Early Perfusion of a Free RPE–Choroid Graft in Patients with Exudative Macular Degeneration Can Be Imaged with Spectral Domain–OCT

Abstract

Purpose.: To study early flow and revascularization in a free, autologous, retinal pigment epithelium (RPE)–choroid graft.

Methods.: This prospective cohort study used spectral domain–optical coherence tomography (SD-OCT) after RPE–choroid graft surgery in 12 patients. This SD-OCT was combined with fluorescein angiography (FA) and indocyanine green angiography (ICGA) in 5 patients.

Results.: SD-OCT revealed that vessel diameter, number of vessels, and graft thickness increased in 10 of 12 patients, starting between 3 and 10 days after surgery. A subsequent decrease in thickness was found in all 10 patients, beginning as early as 8 days after surgery. Initially, the graft vessels were optically clearer than the underlying choroidal recipient vessels. Between 8 and 30 days after surgery, the optically clear vessels became gray, similar to the recipient choroid. FA and ICGA revealed perfusion in 4 of 5 patients between postoperative days 6 and 15. Between postoperative days 12 and 60, the entire choroidal structure of the graft was visible on ICGA.

Conclusions.: These data suggest that enlargement of vessel diameter, increase in the number of choroidal vessels, and graft thickening visualized by SD-OCT correspond with the ingrowth of afferent vessels, as demonstrated by ICGA. The subsequent establishment of efferent vessels results in flow, imaged as a change in color of the graft's vessels from optically clear to gray, graft thinning on SD-OCT, and complete revascularization on ICGA. SD-OCT, a noninvasive examination, can be used to demonstrate early graft perfusion in patients (trialregister.nl/trialreg/admin/rctview.asp number, NTR1768).

The leading cause of irreversible legal blindness among senior citizens in the industrialized world is age-related macular degeneration (AMD). AMD is also the third most common cause of blindness worldwide. ^1,2 Exudative AMD involves choroidal neovascularization, whereby new choroidal blood vessels cross Bruch's membrane and grow into the space underneath the retinal pigment epithelium (RPE) and/or retina. ^3,4 These newly formed vessels are prone to leakage and bleeding into the submacular area; if left untreated, the condition can ultimately lead to irreversible damage to RPE cells and the retina. ^5,6  

In randomized controlled trials, intravitreal injections of anti–vascular endothelial growth factor (anti-VEGF) have been shown to be effective in patients with exudative AMD with classic, occult, and mixed choroidal neovascular membranes. ^7,8 However, patients with large hemorrhages and RPE tears were excluded from these studies. Additionally, these studies identified a small percentage of nonresponders: patients who had lost ≥15 letters after 24 months of ranibizumab treatment ⁹ and patients with a 3-line drop after three injections, with no later improvement (unpublished data; Morse L, et al. 2-Year ANCHOR Study, World Ophthalmology Congress 2008, Hong Kong). Other methods of treatment may be indicated for these three patient groups. ^{10 –13}  

In patients with a large hemorrhage, the submacular hemorrhage alone can be removed. Although results from the Submacular Surgery Trial Hemorrhage group showed no overall better outcome of blood removal in AMD patients compared with untreated patients, the percentage of patients with >6 lines of visual acuity (VA) loss was statistically significantly smaller in the surgery group. ¹⁴ Another method is rotation of the macula, after a 360° retinectomy, to an area with less damaged RPE. ¹⁵ A third method is the transplantation of an autologous free graft of RPE, Bruch's membrane, choroicapillaris, and choroid. ¹⁶  

In free grafts, the best functional outcome is likely to relate to preserved neuroretina in the macula combined with a functioning RPE–choroid graft. For neuroretina and graft alike, surgery with minimal trauma would limit damage to these tissues. Furthermore, patient selection is important for a preserved neuroretina, and early revascularization is likely to help minimize ischemic damage to the graft. 

