Chua et al. evaluated the diagnostic ability of numerous optical coherence tomography (OCT) metrics in distinguishing eyes with early, moderate, and advanced glaucoma from normal eyes. The authors combined OCT data from three different prospective studies conducted in Singapore to compare the ability of macular thickness in different elliptical sectors and circumpapillary retinal nerve fiber layer (RNFL) thickness in detecting disease. They employed a standardized OCT layer segmentation program to accurately measure layer thickness, while accounting for various potential confounders using multivariable regression.
Using area under the receiver operating curve (AUC), the authors determined that the two best parameters to differentiate normal eyes from diseased eyes were circumpapillary RNFL and macular ganglion cell layer (GCL) in early glaucoma, circumpapillary RNFL and macular GCL-inner plexiform layer (IPL) in moderate glaucoma, and circumpapillary RNFL and macular GCL-IPL in advanced glaucoma. The authors also evaluated combining macular and circumpapillary RNFL, which exhibited improved diagnostic performance for early and moderate glaucoma, but not in advanced disease.
This study benefited from a large sample size, with over 400 participants of different ethnicities. In addition, the methodology of segmentation analysis and thickness measurement was standardized, which significantly strengthens the conclusions of the study. By using multivariable regression, the authors also accounted for important differences between eyes, such as magnification differences due to varying axial length as well as altered thickness values due to differences in the angle between the optic nerve and fovea. The latter is vitally important in this study since this angle impacts how the sectors of the macular elliptical annulus are defined.
While control and glaucomatous eyes were matched on age and gender, it appears as though they were not matched on ethnicity. The authors mention that Chinese eyes comprised 53% of the control eyes but 92% of the glaucomatous eyes. Arguably, this difference is important, as the authors have previously demonstrated thinner RNFL in healthy Indian eyes and thicker RNFL in healthy Chinese eyes.1,2 Comparisons may thus be affected by the difference in the ethnic composition of the control and glaucoma eyes. In addition, while the sample size most certainly makes the study more robust, we wonder if combining different types of glaucoma as was done in this study (angle-closure glaucoma, primary open-angle glaucoma (POAG), and possibly normal tension glaucoma) might lead to a mixture of different patterns of macular and RNFL thinning. As is well known, the pattern of structural loss in normal tension glaucoma is more likely to affect structures closer to the fovea than in traditional POAG.
While the current paper clearly demonstrates the power of macular parameters combined with circumpapillary RNFL thickness in identifying glaucoma at different stages of disease, other publications have demonstrated these findings, admittedly in smaller cohorts.3-5 Perhaps the authors could consider which parameters are the most sensitive to distinguish between different stages of disease (early, moderate, and advanced), which could have greater clinical application. Nonetheless, the stringent methodology utilized in this study and large sample size highlight the importance of using both macular and circumpapillary OCT parameters in the diagnosis of glaucoma.