Nowadays, one of the most common surgical procedures for idiopathic macular hole (IMH) management is based on vitrectomy with internal limiting membrane (ILM) peeling [
On the other hand, ILM peeling itself may induce visible changes of the inner retinal surface, although no changes in retinal nerve fiber layer (RNFL) thickness have been detected [
To our knowledge, only a few groups have evaluated the effect of ILM peeling on the RGCC after idiopathic macular hole surgery using the RTVue-100 SD OCT (Optovue, Fremont, CA, USA) with different results [
This study was a multicenter (
Macular analysis of the left eye and ganglion cell analysis of both eyes of a 78-year-old woman at 6 months after vitrectomy with Brilliant Blue G-assisted internal limiting membrane peeling for idiopathic macular hole. (a) OCT cross-section analysis showing cube volume and cube thickness. (b) Average, minimum, and sectorial macular thickness of the ganglion cell-inner plexiform layer of both eyes. A macular temporal defect may be appreciated in the left eye.
Demographic information collected from the clinical chart included patient age, sex, combined cataract surgery, macular hole stage, preoperative best-corrected visual acuity (BCVA), postoperative BCVA, intraocular hypertension after surgery (>25 mmHg), history of glaucoma, and failure to close the macular hole. Best-corrected visual acuity was measured using a decimal visual acuity chart, and the decimal visual acuity was converted to the logarithm of the minimum angle of resolution (logMAR) units for statistical analysis.
Three-dimensional cube OCT data were obtained with the Cirrus HD-OCT device using the Macular Cube 200 × 200 scan protocol. This protocol performs 200 horizontal B-scans comprising 200 A-scans per B-scan over 1024 samples within a cube measuring 6 × 6 × 2 mm. The GCA software (6.0 version) evaluates the thickness of the ganglion cell plus inner plexiform layers. The average, minimum, and sectorial thicknesses of the GCIPL are measured in an elliptical annulus (vertical inner and outer radius of 0.5 mm and 2.0 mm; horizontal inner and outer radius of 0.6 and 2.4 mm, resp.) around the fovea. In order to avoid segmentation errors, OCT measurements with signal strength (SS) below 5 were excluded (0: lowest SS; 10: highest SS).
All OCT images were obtained by experienced clinical technicians. Eyes were dilated with tropicamide 1% and phenylephrine 2.5%. Average GCIPL thickness, macular cube average thickness (MCAT), and macular cube volume (MCV) values of the patients included in this study were measured preoperatively, at 1 and at 6 months after macular hole surgery by scanning with the Cirrus HD-OCT system (Carl Zeiss Meditec, Dublin, CA) (Figure
The main outcome measure was the comparison of average GCIPL thickness preoperatively and at 6 months after macular hole surgery with BBG-assisted ILM peeling. Comparison of MCAT and MCV preoperatively and at 6 months after macular hole surgery with ILM peeling was the secondary outcome measures. Moreover, all values were obtained at 1 month after surgery. Average, minimum, and sectorial (superior, inferior, superonasal, inferonasal, superotemporal, and inferotemporal) GCIPL thickness values were obtained and compared in every patient preoperatively and at 1 and 6 months after surgery (Figure
Each GCIPL scan was evaluated in order to identify how many cases had a greater GCIPL thickness after surgery compared to before. This data was studied to evaluate the quality of the measurements, as the real GCIPL thickness should not be higher in the postoperative period.
A comparison between preoperative and postoperative macular GCIPL thickness values was also performed by semimanual segmentation. The Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, CA) GCIPL analysis software is not capable of real manual segmentation of the macular layers, but it does allow relocation of the area of analysis (Figure
This example shows the relocation of the area of analysis (semimanual segmentation). The center was manually displaced following the direction of the black arrow (b).
Surgery was performed using a standard 23- or 25-gauge 3-port pars plana vitrectomy. The infusion cannula was placed in the inferotemporal quadrant. If the posterior hyaloid was still attached to the optic disc, its detachment was induced by suction with the vitrectomy probe. A volume of 0.1 mL BBG (Fluoron GmbH, Ludwigsfeld, Germany) at a concentration of 0.25 mg/mL was injected into the vitreous cavity over the posterior pole for 30 seconds. The ILM was grasped at the temporal quadrant and peeled off with forceps in an area of 2-disc diameter around the macular hole. Fluid-air exchange and intraocular gas tamponade with SF6 at 20% were performed. After surgery, patients were asked to remain in a facedown position for at least 50 minutes per hour for four days. In 12 patients, the crystalline lens was removed by phacoemulsification followed by intraocular lens implantation before pars plana vitrectomy. A topical beta blocker (timolol maleate 0.5% BID) was routinely used to prevent postoperative intraocular pressure (IOP) rise.
