Pegaptanib sodium for macular edema secondary to central retinal vein occlusion.

Abstract

To assess the safety and efficacy of intravitreous pegaptanib sodium for the treatment of macular edema following central retinal vein occlusion (CRVO).
This dose-ranging, double-masked, multicenter, phase 2 trial included subjects with CRVO for 6 months' or less duration randomly assigned (1:1:1) to receive pegaptanib sodium or sham injections every 6 weeks for 24 weeks (0.3 mg and 1 mg, n=33; sham, n=32).
Visual acuity at week 30.
In the primary analysis at week 30, 12 of 33 (36%) subjects treated with 0.3 mg of pegaptanib sodium and 13 of 33 (39%) treated with 1 mg gained 15 or more letters from baseline vs 9 of 32 (28%) sham-treated subjects (P= .48 for 0.3 mg and P= .35 for 1 mg of pegaptanib sodium vs sham). In secondary analyses, subjects treated with pegaptanib sodium were less likely to lose 15 or more letters (9% and 6%; 0.3-mg and 1-mg pegaptanib sodium groups, respectively) compared with sham-treated eyes (31%; P= .03 for 0.3 mg and P= .01 for 1 mg of pegaptanib sodium vs sham) and showed greater improvement in mean visual acuity (+7.1 and +9.9, respectively, vs -3.2 letters with sham; P= .09 for 0.3 mg and P= .02 for 1 mg of pegaptanib sodium vs sham). By week 1, the mean central retinal thickness decreased in the 0.3-mg and 1-mg pegaptanib sodium groups by 269 microm and 210 microm, respectively, vs 5 microm with sham (P< .001).
Based on this 30-week study, intravitreous pegaptanib sodium appears to provide visual and anatomical benefits in the treatment of macular edema following CRVO.
Benefits accrued with intravitreous pegaptanib sodium treatment of macular edema following CRVO suggest a role for vascular endothelial growth factor in the pathogenesis of this condition.
clinicaltrials.gov Identifier: NCT00088283.

