Arm-to-retina time predicts visual outcome of anti-vascular endothelial growth factor treatment for macular edema due to central retinal vein occlusion

To explore the factors associated with best-corrected visual acuity (BCVA) after anti-vascular endothelial growth factor (anti-VEGF) treatment for macular edema secondary to central retinal vein occlusion (CRVO). We retrospectively reviewed the medical charts of 22 eyes of 22 treatment-naïve patients with CRVO diagnosed between September 2014 and December 2020. They received anti-VEGF treatment and follow-up for > 12 months. Mean patient age was 64.3 years; 13 (59.1%) were men. Eyes with baseline arm-to-retina (AR) time ≥ 16 s had better BCVA at 12 months (adjusted for baseline BCVA and age; B, − 0.658; 95% confidence interval − 1.058 to − 0.257; P = 0.003), greater mean BCVA change (P = 0.006), lower frequency of residual macular edema at 12 months (P = 0.026) and recurrent and/or unresolved macular edema during 12 months (P = 0.046), and higher frequency of reduction in central retinal thickness ≥ 150 μm at 1 and 12 months (both P = 0.046). Delayed AR time was associated with a better visual outcome and macular edema improvement in CRVO after anti-VEGF treatment regardless of initial BCVA and age. Our results may help understand the pathogenesis and predict the visual prognosis of patients before anti-VEGF therapy initiation.


Scientific Reports
| (2022) 12:2194 | https://doi.org/10.1038/s41598-022-06281-w www.nature.com/scientificreports/ confounding factors. Reports have also shown that a longer duration from onset to treatment is associated with a worse prognosis in CRVO 1 . However, this may also be related to initial BCVA and OCT findings.
Here we focus on arm-to-retina (AR) time, measured from the administration of fluorescein into the antecubital vein until it becomes visible in the retinal arteries during fluorescein angiography (FA) recording 13 . The AR time is related to arteriosclerosis 14 , coronary slow-flow 15 , as well as carotid artery occlusion 16 , and systemic hemodynamics. Thus, to analyze the impact of AR time in CRVO in which hypertension and blood hyperviscosity are the risk factors 5 , would be of value. In addition, while previous reports have often evaluated BCVA change [17][18][19] , the final BCVA value is also important for lifestyle management after treatment. In this study, we analyzed the data of patients with CRVO to identify factors that affect visual outcome. The study will help to understand the pathogenesis of CRVO and facilitate obtaining appropriate informed consent regarding the visual outcome of the treatment from patients in the daily clinic.
As the outcome values at 12 months varied from individual to individual, we investigated the factors associated with good visual outcome. We found that eyes with initial AR time ≥ 16 s in FA that was longer than the normal range of AR time (10-15 s) 20 had better BCVA at 12 months after adjusting for baseline BCVA and age (B, − 0.658; 95% confidence interval − 1.058 to − 0.257; P = 0.003) ( Table 2). Sex, initial CRT, initial intraocular  www.nature.com/scientificreports/ pressure, actual value of AR time, initial systolic and diastolic blood pressures, and additional pan-retinal photocoagulation treatment during 12 months of follow-up did not significantly affect the visual outcome. Eyes with AR time ≥ 16 s exhibited a significantly better BCVA change (P = 0.006), and a trend of greater CRT reduction at 12 months (P = 0.069) ( Table 3). More importantly, a significantly smaller number of eyes with AR time ≥ 16 s still exhibited macular edema at 12 months (P = 0.026)-only 10% of eyes with AR time ≥ 16 s exhibited macular edema, while this percentage was 58.3% in the other group. We also analyzed whether the eyes with AR time ≥ 16 s had a lower risk of recurrent or unresolved macular edema during 12 months. Our results showed that only 30% of the eyes with AR time ≥ 16 s had recurrent or unresolved edema during the entire 12-month period, while this percentage was 75% in the other group (P = 0.046). In addition, reduction in CRT ≥ 150 μm both at 1 and 12 months after initial anti-VEGF treatment was achieved more frequently in eyes with AR time ≥ 16 s (both P = 0.046), indicating that these eyes had better responsiveness to anti-VEGF treatment.
Representative FA images of eyes with AR time ≥ 16 s and < 16 s are shown in Fig. 2.

