Data collected with a tracking scanning laser ophthalmoscope suggests that the small, jerk-like, involuntary eye movements that occur during prolonged visual focus can provide measurements of disability and disease worsening in MS.
Christy Sheehy, PhD
New study results suggest that monitoring fixational microsaccades—small, jerk-like, involuntary eye movements that occur during prolonged visual focus—can provide an objective measurement of disability levels and disease worsening in multiple sclerosis (MS).
The data was compiled by Christy Sheehy, PhD, postdoctoral scholar of neurology, UCSF Weill Institute for Neurosciences, and colleagues, using a custom-built retinal eye tracker, dubbed the tracking scanning laser ophthalmoscope (TSLO). They recorded fixation in 111 participants with MS and 100 unaffected controls.
Ultimately, a greater number of microsaccades showed significant association with a higher Expanded Disability Status Scale (EDSS) score (P <.001), as well as higher non-dominant 9-hole peg test (9HPT; P = .006) scores, Symbol Digit Modalities Test (SMDT; P = 0.014), and Functional Systems Scores (FSS) including brainstem (P = .005), cerebellar (P = .011), and pyramidal (P = .009) aspects.
“With microsaccade generation, timing, and execution controlled primarily by the brainstem and cerebellum, and the overall bulk motor function assessed at the bedside through the pyramidal FSS testing, fixational eye motion could provide insight into motor function at the micron scale prior to ambulatory worsening,” Sheehy and colleagues detailed. “The lack of associations between more microsaccadic metrics and the visual FSS is likely due to our exclusion of participants with best-corrected visual acuity worse than 20/40, with only 6% of the participants having an acuity worse than or equal to 20/30.”
Sheehy et al. also wrote that interesting, the microsaccadic data increasing with EDSS scores “clearly illustrates the weaknesses of current clinical tools and the categorical EDSS scoring system in identifying ocular motor deficits early on in patients with low disability.”
Additionally, both brainstem FSS and patient-reported fatigue showed significant associations with microsaccade number, amplitude, and peak acceleration. Those with MS showed a statistically different average number (P = .020), peak vertical acceleration (P = .003), and vertical amplitude (P <.001) compared to controls.
In total, the MS cohort consisted of 76 patients with relapsing MS, 30 with progressive MS, 1 with radiologically isolated syndrome (RIS), and 4 with unknown MS subtype. Of the total cohort, 49 were being treated with a high-efficacy disease-modifying therapy (DMT) and 22 with low-efficacy DMT. The Median EDSS score at baseline was 3 (range, 2—5).
The investigators built a pair of logistic regression models with EDSS as the outcome variable, the first of which utilized eye motion only, and the second which paired eye motion with paraclinical tests. The highest performing version of the first model determined EDSS as equal to the average number of microsaccades plus the average peak acceleration, with an ROC area under the curve of 79.6% (specificity, 0.938; sensitivity, 0.47). The highest performing version of the second model determining EDSS as equal to the 25-foot walk test plus the average number of microsaccades plus the peak vertical acceleration, with an ROC area under the curve of 87.1% (specificity = 1; sensitivity, 0.47).
The group also noted that there are some limitations to the TSLO despite its sensitivity, accuracy, and noninvasive nature. Most markedly being that its motion traces can suffer from reference frame artifacts, though the group combatted this by establishing a frame of reference with offline eye-tracking software. Secondly, Sheehy and colleagues acknowledged that the optical design used to maximize resolution can cause a rotation of the imaging raster, although all 211 participants had data recorded with static TSLO alignment, providing an equal amount of rotation (roughly 7° counterclockwise). Finally, they noted that the cross-sectional cohort may underestimate the magnitude of the relationship between measurements, and thus may be incapable of establishing the true predictive capacity of TSLO.
Sheehy CK, Bensinger EES, Romeo A, et al. Fixational microsaccades: A quantitative and objective measure of disability in multiple sclerosis. Mult Scler J. 2020;26(3):343—353. doi: 10.1177/1352458519894712