Conjunctival and Limbal Conjunctival Autograft vs. Amniotic Membrane Graft in Primary Pterygium Surgery: A 30-Year Comprehensive Review

Pterygium is a UV-sunlight-induced benign abnormal growth of epithelial and fibrovascular tissue migrating towards the central cornea [1]. Its global prevalence increases with age, ranging between 3% and 19.5% (mean of 12%) [2]. Impaired vision, induced astigmatism, and recurrent inflammation are among the common complications of a pterygium [3]. Even though medical treatment is crucial to relieve eye irritation and discomfort, surgical treatment remains the only option to remove a pterygium. Indications for surgical intervention include loss of vision secondary to induced astigmatism or to the involvement of the central part of the cornea; a restriction in ocular motricity; cosmetic problems; and suspicious features which can reveal an underlying squamous cell carcinoma [4].

Several surgical modalities for the treatment of pterygia have been developed over the years, ranging from the bare sclera technique first documented by Celsus in 25 AD to the PERFECT technique (pterygium extended removal followed by extended conjunctival transplant) described by Hirst et al. in 2009 [5, 6]. Growth over the cornea is proposed to be due to a destruction of the barrier limbal tissue, which forms the basis for the inclusion of limbal stem cells in the limbal conjunctival autograft (LCA) technique, with lower recurrence rates for both primary and secondary pterygia [7]. Recently, human amniotic membranes have gained increased popularity for the management of many ocular surface disorders, including coverage of the conjunctival defect after pterygium excision, or even for their use in combination with other techniques such as conjunctival autograft [8]. However, the efficacy of amniotic membrane graft (AMG) in preventing recurrences varies widely in the literature and depends on the surgical procedure, fixation technique, adjuvant treatment and/or study population.

Meta-analysis and systemic reviews of randomized controlled trials (RCTs) have failed to show a statistical superiority of LCA over AMG for primary pterygium surgery [9, 10]. Furthermore, RCTs have strict inclusion criteria and may not be representative of patients seen in routine clinical practice. Therefore, the place of AMG in the surgical therapeutic arsenal remains to be defined. The main objective of this review is to compare the “real-life” effectiveness of free conjunctival autograft (CAT) with LCA, AMG and combined surgical techniques in terms of pterygium recurrence based on the retrospective published data in the literature, providing new insights and perspective for the management of this very common disease.

This section is divided into subsections that treat the results of each surgical technique separately. Overall, 35 subgroups comprising a total of 3747 treated eyes met all the inclusion criteria and were included in the final review. Mean pterygium recurrence was assessed globally (global recurrence rate; Fig. 2), taking into account all included papers, and was then recalculated for each fixation technique separately (specific recurrence rate; Fig. 3). Except for one publication by Okoye et al. [33], all included papers had a minimum follow-up of 6 months (range, 3–48 months).

Pterygium recurrence is a complication of pterygium surgery. Even though the predicting factors are not fully understood, some have been identified, including ethnicity (Hispanic and dark-skinned patients), younger age, current active growth, preexisting disfiguration of the lacrimal caruncle, ocular motility restriction, ocular surface inflammation and genetic predisposition [40]. Surgery-related factors have also been studied. They include appropriate removal of the proliferative epithelial cells that alter the corneal limbal stem cells; thorough removal of the subconjunctival fibrovascular tissue with or without the use of adjunctive therapies; appropriate covering of the bare sclera; and adequate postoperative inflammation control [4].

