Repeat Versus Primary Photorefractive Keratectomy for Treatment of Myopia

Abstract

Although effective, a portion of photorefractive keratectomy (PRK) patients will suffer residual myopia or relapse to myopic regression. This retrospective, non-randomized, comparative study, aimed to compare the efficacy of primary PRK versus PRK performed as retreatment after previous surgery for myopia. Data regarding the right eye of 220 consecutive myopic patients undergoing repeat or primary PRK in 2013–2017 were extracted. Groups were matched for demographics and preoperative spherical equivalent, sphere, astigmatism, uncorrected and corrected distance visual acuity (UDVA and CDVA). Primary outcomes were an efficacy index (ratio between the postoperative UDVA and the preoperative CDVA), a safety index (ratio between the postoperative and the preoperative CDVA), postoperative UDVA and CDVA, and deviation from target refraction. Primary PRK showed significant superiority in logMAR UDVA (0.01 ± 0.05 versus 0.05 ± 0.10, p = 0.001), logMAR CDVA (0.01 ± 0.05 versus 0.04 ± 0.08, p = 0.01), efficacy index (1.00 ± 0.05 versus 0.97 ± 0.09, p = 0.003) and safety index (1.00 ± 0.06 versus 0.98 ± 0.08, p = 0.04) compared to repeat PRK, but had a significantly higher share of patients with postoperative spherical equivalent (74.5% versus 67.3%) and cylinder (74.5% versus 68.2%) in the range of ±0.5 D. To conclude, enhancement PRK leads to inferior efficacy and safety with greater deviation from target refraction. Adjusted nomograms for repeat PRK may be warranted.

1. Introduction

Refractive surgery using excimer laser has been proven to be an effective and safe technique to correct low to medium refractive errors [1]. Photorefractive keratectomy (PRK) was the first widely used refractive procedure utilizing excimer laser technology. Since then, many alternatives have been introduced [2]; however, PRK remains an excellent option for the correction of low to moderate myopia [3,4].

It is well known that a percentage of PRK-treated eyes will relapse to myopic regression of myopia higher than −1.0 D. It is estimated that between 10% and 20% of patients undergoing primary PRK will have residual myopia to a level they will seek to eliminate it [5,6,7,8]. Moreover, when examining the long term, it has recently been shown that while PRK is a safe refractive procedure, its efficacy may decrease after a 15-year period in some cases [9].

Several factors are known to influence the predictability of excimer laser surgery such as more severe initial refractive errors, astigmatism, older age, variations in wound healing, as well as the differences in ablation profiles and laser nomograms. Thus, at times, a revision treatment is necessary [10,11]. However, surgeon experience, innovations in laser technologies, and nomogram adjustments have led to improvements in visual and refractive outcomes, of which the most significant effect on the need for retreatment was found to be in the difference between experienced surgeons and ones with lower volume of surgeries and no significant differences were found in retreatment rates when comparing laser-assisted in situ keratomileusis (LASIK) with PRK [11,12]. Altogether, retreatment rates have been declining significantly throughout the years [12].

The results of enhancements following refractive surgery have been mostly studied on cases of repeat PRK after previous LASIK [13,14,15,16], in which the development of haze was the most frequent limitation [17,18]. However, surprisingly few reports can be found for repeat PRK due to under-correction and regression after primary PRK [1,19].

In this study, we compare the refractive outcomes of primary PRK to that of repeat PRK performed secondary to refractive surgery for myopia, in thought that such a comparison should reveal the true results of these treatments as it compares the same operational procedure performed in two distinct baseline scenarios a patient can present with. Thus, this also allows better setting of patient expectation and education as to what to expect relative to their previous experience.

2. Materials and Methods

This was a retrospective, non-randomized, comparative study. All data for the study were collected and analyzed in accordance with the policies and procedures of the Institutional Review Board of the Barzilai Medical Center and the tenets set forth in the declaration of Helsinki.

2.1. Study Participants

Medical records of patients who underwent primary PRK or those who had repeat PRK following earlier primary myopic PRK between 1 January 2013 and 31 December 2017 at the Care-Vision Laser Centers, Tel-Aviv, Israel, were screened. Included were the right eyes of patients with myopia of up to −12 D with a cylinder of up to 6 D. Only patients with a minimum follow-up time of 90 days post-surgery were included. Even if both eyes needed retreatment, only one eye was included. Exclusion were patients under 18 years of age, a non-stable refraction for in the 12 months prior to surgery, intra-ocular pressure (IOP) of more than 21 mmHg, the wear of contact lenses in a predefined period prior to surgery (2 weeks for rigid contact lenses and 4 days for soft contact lenses), a history of autoimmune disease or diabetes, or a history of ocular surgery (other than primary PRK in the repeat group). The study arms comprised the right eyes (OD) of the enrolled patients.

