advertisement
Advances are underway for polarimetric imaging in glaucoma
Polarimetric imaging has garnered clinical interest especially in the field of glaucoma owing to its distinct property that allows for characterizing the parallel alignment of molecules in the living ocular tissues, such as peripapillary retinal nerve fibers (pRNFs), which, were studied recently by Parakkel and colleagues. In their report, these authors quantified pRNF retardance (actually full thickness retinal retardance, a minor difference) in early, moderate and severe glaucoma (primary open-angle glaucoma, POAG) using their prototype polarization-sensitive OCT (PS-OCT) system. One strength of their study, self-designated as a pilot study, was its relatively large sample size, including 80 healthy eyes of 49 participants comprising the control group and 90 eyes of 68 patients diagnosed with glaucoma. They also conducted intra- and inter-visit repeatability analysis in subsets of these cohorts. The study also benefitted from a formal comparison to standard clinical SD-OCT imaging (measurement of pRNFL thickness), performed using a commercially available CIRRUS HD-OCT 6000 system. One important limitation of the study was a large difference in the age of the control groups versus the study cohort (43 ± 16 years versus 67 ± 9 years, respectively), although, post-hoc analyses included adjustments for age, gender and axial length. Their results demonstrated excellent repeatability of retardance measured using their prototype PS-OCT system, as well as a very strong correlation between retardance and pRNFL thickness. Both retardance and thickness provided very high levels of diagnostic accuracy, with AUC values of 0.98 and 0.97, respectively. Retardance exhibited a slightly stronger correlation with visual field mean deviation (VF-MD) than thickness, overall, and among the sub-group of eyes with severe POAG (VF-MD values worse than -12 dB).
However, among eyes with early and moderate stage POAG, structure-function correlations were generally weaker for both retardance and thickness, and not statistically significantly different from each other. The authors argued that the stronger structure-function correlation for retardance among eyes with worse glaucoma was likely due to its reduced dependence on image segmentation accuracy, especially at the locations of larger blood vessels.
One drawback of this study was its lack of measurement of depth-resolved birefringence (a lost opportunity given the technological advance over prior clinical instruments used for measurement of retardance, e.g., scanning laser polarimetry as implemented in the GDx device). Thus, the field still lacks unequivocal evidence from clinical studies that birefringence abnormalities are a harbinger of subsequent RNFL thinning in neurodegenerative diseases affecting the optic nerve, such as glaucoma, as predicted decades ago by Knighton and colleagues1-3 and demonstrated in non-human primate experimental models of glaucoma and optic nerve injury.4-7 In this regard, it is slightly disappointing to find that the results of this most recent study by Parakkel et al. reaffirm that, despite significant technological development, the diagnostic performance of PS-OCT remains similar to OCT-based RNFL thickness measures.
It is slightly disappointing to find that the results of this most recent study by Parakkel et al. reaffirm that, despite significant technological development, the diagnostic performance of PS-OCT remains similar to OCT-based RNFL thickness measures.
However, it is quite intriguing that Parakkel et al. found a stronger correlation between retardance and VF-MD in the eyes with severe glaucoma. If true, this could lead to an important implication in the late stage of glaucoma along with further development of in vivo polarimetry. Future work with PS-OCT is needed to address costs and complexities while finding applications that clearly benefit patients.