●

ieo●

●

ccsahTipidw●

t7iirm3●

pmcdmCfiemr

A

UW

OJ

0d

Viral Retinitis After Intravitreal Triamcinolone Injectionin Patients With Predisposing Medical Comorbidities

ANKUR M. SHAH, STEPHEN F. OSTER, AND WILLIAM R. FREEMAN

VtesvcifilttfG(1avtveidsbiitatrtasnnta

ioA(ge

PURPOSE: To review the cases of viral retinitis afterntravitreal steroid administration at a single center, tostimate the incidence, and to propose risk factors for itsccurrence.DESIGN: Retrospective, observational case series.METHODS: Seven hundred thirty-six intravitreal triam-

inolone (IVTA) injections were administered in thelinic and operating room by 3 retina specialists at aingle academic medical center between September 2002nd November 2008. Inclusion criteria were simply aistory of 1 or more IVTA injections during the period.he overall incidence of viral retinitis after IVTA

njection was calculated. Subsequently, a chart audit waserformed to estimate the number of patients withmmune-altering conditions who had received IVTAuring the period, and the incidence within this subgroupas calculated.RESULTS: Viral retinitis developed after IVTA injec-

ion in 3 patients, yielding an overall incidence of 3 in36 or 0.41%. An estimated 334 injections were admin-

stered to patients with an immune-altering condition,ncluding diabetes. All 3 of the patients in whom viraletinitis developed after IVTA injection possessed abnor-al immune systems, yielding an incidence rate of 3 in34 or 0.90% within this subgroup.CONCLUSIONS: Our high reported incidence for this

otentially devastating complication can be attributed toultiple factors, including coexisting medical immuno-

ompromising comorbidities, a higher dose with a longeruration of local immunosuppression in the vitreous,ultiple injections, as well as previous viral retinitis.aution with a high index of clinical suspicion and

requent follow-up is advised in patients receiving IVTAnjection with potentially immune-altering conditions,ven after apparent immune recovery. (Am J Ophthal-ol 2010;149:433–440. © 2010 by Elsevier Inc. All

ights reserved.)

ccepted for publication Oct 15, 2009.From the Jacobs Retina Center, Department of Ophthalmology,niversity of California, San Diego, La Jolla, California (A.M.S., S.F.O.,.R.F.).Inquiries to William R. Freeman, Jacobs Retina Center, Department of

rphthalmology, Shiley Eye Center 0946, 9415 Campus Point Drive, La

olla, CA 92093; e-mail: freeman@eyecenter.ucsd.edu

© 2010 BY ELSEVIER INC. A002-9394/10/$36.00oi:10.1016/j.ajo.2009.10.019

IRAL RETINITIS IS A DEVASTATING CONDITION

with disastrous consequences. Depending on theetiologic agent and immune status of the victim,

he infection often can progress rapidly to involve thentire retina, including the macula, ultimately leading toevere loss of vision. Cytomegalovirus (CMV) was the firstiral retinitis to be described and remains the mostommon etiologic agent. It generally affects severelymmunocompromised patients, such as human immunode-ciency virus (HIV)-positive individuals with CD4 T-ymphocyte counts of less than 50 cells/mm3 or, rarely,hose with iatrogenic immunosuppression after chemo-herapy, organ transplantation, or systemic corticosteroidsor other indications.1–5 In their Focal Points module,oldstein and associate indicate that acute retinal necrosis

ARN) was initially reported by Urayama and associates in971, and is most common in young immunocompetentdults.6 Freeman and associates demonstrated herpes groupirus in endoretinal biopsy specimens in the acute phase ofhe disease, with varicella zoster virus or herpes simplexirus being the usual culprits.7 Patients left untreated canxperience contralateral eye involvement within 1 monthn up to one third of cases. More recently, a highlyestructive, rapidly progressive variant of ARN that occursolely in severely immunocompromised individuals haseen described. Progressive outer retinal necrosis syndromes caused by the same viruses as ARN, and leads to rapidnvolvement of the central and peripheral retina by full-hickness necrosis, ultimately leading to severe vision losss a result of retinal detachment or optic atrophy.6 Al-hough the virulence and rate of destruction of viraletinitis can vary, prompt, appropriate treatment is the keyo a favorable visual outcome. However, the therapeuticgents used often are associated with frequent and severeystemic side effects, namely bone marrow suppression andephrotoxicity. Patients, especially those who are immu-ocompromised, often require lifelong prophylaxis withhese toxic agents to avoid recurrent infections in the samend contralateral eye.