2.2. Data Collection

The following preoperative data were collected: age, gender, sphere, spherical equivalent, cylinder, uncorrected distance visual acuity (UDVA) and corrected distance visual acuity (CDVA). The following data were extracted in the final follow-up post-surgery: sphere, cylinder, spherical equivalent, UDVA and CDVA. The efficacy index was calculated as the ratio between postoperative UDVA and preoperative CDVA, while the safety index was calculated as the ratio between the postoperative and the preoperative CDVA [20].

The primary outcomes were postoperative UDVA and CDVA, deviation from target refraction, the efficacy index, and safety index. Secondary outcomes were postoperative cylinder, and spherical equivalent.

2.3. Surgical Technique

All patients underwent alcohol-assisted PRK performed by one of five highly experienced surgeons, using the WaveLight^® EX500 platform (Alcon Laboratories, Fort Worth, TX, USA) and each patient received a drop of topical antibiotic, dexamethasone 0.1%, and artificial tears in the immediate postoperative scenario as previously published by our team [12,21]. According to clinic protocol, patients were then routinely examined at 1 day; 1 week; 1, 3, and 6 months; and as necessary post-surgery. Patients were also encouraged to return for examination if vision deteriorated at any time after surgery. Retreatment surgery was offered free of charge. The same nomogram was used for all treatments.

2.4. Matching Primary and Repeat PRK Arms

The primary and repeat PRK groups were matched for preoperative age, gender, spherical equivalence, sphere, astigmatism, and preoperative UDVA and CDVA [21].

2.5. Statistical Analyses

Data were analyzed with the Minitab Software, version 17 (Minitab Inc., State College, PA, USA) and MedCalc version 12.7.1.0 (MedCalc Software, Mariakerke, Belgium). A student’s t-test was used for the analysis of continuous data, and a Kruskal–Wallis for non-parametric variables. For the analysis of categorical variables, a Chi-Square or Fisher’s exact tests were used when applicable. Standardized graphs of refractive outcomes were produced using the mEYEstro Software Version 1.8 [22]. In all analyses, a two-sided p value of less than 0.05 was considered statistically significant. All data are presented as means accompanied by their respective standard deviations unless stated otherwise.

3. Results

Overall, the study included 220 eyes of 220 patients. Mean age was 29.2 ± 4.2 years (range 19 to 40 years), of which 68.2% (150 individuals) were of male gender. The primary and repeat PRK groups comprised 110 eyes each and had a mean follow-up period of 113 ± 131 and 107 ± 120 days, respectively (p = 0.14). There were no significant differences between the groups in terms of age (28.9 ± 4.4 versus 29.5 ± 4.0 years, p = 0.28), spherical equivalent (−1.55 ± 0.60 versus −1.47 ± 0.58 D, p = 0.30), sphere (−1.28 ± 0.61 versus −1.22 ± 0.59 D, p = 0.41), cylinder (−0.54 ± 0.46 versus −0.51 ± 0.44 D, p = 0.58), logMAR UDVA (0.55 ± 0.37 versus 0.52 ± 0.37, p = 0.59) and logMAR BCVA (0.01 ± 0.04 versus 0.02 ± 0.04, p = 0.71), indicating adequate matching (

Table 1. A comparison of baseline parameters of the primary and repeat PRK groups.

Parameter	Primary PRK (n = 110)	Repeat PRK Before Enhancement Surgery (n = 110)	p-Value
Age (years)	28.9 ± 4.4	29.5 ± 4.0	0.28
Gender (%male)	70.9%	65.5%	0.39
Sphere (D)	−1.28 ± 0.61	−1.22 ± 0.59	0.41
Cylinder (D)	−0.54 ± 0.46	−0.51 ± 0.44	0.58
Spherical equivalent (D)	−1.55 ± 0.60	−1.47 ± 0.58	0.30
UDVA (logMAR)	0.55 ± 0.37	0.52 ± 0.37	0.59
CDVA (logMAR)	0.01 ± 0.04	0.02 ± 0.04	0.71

PRK: photorefractive keratectomy, D: diopter, UDVA: uncorrected distance visual acuity, CDVA: corrected distance visual acuity.

