Wide-field swept-source OCT angiography of the periarterial capillary-free zone before and after anti-VEGF therapy for branch retinal vein occlusion
(0 reviews)
Classification
Eye anatomy
Language
English
Abstract
Abstract Background The aim of the study was to investigate the changes in the periarterial capillary-free zone (paCFZ) after anti-vascular endothelial growth factor (VEGF) therapy in patients with branch retinal vein occlusion (BRVO) by wide-field swept-source optical coherence tomography angiography (SS-OCTA) and assess their associations with clinical outcomes. Methods In this retrospective observational study of 54 treatment-naïve BRVO patients with macular edema, we reviewed the findings of 12 × 12 mm2 SS-OCTA at baseline, 3, 6, and 12 months after intravitreal ranibizumab injections. The paCFZ and major retinal artery areas were measured on SS-OCTA images. The paCFZ area to artery area (P/A) ratio was calculated. Results The paCFZ areas and P/A ratios of first- and second-order arteries were significantly greater in BRVO eyes than in contralateral eyes (all P < 0.01), but there were no differences in the first- and second-order artery areas (P = 0.20 and 0.25, respectively). The paCFZ areas and P/A ratios decreased significantly at 3, 6, and 12 months after anti-VEGF therapy (all P < 0.01). The baseline P/A ratio was significantly correlated with the baseline best-corrected visual acuity (BCVA), central retinal thickness, and their improvements at 3, 6, and 12 months (all P < 0.05). Baseline BCVA and P/A ratios of first- and second-order arteries were independently associated with the final BCVA in multivariate linear regression. Conclusions Wide-field SS-OCTA shows that anti-VEGF therapy can lead to a significant improvement in the paCFZ parameters in BRVO. Smaller baseline P/A ratios on SS-OCTA tend to predict better visual outcomes at 12 months after anti-VEGF therapy. Background Branch retinal vein occlusion (BRVO) is a common retinal vascular disease worldwide, where macular edema is the main cause of visual impairment [1]. Macular edema is characterized by central retinal thickening caused by vascular leakage and fluid accumulation, which is triggered by the release of mediators, particularly vascular endothelial growth factor (VEGF) [2]. Large randomized clinical trials have demonstrated that intravitreal injection of anti-VEGF agents can reduce macular edema and improve visual acuity. Thus, anti-VEGF agents have become the standard therapy for patients with macular edema secondary to BRVO [3, 4]. However, some patients show resistance to anti-VEGF therapy and have poor prognosis [5]. To date, few clinical biomarkers that have been identified reflect disease activity and can be used to predict prognosis [6]. Fundus fluorescein angiography (FA) is considered the gold-standard method for evaluating retinal microvascular abnormalities and macular nonperfusion. Optical coherence tomography angiography (OCTA), a recently developed technique, offers many advantages for non-invasive retinal evaluation including high resolution and layered imaging of the posterior pole vasculature [7]. An increasing number of studies have utilized OCTA to evaluate the microvascular dysfunction in BRVO eyes with qualitative and quantitative analyses [8, 9]. The retinal periarterial capillary-free zone (paCFZ), first described by His in 1880, is a physiologically avascular area surrounding the retinal arteries exclusively in the superficial capillary plexus layer [10, 11]. Its formation is based on oxygen saturation during embryonic development because the retinal arteries contain a high concentration of oxygen, and vasoobliteration of capillaries in the periarterial area can be induced upon exposure to hyperoxia [12]. We previously reported that the retinal paCFZ was enlarged along unaffected major retinal arteries in eyes with BRVO [13]. Li et al. reported that the paCFZ was significantly larger in patients with severe nonproliferative diabetic retinopathy (DR) compared with those in normal subjects [14]. Based on those studies, it was suggested that the paCFZ might be a useful biomarker for monitoring retinal diseases associated with changes in the retinal microvasculature [15]. However, those studies were cross-sectional in design and, to our knowledge, no studies have reported the longitudinal changes in the paCFZ in retinal vascular disorders. In this retrospective study, we assessed the changes in the paCFZ before and after anti-VEGF therapy in patients with macular edema due to BRVO by using wide-field swept-source OCTA (SS-OCTA) images. We also examined the associations between these changes and other clinical parameters, including macular thickness and visual acuity, to investigate whether paCFZ parameters can predict the response and functional outcomes after anti-VEGF treatment in BRVO eyes. Methods Patients This retrospective, observational study was approved by the institutional review board of the Eye and ENT Hospital of Fudan University (Shanghai, China), and it followed the tenets of the Declaration of Helsinki. The study included consecutive patients with macular edema due to BRVO who were treated with intravitreal ranibizumab (IVR; 0.5 mg, Lucentis®; Novartis Pharma AG, Basel, Switzerland) at