CLINICAL MANIFESTATIONS OF CENTRAL RETINAL ARTERY OCCLUSION IN EYES OF PROLIFERATIVE DIABETIC RETINOPATHY WITH PREVIOUS VITRECTOMY AND PANRETINAL PHOTOCOAGULATION SAN-NI CHEN, MD,*†‡ CHEN-CHENG CHAO, MD,* JIUNN-FENG HWANG, MD,*† CHUNG-MAY YANG, MD§¶ Purpose: To report the clinical characteristics of central retinal artery occlusion in eyes of proliferative diabetic retinopathy with previous vitrectomy and panretinal photocoagulation. Methods: Retrospective case series. Results: Twelve eyes in 12 patients (4 women and 8 men) with a mean age of 55.3 ± 6.2 years of age were included in this study. All patients had successful previous surgery for complications of proliferative diabetic retinopathy and complete panretinal photocoagulation performed. All patients had sudden visual deterioration and fluorescein angiography confirmed central retinal artery occlusion. Prominent cherry-red spot was noted in only two eyes. The other eyes either had less prominent or incomplete, or invisible cherry-red spots. Sequential optical coherence tomography in 7 of the 12 eyes showed increased optical reflectivity at inner retinal layer in 6 of the 7 eyes. Rubeosis iridis developed later in six eyes and neovascular glaucoma in four eyes. Vision worsening was noted in five eyes, improved in three eyes, and stable in four eyes after an averaged follow-up of 4 months. Conclusion: Less typical cherry-red spots, higher possibility of rubeosis despite previous complete panretinal photocoagulation, are the characteristic features of central retinal artery occlusion in patients with proliferative diabetic retinopathy with previous vitrectomy. RETINA 34:1861–1866, 2014
C
include hypertension, carotid atherosclerosis, structural cardiac pathology, coronary heart disease, cerebral vascular accident, and diabetes mellitus.1–3 Patients who underwent vitrectomy for complications of proliferative diabetic retinopathy (PDR) are often patients with more severe degree of retinopathy and vasculopathy, which may render them even more susceptible to vasoocclusive disease and different clinical pictures may be presented. The aim of this study is to investigate the clinical features of CRAO in patients with PDR with previous vitrectomy and panretinal photocoagulation.
entral retinal artery occlusion (CRAO) is a devastatingly vision-threatening disease. The sudden onset of visual loss, retinal cherry-red spot, multiple cotton-wool spots or whitening around the fovea, retinal swelling, arterial constriction with box-caring of blood flow, with or without visible arterial emboli are the typical clinical presentation. Risk factors of CRAO
From the *Department of Ophthalmology, Changhua Christian Hospital, Changhua City, Taiwan; †School of Medicine, Chung-Shan Medical University, Taichung, Taiwan; ‡School of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan; §Department of Ophthalmology, National Taiwan University Hospital, Taipei, Taiwan; and ¶School of Medicine, National Taiwan University, Taipei, Taiwan. None of the authors have any financial/conflicting interests to disclose. Reprint requests: Chung-May Yang, MD, Department of Ophthalmology, National Taiwan University Hospital, No.7, Chung Shan South Road, Taipei, Taiwan; e-mail: [email protected]
Materials and Methods From January 2009 to July2012, a retrospective case series study of CRAO was performed in eyes with 1861
1862 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
PDR that had previous vitrectomy and panretinal photocoagulation in Changhua Christian Hospital and National Taiwan University Hospital. Inclusion criteria were: patients who had undergone successful vitrectomy and panretinal photocoagulation for complications of PDR, such as tractional retinal detachment, tractional maculopathy, or persistent vitreous hemorrhage; the sudden onset of visual loss in the operated eye; and delayed arterial filling for .30 seconds by fluorescein angiography (FA). Patients with previous neovascular glaucoma (NVG) and clinical follow-up for ,4 months were excluded. Visual acuity, slit-lamp biomicroscopoy, funduscopy, and intraocular pressure (IOP) measurement were performed at each visit. Patients were followed up at outpatient clinic for at least 4 months. The development of rubeosis or NVG was specifically looked for by slit-lamp (Slit Lamp BM, Hagg-Streit Company, Koeniz, Switzerland) and gonioscopy contact lens. Patients’ age, gender, medical history, visual acuity at initial and later stages, and the duration from seeing a doctor for the problem of possible CRAO to the definite diagnosis of CRAO, were recorded. All patients had carotid artery Doppler performed after the diagnosis of CRAO. Operative procedures including the use of silicone oil, gas tamponade, and scleral buckling were documented. Postoperative complications including high IOP (.30 mmHg) for .1 week or vitreous hemorrhage were also recorded. Fundus features including the presence of cherry-red spot, retinal swelling, arterial attenuation, vascular sheathing, and arterial emboli were recorded. Findings of optical coherence tomography (OCT) (Cirrus OCT; Zeiss Meditec, Inc, Dublin, CA) including optical reflectivity of the inner retina and macular thickness (the average thickness of the central macular area with a diameter of 1.0 mm) were recorded at the onset and follow-up. All patients had antivascular endothelial growth factor agent injection intravitreally preoperatively. Only one patient (Case 7, the only case from National Taiwan University Hospital) had antivascular endothelial growth factor agent injection at the end of surgery. No cases received any intravitreal injection of antivascular endothelial growth factor during the follow-up period before the onset of CRAO.