In recent years, case reports have surfaced identifyingmmunocompetent patients in whom viral retinitis devel-ped after the administration of intravitreal steroids.8–14

s the popularity of intravitreal triamcinolone acetonideIVTA) injection and the indications for its use haverown over the years, deleterious side effects such aslevated intraocular pressure, posterior subcapsular cata-

acts, vitreous hemorrhage, and endophthalmitis have

LL RIGHTS RESERVED. 433

bshhcavasawdmsaUnpavTmrpsif

hteptctp

oIi

W

ffNtdceiotncpdtss

A

o3I0at

4

ecome well known to ophthalmologists, and intravitrealteroids often are avoided in patients deemed to be atigher risk of these complications.15–19 Viral retinitis,owever, has received far less attention as a potentialonsequence of the medication. In 2005, Saidel andssociates first described CMV retinitis (diagnosed byitreous fluid polymerase chain reaction [PCR]) 4 monthsfter IVTA injection in a patient treated for clinicallyignificant macular edema resulting from diabetes.8 Delyfernd associates reported 2 immunocompetent patients inhom CMV retinitis developed after IVTA use for exu-ative age-related macular degeneration in one case andacular edema resulting from a central retinal vein occlu-

ion in the second case.9 The cases were diagnosed vianterior chamber paracentesis and PCR. A recent paper byfret-Vincenty and associates presented a patient diag-osed with CMV retinitis (by clinical appearance) afterlacement of a fluocinolone acetonide (Retisert; Bauschnd Lomb Inc., Rochester, New York, USA) implant foritreitis and macular edema secondary to Behçet disease.10

he patient had not received systemic immunosuppressiveedication for 5 months before the development of

etinitis. Other articles have described ARN resulting fromresumed herpes simplex virus (on the basis of positiveerum antibody results) after IVTA injection in previouslymmunocompetent patients, although PCR was not per-ormed on aqueous or vitreous fluid (Table 1).11,12

Although reports of viral retinitis after IVTA injectionave begun to surface in the literature, no information onhe incidence of this potentially severe complicationxists. Furthermore, minimal discussion of risk factorsredisposing individuals to viral retinitis after IVTA injec-ion has been presented. Our purpose was to review theases of viral retinitis after intravitreal steroid administra-ion at a single center, to estimate the incidence, and toropose risk factors for its occurrence. Following are 3 cases

TABLE 1. Published Cases of Viral Reti

Authors Dose (mg/0.1 mL) Indication for IV

Saidel and associates8 4 CSME

Toh and associates12 4 Exudative AMD (w

Aggermann and associates11 4 CME resulting from

Delyfer and associates9 20 Exudative AMD (w

Delyfer and associates9 8 � 3 CME resulting from

Ufret-Vincenty and

associates10

Retisert � 2 CME/vitreitis resul

Behçet disease

Sekiryu and associates14 4 CME resulting from

Park and associates13 4 CME resulting from

AC � anterior chamber; AMD � age-related macular degeneration

CMV � cytomegalovirus; CRVO � central retinal vein occlusion; CS

IVTA � intravitreal triamcinolone; NIDDM � non–insulin-depend

photodynamic therapy.

AMERICAN JOURNAL OF34

f primary or reactivated viral retinitis that developed afterVTA injection from our institution in patients withmmune-altering medical comorbidities.

METHODS

E RETROSPECTIVELY REVIEWED THE CHARTS OF PATIENTS

rom our institution who had received IVTA injectionsrom 3 retina specialists from September 2002 throughovember 2008. No patients received IVTA at our insti-

ution before 2002. All cases were included, regardless ofiagnosis code or indication for the treatment. The mostommon diagnoses were cystoid macular edema, maculardema, and diabetic macular edema. Settings for thenjections included the clinic and operating room. Theverall incidence of this complication was calculated ashe number of viral retinitis cases divided by the totalumber of IVTA injections administered. Subsequently, ahart audit was performed to estimate the number ofatients with altered immunity who had received IVTAuring the period. Responsible conditions included diabe-es mellitus (DM), HIV or acquired immunodeficiencyyndrome, or use of long-term oral steroids or steroid-paring immunomodulatory or chemotherapeutic agents.

