Study Groups and Protocol

Twenty eyes of 20 patients with documented progressive primary keratoconus (progression was defined as an increase of K _max of at least 1 diopter [D] within up to 1 year, measured by corneal tomography [Pentacam HR 70700; Oculus, Wetzlar, Germany]) were enrolled in this study. Excluded were eyes with previous ocular surgery, penetrating trauma, glaucoma, or aphakia; endothelial cell count less than 2300 cells/mm ² ; corneas with no distinctive posterior float area; corneas with pellucid marginal degeneration; corneal scars that may interfere with Scheimpflug photography; history of recurrent erosions; patients younger than 18 years; pregnancy and breastfeeding; nonavailability for follow-up examinations during 1 year; neurodermatitis; and connective tissue diseases. The study protocol was approved by the institutional review board of the Institut für Refraktive und Ophthalmo-Chirurgie (IROC). Of the 20 patients, 19 completed the 1-year follow-up (dropout rate: 5%), all patients completed the 6-month follow-up. After completion of the study group, standard epi-off CXL was applied to a control group that was matched to the study group regarding group size, inclusion criteria, age, and sex. Also in this group, only 1 eye per patient was included.

The patients were examined preoperatively, early postoperatively (daily until epithelium healing), and at 1 month, 6 months, and 12 months after customized CXL or standard CXL. At every follow-up, except early postoperative, a standard examination was performed consisting of autorefractometry and autokeratometry (Humphrey Model 599; Zeiss, Jena, Germany), corneal topography (Keratograph C; Oculus), Scheimpflug imaging (Pentacam HR 70700; Oculus), optical coherence tomography (OCT) (SS-1000; Tomey, Nagoya, Japan), manifest refraction using the fogging technique, uncorrected (UDVA) and best spectacle-corrected visual acuity (BSCVA), applanation tonometry, endothelial cell count (SP-02; Costruzione Strumenti Oftalmici, Scandicci Firenze, Italy) and slit-lamp examination. Patients were asked not to use contact lenses during the first postoperative month and 2 weeks before each examination.

Treatments

The control group underwent a standard CXL procedure including a circular epithelium debridement with a diameter of 9.0 mm (Epi clear; Orca Surgical, Caesarea, Israel) and application of 0.1% riboflavin in 16% dextran T500 for 30 minutes. Ultrasound pachymetry at the thinnest point was repeated every 10 minutes (SP-100; Tomey). If pachymetry fell below 400 μm the imbibition with riboflavin was continued using hypo-osmolar riboflavin 0.1% drops until the thickness was 400 μm. The cornea was irradiated with 9 mW/cm ² for 10 minutes (UVX 2000; IROC Innocross, Zurich, Switzerland).

The customized CXL group underwent a similar procedure except using a smaller eccentric debridement area and a customized UV irradiation profile (Mosaic System; Avedro Inc, Waltham, Massachusetts, USA). The profiles were designed as a concentric superposition of 3 circular areas ( Figure 1 ) centered on the maximum of the posterior float (reference: sphere, 8 mm approximation). The inner circle’s diameter was equal to the shortest diameter of the posterior float minus 0.5 mm (range, 1.9–2.9 mm); the outer circle’s diameter was 1 mm bigger than the maximal diameter of the posterior float (range, 5.2–6.5 mm). The middle circle’s diameter was the mean of the other 2 diameters. The radiance exposure in the inner circle was 10 J/cm ² , in the middle 7.5 J/cm ² , and in the outer 5.4 J/cm ² . The fluence was 9 mW/cm ² . Small eye movements were compensated by the eye tracker of the Mosaic System.

Typical evolution of a patient treated with customized corneal cross-linking with preoperative axial curvature (top left), axial curvature at the 12-month follow-up (top middle), difference map in axial curvature (top right), preoperative posterior float with an 8 mm approximated sphere and the irradiation pattern (bottom left), demarcation line at the 1-month follow-up emphasized with arrows (bottom right). ΔK _max is −3.6 diopters (D); regularization index is 5.3 D.

For both groups, antibiotic ointment (Floxal; Bausch & Lomb, Zug, Switzerland) and a bandage contact lens were applied to the eye and patched until epithelial healing. Thereafter, topical treatment consisted of fluorometholone eye drops (FML; Allergan, Zurich, Switzerland) 3 times per day for 1 week and taper during weeks 2 and 3.

Evaluation

The following parameters were used for quantitative evaluation: BSCVA, K _max , maximal posterior float, central endothelial cell count, pachymetry of the thinnest point, and maximal depth of the demarcation line. Optical pachymetry using Scheimpflug tomography was discarded for follow-up evaluation because information from the rear surface may be erroneous in hazy corneas. A new parameter (regularization index; RI) was introduced to describe the regularization of the corneal surface within a 6 mm circle around the apex ( Figure 1 ): absolute amount of maximal flattening (usually in steep areas) plus absolute amount of maximal steepening (usually in flat areas). Comparisons of parameters between the 2 groups were performed using the Mann-Whitney U test. Comparisons of parameters at different times were performed by means of the Wilcoxon rank sum test. Correlations and their significance were tested using the Spearman rank correlation test. All calculations were performed with WinSTAT for Excel (R. Finch Software, 2015, Bad Krozingen, Germany). Statistical significance was accepted if P < .05.