Using longitudinal data from a prospective study, investigators compared detection rates of glaucoma progression using OCT and visual field (VF) testing.

The cohort comprised 356 glaucoma suspect (GS)/preperimetric glaucoma (PPG) eyes and 153 perimetric glaucoma (PG) eyes from the multicenter Advanced Imaging for Glaucoma Study.

OCT progression was defined as a downward trend in average retinal nerve fiber layer (RNFL) or ganglion cell complex (GCC) thickness. VF progression was classified as event-based (Guided Progression Analysis) or trend-based (downward slope of mean deviation and/or visual field index).

OCT detected RNFL or GCC progression in 59.8% of GS/PPG eyes and in 62.1% of PG eyes. In contrast, the rate of progression detected using VF (event or trend) was significantly lower: 27.3% GS/PPG eyes and 41.8% PG eyes.

The discrepancy in detecting progression was most pronounced in mild glaucoma, but there were no significant differences between detection of moderate and advanced glaucoma. The rate of RNFL thinning slowed as the disease severity advanced, but the rate of GCC thinning remained steady.

In this study, multiple tests were generated at each visit so to reduce the noise in the data, eliminating a large source of error. This is not routinely done in clinical practice, so caution must be used when extrapolating data to a clinical scenario.

In determining change on the RNFL OCT, a trend in the average thickness was used which may mask a thinning in only 1 segment. It is also difficult to determine if a test result was a false-positive or negative since there is no gold standard comparison for detecting glaucoma progression. Finally, there were relatively few advanced glaucoma eyes in this study, limiting the conclusions that you can draw about testing in this subgroup.

The Venn diagrams in the article illustrate that while OCT may be superior overall in detecting glaucoma progression, in any given patient, both tests need to be done. This is because, in some patients, only VF tests can detect progression.

In order to detect as many progressors as possible, it is important to look at the RNFL and GCC data from the OCT, and to analyze the VF using both trend and event analysis methods.