PURPOSE: To evaluate the ability of a noncontact, prototype scanning laser confocal microscope to image microstructural features of filtering blebs in vivo and to correlate these with the clinical features. METHODS: Thirty-one blebs of 24 patients underwent noncontact in vivo confocal microscopy using the Rostock Cornea Module of the Heidelberg Retina Tomograph II (50x noncontact Nikon lens, 1-2 mum resolution, field of view: 500x500 mum). Blebs were clinically classified as successful (low, diffuse, microcystic, and hypovascular) (n=22, mean intraocular pressure<20 mm Hg without hypotensive medications) or failed (flat, cicatrized, and nonfiltering) (n=9, mean intraocular pressure>20 mmHg with or without hypotensive medications). Multiple transverse images were taken over the scleral flap at 1 to 2 mm from the limbus. Images were evaluated by 2 observers masked to clinical appearance and function of the bleb. RESULTS: Transverse images of the successful blebs showed epithelial microcysts (50 to 100 microm) and small-sized to medium-sized (100 to 150 microm) subconjunctival cystic spaces with scattered loose collagen-like networks around the scleral flap. Failed blebs showed diffuse hyperreflective dense stroma, increased and tortuous vascularization, and no cystic spaces. The microscopic findings of all blebs were consistent with previous reports of contact in vivo confocal microscopy. CONCLUSIONS: Noncontact in vivo confocal microscopy of filtering blebs correlated with the clinical morphology. This noncontact approach provides information about bleb morphology without the risk of bleb microinjury, leak, or infection. Further study and optimization for imaging surface tissues are needed to improve our understanding of bleb maturation and wound healing.
PURPOSE: To evaluate the ability of a noncontact, prototype scanning laser confocal microscope to image microstructural features of filtering blebs in vivo and to correlate these with the clinical features. METHODS: Thirty-one blebs of 24 patients underwent noncontact in vivo confocal microscopy using the Rostock Cornea Module of the Heidelberg Retina Tomograph II (50x noncontact Nikon lens, 1-2 mum resolution, field of view: 500x500 mum). Blebs were clinically classified as successful (low, diffuse, microcystic, and hypovascular) (n=22, mean intraocular pressure<20 mm Hg without hypotensive medications) or failed (flat, cicatrized, and nonfiltering) (n=9, mean intraocular pressure>20 mmHg with or without hypotensive medications). Multiple transverse images were taken over the scleral flap at 1 to 2 mm from the limbus. Images were evaluated by 2 observers masked to clinical appearance and function of the bleb. RESULTS: Transverse images of the successful blebs showed epithelial microcysts (50 to 100 microm) and small-sized to medium-sized (100 to 150 microm) subconjunctival cystic spaces with scattered loose collagen-like networks around the scleral flap. Failed blebs showed diffuse hyperreflective dense stroma, increased and tortuous vascularization, and no cystic spaces. The microscopic findings of all blebs were consistent with previous reports of contact in vivo confocal microscopy. CONCLUSIONS: Noncontact in vivo confocal microscopy of filtering blebs correlated with the clinical morphology. This noncontact approach provides information about bleb morphology without the risk of bleb microinjury, leak, or infection. Further study and optimization for imaging surface tissues are needed to improve our understanding of bleb maturation and wound healing.