Intravitreal Triamcinolone for the Management of Macular Edema Dueto Nonischemic Central Retinal Vein Occlusion

Objective  To evaluate the efficacy of intravitreal triamcinolone acetonide inthe management of persistent macular edema secondary to nonischemic centralretinal vein occlusion (CRVO).

Methods  Twenty consecutive patients were selected with a 3- to 4-month historyof nonischemic CRVO and persistent macular edema. These patients receiveda single intravitreal injection of 4 mg of triamcinolone acetonide (40 mg/mL).The follow-up period ranged from 10 to 12 months. The amount of macular edemawas assessed by the amount of retinal thickening on clinical examination usingthe Goldmann contact lens and by the area and intensity of staining on fluoresceinangiography. Treated patients were compared with a retrospectively matchedgroup of patients who were managed with observation only.

Main Outcome Measures  Changes in visual acuity and amount of macular edema were assessed inthe treated patients and compared with the observation group.

Results  The mean baseline visual acuity in the treatment group was 20/132 vs20/123 for the observation group (P = .57). After1 week, the treated group had a mean visual acuity of 20/51. At final follow-up,the treated group had a mean visual acuity of 20/37 while the observationgroup had a mean visual acuity of 20/110 (P = .001).Sixty percent of treated patients had a final visual acuity of 20/40 or bettervs only 20% in the observation group (P = .01). Fortypercent of the untreated patients had a final visual acuity worse than 20/200while none of the treated patients did (P<.001).At final follow-up, 75% of treated patients had complete resolution of macularedema on clinical examination vs only 20% of the untreated patients (P<.001). Two of the treated patients had recurrenceof macular edema at 6 months, and 3 had elevated intraocular pressure.

Conclusion  This study shows a treatment benefit from intravitreal triamcinolonein terms of visual acuity and macular edema for nonischemic CRVO.

Central retinal vein occlusion (CRVO) is one of the most common retinalvascular diseases and can be associated with significant visual morbidity.Decrease in visual acuity (VA) due to CRVO may be secondary to retinal hemorrhagesover the fovea, ischemia, or macular edema.¹ Thestandard of care currently remains limited to the management of the neovascularsequelae with panretinal photocoagulation. The Central Vein Occlusion Studyshowed that grid-pattern photocoagulation significantly reduced angiographicmacular edema but did not result in visual improvement.²

Recent reports have suggested various treatment options to improve thecirculation status of the retina after CRVO. Some of these include intravitrealtissue plasminogen activator, radial optic neurotomy, surgical recanalizationof the retinal vein, and laser-induced chorioretinal anastomosis.^3-6 Suchreports, although encouraging, are pilot studies, and larger well-controlledstudies are needed to establish the efficacy of these procedures.

Corticosteroids have been proposed for the management of macular edemadue to various retinal vascular disorders. Intravitreal triamcinolone acetonidewas investigated for persistent macular edema in diabetic retinopathy, chronicuveitis, and postsurgical cystoid macular edema.^7-10 Recently,Ip and Kumar¹¹ described 2 patients (2 eyes)and Greenberg et al¹² described 1 patient (2eyes) in whom macular edema due to CRVO was managed with intravitreal triamcinolone.The eyes with nonischemic CRVO had both reduction in macular edema and improvementin VA.

The purpose of this study is to evaluate the efficacy of intravitrealcorticosteroids in the management of persistent macular edema secondary tononischemic CRVO.

Between September 2001 and March 2002, patients with CRVO of 3 to 4months' duration were considered for enrollment in the study. All patientsreceived an eye examination that included best-corrected VA using the Snellenacuity chart, slitlamp inspection of the anterior segment, and dilated fundusexamination with indirect ophthalmoscopy and Goldmann 3-mirror contact lens.All patients had fluorescein angiography to the affected eye. Those with nonischemicCRVO and persistent macular edema were considered for admission into the study.A patient was considered to have a nonischemic CRVO if there was no afferentpupillary defect greater than 0.3 log unit, no rubeosis iridis, and capillarynonperfusion on fluorescein angiography less than 10 disc areas. Table 1 summarizes the inclusion and exclusioncriteria. The first 20 consecutive patients who met the inclusion requirementswere assigned to the treatment group.