RESULTS

TOTAL OF 736 IVTA INJECTIONS WERE ADMINISTERED

ver this period. Viral retinitis subsequently developed inpatients. The overall incidence of viral retinitis after

VTA injection at our institution was 3 in 736 patients, or.41%. Of the 736 total injections, an estimated 334 weredministered to patients with an immune-altering condi-ion, including diabetes. All 3 of the patients at our center

fter Intravitreal Triamcinolone Injection

Cause

Diagnosis

Confirmation Interval to Retinitis

Medical

Comorbidities

CMV Vitreous PCR 4 mos NIDDM

) HSV Serum IgG

and IgM

5 mos None

O HSV Serum IgG 3 wks None

) CMV AC PCR 3 mos NIDDM

O CMV AC PCR 3 mos NIDDM

om CMV Clinical 5 mos after Retisert

no. 2

Behçet disease

O CMV AC PCR 7 mos NIDDM

O CMV AC PCR 4 mos None

O � branch retinal vein occlusion; CME � cystoid macular edema;

clinically significant macular edema; HSV � herpes simplex virus;

diabetes mellitus; PCR � polymerase chain reaction; PDT �

nitis a

TA

/ PDT

CRV

/ PDT

CRV

ting fr

BRV

CRV

; BRV

ME �

ent

OPHTHALMOLOGY MARCH 2010

ipdsTmw1eomumw

ibgtasphrysnwnefltChcr

CatclnvITacac

mgrs

TA

BLE

2.C

hara

cter

istic

sof

Pat

ient

sin

Who

mV

iralR

etin

itis

Dev

elop

edaf

ter

Intr

avitr

ealT

riam

cino

lone

Inje

ctio

n

Age

(yrs

)

Med

ical

Com

orb

iditi

es

Ind

icat

ion

for

IVTA

No.

of

Inje

ctio

ns

Dos

e

(mg/

0.1

mL)

Pre

viou

s

Vitr

ecto

my

Inte

rval

to

Ret

initi

s(w

ks)

Cau

seD

iagn

osis

Con

firm

atio

n

Prio

r

Ret

initi

s

VA

bef

ore

IVTA

Inje

ctio

n

VA

afte

r

Ret

initi

s

62N

IDD

MC

ME

from

BR

VO

220

Yes

26C

MV

Vitr

eous

PC

RN

one

20/2

0020

/400

43H

IVC

ME

from

IRU

120

No

13C

MV

Prio

rC

MV

retin

itis,

clin

ical

app

eara

nce

CM

V20

/40

20/4

0

73M

etas

tatic

ovar

ian

canc

er

rece

ivin

gch

emot

hera

py

IRF

from

ER

M1

20Y

es10

VZ

VR

esp

onse

totr

eatm

ent

Non

e20

/200

20/4

00

BR

VO

�b

ranc

hre

tinal

vein

occl

usio

n;C

ME

�cy

stoi

dm

acul

ared

ema;

CM

V�

cyto

meg

alov

irus;

ER

M�

epire

tinal

mem

bra

ne;H

IV�

hum

anim

mun

odefi

cien

cyvi

rus;

IRF

�in

trar

etin

alflu

id;

IRU

�im

mun

ere

cove

ryuv

eitis

;IV

TA�

intr

avitr

ealt

riam

cino

lone

;NID

DM

�no

n–in

sulin

-dep

end

entd

iab

etes

mel

litus

;PC

R�

pol

ymer

ase

chai

nre

actio

n;V

A�

visu

alac

uity

;VZ

V�

varic

ella

zost

er

viru

s;w

ks�

wee

ks;

yrs

�ye

ars.

VOL. 149,

n whom viral retinitis developed after IVTA injectionossessed abnormal immune systems. The calculated inci-ence of viral retinitis after IVTA injection among thisubgroup at our institution was 3 in 334 patients, or 0.90%.able 2 describes the clinical data for the 3 patients. Theean age of the patients in whom viral retinitis developedas 59.3 years (range, 43 to 73 years). Two were male andwas female. Indications for IVTA injection were macular

dema from various causes. The patients received either 1r 2 injections, and the dose of IVTA administered was 20g/0.1 mL in all 3 cases. Two patients previously had

ndergone vitrectomy. The mean interval to the develop-ent of retinitis after the final IVTA injection was 16eeks (range, 10 to 26 weeks).Patient 1 was a diabetic male who received IVTA

njection for recurrent macular edema secondary to aranch retinal vein occlusion that had been refractory torid laser and anti–vascular endothelial growth factorherapy. Six months after a second IVTA injection during

