Reconstruction of the Limbal Vasculature After Limbal-Conjunctival Autograft Transplantation in Pterygium Surgery: An Angiography Study

Abstract

Purpose.: We evaluated the angiographic features of the affected limbus in patients with pterygia and assessed limbal reconstruction outcomes after limbal-conjunctival autograft (LCA) transplantation in terms of vascular remodeling.

Methods.: We studied prospectively 31 eyes of 31 patients who underwent pterygium excision and LCA transplantation; 28 eyes of 28 normal participants served as controls. Anterior segment indocyanine green angiography (ICGA) was performed for each participant preoperatively and at 1 week, and 1 and 3 months postoperatively. The perioperative angiographic features of the pterygium were compared to those in normal eyes. The structural changes of the marginal corneal vascular arcades (MCAs) and LCA were quantitatively assessed postoperatively in terms of vascular density and lacunarity.

Results.: Deteriorated MCAs that extended beyond the pterygium head were observed in the pterygium group. The pterygium had a dual blood supply from the conjunctival and episcleral circulations. In terms of limbal reconstruction, the engorged reperfusion vessels arose from the adjacent episcleral vessels along the limbus at 1 week postoperatively. The reconstructed MCAs had begun to appear at 1 month postoperatively and became apparent 3 months postoperatively in 26 (83.9%) of 31 eyes of the pterygium group, resulting in a successful clinical outcome. Higher vascular density and lower lacunarity were measured in the limbus and the graft at 3 months than at 1 month (P < 0.001 for all), which indicated fine reorganization of the reconstructed vessels.

Conclusions.: The pterygium had a dual blood supply, and the remodeling of the affected limbus and LCA continued up to 3 months postoperatively.

Introduction

Pterygium formation is an active, inflammatory ocular surface disease. Pterygia begin growing from the limbal epithelium and invade the cornea, and subsequently the conjunctival migration.¹ With respect to epidemiological features, many investigators have suggested a strong correlation between pterygia and ultraviolet light exposure, dryness, chronic inflammation, and exposure to wind.^2,3 Many fibroangiogenic growth factors have been implicated in the pathogenesis of pterygium, such as TNF-α, platelet-derived growth factor, TGF-β, and VEGF.^4–6 

Pterygia recently have been thought to originate from limbal stem cells (LSCs) that have been altered by chronic ultraviolet light exposure.^7,8 The healthy limbal epithelium serves as a junctional barrier to conjunctival migration.⁹ The establishment of these concepts has resulted in efforts aimed at limbal reconstruction.^10,11 The superiority of limbal-conjunctival autograft (LCA) transplantation has been confirmed in several prospective studies and randomized controlled trials.^10,12 Inclusion of healthy limbal epithelium in the conjunctival graft restores the barrier function of the limbus, thereby leading to a decrease in the incidence of recurrence. 

Although many studies have been conducted to analyze the molecular aspects of pterygia and the clinical outcomes of LCA transplantation, few have evaluated the angiographic features of the affected limbus before and after surgery. The palisades of Vogt develop adjacent to the limbus, and the region of the limbus internal to the palisades and extending into corneal margin contains complex vascular loops and projections called the marginal corneal vascular arcades (MCAs).^13–15 The MCAs lie outside the peripheral edge of Bowman's layer and are associated with the location of the LSC niche. The MCAs also are known to be involved in disease processes that affect the peripheral cornea and may account for the characteristic features of these conditions.¹⁶ Given this background, in vivo angiographic findings would provide a better understanding of the disease process and reconstruction process of the affected area after pterygium surgery. The primary aim of the present study was to evaluate the angiographic features of the affected limbus in pterygia and assess the limbal reconstruction in terms of vascular remodeling. 

Materials and Methods

Consecutive patients from the Ophthalmology Department in our institution were prospectively recruited for pterygium surgery from March 1, 2012 to April 31, 2013. The healthy participants with sound eyes also were recruited for this study as control. The research protocol adhered to the tenets of the Declaration of Helsinki. Institutional Review Board and Ethics Committee approval was obtained for this prospective study, and informed consent was obtained from all participants before pterygium surgery. The inclusion criteria were confirmation of pterygium on slit-lamp biomicroscopy, and absence of serious ocular surface disease (e.g., severe dry eye, Sjögren's syndrome, abnormal lid function or position, significant conjunctival scarring), glaucoma, retinal vascular disease, and severe systemic vascular complications. We took a detailed medical history from each participant at the first visit. 