Although several surgical techniques to cover the surgical site have been described over the years, CAT, LCA and AMG have gained worldwide acceptance. In this “real-world” review of 35 subgroups of patients, comprising a total of 3747 primary pterygia, we compared the effectiveness of these techniques with respect to pterygium recurrence. Meta-analysis and systemic reviews have already been conducted and did not show a statistical superiority of LCA over AMG or the converse [9, 10]. A meta-analysis of 13 RCTs in Cornea (878 eyes) compared LCA to other surgical techniques in primary and recurrent pterygia [9]. It excluded non-randomized, non-controlled trials and those combining two different techniques. The authors found that the recurrence rate of LCA was not statistically different than that of AMG, with an odds ratio of 0.96 for primary pterygia. Similarly, a Cochrane systemic review published in 2017 did not show any superiority of CAT or AMG for primary pterygia at 3 and 6 months (6 RCTs, 538 eyes and 10 RCTs, 1021 eyes, respectively), with a relative risk of 0.58 (0.27–1.27) at 6 months [10]. On the other hand, another meta-analysis published in 2012 showed different results, with a superiority of the CAT technique over AMG (hazard ratio 0.30) [41]. However, the paper included both RCTs and retrospective cohorts, including a study by Tananuvat et al. [42] with an abnormally high recurrence rate (40.9%) for patients treated with AMG fixated with vicryl 8.0 sutures, which are known to be pro-inflammatory. According to the authors, this particularly high recurrence rate could be partly explained by the high sun exposure of the studied population, which was mainly composed of agricultural workers in a tropical area.

In our review, the real-world recurrence rates for AMG, CAT and LCA were 9.0%, 7.61% and 5.50%, respectively. Amniotic membranes have several biological properties, including anti-angiogenic, anti-inflammatory and anti-fibrotic effects. Also, because of their structure and low immunogenicity, they work as a substrate for conjunctival epithelial adherence, migration, and differentiation [43]. Different mechanisms by which amniotic membranes can reduce pterygium recurrence have been described. Wound healing can be generally considered an inflammatory process in its earlier stages. The use of an amniotic membrane provides an anti-inflammatory environment and promotes epithelialization [44]. Also, the stroma of the amniotic membrane is thought to inhibit tissue growth factor beta (TGF-β) signaling and reduce extracellular matrix production [28]. When placed directly in contact with the Tenon’s fibroblasts, it suppresses myofibroblast differentiation and reduces scarring [45]. Amniotic membranes can therefore be considered an alternative to conjunctival grafting without compromising efficacy. In our review, only one study used lyophilized amniotic membranes [25], and further studies are therefore needed to compare their efficacy to that of cryopreserved ones in pterygium management. However, it is noteworthy that experimental studies have shown that lyophilized and cryopreserved membranes have similar amounts of the major growth factors that play a role in epithelial healing [46‐48], and recent clinical studies showed similar rates of superficial ulcer healing for the two types of membranes [49]. Furthermore, a comparative study showed that both can be safe and effective adjunctive treatments for primary pterygium over a 22-month follow-up [50].

The recurrence rate after surgery does not depend solely on the biomaterial used, but also on the concomitant use of adjunctive treatments such as MMC and/or on the fixation technique [51].

In our review, MMC 0.02% was used in only two subgroups of patients. However, wide variations of the recurrence rates were found when sutures and fibrin glue were treated separately. Despite its anti-inflammatory effect, the fixation of the amniotic membrane with vicryl sutures yielded very high specific recurrence rates (23.0%), but MMC was used in none of these studies. In fact, a meta-analysis by Song et al. showed that amniotic membrane transplantation with MMC was associated with lower recurrence rates compared to amniotic membrane alone, but the fixation technique was not specified in the paper [52]. It should be noted, however, that the use of MMC for primary pterygium surgery is arguable, and the lowest effective concentration and contact time should be employed in order to limit possible side effects [53].

In our review, vicryl sutures were used more frequently than nylon 10–0, except for the LCA technique where the two were combined. This is probably due to the fact that vicryl sutures have the advantage of being absorbed within 60 days, eliminating the need for suture removal. While both types of sutures can cause a conjunctival reaction in the postoperative period, vicryl ones are known to be pro-inflammatory and to enhance the TGF-β pathway [54], promoting myofibroblast transformation and pterygium recurrence. Human fibroblasts, on the other hand, produce a fibrin clot that promotes tissue adhesion and minimizes inflammation [55]. The recurrence rate of AMG fixated with fibrin glue is lower than that of CAT and similar to that of LCA (3.60%). Even though only one monocentric study was included, it reported the results of a large cohort of 527 eyes (535 patients) from a high-risk population with a mean follow-up duration of 17.3 months. The reason why sutures are more inflammatory when associated with AMGs rather than with conjunctival grafts remains unclear. One possible explanation could be that, unlike CAT, which does not require an inflammatory re-epithelialization phase, a sutured AMG requires both a conjunctival epithelial healing phase and a suture resorption phase, which are both associated with inflammation.