vitrectomy with epiretinal membrane peel, the patientought treatment for 3� anterior chamber cell, keraticrecipitates, vitreitis, and foci of retinal whitening andemorrhage in the supertemporal and inferotemporal pe-iphery (Figure 1). Although initial vitreous biopsy resultsielded negative cultures and PCR for CMV, herpesimplex virus, and varicella zoster virus, a clinical diag-osis of ARN was made and valacyclovir 1 g 3 times dailyas initiated. However, progressive worsening of the reti-itis led to retinal detachment requiring vitrectomy withndolaser and silicone oil. Repeat PCR analysis of vitreousuid at this time produced positive results for CMV, andherapy was switched to valganciclovir 900 mg twice daily.D4 T-lymphocyte count was 307 cells/�L, with noistory of immunodeficiency other than diabetes. Valgan-iclovir was continued for 6 more weeks, at which time alletinitis lesions were healed.

Patient 2 was an HIV-positive man with a history ofD4 counts of less than 50 cells/�L and CMV retinitis,

fter which highly active antiretroviral therapy was insti-uted. Within months, his CD4 count had risen to 325ells/�L and macular edema and vitreitis had developed,eading to a diagnosis of immune recovery uveitis. Becauseo CMV lesions were present, the maintenance dose ofalganciclovir 450 mg twice daily was discontinued andVTA injection was administered for the macular edema.he patient returned 3 months later with 3� vitreous cellnd apparent reactivation of CMV retinitis despite a CD4ount of 417 cells/�L (Figure 2). Valganciclovir was begunt treatment dose of 900 mg twice daily, and therapyurrently is ongoing.

Patient 3 had visual acuity of 20/200 and an epiretinalembrane and had received IVTA injection during sur-

ery at the end of a vitrectomy and epiretinal membraneemoval. Ten weeks after the IVTA injection, the patientought treatment for 2� cell in the anterior chamber and

new retinitis suspicious for ARN (Figure 3). Medical

VIRAL RETINITIS AFTER IVTA INJECTIONNO. 3 435

cwwvrmtcwi

W

rit

igasoDaiuh1Tipips

Frria

4

omorbidities included metastatic ovarian cancer, forhich the patient had undergone resection and radiationith active chemotherapy. Also of note was a history ofaricella zoster virus keratouveitis in June 2005, withecurrent anterior uveitis in October 2006. Foscarnet 2.4g/0.1 mL was given intravitreally and valacyclovir 1 g 3

imes daily was initiated. Within 2 months, the anteriorhamber was quiet, retinitis had resolved, and valacycloviras decreased to a maintenance dose to be continued

ndefinitely (Figure 3).

DISCUSSION

E REPORT A SINGLE-CENTER SERIES OF CASES OF VIRAL

etinitis after IVTA injection and an estimate of thencidence of this complication. The incidence of 0.90% in

IGURE 1. Patient 1 with active cytomegalovirus (CMV) retetinitis and hemorrhage in the supertemporal periphery 26 weeight) The view of the macula is obstructed because of vitnferotemporal periphery. (Bottom left) Wide-angle view of thesmaller focus supertemporally.

he immune-altered subset is more than double our overall t

AMERICAN JOURNAL OF36

ncidence of 0.41%. These incidences may seem highiven the small number of case reports published thus farnd likely are higher than the experience of most retinapecialists who use IVTA in their clinical practices, manyf whom may not be familiar with this complication. Theiabetic Retinopathy Clinical Research Network evalu-

ted the safety and efficacy of 2 doses of preservative-freentravitreal triamcinolone versus focal or grid photocoag-lation for the treatment of diabetic macular edema.20 Fiveundred ten eyes received a total of 1649 IVTA injections in-mg and 4-mg doses. No cases of viral retinitis were reported.he higher incidence at our institution likely represents an

ncreased susceptibility to this complication in our patientopulation. As previously described, all 3 patients had coex-sting systemic immune-altering conditions. One was HIVositive, one was receiving systemic chemotherapy for meta-tatic ovarian cancer, and another was diabetic. Furthermore,

. (Top left) The patient’s fundus is shown, with active CMVter a second intravitreal triamcinolone (IVTA) injection. (Top. Active CMV retinitis is visible in the supertemporal anddus showing a large area of inferotemporal CMV retinitis and

initisks afreitis

fun

he HIV-positive individual had previous viral retinitis during

OPHTHALMOLOGY MARCH 2010

pcdp

iHtagttar

smclacwaid2pC

Frwo

Ft(

V

eriods when his CD4 T-lymphocyte count was less than 50ells/�L and experienced reactivation after IVTA injectionespite improved CD4 counts and concomitant antiviralrophylaxis.

