Background: Phototherapeutic keratectomy (PTK) has been increasingly used to treat severe recurrent corneal erosion syndrome (RCES) patients who do not respond to other treatments. However, the efficacy and complication of each study are currently uncertain due to varying rates. Objectives: The objective of this study was to investigate the safety and efficacy of the PTK for recurrent corneal erosions. Methods: This article performed a systematic literature research in Cochrane, Embase, PubMed, Scopus, and the Web of Science for the literature on PTK treatment of RCES until December 20, 2022. The extracted data including recurrence rate and the adverse event rate were used for meta-analysis. Results: The recurrence rate was 18% (95% CI, 13%–24%) (129/700 eyes). Subgroup analysis showed that the RCE recurrence was 17% (95% CI, 9%–24%) after trauma and 22% (95% CI, 11%–32%) in the corneal dystrophy group. Treatment-related adverse events included subepithelial haze, hyperopic shift, and decrease of the best spectacle-corrected visual acuity. In this study, the incidence of these events was 13% (95% CI, 6%–21%), 20% (95% CI, 11%–28%), and 11% (95% CI, 5%–16%), respectively. Conclusions: PTK represented a valuable treatment option for patients with recurrent corneal erosions, especially those with traumatic injuries, which had minimal side effects.

Recurrent corneal erosion syndrome (RCES) is a common clinical disorder characterized by episodes of corneal epithelial defect accompanied by pain, photophobia, tearing, and blurred vision [1]. Moreover, most patients have a history of superficial corneal trauma or epithelial basement membrane dystrophy [2]. Conservative treatments, such as topical lubrication, hypertonic agents, and bandage soft contact lenses, were usually effective for recurrent erosion. Despite conservative treatment, a large proportion of patients continue to have minor and major recurrences [3], which can last on a variable time scale ranging from minutes to hours to days, and recur frequently or coalesce into persistent distress [1]. Therefore, the recurrence of corneal erosion remains an important issue. Surgical treatments are available for these refractory patients, including epithelial debridement (ED), anterior corneal stromal needle puncture (ASP), neodymium aluminum-yttrium-garnet laser anterior stromal puncture, superficial keratectomy with diamond burr (DB), and phototherapeutic keratectomy (PTK).

PTK was first approved by the US Food and Drug Administration (FDA) in 1995 for treating anterior corneal pathologies [4]. It has been a useful tool in the treatment of recurrent corneal erosions because it ablates corneal tissue with extremely high precision and minimal adjacent tissue damage [5]. Some scholars have studied the mechanisms [6, 7]. Marshall et al. [8] showed that excimer laser local ablation of Bowman layer provided a smooth new bed for the migrating epithelium, resulting in newly formed hemidesmosomal adhesion complexes. The efficacy and safety of PTK in RCES was first demonstrated by Dausch et al. [9]. Since then, an increasing number of studies have been reported. Most studies reported treating recurrent erosions with a recurrence-free rate of 46%–100% after surgery. Development of haze, hyperopic shift, and loss of vision after surgery were the reported complications. Hyperopic shift was a result of central flattening of the cornea caused by PTK, which was related to the ablation depth. When planning ablation depths, it is important to consider that the objective of the surgery is to remove any remnants of redundant epithelial basement membrane associated with poor adherence of the corneal epithelium to the underlying stroma, which may remain after thorough debridement of the epithelium and underlying basement membrane using a scalpel. Therefore, unless there is associated stromal haze, the PTK treatment should be limited to only a small number of pulses at each location on the exposed stromal surface. This approach helps avoid refractive shifts, and the limited ablation depth may also account for the lower incidence of haze after PTK compared to PRK.

A previous systematic review of interventions for recurrent corneal erosions was conducted to evaluate the effectiveness and adverse effects of protocols for the prophylaxis and treatment of RCE. However, the size and quality of the studies included in Watson et al. [10] were not sufficient to provide firm evidence for the development of management guidelines. Several series have detailed the use of the PTK in patients with recurrent corneal erosions, and the main goal of these clinical trials was to establish the efficacy and safety of the procedure in a carefully selected group of patients. However, these studies were relatively small in scale and only represented early experience, which were still different from practical applications. To assess the technical success rate, efficacy, and safety of PTK in patients with RCE of different etiologies, this article performed a systematic review and meta-analysis of published series.

Search Strategy

The search terms “Phototherapeutic Keratectomy,” “Keratectomies,” “Keratectomy,” “recurrent corneal erosions,” and “Corneal Erosions, Recurring Hereditary” were searched for eligible publications from the database of Cochrane, Embase, PubMed, Scopus, and the Web of Science. There was no time limit for searching until the final search date on December 20, 2022, with most editorials and reviews removed. In addition, the reference list of applicable studies was manually checked for inclusion in other articles. Two researchers commonly completed this search process.

Inclusion and Exclusion

Inclusion criteria were as follows: first, patients with recurrent corneal erosions diagnosed and confirmed by slit lamp microscopy; second, clinical trials or prospective/retrospective cohort series studies; third, patients who had received the PTK after recurrence; fourth, studies that reported the outcomes of patients with at least one of the recurrence rates and surgery-related complications. Exclusion criteria were as follows: first, negative diagnosis or diagnosis that mixed with other influential diseases; second, therapies that included other operational procedures; third, inconsistent patient baseline data; fourthly, unobtained full-text articles or unavailable data; fifth, review articles, guidelines, case reports, meeting abstracts, and comments.

Study Selection Process

Two researchers independently screened these studies by title and abstract. For the articles excluded with uncertainty, the full text was further examined to screen eligible studies. Disagreements were resolved by consensus or a third reviewer. In studies that included treatments other than PTK, data were abstracted only from patients treated with PTK.

