Vision-related quality of life and subscale items following intravitreal ranibizumab injections for central retinal vein occlusion

To evaluate the vision-related quality of life (VR-QOL) before and after intravitreal ranibizumab injections (IVR) for central retinal vein occlusion (CRVO) and to investigate subscale items of VR-QOL in detail. A multicenter, open-label, prospective and comparative study. Twenty-three patients with treatment naïve CRVO and 22 age-matched normal controls were included. VR-QOL was assessed by 25-Item Visual Function Questionnaire (VFQ-25) before and up to 12 months post-treatment. The VFQ-25 composite score and 12 subscales were compared between CRVO patients and normal controls. The VFQ-25 composite scores of CRVO patients showed significant improvement throughout the treatment period compared with baseline. The VFQ-25 composite scores both before and after treatment for CRVO were significantly lower than in normal controls. The subscale items, including general health, general vision, near activities, social functioning, mental health, role difficulties, dependency, and peripheral vision in CRVO, were significantly lower than in the normal controls even after treatment. Vision-related QOL was low in patients with CRVO before treatment and improved with IVR. In spite of the improvements, several subscale items remained lower than in normal controls following treatment. Attention should be paid to the patients with CRVO and their decreased visual, social and psychological QOL.


Introduction
Central retinal vein occlusion (CRVO) causes a significant loss of vision in the elderly [1], yet only limited treatments are available [2]. Since the introduction of antivascular endothelial growth factor (VEGF) therapy [3,4], the visual prognosis of CRVO patients has improved, yet many patients still complain of visual impairment after anti-VEGF treatment. Poor baseline visual acuity, poor visual acuity after the initial injection, coexistence of internal membrane (ERM) [12,13], macular hole (MH) [14], retinal detachment (RD) [15][16][17], proliferative diabetic retinopathy (PDR) [18,19], diabetic macular edema (DME) [19], and branch retinal vein occlusion (BRVO) [20,21]. Deramo et al. assessed the VR-QOL of CRVO patients using the VFQ-25 and found that their VR-QOL was lower than normal controls and was associated with visual acuity in the better-seeing eye and the number of systemic medical conditions [22]. The CRUISE study [3] and the GALILEO/ COPERNICUS study [4,23] used the VFQ-25 to assess VR-QOL of CRVO patients and observed improvement in VR-QOL after treatment. However, there are no reports on the relationship between VR-QOL and visual function in CRVO patients, or a detailed evaluation of changes in subscale after treatment. The purpose of the present study was to evaluate VR-QOL before and after treatment for CRVO in comparison with that of normal controls and investigate the subscale of VR-QOL in detail.

Study design
This was a multicenter, open-label, prospective study conducted in accordance with the Declaration of Helsinki and with the approval of the Institutional Review Board of the University of Tsukuba Hospital and Mito Kyodo General Hospital. Before inclusion, all patients were informed of the nature of the study, and their informed consent was obtained. Treatment naïve CRVO patients referred to the Tsukuba University Hospital or Mito Kyodo General Hospital were enrolled in this study. Patient inclusion criteria were: center-involving macular edema secondary to CRVO, central foveal thickness (CFT) > 250 μm, age over 18 years and under 85 years, and patients with paper-based informed consent. The exclusion criteria were: previous history of vitreoretinal surgery, previous history of ophthalmic disorders except mild refractive errors and mild cataracts, within 90 days of macular edema treatment involving sub-Tenon triamcinolone acetonide, intravitreal bevacizumab, intravitreal ranibizumab (IVR), intravitreal aflibercept, topical steroid, carbonic anhydrase inhibitors, within 90 days of intraocular surgery, the contralateral eye of CRVO, within 30 days of laser treatment. Patients with severe diabetes (HbA1c > 12%), hypertension (systolic blood pressure greater than 160 mmHg and diastolic blood pressure greater than 95 mmHg), cerebrovascular disease, such as cerebral infarction and hemorrhage, and ischemic heart disease, such as cardiac infarction and angina pectoris, were also excluded. We also included age-matched normal controls in this study. This study was designed to be exploratory. We investigated VR-QOL in 28 eyes after vitrectomy for ERM as preliminary study [13]. In this study, VFQ-25 composite score at baseline was 66.2 points, and at 3 months postoperatively was 77.9 points. The SD was 8.8, and the calculated sample size was 20 subjects based on the following assumption for the primary endpoint. This sample provided 80% power to detect a difference between groups (Paired t-test, two-sided α level of 5%).