Results A total of 12 eyes in 12 patients were recruited in this study, including 4 women and 8 men. The average age was 55.3 ± 6.2 years. The average follow-up was 8 ± 6.4 (4–35 months). Systemic hypertension was noted in seven patients. Chronic renal disease
2014 VOLUME 34 NUMBER 9
was noted in two patients and one of them was on hemodialysis (Case 8). Carotid artery Doppler showed 54% stenosis of the internal carotid artery on the side of CRAO in 1 patient, 20% to 35% stenosis in 2 cases (Cases 2 and 6), ,20% stenosis in 4 cases, and normal condition in 5 cases. No previous cerebral vascular accident had been noticed in any case before the episode of CRAO; however, 1 case (Case 9) had a cerebral vascular accident attack 1 year after the CRAO episode. Surgical procedures including vitrectomy, membrane delamination, and panretinal photocoagulation were performed in all eyes, gas tamponade in seven eyes, and silicone oil tamponade in one eye. Elevated IOP .30 mmHg for .1 week was noted in 5 eyes. All the 5 eyes had IOP returned to ,21 mmHg for .3 months before the onset of CRAO. No postoperative rubeosis was noted in any eye before the onset of CRAO. The duration from vitrectomy to the onset of CRAO ranged from 1 month to 17 months (averaged duration, 8.4 months). Decimal bestcorrected visual acuity before the onset of CRAO ranged from 0.05 to 0.4. Best-corrected visual acuity at the onset of CRAO ranged from 0.05 to light perception. The duration from acute visual loss to visiting our clinic ranged from 2 days to 1 week. Delayed arterial filling .30 seconds in FA was noted in all eyes. A diagnosis of CRAO was established immediately after patients’ visiting in eight cases; four eyes did not have the diagnosis of CRAO confirmed until FA was performed. The duration of reaching the definite diagnosis after initial ophthalmologic examinations for patients’ visual loss ranged from immediacy to 1 week. Prominent cherry-red spot was noted in only 2 eyes (Figure 1). Three eyes had less prominent retinal whitening and cherry-red spots (Figure 2), 3 eyes had incomplete cherry-red spots (Figure 3), and 4 had no visible cherry-red spots (Figure 4). Diffuse arterial sheathing with silver wire appearance was present in 9 eyes (Figures 1–4). Delayed retinal arterial fluorescein filling for .30 seconds was noted in all eyes. Delayed choroidal filling .20 seconds was noted in 4 eyes (Cases 3, 4, 6, and 9). In the 4 eyes with delayed choroidal filling, the lag from choroidal to retinal filling was all .10 seconds. Arterial emboli were not observed in any eye. In the seven eyes in which OCT had been done at the onset of symptom, increased optical reflectivity at the inner retina was noted in 6 eyes (Figure 3C). The averaged macular thickness was 246.50 ± 50 mm in the infarction eye and 264.83 ± 4.83 mm in the fellow eye at the diagnosis of CRAO (P = 0.533, paired t-test). Two eyes had thicker macula as compared with the fellow eye (Cases 2 and 11). Two eyes had asymmetrical increased reflectivity and thickening at the inner retina
CRAO IN PDR WITH PREVIOUS VITRECTOMY CHEN ET AL
1863
of CRAO ranged from 1 week to 6 weeks. After intravitreal injection of bevacizumab, supplementary laser, and antiglaucomatous agents, IOP was under control in two of the four eyes with NVG. Ciliary body ablation with cyclocryotherapy or transscleral cyclophotocoagulation was necessary in two eyes (Cases 7 and 9). Demographic data and characteristics of patients were listed in Table 1.
Discussion
Fig. 1. Color fundus in Case 7 showed a prominent cherry-red spot and diffuse arterial sheathing.
(Cases 10 and 12). Percutaneous transluminal thrombolysis was performed in two eyes (Cases 1 and 5). Visual improvement after thrombolysis was observed in one eye (Case 5). The final visual acuity deteriorated in five eyes, improved in three eyes, and was stable in four eyes. Rubeosis iridis developed in 6 eyes and NVG was noted in 4 eyes (Cases 7, 8, 9, and 10). The onset of rubeosis and glaucoma after the diagnosis
Fig. 2. A. Color fundus in Case 1 showed a faint cherry-red spot and silver wire appearance of retinal arteries. B. Fluorescein angiography showed marked delayed filling of artery at 47 seconds after injection of dye.
