Retreatment using LASIK, PRK, or ICL for myopic regression after corneal laser refractive surgery Authors: Robert Edward T Ang, Miguel Roberto C Ang, James Michael D Jacomina, Michelle Marie Q Araneta, Emerson M Cruz Published: 2026 ABSTRACT AbstractPurposeTo describe the outcomes of retreatment for myopic regression following primary LASIK or PRK, and to explore factors associated with pre-retreatment refractive error.MethodsRetrospective cross-sectional study of eyes undergoing retreatment with LASIK, PRK, or ICL implantation after prior LASIK/PRK. Outcomes included UDVA, CDVA, MRSE, and safety and efficacy indices through 6–12 months. Associations between baseline factors and pre-retreatment MRSE were assessed.Results158 eyes were included (136 LASIK, 12 PRK, 10 ICL). Early postoperative outcomes improved significantly for LASIK (UDVA 0.37 to 0.04 logMAR; MRSE − 1.28 D to − 0.12 D) and ICL (UDVA 0.68 to 0.03 logMAR; MRSE − 2.72 D to 0.03 D). PRK also produced significant UDVA gains (0.63 to 0.06 logMAR). Efficacy and safety indices were near or above 1.0 across groups. A longer interval since primary surgery was significantly linked to more negative MRSE, consistent with greater regression. Mild correlation between pre-retreatment MRSE and age, initial MRSE, and tissue ablation were observed. Complications were low (7.9%).ConclusionsRetreatment for myopic regression with LASIK, PRK, or ICL was safe and effective, with stable outcomes through 6–12 months. Regression was multifactorial and not predicted by baseline characteristics, though longer intervals between surgeries were significantly associated with greater myopic regression.Keywords: Retreatment, Regression, Myopia, LASIK, PRK, ICL Introduction Laser Assisted In-situ Keratomileusis (LASIK) and Photorefractive Keratectomy (PRK) used to correct myopia have the longest history and track record in laser refractive surgery. According to Taneri et al., in long-term studies of up to 10 years, most cases of LASIK and PRK have favorable long-term outcomes [1]. However, residual refractive error remains the most common complication in corneal laser refractive surgery and can be due to overcorrection, undercorrection, or regression. Overcorrection means the refractive correction was overdone. A myopic treatment that overshoots the plano target results in a hyperopic outcome, or a hyperopic treatment resulting in a myopic outcome, with both scenarios causing unsatisfactory visual outcomes. In contrast, undercorrection means the refractive correction was incomplete; a myopic treatment that resulted in a significant residual myopia or a hyperopic treatment that resulted in a significant residual hyperopia, similarly causing dissatisfaction. Both overcorrection and undercorrection are diagnosed in the early post-operative period, approximately within one month of treatment, when healing and other factors have not yet affected the residual refraction. Our definition of regression was similar to those mentioned in studies by Lim et al. and Shalchi et al.,. where regression is the gradual return of some amount of the original refractive error after experiencing a period of full correction and good vision after refractive treatment [2, 3]. Several studies, specifically those done by Taneri, Chayet, Lim, Moshirfar, Mohammadi, Pokroy, and Xu et al., all noted the general idea of corneal self-repair and healing after surgery as the main pathophysiology behind regression [1, 2, 4–8] One common factor cited in several studies, such as those done by Vorotnikova, Yushan, and Ramin et al., included the extent of preoperative myopia, where higher levels increased the risk for regression [9–11]. Increased corneal thickness with central corneal steepening [4], preoperative axial length [12], and increased hours of near work [10] have also been mentioned as possible contributors to regression. Progressive myopia in younger patients [2], age-related and environmental changes [13], hormonal changes such as in pregnancy [14], and systemic changes like diabetes may also play a role [6]. Regression may occur as a side effect of any refractive procedure, with no predilection towards any type of surgery. No studies have been done comparing the regression rate between LASIK, PRK, and ICL. It is difficult to determine the actual number of patients who experience regression for several reasons. First, it depends on whether the patient will return to their original site of surgery and be formally diagnosed. Second, patients who experience blurred vision can decide to pursue different courses of action, like eyeglasses, contact lenses, retreatment surgery, or no intervention and accept the blurred vision. A surrogate method of reporting regression is by counting retreatment procedures. Retreatment is typically done by performing the primary procedure as the primary surgery. If the residual corneal stromal bed thickness is adequate, re-lifting