Ophthalmic Examinations

In the preoperative assessment, all participants underwent an ophthalmic examination, including visual acuity, tonometry, keratometry, slit-lamp biomicroscopy, and dilated fundus examination. Routine postoperative examinations were scheduled for 1 day, 1 week, and 1, 3, 6, and 12 months after surgery. All ophthalmic examinations were performed by one experienced corneal specialist (JKC). Pterygia were classified into three stages according to the following method suggested by Yang et al.¹⁷: stage I, the head of the pterygium does not reach the midline between the limbus and pupillary margin; stage II, the head of the pterygium passes the midline but does not reach the pupil; and stage III, the head of the pterygium passes the pupillary margin. Stages I and II are considered to be the early stages of the disease, and stage III is considered to be the advanced stage. Recurrence of the pterygium after surgery was graded as follows: G0, normal appearance; G1, the presence of some fine episcleral vessels, but without any fibrous tissue in the excised area; G2, the presence of additional fibrous tissues without invasion of the cornea; and G3, true recurrence with invasion of fibrovascular tissue into the cornea.¹⁸ Grades G2 and G3 were considered to be clinical recurrences. 

Pterygium Surgery and Postoperative Management

All participants underwent the same pterygium surgery under local anesthesia by a single experienced corneal specialist (JKC). After excision of the head and body of the pterygium with a No. 15 blade, subconjunctival fibrovascular tissues, including Tenon's capsule, were widely removed up to the rectus muscle. Next, 0.02% mitomycin C was placed into the subconjunctival space for 2 to 3 minutes. A free LCA harvested from the superotemporal bulbar area was placed over the sclera and secured with a tissue adhesive (Greenplast-Q; Green Cross Corp., Seoul, Korea) and 10-0 nylon sutures. All patients were treated with a topical 0.5% levofloxacin (Cravit; Santen Pharmaceutical Co., Ltd., Osaka, Japan) and 1% prednisolone (Pred Forte; Allergan, Inc., Irvine, CA, USA) eye drops 4 times daily for 4 weeks postoperatively, respectively. All sutures were removed 1 week postoperatively. 

Angiographic Imaging Acquisition

Indocyanine green angiography (ICGA, Spectralis HRA+OCT; Heidelberg Engineering, Dossenheim, Germany) was conducted immediately preoperatively, and at 1 week, and 1 and 3 months postoperatively on the study eye of each patient. After rapid intravenous injection of 1 mL of 25 mg/mL indocyanine green (ICG) dye (Dianogreen; Daiichi Sankyo Co., Ltd., Tokyo, Japan) through a butterfly cannula on the dorsum of the hand, angiographic images were taken at 2-second intervals for the first minute, at 10-second intervals for the next 4 minutes, and at 1-minute intervals up to the 15 minutes. We assessed the vasculature of the affected limbus before and after surgery. The image exposure quality was continuously assessed on a high-resolution display monitor. 

Study Groups

The eyes that had pterygium surgery with LCA transplantation comprised the pterygium group, and the sound eyes of the normal participants served as the control group. Only one eye of each participant was included in the analysis. 

Outcome Measures

The main outcome measures were the angiographic features of the affected limbus before and after surgery. The MCAs in the limbus were analyzed using ICGA. Before surgery, we drew an imaginary ellipse that exhibited the best fit to the MCAs. We assessed the vascular loop and stromal projections of each area of the MCAs, and the focal area in which the innermost loops and projections markedly deviated toward cornea from the imaginary ellipse was defined as a deteriorated MCA. We also assessed the vasculature and fluorescence pattern of the pterygium. We serially analyzed the changes in the angiographic features of the MCAs as well as the perfusion states and fluorescence pattern of the graft after surgery. The episcleral vessels were identified by their tortuous appearance and correlation with the clinical picture. We also used the device module that allows the user to focus on different layers of the ocular surface to differentiate the episcleral vessels from the conjunctival vessels. Evaluation of the ICGA results of all participants was performed by 2 of us (YJK and SHY) using the built-in analysis program of the device, regarding the presence or absence of deteriorated MCAs, limbal vasculatures, graft reperfusion vessels, and leaking spots. In cases of disagreement, a third reviewer (JKC) made the final decision. 