Central retinal artery occlusion is a vision-threatening disease with poor prognosis. Patients with CRAO typically present with the sudden onset painless visual loss. Cherry-red spot, retinal swelling, retinal arterial attenuation, and delayed arterial filling in FA are typical signs. Cherry-red spot is an important clue for diagnosing CRAO in most of the situations. In patients without chronic ischemic vascular disease, the inner retina at the posterior pole becomes swollen and opacified after acute infarction, except in the region of the foveola, where a cherry-red spot is present.4 In our series, we found prominent cherry-red spot was present in only two eyes. We postulated that in this particular group of patients, there was preexisting atrophic inner retina induced by the chronic ischemic changes, long-term macular edema, or macular grid laser before the onset of CRAO. The fluctuated IOP during and after surgery may further compromise the inner retina. Because cherry-red spot simply manifests swelling of the ganglion cells and nerve fibers around the foveloa; if those are destroyed, cherry-red spot is not likely to develop. This scenario could be well illustrated in our Case 10 where the fundus showed an incomplete cherry-red spot with only retinal opacification at the upper macula at the onset of CRAO. Optical coherence tomography of the same patient 1 month before the onset of CRAO already revealed retinal thinning and atrophic changes at the lower macula. Thus, the OCT at acute stage of CRAO in this case showed prominently increased thickness and reflectivity only at the upper part, and mild changes at the lower half of the macula (Figure 3). Arterial attenuation is another important sign in CRAO.4 Unfortunately, arterial attenuation is widely present in patients with PDR after panretinal photocoagulation because the oxygen demand is much lowered in those treated eyes. This change would make arterial attenuation a less reliable sign for detecting CRAO. Furthermore, diffuse silver wire appearance of arteries, which is not uncommon in end-stage PDR with burned out retina, was present in 9 of the 12 eyes in our series. This preexisting alteration makes CRAO-associated arterial attenuation difficult to be
1864 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2014 VOLUME 34 NUMBER 9
Fig. 3. A. Color fundus in Case 10. The cherry-red spot is incomplete in shape, with more intense whitening at the upper perifoveal area. Retinal artery showed diffuse silver wire appearance. B. Fluorescein angiography showed almost no filling of dye at 48 seconds after injection of dye. C. Optical coherence tomography on the upper taken 1 month before the onset of CRAO showed thicker upper macula and thinner lower macula with atrophic inner retina. Optical coherence tomography on the lower showed increased reflectivity and swelling of the inner retina at the upper macula and only mild changes of reflectivity and no changes of thickness of the lower, thinner atrophic macula at the acute stage of CRAO.
observed, beside, the silver wire changes probably will render the arterial lumen less elastic and smaller, which would make emboli easier to be trapped and CRAO more likely to happen. Arterial emboli, which is present in 20% to 40% of eyes with CRAO, is also an important sign observed in retinal arterial occlusion.1,4,5 However, none of our patients had emboli observed. This might be explained by the following reasons. First, the lumen of central retinal artery in this group of patients may be narrower, which makes the emboli more likely to be stuck in the optic nerve, before reaching to the optic nerve head to be visible. Second, the sheathing of vessels makes emboli less likely to be seen through the arterial wall. Third, may be, a higher percentage of those patients have thrombosis instead of emboli as the primary cause because those patients may have narrower lumen of artery, which is more vulnerable to thrombosis. And fourth, those eyes may have underlying ocular ischemic syndrome, with very low central arterial perfusion pressure, which is readily to be closed by normal IOP without any emboli.
Increased macular thickness compared with the fellow eye observed in OCT is another common phenomenon in other patients at the acute stage of CRAO.6,7 In our 7 cases with OCT available at the acute stage, however, the averaged central macular thickness is about the same in the studied eyes as in their fellow eyes. Two factors may be contributory: first, the preexisting atrophic inner retina before the CRAO episode in the lesion eye, and second, the frequent presence of macular edema or epiretinal membrane in the fellow eye, which make the macula in the fellow eyes thicker than normal. However, increased optical reflectivity from the inner retina in OCT, another important finding in eyes with CRAO,6,7 was observed in six of the seven eyes with OCT available on the acute stage in our series. This change seemed to be a more reliable indicator for CRAO in our patients. Rubeosis iridis was present in 6 and NVG in 4 of the 12 eyes despite previous complete panretinal photocoagulation. Rubeosis iridis is not a common complication in CRAO.3 However, in the report of Vander et al,8
Fig. 4. A. Color fundus in Case 6 taken 1 week after the onset of CRAO showed no cherry-red spot and silver wire appearance of retinal arteries. B. Fluorescein angiography showed almost no filling of retinal arteries at 40 seconds after injection of dye. C. Optical coherence tomography (upper) taken 5 months before the onset of CRAO showed an atrophic inner retina with residual subretinal fluid at the upper macular area. Optical coherence tomography (lower) taken 1 week after the onset of CRAO revealed neither increased reflectivity nor swelling of the inner retina.
CAA, carotid artery atherosclerosis; CF, counting finger; CRF, chronic renal failure; F, female; HM, hand motion; HT, hypertension; LP, light perception; NLP, no light perception; M, male; VA, visual acuity.
