Purchase this article with an account.
Maria Anna Mavrommatis, Saskia Naegele, Jason Nunez, Gustavo Moraes, Robert Ritch, Donald Hood; Deep defects seen on visual fields spatially correspond well to loss of retinal nerve layer seen on OCT circle scans.. Invest. Ophthalmol. Vis. Sci. 2017;58(8):5822.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Structure-function relationships, as typically measured by correlating measures of local defects, are only fair-to-good, and have yet to meet FDA recommendation. Here, we take a different approach and examine the spatial relationship between deep defects on visual fields (VF) and deep losses in the circumpapillary retinal nerve fiber layer (cpRNFL) on optical coherence tomography (OCT) circle scans.
89 eyes with glaucoma and 24-2 VFs with MD better than –6dB also had 10-2 VFs and spectral domain OCT circle scans of the optic disc. 18 of these eyes had deep VF defects as defined by test point locations with pattern deviation values ≤-15 dB on the 10-2 and/or 24-2 VF. These points were enclosed within a contour (red and green outlines in Fig. 1A,B). Based upon RNFL bundle tracings of Jansonius et al.  and a model of the macula, the location of local damage at the disc was predicted, as shown in Fig. 1C by the red slanted lines. The location of these red lines are indicated on the OCT circle scan in Fig. 2A, along with the midpoint (red circle). The agreement between two independent observers for widths and midpoints was r =0.86 and 0.99, respectively. The region (yellow line) and midpoint (yellow circle) of the thinnest cpRNFL (i.e. deepest loss) was determined (Fig. 2A).
For 14 of the 18 eyes with deep VF defects, a region of local cpRNFL dropout was observed, examples of which can be seen in Fig. 2A-B. Interestingly, these 14 were the only eyes with local dropout of the cpRNFL in the total sample of 89. In 3 of the 18 eyes, only a relatively thin (<21um) region was present (Fig. 2C,D). The remaining eye (Fig. 2E) was the only one with a reasonably thick region. It had only a nasal step defect consisting of two deep points on the 24-2. Finally, the midpoint of the VF defect (red circle) and the location of deepest cpRNFL (yellow circle) showed a correlation of 0.99 and were on average 6.1°±6.19° (median: 3.9°) of each other. Finally, the difference between them was poorly correlated with disc to fovea angle (0.127, p=0.62), suggesting that the individualized maps played a negligible role.
The results suggest that structural and functional measures show excellent agreement if assessed in nontraditional ways. 1. Weinreb, Kaufman IOVS, 2011; 2. Jansonius et al EER, 2012; 3. Hood et al, PRER, 2013.
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