記住我
Ocular neovascularization (NV) occurs in eyes with chronic ischemia.1 Ocular NV is common among patients with proliferative diabetic retinopathy (PDR), central retinal vein occlusion, or ocular ischemic syndrome.2,3 In contrast, in central retinal artery occlusion (CRAO), artery obstruction by emboli may cause irreversible damage to the adjacent tissues, and thus, reduce the demand for oxygenation.4 Therefore, ocular NV may not develop under this acute condition. Nonetheless, 10% to 20% of patients with CRAO develop ocular NV within 3 months.4–10 In light of these observations, the exact mechanism for developing NV after CRAO remains under investigation.
The prevalence of CRAO is around 1/100 000.11 Although rare, it is devastating because there is no effective treatment, and the visual prognosis is very poor. Moreover, CRAO increases the risk of developing ocular NV.12 Therefore, it is important to identify the risk factors associated with ocular NV development after CRAO. This knowledge could provide clues for preventing the occurrence of ocular NV.
Some vascular-related risk factors were previously identified in patients with CRAO, including hypertension, stroke, and diabetes mellitus (DM).13 However, little information is available on which patients are most likely to develop ocular NV after CRAO.5,9,14 Here, we aimed to identify risk factors associated with ocular NV development in patients with CRAO.
2. METHODS 2.1. Study populationThis retrospective study was approved by the Institutional Review Board at Taipei Veterans General Hospital in Taiwan. The research was conducted in accordance with the tenets of the Declaration of Helsinki. We reviewed the medical records of all patients that visited Taipei Veterans General Hospital, from January 2006 to May 2020. We identified patients diagnosed with CRAO, based on the International Classification of Diseases, 9th or 10th Revision. After this screening, all records were reviewed by two ophthalmologists (K.J.L. and Y.C.L.). Only patients that presented with acute CRAO were enrolled in this study. The acute phase of CRAO was defined as ocular signs of a cherry-red spot and retinal whitening on fundoscopy and ocular symptoms that presented within two weeks before the patient’s first visit to our outpatient clinics. Patients were excluded when they did not receive a thorough eye exam upon the first visit or when the diagnosis was not definite.
A total of 234 eyes were enrolled in this study. Among these, 109 eyes were excluded, due to a follow-up period <4 months; 25 eyes were excluded, due to concurrent artery and venous occlusions; 3 eyes were excluded because ocular NV occurred before CRAO; 3 eyes were excluded, due to severe non-proliferative diabetic retinopathy (NPDR) or PDR; 2 eyes were excluded because they lacked carotid artery Doppler ultrasound imaging data; and 5 eyes were excluded, due to severe carotid artery stenosis (stenosis >90%). Finally, 87 eyes were eligible for this study (Fig. 1). All medical charts were also reviewed to determine the occurrence and type of ocular NV, including NV of the iris (NVI), NV of the disc (NVD), and NV glaucoma. NVI was defined as a fine network of capillaries that is visible at the pupillary margin on slit lamp or capillaries extending to the iridocorneal angle on gonioscopy. NVD was defined as new vessels located at the disc or within 1 disc diameter from its margin on fundoscopy. NV glaucoma was defined as NVI with an elevated intraocular pressure (IOP) >21 mmHg. Additionally, we recorded the time interval between CRAO and ocular NV development.
Fig. 1: Flowchart of inclusion/exclusion criteria. CRAO = central retinal artery occlusion; NPDR = non-proliferative diabetic retinopathy; NVI = neovascularization of the iris; PDR = proliferative diabetic retinopathy.
2.2. Data collectionWe collected data on all systemic conditions, including hypertension, DM, chronic kidney disease (CKD) and staging, stroke, body height, body weight, and body mass index (BMI, kg/m2). We also collected data on ocular conditions recorded at the first visit, including the best-corrected visual acuity (BCVA), expressed as the logarithm of the minimum angle of resolution (LogMAR); IOP (in mmHg); laterality of the eye; the spherical equivalent, expressed in diopters (D); the lens condition; the central macular thickness (µm); the presence and severity of diabetic retinopathy (DR); glaucoma history; age; and sex. The central macular thickness was measured with Optical Coherence Tomography on an Avanti RTVue XR instrument (OptoVue, Fremont, CA). The severity of DR was graded by two ophthalmologists (K.J.L. and M.J.C.).
