Purchase this article with an account.
SUSY C. PACHON, Gustavo Rosa Gameiro, Gadi Wollstein, Giacinto Triolo, Jean-Claude Mwanza, Donald Budenz, Luis Vazquez, Joel Schuman; Specificity and sensitivity of optical coherence tomography (OCT) parameters to detect glaucomatous visual field progression. Invest. Ophthalmol. Vis. Sci. 2017;58(8):5837.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To evaluate the ability of OCT retinal nerve fiber layer (rNFL) parameters to detect glaucoma visual field progression using an event analysis in a longitudinal cohort.
Both eyes from 40 glaucoma patients from the Pittsburgh Imaging Technology Trial (PITT) were included in the study; inclusion/exclusion criteria were previously described.1 Humphrey 24-2 SITA-standard VF perimetry and Cirrus OCT optic nerve scans were acquired every 6 months. An event analysis was performed in which visual field progression (VFP) was defined using the Hodapp-Anderson-Parrish criteria,2 and progression in rim area (RA), average and vertical cup disc ratio (CDR and VCDR, respectively), and superior, inferior, nasal, and temporal quadrant rNFL thickness (S rNFL, I rNFL, N rNFL and T rNFL, respectively) was defined as a drop from baseline in 2 consecutive visits that exceeded tolerance limits of variability of each parameter.3 Rate of VFP and OCT rNFL thinning per individual parameter was determined using Kaplan-Meyer survival analysis and Log-rank (Mantel-Cox) test. Chi-square analysis was used to calculate the sensitivity and specificity of OCT parameters to detect VFP. Baseline VF mean deviation (VFMD) between true positives (TP) and false positives (FP), and true negatives (TN) and false negatives (FN) was compared using the Student’s t-test. P<0.05 was considered significant.
Baseline age (mean +/- standard deviation) was 71.4 +/- 8.3 years, baseline VFMD was -4.70 +/- 5.4 dB, and follow up time was 69.0 +/- 12.4 months. Kaplan-Meyer survival curves for CDR, VCDR, N rNFL, and T rNFL in glaucoma were not significantly different to control eyes and were excluded from the analysis. At 2 years, the cumulative rate of VFP was 30% compared to 30% S rNFL, 33% I rNFL, and 28% average rNFL progression (p= 0.31, 0.19, 0.88; respectively). Using Chi square analysis, I rNFL had great sensitivity (61%) and specificity (63%) to detect VF progression. Baseline VFMD was -5.82 ±4.82 dB for TP vs. -4.92 ±8.61 dB for FP (p=0.014), and -3.80 ±6.32 dB for TN vs. -8.63 ±3.78 dB for FN (p=0.009).
Cumulative glaucoma progression was similar between VF and OCT RA, S rNFL, I rNFL, and average rNFL. I rNFL had great sensitivity and specificity to detect VF progression, FPs generally had early glaucoma at baseline whereas FNs generally had advanced glaucoma at baseline.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.
This PDF is available to Subscribers Only