PURPOSE: To study the clinical, tomographic and densitometric features of eyes that showed >5 D of corneal flattening following collagen crosslinking (CXL) for progressive keratoconus and to identify preoperative predictive factors for such a response. METHODS: This was a retrospective case control study of 548 eyes with progressive keratoconus which had undergone epithelium-off CXL (Dresden protocol) with a follow-up of 1 to 10 years. Eyes that showed ⩾5 D corneal flattening in maximum keratometry (Kmax) following CXL (group A) were compared with one eye of the remaining patients in the same cohort (group B). Changes in refraction and visual acuity, Kmax and thinnest pachymetry were compared between both groups. Univariate and multivariate regression analysis identified preoperative risk factors for unusual corneal flattening. RESULTS: Forty three eyes in group A were compared with 502 eyes in group B. At the time of maximum flattening, group A showed more flattening (-7.6 ± 3.2 D) and thinning (-53.7 ± 45.2 µ) than group B (-1.69 ± 2.9 D and -26.6 ± 36.7 µ, respectively). Multivariate analysis based on parameters suggested by a univariate regression analysis identified pre-op Kmax to be the most significant predictor of intense corneal flattening. A subgroup analysis of K-matched eyes revealed that the duration of time following CXL was a significant risk factor for extreme corneal flattening following CXL. CONCLUSION: An intense corneal flattening >5 D in Kmax was documented in 7.85% of a cohort of keratoconus patients who underwent CXL. High preoperative Kmax and the duration of time following CXL were significant predictors of this response which was accompanied by significant corneal thinning.
PURPOSE: To study the clinical, tomographic and densitometric features of eyes that showed >5 D of corneal flattening following collagen crosslinking (CXL) for progressive keratoconus and to identify preoperative predictive factors for such a response. METHODS: This was a retrospective case control study of 548 eyes with progressive keratoconus which had undergone epithelium-off CXL (Dresden protocol) with a follow-up of 1 to 10 years. Eyes that showed ⩾5 D corneal flattening in maximum keratometry (Kmax) following CXL (group A) were compared with one eye of the remaining patients in the same cohort (group B). Changes in refraction and visual acuity, Kmax and thinnest pachymetry were compared between both groups. Univariate and multivariate regression analysis identified preoperative risk factors for unusual corneal flattening. RESULTS: Forty three eyes in group A were compared with 502 eyes in group B. At the time of maximum flattening, group A showed more flattening (-7.6 ± 3.2 D) and thinning (-53.7 ± 45.2 µ) than group B (-1.69 ± 2.9 D and -26.6 ± 36.7 µ, respectively). Multivariate analysis based on parameters suggested by a univariate regression analysis identified pre-op Kmax to be the most significant predictor of intense corneal flattening. A subgroup analysis of K-matched eyes revealed that the duration of time following CXL was a significant risk factor for extreme corneal flattening following CXL. CONCLUSION: An intense corneal flattening >5 D in Kmax was documented in 7.85% of a cohort of keratoconus patients who underwent CXL. High preoperative Kmax and the duration of time following CXL were significant predictors of this response which was accompanied by significant corneal thinning.