Myopic regression after photorefractive keratectomy.

Background And Objective: Myopic photorefractive keratectomy (PRK) using the 193 nm excimer laser is an effective and precise surgical procedure to correct myopia, but not without complications, such as corneal haze, over- or undercorrection, halo, glare, and myopic regression. Among these, myopic regression, which can reduce uncorrected visual acuity, is the most common and progressive long-term complication.

Patients And Methods: We evaluated 228 eyes that received successful myopic PRK with the Summit excimer laser (ExciMed UV 200LA, 5.0 mm optical zone) and followed up more than three years at Kangnam St. Mary's hospital. Two hundred twenty eight eyes were divided into 3 groups according to the degree of preoperative myopia and correction: Group I, 79 eyes of moderate myopia (-2.25 to-6.75 diopters [D]) with full correction; Group II, 110 eyes of high myopia (-7.00 to -12.50 D) with full correction by two-zone ablation; Group III, 29 eyes of high myopia (-7.00 to -12.00 D) with partial correction of 6.00 D. Retrospective analysis of the data was performed with statistical evaluation of the incidence of myopic regression, regression formula, coefficient of parameters, and cumulative hazard of regression in each group.

Results: An incidence of myopic regression greater than 1.00 D was noted in 40.5%, 83.3%, and 18.2% of patients in Group I, II, and III, respectively. The regression formula obtained from the repeated measure ANOVA was Y = 2.13 -0.47 InX, Y = 3.02 -0.82 InX, and Y = 3.69 -0.50 InX, in Group I, II, and III, respectively. The amount of correction was a statistically significant correlation factor to the myopic regression (Pearson's correlation coefficient = 0.78, P < 0.001). The amount of preoperative myopia may act as a negative correlation factor to the myopic regression. According to the Kaplan-Meyer survival table, the cumulative correlation hazard was significantly higher in high myopia with full correction than in moderate myopia with full correction.

Conclusion: Our results indicated that the incidence and amount of myopic regression have a tendency to be dependent on the amount of correction, but may be reduced as the preoperative myopia increases, if the amount of correction is the same. A longer term and more detailed study of PRK using a 6.0 mm optical zone will be needed to determine how the amount of preoperative myopia and would profiles affect the myopic regression.