Discussion
The study showed that the outcome of anti-VEGF treatment varied from individual to individual in patients with CRVO-related macular edema. Eyes with AR time ≥ 16 s exhibited better BCVA 12 months after anti-VEGF treatment with or without pan-retinal photocoagulation. These eyes exhibited residual or recurrent macular edema less frequently at 12 months and experienced recurrence or absence of remission of macular edema less frequently during this period. The frequency of reduction in CRT ≥ 150 μm at 1 and 12 months compared to that at baseline was also higher. The mean age of the CRVO patients in the current study was 64.3 years, which was similar to that in previous clinical trials on treatment-naïve patients, such as the CRUISE study (68 years) 18 , COPERNICUS study (66.3 years) 21 , and GALILEO study (61.5 years) 22 . A Japanese population-based cohort study, the Hisayama Table 2. Factors associated with visual outcome 12 months after initial treatment. Forced input analysis adjusted for age and initial best-corrected visual acuity. The selected factors were sex, initial central retinal thickness, initial intraocular pressure, actual value of AR time, initial systolic blood pressure, initial diastolic blood pressure, and additional photocoagulation treatment during 12 months of follow-up. **P < 0.01.  It is reported that initial BCVA as well as initial OCT findings, which may reflect the BCVA, are related to visual outcome [10][11][12] . However, we found that AR time ≥ 16 s in FA at baseline was associated with better BCVA 12 months after anti-VEGF treatment adjusting for initial BCVA and age. A longer AR time is related to worse vascular circulation [14][15][16] , and AR time is prolonged in combined artery and vein occlusion 25 . However, in the absence of significant artery occlusion, as shown in the current study, longer AR time was advantageous for better visual outcome at 12 months.
This may be because residual macular edema at 12 months was less frequently observed in patients who exhibited AR time ≥ 16 s at baseline, as shown in the current study. The patients more frequently achieved resolution of the macular edema after the treatment and less frequently showed recurrent macular edema after the resolution during the follow-up period. Previous reports showed that recurrent macular edema is related to worse visual prognosis in RVO 26,27 . Thus, effective resolution of macular edema may have contributed to the better visual outcome in the patients with AR time ≥ 16 s. In fact, patients with AR time ≥ 16 s at baseline showed a reduction in CRT ≥ 150 μm 1 month after initial anti-VEGF treatment, suggesting that these patients responded well to the anti-VEGF treatment.
Anti-VEGF treatment is believed to improve the vascular obstruction related to inflammatory cells, termed leukostasis 9 . Given that patients with AR time ≥ 16 s at baseline had better response to anti-VEGF treatment, they may have had a major influence of leukostasis. The scenario is consistent with the fact that AR time was correlated with the mean blur rate, which represents blood flow velocity, measured using laser speckle flowgraphy at the optic nerve head in patients with RVO 14 . Thus, longer AR time is related to high velocity, which may be at least in part due to leukostasis. Taken together, patients with AR time ≥ 16 s at baseline may have had a good response to anti-VEGF therapy after leukostasis resolution by VEGF inhibition, and resulting blood flow recovery. Moreover, it has been reported that macular edema recurrence is related to sustained VEGF expression 9 , and suppression of leukostasis may have reduced chronic relative ischemia in the retinal tissue, leading to downregulation of VEGF expression from the retina. In contrast, patients with shorter AR time could have been more influenced by mechanistic occlusion related to vascular anatomy and arteriosclerosis, and less by leukostasis, although further studies are required to clarify these aspects.
The limitations of the current study included a retrospective design and a relatively small sample size; however, considering that the prevalence of CRVO (0.2% in individuals over 40 years of age) is 10 times lower than that of branch RVO 5,6 (2% as reported in the Hisayama Study 5 ), the sample size was considerable for a single-center analysis. Nevertheless, actual AR time did not have a statistically significant effect on the visual outcome, most likely due to the small sample size. We included patients who received either or both ranibizumab and aflibercept therapy and laser therapy during the 12 month study period. However, this was similar to most previous clinical trials involving rescue laser therapies for patients [17][18][19] . Laser therapy also reduces VEGF expression, most likely by reducing oxygen demand and/or VEGF producing retinal cell number; the treatment concept is similar to that of anti-VEGF therapy in this regard. Besides, laser therapy during the 12 months of follow-up was not found to be associated with visual outcome in the current study. Because informed consent was necessary for each www.nature.com/scientificreports/ re-injection, patients could refuse the re-treatment, however, we confirmed that retreatment had been performed properly at most times based on the criterion, if follow-up OCT showed exudative changes. Visual outcome and BCVA at month 12 varied from patient to patient. This is consistent with previous reports on natural progression 8 and anti-VEGF treatments 3 , and could be related to the differences in the pathogenesis of CRVO. AR time ≥ 16 s was associated with better visual outcome regardless of initial BCVA and age, which may help us to better understand one aspect of the pathogenesis of CRVO, as discussed above. In addition, clinicians may refer to the AR time to explore the pathological condition of the individual patients, and to better predict visual outcome, which would facilitate obtaining informed consent from patients before initiating anti-VEGF therapy in daily practice. Nevertheless, further studies are warranted to validate the results of this study.

