| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Head and Neck Imaging |
1 From the Departments of Neuroradiology and Therapeutic Angiography (J.N.V., A.A., J.J.M.) and Ophthalmology (M.P., P.M., P.Y.S., A.G.), Hôpital Lariboisière, University of Paris, France; and Department of Biostatistics and Medical Informatics, University of Lyon, France (P.A.). Received February 7, 2001; revision requested March 26; final revision received August 23; accepted September 14. Address correspondence to J.N.V., Federation of Neuroradiology, Charcot Diagnostic and Interventional Neuroradiology Service, Groupe Hospitalo-Universitaire Pitié-Salpétrière, 43-87 boulevard de l’Hôpital, 75651 Paris Cedex 13, France (e-mail: jean-noel.vallee@psl.ap-hop-paris.fr).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Over a 6-year period, 11 consecutive patients presented with recent combined CRAO and CRVO (
72 hours). Urokinase (300,000 IU) was selectively perfused via the femoral artery into the ophthalmic artery for 40 minutes. Evaluation criteria were Snellen visual acuity with best correction, funduscopic results, and retinal arteriovenous transit time assessed over a mean 3.5-year follow-up. Mean vision and retinal perfusion were tested by means of repeated-measures analysis of variance. The correlation between visual improvement and retinal perfusion improvement was evaluated by means of Spearman rank correlation.
RESULTS: Substantial improvement in vision and retinal perfusion was noted in seven of the 11 patients treated. Mean vision improvement was significant (P = .009) within 24–48 hours after fibrinolysis, increased until 1 month after (P = .006), then remained stable throughout the follow-up (P > .10). Visual improvement correlated with retinal perfusion improvement during the period from before fibrinolysis to 24–48 hours after (P = .028). In all patients with improved results, retinal hemorrhages transiently increased. One patient had intravitreal hemorrhage shortly after fibrinolysis.
CONCLUSION: For this uncommon clinical entity, which typically has a poor visual outcome, these results suggest that ophthalmic arterial fibrinolysis may restore retinal perfusion, which leads to rapid substantial visual improvement in many cases of combined CRAO and CRVO, without systemic complications, but it may be responsible for intravitreal hemorrhage.
© RSNA, 2002
Index terms: Eye, abnormalities, 1724.124, 1724.1265, 1724.752, 2245.124, 2245.492 • Eye, hemorrhage, 1724.1265 • Retina, 1724.752, 2245.1265, 2245.492 • Urokinase, 1724.1265, 2245.1265
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
On the basis of our initial experience with ophthalmic arterial fibrinolysis with urokinase for recent severe CRVO, we inferred that therapy was beneficial, especially in patients with obstruction of the central retinal artery in the presence of CRVO (8). Consequently, the purpose of this study was to investigate the role of urokinase selectively perfused into the ophthalmic artery as an emergency treatment for combined CRAO and CRVO.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The diagnosis of combined CRAO and CRVO was established by two ophthalmologists (P.M., M.P., P.Y.S., or A.G.) in consensus in accordance with the following criteria: (a) a history of relatively sudden unilateral visual loss at visual acuity measurement and (b) ischemic whitening of the retinal posterior pole, diffusely narrowed irregular retinal arteries (signs characteristic of CRAO), dilated or tortuous retinal veins in four quadrants, scattered panretinal hemorrhages, and a swollen optic disk (signs characteristic of CRVO) at funduscopic examination.
Each patient initially underwent a complete ophthalmologic work-up, which included ophthalmologic history, Snellen visual acuity with best correction, applanation tonometry, ophthalmoscopy, slit-lamp biomicroscopy of the anterior and posterior segments of the eye, and fluorescein angiography.
Patient eligibility was based on visual loss less than 72 hours previously; fundus appearance typical of combined CRAO and CRVO; lack of capillary closure, as proved by the persistence of perfused capillaries at fluorescein angiography; and lack of cilioretinal arterial occlusion. Patients underwent no other treatment before fibrinolysis.
Patients with extensive retinal ischemia that involved the peripheral retina were excluded to avoid potential hemorrhagic transformation of the ischemic infarction due to sudden revascularization after thrombolysis. Pregnancy, severe blood coagulation disorders, and contraindications to urokinase were additional exclusion criteria.