We also assessed the details of the vascular network of the pterygium and grafts preoperatively and postoperatively. The MCAs in the limbal region (0.85 × 0.85 mm, from the innermost row of the MCAs) and the vasculature of the central square region of the pterygium body and graft, and corresponding square region of the normal conjunctiva (2.0 × 2.0 mm, measured using the built-in analysis program) were converted to high resolution images and used in the analysis. The quantitative measure of vascular network, including the vascular area, vascular density, and lacunarity, were performed using a semiautomated computational tool (AngioTool; available in the public domain at http://angiotool.nci.nih.gov).¹⁹ In morphological analysis, the lacunarity has been defined as gappiness, inhomogeneity, and translational and rotational invariance.²⁰ It has been known to be useful to characterize and quantitatively analyze vascular networks.¹⁹ For example, if a vascular network has large gaps or holes, it has high lacunarity. Measuring lacunarity enabled us to perform the numerical analysis of the complexity of vascular network. 

We optimized the parameters of the AngioTool for producing skeletons that best represented the vascular networks. The mean values of the two measurement were used in the statistical analysis. The validity of the AngioTool in the analysis of vascular network has been reported previously.¹⁹ 

Statistical Analysis

All statistical analyses were performed with SPSS ver. 19.0 (SPSS, Inc., an IBM company, Chicago, IL, USA). The Mann–Whitney U test or Fisher's exact test was used for comparisons between the two groups. Changes in vascular network after surgery were compared with the Wilcoxon's signed-rank test. All tests were 2-tailed, and P < 0.05 was considered to indicate statistical significance. 

Results

Study Population

In total, 38 consecutive patients underwent pterygium surgery in our institute beginning on March 1, 2012. Of these patients, 31 met the inclusion criteria and completed the study protocol, and their 31 eyes were included in the analysis. A total of 28 sound eyes of 28 normal participants served as control. The details of the clinical characteristics of each group are presented in Table 1. 

Table 1

 

Clinical Characteristics of the Study Groups

Variables	Assessment of Pterygium Before Surgery
	Pterygium Group	Control Group	P Values*
Eyes, n (%)	31 (100)	28 (100)	–
Age, y†	64.0 (54.0–71.0)	62.0 (56.5–70.5)	0.949
Male, n (%)	17 (54.8)	18 (64.3)	1.000
HTN, n (%)	11 (35.5)	12 (42.9)	0.602
Recurrent, n (%)	3 (9.7)	–	–
Stages, n (%)
Stage I	12 (38.7)	–	–
Stage II	11 (35.5)	–	–
Stage III	8 (25.8)	–	–
F/U, mo†	14.0 (12.0–15.0)	–	–

 HTN, hypertension; LCAT, LCA transplantation; F/U, follow-up duration.

*  Statistics: Mann-Whitney U test or Fisher's exact test.

†  Median (interquartile range).

Angiographic Features of the Pterygium and Control Groups

The episcleral circulation appeared from 15.3 to 24.5 seconds after dye injection; this was followed by the appearance of the conjunctival circulation. All eyes in the control group exhibited a complex vascular plexus upon the conjunctiva and episclera. The median vascular area, vascular density, and lacunarity of the healthy conjunctiva were 1.65 mm² (interquartile range [IQR], 1.50–1.78), 41.34% (IQR, 37.53–44.62), and 0.094 (IQR, 0.083–0.119), respectively. No significant dye leakage was found in these eyes. The palisade of Vogt and the MCA were clearly visible around the corneal limbus in all normal eyes (n = 28, 100%), and not deteriorated (Fig. 1A). The median vascular area, vascular density, and lacunarity of the limbal region in the control group were 0.26 mm² (IQR, 0.23–0.31), 36.16% (IQR, 32.30–42.41), and 0.184 (IQR, 0.166–0.204), respectively. 