Yes/Yes No/No Yes/No 1 week 2 days 4 days 184/225 257/237 216/223 Yes Yes Yes Yes No Yes Incomplete Incomplete Incomplete 10 12 7 HM 0.7 LP HM 0.6 LP
VT, Gas VT VT, Gas
No/No No/No No/No No/No Yes/Yes Yes/Yes 207/294 — — — — — 150/221 Yes No No Yes No Yes No Less 1 1.25 7 5 5 15 14 LP 0.05 0.1 HM NLP CF NLP 0.01 HM CF HM 0.02 HM HM
VT, SO VT, Gas VT VT, Gas VT, Gas VT, Gas VT
NLP HM
VT
16
Less
Yes Yes No Yes Yes Yes
— — — — — No prominent
— — — — — 1 week Yes
356/261 Yes No prominent
268/294 Yes Yes prominent Less 3.5 VT, Gas CF 0.05
0.4 No/No 0.2 Yes/No 0.01 0.05 0.2 0.05 0.3 0.05 0.05 Yes/Yes 0.2 0.7 0.1
Final Surgical VA Procedures No./ VA Before Age/Sex CRAO
1/46/F 2 days 2/59/M 3 days 3/60/M 4/53/M 5/46/M 6/58/M 7/46/F 8/50/F 9/71/F 3 days 10/56/M 11/55/M 12/54/M
Increased Optical Reflectivity of Cherry-red Arterial Spot Sheathing Inner Retina Duration From Vitrectomy, months VA at Onset of CRAO
Table 1. Demographic data of patients
Macular Thickness (1.0 mm): Lesion/ Fellow Eye
Duration From Onset of CRAO Rubeosis/ to First OCT NVG
CRAO IN PDR WITH PREVIOUS VITRECTOMY CHEN ET AL
1865
rubeosis iridis developed in four of five eyes with PDR that subsequently developed CRAO despite previous panretinal photocoagulation. This observation indicates that CRAO may superimpose on the preexisting ocular ischemia in those eyes,3 because rubeosis is the hallmark of ocular ischemia syndrome, and diabetes mellitus is a major risk factor for ocular ischemic syndrome.9 This possibility could be partially supported by the fact that delayed choroidal filling (.20 seconds) was noted in 4 of the 12 eyes and 1 patient had cerebral vascular accident attack 1 year after CRAO. Though none of our cases demonstrated severe stenosis of internal carotid artery by Doppler examination, this examination result cannot rule out the possibility of ocular ischemia because the general circulation is always compromised in PDR, and there are limitations in the ability of carotid artery Doppler in detecting the stenosis distal to the neck, especially the ophthalmic artery. Beside, the previous panretinal photocoagulation, hemodynamic changes during vitrectomy, fluctuated IOP in the postoperative course, and other damages brought by surgical procedures may further damage the choroidal circulation locally, with subsequent choroidal ischemia, which cannot be detected by carotid artery Doppler or by FA. It is possible that ocular ischemia was already present in most of our cases, which contribute to the onset of CRAO in some of our cases, and in those eyes developing rubeosis, CRAO may just be a precipitating factor beyond ocular ischemia. In conclusion, it takes more caution to make an early diagnosis of CRAO in eyes with PDR with previous vitrectomy. Cherry-red spot is not a reliable sign in the diagnosis of CRAO. It can be absent or not detectable ophthalmoscopically because of preexisting marked damage of the inner retina or resolving of cherry-red spots that may start soon, and varies markedly from eye to eye. Arterial attenuation and presence of emboli also cannot be served as indicators for diagnosis. OCT, though may help sometimes by demonstrating increased optical reflectivity at the inner retina, is also of not much value in cases with preexisting diffuse inner retinal atrophy. Delayed filling time in FA, thus, is still the most reliable confirming sign. Because ophthalmoscopic signs are not obvious in this group of patients, clinicians should be more alert with this diagnosis when patients have complaints of sudden visual loss and emergent FA should be arranged. Beside, rubeosis and NVG are more frequently noted in this group of patients, close follow-up is necessary for timely intervention. Key words: CRAO, vitrectomy, OCT, rubeosis irides, fluorescein angiography.
1866 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
References 1. Sharma S, Grown GC, Cruess AF, et al, RCEO. The accuracy of visible retinal emboli for the detection of cardio-embolic lesions requiring anticoagulation or cardiac surgery. Br J Ophthalmol 1998;82:655–658. 2. Sharma S, Naqvi A, Sharma SM, et al, RCEO. Transthoracic echocardiographic findings in patients with acute retinal arterial obstruction. Arch Ophthalmol 1996;114:1189–1192. 3. Hayreh SS, Podhajsky PA, Zimmerman MB. Retinal artery occlusion: associated systemic and ophthalmic abnormalities. Ophthalmology 2009;116:1928–1936. 4. Hayreh SS, Zimmerman MB. Fundus changes in central retinal artery occlusion. Retina 2007;27:276–289.
