To evaluate the foveal microstructure using high-density (24-line) radial scans and their correlation with visual recovery following macular hole surgery.
Retrospective, consecutive, interventional case series.
Forty-five eyes of 43 patients with ≥6 months follow-up following surgery were analyzed. Preoperative predictive measures evaluated included basal hole and minimum linear diameters. Outcome measures included best-corrected visual acuity (BCVA), postoperative foveal lucency horizontal and vertical size, external limiting membrane, and ellipsoid zone defect.
Mean basal hole diameter was 642 ± 330 μm and minimum linear diameter was 277 ± 161 μm. BCVA (logMAR) improved from 0.67 ± 0.23 to 0.31 ± 0.16 at 12 months ( P < .001). Foveal lucency horizontal and vertical sizes (μm) improved from 97 ± 81 and 33 ± 18 at 1 month to 26 ± 33 and 19 ± 18, respectively ( P < .05) at 12 months. External limiting membrane recovered in all cases at 1 month. Mean ellipsoid zone defect (μm) reduced from 136 ± 164 at 1 month to 32 ± 33 at 12 months ( P < .05). Preoperative basal hole diameter correlated with horizontal foveal lucency size at all time points ( P < .05). Horizontal foveal lucency size at 1 month correlated ( P < .05) with BCVA at 6 and 12 months. Basal hole diameter ≥700 μm (71% sensitivity and 70% specificity) and minimum linear diameter ≥330 μm (71% sensitivity and 70% specificity) were predictive of foveal lucency development. Cataract surgery did not influence foveal lucency resolution and no holes reopened.
Using radial scans, 71% of eyes demonstrated a foveal lucency at 1 month, whose size correlated with visual recovery. Preoperative basal hole diameter was predictive of foveal lucency development and size.
The foveal microstructure is formed by a dense compaction of cone photoreceptors and their projected axons, which are intertwined with retinal glial Müller cells and follow a centrifugal course toward the inner layers at the edges of the fovea.
Idiopathic macular hole surgery has been shown to result in successful hole closure and significant visual improvement in over 85% of cases. However, the postoperative visual acuity is occasionally poor despite successful anatomic closure. Histopathologic studies from autopsy eyes with idiopathic macular holes revealed some microstructure changes, such as persistent cysts and disruption of the photoreceptor layer. Anatomic factors that predict potential restoration of the photoreceptor microstructure and subsequent visual recovery in surgically closed macular holes have still to be fully elucidated. Recent reports have demonstrated that the postoperative status of the ellipsoid zone layer significantly correlates with the visual outcome of idiopathic macular hole surgery. Most of the studies have a follow-up period of usually 6 months or less, which is relatively short.
We have shown that high-density radial scanning with a greater foveolar scan density has superior detection rates of small full-thickness macular holes compared to standard raster volume scanning and can offer improved visualization and quantification of round geometric structures like macular holes.
In this investigation we examine the long-term changes in foveal microstructural parameters assessed using the radial scan pattern and their correlation with visual recovery following surgery for idiopathic full-thickness macular holes. We also quantify the dynamic restoration of foveal microstructures including foveal lucencies after successful idiopathic macular hole surgery and their impact on visual recovery.
This was a retrospective, cross-sectional analysis in which clinical data and spectral-domain optical coherence tomography (OCT) images were reviewed for patients diagnosed with a macular hole based on receipt of International Classification of Diseases (9th revision) code 362.54, as identified by billing records in patients undergoing vitrectomy for macular hole with and without internal limiting membrane peeling (Current Procedural Terminology code 67042) by a single surgeon (T.M.). This study was approved by the Duke University Institutional Review Board and adhered to the tenets set forth in the Declaration of Helsinki.
OCT imaging was performed with the Heidelberg Spectralis OCT unit (Heidelberg Engineering, Vista, California, USA) using the TruTrack eye-tracking system and the 24-line radial scanning (24 B-scans with 7.5 degrees interscan spacing). The “AutoRescan” function identified previous scan locations and automatically guided the OCT instrument to scan the same location again for follow-up visits. The same scanning location during follow-up is related not to the location of the fixation light but to the eye-tracking system and its reference points. The scans were manually recentered onto the foveola during acquisition in cases with eccentric fixation to correct for decentration of the radial scan away from the foveal center. This can occur because patients with full-thickness macular holes have been shown to preferentially fixate not at the foveolar center, but rather at the margin of the full-thickness neurosensory defect (located at a distance equivalent to one-half the minimum linear hole diameter from the foveolar center).
To be included in the analysis, patients had to have undergone 24-line radial scanning between March 1, 2012, and August 31, 2014, and have had 6 months or more follow-up after macular hole surgery. Exclusion criteria included prior vitreoretinal surgery, pathologic myopia, and a history of neovascular age-related macular degeneration, proliferative diabetic retinopathy, solar retinopathy, or significant ocular trauma.
