Abstract
Purpose: :
OCT provides superb morphologic information about subretinal pathology in disease processes such as choroidal neovascularization (CNV) and retinal angiomatous proliferation (RAP). However, the automated quantification routines included with the Stratus OCTTM device often incorrectly identify the retinal boundaries in CNV and RAP by combining the subretinal and retinal spaces, and also fail to quantify retinal pigment epithelial detachments. Custom software was written at the Doheny Image Reading Center (DIRC) to allow a human grader to accurately measure retinal and subretinal features on OCT B–scans. This software was used on data from patients with CNV and RAP to characterize differences in retinal thickness and volume that might be present in different forms of these diseases.
Methods: :
OCT data from 27 eyes of 25 patients with CNV (11 classic, 11 occult, 5 RAP) were graded using custom DIRC OCT analysis software. Using a computer mouse, graders outlined the internal limiting membrane, the photoreceptor outer segment layer, the inner aspect of subretinal tissue (if present), the RPE hyperreflective band, and the approximate location of Bruch’s membrane. Thicknesses and volumes were calculated for the retinal, subretinal tissue, subretinal fluid, and sub–RPE spaces. Values were compared using the Wilcoxon Rank Sum Test.
Results: :
The mean retinal volume in RAP lesions [8.42 ± 0.78 (7.11, 9.12) mm3] was greater than both classic CNV [7.32 ± 1.46 (6.05, 11.54) mm3, p<0.05] and occult CNV lesions [7.20 ± 0.79 (6.07, 8.42) mm3, p<0.01]. Although the total subretinal space volume in occult CNV [0.34 ± 0.50 (0.00, 1.71) mm3] was less than both classic CNV [0.93 ± 0.89 (0.34, 3.38 mm3, p<0.01] and RAP lesions [0.95 ± 0.63 (0.33, 1.85) mm3, p<0.05], there was no statistically significant difference in the subretinal fluid volume between the three lesion types. As expected, the subretinal tissue volume in classic CNV [0.59 ± 0.77 (0.18, 2.87) mm3] was greater than occult CNV [0.01 ± 0.05 (0.00, 0.15) mm3, p<0.001] and RAP lesions [0.13 ± 0.21 (0.00, 0.49) mm3, p<0.05]. Occult CNV had a larger amount of sub–RPE tissue volume than classic CNV (p<0.05) and RAP lesions (p<0.05).
Conclusions: :
Total retinal volume in patients with RAP is greater than in patients with classic or occult CNV. No statistically significant difference in subretinal fluid production was found between the lesion subtypes. Quantitative sub–analysis of conventional OCT data may provide a useful method of monitoring individual lesion components when studying new therapies or treating patients with CNV or RAP.
Keywords: choroid: neovascularization • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • imaging/image analysis: clinical