Data Extraction and Outcome Measures

Baseline patient characteristics were extracted from each study, including the cause of recurrent corneal erosion, number of patients and eyes treated with PTK, sex, mean age, and length of follow-up. To evaluate the efficacy of PTK, primary and secondary outcomes were identified. Painful erosion after the PTK procedure was chosen as the primary outcome because the primary goal of treating RCES was to prevent further episodes. Safety outcomes were surgical complications, including subcutaneous opacity, hyperopic displacement, and vision loss after surgery. According to the length of follow-up, it can be divided into long-term follow-up and short-term follow-up. Short-term follow-up was defined as ≤1-year follow-up, and >1-year follow-up was considered long-term follow-up. Low-quality studies included those with high risk of bias and those with moderate risk of bias, and those with low risk of bias were considered to be high-quality studies. Recurrence of erosions was defined as either the presence of erosion on examination or severe symptoms of erosion on patient history. The decrease of the BSCAV was considered optimal for correcting vision loss greater than or equal to 1 line.

Study Risk of Bias Assessment

The first 8 item of the MINORS were assessed by two independent authors on the quality of each study included. We focused on the following essentials: a clearly stated aim; inclusion of consecutive patients; prospective collection of data; endpoints appropriate to the aim of the study; unbiased assessment of the study endpoint; follow-up period appropriate to the aim of the study; loss to follow-up less than 5%; prospective calculation of the study size. Each item was 2 points, the total score was 16 points, and studies with 10 points or more entered our research. Table 1 summarizes the literature quality evaluation situation.

Table 1.

MINORS quality evaluation for included studies

StudyClear purposePatient continuityData collectionAppropriate endpointObjective evaluation endpointAdequate follow-up timeLow lost to follow-up rateSample size estimationTotal score
Yang et al. [11] (2022) 12 
Song et al. [12] (2020) 13 
Mehlan et al. [13] (2016) 12 
Dedes et al. [14] (2015) 12 
Chan et al. [15] (2014) 14 
Suri et al. [16] (2013) 13 
Nassaralla et al. [17] (2012) 12 
Baryla et al. [18] (2006) 12 
Pogorelov et al. [19] (2006) 14 
Chow et al. [20] (2005) 14 
Sridhar et al. [5] (2002) 12 
Stewart et al. [21] (2002) 13 
Rashad et al. [22] (2001) 13 
Jain and Austin [23] (1999) 13 
Cavanaugh et al. [24] (1999) 13 
Morad et al. [25] (1998) 12 
Forster et al. [26] (1997) 11 
Lohmann et al. [27] (1996) 11 
Bernauer et al. [28] (1996) 12 
O’Brart et al. [29] (1994) 13 
Fagerholm et al. [30] (1993) 12 
StudyClear purposePatient continuityData collectionAppropriate endpointObjective evaluation endpointAdequate follow-up timeLow lost to follow-up rateSample size estimationTotal score
Yang et al. [11] (2022) 12 
Song et al. [12] (2020) 13 
Mehlan et al. [13] (2016) 12 
Dedes et al. [14] (2015) 12 
Chan et al. [15] (2014) 14 
Suri et al. [16] (2013) 13 
Nassaralla et al. [17] (2012) 12 
Baryla et al. [18] (2006) 12 
Pogorelov et al. [19] (2006) 14 
Chow et al. [20] (2005) 14 
Sridhar et al. [5] (2002) 12 
Stewart et al. [21] (2002) 13 
Rashad et al. [22] (2001) 13 
Jain and Austin [23] (1999) 13 
Cavanaugh et al. [24] (1999) 13 
Morad et al. [25] (1998) 12 
Forster et al. [26] (1997) 11 
Lohmann et al. [27] (1996) 11 
Bernauer et al. [28] (1996) 12 
O’Brart et al. [29] (1994) 13 
Fagerholm et al. [30] (1993) 12 

Statistical Analysis

All analyses were performed using Stata v.15, and forest plots were drawn. The cumulative incidence (event rate per patient at the end of the study) and 95% confidence intervals were estimated for each study. Due to significant heterogeneity of populations and interventions across various included studies, random-effects models were used to summarize incidence across studies. The I2 statistic was used to express the proportion of inconsistency not attributable to chance. p < 0.10 or I2 > 50% suggested significant heterogeneity. Funnel plot and statistical test were used to detect small sample effect size and publication bias. p < 0.05 was considered statistically significant. We check the manuscript carefully according to the PRISMA and revise the manuscript to ensure full compliance with PRISMA standards.

Study Selection and Characteristics

A total of 1,177 articles relevant to the search terms were identified, and 657 articles were selected after removing duplicates. After eliminating nonrelevant articles by title and abstract screening, 71 articles were included for full-text screening. Twenty-one studies were included in our qualitative and quantitative analyses [11‒14, 16‒30], including 4 controlled experiments and 17 single-arm experiments. The intervention method was PTK, and the control was DB [5], PTK + peripheral anterior stromal puncture (ASP) [12], ED + ASP and ED + DB [11], or alcohol delamination [15]. The process of literature identification, screening, and selection is summarized briefly in Figure 1. All studies included in this study were based on moderate- to high-quality evidence. Table 2 provides a brief description of these 21 studies. In the included studies, both controlled and uncontrolled experiment scores were above 10. The quality of the literature supported the meta-analysis.

Fig. 1.

Flowchart of publication search and selection.

Fig. 1.

Flowchart of publication search and selection.

Close modal
Table 2.