the flap is usually performed by lifting the previous flap [6, 15]. PRK retreatment involves removing the epithelium again and applying the laser treatment [9]. In the literature, cross-over treatments such as PRK retreatment after primary LASIK [16], LASIK retreatment after primary PRK [17], and ICL after primary LASIK [18] have been described. An ICL can be implanted in an eye with previous LASIK or PRK if the anterior chamber depth and endothelial cell count qualify for the procedure. ICL exchange can be performed by removing the previous ICL and replacing it with a new one with the updated refractive correction. Proper screening is done before the retreatment so that the ideal surgical option can be identified, with follow-up thereafter to monitor the outcome and safety of the additional surgery. The Singapore National Eye Center performed an audit of its myopic LASIK outcomes. Yuen et al. reported a retreatment rate of 3.8% at 10 years on 37,932 eyes [19]. At 18 years, another audit in the same institution by Chua et al. showed a LASIK retreatment rate of 2.55% with a complication rate of 0.98% with 53,731 eyes [20]. The interval between initial surgery and regression varied. Studies done by Kamiya and Hickenbotham et al. noted regression as early as 6 months after initial refractive surgery and even as late as 22 years after [21, 22]. No risk factors have been identified affecting the interval time between the initial surgery and the regression. A study done by Spadea et al. noted retreatment for residual myopia in PRK has been reported in approximately 3.8% to 20.8% of eyes, with increasing retreatment frequency associated with higher degrees of myopia, with an incidence of 6.8% [23]. Even ICL-implanted eyes experienced regression. In a population of 177 eyes with ICL, Kamiya et al. reported 2 eyes (1.1%) needed retreatment with PRK for regression [21]. This study presents a descriptive analysis of clinical outcomes in eyes with myopic regression that underwent retreatment with LASIK, PRK, or ICL implantation. Patient characteristics and refractive data were examined to explore factors associated with regression and retreatment outcomes, without implying causal or predictive relationships. Methodology This was a retrospective cross-sectional study on patients who had undergone retreatment via LASIK, PRK, or ICL for myopic regression after primary LASIK or PRK. This study followed the tenets of the Declaration of Helsinki and good clinical practice and was approved by the Cardinal Santos Medical Center (CSMC) Research Ethics Review Committee (2024-047). The requirement for informed consent was waived due to the retrospective nature of the study. Medical records of 158 eyes that underwent retreatment with LASIK, PRK, or ICL were identified and retrieved. Inclusion criteria consisted of patients who experienced myopic regression, underwent retreatment within the past ten years (January 1, 2013 to December 31, 2023), and had at least one month of follow-up after retreatment. Exclusion criteria included patients who underwent presbyopic LASIK, those with any complication from the primary surgery, and those with retinal pathology, glaucoma with visual field defects, visually significant cataract as defined by the Lens Opacities Classification System version III (LOCS III), or ocular surface diseases such as corneal scars or ectasia. Perimetry was not routinely included in the standard refractive screening protocol and was performed selectively when suspected glaucoma was present based on history, optic nerve evaluation, intraocular pressure, or other clinical findings. Patients with glaucoma and associated visual field defects were excluded from the study. Patients underwent the following tests to determine eligibility for a refractive procedure: manifest refraction, visual acuity tests, Orbscan topography (Bausch and Lomb, USA), Pentacam tomography (Oculus, Germany), iTrace aberrometry (Tracey Technologies, USA), Anterion Optical Coherence Tomography (OCT) (Heidelberg Engineering, Germany) specular microscopy, cycloplegic refraction, Schirmer’s test, slit lamp examination, intraocular pressure measurement and dilated retina exam. In this study, although Pentacam provides more detailed corneal tomography, Orbscan was consistently used throughout the study because it serves as the reference tomographic platform integrated with the excimer laser ablation planning system in our institution. The use of a single imaging modality ensured consistency of measurements and minimized inter-device variability during follow-up. Patients requiring enhancement underwent retreatment using the same surgical modality as the primary procedure. Specifically, PRK was repeated in eyes initially treated with PRK, and LASIK retreatment was performed only in eyes initially treated with LASIK. No cross-over enhancement procedures were performed. The estimated ablation