The differences in the OCT values between the preoperative time and at 1 and at 6 months after surgery were analyzed using the paired
The study sample was comprised of 25 eyes of 25 participants (mean age 70.48 ± 8.66 years old, range: 49–82). Mean preoperative and postoperative (6 months) BCVA were 0.7 ± 0.32 logMAR units and 0.34 ± 0.32 logMAR units, respectively. The rate of closure of macular holes by OCT evaluation was 100% at 1 and 6 months after surgery. Twelve patients underwent a combined cataract surgery with pars plana vitrectomy. None of the patients had a postoperative retinal detachment. There was no recorded incidence of increased postoperative IOP above 25 mmHg. Demographic data are shown in Table
Demographic characteristics.
Parameter | Values | % |
---|---|---|
Sex | ||
Male | 11 | 44 |
Female | 14 | 56 |
Age* | 70.48 ± 8.66 | |
Eye | ||
Right | 13 | 52 |
Left | 12 | 48 |
Macular hole stage | ||
2 | 6 | 24 |
3 | 8 | 32 |
4 | 11 | 44 |
Visual acuity ( |
||
Preoperative* | 0.70 ± 0.32 | |
Postoperative* | 0.34 ± 0.32 | |
Glaucoma | ||
No | 25 | 100 |
Yes | 0 | 0 |
Macular hole closure | 25 | 100 |
Average MCV was 10.22 ± 0.81
Preoperative and postoperative average GCIPL thickness values obtained by automated segmentation were 60.72 ± 18.20
Comparison between preoperative and postoperative (at 1 month after surgery) macular GCIPL thickness values performed by automated segmentation.
|
Preoperative ( |
SD ( |
Postoperative ( |
SD ( |
Difference ( |
SD ( |
|
---|---|---|---|---|---|---|---|
Average GCIPL thickness | 60.72 | 18.20 | 61.52 | 17.37 | −0.8 | 0.83 | 0.865 |
Minimum GCIPL thickness | 33.56 | 18.20 | 43.60 | 23.83 | −10.04 | −5.63 | 0.134 |
GCIPL superior | 59.76 | 26.37 | 64.04 | 27.73 | −4.28 | −1.36 | 0.531 |
GCIPL inferior | 53.88 | 23.79 | 59.56 | 18.48 | −5.68 | 5.31 | 0.363 |
GCIPL superonasal | 59.04 | 24.77 | 65.64 | 22.01 | −6.6 | 2.76 | 0.243 |
GCIPL superotemporal | 69.48 | 14.87 | 59.20 | 20.37 | 10.28 | −5.5 | 0.076 |
GCIPL inferonasal | 54.48 | 20.62 | 62.80 | 18.90 | −8.32 | 1.72 | 0.116 |
GCIPL inferotemporal | 67.40 | 20.55 | 59.36 | 22.77 | 8.04 | −2.22 | 0.176 |
GCIPL: ganglion cell-inner plexiform layer.
SD: standard deviation.
Comparison between preoperative and postoperative (at 1 month after surgery) macular GCIPL thickness values performed by semimanual segmentation.
|
Preoperative ( |
SD ( |
Postoperative ( |
SD ( |
Difference ( |
SD ( |
|
---|---|---|---|---|---|---|---|
Average GCIPL thickness | 69.23 | 19.50 | 65.00 | 14.80 | 4.23 | 4.70 | 0.466 |
Minimum GCIPL thickness | 49.69 | 23.29 | 47.85 | 23.78 | 1.85 | −0.50 | 0.832 |
GCIPL superior | 66.31 | 21.64 | 63.15 | 19.27 | 3.15 | 2.37 | 0.610 |
GCIPL inferior | 63.31 | 22.35 | 63.54 | 16.49 | −0.23 | 5.86 | 0.973 |
GCIPL superonasal | 71.08 | 25.08 | 68.69 | 23.15 | 2.38 | 1.93 | 0.728 |
GCIPL superotemporal | 75.23 | 14.21 | 62.08 | 18.27 | 13.15 | −4.06 | 0.083 |
GCIPL inferonasal | 67.15 | 24.13 | 71.69 | 17.07 | −4.54 | 7.06 | 0.303 |
GCIPL inferotemporal | 72.00 | 19.20 | 61.46 | 17.16 | 10.54 | 2.03 | 0.198 |
GCIPL: ganglion cell-inner plexiform layer.
SD: standard deviation.