Most studies regarding the revascularization of a free graft show results no earlier than 1 week after surgery. Due to the invasive nature of such studies (histology in animals ¹⁷ and angiography in human patients ^{18 –20} ), such data were neither early nor consecutive. 

In this study, we further analyzed early revascularization by a noninvasive method: spectral domain–optical coherence tomography (SD-OCT) (Spectralis HRA; Heidelberg Engineering, Heidelberg, Germany). SD-OCT was chosen because it allows visualization of the choroid in a repeatable manner thanks to its eye-tracking facility. We combined SD-OCT images with indocyanine green angiography (ICGA) and fluorescein angiography (FA) images obtained at the same time points as those in the SD-OCT scans. We studied revascularization in patients after RPE–choroid graft translocation with invasive and noninvasive techniques, and tried to correlate data between these different imaging tools. 

Materials and Methods

Patients

Twelve consecutive patients with exudative AMD who were ineligible for (further) anti-VEGF or any other treatment were included in this study. Patients could be included if they were nonresponders to anti-VEGF treatment (i.e., if they had a visual loss of ≥15 letters on the Early Treatment Diabetic Retinopathy Study [ETDRS] chart after at least three anti-VEGF injections). Other indications were a massive submacular hemorrhage (>1.5-mm thickness on ultrasound) that was no longer eligible for recombinant tissue plasminogen activator (rTPA) injection (i.e., existing for >2 weeks ^21,22 ), or an RPE tear. Seven of the included patients were at the Rotterdam Eye Hospital, The Netherlands (REH) and five were at the Ospedale Sacro Cuore, Negrar, Verona, Italy (OSC). 

All patients provided informed consent for the surgical procedure and for preoperative and postoperative examinations, in accordance with the tenets of the Declaration of Helsinki. The study was also approved by the Medical Ethical Committee of the Erasmus University, The Netherlands. In accordance with institutional guidelines, the surgical procedure and examinations are considered part of standard care in the OSC. 

Surgery

At two centers, the REH and the OSC, we examined 12 patients who underwent a full-thickness translocation of autologous midperipheral RPE, Bruch's membrane, choriocapillaris, and choroid (RPE–choroid graft), as described previously. ^19,23,24 At the REH for patients 1 to 6, a retinotomy was made in the raphe, and the midperipheral graft was placed under the retina using a bent forceps. At the OSC and for patient 7 at the REH, a 180° temporal retinotomy was made to permit folding over the temporal retina (and therefore exposure of the macula). The graft could then be dragged over the macular area from its donor site. Before harvesting the graft, we applied linear diathermia or laser to the RPE and choroid in a rectangular or oval shape. The graft was then cut out within the diathermia or laser borders, avoiding diathermized or lasered tissue in the graft itself. At the end of surgery, silicone oil was used as a tamponade. All surgical procedures were performed by one surgeon at each site (JCvM at REH, GP at OSC). The silicone oil was removed in a second procedure approximately 3 months later. Lensectomy or phacoemulsification and insertion of an intraocular lens were performed during the first or second surgery in the 8 phakic patients, as shown in Table 1. We examined the status of the graft shortly after the RPE–graft translocation operation in the patients with Heidelberg SD-OCT at the REH and with Heidelberg SD-OCT, FA, and ICGA at the OSC. 

Table 1.

 