depth of the residual refractive error was calculated. Flap thickness and residual stromal bed values were estimated from the documented intended surgical parameters recorded during the primary refractive procedure and were not directly measured postoperatively with anterior segment OCT or other imaging modalities. LASIK retreatment was allowed if there were no signs of ectasia and the estimated residual stromal bed thickness after ablation would be higher than 280 microns. If the eye was not qualified for LASIK, patients were recommended to undergo PRK retreatment or ICL implantation. PRK retreatment was allowed if there was no corneal scarring and if the total estimated corneal thickness with epithelium after ablation was 360 microns. The option of ICL implantation for retreatment was discussed if the anterior chamber depth was at least 2.80 mm and the endothelial cell count was deemed adequate for age. Surgical procedure LASIK and PRK retreatment were performed under topical anesthesia using 0.5% proparacaine hydrochloride (Alcaine, Alcon Laboratories, Inc.). ICL retreatment was done under the same topical anesthesia plus intravenous sedation. LASIK retreatment was performed by redissecting and lifting the previous corneal flap, ablating the stromal bed with the Teneo 317 Model 2 excimer laser (Bausch & Lomb, Rochester, NY, USA), irrigating the interface, and repositioning the flap. PRK retreatment was performed by trephining the cornea, applying alcohol, mechanically removing the epithelium, and ablating the cornea with the Teneo 317 Model 2 excimer laser (Bausch & Lomb, Rochester, NY, USA), followed by topical mitomycin-C application for 30 s, after which the ocular surface was irrigated with 30 cc of cold balanced salt solution (BSS), and bandage contact lens placement. Postoperative treatment was standardized for procedures, including primary PRK and enhancement cases such as PRK following PRK or LASIK following LASIK. All patients received Prednisolone acetate 1% (Pred Forte, Allergan, Inc.), Gatifloxacin 0.3% (Zymar, Allergan, Inc.), and Ketorolac tromethamine 0.5% (Acular, Allergan, Inc.) four times daily for two weeks. Although PRK enhancement procedures may clinically warrant individualized or prolonged postoperative therapy due to differences in epithelial healing, a uniform regimen was used in this study to ensure consistency across treatment groups. For ICL implantation, a paracentesis incision was made, and lidocaine and methylcellulose viscoelastic were injected into the anterior chamber. A 3.2 mm knife was used to create the main corneal incision. A V4c model ICL (STAAR Surgical Phakic ICL, California, USA) was implanted into the sulcus. Viscoelastic was manually flushed out using BSS. The intraocular pressure was monitored for two hours before discharging the patient. Patients were instructed to use Gatifloxacin 0.3% (Zymar, Allergan, Inc.), Prednisolone acetate 1% (Pred Forte, Allergan, Inc.), and Ketorolac tromethamine 0.5% (Acular, Allergan, Inc.) four times a day until consumed. Data assessment Collected data included the type of initial surgery, type of retreatment, interval between procedures, uncorrected distance visual acuity (UDVA) and corrected distance visual acuity (CDVA) in logMAR, and manifest refraction spherical equivalent (MRSE). All patients underwent slit-lamp examinations, corneal topography using Orbscan II (Bausch & Lomb, Munich, Germany), and undilated and dilated wavefront aberrometry measurements using Zywave II (Bausch & Lomb, Munich, Germany). Lens vault was assessed with Anterior Segment OCT (Visante, Zeiss, Germany). Postoperative UDVA, CDVA, MRSE, and any complications were noted. Data up to one-year post-retreatment were collected. To avoid inter-eye correlation, only one eye per patient was included in the analysis. In cases where both eyes were eligible, a single eye was selected based on a predefined criterion (e.g., the right eye / the eye with worse baseline visual acuity / the first recorded eye in the dataset), ensuring that each patient contributed only one observation to the statistical analysis. Statistical analysis Following the study of Yuen et al. (2010), the overall retreatment rate after the initial surgery is 3.80%. Using a 95% confidence interval and 5% margin of error, the required sample size for the study is 57 [19]. Paired comparisons between preoperative and postoperative measurements were performed using paired t-tests for normally distributed variables and Wilcoxon signed-rank tests for non-normally distributed variables. For outcomes measured at multiple follow-up time points, repeated-measures analysis of variance (ANOVA) was used when assumptions of normality and sphericity were satisfied; otherwise, the Friedman test was applied. Categorical variables were summarized as frequencies and percentages and compared using Pearson’s chi-squared test or Fisher’s exact test as