2014 VOLUME 34 NUMBER 9
5. Brown GC, Magargal LE. Central retinal artery obstruction and visual acuity. Ophthalmology 1982;89:14–19. 6. Chen SN, Hwang JF, Chen YT. Macular thickness measurements in central retinal artery occlusion by optical coherence tomography. Retina 2011;31:730–737. 7. Ozdemir H, Karacorlu S, Karacorlu M. Optical coherence tomography findings in central retinal artery occlusion. Retina 2006;26:110–112. 8. Vander JF, Brown GC, Benson WE. Iris neovascularization after central retinal artery obstruction despite previous panretinal photocoagulation for diabetic retinopathy. Am J Ophthalmol 1990;109:464–468. 9. Mizener JB, Podhajsky P, Hayreh SS. Ocular ischemic syndrome. Ophthalmology 1997;104:859–864.
C
include hypertension, carotid atherosclerosis, structural cardiac pathology, coronary heart disease, cerebral vascular accident, and diabetes mellitus.1–3 Patients who underwent vitrectomy for complications of proliferative diabetic retinopathy (PDR) are often patients with more severe degree of retinopathy and vasculopathy, which may render them even more susceptible to vasoocclusive disease and different clinical pictures may be presented. The aim of this study is to investigate the clinical features of CRAO in patients with PDR with previous vitrectomy and panretinal photocoagulation.
entral retinal artery occlusion (CRAO) is a devastatingly vision-threatening disease. The sudden onset of visual loss, retinal cherry-red spot, multiple cotton-wool spots or whitening around the fovea, retinal swelling, arterial constriction with box-caring of blood flow, with or without visible arterial emboli are the typical clinical presentation. Risk factors of CRAO
From the *Department of Ophthalmology, Changhua Christian Hospital, Changhua City, Taiwan; †School of Medicine, Chung-Shan Medical University, Taichung, Taiwan; ‡School of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan; §Department of Ophthalmology, National Taiwan University Hospital, Taipei, Taiwan; and ¶School of Medicine, National Taiwan University, Taipei, Taiwan. None of the authors have any financial/conflicting interests to disclose. Reprint requests: Chung-May Yang, MD, Department of Ophthalmology, National Taiwan University Hospital, No.7, Chung Shan South Road, Taipei, Taiwan; e-mail: [email protected]
Materials and Methods From January 2009 to July2012, a retrospective case series study of CRAO was performed in eyes with 1861
1862 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
PDR that had previous vitrectomy and panretinal photocoagulation in Changhua Christian Hospital and National Taiwan University Hospital. Inclusion criteria were: patients who had undergone successful vitrectomy and panretinal photocoagulation for complications of PDR, such as tractional retinal detachment, tractional maculopathy, or persistent vitreous hemorrhage; the sudden onset of visual loss in the operated eye; and delayed arterial filling for .30 seconds by fluorescein angiography (FA). Patients with previous neovascular glaucoma (NVG) and clinical follow-up for ,4 months were excluded. Visual acuity, slit-lamp biomicroscopoy, funduscopy, and intraocular pressure (IOP) measurement were performed at each visit. Patients were followed up at outpatient clinic for at least 4 months. The development of rubeosis or NVG was specifically looked for by slit-lamp (Slit Lamp BM, Hagg-Streit Company, Koeniz, Switzerland) and gonioscopy contact lens. Patients’ age, gender, medical history, visual acuity at initial and later stages, and the duration from seeing a doctor for the problem of possible CRAO to the definite diagnosis of CRAO, were recorded. All patients had carotid artery Doppler performed after the diagnosis of CRAO. Operative procedures including the use of silicone oil, gas tamponade, and scleral buckling were documented. Postoperative complications including high IOP (.30 mmHg) for .1 week or vitreous hemorrhage were also recorded. Fundus features including the presence of cherry-red spot, retinal swelling, arterial attenuation, vascular sheathing, and arterial emboli were recorded. Findings of optical coherence tomography (OCT) (Cirrus OCT; Zeiss Meditec, Inc, Dublin, CA) including optical reflectivity of the inner retina and macular thickness (the average thickness of the central macular area with a diameter of 1.0 mm) were recorded at the onset and follow-up. All patients had antivascular endothelial growth factor agent injection intravitreally preoperatively. Only one patient (Case 7, the only case from National Taiwan University Hospital) had antivascular endothelial growth factor agent injection at the end of surgery. No cases received any intravitreal injection of antivascular endothelial growth factor during the follow-up period before the onset of CRAO.