Figure 1

 

(A) Angiographic features of the normal eye. The palisades of Vogt and the MCAs are clearly visible in the limbus. No leakage is visible. (B) The vascular plexus of the pterygium seems to originate from the conjunctival circulation adjacent to the boundaries of the pterygium body. Multiple hyperfluorescent spots are visible. (C) Leakages (arrow) are visible >5 minutes after dye injection. (D–F) Findings of the early phase of the anterior segment ICG angiography. Feeding vessels of pterygium head originating from episcleral circulation appear earlier (D) and are followed by the conjunctival circulations (E). Vascular loops and projections of MCAs extend beyond the pterygium head ([C, F], arrowheads). Time after dye injection: (A) 69.8 seconds, (B) 93.1 seconds, (C) 355.7 seconds, (D) 18.9 seconds, (E) 23.0 seconds, (F) 34.0 seconds.

The vascular plexus of the pterygium seemed to originate from the multiple conjunctival vessels adjacent to the boundaries of the pterygium body and multiple hyperfluorescent spots were detected in the head and body of the pterygium (Fig. 1B). There was no identifiable main feeding vessel in each eye in the pterygium group. Leakage from these spots was visible >5 minutes after dye injection (Fig. 1C). Feeding vessels originating from episcleral circulation appeared earlier under the pterygium head and body (Fig. 1D), and conjunctival circulations were visible later (Fig. 1E and Supplemental Video S1). Deteriorated MCAs were observed around the affected limbus in the eyes with pterygium. The innermost vascular loops and projections of the MCAs extended beyond the pterygium head (Fig. 1F). Although variable in terms of intensity and timing of appearance, all eyes (n = 31, 100%) revealed these common angiographic features (Table 2). 

Table 2

 

Angiographic Features of the Study Groups

ICGA Features	Pterygium Group, n = 31		Control Group, n = 28		P Value*
Angiographic characteristics of the pterygium, n (%)
Presence of the LV	31 (100)		28 (100)		–
Deteriorated MCAs	31 (100)		0 (0)		< 0.001
Leaking spots	31 (100)		0 (0)		< 0.001
Reconstruction of the vasculature after LCA transplantation in the pterygium group, n = 31
ICGA features	Preoperative, n (%)	Postoperative, n (%)			P Value*
		1 wk	1 mo	3 mo
Presence of the LV	31 (100)	2 (6.5)	29 (93.5)	29 (93.5)	< 0.001†
Deteriorated MCAs	31 (100)	0 (0)	0 (0)	2 (5.1)	< 0.001‡
Reperfusion of the LCA	–	30 (97.4)	30 (97.4)	30 (97.4)	–

 LV, limbal vasculature.

*  Statistics: Fisher's exact test.

†  Postoperative 3-month value is compared to the postoperative 1-week value.

‡  Postoperative 3-month value is compared to the preoperative value.

Structural Changes of the Limbus and Graft After Pterygium Surgery

Neither the palisades of Vogt nor the MCAs appeared in the affected limbus 1 week postoperatively (Figs. 2A, 2B). The engorged vessels began to arise from the adjacent episcleral vessels along the limbus 1 week after LCA transplantation (Fig. 2B). These vessels became finer and more complex at 1 month postoperatively (Fig. 2C). After 3 months, the vascular loops and stromal projection reappeared, and the vasculature of the affected limbus seemed to be reconstructed (Fig. 2D). 

Figure 2

 

Reconstruction of MCAs after surgery. The deteriorated MCAs are clearly visible beyond the pterygium head ([A], arrowheads). At 1 week after surgery, the MCAs do not appear in the affected limbus, and the engorged vessels arise from the adjacent episcleral vessels along the temporal edge of the graft ([B], arrows). The scanty vascular loops and projections are visible at 1 month postoperatively (C), and become finer and more complex at 3 months after surgery (D). Time after dye injection: (A) 153.5 seconds, (B) 93.1 seconds, (C) 133.8 seconds, (D) 62.2 seconds.

