Multi-parametric evaluation of autologous cultivated Limbal epithelial cell transplantation outcomes of Limbal stem cell deficiency due to chemical burn

Limbal stem cell deficiency (LSCD) is a condition induced by a number of diverse factors, ranging from genetic abnormalities to accidental limbal injuries. It develops when the epithelial stem and precursor cell systems of the limbal zone undergo numerical or functional degradation, thereby compromising the regenerative capacity of the corneal epithelium [1‐5]. Ensuing central corneal epitheliopathy results in reduced visual acuity, chronic inflammation, corneal pain, photophobia, epiphora, and blepharospasm, which when bilateral, may progress to full blindness. A consensus for the classification and grading of LSCD has been recently proposed [6]. Worldwide, the prevalence of LSCD is estimated to be between 1 to 9 cases per 100,000 individuals [7]. For unilateral LSCD the autologous transplantation of culture-expanded limbal epithelial cell (CLET) from these biopsies, represents a promising treatment with success rates of about 70–80% [8‐17]. For the more critical bilateral LSCD in spite of the hindrance presented by potential immune-rejection, allografts from semi-compatible donors are being implemented [18, 19]. To date, no specific international consensus on standards for the use of objective clinical parameters has been established. To contribute to the development of such standards, we have now subjected our clinical CLET work, completed under the guidelines for good laboratory practices (GLP) to a battery of recognized objective and patient-declared subjective outcome criteria.

All contralateral limbal donor eyes showed no change in visual acuity during the follow-up period or resulted in reports of pain or discomfort by the subjects. Figure 1 depicts representative results of the characterization of the six successful, in-parallel cultures that were not used for transplantation. Figure 1, a shows an H&E staining in the proximity of the biopsy. The epithelium consists of 2–3 layers of cells. There were no visible flattened cell profiles typical of the stratified cell when the epithelium is at rest. Figure 1, B shows an immunofluorescence micrograph of p63α (green) stain side-by-side with its corresponding nuclear DAPI blue counterstain. The percent of p63α-positive cells within the basal cell layer was high and quite similar for all six explants. The highest and lowest percentiles were, 56 and 47%, respectively. A clear feature of the staining pattern was the complete absence of protein staining in suprabasal cells; either p63α-positive cells were resilient to the upward mobility pressures. A second possibility, an extremely rapid down-regulation and degradation upon stratification seem less likely. Figure 2 shows micrographs of pre-operative (Preop) and 12 months post-operative (Postop) corneal surfaces for 5 of the treated subjects. The corresponding BCVAs (VA) are stated in the micrographs.

Table 1 displays the score ranges and average ± SD scores for all six criteria used. Pre- (control), and postoperative mean best-corrected visual acuity at 3, 6 and12-month were 1.15 ± 0.70; 0.28 ± 0.07; 0.29 ± 0.21 and 0.21 ± 0.13 LogMAR, respectively (n = 10; p < 0.05 vs all other times). Corresponding mean OSSs were 7.40 ± 2.01; 2.60 ± 0.84; 2.40 ± 0.96; 2,50 ± 1.08, (p < 0.05 vs. control). Pre-, and 1st year post-operative mean subjective visual losses were 7.30 ± 1.49 and 3.50 ± 1.92, respectively (p = 0.018). Patients’ responses to all parts of the 25-Item National Eye Institute Visual Function Questionnaire (Turkish version) were consistent with significant visual improvement (p < 0.05). Ocular pain score had no significant difference. In brief, objective visual criteria, BCVA and OSS, were significantly improved from the first determination after 1 month onward and continued to improve over the next 2 months. Three months after the surgery the LogMAR has undergone a 75% decrease (corresponding to a change from 20/290 to 20/35 in the Snellen scale) with a corresponding decrease of 60% in OSS. Notably, there were no further statistically significant improvements over the next 9 months. Lacrimation as measured by the Schirmer’s test was in the normal range prior to the surgical intervention and was not modified by it. Overall, in line with the changes in the objective criteria, the patient’s perceptions of vision and pain decreased by half from by the third Postop month. The scores of the OSDI and first two subparts (P1 and P2) of the NEI-VFQ criteria underwent moderate decreases. The high standard deviation intrinsic to these subjective tests, though, impacted the chances for statistically significant differences at many time points. The changes in two BCVA and OSS, for each individual patient, are shown in Fig. 3 A. and the correlation between the preoperative scores and the average for 3, 6 and 12-month scores is shown in Fig. 3 B for both BCVA (top) and OSS (bottom). These latter panels clearly show that there was no correlation between Preop scores and Postop outcome (p > 0.5).

CLET is an advanced treatment method for LSCD. In autologous transplantations, where systemic immunosuppression and its attendant sequel are avoided, success rates approach 80% [6]. Since its first reports in the late nineties, there have been raising amounts of publications reporting the clinical outcomes of the modality. However, the self-reporting of outcomes in these studies, are usually based on subjective grading, or do not depend on clear criteria, which causes bias of the success dependability. Given the increased frequency of this limbal stem cell transference procedure and its expansion to an increasing number of countries as national regulatory organisms approve its implementation, the development of universally accepted outcome criteria becomes essential to objectively compare the relative advantage of alternative CLET approaches for patient safety, vision improvement, and cost-effectiveness.

Shortt and collaborators have proposed and exemplified a scoring system for evaluating the LSCD inter-, and intra-observer agreements to evaluate the outcome of allogeneic CLET treatment for Steven Johnson Syndrome-induced LSCD [19]. In the present study, to evaluate the outcomes of autologous tissue-based CLET interventions, we applied an expanded system of scoring that incorporates that scoring approach and a battery of widely accepted criteria in ophthalmic practice. The objective criteria used objective provided a sound manner to quantitate the improvement in patient vision. In addition, our scoring system and its analysis identified several valuable features for future scoring improvement and surgical decision-making. The first observation to be made is that within the first year of follow up, the lion share of vision improvement occurred during the first trimester. The second notable, and somewhat counterintuitive observation was that the Postop score attained by the surgical intervention was not dependent on the degree of initial visual and ocular surface degradation These two features suggest that while the loss of function is primarily related to the catastrophic changes on the corneal epithelium/ supra stromal zone, stromal changes that affect vision occur simultaneously in a manner not directly correlated with the degree of changes at the corneal surface. Additionally, in the present study, other parametric tests were included such as OSDI, NEI-VFQ, subjective pain and visual loss scale in order to evaluate the outcomes. The meager improvement in pain sensation may relate to the latter concept. Although, the availability of a robustly documented set of scores will be very useful in the comparison across implementation of cell expansion, surgical methods and assessment of patient status over long periods in patients with LSCD due to ocular surface burns, it may not be applied to other causes of LSCD including aniridic keratopathy.

Objective clinical metrics demonstrate that in patients with limbal stem cell deficiency, cultivated limbal epithelial cell transplantation improves vision and ocular surface health and subjective visual perceptions.