Abstract
Purpose :
To examine the diagnostic power of two new forms of rapid objective perimetry in patients with multiple sclerosis (MS).
Methods :
We tested 44 MS patients aged 60.7 ± 10.1 y (mean ± SD), 13 of them with progressive disease, the remaining 31 with relapsing remitting. The 40 normal control patients were aged 60.2 ± 12.5 y. EDSS scores were assessed by J B-G who is a neurologist. We used two variants of multifocal pupillographic objective perimetry: W12 and W20, which had 12 or 20 stimuli/eye that sampled the central 60 degrees of each visual field. Both eyes were tested concurrently in 80 seconds. We used an FD-cleared prototype of the Konan objectiveFIELD Analyser (OFA). The main variable of interest was the asymmetry between eyes in the per-region response delays. We calculated % Areas under Curve (%AUC) for Receiver Operating Characteristic (ROC) plots. Separate %AUCs were calculated for the mean of the worst 4, and worst 12, per-region response delay asymmetries for each of 3 EDSS groups: < 2.5; 2.5 to <4.5; and >=4.5, compared to controls. We also estimated Effect-sizes as Hedge’s g.
Results :
The measured %AUCs increased with disease severity, reaching 94.4 ± 3.16 for EDSS3 (Table). Both methods had similar performance although W20 was better for the less severe EDSS1 patients. The Hedge’s g values indicated ‘large’ to ‘huge’ Effect-sizes. An interesting sub-analysis compared subjects who had experienced Optic Neuritis (ON, N=30) and those who had not (no-ON, N=14). The resulting %AUCs were very similar for both tests. For No-ON and ON, W12: 86.9 ± 5.24 vs 86.0 ± 4.46; and W20: 87.2 ± 5.45 vs 86.8 ± 4.36.
Conclusions :
The relatively large %AUCs and Effect-sizes for the EDSS 1 group possibly indicates an ability to diagnose and monitor disease condition in early-stage disease. The results for ON vs. no-ON suggest that the tests are more biased towards disease progression rather than the history of acute inflammation. This is similar to our earlier multifocal VEP data that used related stimuli [Annals Neurology, 57, 904-913].
This is a 2021 ARVO Annual Meeting abstract.