).

3.1. Repeat PRK Baseline

Prior to their primary PRK, the repeat PRK group originally had a sphere of 3.6 ± 2.0 D (range −9.5 to 0 D), cylinder of −0.8 ± 0.6 D (range −3 to 0 D) and underwent an ablation depth of 65.7 ± 26.1 microns (range 17 to 132 microns). In the repeat PRK group- time from primary to repeat PRK was 91.3 ± 32 months (range 12 to 171 months).

3.2. Visual Acuity Outcomes

Compared to the enhancement PRK, the primary PRK group showed a significantly lower mean logMAR UDVA (0.01 ± 0.05 versus 0.05 ± 0.10, p = 0.001), higher mean decimal UDVA (0.98 ± 0.11 versus 0.91 ± 0.17, p < 0.001), and significantly superior mean efficacy index (1.00 ± 0.0.05 versus 0.97 ± 0.09, p = 0.003) post-surgery. In addition, the primary PRK group showed a significantly lower mean logMAR CDVA (0.01 ± 0.05 versus 0.04 ± 0.08, p = 0.01) and significantly superior mean safety index (1.00 ± 0.06 versus 0.98 ± 0.08, p = 0.04), but not a higher mean decimal CDVA (0.99 ± 0.10 versus 0.97 ± 0.11, p = 0.23) (

Table 2. A comparison of outcomes of the primary and repeat PRK groups.

Parameter	Primary PRK (n = 110)	Repeat PRK Before Enhancement Surgery (n = 110)	p-Value
Sphere (D)	0.01 ± 0.50	−0.13 ± 1.16	0.30
Cylinder (D)	−0.48 ± 0.39	−0.59 ± 0.46	0.09
Spherical equivalent (D)	−0.23 ± 0.52	−0.42 ± 1.25	0.18
UCVA (logMAR)	0.01 ± 0.05	0.05 ± 0.10	0.001
BCVA (logMAR)	0.01 ± 0.04	0.02 ± 0.05	0.20
Efficacy index	1.02 ± 0.13	0.95 ± 0.19	0.005
Safety index	1.02 ± 0.12	1.01 ± 0.15	0.48

PRK: photorefractive keratectomy, D: diopter, UDVA: uncorrected distance visual acuity, CDVA: corrected distance visual acuity.

).

3.3. Refractive Outcomes

No significant differences in mean postoperative spherical equivalent (p = 0.93) or cylinder (p = 0.67) were found. However, the primary PRK group had a higher share of patients with postoperative spherical equivalent (74.5% versus 67.3%) and cylinder (74.5% versus 68.2%) in the range of ±0.5 D (Table 2). Standardized graphs are displayed in Figure 1.

4. Discussion

PRK is an excellent option for the correction of low to moderate myopia [3,4] and has recently been shown to be safe in the long term as well. However, the procedure’s efficacy is known to be unsatisfactory in 10–20% of cases [5,6,7,8], and to decrease with time, especially in high myopia. Surprisingly, very little has been published about the refractive predictability of enhancement PRK for under-correction or regression after a previous PRK. One study reported repeated PRK as safe and, in most cases, effective in treating regression and under-correction in eyes with low residual myopia after initial PRK [19]. Other studies have stressed the importance of a sufficient time interval between refractive procedures to minimize keratocytes activity for PRK retreatment [23], and that enhancements using topographically guided excimer-laser-customized photoablation methods resulted in satisfactory and stable visual outcomes, with good safety and efficacy after unsuccessful PRK and LASIK [1]. However, the latter included only 32 patients after PRK and compared only pre and postoperative results.

In the last twenty years, only four papers have studied retreatment with PRK after primary PRK. Two of these were in the sample size range of case series [16,24,25,26]. Of the other two, the first [24] reported a UDVA of at least 20/20 in around 70% and CDVA within 1 line of pre-treatment in essentially all eyes. The most recent study [26], included 188 eyes (141 patients) and had published a more substantial improvement with a UDVA of 20/20 or better in 75% and a UDVA of 20/40 or better in 99% of eyes after 12 months, with a residual refraction within 1 D of the intended in 98% of the eyes.