Associating AMG with fibrin glue for primary pterygia has a double advantage: the use of AMG preserves the conjunctiva of the patient in case they require a future glaucoma procedure or in case of recurrences, and the use of fibrin glue reduces surgical time and minimizes postoperative discomfort and inflammation [56, 57]. However, it is noteworthy that fibrin glue is more expensive than sutures and is not available in certain countries or surgical settings.

Recent evidence suggests that the combination of AMG with CAT or LCA can achieve an even lower recurrence rate than what can be achieved with either technique alone. In our review, the mean recurrence rate for papers combining CAT or LCA with AMG was the lowest (1.83%), and dropped even further to 1.22% when the biomaterial was glued, offering what is possibly the best available surgical option to date. Further evaluation will be required to confirm these results.

Different surgical techniques combining AMG with conjunctival graft for primary pterygia have been described over the years. In their large cohort of 989 eyes, Fernandes et al. described 20 patients in which the CAT was harvested and transferred to the recipient bed over the AMG with the epithelial side up and then anchored to the surrounding conjunctiva with 8–0 vicryl sutures, with a 15% recurrence rate [28]. In 2011, Taylan et al. prospectively followed 30 patients with primary pterygia treated with excision followed by an inlay AMG to cover the conjunctival side of the bare sclera combined with a narrow-strip CAT to cover the limbal part of the bare sclera; both tissues were fixated using fibrin-glue tissue adhesive (recurrence rate 6.67%) [58]. Other surgical techniques have also been described, including CAT with an overlay cryopreserved or lyophilized AMG (recurrence rates of 7.10% and 3.00%, respectively); mini simple limbal epithelial transplantation followed by an overlay AMG (recurrence rate of 2.50%); and more recently, LCA and dehydrated AMG inserted into a subconjunctival pocket created around the periphery of the excision site in 493 eyes (recurrence rate of 1.20%) [36, 59‐61].

Regardless of the technique used, extensive removal of the Tenon can be beneficial. In fact, Hirst et al. followed 1000 consecutive patients (806 primary pterygium) undergoing extended pterygium removal encompassing the Tenon layer followed by an extended conjunctival transplant, and noted a near-zero recurrence rate [62]. The extent of Tenon resection is barely described in the different studies. This could be a confounding factor when interpreting their results.

Our review has several limitations. “Real-life” observational studies better represent real-life patients but have a lower level of evidence than interventional trials. Data from such papers can be statistically poor because of their retrospective design but also due to their heterogenicity. In fact, the follow-up time and the definition of recurrence varied widely between authors. We tried to compensate for this bias and to strengthen the relevance of our findings by analyzing a significant number of papers. Also, except for one paper, the follow-up was at least 6 months for all studies and reached 46 months for Masters et al. [30]. We tried to homogenize the different recurrence definitions by recalculating the corneal recurrence rate for all papers. Some factors, such as ethnicity, sun exposure or surgeon’s experience, could still have affected our results. Moreover, our review only considered the pterygium recurrence rate, not the postoperative esthetic appearance. Other complications have been reported in the literature, such as graft dehiscence, graft detachment and granulomas. Finally, another limitation would be the small sample of patients in some subgroups. Nevertheless, this bias was minimized by the overall number of included eyes (3747 eyes).

In conclusion, based on our review, AMG seems like a reasonable option that could be considered in primary pterygium surgery. It can be used alone, glued to the underlying sclera with its epithelial side up, with the advantage of preserving the patient’s conjunctiva, particularly in the case of subsequent glaucoma surgery or in the case of double-headed pterygia. If the AMG is to be sutured, the use of thinner 10.0 sutures with MMC is encouraged, but physicians should be aware of its various side effects. The combination of AMG with CAT or LCA seems to yield very low recurrence rates, but further studies are needed to confirm these results.