Active learning of contrast sensitivity function as a clinical endpoint in cataract disease

Abstract

Purpose : Traditional letter visual acuity does not adequately describe a patient’s visual limitations in cataract disease. There is a need for a patient-centered, functional endpoint that can be tested efficiently in clinic and appropriately reflects patients’ subjective vision. We performed a prospective, observational, IRB-approved study to evaluate contrast sensitivity using a quick contrast sensitivity function (CSF) test in eyes with nuclear sclerotic cataract.

Methods : Eyes with cataract status >/= 2+ nuclear sclerosis were included in the study. All patients had subjective visual complaints. Eyes with best-corrected Snellen visual acuity less than 20/30 or presence of additional ocular pathology were excluded from analysis. CSF was tested using the Manifold Platform (Adaptive Sensory Technology, San Diego, CA). This active learning approach estimates CSF using an information-gain strategy. The main outcome measure is a vision metric represented by the area under the CSF curve (AULCSF). Secondary outcome measures included contrast sensitivity thresholds at six spatial frequencies and high contrast acuity. Outcome measures were compared to data from 40 age-matched control eyes. A small subset of eyes that underwent cataract surgery also had qCSF testing done 1-2 weeks postoperatively.

Results : 36 eyes with cataract from 27 patients were included in the study. The mean age was 68 years ± 7. Mean visual acuity was 0.098 logMAR±0.06 (~20/25) and mean AULCSF was 0.997±0.21. Compared to age-matched controls (mean AULCSF = 1.164±0.26), we found a statistically significant reduction in CSF in cataract eyes (p<0.0001). High contrast acuity did not show a statistically significant reduction (p=0.121). In the 8 eyes that underwent cataract surgery, mean AULCSF increased from 1.05±0.23 to 1.33±0.14 (p=0.003). Following surgery, AULCSF increased an average of 32% with less than 1 line improvement (4.7 letters) in best-corrected visual acuity.

Conclusions : This active learning platform confirms a visually significant decrease in CSF in cataract eyes despite visual acuity better than or equal to 20/30. We observed a significant increase in CSF and subjective marked improvement in visual acuity with seemingly small improvement in traditional Snellen testing following cataract surgery. This demonstrates the potential for measuring treatment effects in eyes with cataract and other eye diseases using this platform.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.