Study information and patient characteristics

StudyCountryEtiologyInterventionPatients/eyesSex (female/male)Age, yearsLength of follow-up, monthsRecurrences (eyes)Complication
hazehyperopic shiftvisual activity decrease
Yang et al. [11] (2022) Canada EBMD PTK NA/30 19/11 53±16 17.8 (51–3,400 days) 3/30 1/30 2/30 NA 
Song et al. [12] (2020) Korea TA PTK 23/23 11/12 43.57±13.20 19.75±6.83 5/23 NA NA NA 
ABMD 
ID 
Mehlan et al. [13] (2016) Germany TA PTK 37/37 NA 41±11 12±12 (3–70 months) 3/37 NA NA NA 
EBMD 
Dedes et al. [14] (2015) Switzerland TA PTK 82/89 47/35 45.5±11.6 51.56±21.22 (24–91 months) TA: 8/55 NA NA NA 
EBMD EBMD: 14/29 
ID ID: 3/5 
Chan et al. [15] (2014) Australia TA PTK 16/16 7/9 47.9 17.25±8.4 6/16 8/16 NA 0/16 
EBMD 
Other 
Suri et al. [16] (2013) USA EBMD PTK NA/7 NA 46.5±14.6 9 (2–52 months) 2/7 2/7 NA NA 
Other 
Nassaralla and Nassaralla Junior [17] (2012) Brazil TA PTK 23/26 17/9 45.07±13.19 144±18.24 EBMD: 6/26 2/26 NA 2/26 
EBMD 
ID 
Baryla et al. [18] (2006) Canada ID PTK 33/39 13/20 47.1±11.24 17.4 (0.4–67.6 months) 16/39 4/39 NA NA 
TA 
EBMD* 
Pogorelov et al. [19] (2006) Germany EBMD PTK 11/15 8/3 31–68 57.6±36 0/15 1/15 0/15 0/15 
Chow et al. [20] (2005) China TA PTK 13/13 9/4 38.4 49.5 (24–48 months) 2/13 0/13 0/13 0/13 
EBMD 
Sridhar et al. [5] (2002) USA ABMD PTK 14/15 10/4 46.1±3.1 17.6±5.7 (0.7–82.2 months) 4/15 5/15 NA 0/15 
Stewart et al. [21] (2002) UK ABMD PTK 13/19 NA 48.6 22.3 (12–48 months) 2/19 NA 4/19 2/19 
Rashad et al. [22] (2001) Egypt TA PTK 41/43 24/17 37.6±8.3 23.3±9.1 (12–48 months) 4/43 0/43 10/43 0/43 
EBMD 
ID 
Jain et al. [23] (1999) UK TA PTK 62/71 NA 39 24 (6–55 months) TA: 8/40 NA 17/71 9/71 
CD CD: 10/19 
ID ID: 6/12 
Cavanaugh et al. [24] (1999) USA ABMD PTK 34/36 23/11 NA 12 5/36 2/36 10/36 0/36 
Morad et al. [25] (1998) Israel TA PTK 23/23 19/4 50.56±13.87 38.43±14.08 (12–60 months) 4/23 0/23 NA 0/23 
EBMD 
Other 
Forster et al. [26] (1997) Germany TA PTK 92/103 62/41 24–84 12–36 months 9/98 (5 lost) 0/98 NA 0/98 
Other 
Lohmann et al. [27] (1996) Germany TA PTK 24/31 NA 43 3–12 months 1/31 0/31 NA 0/31 
EBMD 
Bernauer et al. [28] (1996) UK TA PTK 12/15 6/6 44.6 12.8 (3–28 months) CD: 3/10 0/15 0/15 0/15 
TA: 1/4 
CD 
ID 
Obrart et al. [29] (1994) UK TA PTK 15/17 11/4 43 11 (6–24 months) 4/17 5/17 NA 0/17 
ID 
Fagerholm et al. [30] (1993) Sweden NA PTK 37/37 NA 47 11.8 (6–28 months) 0/37 0/37 0/37 NA 
StudyCountryEtiologyInterventionPatients/eyesSex (female/male)Age, yearsLength of follow-up, monthsRecurrences (eyes)Complication
hazehyperopic shiftvisual activity decrease
Yang et al. [11] (2022) Canada EBMD PTK NA/30 19/11 53±16 17.8 (51–3,400 days) 3/30 1/30 2/30 NA 
Song et al. [12] (2020) Korea TA PTK 23/23 11/12 43.57±13.20 19.75±6.83 5/23 NA NA NA 
ABMD 
ID 
Mehlan et al. [13] (2016) Germany TA PTK 37/37 NA 41±11 12±12 (3–70 months) 3/37 NA NA NA 
EBMD 
Dedes et al. [14] (2015) Switzerland TA PTK 82/89 47/35 45.5±11.6 51.56±21.22 (24–91 months) TA: 8/55 NA NA NA 
EBMD EBMD: 14/29 
ID ID: 3/5 
Chan et al. [15] (2014) Australia TA PTK 16/16 7/9 47.9 17.25±8.4 6/16 8/16 NA 0/16 
EBMD 
Other 
Suri et al. [16] (2013) USA EBMD PTK NA/7 NA 46.5±14.6 9 (2–52 months) 2/7 2/7 NA NA 
Other 
Nassaralla and Nassaralla Junior [17] (2012) Brazil TA PTK 23/26 17/9 45.07±13.19 144±18.24 EBMD: 6/26 2/26 NA 2/26 
EBMD 
ID 
Baryla et al. [18] (2006) Canada ID PTK 33/39 13/20 47.1±11.24 17.4 (0.4–67.6 months) 16/39 4/39 NA NA 
TA 
EBMD* 
Pogorelov et al. [19] (2006) Germany EBMD PTK 11/15 8/3 31–68 57.6±36 0/15 1/15 0/15 0/15 
Chow et al. [20] (2005) China TA PTK 13/13 9/4 38.4 49.5 (24–48 months) 2/13 0/13 0/13 0/13 
EBMD 
Sridhar et al. [5] (2002) USA ABMD PTK 14/15 10/4 46.1±3.1 17.6±5.7 (0.7–82.2 months) 4/15 5/15 NA 0/15 
Stewart et al. [21] (2002) UK ABMD PTK 13/19 NA 48.6 22.3 (12–48 months) 2/19 NA 4/19 2/19 
Rashad et al. [22] (2001) Egypt TA PTK 41/43 24/17 37.6±8.3 23.3±9.1 (12–48 months) 4/43 0/43 10/43 0/43 
EBMD 
ID 
Jain et al. [23] (1999) UK TA PTK 62/71 NA 39 24 (6–55 months) TA: 8/40 NA 17/71 9/71 
CD CD: 10/19 
ID ID: 6/12 
Cavanaugh et al. [24] (1999) USA ABMD PTK 34/36 23/11 NA 12 5/36 2/36 10/36 0/36 
Morad et al. [25] (1998) Israel TA PTK 23/23 19/4 50.56±13.87 38.43±14.08 (12–60 months) 4/23 0/23 NA 0/23 
EBMD 
Other 
Forster et al. [26] (1997) Germany TA PTK 92/103 62/41 24–84 12–36 months 9/98 (5 lost) 0/98 NA 0/98 
Other 
Lohmann et al. [27] (1996) Germany TA PTK 24/31 NA 43 3–12 months 1/31 0/31 NA 0/31 
EBMD 
Bernauer et al. [28] (1996) UK TA PTK 12/15 6/6 44.6 12.8 (3–28 months) CD: 3/10 0/15 0/15 0/15 
TA: 1/4 
CD 
ID 
Obrart et al. [29] (1994) UK TA PTK 15/17 11/4 43 11 (6–24 months) 4/17 5/17 NA 0/17 
ID 
Fagerholm et al. [30] (1993) Sweden NA PTK 37/37 NA 47 11.8 (6–28 months) 0/37 0/37 0/37 NA 