Comparison between preoperative and postoperative (at 6 months after surgery) macular GCIPL thickness values performed by automated segmentation.
|
Pre-operative ( |
SD ( |
Postoperative (6 m) ( |
SD ( |
Difference ( |
SD ( |
|
---|---|---|---|---|---|---|---|
Average GCIPL thickness | 60.72 | 18.20 | 61.20 | 14.71 | −0.48 | 3.49 | 0.912 |
Minimum GCIPL thickness | 33.56 | 18.20 | 44.20 | 19.38 | −10.64 | −1.18 | 0.053 |
GCIPL superior | 59.76 | 26.37 | 58.32 | 19.46 | 1.44 | 6.91 | 0.808 |
GCIPL inferior | 53.88 | 23.79 | 60.16 | 15.07 | −6.28 | 8.72 | 0.292 |
GCIPL superonasal | 59.04 | 24.77 | 64.28 | 20.28 | −5.24 | 4.49 | 0.353 |
GCIPL superotemporal | 69.48 | 14.87 | 59.52 | 14.83 | 9.96 | 0.04 |
|
GCIPL inferonasal | 54.48 | 20.62 | 64.08 | 15.56 | −9.6 | 5.06 | 0.080 |
GCIPL inferotemporal | 67.40 | 20.55 | 60.72 | 17.31 | 6.68 | 3.24 | 0.070 |
GCIPL: ganglion cell-inner plexiform layer.
SD: standard deviation.
Comparison between preoperative and postoperative (at 6 months after surgery) macular GCIPL thickness values performed by semimanual segmentation.
|
Preoperative ( |
SD ( |
Postoperative (6 m) ( |
SD ( |
Difference ( |
SD ( |
|
---|---|---|---|---|---|---|---|
Average GCIPL thickness | 69.23 | 19.50 | 63.77 | 10.14 | 5.46 | 9.36 | 0.241 |
Minimum GCIPL thickness | 49.69 | 23.29 | 45.08 | 16.94 | 4.62 | 6.35 | 0.499 |
GCIPL superior | 66.31 | 21.64 | 60.69 | 15.80 | 5.62 | 5.84 | 0.292 |
GCIPL inferior | 63.31 | 22.35 | 62.31 | 14.40 | 1.00 | 7.95 | 0.866 |
GCIPL superonasal | 71.08 | 25.08 | 67.85 | 17.23 | 3.23 | 7.85 | 0.590 |
GCIPL superotemporal | 75.23 | 14.21 | 60.08 | 10.70 | 15.15 | 3.51 |
|
GCIPL inferonasal | 67.15 | 24.13 | 72.54 | 11.13 | −5.38 | 13.01 | 0.313 |
GCIPL inferotemporal | 72.00 | 19.20 | 58.62 | 10.19 | 13.38 | 9.01 |
|
GCIPL: ganglion cell-inner plexiform layer.
SD: standard deviation.
Quality measurement analysis performed by automated segmentation showed that GCIPL thickness was higher in the postoperative period in around 50% of the scans (Table
Quality measurement analysis between the preoperative and postoperative macular GCIPL thickness values performed by automated segmentation.
|
Postoperative (1 m) | Postoperative (6 m) | ||||||
---|---|---|---|---|---|---|---|---|
GCIPL (< or =)* | GCIPL (>)† | GCIPL (< or =)* | GCIPL (>)† | |||||
Average GCIPL thickness | 12 | 48% | 13 | 52% | 11 | 44% | 14 | 56% |
Minimum GCIPL thickness | 9 | 36% | 16 | 64% | 9 | 36% | 16 | 64% |
GCIPL superior | 11 | 44% | 14 | 56% | 12 | 48% | 13 | 52% |
GCIPL inferior | 10 | 40% | 15 | 60% | 10 | 40% | 15 | 60% |
GCIPL superonasal | 14 | 56% | 11 | 44% | 12 | 48% | 13 | 52% |
GCIPL superotemporal | 15 | 60% | 10 | 40% | 18 | 72% | 7 | 28% |
GCIPL inferonasal | 11 | 44% | 14 | 56% | 12 | 48% | 13 | 52% |
GCIPL inferotemporal | 15 | 60% | 10 | 40% | 16 | 64% | 9 | 36% |
Average |
|
|
|
|
|
|
|
|
†GCIPL (>): number of cases where the ganglion cell-inner plexiform layer thickness is higher than before surgery.
