PURPOSE: To investigate the utility of optical coherence tomography angiography (OCTA) for detecting pathologic vascularization within pigment epithelial detachments (PEDs).
METHODS: This was a retrospective, cross-sectional, consecutive case series. Multimodal imaging (structural OCT, fluorescein, and indocyanine green angiography) was used as the gold standard to classify PEDs as nonvascularized or vascularized. Optical coherence tomography angiography imaging of the PED was subsequently and independently evaluated to classify PEDs as vascularized or nonvascularized. Specifically, OCTA images were evaluated for the presence of abnormal flow on cross-sectional OCTA and the presence of a vascular complex on en face OCTA. Comparisons between OCTA and the gold standard were determined.
RESULTS: Sixty-four eyes of 49 patients were evaluated. A total of 18 eyes were classified as nonvascularized PED, and 46 eyes were classified as vascularized PED using the gold standard. Optical coherence tomography angiography was found to have a sensitivity of 76%, specificity of 61%, positive predictive value of 83%, and negative predictive value of 50% for detecting vascularized PEDs. False positive cases in the nonvascularized PED group were due to projection or flow artifacts from hyperreflective material overlying the PED. False negative cases were seen in eyes with minimal exudation on structural OCT and also those manifesting retinal pigment epithelial tears.
CONCLUSION: Our proposed two-step approach of OCTA interpretation, first using cross-sectional OCTA and then en face OCTA, may allow the detection of vascularization within PEDs and, in some cases, reduce the need for conventional angiography. Increased awareness about potential artifacts and limitations of OCTA may help clinicians interpret OCTA more accurately.