To demonstrate the posterior vitreous mobility following eye movements in patients forming macular holes, using an optical coherence tomography (OCT) eye-tracking system.
Prospective consecutive interventional case series.
Twenty-four eyes of 17 consecutive patients with idiopathic macular hole and the remaining fellow eyes were recruited. Following acquisition of baseline fundus scans, all patients were instructed to perform sets of standardized full-excursion vertical and horizontal eye movements. Immediately after each set of movements an OCT scan registered to the baseline fundus image was obtained using the eye-tracking system. Three images were then overlaid using picture-editing software, thereby documenting the changing configuration of the posterior vitreous as well as its position relative to the static retinal structures.
In patients with macular hole, 22 of 24 eyes (92%) had duplication of the posterior cortical vitreous in overlaid images. The extent of duplication (indicating mobility) was increased with increasing vitreous separation and as the stage of macular hole increased (75% in stage 0; 80% in stage 1; 100% in both stage 2 and stage 3). In all eyes in which there was a wavy configuration to the vitreous face, or in which there was a greater angle of vitreous insertion into the peripheral retina, vitreous mobility was evident on overlaid images ( P = 2.7 × 10 −17 and P = 1.7 × 10 −13 , respectively).
By dynamically evaluating patients, we offer potential mechanistic insights that may further implicate mobile vitreous and associated fluid currents in the pathogenesis and progression of macular holes.
The present understanding of the pathogenesis of idiopathic macular hole clearly implicates tractional forces in the development and progression to full-thickness macular hole. The nature of these forces is not yet fully understood. Gass proposed an early hypothesis of macular hole formation based primarily on biomicroscopic observations suggesting shrinkage of perifoveal cortical vitreous with resulting tangential traction. In his theory Müller cells in the normal fovea proliferate and migrate through the internal limiting membrane, inducing focal tangential contraction of the prefoveolar vitreous cortex.
More than 2 decades ago, optical coherence tomography (OCT) was developed, and it is now capable of routinely capturing high-resolution tomographic images of the vitreoretinal interface. This rapidly evolving technology provides a useful means by which to study the anatomic information that relates to the development of idiopathic macular holes. Several groups, including our own, have previously demonstrated convex perifoveal detachment of the posterior hyaloid in eyes with early-stage macular holes, and have also noted this finding in unaffected fellow eyes in patients with macular hole. These OCT and B-scan ultrasonography findings have contributed to an evolving theory of macular hole formation that also implicates oblique anteroposterior traction attributable to age-related perifoveal vitreous detachment rather than an exclusive tangential traction secondary to vitreous shrinkage mechanism. However, the mechanistic origin of the vitreous traction in macular hole formation is still not fully accounted for and remains controversial, as there has not yet been direct observational proof of this proposed mechanism. Guyer and Green have further suggested that the pathogenic vitreous traction may occur by 3 possible mechanisms: fluid movements, cellular remodeling of cortical vitreous, and contraction of a cellular membrane on the inner surface of the cortical vitreous. Since fibrocellular membrane fragments were found in only a minority or 10% of the surgical specimens obtained from impending macular holes, they have hypothesized that fluid currents are a primary mechanism. Johnson also presented the hypothesis in his selected literature review that dynamic traction attributable to vitreous movement associated with ocular saccades is the most important type of vitreomacular traction in macular hole formation and various other vitreomacular traction disorders. Our group recently presented montaged OCT images from the macula to the periphery and demonstrated “wavy” posterior vitreous folds or redundancy, indicating greater posterior vitreous mobility potentially suggesting less tension. However, these findings do not exclude a role for cellular shrinkage of residual cortical vitreous.
OCT has historically been a static imaging modality, and dynamic studies have not, to date, clearly documented the motion of the posterior vitreous relative to macular hole formation. In order to directly assess whether the posterior vitreous is fixed or mobile during macular hole development, we imaged and examined the posterior vitreous before and after eye movement using the TruTrack eye-tracking system present in the Spectralis OCT (Heidelberg Engineering, Vista, California, USA). This tracking system enables image registration of a select area of the fundus while allowing longitudinal imaging of the cortical vitreous. Merging of the in-register fundus images serially obtained over time clearly demonstrates that the configuration and relative position of the corresponding posterior vitreous is highly variable (resulting from movement of the vitreous body) following eye motion during macular hole development.