2.3. Statistical analysisCategorical variables are expressed as the number of patients and percentages, and continuous variables are expressed as the mean ± SD. The time interval between the CRAO presentation and the occurrence of ocular NV was evaluated with a Kaplan-Meier survival analysis and Cox proportional hazards regression. All statistical analyses were performed with SPSS software version 22.0 (SPSS, Inc., Chicago, IL, USA). p values < 0.05 were considered significant.
3. RESULTS 3.1. Baseline characteristicsAmong 87 eyes, 13 (15%) developed ocular NV at an average of 112 days (range: 5-314 days) after CRAO, during a mean follow-up period of 739 (161-3145) days. The other 74 (85%) eyes did not develop ocular NV during a mean follow-up period of 797 (181-3245) days. NVI were found in all eyes developed ocular NV. The mean age of patients that developed ocular NV was 81.54 ± 8.46 years, and the mean age of patients that did not develop ocular NV was 69.65 ± 17.68 years. The male-to-female ratio was 2:1 in both groups (Tables 1 and 2).
Table 1 - Clinical characteristics of categorical variables of the acute CRAO patients Categorical variables Value All patients CRAO with NVI CRAO without NVI Gender Male 59 (68%) 9 (69%) 50 (68%) Female 28 (32%) 4 (31%) 24 (32%) Laterality of eye OD 46 (53%) 7 (54%) 39 (53%) OS 41 (47%) 6 (46%) 35 (47%) Lens condition Phakic 62 (71%) 9 (69%) 53 (72%) Pseudophakic 19 (22%) 3 (23%) 16 (22%) Unknown 6 (7%) 1 (8%) 5 (6%) Hypertension Yes 58 (67%) 11 (85%) 47 (64%) No 29 (33%) 2 (15%) 27 (36%) DM Yes 27 (31%) 7 (54%) 20 (27%) No 60 (69%) 6 (46%) 54 (73%) DR Yes 9 (10%) 5 (38%) 4 (5%) No 78 (90%) 8 (62%) 70 (95%) Severity of DR Mild 2 (22%) 1 (20%) 1 (25%) Moderate 7 (78%) 4 (80%) 3 (75%) Stroke Yes 17 (20%) 4 (31%) 13 (18%) No 70 (80%) 9 (69%) 61 (82%) CKD Yes 5 (6%) 4 (31%) 1 (1%) No 82 (94%) 9 (69%) 73 (99%) CKD staging 1 0 (0%) 0 (0%) 0 (0%) 2 0 (0%) 0 (0%) 0 (0%) 3 1 (20%) 1 (25%) 0 (0%) 4 1 (20%) 0 (0%) 1 (100%) 5 3 (60%) 3 (75%) 0 (0%) Dialysis 3 (60%) 3 (75%) 0 (0%) Glaucoma history Yes 5 (6%) 2 (15%) 3 (4%) No 82 (94%) 11 (85%) 71 (96%)CKD = chronic kidney disease; CRAO = central retinal artery occlusion; DM = diabetes mellitus; DR = diabetic retinopathy; NVI = neovascularization of the iris.
BCVA = best-corrected visual acuity; BMI = body mass index; CRAO = central retinal artery occlusion; IOP = intraocular pressure; LogMAR = logarithm of the minimum angle of resolution; NVI = neovascularization of the iris.
At the last follow-up, the LogMAR BCVA was significantly worse in patients with ocular NV (1.89 ± 0.79) than in patients without ocular NV (1.63 ± 0.96, p < 0.001). However, the two groups did not differ significantly in other ocular conditions, including the first visit BCVA (1.98 ± 0.42 vs 2.60 ± 0.44, p = 0.83), IOP (14.58 ± 3.60 mmHg vs 13.54 ± 3.17 mmHg, p = 0.28), spherical equivalent (−2.30 ± 1.76 D vs −0.49 ± 2.83 D, p = 0.16), and central macular thickness (240.17 ± 47.78 µm vs 257.48 ± 70.09 µm, p = 0.97). There was no difference between groups in the laterality of the eye. The ratio of phakic to pseudophakic eyes was 3:1 in both groups. DR was found in 5 (38%) patients in the ocular NV group, and in 4 (5%) patients in the group without ocular NV. In addition, most patients had moderate NPDR. A glaucoma history was reported in 2 (15%) patients in the ocular NV group and in 3 (4%) patients in the group without ocular NV. In the ocular NV group, One is primary open-angle glaucoma (POAG) and another is normal-tension glaucoma (NTG). In the group without ocular NV, Two of them are POAG and another is NTG.