We manually measured anatomic parameters before and after surgery. Using the digital caliper tool built into the OCT system software (Heidelberg Eye Explorer, Version 3.0; Heidelberg Engineering GmbH, Heidelberg, Germany), each individual preoperative radial B-scan was analyzed to evaluate the preoperative macular hole basal hole diameter, minimum linear diameter, macular hole height (maximal height of the elevated foveal retina), and inner opening diameter.
Using postoperative OCT images we manually measured the horizontal and vertical dimensions of the foveal lucency in cases where it occurred, as well as the size of the external limiting membrane and ellipsoid zone defect, and also calculated the percentage recovery, defined as the proportion of radial scans with intact external limiting membrane and ellipsoid zone. Images obtained at 1, 3, 6, and 12 months postoperatively were analyzed. At each visit, best-corrected distance visual acuity was measured using retroilluminated Bailey-Lovey Early Treatment Diabetic Retinopathy Study charts with standardized refraction and visual acuity protocols and converted to logarithm of the minimal angle of resolution (logMAR) for analysis.
In all cases, a single surgeon (T.M.) performed 23 or 25 gauge, 3-port pars plana vitrectomy including the induction of a posterior vitreous detachment when required. The posterior vitreous detachment was extended around the optic nerve, across the macula, and into the midperiphery. Peripheral vitreous removal with the vitreous cutter and scleral depression were typically performed. At the surgeon’s discretion, a dilute 50% solution of triamcinolone acetate was injected into the vitreous cavity to stain any remaining unopacified vitreous. Indocyanine green, diluted to 5 mg/mL concentration using 5% dextrose water, was used to stain the internal limiting membrane. Internal limiting membrane removal was performed at the discretion of the surgeon. Balanced salt solution (BSS; Alcon, Fort Worth, Texas, USA) was used as an irrigation solution. Fluid-air exchange was performed with either 30% SF6 or 14% C3F8 gas and sclerotomies were sutured if they were leaking. All patients were positioned 45 degrees chin down while awake postoperatively for 3–5 days.
Statistical analysis was carried out using SPSS Statistics, v 20.0 (IBM, Armonk, New York, USA). Categorical variables were reported as proportions and continuous variables as median ± standard deviation (SD). Nonparametric statistics were used: Kruskal-Wallis test and Spearman correlation coefficient.
Changes of visual acuity over time as well as association between visual acuity and the OCT findings were evaluated using linear regression analysis. P < .05 was considered significant.
Receiver operating characteristic curve analysis was performed to obtain cutoff values with optimum sensitivity and specificity of macular hole minimum linear diameter, height, basal hole diameter, and inner opening diameter to predict the development of postoperative foveal lucency.
Baseline Macular Hole Anatomic Characteristics and Postoperative Changes in Visual Acuity
A total of 45 eyes of 43 patients with registered serial 24-line radial OCT scans were analyzed. The mean age (± SD) of included patients was 68.5 ± 8.8 (range 41–90) years. Twenty-seven patients were female (63%) and 16 (37%) were male. Twenty-six patients (60.5%) were white, 16 (37.2%) were African-American, and 1 (2.3%) was Asian. Thirty-nine eyes (87%) underwent internal limiting membrane peel during macular hole repair while 6 eyes underwent pars plana vitrectomy with gas tamponade alone. Forty-three eyes (95.6%) received 30% SF6 and 2 eyes (4.4%) received 14% C3F8 intraocular gas tamponade. Four eyes (8.9%) had a detached posterior hyaloid over the optic nerve noted intraoperatively and 41 eyes underwent induction of a posterior vitreous detachment. Mean follow-up was 8.8 ± 3 months (range 6–12 months). Mean preoperative spherical equivalent was −0.91 ± 2.19 diopters (D) (range −6 to +1.25 D). Hole closure was achieved after the initial surgery in 44 of 45 eyes (97.8%) at the 1-month-postoperative visit, while 1 patient required a reoperation with a wider internal limiting membrane peel, following which the macular hole closed. The preoperative demographic, visual, and anatomic characteristics of the study population are provided in Table 1 .
|Age (y), mean ± SD (range)||68.5 ± 8.8 (41–90)|
|Sex||27 female, 16 male|
|Internal limiting membrane peeling||Yes = 39 (87%); no = 6 (13%)|
|Intravitreal gas tamponade||30% SF6 (43 eyes, 95.6%); 14% C3F8 (2 eyes, 4.4%)|
|Lens status at baseline||32 phakic, 13 pseudophakic|
|Basal hole diameter (μm), mean ± SD (range)||642 ± 329.6 (118.6–1466)|
|Minimum linear diameter (μm), mean ± SD (range)||276.5 ± 161.4 (45.9–718)|
|Macular hole height (μm), mean ± SD (range)||386.9 ± 134.5 (124.4–842)|
|Inner opening diameter (μm), mean ± SD (range)||386.7 ± 126 (64.3–638.8)|
Mean preoperative basal hole diameter was 642.1 ± 329.6 μm (range 118.6–1466 μm), minimum linear diameter was 276.5 ± 161.4 μm (range 45.9–718 μm), macular hole height was 386.9 ± 134.5 μm (range 124.4–842 μm), and inner opening diameter was 386.7 ± 126 μm (range 64.3–638.8 μm).