Methods
This retrospective study adhered to the tenets of the Declaration of Helsinki and was approved by the St. Luke's International University Ethics Committee (approval number: 20-R048).
Patients. The analyses were based on a detailed medical chart review of 22 eyes of 22 patients. All patients were treatment-naïve and were diagnosed with CRVO at the Vitreo-Retina Division Clinic of the Department of Ophthalmology at St. Luke's International Hospital in Tokyo, Japan, between September 2014 and December 2020 and treated and followed-up for more than 12 months. Patients without FA records at baseline or had undergone prior treatment for CRVO, including anti-VEGF therapy, steroid therapy, intraocular surgery in the previous 3-month period, and laser photocoagulation, were excluded. Informed consent has been obtained from the participants.
Eye examinations. All patients underwent BCVA measurement based on refraction tests, slit-lamp examinations, and binocular indirect ophthalmoscopy after pupil dilation with 0.5% tropicamide at each time-point. Data at baseline and 1, 3, 6, and 12 months after initial anti-VEGF treatment were analyzed.

FA.
For FA, we injected 10 mL containing 500 mg sodium fluorescein (Fluorescite; Novartis Pharma). The dyes were injected through an intravenous line over 10 s at the same rate for each patient and flushed through with sterile 5% glucose (Terumo Corporation, Tokyo Japan). Angiography was recorded from the time point that the dye reached the injection inserted point using a Heidelberg Spectralis HRA + OCT instrument (Heidelberg Engineering). The AR time was measured using the device equipped in the instrument. Fundus photographs for CRVO diagnosis were obtained using a Topcon TRC-50DX retinal camera (Topcon Corporation, Tokyo, Japan). Ischemic CRVO was defined as the presence of at least ten disc areas of retinal non-perfusion 28 . Macular ischemia was defined as the presence of retinal non-perfusion in the Early Treatment Diabetic Retinopathy Study grid center subfield 29 . OCT. OCT images were recorded at every follow-up visit using a Cirrus HD-OCT system (Zeiss, Oberkochen, Germany) or a Heidelberg Spectralis OCT system (Heidelberg Engineering GmbH, Dossenheim, Germany).
Treatments. All patients were treatment naïve for CRVO and were initially treated either with intraocular injection of ranibizumab (0.5 mg [0.05 mL]) or aflibercept (2 mg [0.05 mL]). Re-injections were recommended if the OCT image and/or fundus examination showed evidence of any exudative changes in the macula, identified as macular edema and/or subretinal fluid at the time of the follow-up examinations, and performed under informed consent. Application of the drug was changed from ranibizumab to aflibercept in some of the patients. Pan-retinal laser photocoagulation, were provided at each doctor's discretion.
Statistical analyses. Data are expressed as the mean ± standard deviation (SD). Generalized mixed model analyses, forced input analyses adjusted for age and initial BCVA, the Mann-Whitney U test, and Fishers' exact test (one-sided) were performed using SPSS version 27.0 (SPSS Japan, Tokyo, Japan). P values < 0.05 were considered statistically significant.
Ethics declarations. This retrospective study adhered to the tenets of the Declaration of Helsinki and was approved by the St. Luke's International University Ethics Committee (approval number: 20-R048).
Consent to participate/consent to publish. This retrospective study was approved by the St. Luke's International University Ethics Committee (approval number: 20-R048).

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.