appropriate. Continuous variables between independent groups were compared using one-way ANOVA or the Kruskal–Wallis H test depending on distributional assumptions. A significance level of 5% (p < 0.05) was used for all analyses. Statistical analyses were conducted using STATA version 16.1 and Microsoft Excel. Results Table 1 presents baseline characteristics and demographics of patients stratified by retreatment type (LASIK, PRK, or phakic intraocular lens [ICL]). The mean age at the time of retreatment ranged from 36 to 39 years. The mean interval from the primary procedure to retreatment was 70.1 months for LASIK, 46.8 months for PRK, and 158.5 months for ICL. No cases were identified in which primary ICL implantation was followed by subsequent retreatment. Among the cohort, 129 of 158 eyes had their initial procedure performed at our institution, allowing analysis of available preoperative data prior to the initial LASIK or PRK. Table 1Baseline characteristics of eyes prior to retreatment with LASIK, PRK, or ICLVariablesN= 158 Age (mean ± standard deviation) LASIK 36.6 ± 9.2 (n =136) PRK 39.3 ± 8.8 (n = 12) ICL 36.6 ± 6.8 (n = 10) Gender LASIK (%) Male 41.2% (n = 56) Female 58.8% (n = 80)PRK (%) Male 50% (n= 6) Female 50% (n= 6)ICL (%) Male 20% (n= 2) Female 80% (n= 8) Interval from Primary Treatment to Retreatment (months ± standard deviation) LASIK70.1 ± 62.62PRK46.8 ± 49.8ICL158.5 ± 48.2 Initial Surgery Venue Within our institution81.6% (n = 129)Done outside18.4% (n = 29) Baseline characteristics of eyes prior to retreatment with LASIK, PRK, or ICL At the time of the initial procedure, the mean age was 31.2 ± 9.4 years in eyes later undergoing LASIK retreatment, 35.9 ± 11.6 years in the PRK retreatment group, and 24 ± 5.7 years in the ICL group. Most eyes that underwent LASIK or PRK as the primary procedure received the same modality at retreatment. In the initial LASIK group, mean preoperative MRSE was − 3.98 ± 2.09 D, pachymetry was 560.72 ± 34.20 μm, and ablation depth was 90.79 ± 29.98 μm. In the initial PRK group, corresponding values were − 4.30 ± 1.82 D, 540.00 ± 0.0 μm, and 95.70 ± 30.46 μm. Table 2 includes only preoperative data from patients undergoing excimer laser refractive procedures (LASIK and PRK). Implantable Collamer Lens (ICL) cases were not included because ICL surgery represents a fundamentally different refractive correction approach involving intraocular lens implantation rather than corneal ablation, making direct comparison less methodologically appropriate. Table 2Preoperative measurements prior to initial LASIK & PRK (within institution)LogMAR UDVAInitial LASIK(n = 116)Initial PRK(n = 11)p-value1.31 ± 0.61 (20/400)1.56 ± 0.56 (20/400)0.210LogMAR CDVA0.02 ± 0.05 (20/20)0.00 ± 0.00 (20/20)0.165Sphere (D)-3.60 ± 2.10-3.78 ± 2.190.730Cylinder (D)-0.94 ± 0.83-0.95 ± 1.350.980MRSE (D)-3.98 ± 2.09-4.30 ± 1.820.507Pachymetry (µm)560.72 ± 34.20540.00 ± 0.00.620Max ablation (µm)90.79 ± 29.9895.70 ± 30.460.617 Preoperative measurements prior to initial LASIK & PRK (within institution) Within each retreatment group, improvements in visual and refractive measures were observed (Table 3). In the LASIK retreatment group, mean UDVA changed from logMAR 0.37 to 0.04, and MRSE from − 1.28 D to − 0.12 D, with values remaining stable over follow-up. In the PRK retreatment group, UDVA changed from 0.63 to 0.06 and MRSE from − 0.32 D to − 0.14 D. In the ICL group, UDVA reached 0.03 and MRSE changed from − 2.72 D to − 0.03 D, with a mild myopic shift noted at later follow-up. Residual myopia was reduced in all groups. Table 3Visual and refractive outcomes of before and after LASIK retreatmentVariablesPre-operativeEarly post-operative(1-6 months)p-values(Pre vs. Early)Late post-operative(6-12 months)P-values(Early vs. Late) LASIK UDVA 0.37 ± 0.29(20/50)0.04 ± 0.11(20/20)< 0.01**0.04 ± 0.15(20/20)0.513 CDVA 0.01 ± 0.06(20/20)0.01 ± 0.07(20/20)0.6570.01 ± 0.07(20/20)0.057 Sphere -1.18 ± 0.86-0.03 ± 0.41< 0.01**-0.05 ± 0.470.678 Cylinder -0.42 ± 0.35-0.27 ± 0.23< 0.01**-0.29 ± 0.260.695 MRSE -1.28 ± 0.93-0.12 ± 0.33< 0.01**-0.19 ± 0.460.726 PRK UDVA 0.63 ± 0.24(20/80)0.06 ± 0.05(20/20)< 0.01**-0.04 ± 0.05(20/20)0.099 CDVA 0.00 ± 0.00(20/20)0.01 ± 0.04(20/20)0.3560.00 ± 0.00(20/20)0.990 Sphere -1.00 ± 1.68-0.25 ± 0.290.073-0.25 ± 0.250.990 Cylinder -0.20 ± 0.27-0.30 ± 0.330.7170.00 ± 0.000.252 MRSE -0.32 ± 2.07-0.14 ± 0.330.5730.23 ± 0.220.666 ICL UDVA 0.68 ± 0.21(20/100)0.02 ± 0.07(20/20)< 0.01**0.03 ± 0.06(20/20)0.678 CDVA 0.00 ± 0.00(20/20)0.00 ± 0.00(20/20)0.9900.00 ± 0.00(20/20)0.990 Sphere -2.43 ± 1.04-0.15 ± 0.54< 0.01**-0.22 ± 0.280.057 Cylinder -0.40 ± 0.34-0.25 ± 0.260.297-0.32 ± 0.310.279 MRSE -2.72 ± 1.07-0.03 ± 0.56< 0.01**-0.48 ± 0.39< 0.05***significant at the 1% level Visual and refractive outcomes of before and after LASIK retreatment 0.37 ± 0.29 (20/50) 0.04 ± 0.11 (20/20) 0.04 ± 0.15 (20/20) 0.01 ± 0.06 (20/20) 0.01 ± 0.07 (20/20) 