Results A total of 12 eyes in 12 patients were recruited in this study, including 4 women and 8 men. The average age was 55.3 ± 6.2 years. The average follow-up was 8 ± 6.4 (4–35 months). Systemic hypertension was noted in seven patients. Chronic renal disease
2014 VOLUME 34 NUMBER 9
was noted in two patients and one of them was on hemodialysis (Case 8). Carotid artery Doppler showed 54% stenosis of the internal carotid artery on the side of CRAO in 1 patient, 20% to 35% stenosis in 2 cases (Cases 2 and 6), ,20% stenosis in 4 cases, and normal condition in 5 cases. No previous cerebral vascular accident had been noticed in any case before the episode of CRAO; however, 1 case (Case 9) had a cerebral vascular accident attack 1 year after the CRAO episode. Surgical procedures including vitrectomy, membrane delamination, and panretinal photocoagulation were performed in all eyes, gas tamponade in seven eyes, and silicone oil tamponade in one eye. Elevated IOP .30 mmHg for .1 week was noted in 5 eyes. All the 5 eyes had IOP returned to ,21 mmHg for .3 months before the onset of CRAO. No postoperative rubeosis was noted in any eye before the onset of CRAO. The duration from vitrectomy to the onset of CRAO ranged from 1 month to 17 months (averaged duration, 8.4 months). Decimal bestcorrected visual acuity before the onset of CRAO ranged from 0.05 to 0.4. Best-corrected visual acuity at the onset of CRAO ranged from 0.05 to light perception. The duration from acute visual loss to visiting our clinic ranged from 2 days to 1 week. Delayed arterial filling .30 seconds in FA was noted in all eyes. A diagnosis of CRAO was established immediately after patients’ visiting in eight cases; four eyes did not have the diagnosis of CRAO confirmed until FA was performed. The duration of reaching the definite diagnosis after initial ophthalmologic examinations for patients’ visual loss ranged from immediacy to 1 week. Prominent cherry-red spot was noted in only 2 eyes (Figure 1). Three eyes had less prominent retinal whitening and cherry-red spots (Figure 2), 3 eyes had incomplete cherry-red spots (Figure 3), and 4 had no visible cherry-red spots (Figure 4). Diffuse arterial sheathing with silver wire appearance was present in 9 eyes (Figures 1–4). Delayed retinal arterial fluorescein filling for .30 seconds was noted in all eyes. Delayed choroidal filling .20 seconds was noted in 4 eyes (Cases 3, 4, 6, and 9). In the 4 eyes with delayed choroidal filling, the lag from choroidal to retinal filling was all .10 seconds. Arterial emboli were not observed in any eye. In the seven eyes in which OCT had been done at the onset of symptom, increased optical reflectivity at the inner retina was noted in 6 eyes (Figure 3C). The averaged macular thickness was 246.50 ± 50 mm in the infarction eye and 264.83 ± 4.83 mm in the fellow eye at the diagnosis of CRAO (P = 0.533, paired t-test). Two eyes had thicker macula as compared with the fellow eye (Cases 2 and 11). Two eyes had asymmetrical increased reflectivity and thickening at the inner retina
CRAO IN PDR WITH PREVIOUS VITRECTOMY CHEN ET AL
1863
of CRAO ranged from 1 week to 6 weeks. After intravitreal injection of bevacizumab, supplementary laser, and antiglaucomatous agents, IOP was under control in two of the four eyes with NVG. Ciliary body ablation with cyclocryotherapy or transscleral cyclophotocoagulation was necessary in two eyes (Cases 7 and 9). Demographic data and characteristics of patients were listed in Table 1.
Discussion
Fig. 1. Color fundus in Case 7 showed a prominent cherry-red spot and diffuse arterial sheathing.
(Cases 10 and 12). Percutaneous transluminal thrombolysis was performed in two eyes (Cases 1 and 5). Visual improvement after thrombolysis was observed in one eye (Case 5). The final visual acuity deteriorated in five eyes, improved in three eyes, and was stable in four eyes. Rubeosis iridis developed in 6 eyes and NVG was noted in 4 eyes (Cases 7, 8, 9, and 10). The onset of rubeosis and glaucoma after the diagnosis
Fig. 2. A. Color fundus in Case 1 showed a faint cherry-red spot and silver wire appearance of retinal arteries. B. Fluorescein angiography showed marked delayed filling of artery at 47 seconds after injection of dye.