3.3. Systemic conditionsThe mean BMIs were not significantly different between patients with and without ocular NV (24.81 ± 4.03 kg/m2 vs 24.49 ± 3.45 kg/m2, respectively, p = 0.82). We found concurrent hypertension, DM, stroke, and CKD, respectively, in 11 (85%), 7 (54%), 4 (31%), and 4 (31%) patients in the ocular NV group and in 47 (64%), 20 (27%), 13 (18%), and 1 (1%) patients in the group without ocular NV. Among the 5 patients with CKD, 3 (75%) patients were in the ocular NV group and were classified as CKD stage 5, and all 3 were undergoing dialysis. In addition, 1 (25%) patient in the ocular NV group was classified as CKD stage 3, and 1 patient in the group without ocular NV was classified as CKD stage 4 (Tables 1 and 2).
3.4. Statistical analysisA univariate analysis showed that the group with ocular NV had significantly higher frequencies of DR (HR: 6.20, 95% CI, 2.02-19.02, p = 0.001), CKD (HR: 7.84, 95% CI, 2.40-25.64, p = 0.001), and older age (HR: 1.05, 95% CI, 1.01-1.10, p = 0.027) compared to the group without ocular NV. The multivariate analysis showed that only CKD (HR: 9.27, 95% CI, 1.87-46.05, p = 0.006, Fig. 2) and a glaucoma history (HR: 7.52, 95% CI, 1.14-49.46, p = 0.036, Fig. 3) were independent risk factors for developing ocular NV (Tables 3 and 4).
Table 3 - Categorical variables for risk factors of developing NVI following CRAO Categorical variables Univariate analysis Multivariate analysis Cox-regression p HR 95% CI Cox-regression p HR 95% CI Gender 0.971 0.98 0.30-3.12 0.595 0.66 0.15-3.03 Laterality of eye 0.810 1.14 0.38-3.40 0.627 1.37 0.39-4.80 Lens condition 0.886 0.91 0.25-3.36 0.630 0.67 0.13-3.47 Hypertension 0.152 3.01 0.67-13.61 0.260 3.99 0.36-44.40 DM 0.065 2.79 0.94-8.31 0.972 1.03 0.15-6.94 DR 0.001* 6.20 2.02-19.02 0.388 2.43 0.32-18.32 Stroke 0.260 1.97 0.61-6.40 0.170 3.02 0.62-14.67 CKD 0.001* 7.84 2.40-25.64 0.006* 9.27 1.87-46.05 Glaucoma history 0.171 2.87 0.63-12.97 0.036** 7.52 1.14-49.46CKD = chronic kidney disease; CRAO = central retinal artery occlusion; DM = diabetes mellitus; DR = diabetic retinopathy; HR = hazard ratio; NVI = neovascularization of the iris.
*p < 0.01.
**p < 0.05.
BCVA = best-corrected visual acuity; BH = body height; BMI = body mass index; BW = body weight; CRAO = central retinal artery occlusion; HR = hazard ratio; IOP = intraocular pressure; LogMAR = logarithm of the minimum angle of resolution; NVI = neovascularization of the iris.
*p < 0.05.
Fig. 2: Kaplan-Meier survival curves for NVI after CRAO, CKD as a risk factor. CKD = chronic kidney disease; CRAO = central retinal artery occlusion; NVI = neovascularization of the iris.
Fig. 3: Kaplan-Meier Survival Curves for NVI after CRAO, glaucoma history as a risk factor. CRAO = central retinal artery occlusion; NVI = neovascularization of the iris.
4. DISCUSSIONTo the best of our knowledge, this study was the first to demonstrate that a glaucoma history and CKD were risk factors for developing ocular NV after CRAO. Tripathi et al16 and Hu et al15 pointed out that vascular endothelial growth factor (VEGF) levels in the aqueous humor were increased in patients with glaucoma. This increase in VEGF was thought to be the main mechanism for inducing NV in ocular tissue.17 Therefore, patients with CRAO that had a glaucoma history may be prone to the development of ocular NV.
Increases in plasma VEGF levels were also observed in patients with stages 3 to 5 CKD.18 In our study, most patients with CKD were undergoing dialysis when they were diagnosed with ocular NV. The development of CRAO-related ocular NV in patients undergoing dialysis could be explained in two ways. First, intradialytic hypotension can cause a reduction in the ocular perfusion pressure in retinal vessels. McGuire et al19 showed that the prevalence of hemodynamic instability due to intradialytic hypotension was 20% to 30% during hemodialysis. Hence, a recurrent ischemic state during dialysis may account for ocular NV. Second, the period of hemodialysis was associated with increased IOP, reduced ocular perfusion pressure, and low mean arterial pressure.20 The increase in IOP may press on the retinal arteriole, and this vasoconstriction could lead to ischemia and the development of NV.