0.01 ± 0.07 (20/20) 0.63 ± 0.24 (20/80) 0.06 ± 0.05 (20/20) -0.04 ± 0.05 (20/20) 0.00 ± 0.00 (20/20) 0.01 ± 0.04 (20/20) 0.00 ± 0.00 (20/20) 0.68 ± 0.21 (20/100) 0.02 ± 0.07 (20/20) 0.03 ± 0.06 (20/20) 0.00 ± 0.00 (20/20) 0.00 ± 0.00 (20/20) 0.00 ± 0.00 (20/20) **significant at the 1% level Efficacy and safety indices are presented descriptively for each modality. The efficacy index was 0.970 ± 0.176 in the LASIK group, 1.022 ± 0.228 in the PRK group, and 1.000 ± 0.000 in the ICL group. The safety index was 1.023 ± 0.080 for LASIK, and 1.000 ± 0.000 for both PRK and ICL. The overall complication rate was 7.89%, including eight cases of epithelial ingrowth (EI) and one case of haze. The haze occurred following LASIK retreatment and resolved without intervention. None of the reported complications were associated with loss of lines of CDVA. In cases of epithelial ingrowth, flap lift and stromal cleaning were performed, with no documented recurrence. The interval between the primary LASIK procedure and flap relifting for enhancement ranged from 3 to 180 months. Only one case occurred after a short interval of 3 months, whereas the remaining cases occurred after longer intervals (82, 83, 95, 95, and 180 months), indicating a predominance of epithelial ingrowth events following extended postoperative periods in this small cohort. Although previous reports [35] suggest that longer intervals may be associated with increased EI incidence after flap relifting, the limited sample size in our study precludes any definitive conclusion regarding this temporal relationship. Table 4 summarizes linear regression analyses examining associations between baseline variables and MRSE values. For initial MRSE, age at initial surgery (β = 0.014, 95% CI: −0.033 to 0.062, p = 0.546) and sex (β = −0.771, 95% CI: −1.596 to 0.054, p = 0.067) were not statistically associated with MRSE. For pre-retreatment MRSE, age (β = −0.016, 95% CI: −0.032 to 0.001, p = 0.062) and sex (β = −0.138, 95% CI: −0.509 to 0.232, p = 0.460) were also not statistically associated. The interval between surgeries (β = −0.008, 95% CI: −0.010 to − 0.005, p < 0.01) demonstrated a statistically significant association with pre-retreatment MRSE, indicating an inverse relationship in this sample. Table 4Linear regression of preoperative and pre-retreatment MRSE predictorsDependent VariableIndependent VariableCoefficient95% CIp-value1Initial MRSEAge before initial surgery0.014-0.033 – 0.0620.546Gender-0.771-1.596 – 0.0540.067Pre-retreatment MRSEAge-0.016-0.032 – 0.0010.062Gender-0.138-0.509 – 0.2320.460Interval between surgeries-0.008-0.010 – -0.005< 0.01**1 ** - significant at the 1% level; * - significant at the 5% level; CI = Confidence Interval Linear regression of preoperative and pre-retreatment MRSE predictors 1 ** - significant at the 1% level; * - significant at the 5% level; CI = Confidence Interval Figure 1 illustrates the relationship between age at initial surgery and pre-retreatment MRSE, showing a weak positive trend (R² = 0.0471) without statistical significance. Fig. 2 depicts the relationship between initial MRSE and pre-retreatment MRSE, also demonstrating a weak positive trend (R² = 0.0541). Fig.3 shows the relationship between percentage of tissue ablation and pre-retreatment MRSE, with a similarly weak positive trend (R² = 0.0458). These observed relationships did not reach statistical significance. Correlation between age and Pre-Retreatment MRSE Correlation between Initial MRSE and Pre-Retreatment MRSE Correlation between Tissue Ablation and Pre-retreatment MRSE Discussion In our practice, adult patients of different age groups come for refractive surgery. They undergo a comprehensive screening test after which a treatment plan is recommended. Based on the surgeon’s evaluation, patients are qualified for LASIK if their estimated residual stromal bed thickness is considered sufficient and the corneal curvature evaluation is within normal limits. Patients disqualified from LASIK and PRK due to thin corneas, low residual stromal bed thickness, or corneal topographic findings suggestive of ectasia, but who have adequate anterior chamber depth measured from the corneal endothelium, will be considered suitable candidates for ICL implantation as an alternative refractive correction option. Most patients undergoing refractive surgery aspire or even expect that they will be spectacle-free for the rest of their lives. However, over time, uncorrected far vision can change and become blurry. Regression is the gradual return of some amount of the original refractive error after experiencing a period of full correction and good vision after refractive treatment [2, 3]. The exact cause or mechanism is unknown, but factors related to the patient’s demographics, physiology, or treatment parameters have been explored. In a study by Chua et al., they noted a mean age of 32.6 ± 7.3 years for those undergoing