TA, trauma; ABMD, anterior basement membrane dystrophy; ASP, anterior stromal puncture; RCES, recurrent corneal erosion syndrome; EBMD, epithelial basement membrane dystrophy; ID, idiopathic; ALD, alcohol delamination; DB, diamond burr; CD, corneal dystrophy; NA, not available.

*One patient had both trauma and a dystrophy.

In this study, a total of 642 patients with 705 eyes suffering from recurrent corneal erosion were included (Table 2). The sample sizes in these studies ranged from 7 to 103. These studies were conducted in six continents: Africa (n = 1), Europe (n = 10), North America (n = 5), South America (n = 1), Oceania (n = 1), and Asia (n = 3). The recurrence rate in all studies ranged from 0 to 41%, and the follow-up time ranged from 3 months to 14 years. Most of them had a history of more than three symptomatic episodes of RCE despite conventional treatments, such as artificial lubricants, patching, ED, and therapeutic contact lenses. Data from the included studies are summarized in the Table 2.

Recurrence Outcomes

All included studies reported values for recurrent rates. A total of 571 patients showed complete resolution of RCE symptoms after PTK. The mean length of follow-up ranged from 9 months to 12 years. It should be noted that 5 patients (five eyes) were not available for postoperative follow-up in the study of Forster et al. [26]. The meta-analysis showed that the primary outcome (recurrence after PTK) occurred in 18% of cases (95% CI, 13%–24%) (129/700 eyes), with a random-effects model (I2 = 70.4%, p = 0.000; Fig. 2). Median follow-up time was reported in all studies. The study was divided into long-term and short-term follow-ups depending on whether the median follow-up time was more than 1 year. Subgroup analysis showed higher recurrence rates at long-term follow-up (long-term follow-ups: 21% [95% CI, 15%–27%; 114/535 eyes; Fig. 3]; short-term follow-ups: 10% [95% CI, 3%–17%; 15/165 eyes; Fig. 4]).

Fig. 2.

Pooled recurrence rate after PTK.

Fig. 2.

Pooled recurrence rate after PTK.

Close modal
Fig. 3.

Recurrence rate in patients with long-term follow-up.

Fig. 3.

Recurrence rate in patients with long-term follow-up.

Close modal
Fig. 4.

Recurrence rate in patients with short-term follow-up.

Fig. 4.

Recurrence rate in patients with short-term follow-up.

Close modal

Of these 21 studies, 10 reported the underlying etiology of RCE (10 studies included corneal dystrophy, 3 studies included post-traumatic RCE) [5, 11‒14, 17, 19, 21, 23, 24]. Therefore, two etiology of RCE were classified, which were caused by trauma in 99 eyes and corneal dystrophy in 230 eyes. Subgroup analysis showed a recurrence rate of 17% post-traumatic RCE (95% CI, 9%–24%; 17/99 eyes) and 22% (95% CI, 11%–32%; 48/230 eyes) in the patient group with corneal dystrophy. Compared to patients with a traumatic cause of epithelial erosions, the likelihood to experience a recurrence was higher in patients with corneal dystrophy (Fig. 5).

Fig. 5.

Recurrence rate in the post-traumatic or corneal dystrophy group. TA, trauma; CD, corneal dystrophy.

Fig. 5.

Recurrence rate in the post-traumatic or corneal dystrophy group. TA, trauma; CD, corneal dystrophy.

Close modal

Furthermore, we reviewed the included literature and extracted information related to survival curves. Based on the information provided in the articles, we generated Kaplan-Meier curves (Fig. 6).

Fig. 6.

Kaplan-Meier survival curve demonstrating the time to recurrence of recurrent corneal erosion syndrome symptoms after PTK.

Fig. 6.

Kaplan-Meier survival curve demonstrating the time to recurrence of recurrent corneal erosion syndrome symptoms after PTK.

Close modal

Safety Outcomes

In all 21 studies, 7 literature works showed that no patients developed haze after PTK, and 9 publications [5, 11, 15‒19, 24, 29] reported the occurrence of subepithelial haze events after PTK (Fig. 7) with an overall incidence of 13% (95% CI, 6%–21%), which was developed in 30 of the 461 reported cases. But most complications were completely reversible during follow-up, and only a minority resulted in a permanent corneal haze. Nine studies [11, 19‒24, 28, 30] reported spherical equivalent manifest refraction change after PTK (Fig. 8), among which 5 publications [11, 21‒24] indicated a slight trend toward mild hyperopia with an overall incidence of 20% (95% CI, 11%–28%), while other four studies reported no significant change in apparent refraction after PTK.

Fig. 7.

Overall incidence of postoperative subepithelial haze.

Fig. 7.

Overall incidence of postoperative subepithelial haze.

Close modal
Fig. 8.

Overall incidence of postoperative hyperopic shift.

Fig. 8.

Overall incidence of postoperative hyperopic shift.