All patients selected for fibrinolysis underwent routine investigation (Fig 1) for systemic diseases (9–14) and received complete information about this therapy. After we obtained their informed consent, 11 consecutive patients (five women and six men; mean age, 54 years; age range, 42–69 years) who met the inclusion criteria were referred to the interventional neuroradiology department.
|
Review of the arteriogram allowed us, by using steam, to give an S shape to the tip of the microcatheter, in accordance with the geometry of the carotid arterial siphon and ophthalmic arterial ostium (Fig 2). A 1.8- or 1.5-F flow-guide microcatheter (Magic, Balt Extrusion, Montmorency, France; or Spinnaker, Target Boston Scientific, Fremont, Calif) was advanced coaxially into the ophthalmic artery ostium through the 5-F vertebral catheter with fluoroscopic guidance. To prevent thrombosis due to the presence of catheter systems, heparin sodium (Héparine Choay; Sanofi-Synthelabo, Le Plessis-Robinson, France) was administered initially as an intravenous bolus of 30 IU/kg. Continuous pressurized flushing of both coaxial systems with 0.9% physiologic saline was performed throughout the procedure to reduce friction, to avoid intraarterial air embolism, and to eliminate clot formation due to the presence of catheter systems.
|
|
Follow-up and Assessment of Outcome
The examinations performed were (a) Snellen visual acuity measurement with best correction for distance with use of an 18-line logarithmic table; (b) funduscopic examination to evaluate hemorrhage, retinal and papillary edema, venous and arterial calibers, and the overall quality of the capillary bed, with the amount of intraretinal blood classified as small, moderate, or abundant; and (c) intravenous fluorescein angiography to estimate retinal circulation, which was evaluated with the retinal arteriovenous transit time (normal, 2 seconds). The latter was defined as the time, in seconds, between the first appearance of dye at the origin of the central retinal artery and the first laminar flow in any of the retinal veins that received peripheral blood at the edge of the disk.
Patients were examined by two ophthalmologists (P.M., M.P., P.Y.S., or A.G.) in consensus before fibrinolysis therapy, 24–48 hours after, and then during follow-up at 7, 14, 30, 90, and 180 days, at 1 year, and then at every 6 months.
Statistical Analysis
The mean evolutionary trends for vision and retinal perfusion, which were assessed at six time points, were analyzed by means of repeated-measures analysis of variance. The significance of the univariate repeated-measures F test was evaluated with Huynh-Feldt correction. Trend analyses were performed with polynomial and difference contrasts. For difference contrasts, the mean visual acuity and mean retinal arteriovenous transit time at each time point but the first were compared with the mean of the means of visual acuity and the means of the retinal arteriovenous transit time at the previous time points. All computations were performed with software (SPSS, version 10; SPSS, Chicago, Ill).
The correlations between the change score in visual acuity and that in retinal arteriovenous transit time assessed between one time point and the previous time point were evaluated by means of Spearman rank correlation.
Correlations between patients’ baseline characteristics (patient age, sex, duration of visual loss prior to fibrinolysis, retinal hemorrhages, retinal arteriovenous transit time, and visual acuity) and visual outcome and between these baseline characteristics and retinal arteriovenous transit time were evaluated by means of Spearman rank correlation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
|
|
The vision of seven of the 11 patients improved at some point following treatment, with substantial improvement in vision within 24–48 hours in six patients (Table 2). Of the six patients, vision in patients 3 and 11 had returned to normal and that in patients 4–7 showed marked improvement. The vision in patients 5–7 was normal at 1 week, 1 month, and 6 months, respectively. In patient 4, visual acuity improved from counting fingers at 3 feet to 20/40 within 24–48 hours of fibrinolysis, but there was no further improvement. Vision in patient 2 improved from counting fingers at 3 feet to 20/100 within 1 week and then to 20/25 within 1 month of fibrinolysis. Among these seven patients with improved vision, retinal arteriovenous transit time improved markedly in patients 3, 7, and 11, with normalization in patient 11 (Fig 5) 24–48 hours after fibrinolysis; returned to normal in patients 4 and 7; improved substantially in patients 3, 5, and 6 within 2 weeks; and returned to normal in patient 2 within 1 month.
|
|
|
|
|
Among the patients without improved vision, visual acuity remained unchanged in patients 8–10. Patient 1 developed an intravitreal hemorrhage 3 days after fibrinolysis, even though she was not given too much heparin; she underwent pars plana vitrectomy combined with endoretinal panretinal photocoagulation, which resulted in final vision of no light perception. Patient 10 developed cystoid macular edema at 1 month after the onset of initial visual loss, and no patients developed iris or angle neovascularization. No systemic complications due to fibrinolysis were observed.
Risk factors for thrombosis were found in eight of the 11 patients (Table 2). They included hypertension in five patients, hypercholesterolemia in two, an elevated peak of gammoglobulins at serum protein electrophoresis that revealed monoclonal gammopathy in one, a heterozygous state for the factor V Leiden that indicated resistance to activated protein C in one, a heterozygous protein S deficiency in one, and an elevated anticardiolipin antibody concentration in one.