Reactivation of CMV retinitis after IVTA injection formmune recovery uveitis macular edema developed in theIV-positive individual. Immune recovery uveitis is rela-

ively common in patients with CMV retinitis, with Jabsnd associates reporting a prevalence of 15.5%.21 Ourroup previously described the effective use of IVTA toreat immune recovery uveitis macular edema in 7 pa-ients, none of which experienced CMV reactivation over

minimum follow-up of 9 months.22 However, viral

IGURE 2. Patient 2 with reactivated cytomegalovirus (CMV)etinitis inferior to the optic nerve before intravitreal triamcinoloeeks after IVTA injection, the fundus shows multiple areas original healed CMV scar is visible inferior to the optic nerve.

IGURE 3. Patient 3 with triamcinolone and acute retinal necriamcinolone is visible 10 weeks after intravitreal triamcinolonRight) Healed ARN with chorioretinal scarring remains 10 w

etinitides are known to remain latent after infection, and F

VIRAL RETINITIS AFTEROL. 149, NO. 3

ubsequent local immunosuppression with IVTA injectionay allow these dormant organisms to proliferate and

ause reinfection despite presumed immune recovery. Da-essandro and Bottaro reported a case of CMV reactivationfter posterior sub-Tenon triamcinolone for immune-re-overy uveitis.23 Our group previously reported a patientith acquired immunodeficiency syndrome in whom newlyctive CMV retinitis developed 19 months after thenitiation of highly active antiretroviral therapy, despite aocumented median CD4 count of 204/�L (range, 125 to84/�L) during the treatment period.24 In this particularatient, lymphoproliferative assays revealed that theMV-specific T-cell response was found to be negligible.

itis. (Left) The patient’s fundus shows an area of healed CMVIVTA) injection and subsequent reactivation. (Right) Thirteeninal whitening involving the posterior pole and periphery. The

(ARN) resulting from varicella zoster virus. (Left) Residualection, along with inferior active retinitis resulting from ARN.later, after treatment.

retinne (f ret

rosise injeeks

urthermore, Jacobson and associates described active

IVTA INJECTION 437

Ctrrhowtdaap

sfios1fihop

cDtlmreivoritdpowlsrcptmmatI

ir

RrpmarirpmiciadacgctpJosidtasopmsIrtoiorbdwsdisahtrori

4

MV retinitis in patients with CD4 cell counts of morehan 195/�L during the early stages of response to antiret-oviral therapy.25 In addition, there have been previouseports of reactivation of CMV retinitis in patients despiteigh CD4 cell counts.26–28 However, our group alsobserved 4 acquired immunodeficiency syndrome patientsho were nonresponders to highly active antiretroviral

herapy, yet remained free of CMV retinitis recurrenceespite CD4 cell counts of less than 50/�L and lack ofnti-CMV therapy.24 These patients illustrate that anrbitrary cutoff value cannot be established to predictrotection from CMV retinitis.The propensity for intraocular steroids to cause infection

econdary to local immunosuppression is supported by datarom a retrospective, multicenter case series of 922 IVTAnjections that reported an endophthalmitis incidence ratef 0.87%.29 In contrast, a retrospective, multicenter, con-ecutive case series of 12 585 intravitreal bevacizumab and4 320 ranibizumab injections from 4 clinical centersound an endophthalmitis rate of only 0.02%.30 A highndex of suspicion may be warranted in patients withistories of viral retinitis before IVTA injection, regardlessf apparent immune recovery or appropriate antiviralrophylaxis.The role of diabetes mellitus as an immune-altering

ondition often may be overlooked in clinical practice.iabetics are frequent victims of macular edema because of

heir damaged vascular endothelium and increased capil-ary permeability. The use of IVTA to treat diabeticacular edema has grown in popularity considerably in

ecent years, especially in patients with diffuse maculardema for which grid or focal laser therapy is impractical orneffective. Similar to one of our diabetic patients in whomiral retinitis developed after IVTA injection with nother immunocompromising conditions, many of the caseeports of viral retinitis after IVTA injection in supposedlymmunocompetent patients occurred in diabetics.8,9,14 Pa-ient 1 in our study had been diagnosed with non–insulin-ependent diabetes mellitus only months before his initialresentation with a branch retinal vein occlusion. Becausef the very recent diagnosis, no hemoglobin A1c valuesere available, but the patient’s morning fasting glucose

evels varied between the mid 100s and low 200s (mea-ured in mg/dL) during subsequent follow-up visits in theetina clinic. These values not only suggested poorlyontrolled diabetes, but also raised the possibility that theatient maintained high blood glucose levels for an ex-ended period before an official diagnosis of diabetes. Thisight have contributed to previously undiagnosed diabeticicroangiopathy that predisposed the patient not only tovascular occlusive event, but also subsequently to persis-

ent macular edema and vulnerability to viral retinitis afterVTA injection.