Close modal

A total of 438 patients from 14 studies [5, 15, 17, 19‒29] reported postoperative visual acuity outcomes. Most of them had a retained or improved best spectacle-corrected visual acuity (BSCVA), while 13 of 438 patients had deterioration of more than or equal to 1 line on the Snellen test (11%; 95% CI, 5%–16%). Eleven of these studies indicated that there were no vision-threatening complications (Fig. 9), and the statistical analysis showed no significant difference in BSCVA before and after PTK in most of studies (Table 3).

Fig. 9.

Overall incidence of postoperative visual acuity deterioration.

Fig. 9.

Overall incidence of postoperative visual acuity deterioration.

Close modal
Table 3.

Efficacy and safety of PTK treatment in patients with RCE

Studies, nSample size (eyes)Occurrence rate, % (95% CI)I2, %
Recurrence 
 Length of follow-up 
  ≤1 year 165 0.10 (0.03, 0.17) 42.9 
  >1 year 15 535 0.21 (0.15, 0.27) 68.6 
 Total 21 700 0.18 (0.13, 0.24) 70.4 
 Etiology of RCE 
  Trauma 99 0.17 (0.09, 0.24) 0.0 
  Corneal dystrophy 10 230 0.22 (0.11, 0.32) 80.3 
 Complication 
  Subepithelial haze 16 461 0.13 (0.06, 0.21) 65.9 
  Hyperopic shift 279 0.20 (0.11, 0.28) 60.2 
  Visual acuity deterioration 14 438 0.11 (0.05, 0.16) 0.0 
Studies, nSample size (eyes)Occurrence rate, % (95% CI)I2, %
Recurrence 
 Length of follow-up 
  ≤1 year 165 0.10 (0.03, 0.17) 42.9 
  >1 year 15 535 0.21 (0.15, 0.27) 68.6 
 Total 21 700 0.18 (0.13, 0.24) 70.4 
 Etiology of RCE 
  Trauma 99 0.17 (0.09, 0.24) 0.0 
  Corneal dystrophy 10 230 0.22 (0.11, 0.32) 80.3 
 Complication 
  Subepithelial haze 16 461 0.13 (0.06, 0.21) 65.9 
  Hyperopic shift 279 0.20 (0.11, 0.28) 60.2 
  Visual acuity deterioration 14 438 0.11 (0.05, 0.16) 0.0 

Although PTK has been used to treat RCES for nearly 40 years, and many studies have assessed its outcomes, the conclusions of each study were inconsistent and inconclusive. As these studies are relatively small and represent early experience, it is necessary to summarize previous studies to better understand the efficacy and safety of PTK. Therefore, this article performed a systematic review and meta-analysis of the published literature to more precisely estimate the success rate, efficacy, and safety of PTK treatment in patients with RCE of different etiologies.

This meta-analysis combined the outcomes of 642 patients with 705 eyes from 21 individual studies, indicating that PTK was safe and effective in treating patients with intractable recurrent corneal erosion, with an overall recurrence rate of approximately 18% (95% CI, 13%–24%). During long-term follow-up, more than 78% of patients disappeared completely without any erosion, suggesting that PTK can be an effective solution for RCES in general. By observing the forest plot, it can be seen that Baryla et al. [18] had the highest recurrence rate among the included studies, reporting a rate of 41%. On the other hand, two studies reported the lowest recurrence rate of 0%, namely, Pogorelov et al. [19] and Fagerholm et al. [30]. It is important to note that the variation in baseline characteristics of patients and the heterogeneity of PTK treatment within the same case series may introduce bias to these studies.

As stated by Baryla et al. [18], the higher recurrence rate may be due to a greater proportion of RCES caused by corneal dystrophy rather than trauma. The efficacy of PTK in RCES patients with different etiologies was controversial. Baryla et al. [18] pointed out that a high proportion of dystrophic cases resulted in an increased recurrence rates, while Pogorelov et al. [19] observed no recurrences in map-dot-fingerprint corneal dystrophy patients. Jain et al. [23] found no difference in the recurrence rate of RCE retreatment after PTK caused by two causes. Our subgroup analysis showed that patients with RCES secondary to corneal dystrophy may have a higher recurrence rate than patients with RCES due to trauma (corneal dystrophy recurrence rate = 22% [95% CI, 11%–32%], trauma recurrence rate = 17% [95% CI, 9%–24%]). Dedes et al. [14] studied the impact of etiology on postoperative PTK recurrence rate in RCES patients and found that map-dot-fingerprint corneal dystrophy patients were more likely to relapse than traumatic ones (odds ratio = 5.48; 95% CI, 1.27–61.32; p = 0.028), which may be because corneal dystrophy was a diffuse corneal pathology involving abnormalities in basement membrane and epithelial cell conversion with hemispherical body adhesion defects throughout the entire epithelium, whereas traumatic corneal erosions had more limited pathologic area, intact hemidesmosomes bodies, and basement membrane-bowman-complex [14]. However, these concepts require further confirmation.

For patients with recurrent corneal erosion, management of an acute episode varied depending on the severity of the recurrence, the presence of an epithelial defect, the size and location of the defect, and the type of previous management. Many patients first choose conservative therapy as first-line management, but the recurrences are common. Reidy et al. [2] showed that conservative management was effective in the short term in about 50% of patients, while long-term surgery was required for recurrent erosion in another 50% of patients. It may be that conservative treatment regimen helped temporarily relieve symptoms but did not alter the underlying disease process. Since RCES were the result of abnormal adhesion of the corneal epithelium to the epithelial basement membrane, surgical interventions can remove the unstable epithelium and promote the formation of a stable epithelium with strong adhesion complexes. PTK was an excimer laser-based surgical procedure widely performed by corneal surgeons for treating anterior corneal stromal pathologies, whose inherent ability to ablate corneal tissue with extremely high precision and minimal adjacent tissue damage, and it had been a useful tool in the treatment of recurrent corneal erosions [31]. Since it can provide a greater stability and stronger epithelial anchorage, and its recurrence rate ranges from 25–50% [31], it was generally superior to other treatment options, such as ASP, alcohol delamination of epithelium, and superficial keratectomy. The optimal timing for selecting PTK as a treatment option for patients with RCES remains uncertain. However, based on a comprehensive literature review, PTK has been found to be suitable for patients who have not responded to conventional treatments such as topical medications, bandage contact lenses, ED, and even superficial stromal puncture. Therefore, we recommend considering PTK if patients have undergone conventional treatments but continued to experience more than three episodes of recurrent epithelial erosion.