The logMAR best-corrected visual acuity (BCVA) improved from 0.67 ± 0.23 (range 0.2–1.3) preoperatively to 0.47 ± 0.20 (range 0.2–0.9; P = .001) at 1 month, 0.42 ± 0.25 (range 0.2–1.3; P = .0002) at 3 months, 0.39 ± 0.24 (range 0–0.9; P = .0001) at 6 months, and 0.31 ± 0.16 (range 0–0.8; P = .0002) at 12 months ( Table 2 ).
|Parameter||POM 1||POM 3||POM 6||POM 12||P Value|
|Horizontal foveal lucency (μm)||96.7 ± 80.9||68 ± 55.8||51.7 ± 48.4||25.5 ± 33.1||<.05|
|Vertical foveal lucency (μm)||32.9 ± 18.1||26.4 ± 17.1||22.3 ± 21.5||18.7 ± 18||<.05|
|EZ recovery (%)||65.03 ± 34.9||78.1 ± 28.3||81 ± 27.1||88.39 ± 17.9||<.001|
|Mean EZ defect (μm)||136.1 ± 163.5||85.1 ± 119.1||36.8 ± 34.7||31.9 ± 33.4||<.05|
External Limiting Membrane and Ellipsoid Zone Recovery
Complete external limiting membrane recovery, defined as an intact external limiting membrane seen in all 24 radial scans, was observed in all cases at 1 month. The mean percentage ellipsoid zone recovery, defined as the proportion of scans out of 24 with an intact ellipsoid zone, was 65.03% ± 34.9% at 1 month, improving to 78.1% ± 28.3% at 3 months ( P = .027), 81% ± 27.1% at 6 months ( P = .0004), and 88.39% ± 17.9% at 12 months ( P = .0001).
The mean ellipsoid zone defect, averaged over 24 radial scans, was 136.1 ± 163.5 μm at 1 month; that decreased to 85.1 ± 119.1 μm at 3 months ( P = .003), 36.8 ± 34.7 μm at 6 months ( P = .03), and 31.9 ± 33.4 μm at 12 months ( P = .03) ( Table 2 ).
Foveal Lucency and its Correlation With Preoperative Anatomic Parameters and Recovery of Visual Function
Thirty-two of 45 eyes (71.1%) had a visible foveal lucency at 1 month postoperatively visualized on 1 or more radial scans. This reduced to 25 eyes (55.5%) at 3 months, 12 eyes (26.7%) at 6 months, and 8 eyes (17.8%) at 12 months.
At 1 month, the mean postoperative horizontal foveal lucency size was 96.7 ± 80.9 μm (range 33–464 μm) and mean vertical foveal lucency size was 32.9 ± 18.1 μm (range 15.3–81.7 μm), reducing to, respectively, 68 ± 55.8 μm ( P = .02) and 26.4 ± 17.1 μm ( P = .04) at 3 months, 51.7 ± 48.4 μm ( P = .0035) and 22.3 ± 21.5 μm ( P = .0049) at 6 months, and 25.5 ± 33.1 μm ( P = .003) and 18.7 ± 18 μm ( P = .005) at 12 months ( Table 2 ).
Using linear regression analysis, there was a significant correlation between the preoperative basal hole diameter and mean postoperative horizontal foveal lucency size at 1 month (r = 0.4, P = .02), 3 months (r = 0.44, P = .01), 6 months (r = 0.5, P = .09), and 12 months (r = 0.46, P = .01) ( Figure 1 ).
The mean horizontal foveal lucency size at 1 month following surgery was predictive of BCVA at 6 months (r = 0.5, P = .04) and 12 months (r = 0.57, P = .02). There was also a significant correlation between the mean vertical foveal lucency size and BCVA at 12 months (r = 0.64, P = .007) but not at 6 months (r = 0.46, P = .06) ( Figure 2 ).
Receiver operating characteristic curves were constructed to assess the predictive value of the preoperative anatomic variables in predicting the development of a foveal lucency postoperatively following macular hole closure. The area under the receiver operating characteristic curves is provided in Table 3 . A preoperative basal hole diameter of 699.5 μm had 71% sensitivity and 70% specificity as predictive for developing a postoperative foveal lucency at 1 month ( Figure 3 ). Minimum linear diameter of 330.4 μm had 71% sensitivity and 70% specificity for developing a foveal lucency. Figure 4 provides a representative example of the development and resolution of foveal lucency postoperatively.
|Macular Hole Anatomic Parameter||AUC||SE||95% CI for AUC||P Value|
|Basal hole diameter||0.73||0.09||0.55–0.91||.019|
|Minimum linear diameter||0.72||0.09||0.54–0.9||.02|
|Inner opening distance||0.68||0.09||0.49–0.86||.066|
|Macular hole height||0.65||0.09||0.48–0.82||.12|