Central retinal artery occlusion is a vision-threatening disease with poor prognosis. Patients with CRAO typically present with the sudden onset painless visual loss. Cherry-red spot, retinal swelling, retinal arterial attenuation, and delayed arterial filling in FA are typical signs. Cherry-red spot is an important clue for diagnosing CRAO in most of the situations. In patients without chronic ischemic vascular disease, the inner retina at the posterior pole becomes swollen and opacified after acute infarction, except in the region of the foveola, where a cherry-red spot is present.4 In our series, we found prominent cherry-red spot was present in only two eyes. We postulated that in this particular group of patients, there was preexisting atrophic inner retina induced by the chronic ischemic changes, long-term macular edema, or macular grid laser before the onset of CRAO. The fluctuated IOP during and after surgery may further compromise the inner retina. Because cherry-red spot simply manifests swelling of the ganglion cells and nerve fibers around the foveloa; if those are destroyed, cherry-red spot is not likely to develop. This scenario could be well illustrated in our Case 10 where the fundus showed an incomplete cherry-red spot with only retinal opacification at the upper macula at the onset of CRAO. Optical coherence tomography of the same patient 1 month before the onset of CRAO already revealed retinal thinning and atrophic changes at the lower macula. Thus, the OCT at acute stage of CRAO in this case showed prominently increased thickness and reflectivity only at the upper part, and mild changes at the lower half of the macula (Figure 3). Arterial attenuation is another important sign in CRAO.4 Unfortunately, arterial attenuation is widely present in patients with PDR after panretinal photocoagulation because the oxygen demand is much lowered in those treated eyes. This change would make arterial attenuation a less reliable sign for detecting CRAO. Furthermore, diffuse silver wire appearance of arteries, which is not uncommon in end-stage PDR with burned out retina, was present in 9 of the 12 eyes in our series. This preexisting alteration makes CRAO-associated arterial attenuation difficult to be
1864 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2014 VOLUME 34 NUMBER 9
Fig. 3. A. Color fundus in Case 10. The cherry-red spot is incomplete in shape, with more intense whitening at the upper perifoveal area. Retinal artery showed diffuse silver wire appearance. B. Fluorescein angiography showed almost no filling of dye at 48 seconds after injection of dye. C. Optical coherence tomography on the upper taken 1 month before the onset of CRAO showed thicker upper macula and thinner lower macula with atrophic inner retina. Optical coherence tomography on the lower showed increased reflectivity and swelling of the inner retina at the upper macula and only mild changes of reflectivity and no changes of thickness of the lower, thinner atrophic macula at the acute stage of CRAO.
observed, beside, the silver wire changes probably will render the arterial lumen less elastic and smaller, which would make emboli easier to be trapped and CRAO more likely to happen. Arterial emboli, which is present in 20% to 40% of eyes with CRAO, is also an important sign observed in retinal arterial occlusion.1,4,5 However, none of our patients had emboli observed. This might be explained by the following reasons. First, the lumen of central retinal artery in this group of patients may be narrower, which makes the emboli more likely to be stuck in the optic nerve, before reaching to the optic nerve head to be visible. Second, the sheathing of vessels makes emboli less likely to be seen through the arterial wall. Third, may be, a higher percentage of those patients have thrombosis instead of emboli as the primary cause because those patients may have narrower lumen of artery, which is more vulnerable to thrombosis. And fourth, those eyes may have underlying ocular ischemic syndrome, with very low central arterial perfusion pressure, which is readily to be closed by normal IOP without any emboli.
Increased macular thickness compared with the fellow eye observed in OCT is another common phenomenon in other patients at the acute stage of CRAO.6,7 In our 7 cases with OCT available at the acute stage, however, the averaged central macular thickness is about the same in the studied eyes as in their fellow eyes. Two factors may be contributory: first, the preexisting atrophic inner retina before the CRAO episode in the lesion eye, and second, the frequent presence of macular edema or epiretinal membrane in the fellow eye, which make the macula in the fellow eyes thicker than normal. However, increased optical reflectivity from the inner retina in OCT, another important finding in eyes with CRAO,6,7 was observed in six of the seven eyes with OCT available on the acute stage in our series. This change seemed to be a more reliable indicator for CRAO in our patients. Rubeosis iridis was present in 6 and NVG in 4 of the 12 eyes despite previous complete panretinal photocoagulation. Rubeosis iridis is not a common complication in CRAO.3 However, in the report of Vander et al,8
Fig. 4. A. Color fundus in Case 6 taken 1 week after the onset of CRAO showed no cherry-red spot and silver wire appearance of retinal arteries. B. Fluorescein angiography showed almost no filling of retinal arteries at 40 seconds after injection of dye. C. Optical coherence tomography (upper) taken 5 months before the onset of CRAO showed an atrophic inner retina with residual subretinal fluid at the upper macular area. Optical coherence tomography (lower) taken 1 week after the onset of CRAO revealed neither increased reflectivity nor swelling of the inner retina.
CAA, carotid artery atherosclerosis; CF, counting finger; CRF, chronic renal failure; F, female; HM, hand motion; HT, hypertension; LP, light perception; NLP, no light perception; M, male; VA, visual acuity.
