Abstract
Purpose: :
To determine if features that may be used to guide treatment decisions with anti-VEGF therapy are seen with equal frequency using conventional time-domain OCT (TD-OCT) and 3D Fourier Domain OCT (3D-OCT).
Methods: :
Retrospective review of patients receiving at least one intravitreal injection of either ranibizumab or bevacizumab who had simultaneous TD-OCT and 3D-OCT imaging within 1 week of their therapy. The presence and approximate size (small, medium, large) of subretinal fluid (SRF), retinal pigment epithelial detachments (PED) and cystoid macular edema (CME) were independently assessed by a single grader in each set of scans. Differences in grading were adjudicated by a head-to-head comparison between scans.
Results: :
62 eyes of 45 patients with simultaneous TD-OCT and 3D-OCT imaging undergoing ranibizumab (44/62) or bevacizumab (18/62) therapy between July, 2006 and November, 2007 were found. 96% of patients had choroidal neovascular membranes and 4% had central retinal vein occlusions. All 3D-OCT scans were acquired with a 512 x 128 A-scan protocol. 52/62 (84%) of TD-OCT scans were acquired as high resolution radial line scans and 10/62 (16%) were acquired as fast macular thickness scans.33/62 eyes (53%) had CME, 40/62 eyes (65%) had SRF, and 53/62 eyes (85%) had PEDs. All features detected on TD-OCT were present on 3D-OCT. At least one graded feature was larger on 3D-OCT in 38/62 cases (61%) and larger on TD-OCT in 1/62 (2%) cases. CME was more extensive on 3D-OCT in 12/33 (36%) cases and on TD-OCT in 1/33 (3%) cases. SRF fluid was more extensive on 3D-OCT in 21/40 (53%) cases and none of the TD-OCT cases. PEDs were more extensive on 3D-OCT in 17/53 (32%) cases and none of the TD-OCT cases. 3D-OCT detected at least one feature that was not detected by TD-OCT in 12/62 cases (19%). 15% (5/33) of cases with CME on 3D-OCT had no CME on TD-OCT. 18% (7/40) of cases with SRF on 3D-OCT had no SRF on TD-OCT. 4% (2/53) of cases with PEDs on 3D-OCT had no PED on TD-OCT. If presence any graded feature was used as the only treatment indication, 3D-OCT would have changed management in 6/62 cases (10%). 7/10 (70%) of fast macular thickness scans missed findings when compared to 3D-OCT while only 22/52 (42%) of radial line scans had deficiencies.
Conclusions: :
Features associated with anti-VEGF treatment decisions and seen on 3D-OCT scans were missing in 19% of TD-OCT cases. Features appeared more extensive on 3D-OCT scans in 61% of cases. In this study, the sensitivity of TD-OCT for CME was 85%, for SRF was 83% and for PEDs was 96%. 3D-OCT has a higher sensitivity than TD-OCT for findings that guide treatment decisions in anti-VEGF therapy.
Keywords: age-related macular degeneration • imaging/image analysis: clinical • retina