LASIK retreatment [20]. Pokroy et al. reported a mean age of 25.9 ± 7.3 years, which was a younger-aged demographic undergoing PRK retreatment [7]. In a study by Chen et al., the average age of those undergoing ICL retreatment was 37.54 ± 7.59 years [18]. In our study, the mean age at which patients came for retreatment was 36.6 ± 9.2 years for the LASIK group, 39.3 ± 8.8 years for the PRK group, and 36.6 ± 6.8 years for the ICL group. There was no study comparing the average age when regression occurred between these three retreatment groups. Our study shows that for patients who came for retreatment, the mean age during their initial LASIK procedure was 31.2 ± 9.4 years and 35.9 ± 11.6 years during their initial PRK. Our trend analysis showed that the older the age at initial treatment, the higher the amount of refractive error. We observed that patients who came for primary LASIK or PRK in our institution tended to be in the older age group. This may not be reflective of the age of the patient population in other institutions. The total percent ablation is reflective of the magnitude of refractive error during the initial surgery. A non-significant, weak positive trend was observed between higher initial refractive error and greater refractive change at retreatment; however, this did not reach statistical significance and should be interpreted as exploratory. Our study calculated the interval between initial and retreatment surgery and determined if that was a factor in the magnitude of regression. We observed that PRK had the shortest interval between retreatment at 46.8 ± 49.8 months, followed by LASIK 70.1 ± 62.62 months, and lastly ICL 158.5 ± 48.2 months. Patients with longer intervals between surgeries demonstrated more myopic pre-retreatment MRSE; however, this relationship was not statistically significant and may reflect time-related changes or other unmeasured confounders. The observation is exploratory in nature. Overall, most observed correlations were weak and did not reach statistical significance; therefore, these findings should be interpreted as exploratory patterns rather than evidence of causal or predictive relationships. In our study, more women experienced regression and underwent retreatment as compared to men. Out of 136 eyes undergoing LASIK retreatment, 56 (41.2%) of those were male patients, and 80 (58.8%) were women. This is consistent with the study by Xu et al., where they postulated that increased near work done by women was a significant factor in the occurrence of regression [8]. Mohammadi et al. also noted an association between female patients and risk factors for regression with an odds ratio of 1.03 [5]. Another study by Hashemi et al. reported that females had higher keratometry values compared to men, which could be due to hormonal factors, although the exact cause is not well known [24]. Regression can be due to many different factors, although none are yet proven to be the predominant one. One of the many factors being considered is the preoperative axial length and/or refractive error. In a study by Torii et al., 67 eyes underwent myopic LASIK and were compared with 66 eyes undergoing phakic intraocular lens implantation. Their study noted that there was a significant difference in myopic regression postoperatively between groups matched for preoperative axial length [12]. Lim et al. identified higher preoperative myopia and deeper ablation as risk factors for long-term regression [2]. Pokroy et al. likewise found that greater baseline refractive error strongly predicted the need for PRK retreatment [7]. In our series, a similar trend was observed: higher pre-retreatment MRSE and greater tissue ablation were both associated with regression. The observed association between a longer interval between surgeries and greater magnitude of refractive regression is biologically plausible and may, in part, reflect age-related lenticular changes such as gradual increases in nuclear sclerosis that can induce a myopic shift over time. However, in this cohort, chronological age was not significantly associated with pre-retreatment MRSE and demonstrated only a weak, non-significant trend, suggesting that lenticular ageing alone is unlikely to be the primary driver of this relationship. More plausibly, the interval effect reflects time-dependent postoperative biomechanical or epithelial remodeling, or cumulative regression phenomena following refractive surgery. Taken together, these findings support a multifactorial mechanism, with lenticular ageing potentially playing a minor but non-dominant role. Non-anatomical factors have also been linked to regression. Yushan et al. reported that eyes with regression had worse preoperative UDVA and were associated with greater near work exposure, parental myopia, time outdoors, education, and physical activity. In their series of 323 patients, near work exceeding 8 h per day was