Study visits and assessments
Best-corrected visual acuity (BCVA) and retinal microstructure were assessed monthly starting at pre-treatment up to 12 months post-treatment. VR-QOL was examined pre-treatment and at 3, 6, and 12 months post-treatment. BCVA measured on the Landolt chart was converted to the logarithm of the minimum angle of resolution (logMAR) and used for subsequent analysis. The retinal microstructure was assessed using spectral-domain OCT (Cirrus highdefinition OCT; Carl Zeiss). Five-line Raster Cross scans were performed using the Cirrus analysis software version 3.0., and scans with signal strength of more than 6/10 were considered appropriate. The CFT was evaluated based on the OCT images.
The VFQ-25 was used to assess VR-QOL in CRVO patients; it consists of 25 items that enable patients to assess visual symptoms and difficulties in daily activities. The answers to each of the 25 questions were sorted into 12 subscales, including general health, general vision, ocular pain, near activities, distance activities, social functioning, mental health, role difficulties, dependency, driving, color vision, and peripheral vision. The subscales were scored on a scale of 0-100 points, with 100 indicating the highest possible function or the least subjective impairment. We calculated the VFQ-25 composite score as the mean of the scores for the 11 subscales and excluded the general health subscale. The VFQ-25 used in this study was the Japanese version, modified to fit Japanese culture and lifestyle. The modified NEI VFQ-25 questionnaire has been assessed for reliability and validity and is proven to accurately measure VR-QOL in a Japanese population [24].

Intraocular injections
We treated the patients using three consecutive monthly intravitreal ranibizumab injections (IVR) (0.5 mg. Lucentis; Genetech), followed by pro re nata (3 + PRN). After three injections, the subjects were evaluated monthly and treated with intravitreal injections on PRN basis according to the protocol retreatment criteria. Re-administration criteria for PRN were as follows: CFT as assessed by OCT of ≥ 300 μm in the study eye, new cystoid retinal changes, retinal bleeding or subretinal fluid on OCT, logMAR visual acuity decrease of > 0.1 compared with values measured on the last visit.

Statistical analysis
The mean values and standard deviations were calculated. The Shapiro-Wilk test was used to test whether the data are normally distributed. A Mann-Whitney U test was performed to compare age, BCVA, and VR-QOL scores of CRVO patients and normal controls. The chi-square test was used to check for sex-difference between CRVO patients and normal controls. A Wilcoxon signed-rank test was used to compare baseline and each time point values for BCVA, CFT, and VFQ-25 composite score. Associations between VFQ-25 composite scores and 12 subscales and BCVA at baseline and 12 months post-treatment, the number of IVR injections and age were examined by the Spearman rank correlation test. All analyses were conducted using SPSS (version 27, IBC Corp). A P-value < 0.05 was considered statistically significant.

Results
We included 23 patients with CRVO and 22 normal controls. Clinical characteristics of patients with CRVO and normal controls are mentioned in Table 1. All parameters were normally distributed except for the duration of disease and the number of IVR injections. There was no difference between the two groups in terms of age or gender. The VFQ-25 composite score of CRVO patients was significantly worse than that of normal controls. No patient with CRVO discontinued treatment during the study period. The mean number of injections during the treatment period was 5.6 ± 2.0 (range 3-8 injections). None of the patients had ocular treatmentemergent serious adverse events such as retinal detachment, endophthalmitis, vitreous hemorrhage, iris neovascularization, and glaucoma. None of the patients underwent cataract surgery or panretinal laser photocoagulation in the affected eye or in the other eye during the follow-up period. Figure 1 shows changes from baseline BCVA and CFT. There was a significant improvement in BCVA and CFT after 1 month of treatment.

Changes in VR-QOL in patients with CRVO
The VFQ-25 composite score improved significantly at all observation periods compared with the baseline score (p < 0.005 in all) (Fig. 2). The VFQ-25 composite score in patients with CRVO at baseline was 62.6 ± 16.9, at 3 months it was 70.6 ± 15.6, at 6 months it was 72.2 ± 15.8, and at 12 months after treatment it was 74.3 ± 14.6; the mean increase from baseline VFQ-25 composite score was 8.0 points at 3 months, 9.6 points at 6 months, and 11.7 points at 12 months (Fig. 2). Table 2 shows the VFQ-25 composite scores and 12 subscales in normal controls and CRVO patients before and after treatment. The average VFQ-25 composite score in normal controls was 89.4 ± 7.4, and VFQ-25 composite scores in CRVO patients before and after treatment were worse than those of normal controls (p < 0.001 in both cases). The IVR treatment for 12 months significantly improved the subscale of general vision, ocular pain, near activities, distance activities, mental health, role difficulties, dependency, and composite scores. Several subscale items, including general health, general vision, near activities, social functioning, mental health, role difficulties, dependency, and peripheral vision in CRVO patients, were significantly lower than in the normal controls after treatment as well. Table 3 shows the relationship between the VFQ-25 composite scores and 12 subscales and BCVA in CRVO patients before and after treatment. At baseline, the VFQ-25 composite score and all subscales were not significantly associated with BCVA. The VFQ-25 composite score and subscale items of social functioning and dependency