Full text

Page 1

CLINICAL TRIALS
Pegaptanib Sodium for Macular Edema
Secondary to Central Retinal Vein Occlusion
John J. Wroblewski, MD; John A. Wells III, MD; Anthony P. Adamis, MD; Ronald R. Buggage, MD;
Emmett T. Cunningham Jr, MD, PhD; Mauro Goldbaum, MD; David R. Guyer, MD; Barrett Katz, MD, MBA;
Michael M. Altaweel, MD; for the Pegaptanib in Central Retinal Vein Occlusion Study Group
Objectives: To assess the safety and efficacy of intra-
vitreous pegaptanib sodium for the treatment of macu-
laredemafollowingcentralretinalveinocclusion(CRVO).
Design: This dose-ranging, double-masked, multi-
center, phase 2 trial included subjects with CRVO for 6
months’orlessdurationrandomlyassigned(1:1:1)tore-
ceivepegaptanibsodiumorshaminjectionsevery6weeks
for 24 weeks (0.3 mg and 1 mg, n=33; sham, n=32).
Main Outcome Measure: Visual acuity at week 30.
Results: In the primary analysis at week 30, 12 of 33
(36%) subjects treated with 0.3 mg of pegaptanib so-
diumand13of33(39%)treatedwith1mggained15or
more letters from baseline vs 9 of 32 (28%) sham-
treated subjects (P=.48 for 0.3 mg and P=.35 for 1 mg
of pegaptanib sodium vs sham). In secondary analyses,
subjectstreatedwithpegaptanibsodiumwerelesslikely
to lose 15 or more letters (9% and 6%; 0.3-mg and 1-mg
pegaptanibsodiumgroups,respectively)comparedwith
sham-treatedeyes(31%;P=.03for0.3mgandP=.01for
1 mg of pegaptanib sodium vs sham) and showed greater
improvement in mean visual acuity (?7.1 and ?9.9, re-
spectively,vs−3.2letterswithsham;P=.09for0.3mgand
P=.02 for 1 mg of pegaptanib sodium vs sham). By week
1, the mean central retinal thickness decreased in the
0.3-mg and 1-mg pegaptanib sodium groups by 269 µm
and 210 µm, respectively, vs 5 µm with sham (P?.001).
Conclusions: Based on this 30-week study, intravitre-
ouspegaptanibsodiumappearstoprovidevisualandana-
tomical benefits in the treatment of macular edema fol-
lowing CRVO.
Application to Clinical Practice: Benefitsaccruedwith
intravitreous pegaptanib sodium treatment of macular
edema following CRVO suggest a role for vascular endo-
thelialgrowthfactorinthepathogenesisofthiscondition.
Trial Registration: clinicaltrials.gov Identifier:
NCT00088283
Arch Ophthalmol. 2009;127(4):374-380
M
istration–approved pharmacologic treat-
ments exist for macular edema in the set-
ting of CRVO.4,5Grid photocoagulation
doesnotimprovevisualacuitywhencom-
pared with controls,3and surgical inter-
ventions have not been evaluated in con-
trolled randomized trials.
Centralretinalveinocclusionisthought
to result from an obstruction of unknown
etiologyinthecentralretinalveinatorvari-
able distances posterior to the lamina
cribrosa.6Based on histopathology of eyes
withCRVOthatwereenucleatedowingto
neovascularglaucoma,ischemicCRVOmay
result from the formation of focal thrombi
at or just posterior to the lamina cribrosa.7
ACULAR EDEMA FOL-
lowing central retinal
veinocclusion
(CRVO) is a major
cause of vision loss.1-3
No United States Food and Drug Admin-
ThesiteofocclusioninnonischemicCRVO
islikelytobefurtherposteriortothelamina
cribrosa. Subsequent hypoxia leads to up-
regulation of vascular endothelial growth
factor(VEGF),8resultinginincreasedreti-
nal capillary permeability9and leakage of
fluid and blood into the intraretinal
space.10,11In addition, VEGF is a key pro-
moter of angiogenesis,12potentially con-
tributingtothedevelopmentoftheneovas-
cularization associated with CRVO. In
patients with ischemic CRVO, defined by
thepresenceof10to75ormorediscareas
of capillary nonperfusion determined by
fluorescein angiography,4aqueous humor
levels of VEGF show a close temporal cor-
relation with the course of iris neovascu-
larization and retinal capillary permeabil-
ity.13Inhibition of VEGF in a nonhuman
primatemodelofCRVOcanpreventthede-
velopment of iris and angle neovascular-
ization.14Vascular endothelial growth fac-
Author Affiliations are listed at
the end of this article.
Group Information: The
Pegaptanib in Central Retinal
Vein Occlusion Study Group
Investigators are listed at the
end of this article.
(REPRINTED) ARCH OPHTHALMOL/VOL 127 (NO. 4), APR 2009WWW.ARCHOPHTHALMOL.COM
374
©2009 American Medical Association. All rights reserved.
on April 21, 2009 www.archophthalmol.comDownloaded from