Another possible risk factor for viral retinitis after IVTAnjection could be the dose given. Whereas most case

eports of this complication, as well as the Diabetic v

AMERICAN JOURNAL OF38

etinopathy Clinical Research Network study previouslyeferenced, used a dose of 1 mg/0.1 mL or 4 mg/0.1 mL, thehysicians at our institution most frequently administer 20g decanted triamcinolone because of its presumed longer

nd stronger effect, potentially leading to fewer cumulativeeinjections over time. Simple decantation allows thenjection of a higher dose in a small volume after theemoval of a significant portion of the benzyl alcoholreservative as verified by high-performance liquid chro-atography.31 Our group directly compared the effect on

ntraocular pressure of the standard 4-mg versus the de-anted 20-mg IVTA and did not find a higher risk ofntraocular pressure elevation with the higher dose.32 Jonasnd associates similarly used a high dose of 20 to 25 mg andid not report a higher incidence of ocular adverse effectsfter their injections.15 Furthermore, they reported noases of viral retinitis after high-dose IVTA injection. Thisroup has described measurable concentrations of triam-inolone after 25-mg IVTA injection up to 1.5 years afterhe injection in nonvitrectomized eyes, despite the disap-earance of clinically visible crystals after 9 months.33

onas also reported measurable concentrations of triamcin-lone in silicone oil up to 8 months after vitrectomy withilicone oil instillation and simultaneous 25-mg IVTAnjection.34 Our group reported that the median time toisappearance of triamcinolone after 20-mg injection inhe vitrectomized eye was 113 days.35 However, Beer andssociates examined the pharmacokinetics of IVTA of thetandard 4-mg dose and found a mean elimination half-lifef 18.6 days in nonvitrectomized patients, suggestingersistent measurable concentrations for approximately 3onths.36 The half-life in vitrectomized patients was much

horter, only 3.2 days. It is apparent that a higher dose ofVTA remains in the vitreous for a much longer period,egardless of the vitrectomy status of the patient. However,he benefit of a longer duration of action of the higher dosef IVTA may come with the risk of prolonged localmmunosuppression and susceptibility to primary infectionr reactivation of endogenous latent viral retinitis. Viraletinitis developed in Patient 1 in our series sometimeetween 2 and 6 months after 20 mg IVTA injection,espite a previous vitrectomy. Although the viral retinitisas not detected until an examination 6 months after the

econd IVTA injection, we suspect that the retinitiseveloped shortly after a visit 2 months after the secondnjection. The combination of retinitis involving theupertemporal and inferotemporal periphery, along withlready decreased visual acuity from macular edema, mayave allowed the patient to remain asymptomatic, even athe time that the retinitis finally was detected. Viraletinitis developed in the other 2 patients within 3 monthsf IVTA injection. Multiple IVTA injection and theesulting prolonged immunosuppression also might havencreased Patient 1’s vulnerability to the development of

iral retinitis.

OPHTHALMOLOGY MARCH 2010

crcmavr

rwfteoI

SuFUsco

1

1

1

V

In conclusion, we presented a single-center series ofases of viral retinitis after IVTA injection. Our higheported incidence for this potentially devastating compli-ation can be attributed to multiple factors. Coexistingedical immune-altering comorbidities, a higher dose withlonger duration of local immunosuppression in the

itreous, multiple injections, as well as previous viral

etinitis, alone or in combination, may account for the h

382.

1

1

1

1

1

1

1

2

2

2

2

2

VIRAL RETINITIS AFTEROL. 149, NO. 3

elatively high incidence obtained in our series. Cautionith a high index of clinical suspicion and frequent

ollow-up is advised in patients receiving IVTA injec-ion with potentially immunocompromising conditions,ven after apparent immune recovery. Furthermore,ngoing antiviral prophylaxis both before and afterVTA injection may be prudent in patients with a

istory of viral retinitis.