The standard PTK procedure involves manual debridement of the loose epithelium, followed by ablation of the central superficial corneal surface using a large spot size, typically around 6.5 mm. Varying techniques are used to expose the peripheral cornea outside the central zone. With the VISX laser, a smaller spot is manually scanned around the periphery by moving the patient’s head. Some smaller spot-scanning lasers can be programmed so the scan will include the periphery as part of areas treated. The amount of tissue ablated to maximize the effect with minimal refractive change has not been elucidated. While there may be slight variations in techniques among different centers, the primary objective of this approach is to remove a 6.0 μm thick anterior stromal layer from Bowman’s layer [32]. The removal of abnormal epithelium leads to formation of new basement membrane and regeneration of basal epithelial cells. In addition, the ablation of Bowman’s membrane facilitates direct contact between the epithelial cells and stromal keratocytes and stimulates the formation of new hemidesmosomes and anchoring fibrils [29].

The most common complication after PTK was the occurrence of subepithelial haze. In this meta-analysis, the incidence of haze was 13% (95% CI, 6%–21%), but most of them was completely reversible during the follow-up. Chan et al. [15] reported the highest incidence of as high as 50%; they noticed that postoperative subepithelial haze was seen in 8 eyes in the PTK group at 1 month after surgery and disappeared completely in all eyes at 12 months. Suri et al. [16] reported the incidence was 29%, the mild haze appeared in 2 eyes after PTK, but the haze was not visually significant and was not treated with steroids. At last, vision improved, and haze essentially resolved in all eyes. Obrart et al. [29] found that only the slightest subepithelial haze appeared in 5 eyes during the first 3 months, which improved rapidly, and there was no evidence of persistent corneal opacity disturbances associated with photoablative 6 months after surgery. Cases with delayed epithelial healing and deeper ablation had an increased risk of postoperative haze formation [33]. Unlike PRK [27, 28], haze was not visible or was only occasionally observed after PTK, which may be attributed to the limited ablation depths used in the treatment of RCES patients. Although PTK-induced haze may occur after surgery, it has a very low probability of affecting vision, as most ocular symptoms were completely reversible during follow-up.

All types of refractive errors may occur after PTK, but the greatest risk was induced hyperopia (incidence was 20% [95% CI, 11%–28%]), which may be caused by a central flattening of the cornea and unacceptable to patients who were not myopic preoperatively. Although most studies showed a trend toward hyperopia, the change in spherical equivalent was not statistically significant. Rashad et al. [22] found that at 1 year after PTK, the spherical equivalent manifest refraction of all eyes was +0.15 ± 0.39 D compared with that before surgery, whose difference was not statistically significant, and the induced hyperopia changes in 10 eyes ranged +0.37 to +1.00 D. Amm et al. [34] argued that the curvature of the cornea and the increase in the inclination of incident radiation at the edge of the beam may result in irregular and lower ablation rate at the periphery of the treated area than at the center, thus forming an ablation profile similar to that of myopia correction. The degree of hyperopic shift induced was related to the depth of ablation performed. In the treatment of RCE, only 6-μm depth ablation was attempted. This series confirmed that this superficial ablation has no significant dioptric effect and does not cause eccentric treatment of irregular astigmatism. It can be used either on or off the visual axis [35]. Dogru et al. [36] analyzed 112 eyes of 80 patients who underwent PTK for superficial corneal opacities to investigate factors affecting refractive changes. Their findings revealed that eyes treated with a transitional zone setting, shallow ablation depth, and the utilization of a masking agent exhibited a significantly smaller hyperopic shift. The ablation diameter depends on the size and location of the lesion. If recurrent corneal erosion is attributed to corneal injury and consistently occurs in a specific position on the corneal surface, then treatment with ED and localized PTK can be performed. However, in most clinical cases, recurrent erosions are either not observed during examination or recurrences are noted in several locations within the same cornea. In such cases, it is preferable to treat the entire corneal surface [35]. ED can be performed across the entire corneal surface, extending up to 1 mm within the limbus. Subsequently, an approximately 7 mm diameter excimer laser is applied centered over the entrance pupil to avoid inducing clinically significant central irregular astigmatism. The remaining exposed stroma is treated with a smaller excimer laser. It is believed that utilizing a transition zone setting would prevent a sharp junction between the ablated and non-ablated corneal zones and promote reepithelialization, and the key point is to avoid placing the edges of the ablation laser spot within the pupil area to prevent postoperative corneal irregular astigmatism. It is important to note that the ablation diameter, depth, and transition zone size for PTK can vary among patients and should be customized for each individual and the specific excimer laser system being used.

The incidence of postoperative vision loss was 11% (95% CI, 5%–16%), and most patients showed only line 1 deterioration on the Snellen test. Nassaralla et al. [17] found that there was no significant difference in BSCVA before and after PTK, and only 2 cases had one line of vision loss in each eye after operation. Stewart et al. [21] reported that the patients had a recurrence of bilateral corneal dystrophy, resulting in a one-line drop of the best corrected vision of both eyes. The results of this study indicated that PTK was a safe and effective procedure for the treatment of recurrent corneal erosions.

