Optic disc margin anatomy in patients with glaucoma and normal controls with spectral domain optical coherence tomography

Abstract

Objective: To characterize optic nerve head (ONH) anatomy related to the clinical optic disc margin with spectral domain-optical coherence tomography (SD-OCT).

Design: Cross-sectional study.

Participants: Patients with open-angle glaucoma with focal, diffuse, and sclerotic optic disc damage, and age-matched normal controls.

Methods: High-resolution radial SD-OCT B-scans centered on the ONH were analyzed at each clock hour. For each scan, the border tissue of Elschnig was classified for obliqueness (internally oblique, externally oblique, or nonoblique) and the presence of Bruch's membrane overhanging the border tissue. Optic disc stereophotographs were co-localized to SD-OCT data with customized software. The frequency with which the disc margin identified in stereophotographs coincided with (1) Bruch's membrane opening (BMO), defined as the innermost edge of Bruch's membrane; (2) Bruch's membrane/border tissue, defined as any aspect of either outside BMO or border tissue; or (3) border tissue, defined as any aspect of border tissue alone, in the B-scans was computed at each clock hour.

Main outcome measures: The SD-OCT structures coinciding with the disc margin in stereophotographs.

Results: There were 30 patients (10 with each type of disc damage) and 10 controls, with a median (range) age of 68.1 (42-86) years and 63.5 (42-77) years, respectively. Although 28 patients (93%) had 2 or more border tissue configurations, the most predominant one was internally oblique, primarily superiorly and nasally, frequently with Bruch's membrane overhang. Externally oblique border tissue was less frequent, observed mostly inferiorly and temporally. In controls, there was predominantly internally oblique configuration around the disc. Although the configurations were not statistically different between patients and controls, they were among the 3 glaucoma groups. At most locations, the SD-OCT structure most frequently identified as the disc margin was some aspect of Bruch's membrane and border tissue external to BMO. Bruch's membrane overhang was regionally present in the majority of patients with glaucoma and controls; however, in most cases it was not visible as the disc margin.

Conclusions: The clinically perceived disc margin is most likely not the innermost edge of Bruch's membrane detected by SD-OCT. These findings have important implications for the automated detection of the disc margin and estimates of the neuroretinal rim.

Figure 1

Schematic representation of border tissue (BT) configurations. A. Internally oblique: BT extends internally from the posterior BT/scleral junction to fuse with Bruch’s membrane (BM). B. Internally oblique with BM overhang. C. Externally oblique: BT extends externally from the posterior BT/scleral junction to fuse with BM. D. Non-oblique: BT is perpendicular relative to the scleral opening. For simplicity, the retinal pigment epithelium overlying Bruch’s membrane is not shown. BMO, Bruch’s membrane opening.

Figure 2

Co-localization of optic disc images. A. Fundus photograph (source image). B. Infrared image (target image) extracted from spectral domain optical coherence tomography raw data. C. Registered photograph after co-localization. D. Two-layer image of the infrared image (background) and registered photograph (65% transparency).

Figure 3

Registered photograph with optic disc margin delineation (green dots), and 12-clock hour radial B-scans (displayed with the same orientation), acquired in the positions indicated by black dashed lines. The green dots on the B-scans show the spectral domain optical coherence tomography structure corresponding to the clinical visible disc margin.

Figure 4

Schematic representation and spectral domain optical coherence tomography examples of optic nerve head delineations. Black dots (white arrows) in the B-scans indicate the spectral domain optical coherence tomography structure corresponding to the clinically visible optic disc margin. A. Bruch’s membrane opening (BMO) corresponding to the optic disc margin when BMO was coincident with the innermost edge of Bruch’s membrane and border tissue (BM/BT, A1 and A2), or when there was a BM overhang (A3). B. BM/BT corresponding to the optic disc margin when the innermost edges of BM and BT did not coincide. C. BT alone corresponding to the optic disc margin.

Figure 5

Polar plots with connected points showing the frequency of border tissue configuration by clock hour, for all glaucoma patients (n = 30) and healthy control subjects (n = 10). The distance from the origin at each clock hour represents the frequency of each configuration by clock hour.

Figure 6

Polar plots with connected points showing the frequency of border tissue configuration by clock hour, for the focal (n = 10), diffuse (n = 10) and sclerotic (n = 10) glaucomatous optic disc damage groups. The distance from the origin at each clock hour represents the frequency of each configuration by clock hour.

Figure 7

Polar plots with connected points showing the frequency of spectral domain optical coherence tomography structure corresponding to the optic disc margin by clock hour, for all glaucoma patients (n = 30) and healthy control subjects (n = 10). The distance from the origin at each clock hour represents the frequency of each spectral domain optical coherence tomography structure analyzed by clock hour.

Figure 8

Polar plots with connected points showing the frequency of Bruch’s membrane (BM) overhang by clock hour, for all glaucoma patients (n = 30) and healthy control subjects (n = 10). The external areas (light grey) show the frequency of BM overhang by clock hour. The internal areas (dark grey) show the frequency of Bruch’s membrane opening (BMO) identified as the optic disc margin in those subjects with BM overhang. The distance from the origin at each clock hour represents frequency.

Figure 9

Right eye of a glaucoma patient in which the presence of a clinically visible cilioretinal artery (cra) within vertical spectral domain optical coherence tomography (SD-OCT) B-scans provides evidence that the SD-OCT detected region of Bruch’s membrane (BM) overhang adjacent to it is invisible by optic disc stereo-photography. A. Optic disc photograph registered to B, infrared image. Inset in A. Magnification of the rectangular area (broken white lines) showing the clinical disc margin determined with stereo-disc photography (green dots) and BM opening determined with SD-OCT B-scans and co-localized to the disc photograph (red dots). C, D and E. Equidistant line B-scans (x, y and z in A and B, respectively). The line scan × (shown in C) is at a location where BM is intact while the line scan z (shown in E) is at a location just inside BM. These B-scans were not part of the study protocol and obtained additionally (left to right represent bottom to top in A and B). The locations of only x and z are shown in A and B for clarity. BM is continuous in x and y (shown in C and D) indicating that the scan locations are outside the neural canal. BM shows a break in z (shown in E) verifying that the central locations of this scan are inside the neural canal. A cross-section of the cra appears beneath BM in x and y (shown in C and D) and curves around BM overhang in z (shown in E). Since the cra beneath BM is visible in the photograph, it indicates that the temporal region of BM overhang is clinically transparent and invisible. Because in this region BM opening is internal to the clinical disc margin and is clinically invisible, conventional clinical examinations overestimate the amount of remaining neuroretinal rim tissue.