The procedure was explained to the patients who then signed a consentform. A peribulbar injection of 1% lidocaine hydrochloride was given in theaffected eye, which was then prepared with 5% Betadine Solution (Purdue Frederick,Stamford, Conn). An intravitreal injection of 4 mg in 0.1 mL of triamcinoloneacetonide was given with a 30-gauge needle 3.5 mm posterior to the limbus.The intraocular pressure was examined immediately, and an anterior chamberparacentesis was performed if the pressure was greater than 25 mm Hg. Patientsin the treatment group were seen 24 hours after the procedure, 1 week later,and monthly thereafter. At each follow-up visit, a complete eye examinationwas carried out including best-corrected VA, slitlamp examination of the anteriorsegment, dilated fundus examination, and fluorescein angiography. The follow-upperiod ranged from 10 to 12 months. Retinal examination was done by the sameophthalmologist (Z.F.B.) using the Goldmann 3-mirror contact lens. Macularedema was assessed by the amount of retinal thickening on clinical examinationand by the area and intensity of staining on angiography.

The observation group was selected from our databases of CRVO patientsseen over the previous 2 years. We selected 20 consecutive patients who, duringtheir initial follow-up visits, met the inclusion criteria for the study.Their records were reviewed and all the pertinent data were tabulated. Patientswho did not have adequate follow-up were excluded.

The main outcome measures in this study were VA and the amount of macularedema. A decrease in macular edema was considered if there was a decreasein retinal thickness on clinical examination accompanied by a decrease inthe area or intensity of fluorescein staining. Complete resolution occurredif no retinal thickening could be identified on dilated fundus examinationwith the contact lens and if there was no leakage on angiography.

Because Snellen charts were used to measure vision for both groups,the acuities were converted to logMAR units to perform the appropriate statisticalmanipulation. The statistical tests used for data analysis were the χ² and t tests. The level of statistical significancewas fixed at P<.05 with a 95% confidence interval.

In general, both groups were well matched. The average age was 64.3years and 64.7 years in the treatment and observation groups, respectively(P = .94). Similarly, there was no statisticallysignificant difference with respect to baseline VA, sex, and all other majorbaseline characteristics. Table 2 summarizesthe baseline criteria for patients in both treatment and observation groups.

All patients in the treatment group tolerated the intravitreal injectionwell with no immediate complications. A paracentesis was performed on 18 patientsto decrease the intraocular pressure immediately after the injection. Throughoutthe entire follow-up period, no untoward effects were noted except for elevatedintraocular pressure in 3 patients at the 1-month follow-up. The pressurewas not greater than 25 mm Hg and responded well to topical 0.5% timolol maleate,treatment with which was discontinued at the 3-month follow-up.

The mean baseline best-corrected VA in the treatment group was 20/132vs 20/123 for the observation group (P = .57). At1 week from entry into the study, the treated group had a mean VA of 20/51;however, it was not possible to compare this with that of the observationgroup because this data was lacking for most patients. At 1 month, the treatedpatients improved to a mean VA of 20/36 while there was no change in the observationgroup (P<.001). The treated patients continuedto show a stable mean VAof 20/36 at 3 months, and the observation group hada mean VA of 20/96 (P = .001). At the final follow-up,the treated group had a mean VA of 20/37 while the observation group had amean VA of 20/110 (P = .001). Table 3 shows the average VA for both groups at various follow-ups.At the conclusion of the study, 12 (60%) of the 20 treated patients had aVA of 20/40 or better while only 4 (20%) of the 20 patients in the observationgroup had a final VA of 20/40 or better (P = .01).Eight (40%) of the untreated patients had a final VA worse than 20/200 whilenone of the treated patients did (P<.001). Table 4 summarizes the differences in finalVA between the 2 groups.