At present, there was no meta-analysis of PTK treatment of recurrent corneal erosions, so the conclusions of this meta-analysis had some significance and value. However, there were still certain limitations. In this study, most of the included studies were non-controlled, and many of them were retrospective and single-center. Therefore, this meta-analysis was limited by selection bias and the heterogeneity that arises from the variability in surgical techniques, different excimer lasers, and varying inclusion criteria. The level of evidence may be limited due to the lack of a control group, the limited follow-up time, inadequate documentation, and grading of adverse events. With the fact that not all safety and efficacy indicators were available from every study, our evaluation was limited. Despite our efforts to exclude overlapping patient populations, the possibility of overlapping patients in the studies remains. Another important limitation of this meta-analysis was the lack of sufficient information in the included literature regarding specific medication usage during the treatment process. While it was widely acknowledged that the use of adjunct medications during surgical interventions can potentially influence the outcome measures, the limited details provided in the literature precluded us from conducting a comprehensive analysis of the therapeutic effects of these medications. The inability to assess the specific impact of adjunct medications on the outcomes of interest was a notable limitation of this study. Future research should stress more attention on providing comprehensive and detailed information on medication usage to allow for a more thorough analysis of their therapeutic effects in surgical comparisons.

This meta-analysis of 21 studies, including 642 patients (705 eyes) with recurrent corneal erosions treated with PTK, demonstrated the safety and efficacy of PTK in the management of RCES, especially in patients who did not respond to conservative medical treatment. PTK was more successful in treating trauma than treating corneal dystrophy. Further studies and randomized clinical trials with longitudinal follow-up directly comparing the PTK with other established techniques were required.

The authors thank all the colleagues in Tianjin Medical University Eye Hospital who once offered the valuable courses and advice during this study.

An ethics statement is not applicable because this study is based exclusively on published literature.

The authors have no conflicts of interest to declare.

This work is supported by the Tianjin Key Medical Discipline (Specialty) Construction Project (TJYXZDXK-037A).

Sijing Chen and Xiaoran Chu contributed equally to this work and should be regarded as co-first authors. Sijing Chen and Xiaoran Chu extracted data from published studies and wrote the main manuscript text. Chen Zhang, Zhe Jia, and Liu Yang provided oversight for the extraction. Ruibo Yang and Yue Huang analyzed and identified the accuracy of the results and discussions presented in the manuscript. Shaozhen Zhao designed the study and was responsible for review and revision of the manuscript.

All data are within the paper and fully available without restriction. Further inquiries can be directed to the corresponding author.