Jung et al9 and Mason III et al14 showed that DM was a risk factor in patients that developed ocular NV after CRAO. They proposed that the elevated VEGF levels in patients with DM could be the main contributor to ocular NV. However, NV is not uncommon among patients with severe NPDR or PDR.21 In the present study, we excluded patients with severe NPDR or PDR because, in these patients, it is difficult to identify the origin of ocular NV after CRAO. Therefore, our analyses did not identify DM or DR as a risk factor for ocular NV after CRAO.
Another common cause of ocular NV is ocular ischemic syndrome.3 Most patients with this syndrome present with ipsilateral, common, or internal carotid artery stenosis (>90% occlusion). Thus, ocular ischemic syndrome was positively correlated with severe carotid artery stenosis.22 In our study, patients with severe carotid artery stenosis (stenosis >90%) were excluded to eliminate potential confounding factors.
Previous studies have reported that the incidence of ocular NV among patients with CRAO ranged from 2.5% to 19.4% and that ocular NV developed within 2 to 13 weeks after CRAO.4–7,9,10,12,23 Our results (15% incidence and a mean interval of 16 weeks) were consistent with previous studies. However, due to the potential 13-week delay in ocular NV development after CRAO, we excluded patients diagnosed with CRAO that were followed for <4 months. Nevertheless, our follow-up period was around 2 years, which was, by far, the longest follow-up period reported in the literature.5,9,24,25
The exact mechanism underlying the development of ocular NV after CRAO remains an issue of debate. Hayreh et al4 proposed that CRAO was not causally related to ocular NV because ocular NV can originate from internal carotid artery occlusion or severe stenosis, rather than CRAO, per se. In contrast, Rudkin et al8 and Mason III et al14 reported that ocular NV could occur after CRAO in the absence of significant carotid artery stenosis. The most important factor for the development of NV was a failure to gain reperfusion and the resultant chronic retinal ischemia.9 Our results showed that 13 eyes developed ocular NV after CRAO without significant carotid artery stenosis or any other clinical features of ocular ischemia. Therefore, our results supported the findings of Rudkin and Mason III. Hayreh et al26 described that even in the absence of central retinal artery flow, the fluorescein fundus angiography still showed some slow filling of the retinal circulation in almost all his patients. These remnant retinal circulations may either be from collateral circulation via cilio-retinal capillary anastomoses within the optic nerve head or from collateral circulation via pial and intraneural anastomoses of the central retinal artery.26 In this situation, although most of cells are infarct, there are still some ischemic but vital cells. These cells may release VEGF and cause the further development of ocular NVs in patients with CRAO.
On initial presentation, the LogMAR BCVAs were similar between the groups with and without ocular NV (1.98 ± 0.42 and 1.89 ± 0.79, respectively, p = 0.83). However, after 2 years of follow-up, the final LogMAR BCVA in patients with ocular NV (2.60 ± 0.44) indicated that their vision was inferior to the vision of patients without ocular NV (1.63 ± 0.96). Although treatment varied, no treatment has been shown to treat CRAO effectively, not to mention ocular NV.13,24,25,27 It is believed that the inner retina is completely destroyed by the complete ischemia caused by CRAO.28 Therefore, the visual prognosis in CRAO is poor. Furthermore, the occurrence of ocular NV after CRAO causes catastrophic damage to retinal tissues. Consistent with those findings, we also demonstrated a worse prognosis in the ocular NV group than in the group without ocular NV.
The central macular thickness in the ocular NV group (240 ± 48 µm) was thinner than the that in the group without ocular NV (258 ± 70 µm), but the difference was not significant. In previous studies, the central macular thickness was relatively thicker shortly after CRAO, compared to our findings after CRAO.27,29 Chen et al27 demonstrated an increase in retinal thickness (299 ± 76 µm) in patients with acute CRAO (interval between initial symptoms and presentation <3 days). This thickening was due to intraretinal edema in the inner retinal layers; however, a marked reduction in fovea thickness (167 ± 30 µm) was noted in the 4th week of follow-up. Hayreh et al30 revealed that the retina was irreversibly damaged at about 240 minutes after inducing CRAO in Rhesus monkeys. The discrepancy in retinal thicknesses measured in different studies might be due to the different average intervals between onset and presentation. In our study, this interval was longer (6.9 days and 6.4 days in the groups with and without ocular NV, respectively), compared to the intervals in previous studies.