Discussion
The VFQ-25 composite score improved from 62.6 points before treatment to 74.3 points 12 months after treatment in this study. In a study by Deramo et al. the VFQ-25 composite score in CRVO at baseline was 66.9 points [22], and in the COPERNICUS study it was 77.4 points [23]; the scores showed a significant correlation with BCVA 12 months after treatment. In addition, the VFQ-25 composite scores were not significantly associated with the number of IVR injections (r = 0.025, p = 0.920) and age (r = 0.333, p = 0.177).   [23], and in the CRUISE study it was 7.1 points [3, and after 12 months these scores were 7.8, 7.5, and 7.1 points respectively. In our study, the mean number of injections for 12 months was 5.6, lower than 9.8 in the CRUISE study [3], and 11.8 in the GALILEO study [4]. Despite these results, there was a greater improvement in the VFQ-25 composite score in the present study than in other studies, suggesting that 3 + PRN is acceptable as a standard treatment for CRVO in terms of QOL.
The IVR treatment for CRVO improved BCVA and CFT immediately after treatment and was maintained throughout the following 12 months. In the present study, the improvement in BCVA at 12 months was 0.2 logMAR (10 letters). In the CRUISE 3 BCVA at 12 months after anti-VEGF treatment was 13.9 and in the GALILEO study [4], it was 16.9 letters from baseline. The reason for less improvement in BCVA in our study compared to that in other studies may be the inclusion criteria. As mentioned earlier, CRVO patients observed in our study were lower than those of the COPER-NICUS study.
In this study the mean increase from baseline VFQ-25 composite score was 9.5 points after 6 months, and 11.7 points after 12 months of treatment. In the GALILEO study the mean increase from the baseline VFQ-25 composite score after 6 months of treatment was 7.1, points [4], in the  of vision (general vision, near activities, and peripheral vision) and psychosocial functioning (social functioning, mental health, role difficulties, and dependency) were more impaired than those in normal controls even after treatment. Attention should be paid to the patients with CRVO and their decreased visual, social and psychological QOL.
The limitations of this study were a small sample size, short follow-up duration and the lack of classification of patients into ischaemic or non-ischaemic CRVO. We evaluated the patients for 12 months after treatment. In the COPERNICUS study, BCVA in CRVO patients treated with intravitreal aflibercept injection at 24 months was 3.2 letters less than that at 12 months [30]. Since CRVO is a retinal disease with a poor prognosis, it is important to monitor visual functions and QOL in the long term. In our study, we could not classify CRVO into ischemic or non-ischemic types because we did not perform fluorescein angiography or OCT-angiography. Changes in VR-QOL following treatment are believed to differ between ischemic and non-ischemic CRVO. In addition, CRVO patients are relatively more likely to have diabetes, hypertension, and stroke. Although this study excludes patients with severe diabetes, hypertension and stroke, it includes patients with mild diabetes and hypertension. It is difficult to accurately investigate QOL based on CRVO alone, because patients with mild DM or HT may have poorer QOL than healthy subjects even if they do not have CRVO. Future studies with larger sample size, longer follow-up duration and classification of CRVO will further improve our understanding of VR-QOL in patients with CRVO.
in the CRUISE and GALILEO studies had a BCVA of 20/40 to 20/320 [3,4], whereas, in the present study, the range of BCVA of the patients at baseline was 20/20 to 20/2000. When CRVO patients were divided into the good VA group (baseline BCVA < 1 logMAR, 12 cases) and the poor VA group (baseline BCVA ≥ 1 logMAR, 11 cases), the improvement in BCVA at 12 months in the good VA group was 0.89 logMAR (18 letters) and in the poor VA group was 0.37 log-MAR (8 letters). There was a significant difference between the two groups (p < 0.0001, unpaired t-test). The improvement in BCVA was lower than in the abovementioned studies because our study included patients with extremely poor vision and patients with good vision.
We observed that the baseline VFQ-25 composite score was not associated with BCVA. Consistently, Deramo VA et al. also report that the VR-QOL of pre-treatment CRVO patients was not associated with visual acuity [22]. Visual acuity and VR-QOL are related in common eye diseases [26]; however, the relationship between VR-QOL and visual functions except for visual acuity has been indicated in many retinal diseases. The VR-QOL in ERM [13,19] and MH [14] was associated with metamorphopsia and not visual acuity. In PDR [18,19], DME [19], after RD [15], and vitreous floaters [27], contrast sensitivity affected the VR-QOL. Moreover, stereopsis was associated with VR-QOL after RD surgery [17]. Visual dysfunctions such as metamorphopsia [28] and aniseikonia [29] are reported in CRVO patients; hence, visual function factors other than visual acuity may be involved in the deterioration of VR-QOL. On the other hand, the post-treatment VFQ-25 composite score showed a significant correlation with posttreatment BCVA. This discrepancy cannot be explained. No significant correlation was found between the posttreatment BCVA and age (r = 0.312, p = 0.223), duration of disease (r = 0.283, p = 0.272), the number of IVR injections (r = 0.343, p = 0.178).
In the present study, IVR treatment for CRVO patients significantly improved subscale items of general vision, ocular pain, near activities, distance activities, mental health, role difficulties, and dependency. At baseline, all subscale items were lower in patients with CRVO than in normal controls. Even after treatment, subscale items, including general health, general vision, near activities, social functioning, mental health, role difficulties, dependency, and peripheral vision in CRVO, were significantly lower than those in the normal controls. In addition, the VFQ-25 composite score and subscale items social functioning, and dependency after treatment showed a significant correlation with BCVA. Most subscale items in pre-treatment in patients with CRVO were lower than in normal controls, consistent with previous reports [22]. Although anti-VEGF treatment for CRVO improved many subscale items, those related to the quality