Statistical Results
Improvement in mean visual acuity with the repeated-measures F test with Huynh-Feldt correction was significant for the period from before fibrinolysis to 1-year follow-up after fibrinolysis (P = .012) and to the mean 3.5-year final follow-up (P = .018) (Fig 6a). F test results revealed a significant linear trend over time (P = .046 at 1 year; P = .063 at 3.5 years) and a significant quadratic trend over time (P = .009 at 1 year; P = .005 at 3.5 years). According to difference contrasts, mean visual acuity improved significantly 24–48 hours after fibrinolysis (P = .009) and at 1-month follow-up (P = .006) and did not change significantly after 1-month follow-up (P > .10).
|
|
Visual acuity improvement correlated with the improvement in retinal arteriovenous transit time during the period from before fibrinolysis to 24–48-hour follow-up (Spearman rank correlation coefficient, 0.696; P = .028). Patients’ baseline characteristics did not correlate with either visual outcome or retinal perfusion outcome.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Brown et al (5) described the cases of 23 patients with unilateral combined CRAO and CRVO. Systemic abnormalities, particularly systemic arterial hypertension, were found in some of these patients. Among the 21 patients with follow-up for at least 6 months, or until the onset of neovascularization of the iris, 17 (81%) developed rubeosis iridis. Three of these 21 patients were treated with argon green laser panretinal photocoagulation prior to the onset of rubeosis iridis. Despite this treatment, two of the three treated patients developed iris neovascularization. The mean time to the onset of rubeosis iridis was 11.8 weeks, and the median time was 6.0 weeks. In many cases, the onset was fulminant. In each instance, the iris neovascularization led to neovascular glaucoma (rubeosis iridis with an intraocular pressure of 22 mm Hg or higher). The condition usually progressed to total angle closure by fibrovascular tissue within 1–2 weeks after the clinical observation of new vessels on the iris, with consequent blindness and even loss of the eye.
The clinical features of the subjects in this series define a subset that tends to develop severe complications. In comparison, in a randomly assigned series of eyes followed up for 3 years and reported by the Central Vein Occlusion Group (16,17), iris or angle neovascularization occurred in 16.4% (117 of 714) of eyes with CRVO alone, including 10.8% (58 of 538) of the initially nonischemic eyes and 34.6% (61 of 176) of the initially ischemic or indeterminate eyes. Neovascular glaucoma that was unsuccessfully managed with medical treatment developed in only 10 eyes, four of which were classified as ischemic. In a prospective study of acute CRAO alone, Duker et al (18) reported a rate of 18% (six of 33 patients) for secondary ocular neovascularization that appeared between 12 days and 15 weeks after the artery obstructions and resulted in neovascular glaucoma in 15% (five of 33 patients).
The visual prognosis of patients with combined CRAO and CRVO was uniformly poor in the 15 eyes reported by Brown et al (5) with available follow-up acuities at 6 months, except in one case, in which the prognosis improved from hand motion to 20/200. Six patients with combined CRAO and CRVO described by Richards (7) experienced permanent visual loss, and two patients developed neovascular glaucoma despite one being treated with systemic corticosteroids and the other being treated with radiation therapy. Systemic abnormalities, particularly inflammatory diseases (Behçet disease, vasculitis of unknown origin, septic cavernous sinus thrombosis, and subacute bacterial endocarditis) or myeloproliferative disorders were found in these patients. However, other reports showed improved vision in a few cases (2,6). Jorrizo et al (2) demonstrated that in the course of several months, vision returned to its premorbid state in two such cases. In one case reported by Iijima and Tsumara (6), vision improved 6 months after intravenous administration of urokinase for 3 days, starting a few hours after the onset of sudden visual loss.
The pathophysiologic process of combined CRAO and CRVO is not well understood. According to Fujino et al (19), CRVO might be the primary disorder, and impaired retinal arterial filling might be a consequence of the transfer of increased venous pressure across the capillary bed to the arterial side. The intravenous thrombus formation, histopathologically documented by Green et al (20), is accelerated by the reduction in the blood flow induced by the decrease in arterial inflow. The only collateral vessels available are the retinociliary veins in the prelaminar region, and they may not be sufficient to drain the retinal blood supply from the retinal to the choroidal veins, which would result in impairment of retinal arterial inflow (21). In addition, the suddenness of onset of CRVO may play a role in the development of a collateral blood flow (19,21). Hayreh et al (21) suggested that arterial obstruction probably occurs first, with or without arteriosclerotic changes in the central retinal arterial wall, and leads to the slowing down of the retinal circulation, which in turn leads to venous stasis and then venous thrombosis.