The vascular patterns of the graft exhibited structural changes for 3 months postoperatively. Most areas under each graft, particularly the central area, showed engorged and tortuous vessels, and multiple hyperfluorescent spots on the ICGA images 1 week postoperatively (Figs. 3A, 3B). Although the hyperfluorescent spots were present in the surrounding conjunctival vessels, the edges of the grafts remained hypofluorescent without remarkable anastomosis. This implies the presence of early reperfusion from the episcleral circulation. The vascular plexus of the grafts became finer and more complex 1 and 3 months postoperatively (Fig. 3C). Some hyperfluorescent spots were observed across the grafts on the ICGA images up to 3 months postoperatively in the each eye. Similar postoperative angiographic features were observed throughout the first 3 months postoperatively in 26 (83.9%) of 31 eyes regardless of the stage of the pterygium (Fig. 4). Details of reconstruction and reorganization of the vasculature are presented in Tables 2 and 3. 

Figure 3

 

Reperfusion of the graft after surgery. (A) Preoperative features of the pterygium. (B) Engorged and tortuous reperfusion vessels are visible in the central region of the graft (arrows), while the boundaries of the graft reveal hypofluorescence (arrowheads). (C) The vascular plexus of the graft becomes finer and more complex at 3 months postoperatively. Time after dye injection: (A) 153.5 seconds, (B) 83.1 seconds, (C) 105.7 seconds.

Figure 4

 

Angiographic features before (A–C), and 1 (D–F) and 3 (G–I) months after surgery in various pterygia. New marginal corneal vascular arcades and the graft vasculature are reconstructed 3 months after LCA transplantation. (A, D, G) Primary and stage I pterygium. (B, E, H) Primary and stage II pterygium. (C, F, I) Primary and stage III pterygium. Time after dye injection: (A) 56.5 seconds, (B) 56.5 seconds, (C) 58.6 seconds, (D), 59.5 seconds, (E) 58.1 seconds, (F) 56.7 seconds, (G) 58.2 seconds, (H) 61.5 seconds, (I) 63.1 seconds.

Table 3

 

Quantitative Analysis of the Reorganization of the Vessels After Surgery

Parameters	Pterygium Body and Graft			Limbus
	Preoperative*	1 mo†	3 mo‡	1 mo§	3 mo||
Vascular area, mm2	1.61 (1.55–1.66)	1.15 (0.98–1.24)	1.47 (1.40–1.55)	0.09 (0.08–0.13)	0.17 (0.13–0.20)
Vascular density, %	40.35 (38.75–41.53)	28.85 (24.58–31.11)	36.76 (35.08–38.73)	12.11 (11.88–18.19)	22.89 (17.59–28.35)
Lacunarity	0.108 (0.093–0.116)	0.235 (0.203–0.324)	0.134 (0.114–0.158)	1.406 (0.930–2.005)	0.784 (0.357–1.015)

 Values are median (interquartile range). Wilcoxon's signed-rank test: P < 0.001 between * and †, † and ‡, and § and ||.

Variants in Angiographic Features and Related Clinical Features

Excessive bleeding developed intraoperatively in one eye (primary and stage II pterygium) and was controlled with bipolar cauterization. Reperfusion vessels did not appear 1 week after surgery in this eye. Reperfusion of the graft was not achieved on ICGA at 1 month postoperatively, and the MCAs were absent in the affected limbus (Figs. 5A–C). Biomicroscopic examination revealed contracture of the graft. In two eyes (primary and stage I pterygium), the remaining pre-existing episcleral vessels in the affected limbus were clearly visible on ICGA whereas the graft was hypofluorescent 1 week postoperatively. One eye developed G3 recurrence and another developed G2 recurrence within 3 months postoperatively (Figs. 5D–F). 