Macular edema was noted to decrease within 1 week of treatment. By 1month of follow-up, macular edema completely resolved in 16 (80%) of the treatedpatients (Figure 1). At the 6-monthfollow-up, 2 patients in the treatment group showed recurrence of macularedema both angiographically and clinically. One had a drop in VA from 20/20to 20/30, and we elected not to re-treat (Figure 2). This patient remained stable as such until the end ofthe study. The second patient had a drop in VA from 20/20 to 20/50. He wasre-treated, and his VA improved to 20/30 within 1 week along with completeresolution of the macular edema. He maintained that VA until the conclusionof the study. By the final follow-up, 15 (75%) of the treated patients hadcomplete resolution of macular edema vs only 4 (20%) of the untreated patients(P<.001). Sixteen (80%) of the untreated patientshad persistent macular edema at final follow-up vs only 5 (25%) in the treatedgroup.

Retinal hemorrhages and vascular retinal changes did not improve morerapidly in the treatment group. However, none of the treated patients developediris neovessels while 2 of the untreated patients needed panretinal photocoagulationfor rubeosis iridis. Statistical analysis was not possible owing to the limitedsample size.

The mechanism of action of corticosteroids for macular edema in CRVOis unclear. Triamcinolone has been shown experimentally to reduce the breakdownof the blood-retina barrier.¹³ It may act byinhibiting such factors as prostaglandins and interleukins, which are inflammatorymediators implicated in the pathogenesis of cystoid macular edema.^14,15 Elevated levels of vascular endothelialgrowth factor (VEGF) have been reported in CRVO.¹⁶ Vascularendothelial growth factor is known to play a role in retinal neovascularizationand the breakdown of the blood-retina barrier.^17,18 Invitro studies have shown that corticosteroids significantly reduce the expressionof VEGF.^19,20 Therefore, intravitrealtriamcinolone may block the expression of VEGF induced by CRVO and preventthe breakdown of the blood-retina barrier. Another possible mechanism is thattriamcinolone can interfere directly with protein kinase C, which has beenassociated with increased retinal vascular permeability.²¹

This study presents the anatomic and visual outcome of 20 patients withnonischemic CRVO treated with intravitreal triamcinolone. This data was comparedwith a retrospectively matched group of patients who were managed by observationalone. The data show a treatment benefit in terms of VA and macular edema(clinically and angiographically). Improvement was noted by the first weekafter treatment and usually remained stable for the rest of the follow-up.With the exception of elevated intraocular pressure in 3 patients, the procedurewas well tolerated and without adverse effects. However, potential complicationsof intravitreal corticosteroid injection include retinal detachment, vitreoushemorrhage, bacterial endophthalmitis, sterile endophthalmitis, and cataractformation.^8,9,22,23 Theprocedure can be further simplified by using topical anesthesia instead ofperibulbar anesthesia and by avoiding a paracentesis to control the intraocularpressure which may normalize spontaneously within a few minutes. For the purposeof the study, we wanted aggressive intraocular pressure control because thesepatients had a compromised retinal circulation, and we feared that even abrief elevation of intraocular pressure could worsen retinal ischemia andconfound our results.

Patients with ischemic CRVO were excluded from this study because benefitfrom the treatment was thought to be unlikely. Severe ischemia may have alreadycaused irreversible damage to the inner retina, and this may prevent improvementin vision even if the macular edema resolved. However, intravitreal corticosteroidsmay have a role in preventing the neovascular sequelae in ischemic CRVO. Wedid not observe any neovascularization in the treated eyes while it occurredin 2 of the control eyes, but the sample size was too small for any statisticalanalysis. By blocking the expression of VEGF, intravitreal corticosteroidscan potentially reduce the likelihood of neovascular proliferation on theretina and iris.

In conclusion, intravitreal triamcinolone may be a safe treatment formacular edema in nonischemic CRVO and can lead to rapid improvement in VAwith resolution of macular edema. Triamcinolone is readily available, inexpensive,and well tolerated in the vitreous. The main shortcoming of this study isthat the control group was not selected in a randomized prospective manner.However, both groups were well matched with respect to major criteria, especiallybaseline VA, age, and sex. Larger randomized prospective studies are necessaryto better ascertain the benefit of intravitreal corticosteroids in the managementof CRVO.

Correspondence: Ziad F. Bashshur, MD, Department of Ophthalmology, AmericanUniversity of Beirut–Medical Center, PO Box 113-6044, Beirut, Lebanon(zb00@aub.edu.lb).

Submitted for publication March 25, 2003; final revision received January6, 2004; accepted January 6, 2004.