Our current study included a relatively large number of patients who underwent enhancement PRK after primary PRK. In terms of VA, both primary and enhancement PRK were found safe and effective. However, primary PRK group showed significantly better LogMAR UDVA, LogMAR CDVA, and efficacy and safety indexes post-surgery. While the results do not contradict the broad conclusion of previous studies regarding enhancement PRK’s proven efficacy, they do show a more objective picture as they are derived from means rather than proportions of patients. Nevertheless, the portion of patients reaching a certain goal is clinically relevant to the refractive clinic in terms of patient education and satisfaction. However, while this study’s results for astigmatic outcomes were more subtle, they derived from an analysis with a lower tolerance for Seq and Cylinder (±0.5D rather than 1D) and are thus more relevant to practical clinical practice from that point, as they assess a more patient satisfaction-associated outcome. As note, significance of visual outcomes remained superior in analysis of decimal visual acuities for UDVA, but not for CDVA, emphasizing the lesser sensitivity of decimal visual acuity to small changes compared to LogMAR visual acuity.

Patient satisfaction is of utmost importance in the field of refractive surgery and previous studies have shown that residual refractive errors and astigmatism are among the dominant factors determining it [27]. Hence, counseling patients as to the expected inferior improvement when undergoing repeat PRK is crucial, even more so considering that these are patients who have undergone such counseling and surgery before and may be wrongfully assuming the same probabilities for flawless refractive correction.

The differences in this study may be postulated to be due to epithelial remodeling, as previously treated myopic corneas may present an altered thickness profile at the mid-periphery and center, which may cause regression and induce overcorrection after the enhancement [28].

Moreover, astigmatic correction is postulated to be less in patients with thick epithelium [29]. Unfortunately, given the nature of this retrospective study, epithelial maps prior to and following surgery were not available for analysis.

Corneal thickness, which is reduced by the primary PRK and even more so after enhancement treatment, may play a role as well, as it may result in difficulty to pinpoint the required treatment. Moreover, retreatment surgery implies manipulation of what was originally more posterior stromal tissue, and tissue which was operated on and altered before, and may have different biophysical properties and thus react differently to PRK, as opposed to the previously untouched, and anterior part of the stroma treated in primary PRK.

Another theory for the differences between the treatments is that they could be rooted in the predisposing factors influencing the need for retreatment to begin with.

The following factors have been reported at times to significantly increase the need for refractive enhancement surgery: a higher degree of astigmatism, hyperopia, lower operating room temperature, contact lens use, a less experienced surgeon [12,30], and age and gender, which were controversial in reports with regard to their effect on the visual outcomes and retreatment rates [10,11,12,30,31,32,33]. However, the control of these has mostly been matched in our study as reported. Nevertheless, it is most certainly possible that a subgroup analysis of a combination of these in future studies may reveal more information as to the reason for enhancement PRK’s inferior outcomes.

As a sidenote, hyperopia was also associated with retreatment in previous studies [11,12,29], which may be the result of hyperopic refractive surgery’s increase in corneal optical aberrations as compared with myopic procedures and peripheral epithelial proliferation [34], and as for astigmatism- association to retreatment has shown contradicting results [10,11,30]. However, in this study only myopic patients were included.

This study has several limitations, of which the most obvious is its retrospective nature, which also affected our ability to fully match between groups (e.g., parameters such as operating room temperature were not documented) and to fully report possibly contributing factors, as we did not have epithelial maps and corneal wavefront analysis.

It should also be noted that final follow-up values were used as outcomes rather than change scores from a baseline. However, groups were matched successfully for significant indifferences preoperatively, and changes from baseline were given weight in the form of the calculated indices. Also, several surgeons took part in the study, which may create a bias originating in variation of surgical methods and experience. However, all performed an alcohol-assisted PRK method as described by a unified protocol. An important note is that this study included only patients undergoing PRK for the correction of myopic errors, and therefore, its conclusions do not apply to hyperopic patients.

5. Conclusions

To the best of our knowledge, this is one of the few studies directly comparing the refractive results of primary and enhancement PRK in this manner. While enhancement PRK demonstrates overall safety and efficacy and remains an effective option for retreatment when needed, it tends to result in inferior results to primary PRK, with a greater deviation from target refraction. Designing refined nomograms for repeat PRK may be warranted, as well as mitigating patient expectations accordingly prior to surgery.