Figure 5

 

(A–C) A case of reperfusion failure. (A) A primary and stage II pterygium is visible. Reperfusion vessels are not visible in the graft or affected limbus at 1 week (B) and 1 month (C) postoperatively. (D–F) A case of G3 recurrence after surgery. (D) A primary and stage I pterygium is visible. (E) The remaining vessels in the affected limbus are clearly visible (arrowheads), while the graft is hypofluorescent at 1 week postoperatively. (F) G3 recurrence of the pterygium and deteriorated marginal corneal vascular arcades (arrows) are noted 3 months postoperatively. Time after dye injection: (A) 62.9 seconds, (B) 70.2 seconds, (C) 76.7 seconds, (D), 54.1 seconds, (E) 58.1 seconds, (F) 56.7 seconds.

Another eye (primary and stage II pterygium) exhibited complete reperfusion of the graft 3 months postoperatively, but poorly restored MCAs were observed in the affected limbus. In another eye (primary and stage II pterygium), the restored MCAs developed 2 mm posterior to the imaginary ellipse. In both of these eyes, no apparent abnormalities were observed on biomicroscopic examination. 

Discussion

Pterygium formation recently has been regarded to be a corneal LSC disorder with premalignant features.⁷ Mutation in the p53 pathway of apoptosis of LSCs has been suggested as an early event in the pathogenesis of pterygia. Altered LSCs produce matrix metalloproteinases, thereby leading to the initial cleavage of collagen in Bowman's layer and having a key role in the centripetal migration of the limbal cells.²¹ These cells also produce various fibroangiogenic cytokines, thereby leading to subsequent migration of the conjunctival epithelium.^4–6 In the era of LCA transplantation, complete removal of altered LSCs and fibrovascular tissue, and successful transplantation of healthy LSCs are the keys to successful treatment of pterygia.^12,22 

Although the genetic and molecular features of pterygia and the efficacy of LCA transplantation in pterygium surgery have been widely studied, there has been little further work on the in vivo angiographic features of the pterygium and affected limbus, or limbal reconstruction and graft reperfusion after surgery. These were the central questions of the present study. Previous studies using ICGA evaluated only the graft perfusion after pterygium surgery or had too few cases.^23,24 

The ICG has an absorption and emission spectrum in the near-infrared wavelength, allowing for more efficient transmission through the tissues than achieved with fluorescein.^25,26 The device used in our study enabled us to focus on the conjunctival and episcleral layers, and provided higher-quality images than devices used in previous studies. These advantages helped us to efficiently evaluate the structures of the MCAs in the affected limbus and the angiographic features of the pterygium. 

The angiographic characteristics revealed in the present study differed from those of previous reports. Chan et al.²³ reported that a single feeder vessel originating from the anterior conjunctival circulation was found in six of nine patients and that no dye leakage was demonstrable in the late phase of ICGA in all cases. However, the vascular plexus of most pterygia originated from multiple conjunctival vessels at each edge of the pterygium body in the present study. Multiple hyperfluorescent spots were present in the head and body of the pterygium, and leakage of dye was demonstrated later. We believe that these findings correspond to the established features of pterygia, such as chronic inflammation and fibrovascular proliferation. 

Deteriorated MCAs were detected in the affected limbus, and the vessel loops and stromal projection of the MCAs extended toward the cornea and beyond the head of the pterygium. These findings support growth of the pterygium from altered LSCs.²¹ These findings also imply that the blood supply of the innermost head of the pterygium originates from the episcleral vessels because the MCAs arise from the anterior branches of the episcleral circle.^13,14 The relevant literature supports our findings. The altered LSCs induce the dissolution of Bowman's layer by producing matrix metalloproteinases and migrate toward the cornea.²¹ An elevated corneal epithelium and dissolution of Bowman's layer were demonstrated in previous histological and optical coherence tomography (OCT) studies.^21,27 The migrated LSCs produce angiogenic cytokines, resulting in neovascularization and remodeling of the MCAs in the affected limbus. Subsequent migration of the conjunctival epithelium is facilitated by a blood supply from the anterior conjunctival vessels; therefore, the head of the pterygium has a dual blood supply from the conjunctival and episcleral circulations. Neovascularization below the corneal epithelium under the head of the pterygium also has been suggested previously.²¹ Our variant cases also support this concept. Although we attempted to completely excise the pterygium grossly under the surgical microscope, the remaining vessels in the affected limbus were demonstrated clearly by ICGA 1 week postoperatively in the two above-described cases. Such findings imply incomplete removal of the pre-existing altered LSC niche and its associated vasculature, which are the components of the pterygium. These eyes developed G2 and G3 recurrence, respectively, within 3 months. In contrast to these cases, the pterygium did not recur in any of the 26 eyes exhibiting typical reconstruction pattern throughout 12 months postoperatively. Taken together, we considered that the remaining altered LSC contributed to the early recurrence of the pterygium. 