Page 2

tor induces expression of thromboplastin,15-17a potent
procoagulant tissue factor that may aggravate the retinal
ischemia induced by the CRVO. Furthermore, VEGF in-
duces retinal endothelial swelling, which may lead to fur-
thercapillarynonperfusion.18,19GiventheroleofVEGFin
the pathophysiology of macular edema and neovascular-
ization following CRVO and the lack of proven and safe
methods of rapid restoration of blood flow, inhibition of
VEGF is a rational therapeutic strategy for this disease.
Pegaptanib,a40-kDaRNAaptamer,bindstoVEGF165,
the predominant pathological isoform in ischemia-
mediatedocularneovascularizationandindiseasessuch
as diabetic macular edema.20-23Clinical trials have sug-
gested that intravitreous injection of pegaptanib so-
dium can be effective in the treatment of diabetic macu-
lar edema,24proliferative diabetic retinopathy,25and
neovascularage-relatedmaculardegeneration.26,27Asin-
creased intraocular VEGF contributes to the pathogen-
esis of vision loss following CRVO, a phase 2 sham-
controlled trial was conducted to examine the 30-week
safetyandefficacyofintravitreouspegaptanibforthetreat-
ment of macular edema secondary to CRVO.
METHODS
Thiswasarandomized,dose-ranging,double-masked,parallel-
group,sham-controlled,multicenter,phase2clinicaltrialcon-
ductedinpractitioners’officesandclinicsinAustralia,France,
Germany, Israel, Spain, and the United States. The study pro-
tocolwasreviewedandapprovedbyaninstitutionalreviewboard
ateachstudysiteinaccordancewiththeguidelinesforthecon-
duct of clinical research in the 1964 Declaration of Helsinki.
All study participants provided signed informed consent be-
fore any baseline procedures were performed, and the study
sitesintheUnitedStateswereHealthInsurancePortabilityand
Accountability Act compliant. Subjects 18 years or older diag-
nosed with CRVO with onset of symptoms within 6 months
prior to baseline and with a best-corrected visual acuity in the
study eye of between 65 and 20 Early Treatment of Diabetic
RetinopathyStudylettersinclusive(approximately20/50to20/
400 Snellen equivalents) and better than or equal to 35 letters
(approximately 20/200) in the fellow eye were eligible. Macu-
laredemawasassessedbyopticalcoherencetomography(OCT).
Study eyes were required to have a central retinal thickness at
the center point of 250 µm or more at both baseline and the
firsttreatmentday.Subjectswithanophthalmichistoryofany
of the following were excluded: subtenon corticosteroid ad-
ministrationforanyophthalmiccondition;priorpanretinalor
sectorscatterphotocoagulation;signsofoldbranchretinalvein
occlusion or CRVO in the study eye, or any other retinal vas-
cular disease including diabetic retinopathy. Eyes with a brisk
afferentpupillarydefect;vitreoushemorrhageexceptforbreak-
through hemorrhage from intraretinal hemorrhage; evidence
of any neovascularization involving the iris, disc, or retina; or
anyotherclinicallysignificantconcomitantoculardiseaseswere
also excluded. There were no angiographic criteria for inclu-
sion in the study. The exclusion of subjects with visual acuity
of less than 20 letters or with brisk afferent pupillary defect,
features more reliably predictive of ischemic CRVO, was in-
tended to limit the inclusion of subjects in whom vision may
not improve despite resolution of macular edema.6,28
The University of Wisconsin Fundus Photograph Reading
Center, the independent reading center for this study, deter-
minedangiographicandOCTbaselinemeasuresandendpoints
and confirmed the eligibility of all subjects prior to their en-
rollment and randomization. Eligible subjects were allocated
equally (1:1:1) to 1 of 3 treatment arms, 0.3 mg of pegaptanib
sodium, 1 mg of pegaptanib sodium, or sham injections, with
randomization stratified by center and baseline visual acuity
(?34 letters vs ?34 letters; approximately 20/200); a maxi-
mum of 40 subjects with baseline visual acuity of 34 letters or
less were to be randomized. Treatment assignment was based
on a dynamic minimization procedure that used a stochastic
treatment allocation algorithm based on the variance method.
Medication kits were identified by randomization number. All
kits were similar in appearance, regardless of dose. To main-
tain investigator masking, the injection was not administered
by the study ophthalmologist responsible for patient care and