Patient Characteristics

Patient	Indication Surgery	Age (y)	Eye	M/F	Baseline VA (logMAR)	SD-OCT Timing	Silicone Oil Removal	IOL Insertion	FA/ICGA	VA 3 mo (logMAR)
1	RPE tear	85	OD	F	0.46	Preoperatively. Days 1, 2, and 8, and 2 mo PO.	6 wk PO	Pseudophakia	N/A	0.7
2	Submacular hemorrhage	79	OD	F	1.2	Preoperatively. Days 1–4 and 60 PO.	15 wk PO	15 wk PO, lensectomy at graft surgery	N/A	1.5
3	Submacular hemorrhage	82	OD	F	0.8	Preoperatively. Days 1–4 and wk3 and 6 PO.	9 wk PO	Pseudophakia	N/A	1.5
4	Nonresponsive to anti-VEGF	52	OD	M	0.66	Preoperatively. Days 1–4, 6, and 8, and wk 2 and 5 PO.	20 wk PO	20 wk PO, lensectomy at graft surgery	N/A	0.32
5	Submacular hemorrhage	79	OD	F	1.4	Preoperatively. Days 1, 3, 4, 8, 9 PO. (New hemorrhage developed at day 9 PO.)	10 days PO, new submacular hemorrhage	10 days PO, lensectomy at graft surgery	N/A	2.1
6	RPE tear	86	OS	F	0.58	Preoperatively. Days 1, 4, 6, 8, 11, 14, 20, and 60 PO.	12 wk PO	Pseudophakia	N/A	1.5
7	Fibrotic scar after anti-VEGF	68	OD	M	1.58	Preoperatively. Days 1, 4–8, 11, 13, 15, 19, 30, and 60 PO.	11 wk PO	Pseudophakia	N/A	1.4
8	Nonresponsive to anti-VEGF	68	OS	M	0.9	Preoperatively. Days 1, 3, 5, 7, 10, 15, and 21 PO. 1, 2, 3, and 5 mo PO.	3 mo PO	Lensectomy and IOL insertion at graft surgery.	Preoperatively. Days 7, 10, 15, and 21 PO. 1 and 3 mo PO	1.3
9	Nonresponsive to anti-VEGF	68	OS	F	0.8	Preoperatively. Days 1, 3, 7, 10, 14, and 21 PO. 1, 2, 3, and 4 mo PO.	3 mo PO	Lensectomy and IOL insertion at graft surgery.	Preoperatively. Days 3, 7, 10, 14, and 21 PO. 1 and 2 mo PO.	0.5
10	Submacular hemorrhage and RPE tear	75	OS	F	1.5	Preoperatively. Days 1, 2, 6, 9, 13, 18, and 25 PO. 1, 2, 3, and 4 mo PO.	3 mo PO	Lensectomy and IOL insertion at graft surgery.	Preoperatively. Days 2, 6, 9, 13, 18, and 25 PO. 1, 2, and 4 mo PO.	1.5
11	Submacular hemorrhage	57	OD	M	1.0	Preoperatively. Days 1–3, 6, 13, and 20 PO. 1, 2, and 3 mo PO.	3 mo PO	Lensectomy and IOL insertion at graft surgery.	Preoperatively. Days 6, 13, and 20 PO. 1 and 2 mo PO.	0.4
12	Nonresponsive to anti-VEGF	64	OS	M	0.5	Preoperatively. Days 1, 2, 5, 7, 12, and 20 PO. 1, 2, and 3 mo PO.	2 mo PO	Lensectomy and IOL insertion at graft surgery.	Preoperatively. Days 5, 7, 12, and 20 PO. 1, 2, and 3 mo PO.	0.6

 IOL, intraocular lens; OD, oculus dexter (right eye); OS, oculus sinister (left eye); PO, posteroperatively; N/A, not applicable.

Spectral Domain–Optical Coherence Tomography

Twelve patients with exudative AMD treated with an RPE–choroid graft were scanned by SD-OCT pre- and postoperatively (timing of examination for each patient summarized in Table 1). For SD-OCT scanning, the software provides an automatic real time (ART) function to increase image quality and reduce noise. Multiple frames (B-scans) of the same scanning location are performed during the scanning process with ART activated, and images are averaged for noise reduction. In this study, 51 frames were acquired for each single B-scan. Preoperative single B-scans were made through the macula, and follow-up scans were performed postoperatively. Four types of single B-scans were performed on each patient: horizontal, vertical, 45°, and 135°. 

