PURPOSE: To correlate in vivo confocal microscopy and impression cytology features of the corneal surface epithelia in patients with clinical features of partial or total limbal stem cell deficiency and to examine the limbal morphology. DESIGN: Prospective case-control observational study. METHODS: Twenty eyes of 17 consecutive patients (mean age 53.9 ± 9.2 years) presenting with clinical suspect of limbal stem cell deficiency and 10 eyes of 10 healthy control subjects were enrolled. In vivo confocal microscopy and impression cytology (PAS, cytokeratin 12, and cytokeratin 19) staining were performed in the central cornea. The inter-examination agreement was determined. Confocal microscopy scans were obtained in all patients to assess microscopic structure of the corneoscleral limbus, in all quadrants. RESULTS: Confocal microscopy and impression cytology agreement in testing the diagnostic hypotheses was high (κ = 0.85). The 2 methods were concordant in 18 out of 20 examined eyes (90%), revealing the presence of just corneal epithelium in 7 cases, just conjunctival epithelium (total limbal stem cell deficiency) in 5 cases, and mixed epithelium in 6 cases (partial limbal stem cell deficiency). Confocal imaging of the limbus revealed normal palisades of Vogt structure and epithelial transition in the healthy eyes while demonstrating a variable degree of alterations, including loss of the limbal palisades and of the normal epithelial mosaic, cystic epithelial changes, and subepithelial fibrosis, in the eyes affected by partial or total limbal stem cell deficiency. CONCLUSIONS: Confocal microscopy was useful for the noninvasive in vivo diagnosis of limbal stem cell deficiency, with a high degree of concordance with impression cytology, and to detect limbal alterations associated with partial or total conjunctivalization of the cornea.
PURPOSE: To correlate in vivo confocal microscopy and impression cytology features of the corneal surface epithelia in patients with clinical features of partial or total limbal stem cell deficiency and to examine the limbal morphology. DESIGN: Prospective case-control observational study. METHODS: Twenty eyes of 17 consecutive patients (mean age 53.9 ± 9.2 years) presenting with clinical suspect of limbal stem cell deficiency and 10 eyes of 10 healthy control subjects were enrolled. In vivo confocal microscopy and impression cytology (PAS, cytokeratin 12, and cytokeratin 19) staining were performed in the central cornea. The inter-examination agreement was determined. Confocal microscopy scans were obtained in all patients to assess microscopic structure of the corneoscleral limbus, in all quadrants. RESULTS: Confocal microscopy and impression cytology agreement in testing the diagnostic hypotheses was high (κ = 0.85). The 2 methods were concordant in 18 out of 20 examined eyes (90%), revealing the presence of just corneal epithelium in 7 cases, just conjunctival epithelium (total limbal stem cell deficiency) in 5 cases, and mixed epithelium in 6 cases (partial limbal stem cell deficiency). Confocal imaging of the limbus revealed normal palisades of Vogt structure and epithelial transition in the healthy eyes while demonstrating a variable degree of alterations, including loss of the limbal palisades and of the normal epithelial mosaic, cystic epithelial changes, and subepithelial fibrosis, in the eyes affected by partial or total limbal stem cell deficiency. CONCLUSIONS: Confocal microscopy was useful for the noninvasive in vivo diagnosis of limbal stem cell deficiency, with a high degree of concordance with impression cytology, and to detect limbal alterations associated with partial or total conjunctivalization of the cornea.