in MS. Despite the availability of more MS-specific QoL tools we opted for the SF36 as this tools has been used for the initial validation of the MSFC and in numerous other trials [12, 14]. In keeping with the observation of Heesen at al. that in the self-concept of MS patients ambulation has the highest significance  the correlation between standard MSFC and the PCS of the SF36 in our cohort was mainly driven by TWT. Although vision was the second important function in the study by Heesen , in our cohort this observation was not reflected by SF36-determined QoL when SLCLA was included in a model controlling for MSFC3. The most likely explanation and the rationale for thereafter using the NEIVFQ is that SF36 is too insensitive a measure with respect to vision. Using NEIVFQ in the same model controlling for functions captured by standard MSFC3 revealed that the inclusion of SLCLA testing at the 2.5% (and 1.25%) level captured additional aspects of disability that affect at least some aspects of vision-related QoL, reflected by the “general vision” and “near activities” subscores. As a conclusion, these cross-sectional findings may cautiously suggest implementation of SLCLA testing in the MSFC as an option to increase the ability of the MSFC to capture MS-relevant disability.
Several limitations of our study need to be addressed. First, data are based on cross-sectional analyses of the baseline data set of an ongoing clinical trial. Whether MSFC4 is also superior to MSFC3 with respect to longitudinal intra-individual changes of LCVA remains to be evaluated; longitudinal data are underway. Second, the majority of included patients had either CIS or RRMS with a median EDSS of 2.0. Thus, our cohort represents rather mildly disabled MS patients, and more severely disabled patients or patients with progressive disease are certainly under-represented. Third, we observed a few outliers with rather unaffected SLCLA but low NEIVFQ39-scores. Having a closer look at these particular patients, mainly the NEIVFQ39 subscores “role”, “peripheral vision”, “mental health”, “eye pain” and “dependency” were the worst rated. Although these aspects are probably not closely related to MS-specific aspects of visual disability we did not find any other relevant ocular disease in these patients. Hence, we refrained from excluding these patients from the analyses.