the Eye and ENT Hospital of Fudan University between December 2018 and September 2020. Healthy volunteers with no retinal disorders were also recruited. Written informed consent for the treatment and the use of clinical data for research aims was obtained from all the participants. The inclusion criteria included the following: (1) unilateral treatment-naïve macular edema secondary to BRVO diagnosed based on fundus examinations and FA; (2) symptom duration of less than 3 months; (3) minimum follow-up of 12 months; and (4) SS-OCTA images obtained before and after IVR. The exclusion criteria were as follows: (1) presence of other coexisting retinal disorders such as DR; (2) severe media opacities that hindered fundus examination; (3) high myopia [− 6 diopters (D) or lower]; (4) high hyperopia (+ 5.0 D or greater); (5) history of intravitreal injection of corticosteroids or anti-VEGF drugs; (6) history of laser photocoagulation or vitreoretinal surgery; (7) presence of any retinal disease or media opacity in the contralateral eye; and (8) poor-quality SS-OCTA images (signal strength less than 7) or obvious motion artifacts in either eye. The control group consisted of subjects with unnoticeable findings on fundus examinations and the study eye was randomly selected. Ocular examinations and treatment The patients’ demographics and medical history were collected from their medical records. All patients underwent complete ophthalmic examinations, which included the measurement of best-corrected visual acuity (BCVA) and intraocular pressure (IOP), manifest refraction (NIDEK RT-5100; NIDEK, Tokyo, Japan), slit-lamp biomicroscopy, fundus photography, spectral-domain optical coherence tomography (SD-OCT; Spectralis HRA + OCT; Heidelberg Engineering, Heidelberg, Germany), SS-OCTA (Plex Elite 9000; Carl Zeiss Meditec, Inc., Dublin, CA, USA), and ultra-wide-field FA (Optos 200Tx imaging system; Optos PLC, Dunfermline, UK). Ischemic BRVO was defined as BRVO with a capillary nonperfusion area of greater than 5 disk diameters on FA and nonischemic BRVO was defined as BRVO with a capillary nonperfusion area of ≤ 5 disk diameters [16]. The occlusion area in BRVO eyes defined as the quadrant affected by occluded venules was recorded after fundus examination. BCVA was measured using the Early Treatment Diabetic Retinopathy Study visual acuity testing charts. Central retinal thickness (CRT) was measured in a circular region of 1 mm diameter centered on the fovea using the built-in software. Macular edema was defined as a CRT of ≥ 300 μm or intraretinal or subretinal fluids in the parafoveal region on SD-OCT [17]. The patients received three injections at intervals of 1 month and subsequent pro re nata (PRN) injections at monthly visits, alongside comprehensive ophthalmic examinations, which included BCVA, slit-lamp biomicroscopy, fundus photography, SD-OCT, and SS-OCTA. Additional IVR injections were administered if intraretinal or subretinal fluid was observed, or if the CRT had increased by more than 50 μm from the lowest recorded value. SS-OCTA analysis SS-OCTA scans (3 × 3 mm2 and 12 × 12 mm2) centered on the fovea were performed with a central wavelength of 1050 nm and an A-scan speed of 100 kHz. The axial and transverse resolutions were 6.3 μm and 20 μm, respectively. Each 3 × 3 mm2 or 12 × 12 mm2 volume scan comprised 500 A-scans per B-scan with a total of 500 B-scan locations. All OCTA B-scans were examined and manual adjustments were performed for segmentation errors if present. The area of the foveal avascular zone (FAZ) of the superficial capillary plexus layer on the 3 × 3 mm2 SS-OCTA images were measured with the software ImageJ (http://imagej.nih.gov/ij/) as described previously [13, 18]. The paCFZ parameters in BRVO eyes were measured in the horizontally opposed quadrant, as any hemorrhage or edema in the involved quadrant was likely to make the measurement difficult and unreliable. The corresponding quadrants of the contralateral eyes and the healthy control eyes were also measured. We used custom software (Medraw; Image Medraw Technology Co., Ltd., Shanghai, China) to semiautomatically identify the paCFZ areas and the corresponding artery areas of the major retinal arteries on 12 × 12 mm2 en face images of the superficial capillary plexus layer. We previously reported that the software and measurement approach had high reproducibility and repeatability in healthy and BRVO eyes [13]. Briefly, two trained graders (WYT and JLG) identified the first- and second-order arteries on the fundus photographs and SS-OCTA images, as previously reported [19]. Next, the SS-OCTA images were imported into the software and a seed point was added to the paCFZ or the corresponding artery by the graders independently. The software detected the pixel of the seed point (SeedPix). The target area was defined as the area which contained pixels between 0 and SeedPix surrounding the seed point, and the software calculated the pixels of the target area (AreaPix) automatically. The value was then converted to square millimeters using the formula Area (mm2) = AreaPix × [(realH × realW)/(pixH × pixW)], where the real height (realH), real width (realW), pixel height (pixH), and pixel width (pixW) were 12 mm, 12 mm, 1024 pixels, and 