Yes/Yes No/No Yes/No 1 week 2 days 4 days 184/225 257/237 216/223 Yes Yes Yes Yes No Yes Incomplete Incomplete Incomplete 10 12 7 HM 0.7 LP HM 0.6 LP
VT, Gas VT VT, Gas
No/No No/No No/No No/No Yes/Yes Yes/Yes 207/294 — — — — — 150/221 Yes No No Yes No Yes No Less 1 1.25 7 5 5 15 14 LP 0.05 0.1 HM NLP CF NLP 0.01 HM CF HM 0.02 HM HM
VT, SO VT, Gas VT VT, Gas VT, Gas VT, Gas VT
NLP HM
VT
16
Less
Yes Yes No Yes Yes Yes
— — — — — No prominent
— — — — — 1 week Yes
356/261 Yes No prominent
268/294 Yes Yes prominent Less 3.5 VT, Gas CF 0.05
0.4 No/No 0.2 Yes/No 0.01 0.05 0.2 0.05 0.3 0.05 0.05 Yes/Yes 0.2 0.7 0.1
Final Surgical VA Procedures No./ VA Before Age/Sex CRAO
1/46/F 2 days 2/59/M 3 days 3/60/M 4/53/M 5/46/M 6/58/M 7/46/F 8/50/F 9/71/F 3 days 10/56/M 11/55/M 12/54/M
Increased Optical Reflectivity of Cherry-red Arterial Spot Sheathing Inner Retina Duration From Vitrectomy, months VA at Onset of CRAO
Table 1. Demographic data of patients
Macular Thickness (1.0 mm): Lesion/ Fellow Eye
Duration From Onset of CRAO Rubeosis/ to First OCT NVG
CRAO IN PDR WITH PREVIOUS VITRECTOMY CHEN ET AL
1865
rubeosis iridis developed in four of five eyes with PDR that subsequently developed CRAO despite previous panretinal photocoagulation. This observation indicates that CRAO may superimpose on the preexisting ocular ischemia in those eyes,3 because rubeosis is the hallmark of ocular ischemia syndrome, and diabetes mellitus is a major risk factor for ocular ischemic syndrome.9 This possibility could be partially supported by the fact that delayed choroidal filling (.20 seconds) was noted in 4 of the 12 eyes and 1 patient had cerebral vascular accident attack 1 year after CRAO. Though none of our cases demonstrated severe stenosis of internal carotid artery by Doppler examination, this examination result cannot rule out the possibility of ocular ischemia because the general circulation is always compromised in PDR, and there are limitations in the ability of carotid artery Doppler in detecting the stenosis distal to the neck, especially the ophthalmic artery. Beside, the previous panretinal photocoagulation, hemodynamic changes during vitrectomy, fluctuated IOP in the postoperative course, and other damages brought by surgical procedures may further damage the choroidal circulation locally, with subsequent choroidal ischemia, which cannot be detected by carotid artery Doppler or by FA. It is possible that ocular ischemia was already present in most of our cases, which contribute to the onset of CRAO in some of our cases, and in those eyes developing rubeosis, CRAO may just be a precipitating factor beyond ocular ischemia. In conclusion, it takes more caution to make an early diagnosis of CRAO in eyes with PDR with previous vitrectomy. Cherry-red spot is not a reliable sign in the diagnosis of CRAO. It can be absent or not detectable ophthalmoscopically because of preexisting marked damage of the inner retina or resolving of cherry-red spots that may start soon, and varies markedly from eye to eye. Arterial attenuation and presence of emboli also cannot be served as indicators for diagnosis. OCT, though may help sometimes by demonstrating increased optical reflectivity at the inner retina, is also of not much value in cases with preexisting diffuse inner retinal atrophy. Delayed filling time in FA, thus, is still the most reliable confirming sign. Because ophthalmoscopic signs are not obvious in this group of patients, clinicians should be more alert with this diagnosis when patients have complaints of sudden visual loss and emergent FA should be arranged. Beside, rubeosis and NVG are more frequently noted in this group of patients, close follow-up is necessary for timely intervention. Key words: CRAO, vitrectomy, OCT, rubeosis irides, fluorescein angiography.
1866 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
References 1. Sharma S, Grown GC, Cruess AF, et al, RCEO. The accuracy of visible retinal emboli for the detection of cardio-embolic lesions requiring anticoagulation or cardiac surgery. Br J Ophthalmol 1998;82:655–658. 2. Sharma S, Naqvi A, Sharma SM, et al, RCEO. Transthoracic echocardiographic findings in patients with acute retinal arterial obstruction. Arch Ophthalmol 1996;114:1189–1192. 3. Hayreh SS, Podhajsky PA, Zimmerman MB. Retinal artery occlusion: associated systemic and ophthalmic abnormalities. Ophthalmology 2009;116:1928–1936. 4. Hayreh SS, Zimmerman MB. Fundus changes in central retinal artery occlusion. Retina 2007;27:276–289.
2014 VOLUME 34 NUMBER 9
5. Brown GC, Magargal LE. Central retinal artery obstruction and visual acuity. Ophthalmology 1982;89:14–19. 6. Chen SN, Hwang JF, Chen YT. Macular thickness measurements in central retinal artery occlusion by optical coherence tomography. Retina 2011;31:730–737. 7. Ozdemir H, Karacorlu S, Karacorlu M. Optical coherence tomography findings in central retinal artery occlusion. Retina 2006;26:110–112. 8. Vander JF, Brown GC, Benson WE. Iris neovascularization after central retinal artery obstruction despite previous panretinal photocoagulation for diabetic retinopathy. Am J Ophthalmol 1990;109:464–468. 9. Mizener JB, Podhajsky P, Hayreh SS. Ocular ischemic syndrome. Ophthalmology 1997;104:859–864.