Patients and Study Design
This is a prospective consecutive interventional case series. Twenty-four consecutive eyes of 17 patients with a clinical diagnosis of idiopathic macular hole, but without other fundus complications or a history of ocular surgery, were enrolled in this study. Patients with mild cataract allowing OCT examination were included in this study. All investigations adhered to the tenets of the Declaration of Helsinki. This study was approved by the institutional review board of the Saitama Medical University (#11-041-01, approved on January 27, 2012). The composition of the patient population was 10 female and 7 male subjects, ranging in age from 57 to 89 years (68.9 ± 7.5, mean ± SD). All patients gave informed consent to participate in this study. All patients were examined by indirect ophthalmoscopy, contact lens slit-lamp biomicroscopic examination, fundus photography, and visual acuity testing and had a Watzke-Allen test performed. Data included identification of the study eye, date of onset of symptoms, refraction, best-corrected visual acuity, fellow eye diagnosis, classification of the macular hole, description of any vitreous (traction), and the acquired study OCT image sets.
The diagnosis and classification of macular holes was based on slit-lamp biomicroscopic examination using the criteria described by Gass, and these were supported by OCT findings. The examined groups consist of 3 eyes with a stage 1A hole, 2 eyes with a stage 1B hole, 6 eyes with a stage 2 hole, and 9 eyes with a stage 3 hole ( Table 1 ). In addition, with respect to the clinically normal fellow eyes, 4 fellow eyes had perifoveal posterior vitreous detachment with normal foveal structure and were examined in this study. Fellow eyes of patients with existing macular holes have previously been noted to have precursor findings (so-called stage 0 holes ) and are known to have a higher rate of macular hole formation than eyes of patients without macular holes.
|Stage||N||Superior Direction||Inferior Direction||Nasal Direction||Temporal Direction||Total a|
|0||4||2/4 (50%)||1/4 (25%)||1/4 (25%)||0/4 (0%)||3/4 (75%)|
|1||5||4/5 (80%)||2/5 (40%)||1/5 (20%)||2/5 (40%)||4/5 (80%)|
|2||6||6/6(100%)||5/6 (83%)||5/6 (83%)||5/6 (83%)||6/6 (100%)|
|3||9||9/9 (100%)||9/9 (100%)||9/9 (100%)||9/9 (100%)||9/9 (100%)|
|Total||24||21/24 (88%)||17/24 (71%)||16/24 (67%)||16/24 (67%)||22/24 (92%)|
Additionally, 11 eyes of 11 patients ranging in age from 52 to 81 years (68.9 ± 8.2, mean ± SD) with idiopathic epiretinal membrane were included to demonstrate and compare other macular pathology using this novel technique of tracking OCT images before and after a series of ocular saccades.
Spectral-Domain Optical Coherence Tomography
The OCT images were taken through a dilated pupil by a trained examiner without prior knowledge of clinical retinal findings. All OCT examinations were performed using spectral-domain OCT (SD OCT; Spectralis; Heidelberg Engineering, Vista, California, USA). Standardized horizontal and vertical vitreoretinal sections through the macular hole were collected for each patient. In order to examine the morphologic features of the entire posterior vitreous cortex and the vitreoretinal interface, wide-angle montage images of OCT from the macula to the periphery (approximately the equator) were obtained. Montage images were assembled using picture-editing software (Photoshop version 5.5; Adobe, San Jose, California, USA).
To determine whether the posterior vitreous is fixed or mobile during macular hole development, we imaged the posterior vitreous before and after eye movements using the TruTrack eye-tracking system of the Spectralis OCT (Heidelberg Engineering). This tracking system enabled accurate registration of fundus images over time. Following acquisition of baseline posterior fundus single scans, each patient was instructed to move his or her eyes rapidly back and forth through the extremes of horizontal gaze 10 times, at which point a single scan of the same portion of the fundus was acquired using the TruTrack eye-tracking system. The same routine was applied to vertical gaze. A set of 3 images was acquired in this way, following both horizontal and vertical eye movements. The images taken before and after horizontal and vertical eye movements were merged and overlaid using picture-editing software (Photoshop 5.5). Retinal vessels, choroidal vascular structure, and the curvature of the retinal pigment epithelium lines were used as landmarks to register and overlay the 3 images. Before and after each set of eye movements, standardized horizontal and vertical fundus sections through the macular hole were collected. Correlation between configurations and mobility of the posterior cortical vitreous was statistically evaluated by χ 2 test with commercially available software (SSRI ver. 1.02; SSRI, Tokyo, Japan). P values less than .05 were prospectively assigned to be considered as statistically significant.
The data described the configuration of detached posterior vitreous as observed by OCT montaged images and also its mobility or flexibility, as demonstrated by merged images acquired before and after scripted eye movements. A series of images is presented and, in most examples, there are 1 of 2 characteristic patterns of posterior vitreous described, either “smooth” ( Figures 1–3 ) or “wavy or scalloped” ( Figures 4 and 5 ).