UPPORTED BY GRANT EY07366 FROM THE NATIONAL INSTITUTES OF HEALTH, BETHESDA, MARYLAND (W.R.F.); ANnrestricted grant from Research to Prevent Blindness, Inc, New York, New York (University of California, San Diego); and a Heed Ophthalmicellowship, Heed Foundation, Cleveland Clinic, Cleveland, Ohio (S.F.O.). The authors indicate no financial conflict of interest related to this study.nrelated financial disclosures include the following: William R. Freeman is a consultant to Opko, Miami, Florida and to OD OS and has received grant

upport from Genentech, Geneneron, and Sirion. Involved in design of study (A.M.S., S.F.O., W.R.F.); Conduct of study (A.M.S., S.F.O., W.R.F.); Dataollection (A.M.S., S.F.O., W.R.F.); Data analysis (A.M.S., S.F.O., W.R.F.); Data interpretation (A.M.S., S.F.O., W.R.F.); and Preparation and reviewf the manuscript (A.M.S., S.F.O., W.R.F.). Institutional review board approval was obtained from the University of California, San Diego, before the study.

REFERENCES

1. Wiegand TW, Young LH. Cytomegalovirus retinitis. IntOphthalmol Clin 2006;46:91–110.

2. Kuo IC, Kempen JH, Dunn JP, et al. Clinical characteristicsand outcomes of cytomegalovirus retinitis in persons withouthuman immunodeficiency virus infection. Am J Ophthalmol2004;138:338–346.

3. Crippa F, Corey L, Chuang EL, et al. Virological, clinical,and ophthalmologic features of cytomegalovirus retinitisafter hematopoietic stem cell transplantation. Clin Infect Dis2001;32:214–219.

4. Benz MS, Glaser JS, Davis JL. Progressive outer retinalnecrosis in immunocompetent patients treated initially foroptic neuropathy with systemic corticosteroids. Am J Oph-thalmol 2003;135:551–553.

5. Browning DJ. Acute retinal necrosis following epiduralsteroid injections. Am J Ophthalmol 2003;136:192–194.

6. Goldstein DA, Pyatetsky D. Necrotizing herpetic retinopa-thies. Focal Points: Clinical Modules for Ophthalmologists2008;26:1–16.

7. Freeman WR, Thomas EL, Rao NA, et al. Demonstration ofherpes group virus in acute retinal necrosis syndrome. Am JOphthalmol 1986;102:701–709.

8. Saidel MA, Berreen J, Margolis TP. Cytomegalovirus retini-tis after intravitreal triamcinolone in an immunocompetentpatient. Am J Ophthalmol 2005;140:1141–1143.

9. Delyfer MN, Rougier MB, Hubschman JP, et al. Cytomega-lovirus retinitis following intravitreal injection of triamcin-olone: report of two cases. Acta Ophthalmol Scand 2007;85:681–683.

0. Ufret-Vincenty RL, Singh RP, Lowder CY, Kaiser, PK. Cyto-megalovirus retinitis after fluocinolone acetonide (Retisert™)implant. Am J Ophthalmol 2007;143:334–335.

1. Aggermann T, Stolba U, Brunner S, Binder S. Endoph-thalmitis with retinal necrosis following intravitreal triam-cinolone acetonide injection. Ophthalmologica 2006;220:131–133.

2. Toh TY, Borthwick JH. Acute retinal necrosis post intrav-itreal injection of triamcinolone acetonide. Royal Australianand New Zealand College of Ophthalmologists 2006:380–

3. Park YS, Byeon SH. Cytomegalovirus retinitis after intrav-itreal triamcinolone injection in a patient with centralretinal vein occlusion. Korean J Ophthalmol 2008;22:143–144.

4. Sekiryu T, Tomohiro I, Kaneko H, Saito M. Cytomegalovi-rus retinitis after intravitreal triamcinolone acetonide in animmunocompetent patient. Jpn J Ophthalmol 2008;52:414–416.

5. Jonas JB, Kreissig I, Sofker A, Degenring RF. Intravitrealinjection of triamcinolone for the treatment of diffusediabetic macular edema. Arch Ophthalmol 2003;121:57–61.

6. Jonas JB, Kreissig I, Degenring RF. Intravitreal triamcinoloneacetonide for pseudophakic cystoid macular edema. Am JOphthalmol 2003;136:384–386.

7. Antcliff RJ, Spalton DJ, Stanford MR, et al. Intravitrealtriamcinolone for uveitis cystoid macular edema: an opticalcoherence tomography study. Ophthalmology 2001;108:765–772.