SD-OCT Analysis

For quantitative analysis the number and diameter of the vessel lumina, as well as thickness of the graft and fluid under the graft, were measured in the horizontal and vertical OCT images of every patient at each visit. The thickness of the graft and subgraft fluid were both measured precisely under the fovea and at 500 μm on either side of the fovea. These six measurements were averaged. The intragraft vessels visualized on the vertical and horizontal OCT scans were counted and averaged. The maximal vessel diameter was measured in each of the three largest vessels of each graft. Six such diameter measurements per graft were averaged. ImageJ software (developed by Wayne Rasband, National Institutes of Health, Bethesda, MD; available at http://rsb.info.nih.gov/ij/index.html.) was used for the measurements. 

Two masked observers (EJTvZ and MCG) independently graded changes in gray shading of the vessel lumina in the graft compared with the gray shading of the vessels in the underlying choroid. 

Fluorescein Angiography and Indocyanine Green Angiography

Both FA and ICGA (HRA Spectralis; Heidelberg Engineering) were performed pre- and postoperatively on the patients at the OSC. In FA, capillary flush (diffuse filling of the choriocapillaris at the early arterial phase) was regarded as a sign of revascularization. In ICGA, video-angiography in the first 40 seconds was used to identify parallel-oriented, ladder-like choroidal vessels to recognize graft perfusion. ²⁰ Stereo images were made to confirm that the parallel-oriented, ladder-like vessels were located between the graft and the recipient choroid. The perfused area of the graft on FA and ICGA was measured with ImageJ software. 

Results

Patients

The mean age of the patients was 71.9 ± 3.1 years (mean ± SD). Indication for surgery in five patients was a submacular hemorrhage; one of these five patients with a submacular hemorrhage also had an RPE tear. Two other patients had an RPE tear only. Five patients were nonresponders to anti-VEGF treatment. These patients had a visual loss of ≥15 letters on the ETDRS chart after at least three anti-VEGF injections. One of these patients had developed a fibrotic scar. 

VA at baseline ranged from 20/58 to 20/760 (0.46–1.58 logarithm of minimal angle of resolution [logMAR]), with a median of 0.85 logMAR (20/142) and mean 0.94 logMAR (SD ± 0.11). VA after 3 months ranged from 20/42 to 20/2518 (0.32–2.1 logMAR), with a median of 1.35 logMAR (20/448) and mean 1.11 logMAR (SD ± 0.16). The characteristics of each patient are summarized in Table 1. SD-OCT, FA, and ICGA images from patient 9 are shown in Figure 1. SD-OCT images from patient 7 are shown in Figure 2. 

Figure 1.

 

Patient 9: SD-OCT, ICGA, and FA images at postoperative (Postop) days 7, 10, 21, and 60. The SD-OCT images are vertical B-scans. The vessel lumina are marked by an asterisk (*) and the numbers of these vessel lumina are shown by small arrows (↑). The graft thickness is depicted by a double-pointed arrow (↕) and the fluid under the graft is shown by a delta (Δ). Large arrows () point at the vessel lumina of the graft and the vessel lumina of the underlying choroid to show the difference or likeness in gray shading between these lumina. White arrows (←) point at the capillary flush in FA and the ladder-like choroidal vessels of the graft in the ICGA images. At Postop day 10, there was a significant increase in vessel diameter, number of vessels, and graft thickness, combined with first ICGA lines visible at the inferior edge of the graft. At Postop day 14, there was a further increase in vessel diameter and graft thickness and more ICGA lines were visible. At Postop day 21, there was a significant increase in fluid under the graft. Optically clearer vessels in the graft compared with vessels of the choroid could be seen until Postop day 21. The SD-OCT at Postop day 21 shows the same gray shading in the vessel lumina as that of the underlying choroid vessels in the left half of the graft, whereas the right half of the graft has optically clearer vessels compared with the underlying choroid vessels. This gray shading on the left half of the graft coincides with the inferior part of the graft, in which FA flush and ICGA lines are visible. At Postop day 60, the vessel lumina had gray shading comparable to that of the choroid, the diameter of the vessels had decreased, thickness and fluid under the graft had decreased, all the vessels of the graft were visible on ICGA, and FA showed fluorescence comparable to that of the surrounding choroid.