significantly correlated with regression in 52 eyes, while time outdoors was not. Sleep duration was noted to be more relevant in women [10]. Xu et al. reported similar findings, noting that near work and shorter sleep duration were significant predictors of long-term regression after SMILE and LASIK [8]. Regression can also arise from secondary compensatory mechanisms and morphological remodeling. Following refractive surgery, epithelial hyperplasia may develop over the flattened corneal zone, which can increase stromal thickness in an already thinned cornea, contributing to regression. Reinstein et al. used a very high-frequency (VHF) ultrasound to evaluate the central cornea and demonstrated an average increase of approximately 2.13 ± 0.44, 7.09 ± 0.85, and 8.78 ± 0.60 μm in epithelial thickness 6 months after LASIK for low, moderate, and high levels of myopia, respectively. They also noted a direct relationship between the degree of attempted correction and the amount of epithelial thickening observed, with higher corrections resulting in more pronounced postoperative epithelial hyperplasia [25]. Our study showed a similar trend with a higher percentage of tissue ablation weakly correlated to the degree of regression refraction. Multiple studies by Gartry et al., Haw et al., Lee et al., Lyle et al. and Spadea et al. showed improved visual acuities and refractive outcomes and concluded that LASIK and PRK retreatments were safe and effective [16, 23, 26–28]. Our results showed that LASIK retreatment resulted in a mean UDVA of logMAR 0.04 ± 0.11 with a mean MRSE of -0.12 ± 0.33. PRK retreatment resulted in a mean UDVA of logMAR 0.06 ± 0.05 with MRSE of 0.14 ± 0.33D. For those we deemed risky to undergo further laser refractive surgery or patients who made specific requests, we offered an ICL as an alternative retreatment as long as they qualified based on anterior chamber depth and endothelial cell count. Kamiya et al. studied the outcomes of ICL implantation after an 8-year period and found that 78% of them had a mean UDVA of logMAR 0 [21]. Chung et al., likewise, saw that 83% of eyes had a UDVA of logMAR 0 or better at 3 months after surgery. In their study, the UDVA of − 0.03 ± 0.11 and MRSE of − 0.06 ± 0.33 were similar to our findings of 0.03 ± 0.07 UDVA and − 0.03 ± 0.56 MRSE, respectively [29]. No other studies were found using ICL as a form of retreatment for regression. Apart from visual and refractive outcomes, another method of reporting good outcomes is by calculating for safety and efficacy. Our study noted that retreatment with LASIK, PRK, and ICL were all safe because the safety indices were 1.023 ± 0.080, 1.000 ± 0.000, and 1.000 ± 0.000, respectively. Our efficacy index for retreatment with LASIK was 0.970 ± 0.176, for PRK it was 1.022 ± 0.228, and for ICL 1.000 ± 0.000. Chung et al. reported the efficacy index and safety index of ICL retreatment after refractive surgery at 1.11 ± 0.22 and 1.13 ± 0.20, respectively [29]. Other surgeries report on primary treatment. Ide et al. monitored the safety profile in LASIK over a 10-year follow-up period and noted a safety index of 0.99 [30]. In a PRK study over a 20-year follow-up period by Cennamo et al., the safety index was 1.00 [31], and for ICL over a 3-year follow-up period reported by Fernandez-Vega-Cueto, the safety index was 1.03 at 36 months postoperatively [32]. Our retreatment outcomes in terms of safety and efficacy are comparable to those of primary laser refractive surgery. Aside from visual outcomes from visual outcomes, other safety indices that should be reported are complications of retreatments. In performing LASIK retreatment, our surgical technique involves lifting the previously-created LASIK flap, then ablating the stroma. The flaps were returned and allowed to heal. We noted eight cases of epithelial ingrowth after LASIK retreatment. The flaps were re-lifted and the stroma cleaned. No recurrence was noted. A study was done by Ting et al. investigating prevalence, risk factors, and management of epithelial ingrowth after LASIK treatment. They noted that the prevalence of ingrowth after primary LASIK was very low, around 0 to 3.9% of eyes. This risk was significantly increased in situations such as when the flap was lifted for retreatment, with a risk of 10–20% compared to our study of 7.89% [33]. This is due to the inadvertent introduction of corneal epithelial cells into the intraocular environment. Risk factors noted for epithelial ingrowth included surgical instruments used, surgical technique, where a direct flap lift was preferred over cleaving with a spatula. Another risk factor was increased age and hyperopic LASIK. Another study done by Letko et al. noted that among 272 retreatments, 132 eyes had flaps created by a microkeratome and 140 eyes had flaps created by a femtosecond laser. Among the microkeratome group, 11 eyes noted ingrowth, and surgical intervention