Evaluation of a combined OCT index in healthy and glaucomatous eyes

Abstract

Purpose : Optical coherence tomography (OCT) provides clinicians with a large number of quantitative parameters that aid in glaucoma diagnosis. The OCT Early Glaucoma Diagnostic Structural Index (EGDSI), was previously proposed and validated in healthy and glaucomatous eyes with no associated disease providing a combined structural index for the detection of early glaucoma^1,2. In this preliminary, ongoing study, we evaluated the sensitivity and specificity of EGDSI in healthy and in a range of glaucomatous eyes that may include comorbidities.

Methods : Retrospective CIRRUS™ 6000 AngioPlex (ZEISS, Dublin, CA) OCT data were sampled from a glaucoma clinic including 76 eyes of 38 subjects that were clinically diagnosed as healthy, glaucomatous, or having other morbidities (see Table 1). Optic Disc 200x200 and Macular 512x128 cube scans from the first clinical visit were used for analyses. Mean Deviation (MD) from SITA Standard or SITA Fast 24-2 visual field data were collected from non-healthy eyes to gauge severity. Sixteen OCT summary parameters were extracted to calculate EGSDI as previously described. Sensitivity and specificity for EGDSI were calculated using 5% and 1% normative cut-off values determined from the CIRRUS reference database.

Results : For 37 healthy eyes, mean age was 56.3 (standard deviation, SD: 18.4; range: 27.1 to 85.7) years. For 39 eyes with glaucoma or other morbidities, mean age was 65.5 (SD: 13.3; range 30.8 to 85.7) years and mean MD was -5.15 (SD: 8.10; range: -28.79 to 0.35) dB. In the full cohort, sensitivities were 78.8% and 60.6% and specificities were 83.7% and 95.3% at the 5% and 1% cutoffs, respectively (see Table 2). There was an additional false positive (1 of 6) and false negative (1 of 3) due to the additional 9 morbidities.

Conclusions : In this preliminary, ongoing study, EGDSI shows comparable sensitivity to detect glaucoma as previously reported in a range of glaucomatous eyes³ despite the presence of a limited number of comorbidities, and exhibited additional Type I and II errors. Additional data sampling more comorbidities may help determine their effect on the EGDSI and its clinical utility to aid in detection of glaucoma.

References
[1] Mwanza et al. IOVS 2013; 54(13).
[2] Mwanza et al. TVST 2018; 7(2).
[3] Wilson et al. IOVS 2021; 62(6): Abstract 995.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.

 

Table 1. Occurrence of non-glaucoma morbidities by eye

 

Table 2. Sensitivity and specificity at 5% and 1% normative cutoffs