This study had the following limitations. First, the retrospective nature of this study could have introduced a selection bias. we excluded patients with characteristics such as short follow-up period, venous occlusion in conjunction with artery occlusion, ocular NV happening before CRAO, severe NPDR or PDR, high grade stenosis on carotid Doppler results to decrease the confounding factors that may hinder the true association between CRAO and subsequent ocular NV. However, these exclusion criteria may also increase the likelihood of biased results. Second, we found that the ocular NV group was somewhat older (mean age: 81.5 years) than the group without ocular NV (mean age: 69.7 years), although the difference was not clinically significant. Third, our cohort of patients was relatively small. However, due to the rarity of CRAO, the numbers of patients in our study were compatible with the numbers included in other major studies. Additionally, we applied strict criteria for enrolling patients with CRAO, which could provide more relevant results.
In conclusion, we demonstrated that a glaucoma history and CKD with dialysis were significant risk factors for developing ocular NV after CRAO. Ocular NV can occur within 4 months after CRAO. Therefore, monthly fundus examinations are highly recommended for these patients, at least for the first half year.
REFERENCES 1. Aiello LP, Avery RL, Arrigg PG, Keyt BA, Jampel HD, Shah ST, et al. Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med. 1994;331:1480–7. 2. Grant MB, Afzal A, Spoerri P, Pan H, Shaw LC, Mames RN. The role of growth factors in the pathogenesis of diabetic retinopathy. Expert Opin Investig Drugs. 2004;13:1275–93. 3. Hayreh SS. Neovascular glaucoma. Prog Retin Eye Res. 2007;26:470–85. 4. Hayreh SS, Podhajsky P. Ocular neovascularization with retinal vascular occlusion. II. Occurrence in central and branch retinal artery occlusion. Arch Ophthalmol. 1982;100:1585–96. 5. Duker JS, Sivalingam A, Brown GC, Reber R. A prospective study of acute central retinal artery obstruction. The incidence of secondary ocular neovascularization. Arch Ophthalmol. 1991;109:339–42. 6. Hayreh SS, Podhajsky PA, Zimmerman MB. Retinal artery occlusion: associated systemic and ophthalmic abnormalities. Ophthalmology. 2009;116:1928–36. 7. Sagong M, Kim J, Chang W. Intravitreal bevacizumab for the treatment of neovascular glaucoma associated with central retinal artery occlusion. Korean J Ophthalmol. 2009;23:215–8. 8. Rudkin AK, Lee AW, Chen CS. Ocular neovascularization following central retinal artery occlusion: prevalence and timing of onset. Eur J Ophthalmol. 2010;20:1042–6. 9. Jung YH, Ahn SJ, Hong JH, Park KH, Han MK, Jung C, et al. Incidence and clinical features of neovascularization of the iris following acute central retinal artery occlusion. Korean J Ophthalmol. 2016;30:352–9. 10. Degoumois A, Miocque S, Denion E. Central retinal artery occlusion without underlying chronic ocular ischemic syndrome may lead to neovascular glaucoma. J Fr Ophtalmol. 2017;40:758–62. 11. Park SJ, Choi NK, Seo KH, Park KH, Woo SJ. Nationwide incidence of clinically diagnosed central retinal artery occlusion in Korea, 2008 to 2011. Ophthalmology. 2014;121:1933–8. 12. Rudkin AK, Lee AW, Aldrich E, Miller NR, Chen CS. Clinical characteristics and outcome of current standard management of central retinal artery occlusion. Clin Exp Ophthalmol. 2010;38:496–501. 13. Varma DD, Cugati S, Lee AW, Chen CS. A review of central retinal artery occlusion: clinical presentation and management. Eye (Lond). 2013;27:688–97. 14. Mason JO 3rd, Patel SA, Feist RM, Albert MA Jr, Huisingh C, McGwin G Jr, et al. Ocular neovascularization in eyes with a central retinal artery occlusion or a branch retinal artery occlusion. Clin Ophthalmol. 2015;9:995–1000. 15. Hu DN, Ritch R, Liebmann J, Liu Y, Cheng B, Hu MS. Vascular endothelial growth factor is increased in aqueous humor of glaucomatous eyes. J Glaucoma. 2002;11:406–10. 16. Tripathi RC, Li J, Tripat
留言 (0)