This study illustrates the benefits of transepithelial PTK for eyes with RCES that have failed multiple conservative treatments. The benefits of transepithelial PTK illustrated in our study include high patient satisfaction (87%), a high proportion no longer requiring conservative treatment (80%), no statistically significant change in CDVA, low risk of cornea haze (grade 2 cornea haze, 1.4%), limited need for invasive retreatment (1%), and improvement or resolution of symptoms (96%). As, we believe PTK may be one of the most effective treatment options for such eyes. PTK requires a laser suite, surgical experience, and may be more expensive than other invasive treatments.

Our study found that 44% of eyes with treatment-resistant RCES had an underlying diagnosis of ABMD in comparison to the literature which reports a rate of 26–29% [1,2,3]. One explanation may be that the likelihood of developing treatment-resistant RCES is higher in those patients with an underlying diagnosis of ABMD. [8] The cohort of patients included in this study were exclusively treatment resistant and therefore represent a unique subset of patients with RCES.

Globally, 50–60% of patients with RCES require invasive intervention as they remain symptomatic despite conservative management [3, 9]. Several invasive treatments for RCES include anterior stromal micro-puncture, alcohol delamination of the epithelium, mechanical corneal epithelial debridement, and diamond burr polishing exist; however, no robust evidence suggests one invasive treatment is superior to another [4].

The potential advantages of anterior stromal micro-puncture in comparison to PTK include decreased cost and the ability to complete this procedure in an outpatient clinic [10]. Disadvantages include a higher percentage of patients requiring a second treatment (17–24%) in comparison to PTK (6%) [10]. Additionally, faint corneal scars are nearly always present secondary to treatment which are not usually visually significant. Epithelial map–guided anterior stromal micro-puncture may have an improved treatment safety and efficacy for RCES; however, large trials are needed to establish reliable outcome [11]. The evidence on stromal micro-puncture is limited which limits the ability to compare efficacy and safety between stromal micro-puncture and PTK. In our study, only 1% of eyes required a repeat surgical treatment.

In a randomized controlled trial of 17 eyes treated with alcohol delamination of the corneal epithelium (ADCE) and 16 eyes treated with PTK, there was complete or partial resolution of symptoms in 65% of eyes in the ADCE group and 63% of eyes in the PTK group [12]. Mean follow-up was 16.25 and 17.25 months, respectively. Recurrence of symptoms occurred in 5 eyes in the ADCE group and 6 eyes in the PTK group, with no long-term complications in either group. In another study of 26 eyes followed for a minimum of 3 years, 11.5% of eyes that underwent ADCE had recurrence [13]. This study was limited by the small sample size. Additionally, no eyes in this study had an underlying diagnosis of ABMD. In our study, 96% of patients achieved complete resolution or improvement of symptoms after PTK and 87% of patients would undergo PTK for treatment-resistant RCES again based on post-PTK telephonic interview.

There is evidence to suggest epithelial debridement and diamond burr polishing may result in resolution of RCES symptoms in up to 88–97% of cases [14, 15]; however, disadvantages of this technique include the risk of developing post-operative corneal haze (14–26%) [14,15,16]. In comparison, only 1.6% of eyes in our study developed grade 2 cornea haze, and none developed grade 3 or more.

In this study, six patients (1%) required RCES retreatment with PTK and 20% of patients required ongoing conservative treatment post-PTK. No significant change was noted between mean pre-operative and mean post-operative CDVA; however, nearly a third of patients (28%) experienced an improved CDVA. There is evidence to suggest that surface treatments have potential to improve vision by treating epithelial irregularity and associated irregular astigmatism, particularly in ABMD [15, 17]. Further research should be done to examine the astigmatism outcomes of patients who undergo PTK for RCES.

Unlike PTK, alcohol delamination of the corneal epithelium, manual epithelial debridement, and diamond burr polishing do not include ablation of the corneal stroma. For this reason, these invasive procedures have zero risk of inducing a hyperopic shift.

In myopic eyes, the hyperopic shift may be a desirable side effect. In emmetropic and hypermetropic eyes, this shift may result in blurring of UDVA and subsequent patient dissatisfaction. In this study, stromal ablation was limited to 10 microns, and treatment diameter was 9 mm, to minimize the risk of hyperopic shift. The excimer laser was operated in transepithelial mode, and a correction factor of + 0.50 D for 65 μm was used to achieve a neutral refractive outcome. This calibration factor was not used for myopic eyes.

Corneal haze is a known complication of PTK, ADCE, mechanical epithelial debridement, and diamond burr polishing. For this reason, eyes undergoing these treatments are commonly treated with a post-operative course of steroids and the use of a bandage contact lens to reduce the risk of corneal haze. It is important to note that use of steroid eye drops and a bandage contact lens are themselves known treatment options for RCES, which may affect treatment outcomes. In our study, only 1.6% of eyes in our study developed grade 2 cornea haze, and none developed grade 3 or more. This is lower than other studies reporting PTK treatment outcomes for RCES [18]. This difference may be due to the use of Mitomycin C 0.02% (MMC) which is used on all eyes for 30 s following corneal ablation to minimize the risk of cornea haze development. Due to concerns about MMC cytotoxic effects, there is large variability in its usage for small refractive errors and standard PTK treatment; however, there were no reported MMC complications in our study.

The strengths of this study include the large sample size, long follow-up, and additional third-party vetting of treatment-resistant RCES by MSP for publicly funded PTK. Limitations of this study include its retrospective nature, no data collection on uncorrected distance visual acuity (UDVA), only a sample (124 eyes) collected on change in CDVA, focus of the study on recurring RCES symptoms, no control group, post-operative refractive data, and non-full patient participation in post-PTK telephonic interviews. This data was not available as TE-PTK in our study was performed for the sole purpose of decreasing RCES symptoms; therefore, this data was not fully available to include in analysis.