In overlaid images the retinochoroidal structure coincided well in all 24 eyes examined ( Figures 1–5 ). In contrast, duplication of posterior cortical vitreous leading to a wavy configuration in overlapped images was frequently observed and was associated with vitreous body mobility and flexibility of the hyaloid face ( Figures 1 and 3–5 ). The posterior vitreous faces that coincided exactly in their overlaid images in configuration were determined to be “still” ( Figure 2 , Bottom; Figure 3 , Bottom). The configurations and flexibility of posterior vitreous by each quadrant (superior, inferior, nasal, and temporal) were judged independently, as findings were not uniformly present in all quadrants.
Stage 0 Macular Holes
In an attempt to examine precursor findings to stage 1 hole, 4 fellow eyes with perifoveal posterior vitreous detachment were examined (stage 0). There were no obvious alterations in retinal structure. In all 4 eyes, posterior vitreous was clearly delineated by SD OCT montaged images. In 2 of 4 eyes (50%), posterior vitreous cortex had a smooth configuration and was adherent to the disc, the macula, and the peripheral retina. The remaining 2 eyes (50%) had a wavy posterior vitreous configuration in at least 1 of 4 quadrants (superior, inferior, nasal, and temporal). Images taken before and after eye movements demonstrated duplication of the posterior vitreous. Duplication occurred in 2 of 4 eyes (50%) in the superior quadrant, in 1 (25%) in the inferior and nasal quadrants, and zero (0%) in the temporal quadrant ( Table 1 , Figure 1 ). In total, 3 of 4 eyes (75%) had duplicated posterior vitreous in at least 1 quadrant ( Table 1 ; Figure 1 , Bottom; Figure 2 , Bottom). Duplication of the posterior vitreous in overlaid images indicated a change in relative position of the cortical vitreous as compared to the fixed retinal structures and, hence, mobility induced by eye movements. All eyes with a “wavy” posterior vitreous configuration or with posterior vitreous inserting at a larger angle (>30 degrees) relative to the surface of the peripheral retina had duplicated posterior vitreous in the merged figures.
Stage 1 Macular Holes
Five eyes with a stage 1 macular hole developed fovea splitting that was associated with oblique anteroposterior vitreous traction ( Figure 3 ). In 3 of 5 eyes (60%) with a stage 1 hole there was shallow perifoveal vitreous separation with a wavy/scalloped contour in at least 1 of the 4 quadrants. The remaining eyes had a posterior vitreous cortex with a smooth contour similar to that shown in Figures 1 and 2 (stage 0). Overlaid images from before and after eye movements demonstrated duplication of the posterior vitreous in 4 of 5 eyes (80%) in the superior quadrant, 2 (40%) in the inferior and temporal quadrants, and 1 (20%) in nasal quadrant ( Table 1 ). In total, 4 of 5 eyes (80%) had duplicated posterior vitreous in at least 1 of 4 quadrants ( Table 1 , Figure 3 ).
Stage 2–3 Macular Holes
In stage 2 macular holes, all 6 eyes (100%) had a wavy or scalloped posterior vitreous configuration in at least 1 quadrant. Typically, vitreous separation extended more peripherally in the superior quadrant than in the other 3 quadrants. While eyes with a smooth configuration (without scalloping) tended to form a shallow, more acute angle with the peripheral retina (<30 degrees), a wavy/scalloped configuration tended to have a greater angle of insertion (>30 degrees). Merged images from before and after eye movements demonstrated duplication of the posterior vitreous in all 6 eyes (100%) in the superior quadrant and 5 eyes (83%) in the other 3 quadrants ( Table 1 ). In total, all eyes (100%) had duplicated posterior vitreous in at least 1 of the 4 quadrants imaged ( Table 1 , Figure 4 ). In stage 3 macular holes all 9 eyes (100%) had wavy posterior vitreous folds. Wavy folds were visible throughout the entire separated posterior vitreous cortex. The incidence ( Table 1 ) and the extent ( Figures 1–5 ) of posterior vitreous duplication in merged images increased with progression of macular hole stage.
All of the patients with idiopathic epiretinal membrane that were imaged had complete posterior vitreous detachment (vitreopapillary separation). The posterior vitreous face was separated anteriorly and was too peripheral to be visible in posterior merged images from before and after ocular saccade. Overlaid images did not show duplication of the epiretinal membrane and there was no change in relative position of the epiretinal membrane relative to the fixed retinal structure ( Figure 6 ).