1024 pixels, respectively, in the 12 × 12 mm2 SS-OCTA images. To reduce the potential influence of the individual variability in the branches of the major retinal arteries, the paCFZ area to artery area (P/A) ratio was calculated. The values of the paCFZ and the corresponding artery areas obtained by two graders (WYT and JLG) were averaged for analysis. Statistical analyses Data were presented as the number or mean ± SD. Paired t-tests and unpaired t-tests were used to compare data between the BRVO eyes and the contralateral eyes and between the contralateral eyes and the healthy control eyes, respectively. Intraclass correlation coefficients (ICCs) were used to test intra-observer repeatability and inter-observer reproducibility. CRT, BCVA, artery areas, and paCFZ parameters (paCFZ area and P/A ratio) at baseline and at 3, 6, and 12 months after anti-VEGF treatment were compared using repeated-measures One-way analysis of variance (ANOVA) with Geisser-Greenhouse correction. The Bonferroni post hoc test was then used to compare the post-treatment values with the baseline values in cases presenting with significant fluctuations. Associations between baseline CRT and BCVA, improvements in CRT and BCVA, and the number of IVR injections with the P/A ratio were evaluated using Pearson’s correlation coefficients. Uni- and multi-variate linear regression analyses were used to determine which baseline variables were significantly associated with the BCVA at 12 months. Variables of P < 0.05 in the univariate analysis were included in the multivariate analysis by a stepwise method. SPSS Statistics (version 21; SPSS, Chicago, IL, USA) was used to perform the statistical analyses. Significance was set at a P value of less than 0.05. Results Fifty-four patients and 30 control subjects were included in this study; their clinical characteristics are summarized in Table 1. No significant difference was observed between the BRVO and the control groups pertaining to the demographics and systemic disease at baseline. Thirty-six patients had ischemic BRVO, and 18 patients had nonischemic BRVO. The IOP (P = 0.15) and refractive error (P = 0.10) were not significantly different between the BRVO eyes and the contralateral eyes. The BRVO eyes had worse BCVA (P < 0.01), thicker CRT (P < 0.01) and larger FAZ areas (P < 0.01) than the contralateral eyes. The paCFZ areas and the P/A ratios of the first- and second-order arteries were significantly greater in BRVO eyes than in the contralateral eyes (all P < 0.01). The first- and second-order artery areas (P = 0.20 and 0.25, respectively) were not significantly different between the BRVO eyes and the contralateral eyes. Furthermore, there were no significant differences in BCVA, CRT, FAZ area, paCFZ area, artery area and the ratio of the paCFZ area to the artery area between the contralateral eyes and the healthy control eyes (all P > 0.05, Table 1). The intra-observer agreement and inter-observer agreement were extremely high for the baseline paCFZ and areas of the first- and second-order arteries, with ICCs of more than 0.9 (Additional file 1: Table S1). The mean number of IVR injections administered over 12 months was 4.98 ± 2.17 (range: 3–10). Table 1 Baseline clinical data of patients with BRVO and healthy control participants in this study Full size table Figure 1 and Table 2 show the time-dependent changes in BCVA, CRT, and paCFZ parameters after anti-VEGF therapy. The BCVA improved significantly after therapy, from 53.56 ± 16.09 letters at baseline to 63.37 ± 16.96 letters at 3 months, 63.98 ± 13.48 letters at 6 months, and 65.67 ± 12.89 letters at 12 months (all P < 0.01). The CRT decreased significantly from 581.70 ± 191.90 μm at baseline to 283.70 ± 78.02 μm at 3 months, 286.90 ± 85.19 μm at 6 months, and 262.5 ± 47.10 μm at 12 months (all P < 0.01). The paCFZ areas and P/A ratios of the first- and second-order arteries decreased significantly during the 12-month follow-up period (all P < 0.01, Fig. 2), the values of which returned almost to the levels of the healthy contralateral eyes. Pairwise comparisons between baseline and 3, 6, and 12 months after treatment revealed significant reductions in the paCFZ areas (all P < 0.01) and P/A ratios (all P < 0.01) of the first- and second-order arteries. However, the areas of the first- (P = 0.05) and second-order arteries (P = 0.20) did not decrease significantly. The FAZ areas (0.39 ± 0.15 mm2) did not significantly change at 12 months after anti-VEGF treatment compared with those at baseline (0.41 ± 0.15 mm2, P = 0.61). Furthermore, a subgroup analysis demonstrated that the P/A ratios of the first- and second-order arteries were significantly larger in the eyes with ischemic BRVO than nonischemic BRVO at baseline (Table 3). Although the paCFZ areas and P/A ratios were both significantly reduced at 12 months after anti-VEGF treatment in these two subgroups, the P/A ratios of the first- and second-order arteries were still significantly larger in the eyes with ischemic BRVO than nonischemic BRVO (Table 3). The paCFZ areas, artery areas and P/A ratios in the healthy contralateral eyes did not change significantly at 12 months compared with those at baseline, respectively (Additional file 1: Table S2).
Authors
- Ahmed Hakim