Figure 2.

 

Patient 7: SD-OCT images at postoperative (Postop) days 1, 6, 13, and 15. The SD-OCT images are horizontal B-scans. The vessel lumina are marked by an asterisk (*) and the numbers of these vessel lumina are shown by small arrows (↑). The graft thickness is depicted by a double-pointed arrow (↕), and the fluid under the graft is shown by a delta (Δ). Large arrows () point at the vessel lumina of the graft and the vessel lumina of the underlying choroid to show the difference or likeness in gray shading between these lumina. At Postop day 6, there was a significant increase in vessel diameter and number, the graft thickness increased, and fluid accumulated under the graft. The vessels of the graft are optically clearer than the vessel lumina of the underlying choroid. At Postop day 13, the diameter of the vessels decreased in size, together with thinning of the graft and decrease of fluid under the graft. The lumina of the vessels in the middle of the graft are optically clearer than the lumina of the choroidal vessels; only some vessels on the very edges of the graft show the same gray shading as that of the vessels of the underlying choroid. At Postop day 15, only a few lumina of the vessels in the middle of the graft are still optically clearer than the lumina of the vessels of the underlying choroid. The graft vessel lumina at both sides of the graft show gray shading more similar to the underlying choroid vessel lumina.

Changes on SD-OCT, FA, and ICGA: Days after Surgery and Measurements

Qualitative and quantitative measurements could be performed for all 12 patients (Table 2). However, the data from only 9 patients were used for the qualitative description of the process of revascularization because 2 patients did not show revascularization and 1 patient developed a hemorrhage 9 days after surgery. When the mean number of days or the mean change of a feature is given in the text and Table 2, it is the mean of these 9 patients (unless the mean is otherwise specified with an n). 

Table 2.

 

Quantitative Measurements of OCT Images

Variable	Days (mean)	Value (mean)	Range (min–max Values)
Diameter of vessels in the graft, μm
Start	1.0	94.6	77–113
Increase	5.6	178.3	105–352
Decrease	25.6	146.3	115–153
Last data	60.0	136.3	107–153
Number of vessels in the graft, n
Start	1.0	19.8	15–27.5
Increase	5.8	26.0	19–36.5
Last data	60.0	35.8	25–54.5
Thickness of the graft, μm
Start	1.0	156.2	79–233
Increase	5.4	243.2	155–463
Increase 2	11.6	316.2	242–511
Decrease	24.6	244.9	167–394
Last data	60.0	231.0	167–356
Fluid under the graft, μm
Start	7.0	182.0	35–304
Maximum	14.0	228.2	138–314
Decrease	25.0	137.2	16–218
Last data	60.0	58.6	0–200
Perfused area of the graft on FA, mm2
Start	10.5	4.8	0.6–11.8
Last data	60.0	18.2	14.2–25
Perfused area of the graft on ICGA, mm2
Start	9.0	2.9	0.2–5
Last data	60.0	17.9	14.1–24

 Values are mean and range of 9 patients.

Immediately after surgery, small vessels are seen in the graft (n = 12) (Figs. 1, 2). The diameter of these small vessels in the graft increased (mean, 94.5–178.3 μm) between postoperative days 3 and 10 (mean, 5.6 days). During the same time period (mean, 5.8 days, n = 8), the number of vessels increased (mean, 19.8–26.0) and graft thickness (mean, 5.4 days) also increased (mean, 156.2–243.2 μm) in most patients (n = 10). After the rapid initial increase, the number of vessels increased progressively over time (mean, from 26.0 to 35.8 at mean 60 days). Among the 10 patients who experienced graft thickening, fluid, and/or material under the graft accumulated in 6 patients. Measurements could be performed for 5 of these patients. The first fluid appeared in each patient between postoperative days 1 and 21 (mean, 7 days), and had a mean thickness of 182 μm. The maximum fluid thickness under the graft (mean, 228.2 μm) of these patients was measured between days 1 and 30 (mean, 14 days). The excluded patient was the one with hemorrhage described previously. 

