Corneal cross-linking using the Sub400 protocol in keratoconus patients with corneal thickness less than 400 μm

To evaluate the 12-month clinical outcomes of first-generation Sub400 corneal cross-linking (CXL) in progressive keratoconus (KC) with intraoperative stromal thickness below 400 μm. This non-randomized interventional case series included 21 eyes from 17 patients with progressive KC and an intraoperative stromal thickness below 400 μm. After epithelial removal, UV-A irradiation at 3 mW/cm² was delivered for an individualized duration selected from the original Sub400 nomogram according to the post-soak stromal thickness. Preoperative and 12-month assessments included corrected distance visual acuity (CDVA), refraction, Scheimpflug tomography, anterior-segment optical coherence tomography, and endothelial cell density (ECD). Individual intraoperative pachymetry readings, exposure durations, and delivered fluences could not be recovered for verification of eye-specific dosing. At 12 months, 17 of 21 eyes (81%) met the prespecified nonprogression criterion (< 1.0 D increase in Kmax); the mean Kmax change was 0.36 ± 2.45 D (postoperative minus preoperative; P = 0.503). Spherical equivalent changed from − 6.30 ± 3.98 D to -4.43 ± 4.40 D (P = 0.009). Mean ECD decreased from 2596 ± 269 to 2568 ± 285 cells/mm² in treated eyes (P = 0.030) and from 2613 ± 259 to 2567 ± 261 cells/mm² in fellow eyes (P = 0.002), without clinical corneal decompensation. A demarcation line was detected in 17 eyes, with a mean depth of 215.18 ± 24.78 μm. Demarcation-line depth was not correlated with baseline thinnest corneal thickness (P = 0.991), ECD change (P = 0.974), or Kmax change (P = 0.544). Exploratory comparisons between the four eyes with continued progression and the 17 nonprogressing eyes identified no statistically significant baseline differences, but power was limited. In this small, uncontrolled cohort, treatment reported as first-generation Sub400 CXL was associated with tomographic nonprogression in 81% of eyes at one year and no observed clinical endothelial decompensation. Missing individual irradiation parameters preclude validation of nomogram execution or the proposed depth-control mechanism. Larger controlled studies with complete procedural documentation and longer follow-up are required to determine comparative efficacy and long-term endothelial safety. Patients with progressive keratoconus whose stromal thickness is below 400 μm at the time of UV-A irradiation are not eligible for the conventional Dresden cross-linking protocol, creating an important treatment gap. In this small, uncontrolled cohort, the Sub400 protocol was associated with tomographic nonprogression in 81% of treated eyes at one year. No clinical corneal decompensation or endothelial failure was observed; however, the small but statistically significant reduction in endothelial cell density requires cautious interpretation. The findings describe one-year outcomes following treatment reported as first-generation Sub400 CXL in selected thin corneas; missing individual irradiation parameters prevent verification of eye-specific nomogram adherence or validation of the proposed depth-control mechanism.