Outcomes of surgical interventions for the treatment of limbal stem cell deficiency

Background & objectives: In the current scenario, with availability of different surgical procedures for limbal stem cell deficiency (LSCD), there exists no common consensus as to the standardization of the management protocol for the same. In addition, there also exists diversity in the views about the clinical diagnosis, ancillary investigations and clinical parameters. The objective of the present study was to evaluate the reported outcomes of surgical interventions for the management of LSCD.

Methods: A systematic review of published literature on limbal stem cell transplantation (LSCT) was performed using Ovid Medline, Embase and PubMed for a duration of 2009 to 2019. Original studies including prospective, retrospective case series and randomized controlled trials, articles in English language, articles with access to full text and studies with more than or at least 10 patients were included in this review. Data related to clinical and visual outcomes were evaluated, and pool estimates of different surgeries were calculated using random-effects model and individually using Pearson's Chi-square test.

Results: A total of 1133 abstracts were evaluated. Finally, 17 studies were included for the analysis. Among these 17 studies, direct limbal lenticule transplantation was performed in five studies, of which autologous tissue from the fellow eye [conjunctival limbal autograft (CLAU)], allograft from a cadaver/live donor [keratolimbal allograft (KLAL)/conjunctival limbal allograft (CLAL)] and combination of CLAU plus KLAL were done in one, three and one studies, respectively. The ex vivo expanded cultivated limbal epithelial transplantation (CLET) was reported in six studies and simple limbal epithelial transplantation (SLET) in four studies. Two were comparative studies comparing CLET and CLAL (living-related CLAL) with cadaveric KLAL, respectively. Outcome analysis of the included studies showed significant heterogeneity. Calculated pool rate for various types of surgeries was calculated. The pool estimate for CLAL was 67.56 per cent [95% confidence interval (CI), 41.75-93.36; I²=83.5%, P=0.002]. For KLAL, this value was 63.65 per cent (95% CI, 31.38-95.91; I²=92.4%, P=0.000). Pool estimate for CLET was 78.90 per cent (95% CI, 70.51-87.28; I²=73.6%, P=0.001). Corresponding values for SLET were 79.08 per cent (95% CI, 74.10-84.07; I²=0.0%, P=0.619). CLAU and combination of CLAU plus KLAL were done in one study each; hence, statistical analysis could not be done. The functional outcome in terms of gain in visual acuity post-operatively was better in KLAL (P<0.005) and SLET group as compared to CLET group.

Interpretation & conclusions: The present analysis suggests that though the anatomical success rates were almost identical between SLET, CLET, CLAL, and KLAL procedures, the functional success rates were better following KLAL and SLET procedures as compared to CLET. Decision for LSCT for cases of ocular burns based on either clinical judgement of the surgeon or individual diagnosis remains a suitable option.