Between postoperative days 8 and 30 (mean, 16 days, n = 8), the gray shading of the vessel lumina changed from clearly more white, compared with the underlying choroid, to a gray shading similar to that of the underlying choroidal vessels (n = 10; Figs. 1, 2). This change in gray shading of the graft coincided with the early choriocapillary flush ²⁰ and ICGA lines of the graft. 

Early choriocapillary flush (mean, 4.8 mm²) could be seen on FA in four of five patients between postoperative days 6 and 15 (mean, 10.5 days). The first vessels of the graft also were visible on ICGA (mean, 2.9 mm²) between postoperative days 5 and 15 (mean, 9 days) in the same four patients. 

A second increase in graft thickness (mean, 243.2–316.2 μm) was observed in six patients, between postoperative days 9 and 15 (mean, 11.6 days). A subsequent significant decrease (mean, 178.3–146.3 μm and later to mean of 136 μm at mean 60 days) in vessel diameter began as early as 8 days, ranging to 40 days (mean, 25.6 days) after surgery (n = 7 of 12). This was combined in these seven patients with an initial decrease (mean, 316.2–244.8 μm) in thickness (mean, 24.6 days) of the graft, in the same time frame. A decrease of fluid/material under the graft (mean, 228.2–137.2 μm) started between days 2 and 60 (mean, 25 days). This decrease in thickness was progressive and lasted through the first weeks to months after surgery (mean, 58.6 μm at mean 60 days). The decrease in graft thickness was also initially progressive, but later stabilized at mean 60 days (mean thickness, 231 μm). 

After the initial choriocapillary flush visible on FA (mean, 10.5 days) and the first vessels visible on the graft with ICGA (mean, 9 days), a progressively increasing area of choriocapillary flush was measured (mean, 4.8–18.2 mm²) on FA (n = 4 of 5, between mean days 10.5 and 28.7) and a progressive area of vessels in the graft (mean, 2.9–17.9 mm²) was measured on ICGA (between mean days 9 and 28.7). The whole choroidal structure of the graft was evident on FA, and ICGA revealed that the entire graft was perfused in four of five patients, both between postoperative days 12 and 60 (mean, 28.7 days). 

In two patients, patients 4 and 10, no significant changes in vessel number and/or diameter or graft thickness were found throughout the follow-up period. Their SD-OCT showed a thin layer of few small-diameter vessels with very small lumina lying under the RPE. In patient 10, the graft also did not become visible on ICGA; however, the original choroidal vessels underlying the graft became weakly visible starting at postoperative day 9 and were clearly visible 4 months after surgery. 

Stereo images confirmed that the choroidal vessels were part of the graft in all patients in which a ladder-like vasculature could be seen (n = 4 of 5). 

Discussion

After surgery, a free RPE and choroid graft is very likely to need revascularization for survival. In this study, we found that the revascularization of an RPE graft can be observed by SD-OCT, given that the SD-OCT findings corresponded with those of FA and ICGA. 