1.
Lin
SR
,
Aldave
AJ
,
Chodosh
J
.
Recurrent corneal erosion syndrome
.
Br J Ophthalmol
.
2019 Sep
103
9
1204
8
.
2.
Reidy
JJ
,
Paulus
MP
,
Gona
S
.
Recurrent erosions of the cornea: epidemiology and treatment
.
Cornea
.
2000 Nov
19
6
767
71
.
3.
Heyworth
P
,
Morlet
N
,
Rayner
S
,
Hykin
P
,
Dart
J
.
Natural history of recurrent erosion syndrome: a 4 year review of 117 patients
.
Br J Ophthalmol
.
1998 Jan
82
1
26
8
.
4.
Rapuano
CJ
.
Phototherapeutic keratectomy: who are the best candidates and how do you treat them
.
Curr Opin Ophthalmol
.
2010 Jul
21
4
280
2
.
5.
Sridhar
MS
,
Rapuano
CJ
,
Cosar
CB
,
Cohen
EJ
,
Laibson
PR
.
Phototherapeutic keratectomy versus diamond burr polishing of Bowman's membrane in the treatment of recurrent corneal erosions associated with anterior basement membrane dystrophy
.
Ophthalmology
.
2002 Apr
109
4
674
9
.
6.
Aitken
DA
,
Beirouty
ZA
,
Lee
WR
.
Ultrastructural study of the corneal epithelium in the recurrent erosion syndrome
.
Br J Ophthalmol
.
1995 Mar
79
3
282
9
.
7.
Wu
WC
,
Stark
WJ
,
Green
WR
.
Corneal wound healing after 193-nm excimer laser keratectomy
.
Arch Ophthalmol
.
1991 Oct
109
10
1426
32
.
8.
Marshall
J
,
Trokel
SL
,
Rothery
S
,
Krueger
RR
.
Long-term healing of the central cornea after photorefractive keratectomy using an excimer laser
.
Ophthalmology
.
1988 Oct
95
10
1411
21
.
9.
Dausch
D
,
Landesz
M
,
Klein
R
,
Schröder
E
.
Phototherapeutic keratectomy in recurrent corneal epithelial erosion
.
J Refract Surg
.
1993 Nov-Dec
9
6
419
24
.
10.
Watson
SL
,
Leung
V
.
Interventions for recurrent corneal erosions
.
Cochrane Database Syst Rev
.
2018 Jul 9
2018
7
Cd001861
.
11.
Yang
Y
,
Mimouni
M
,
Trinh
T
,
Sorkin
N
,
Cohen
E
,
Santaella
G
.
Phototherapeutic keratectomy versus epithelial debridement combined with anterior stromal puncture or diamond burr for recurrent corneal erosions
.
Can J Ophthalmol
.
2023
;
58
(
3
):
198
203
.
12.
Song
MY
,
Chung
JL
,
Kim
KY
,
Hwang
KY
,
Kwon
YA
,
Song
SW
.
Combined phototherapeutic keratectomy and peripheral anterior stromal puncture for the treatment of recurrent corneal erosion syndrome
.
Korean J Ophthalmol
.
2020 Aug
34
4
297
303
.
13.
Mehlan
J
,
Steinberg
J
,
Traber
L
,
Katz
T
,
Linke
SJ
.
Recurrence rate and subjective symptoms after standardized (Hamburg protocol) phototherapeutic keratectomy on recurrent corneal erosions
.
Graefes Arch Clin Exp Ophthalmol
.
2016 Oct
254
10
2005
9
.
14.
Dedes
W
,
Faes
L
,
Schipper
I
,
Bachmann
LM
,
Thiel
MA
.
Phototherapeutic Keratectomy (PTK) for treatment of recurrent corneal erosion: correlation between etiology and prognosis–prospective longitudinal study
.
Graefes Arch Clin Exp Ophthalmol
.
2015 Oct
253
10
1745
9
.
15.
Chan
E
,
Jhanji
V
,
Constantinou
M
,
Amiel
H
,
Snibson
GR
,
Vajpayee
RB
.
A randomised controlled trial of alcohol delamination and phototherapeutic keratectomy for the treatment of recurrent corneal erosion syndrome
.
Br J Ophthalmol
.
2014 Feb
98
2
166
71
.
16.
Suri
K
,
Kosker
M
,
Duman
F
,
Rapuano
CJ
,
Nagra
PK
,
Hammersmith
KM
.
Demographic patterns and treatment outcomes of patients with recurrent corneal erosions related to trauma and epithelial and bowman layer disorders
.
Am J Ophthalmol
.
2013 Dec
156
6
1082
7
.e2.
17.
Nassaralla
BR
,
Nassaralla Junior
JJ
.
Ten-year results of phototherapeutic keratectomy on recurrent corneal erosions
.
Arq Bras Oftalmol
.
2012 Jan-Feb
75
1
33
7
.
18.
Baryla
J
,
Pan
YI
,
Hodge
WG
.
Long-term efficacy of phototherapeutic keratectomy on recurrent corneal erosion syndrome
.
Cornea
.
2006 Dec
25
10
1150
2
.
19.
Pogorelov
P
,
Langenbucher
A
,
Kruse
F
,
Seitz
B
.
Long-term results of phototherapeutic keratectomy for corneal map-dot-fingerprint dystrophy (Cogan-Guerry)
.
Cornea
.
2006 Aug
25
7
774
7
.
20.
Chow
AM
,
Yiu
EP
,
Hui
MK
,
Ho
CK
.
Shallow ablations in phototherapeutic keratectomy: long-term follow-up
.
J Cataract Refract Surg
.
2005 Nov
31
11
2133
6
.
21.
Stewart
OG
,
Pararajasegaram
P
,
Cazabon
J
,
Morrell
AJ
.
Visual and symptomatic outcome of excimer phototherapeutic keratectomy (PTK) for corneal dystrophies
.
Eye
.
2002 Mar
16
2
126
31
.
22.
Rashad
KM
,
Hussein
HA
,
El-Samadouny
MA
,
El-Baha
S
,
Farouk
H
.
Phototherapeutic keratectomy in patients with recurrent corneal epithelial erosions
.
J Refract Surg
.
2001 Sep-Oct
17
5
511
8
.
23.
Jain
S
,
Austin
DJ
.
Phototherapeutic keratectomy for treatment of recurrent corneal erosion
.
J Cataract Refract Surg
.
1999 Dec
25
12
1610
4
.
24.
Cavanaugh
TB
,
Lind
DM
,
Cutarelli
PE
,
Mack
RJ
,
Durrie
DS
,
Hassanein
KM
.
Phototherapeutic keratectomy for recurrent erosion syndrome in anterior basement membrane dystrophy
.
Ophthalmology
.
1999 May
106
5
971
6
.
25.
Morad
Y
,
Haviv
D
,
Zadok
D
,
Krakowsky
D
,
Hefetz
L
,
Nemet
P
.
Excimer laser phototherapeutic keratectomy for recurrent corneal erosion
.
J Cataract Refract Surg
.
1998 Apr
24
4
451
5
.
26.
Förster
W
,
Atzler
U
,
Ratkay
I
,
Busse
H
.
Therapeutic use of the 193-nm excimer laser in corneal pathologies
.
Graefes Arch Clin Exp Ophthalmol
.
1997 May
235
5
296
305
.
27.
Lohmann
CP
,
Sachs
H
,
Marshall
J
,
Gabel
VP
.
Excimer laser phototherapeutic keratectomy for recurrent erosions: a clinical study
.
Ophthalmic Surg Lasers
.
1996 Sep
27
9
768
72
.
28.
Bernauer
W
,
De Cock
R
,
Dart
JK
.
Phototherapeutic keratectomy in recurrent corneal erosions refractory to other forms of treatment
.
Eye
.
1996
10
Pt 5
561
4
.
29.
O’Brart
DP
,
Muir
MG
,
Marshall
J
.
Phototherapeutic keratectomy for recurrent corneal erosions
.
Eye
.
1994
8
Pt 4
378
83
.
30.
Fagerholm
P
,
Fitzsimmons
TD
,
Orndahl
M
,
Ohman
L
,
Tengroth
B
.
Phototherapeutic keratectomy: long-term results in 166 eyes
.
J Refract Surg
.
1993 Mar-Apr
9
2
S76
81
.
31.
Nagpal
R
,
Maharana
PK
,
Roop
P
,
Murthy
SI
,
Rapuano
CJ
,
Titiyal
JS
.
Phototherapeutic keratectomy
.
Surv Ophthalmol
.
2020 Jan-Feb
65
1
79
108
.
32.
Deshmukh
R
,
Reddy
JC
,
Rapuano
CJ
,
Vaddavalli
PK
.
Phototherapeutic keratectomy: indications, methods and decision making
.
Indian J Ophthalmol
.
2020 Dec
68
12
2856
66
.
33.
Rathi
VM
,
Vyas
SP
,
Sangwan
VS
.
Phototherapeutic keratectomy
.
Indian J Ophthalmol
.
2012 Jan-Feb
60
1
5
14
.
34.
Amm
M
,
Duncker
GI
.
Refractive changes after phototherapeutic keratectomy
.
J Cataract Refract Surg
.
1997 Jul-Aug
23
6
839
44
.
35.
Wilson
SE
,
Marino
GK
,
Medeiros
CS
,
Santhiago
MR
.
Phototherapeutic keratectomy: science and art
.
J Refract Surg
.
2017 Mar 1
33
3
203
10
.
36.
Dogru
M
,
Katakami
C
,
Yamanaka
A
.
Refractive changes after excimer laser phototherapeutic keratectomy
.
J Cataract Refract Surg
.
2001 May
27
5
686
92
.