Quality measurement analysis between the preoperative and postoperative macular GCIPL thickness values performed by semimanual segmentation.
|
Postoperative (1 m) | Postoperative (6 m) | ||||||
---|---|---|---|---|---|---|---|---|
GCIPL (< or =)* | GCIPL (>)† | GCIPL (< or =)* | GCIPL (>)† | |||||
Average GCIPL thickness | 13 | 52% | 12 | 48% | 14 | 56% | 11 | 44% |
Minimum GCIPL thickness | 13 | 52% | 12 | 48% | 13 | 52% | 12 | 48% |
GCIPL superior | 13 | 52% | 12 | 48% | 15 | 60% | 10 | 40% |
GCIPL inferior | 14 | 56% | 11 | 44% | 16 | 64% | 9 | 36% |
GCIPL superonasal | 15 | 60% | 10 | 40% | 14 | 56% | 11 | 44% |
GCIPL superotemporal | 15 | 60% | 10 | 40% | 17 | 68% | 8 | 32% |
GCIPL inferonasal | 12 | 48% | 13 | 52% | 12 | 48% | 13 | 52% |
GCIPL inferotemporal | 13 | 52% | 12 | 48% | 16 | 64% | 9 | 36% |
Average |
|
|
|
|
|
|
|
|
Nowadays, ILM peeling combined with pars plana vitrectomy is considered an effective procedure for IMH surgery [
The variables generated by the GCC measuring mode of the original software of the RTVue-100 SD OCT (Optovue, Fremont, CA, USA) employed by Baba et al. and Sevim et al. include the average, superior (0–180 degrees), and inferior (180–360 degrees) thickness of the RGCC, which comprises the retinal nerve fiber layer, the ganglion cell layer, and the inner plexiform layer. In contrast, the ganglion cell analysis (GCA) software of the Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, CA) used in this study allows obtaining the GCIPL thickness, and therefore the retinal nerve fiber layer is not included in the measurement. Furthermore, GCA software generates specific variables for each quadrant of the macular area (superior, inferior, superonasal, inferonasal, superotemporal, and inferotemporal), as well as the average and minimum GCIPL thickness. Therefore, a more specific analysis of the inner retinal thickness area may be performed with this software. Actually, we observed no changes in the average GCIPL thickness at 1 and 6 months after surgery (Tables
Based on our results, we hypothesize that as the RNFL is thinner at the temporal area than at the nasal area [
Several groups have reported nasal visual defects after ICG-assisted ILM peeling in IMH surgery [
The GCC color map provided by the RTVue software may allow identifying specific areas of focal ganglion cell loss. Actually, Baba et al. showed one case of a predominantly temporal GCC loss in the color-coded GCC thickness map reported [
On the other hand, we observed a significant decrease in the MCAT and MCV at 6 months after surgery. Some patients presented a preoperative abnormal increase of both parameters, secondary to microcystic edema usually present at the edge of the macular hole [
The study had some limitations. First, our series of patients is relatively small, in part because the GCA software of the Cirrus HD-OCT has only recently become commercially available. Therefore, future studies should be performed to validate our results. Second, longer observation periods are needed in order to evaluate the GCIPL progress over time. Third, the automatic segmentation performed by the GCA software may be altered in some patients where the macular morphology is distorted due to the IMH (Figure
Examples of scans with incorrect ganglion cell-inner plexiform layer segmentation due to macular morphology distortion. (a) OCT ganglion cell analysis in a patient after idiopathic macular hole surgery. (b) OCT ganglion cell analysis in a patient with an idiopathic macular hole. The arrows show an area where the automated segmentation was incorrectly performed.
In conclusion, only focal temporal changes in the GCIPL thickness may be appreciated after vitrectomy with BBG-assisted ILM peeling for IMH. Furthermore, the results provided by Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, CA) ganglion analysis software should be carefully evaluated in patients with maculopathies where there is a macular morphology distortion. However, semimanual segmentation may slightly help to improve the quality of the GCIPL analysis in these patients.
A significant reduction of GCIPL thickness at the temporal macular quadrants was observed by SD-OCT analysis with the new ganglion cell analysis (GCA) software of the Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, CA) at 6 months after vitrectomy with BBG-assisted internal limiting membrane removal for idiopathic macular hole surgery. This reduction detection was higher when the analysis was performed by semimanual segmentation. On the other hand, an increase in the postoperative GCIPL thickness was observed in a significant number of patients, although this increase was less pronounced when the analysis was performed by semimanual segmentation. As we assume that an increase in GCIPL thickness is not expected after surgery, we hypothesize that these values may be artefacts or even a subtle inner retinal edema after ILM peeling, which could be caused by BBG dye. In this study we have evaluated the capacity of the new software of the Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, CA) to analyze changes in the retinal GCIPL after internal limiting membrane peeling for idiopathic macular hole surgery, but further studies should be performed in patients with other frequent macular pathologies. To our knowledge, no study has previously validated this software in macular pathology. Glaucoma specialists must take into consideration this possible bias in GCIPL thickness analysis in patients with vitreomacular surface changes.
The authors report that they have no proprietary or conflict of interests.
The authors thank Aurora Alvarez, Miriam Camiña, Diego Ruiz-Casas (M.D.), and Ismael Samhan-Arias for their contribution to the SD-OCT image acquisition and data recollection.