Optical coherence tomography findings as a predictor of clinical course in patients with branch retinal vein occlusion treated with ranibizumab

Abstract

Purpose: To examine the relationship between optical coherence tomography (OCT) images and clinical course in eyes with branch retinal vein occlusion (BRVO) treated with intravitreal ranibizumab injection (IVR).

Design: Prospective cohort study.

Participants: Thirty eyes of 30 patients with BRVO treated with IVR.

Methods: All patients received 1 initial IVR followed by repeated injections in the pro re nata (PRN) regimen. Correlations between logarithm of minimum angle of resolution best-corrected visual acuity (logMAR BCVA) or number of IVRs after 12 months and OCT parameters including the external limiting membrane (ELM), ellipsoid zone (EZ), interdigitation zone (IZ), and photoreceptor outer segment (PROS) length at first resolution of macular edema (ME) were assessed. Resolution of ME was defined as central foveal thickness <300 μm and the absence of subretinal fluid. OCT parameters influencing BCVA and number of IVRs were evaluated using multivariate analysis. Correlations between nonperfusion areas (NPAs) and thinning areas and changes in retinal thickness of BRVO-affected areas were assessed.

Results: Of the 30 patients, 27 completed this study and were included in the statistical analyses. The mean logMAR BCVA at 3, 6, and 12 months was 0.16 ± 0.19, 0.09 ± 0.20, and 0.07 ± 0.20, respectively, which improved significantly from baseline at each visit (p < 0.0001, respectively), while the mean number of IVRs at 12 months was 3.9 ± 2.2. The mean number of IVRs for the first resolution of ME was 1.6 ± 0.8. Eyes with ELM and EZ defects at the points of first resolution of ME were correlated with a significantly lower BCVA at 12 months compared with eyes with preserved ELMs and EZs (p = 0.035, p = 0.002, respectively). However, eyes with IZ defects at the points of first resolution of ME were not correlated with a significantly lower BCVA at 12 months compared with eyes with preserved IZs (p = 0.160). Defects in the EZ at the points of first resolution of ME significantly affected the number of IVRs at 12 months (p = 0.042), although the ELM and IZ did not. PROS length at the points of first resolution of ME was significantly correlated with BCVA and number of IVRs at 12 months (p = 0.006, p = 0.0008, respectively). In multivariate analysis, PROS length at the points of first resolution of ME had the most significant effect on BCVA and number of IVRs (p = 0.013, p = 0.012, respectively). NPA size on fluorescein angiography and thinning area on OCT within the macular area showed a significant correlation (p = 0.003, r = 0.971). The retinal thickness of ischemic BRVO-affected areas was significantly less than that of control areas at 10, 11, and 12 months (p = 0.001, p = 0.005, p = 0.003, respectively).

Conclusion: We showed that the 1+PRN regimen may be a useful therapy for ME due to BRVO. In addition, PROS length at points of first resolution of ME appears to be a good indicator of BCVA and number of IVRs in BRVO patients.

Fig 1. Optical coherence tomography (OCT) image before and after intravitreal ranibizumab injection (IVR).

(A) OCT image before IVR. The outer retina was not detected. (B) OCT image at the point of resolution of ME. (C) Enlargement of the OCT image showing the ELM (a), EZ (b), IZ (c), and PROS length (d). ELM = external limiting membrane; EZ = ellipsoid zone; IZ = interdigitation zone; PROS = photoreceptor outer segment.

Fig 2. Fluorescein angiography (FA) images and macular cube analysis of branch retinal vein occlusion (BRVO).

(A, B) FA images and macular cube analysis of ischemic inferotemporal BRVO. These parameters show the retinal thickness of each sector within the macular area as calculated automatically. The retinal thickness of the BRVO-affected area is shown in the lower section (black arrow), and a symmetrical area was examined to determine the retinal thickness of the control area (red arrow). (C, D) FA images and macular cube analysis of nonischemic superotemporal BRVO. In this case, the retinal thickness of the BRVO-affected area is shown in the upper section (black arrow), and a symmetrical area was examined to determine the retinal thickness of the control area (red arrow).

Fig 3. Fluorescein angiography (FA) of ischemic branch retinal vein occlusion (BRVO).

(A) FA image of ischemic BRVO. The nonperfusion area is outlined in white. The area indicated by the red square was identical to the OCT image obtained by macular cube scan (200 × 200) (right, indicated by the yellow square). (B) OCT image showing a layer map of retinal thickness. The thinning area is indicated in blue. OCT = optical coherence tomography.

Fig 4. Changes in the mean central foveal thickness (CFT) and mean logarithm of minimum angle of resolution (logMAR) best-corrected visual acuity (BCVA) from baseline to 12 months after intravitreal ranibizumab injection (IVR).

(A) CFT and (B) logMAR BCVA from baseline to 12 months after IVR. The CFT and visual acuity at 12 months were significantly improved compared with baseline (p < 0.05, respectively). logMAR = logarithm of minimum angle of resolution. IVR = intravitreal ranibizumab injection.

Fig 5. Relationship between optical coherence tomography (OCT) parameters and visual outcome or number of intravitreal ranibizumab injections (IVRs).

(A) Correlation between logMAR BCVA at 12 months and the status of the ELM, EZ, and IZ. (B) Correlation between logMAR BCVA at 12 months and PROS length. (C) Correlation between the number of IVRs and the status of the ELM, EZ, and IZ. (D) Correlation between the number of IVRs and PROS length. OCT = optical coherence tomography; logMAR = logarithm of minimum angle of resolution; BCVA = best-corrected visual acuity; IVR = intravitreal ranibizumab injection; ELM = external limiting membrane; EZ = ellipsoid zone; IZ = interdigitation zone; PROS = photoreceptor outer segment.

Fig 6. Fluorescein angiography (FA) images and color maps of retinal thickness using optical coherence tomography (OCT) images.

(A, C) FA images of eyes with nonperfusion areas (NPAs) (shown in orange). (B, D) Color maps of eyes with NPAs which corresponded to the thinning areas (in blue) detected by OCT. (E) FA of eyes without NPAs. (F) The thinning areas were not detected in OCT.

Fig 7. Correlation between nonperfusion areas (NPAs) in fluorescein angiography (FA) and thinning areas in optical coherence tomography (OCT).

Fig 8. Changes in the retinal thickness of ischemic and nonischemic branch retinal vein occlusion (BRVO).

(A) Changes in the retinal thickness of ischemic BRVO-affected areas and control areas at points of resolution of macular edema (ME) from 1 month to 12 months. Retinal thicknesses at point of recurrence of ME were excluded. Retinal thicknesses of ischemic areas after 10 months were significantly less than those of control areas (p < 0.01). (B) Changes in the retinal thickness of nonischemic BRVO-affected areas and control areas at points of resolution of ME from 1 month to 12 months. The retinal thickness of nonischemic areas tended to be greater than that of control areas. IVR = intravireal ranibizumab.

Fig 9. Changes in the ratio of thinning areas within the macular area at points of resolution of macular edema (ME) compared with the final visit.

The ratio of the thinning area, which is indicated in blue on the macular cube scan (200 × 200), was defined as the measurement of the thinning area at each visit/measurement of the thinning area at the final visit. Only eyes that met the definition of resolution of ME are included. The ratio of thinning areas gradually increased until 9 months (p < 0.05, respectively). IVR = intravitreal ranibizumab.