The restoration of arterial circulation in the presence of venous thrombosis results in the increased hemorrhages and venous dilatation typical of CRVO. This may explain why a slight transient increase of retinal hemorrhages was observed during the immediate postoperative period in all the study patients who improved. Of the 23 patients with combined CRAO and CRVO described by Brown et al (5), two initially presented with CRVO alone and subsequently experienced sudden further visual loss, as well as superficial retinal whitening and a cherry red spot. These authors were not certain whether the central retinal artery became occluded at that time or whether a sudden exacerbation of the venous obstruction alone accounted for the change. None of their patients presented initially with acute CRAO alone.
The underlying mechanism of the initial combination of CRAO and CRVO remains to be identified. Systemic abnormalities in the anticoagulation and fibrinolytic systems that may have a role in retinal vascular occlusive disease (9–14) were found in eight (73%) of our 11 patients.
Considering the high frequency of severe visual impairment in patients with combined CRAO and CRVO, as well as the rapid onset of iris neovascularization and the lack of an alternative therapy, an effective emergency treatment is urgently needed. For this purpose, we performed ophthalmic arterial fibrinolysis as an emergency treatment for this condition. Most medical therapies for CRVO only or CRAO only have focused on pharmacologic agents designed to lyse the thrombus to restore retinal blood flow and visual function (22–26). In patients without life-threatening disease (27–32), selective administration of thrombolytic agents into the ophthalmic artery allows maintenance of a high local concentration of thrombolytic activity in the vicinity of the clot, which results in fewer systemic hemorrhagic side effects than occur in intravenous thrombolysis.
In our series, the improvement in mean visual acuity was substantial within 24–48 hours of fibrinolysis, increased until 1 month after, and then remained stable throughout follow-up. Improvement in mean visual acuity correlated with improvement in mean retinal perfusion during the period from before fibrinolysis to 24–48 hours after. This suggests that partial restoration of retinal perfusion at 24–48 hours after fibrinolysis (patients with improvement, mean gain of 59.0%, range of 33.5%–100.0%; patients without improvement, mean gain of 29.5%, range of 9.0%–75.0%) led to visual improvement and that impairment of retinal perfusion was responsible for visual loss. Improvement in both mean visual acuity and mean retinal perfusion shortly after fibrinolysis corresponded to the mechanism of action of the fibrinolytic drug. This findings suggests that a thrombus may be part of the pathophysiologic process that leads to combined CRAO and CRVO, a process that is still not clear. The histopathologic findings reported by Brown et al (5) in two eyes months after the diagnosis of combined CRAO and CRVO failed to demonstrate that the obstructions were located in either the central retinal artery or central retinal vein. The findings showed only the late sequelae of the probable vascular occlusions. The extensive inner ischemic retinal atrophy seen in both eyes corroborated laboratory test findings of concomitant obstruction of the central retinal vein and artery (33).
In one patient, the occurrence of intravitreal hemorrhage shortly after fibrinolysis resulted in final vision of no light perception. This must be considered a side effect, although the perfused dose of urokinase was low, and there was no capillary closure. The hemorrhage may have been caused by the rupture of ischemic retinal vessels as a result of severely impaired retinal arterial perfusion (retinal arteriovenous transit time, >50 seconds), as well as by the combined sudden increase in retinal venous hydrostatic pressure caused by venous thrombosis as a result of the release by fibrinolysis of the arterial obstruction. This adverse event occurred in a patient whose extensive systemic evaluation disclosed a mild but previously unknown minor von Willebrand syndrome.
No systemic bleeding complications due to fibrinolysis were observed. In accordance with the use of a low total dose of thrombolytic agent, the laboratory data demonstrated the absence of changes in systemic thrombolytic activity due to the procedure.
In conclusion, the results suggest that ophthalmic arterial fibrinolysis may restore retinal perfusion, which leads to rapid substantial visual improvement in many cases of combined CRAO and CRVO, without systemic complications, but it may be responsible for intravitreal hemorrhage. However, these results may be considered encouraging for this uncommon clinical entity, which typically has a poor visual outcome.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, J.N.V., J.J.M., A.G., M.P., P.M., P.Y.S.; study concepts and design, J.N.V., J.J.M., A.G., M.P., P.M., P.Y.S.; literature research, all authors; clinical studies, all authors; data acquisition, M.P., P.M., P.Y.S., A.G.; data analysis/interpretation, all authors; statistical analysis, J.N.V., P.A.; manuscript preparation, J.N.V.; manuscript definition of intellectual content, J.N.V., J.J.M., A.G., M.P.; manuscript editing, J.N.V.; manuscript revision/review and final version approval, J.N.V., M.P., A.G., J.J.M.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H Shahid, P Hossain, and W M Amoaku The management of retinal vein occlusion: is interventional ophthalmology the way forward? Br. J. Ophthalmol., May 1, 2006; 90(5): 627 - 639. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