We have presented the reconstruction process of the MCAs in the affected limbus; this is a novel finding of the present study. The deteriorated MCAs were not detected by ICGA 1 week postoperatively, but the engorged vessels arose from the adjacent episcleral circulation and extended to the limbus with multiple hyperfluorescent spots. Reconstruction of the MCAs had begun to appear 1 month postoperatively and continued for 3 months postoperatively as shown by ICGA. In terms of graft reperfusion, the ICGA findings indicated that early reperfusion originating from the episcleral vessels and the remodeling of the blood vessels continues for 3 months postoperatively. A previous OCT study also showed that the thickness of the graft continued to decrease up to 3 months postoperatively.²⁸ Quantitative analysis of the vasculature showed higher vascular density and lower lacunarity (more complex) in the limbus and graft at 3 months than at 1 month postoperatively. These findings also indicated that the reorganization of the vessels continued for 3 months after surgery. However, limited reorganization of the vasculature was observed in the graft and affected limbus at 3 months postoperatively compared to the control group (P < 0.001 for both). 

The LCA was not reperfused until 1 month postoperatively in one case and became contracted. We agree that excessive cauterization of the episcleral vessels can result in graft failure.²³ Another factor that can affect vascular reconstruction is the mitomycin C. Previous studies reported that a single intraoperative application of mitomycin C causes mild side effects, including delayed conjunctival wound healing (1–2 weeks) and mild conjunctival avascularity.^29,30 Therefore, we believed that the applied mitomycin C would affect the postoperative ICGA features in some cases, particularly in early postoperative periods. It might contribute to the variant cases featuring the reperfusion failure and poorly reconstructed MCAs in the present study. 

The major limitation of this study was that we assessed the angiographic features of the limbus after use of a single surgical method (LCA transplantation). Therefore, we cannot explain the differences in postoperative limbal reconstruction between LCA transplantation and other surgical procedures, such as the bare sclera method, amniotic membrane graft transplantation, or conjunctival autograft transplantation. Although the previous randomized trials suggested that the inclusion of LSCs in the conjunctival autograft would achieve better anatomical and functional reconstruction of limbus by restoring the barrier function, we cannot conclude the role of the transplanted LSCs on reconstruction of the MCAs at this time.¹⁰ 

Despite several limitations, the present investigation had several strengths above and beyond previously published data. First, to our knowledge this is the first documented study to use ICGA to evaluate the affected limbus in eyes with pterygia. Second, our work confirmed the findings of smaller pilot studies and presents several novel findings. We have successfully demonstrated the angiographic features of the affected limbus and head of the pterygium as well as limbal reconstruction after LCA transplantation in vivo. We also demonstrated that complete removal of the altered LSC and its associated vasculature resulted in good surgical outcome without recurrences. These findings are consistent with current concepts in the pathogenesis of pterygia and clinically relevant for pterygium surgery.^7,21 Third, our study implies the clinical usefulness of the ICGA in the in vivo investigation regarding the disease process or efficacy of the ocular surface reconstruction in various anterior segment diseases with limbal deficiencies. 

In conclusion, the MCAs were deteriorated in the affected limbus, and the pterygium head had a dual blood supply from the conjunctival and episcleral circulations. The remodeling of the LCA and affected limbus continued up to 3 months postoperatively. 

Acknowledgments

Supported by the Soonchunhyang University Research Fund. The authors alone are responsible for the content and writing of the paper. 

Disclosure: Y.J. Kim, None; S.H. Yoo, None; J.K. Chung, None