Marie Borderie1,2, Kate Grieve1,2, Kristina Irsch1,2,3, Djida Ghoubay1,2, Cristina Georgeon1,2, Celine De Sousa4, Laurent Laroche1,2, Vincent M Borderie1,2. 1. Quinze-Vingts National Ophthalmology Hospital, Paris VI University, Paris, France. 2. Vision Institute/CIC 1423, UMR_S 968/INSERM, U968/CHNO/CNRS, UMR_7210, UPMC University Paris 06, Paris, France. 3. Laboratory of Ophthalmic Instrument Development, The Wilmer Eye Institute, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA. 4. Tissue Bank, French Blood Establishment, Paris, Ile-de-France, France.
Abstract
PURPOSE: To provide quantitative parameters for assessment of human donor corneal stroma by imaging stromal features of diseased and normal human corneas with full-field optical coherence microscopy (FFOCM), using confocal microscopy (CM) and histology as reference techniques. METHODS: Bowman's layer (BL) thickness and keratocyte density were assessed ex vivo in 23 human donor corneas and 27 human pathological corneas (keratoconus and other corneal disorders) with FFOCM, CM and histology. Stromal backscattering was assessed with FFOCM. Additionally, 10 normal human corneas were assessed in vivo with CM. RESULTS: In FFOCM, the logarithm of the normalized stromal reflectivity was a linear function of stromal depth (R2 = 0.95) in human donor corneas. Compared with keratoconus corneas, human donor corneas featured higher BL thickness (p = 0.0014) with lower coefficient of variation (BL-COV; p = 0.0002), and linear logarithmic stromal reflectivity with depth (higher R2 , p = 0.0001). Compared with other corneal disorders, human donor corneas featured lower BL-COV (p = 0.012) and higher R2 (p = 0.0001). Using the 95% confidence limits of the human donor cornea group, BL thickness < 6.5 μm (sensitivity, 57%; specificity, 100%), BL-COV > 18.6% (79%; 100%) and R2 < 0.94 (93%; 71%) were revealed as indictors of abnormal cornea. In CM, keratocyte density decreased with stromal depth (r = -0.56). The mean overall keratocyte density (cells/mm2 ) was 205 in human donor corneas, 244 in keratoconus, 176 in other corneal disorders and 386 in normal corneas. CONCLUSION: Full-field optical coherence microscopy (FFOCM) provides precise and reliable parameters for non-invasive assessment of human donor corneal stroma during storage, enabling detection of stromal disorders that could impair the results of keratoplasty.
PURPOSE: To provide quantitative parameters for assessment of humandonorcorneal stroma by imaging stromal features of diseased and normal human corneas with full-field optical coherence microscopy (FFOCM), using confocal microscopy (CM) and histology as reference techniques. METHODS: Bowman's layer (BL) thickness and keratocyte density were assessed ex vivo in 23 humandonor corneas and 27 human pathological corneas (keratoconus and other corneal disorders) with FFOCM, CM and histology. Stromal backscattering was assessed with FFOCM. Additionally, 10 normal human corneas were assessed in vivo with CM. RESULTS: In FFOCM, the logarithm of the normalized stromal reflectivity was a linear function of stromal depth (R2 = 0.95) in humandonor corneas. Compared with keratoconus corneas, humandonor corneas featured higher BL thickness (p = 0.0014) with lower coefficient of variation (BL-COV; p = 0.0002), and linear logarithmic stromal reflectivity with depth (higher R2 , p = 0.0001). Compared with other corneal disorders, humandonor corneas featured lower BL-COV (p = 0.012) and higher R2 (p = 0.0001). Using the 95% confidence limits of the humandonor cornea group, BL thickness < 6.5 μm (sensitivity, 57%; specificity, 100%), BL-COV > 18.6% (79%; 100%) and R2 < 0.94 (93%; 71%) were revealed as indictors of abnormal cornea. In CM, keratocyte density decreased with stromal depth (r = -0.56). The mean overall keratocyte density (cells/mm2 ) was 205 in humandonor corneas, 244 in keratoconus, 176 in other corneal disorders and 386 in normal corneas. CONCLUSION: Full-field optical coherence microscopy (FFOCM) provides precise and reliable parameters for non-invasive assessment of humandonorcorneal stroma during storage, enabling detection of stromal disorders that could impair the results of keratoplasty.