Fig. 3: Measures of efficacy from baseline through 18-month visit.

(A) Slit lamp photos (B) Corneal neovascularization tracing and (C) Fluorescein staining for Participant 9, who met criteria for Complete Success at final endpoint (18 months after CALEC transplant). Participant had history of severe alkali burn, and prior to enrollment had ocular surface reconstruction with extensive conjunctivoplasty including resection of corneal pyogenic granuloma. At baseline, participant had stage III limbal stem cell deficiency with best corrected visual acuity of count fingers. From post-operative month 3 through postoperative month 15 participant met criteria for partial success and had waxing and waning small epithelial defects. At post-operative month 18, the participant had a stable and improved ocular surface with improved comfort, and met criteria for a Complete success. Final visual acuity (20/500) with hard contact lens was 20/80, limited by pre-existing corneal stromal scarring. D–G Participants 1, 3, 8, and 15 are marked with an asterisk (*) on all plots to indicate they had an epithelial defect at baseline. The 3-month, 12-month and 18-month visits were prespecified times of outcome assessment. D Efficacy outcome per participants from 3-month to 18-month visit. White color pertains to outcomes that could not be calculated due to missing efficacy component measurements due to missed/telemedicine visits. E Epithelial Defect Surface Area per Participant from Baseline through 18 months (F) Cornea Surface Staining Score per Participant from Baseline through 18 months (G) Neovascular Area per Participant from Baseline through 18 months. H Spearman Correlation among Component Measures of Efficacy (N = 14 participants, at 96 total visits). Spearman correlation coefficients calculated using measures of efficacy at multiple visits (Baseline, 3-month, 6-month, 9-month, 12-month, 15-month and 18-month). Repeated measures were accommodated by using a mixed model approach for p-values and clustered bootstrap methods for estimation of confidence intervals. Epithelial defect, cornea staining, neovascular area and OSDI scores are missing for N = 8 visits, and SANDE scores missing for N = 7 visits due to missed/telemedicine visits.