Studies in free skin transplants (in animals) reveal that, in an early phase, plasma exudes from the recipient site's damaged arteries and veins, fills the lumina of the graft vessels, and supposedly supports graft-tissue metabolism (imbibition phase). ²⁵ Later processes involved in graft survival are neovascularization (vascular ingrowth from the recipient bed) and replacement and/or reconnection of the graft vasculature by endothelial and endothelial-progenitor cells from the recipient bed. ²⁰ Young ²⁶ demonstrated this neovascular ingrowth at the third or fourth day after surgery in the pedicle skin flaps of pigs. This time sequence was also found for wound healing in skin flaps, free skin grafts, and linear incisions. ²⁶  

Revascularization of RPE–choroid grafts has been confirmed in pigs, with connecting vessels between recipient and graft present at 1 week and 3 months after surgery. ¹⁷ To study revascularization of a free RPE–choroid graft in patients, the early phase of FA and or ICGA is recommended. ^19,20 However, angiography remains an invasive technique that is not readily suitable for the repeated study of revascularization in patients. The advent of an improved, more advanced imaging device (Spectralis HRA SD-OCT; Heidelberg Engineering), with better definition of the choroid and an eye-tracking device, allowed repeated, serial, noninvasive studies. It would therefore be desirable to be able to correlate angiographic data with noninvasive imaging, with the subsequent ability to rely on the noninvasive modality alone. 

In our series of patients, we found a consistent pattern of changes at the level of the graft on SD-OCT that coincided with specific changes observed on ICGA and FA. The first days after surgery, choroidal vessel lumina of the graft were relatively small and optically clear on SD-OCT. No filling of the graft vessels was visible on angiography. We suggest that this phase corresponds with serum imbibition, before any revascularization. Several days after surgery, observed on SD-OCT, the thickness of the graft increased and the vessels of the graft started to enlarge in diameter and to increase in number. The vessel lumina in the graft remained optically clear. These events correlated with the appearance on ICGA of a choroidal vessel connecting to the graft. We suggest that the graft was pumped up by this connection of an afferent vessel without the presence of an efferent vessel. Several days later, on SD-OCT, the graft became thinner, the diameter of the vessels decreased slightly, and a gray shading appeared inside the vessels, comparable to that of the recipient choroid. These events on SD-OCT coincided with the perfusion of the entire graft on FA and ICGA. We suggest that these observations were caused by the establishment of an efferent vessel connection, facilitating flow through the graft and a subsiding of the graft-tissue swelling. 

Fluid under the graft does not follow the pattern of changes of the overlying graft. Early fluid may be related to trapped fluid after placement of the graft or an inflammatory exudate after surgical trauma. Later after surgery, subgraft fluid may be secondary to the connection of the afferent vessel before the ingrowth of efferent vessels. 

Fortunately, most patients with exudative AMD respond well to anti-VEGF treatment. However, an RPE–choroid graft may be considered in three patient categories: nonresponders to anti-VEGF treatment; massive submacular hemorrhage, no longer eligible for rTPA injection ^21,22 ; or an RPE tear. 

A successful functional outcome of RPE–choroid graft surgery is dependent not only on graft perfusion, but also on the amount of damage to the RPE during insertion and to the condition of the overlying retina, with patient selection and the preoperative course as major factors. We found that VA increased the first year after surgery, mainly after 3 to 6 months, and stabilized thereafter (van Zeeburg EJT, et al. IOVS 2010;51:ARVO E-Abstract 2058). ²⁷ Therefore, VA at 3 months after surgery in the present study is probably not the best VA that the patient may eventually achieve. At this moment, the small number of patients and short follow-up preclude analysis of a correlation between perfusion and functional outcome. 

We demonstrate in this study that the revascularization of a free RPE and choroid graft follows well-established steps of revascularization, as reported in other free grafts. At the same time, this study demonstrates that noninvasive SD-OCT findings correlate with FA and ICGA findings. Therefore, SD-OCT can be used to monitor the postoperative process of revascularization in a free RPE–choroid graft without the need for invasive imaging techniques. 

Footnotes

 Supported by Royal Visio, The Netherlands and Flieringa Research Foundation, The Rotterdam Eye Hospital, Rotterdam, The Netherlands.

Footnotes

 Disclosure: E.J.T. van Zeeburg, None; M.G. Cereda, None; J. van der Schoot, None; G. Pertile, None; J.C. van Meurs, None