Understanding Long-Term Regression

Long-term regression after corneal refractive surgery remains a significant challenge, with many patients experiencing a gradual return of refractive error that impacts visual durability and may require further surgery. Identifying its many causes and key risk factors is essential to improve patient selection, surgical planning, and long-term refractive stability, said Sara Sella MD.

Dr Sella noted long-term refractive regression arises from complex biological responses involving the corneal epithelium and stromal biomechanics. After myopic LASIK or SMILE, the central epithelium typically thickens to compensate for stromal tissue loss, while it thins centrally and thickens peripherally in hyperopic PRK to smooth the transition zone. Biomechanical changes occur simultaneously, as anterior stromal ablation reduces corneal rigidity, potentially causing posterior stromal steepening that alters refractive power. Cytokine signalling drives keratocyte transformation and stromal remodelling, contributing further to regression and haze formation.

In a study of 236 eyes from 119 civilian pilots in southwest China (mean preoperative SE -2.92 ± 1.11 D), myopic regression was observed in 23.6% of eyes over a 5–16-year period after corneal refractive surgery. Regression risk increased with older age at surgery (OR 1.151 per year, p = 0.022), subepithelial ablation techniques (PRK/LASEK; OR 2.769, p < 0.001), and prolonged, continuous near work (OR 0.635, p = 0.038). Cumulative flying hours and length of postoperative follow-up were not significant predictors, reinforcing the superior long-term refractive stability of intrastromal methods in occupations with high visual demands.¹

A large-scale high myopia cohort (2,460 eyes) supported the pilot study. Yet, in this cohort, regression arrived swiftly and extensively, with rates up to 89% within the first year, suggesting the influence of baseline refractive error in accelerating biomechanical or healing-related shifts. A third study focused solely on modern intrastromal approaches (SMILE, FS‑LASIK), painting a more reassuring five‑year picture (16.1% regression) but reinforcing an environmental warning: prolonged near work can double the risk, particularly in younger patients, women, and FS‑LASIK eyes. Together, these studies reveal that while surgical method and patient age consistently shape long‑term stability, the context of baseline myopia, early healing response, and daily visual habits after surgery can profoundly alter the trajectory of refractive outcomes.^2,3

“The take-home message is, not all procedures are equal,” Dr Sella said.

Regarding hyperopic regression, a retrospective analysis of eyes from 888 patients who underwent hyperopic LASIK at a single centre over 14 years showed older age and earlier-generation platforms as key predictors of retreatment. However, neither corneal thickness nor preoperative refraction appeared to influence regression.⁴

Astigmatic regression presents its own challenges. High‑cylinder cases are more prone to undercorrection or enhancement, with similar susceptibility in both LASIK and PRK. In a 20‑year analysis of 785 mixed‑astigmatism eyes, the overall retreatment rate was 5.2%, with most enhancements performed within 3 years. High preoperative astigmatism increased the likelihood of retreatment almost fourfold, and male gender, significant residual cylinder, and suboptimal platform choice further elevated risk. Although advances in ablation speed, cyclotorsion control, and centration have reduced residual or regressed astigmatism, they have not eliminated it.⁵

“When it comes to long-term regression, there is no universally superior procedure for every refractive error,” Dr Sella said. “For hyperopia, both LASIK and PRK are limited by significant regression in higher corrections, and while SMILE is promising, long-term data is lacking. For astigmatism, LASIK remains the most predictable for high cylinders, but PRK and SMILE are effective for lower corrections. For myopia, all three procedures are effective for low to moderate corrections, but high myopia remains a challenge, with similar regression rates across platforms.”

Treatment for regression should be based on the underlying cause. Diagnosis typically includes refraction, topography, epithelial mapping, and posterior elevation imaging. Myopic regression associated with posterior steepening may be managed with topography-guided PRK, hyperopic regression resulting from epithelial remodelling with PRK using mitomycin-C and an expanded optical zone, and astigmatic regression with vector-corrected wavefront- or topography-guided LASIK or PRK.

“Long-term refractive stability hinges on aligning procedure choice with patient biology, refractive target, and lifestyle factors,” she said.

Dr Sella and her colleagues developed a nomogram that aims to improve retreatment outcomes by adjusting preoperative planning to account for the epithelial contribution to regression. They hope to report their results soon. Meanwhile, a Korean team has reported promising results with machine learning models they developed to identify patients at high risk for refractive regression.⁶

Sara Sella MD is a cataract, refractive, and cornea surgery consultant at the Ein Tal Medical Center, Tel Aviv, Israel. eyesella66@gmail.com

1. Zhang Z, et al. BMC Ophthalmol, 2024; 24: 145.

2. Xu Y, et al. J Refract Surg, 2024 Jan; 40(1): e10–e19. doi:https://doi.org/10.3928/1081597X-20231212-04

3. Zhou J, et al. Int Ophthalmol, 2020 Jan; 40(1): 213–225.

4. Mimouni M, et al. J Refract Surg, 2018 May 1; 34(5): 316–320.

5. Sorkin N, et al. J Refract Surg, 2024 Feb; 40(2): e73–e78.

6. Kim J, et al. Graefes Arch Clin Exp Ophthalmol, 2022; 260(11): 3701–3710.