8. Park CH, Jaffe GJ, Fekrat S. Intravitreal triamcinoloneacetonide in eyes with cystoid macular edema associated withcentral retinal vein occlusion. Am J Ophthalmol 2003;136:419–425.

9. Smithen LM, Ober MD, Maranan L, Spaide RF. Intravitrealtriamcinolone acetonide and intraocular pressure. Am JOphthalmol 2004;138:740–743.

0. Diabetic Retinopathy Clinical Research Network. A ran-domized trial comparing intravitreal triamcinolone acetonideand focal/grid photocoagulation for diabetic macular edema.Ophthalmology 2008;115:1447–1459.

1. Jabs DA, Van Natta ML, Kempen JH. Characteristics ofpatients with cytomegalovirus retinitis in the era of highlyactive antiretroviral therapy. Am J Ophthalmol 2002;133:48–61.

2. Morrison VL, Kozak I, LaBree LD, et al. Intravitreal triam-cinolone acetonide for the treatment of immune recoveryuveitis macular edema. Ophthalmology 2007;114:334–339.

3. Dalessandro L, Bottaro E. Reactivation of CMV retinitisafter treatment with subtenon’s corticosteroids for immunerecovery uveitis in a patient with AIDS. Scand J Infect Dis2002;34:780–782.

4. Song MK, Schrier RD, Smith IL, Plummer DJ, Freeman

WR. Paradoxical activity of CMV retinitis in patients

IVTA INJECTION 439

2

2

2

2

2

3

3

3

3

3

3

3

4

receiving highly active antiretroviral therapy. Retina 2002;22:262–267.

5. Jacobson MA, Zegans M, Pavan PR, et al. Cytomegalovirusretinitis after initiation of highly active antiretroviral ther-apy. Lancet 1997;349:1443–1445.

6. Petit N, Zandotti C, Riss JM, et al. Relapse of CMV retinitisin an AIDS patient with high CD4 counts. J AntimicrobChemother 1998;41:666–667.

7. Angus BJ, et al. HIV-associated CMV retinitis occurring inthe setting of a high CD4 count: what about AIDS defini-tions? Int J STD AIDS 1994;5:71–72.

8. Frame RD, Hutchins RK, Skahan KJ, et al. Progression ofcytomegaloviral retinitis in a patient responding to highlyactive antiretroviral therapy. Invest Ophthalmol Vis Sci1999;40:S873.

9. Moshfeghi DM, Kaiser PK, Scott IU, et al. Acute endoph-thalmitis following intravitreal triamcinolone acetonide in-jection. Am J Ophthalmol 2003;136:791–796.

0. Fintak DR, Shah GK, Blinder KJ, et al. Incidence ofendophthalmitis related to intravitreal injection of bevaci-

zumab and ranibizumab. Retina 2008;28:1395–1399.

AMERICAN JOURNAL OF40

1. Morrison VL, Koh HJ, Cheng L, et al. Intravitreal toxicity ofthe Kenalog vehicle (benzyl alcohol) in rabbits. Retina2006;26:339–344.

2. Tammewar AM, Cheng L, Kayikcioglu OR, et al. Com-parison of 4 mg versus 20 mg intravitreal triamcinoloneacetonide injections. Br J Ophthalmol 2008;92:810 –813.

3. Jonas JB. Intraocular availability of triamcinolone acetonideafter intravitreal injection. Am J Ophthalmol 2004;137:560–562.

4. Jonas JB. Concentration of intravitreally injected triamcin-olone acetonide in intraocular silicone oil. Br J Ophthalmol2002;86:1450–1451.

5. Kosobucki BR, Freeman WR, Cheng L. Photographic esti-mation of the duration of high dose intravitreal triamcino-lone in the vitrectomised eye. Br J Ophthalmol 2006;90:705–708.

6. Beer PM, Bakri SJ, Singh RJ, et al. Intraocular concentrationand pharmacokinetics of triamcinolone acetonide after asingle intravitreal injection. Ophthalmology 2003;110:681–

686.

OPHTHALMOLOGY MARCH 2010

AhMr

V

Biosketch

nkur M. Shah, MD, is a third year ophthalmology resident at the University of California, San Diego, California, wheree serves as chief resident. He received a Bachelor of Science degree from the University of Michigan, Ann Arbor,ichigan, and his medical degree from Northwestern University, Chicago, Illinois. His research interests include medical

etina and uveitis.

VIRAL RETINITIS AFTER IVTA INJECTIONOL. 149, NO. 3 440.e1