was needed in 8 of those eyes. The femtosecond group noted only 2 eyes with ingrowth and neither needed any surgical intervention. The study concluded that the use of a femtosecond laser led to less ingrowth formation and that the rate of epithelial ingrowth could be related to the geometry of the flap edge [34]. Ortega-Usobiaga noted that surgical cleansing of epithelial ingrowth was safe and effective. Their study had 2772 eyes undergo retreatment via lifting of the flap and doing further myopic ablation. Among those eyes, 704 were noted to have epithelial ingrowth, where 634 eyes did not need surgical clean-up, and 70 did. In both these groups, no eyes had worsened VA afterward. Safety index was 1.00 ± 0.12 in the group not needing further clean-up, and 0.97 ± 0.15 in the group where clean-up was necessary. Efficacy index was 0.93 ± 0.20 and 0.86 ± 0.22 in each group, respectively. Epithelial ingrowth, whether in need of further treatment or not, did not decrease the safety or efficacy of the LASIK retreatment groups [35]. We acknowledge several limitations in our paper. This study only included patients who decided to seek consultation and retreatment regarding their regression. The total number of patients who experienced regression is not known since other patients may opt not to come back for follow-up. They may seek another remedy such as eyeglasses. Another limitation is that the patients who were qualified for more than one type of retreatment surgery were not randomized. We explained the options, but most patients chose the same treatment as their initial surgery. Other factors that may have influenced their choice of retreatment were affordability, recommendation by friends, or personal research. Follow-up time was also not standardized among patients. Patients came irregularly based on their convenience, and some did not have extensive follow-up periods. We therefore could not monitor for further regression. The causes, long-term incidence, and severity of regression have not been clearly established, so future longitudinal, multicenter studies with larger sample sizes will be needed to establish more definitive trends. Differences in sample size between the three groups may likewise have contributed to the variability in mean outcomes. The variability in follow-up duration (6 months to 1 year) among patients may introduce bias in intergroup comparisons and limits the ability to draw uniform long-term conclusions. Lastly, this study did not include multivariable regression analyses to adjust for potential confounding variables. Therefore, the observed associations should be interpreted as exploratory and descriptive rather than causal. In addition, although group comparisons and univariate analyses were performed, the absence of adjusted models limits inference regarding independent predictors of outcomes. Future studies with larger sample sizes should incorporate multivariable and longitudinal modeling approaches to strengthen causal interpretation and account for repeated measurements more robustly. Prospective studies with adequately powered regression analyses are required to determine whether any of these observed trends represent true predictive factors for refractive regression. Conclusion Retreatment for myopic regression after corneal refractive surgery appears to be safe and effective across LASIK, PRK, and ICL, with generally favourable refractive outcomes over 6–12 months of follow-up in appropriately selected patients. LASIK and PRK demonstrated overall refractive stability during this period, whereas a mild late myopic shift was observed in the ICL group. A longer interval since the primary surgery was significantly associated with greater myopic regression. There was a weak correlation between pre-retreatment MRSE and initial surgery age, initial MRSE, and percent tissue ablation. Overall, these findings suggest that regression is multifactorial and cannot be reliably predicted from baseline characteristics alone. Acknowledgements None. Author contributions Design and conduct of the study (RTA); collection (MCA), management (RTA), analysis (RTA, MCA, JDJ, MQA, EMC), interpretation of the data (RTA, MCA, JDJ, MQA, EMC); manuscript preparation (MCA, JDJ), manuscript review (RTA, MCA, JDJ, MQA, EMC), manuscript approval (RTA, EMC). Funding RTA (none), MCA (none), JDJ (none), MQA (none), EMC (none). Data availability Available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate The Cardinal Santos Medical Center Ethics Committee approved this retrospective, Helsinki-compliant study (2024-047) and waived informed consent. Consent for publication Not applicable. Consultant / advisory positions RTA: Staar Surgical, Acevision, Acufocus, Bausch and Lomb, Johnson & Johnson, Physiol BVI. MCA, JDJ, MQA, EMC: none Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement Available from the corresponding author on reasonable request.