UNLABELLED: The surgical removal of choroidal neovascularization (CNV) in age-related macular degeneration (AMD) causes a retinal pigment epithelial defect (RPED) corresponding to the area of diffuse RPE damage. We describe angiographic features of recurrent CNV in AMD after surgical membranectomy in order to elucidate the nature of persistence and recurrence. METHODS: After digitalization of the pre- and postoperative fluorescein angiographic images of eight patients with recurrent CNV in AMD we determined the morphology (well or ill-defined) and the area of the CNV and of the subretinal hemorrhage preoperatively and of the recurrent CNV and of the RPE defect postoperatively. RESULTS: The nature of recurrences showed differences between preoperatively well- and ill-defined CNV. Four preoperatively well-defined CNV with surrounding subretinal hemorrhage showed recurrences in the entire area of the preoperative CNV excluding the retinotomy 8-9 weeks postoperatively. Four preoperatively ill-defined CNV with subretinal hemorrhage developed marginal recurrences at the rim of the RPED. There was no background fluorescence in the area of the RPED. CONCLUSION: The nature of recurrences extending over the entire area of the preoperatively well-defined CNV without loss of background fluorescence only a few weeks after surgical removal of well-defined CNV suggests partial persistence. The removal of the subretinal well-defined CNV could leave sub-RPE parts in locations that preoperatively cannot be visualized angiographically. The marginal recurrence of preoperatively ill-defined CNV weeks to months postoperatively shows angiographic similarities to recurrent CNV after laser coagulation.
UNLABELLED: The surgical removal of choroidal neovascularization (CNV) in age-related macular degeneration (AMD) causes a retinal pigment epithelial defect (RPED) corresponding to the area of diffuse RPE damage. We describe angiographic features of recurrent CNV in AMD after surgical membranectomy in order to elucidate the nature of persistence and recurrence. METHODS: After digitalization of the pre- and postoperative fluorescein angiographic images of eight patients with recurrent CNV in AMD we determined the morphology (well or ill-defined) and the area of the CNV and of the subretinal hemorrhage preoperatively and of the recurrent CNV and of the RPE defect postoperatively. RESULTS: The nature of recurrences showed differences between preoperatively well- and ill-defined CNV. Four preoperatively well-defined CNV with surrounding subretinal hemorrhage showed recurrences in the entire area of the preoperative CNV excluding the retinotomy 8-9 weeks postoperatively. Four preoperatively ill-defined CNV with subretinal hemorrhage developed marginal recurrences at the rim of the RPED. There was no background fluorescence in the area of the RPED. CONCLUSION: The nature of recurrences extending over the entire area of the preoperatively well-defined CNV without loss of background fluorescence only a few weeks after surgical removal of well-defined CNV suggests partial persistence. The removal of the subretinal well-defined CNV could leave sub-RPE parts in locations that preoperatively cannot be visualized angiographically. The marginal recurrence of preoperatively ill-defined CNV weeks to months postoperatively shows angiographic similarities to recurrent CNV after laser coagulation.