assessments.Thestudycoordinatorconveyedthetreatmentas-
signment to the study ophthalmologist administering the in-
jection in a fashion that did not inform the treating ophthal-
mologist or the subject.
Intravitreous pegaptanib sodium or sham injections were
administered every 6 weeks for 24 weeks, for a total of 5 in-
jections. Antisepsis procedures were the same for all subjects
including those receiving sham; all subjects received injected
subconjunctival anesthetic. Masking was maintained by treat-
ingsubjectsreceivingshamidenticallytothosereceivingpegap-
tanib sodium, with the exception that sham subjects did not
have scleral penetration; rather, they had blunt pressure ap-
plied to the globe without a needle. During the study, panreti-
nal photocoagulation was permitted at any time point for neo-
vascularizationaccordingtotheCentralVeinOcclusionStudy
protocol;intravitreoussteroidswerenotpermittedatanytime.
STUDY ASSESSMENTS AND END POINTS
An ophthalmologic history and all baseline assessments were
conducted within 14 days prior to the first study treatment. At
baseline and at weeks 0, 6, 12, 18, 24, and 30, vital signs were
recorded,andthefollowingassessmentswereperformedprior
tostudytreatment:best-correctedvisualacuitybycertifiedex-
aminers masked to treatment arm and results of previous vi-
sualacuitymeasurements;applanationtonometry(beforeand
30minutesafterinjection;alsoatleastoncebetweeneach6-week
visit);ophthalmicexamination(alsoatleastoncebetweeneach
6-weekvisit);examinationoftheirisincludinggonioscopy(prior
to dilation); and laboratory tests (eg, hematology, renal func-
tion, hepatic function, and electrolytes). Fluorescein angiog-
raphy was performed at baseline and week 30, and stereo-
scopic color fundus photographs were taken at baseline and at
weeks 12 and 30. Optical coherence tomography was per-
formed at weeks 0, 1, 3, 6, 12, 18, 24, and 30.
Theprimaryefficacyendpointwasthepercentageoftreated
eyes gaining 15 letters or more of visual acuity from baseline
up to 30 weeks for each dose group. The main secondary effi-
cacyendpointsweredifferencesinmeanchangeinvisualacu-
ity over time and from baseline to week 30; percentage of eyes
losing15lettersormoreofvisualacuityfrombaselinetoweek
30; percentage of eyes with visual acuity of 35 letters or more
(betterthanapproximately20/200)and65lettersormore(bet-
terthanapproximately20/50)atweek30;themeanchangefrom
baselineinOCTcenterpointandcentralsubfieldvaluesatweek
30 and over time; and the percentage of eyes developing reti-
nal or iris neovascularization postbaseline before week 30.
Safety end points included all adverse events spontane-
ously reported, elicited, or observed by investigators. Adverse
events were graded as mild, moderate, or severe and were as-
sessed as being either related to the injection procedure or to
thestudydrugsorunrelatedtostudytreatment.Allseriousad-
verse events were recorded whether deemed related to treat-
ment or not.
(REPRINTED) ARCH OPHTHALMOL/VOL 127 (NO. 4), APR 2009WWW.ARCHOPHTHALMOL.COM
375
©2009 American Medical Association. All rights reserved.
on April 21, 2009 www.archophthalmol.comDownloaded from

Page 3

STATISTICAL ANALYSIS
Efficacy analyses were conducted on the intent-to-treat popu-
lation, which included all randomized subjects. Missing data
were imputed using the last-observation-carried-forward
method, except for repeated-measures analyses of variance in
which no imputation of missing data was performed. Subjects
wereanalyzedinthetreatmentgroupandstratumtowhichthey
were assigned by randomization. The significance of associa-
tions between binary end points, including the primary end
point, and treatment was analyzed by applying pairwise com-
parisons using the Cochran-Mantel-Haenszel test adjusted for
vision at randomization (?34 letters or ?34 letters). Analysis
of covariance was used to analyze continuous data or changes
in continuous end points at a given time; the model included
main effects of treatment, vision at randomization (?34 let-
ters or ?34 letters), and the baseline value of the correspond-
ing end points (except for visual acuity). A repeated-measures
analysis of variance that used all visual acuity and OCT data
overtimewasperformedtocompareeachdosegroupwithsham.
All statistical tests were 2-sided, and the ? level was set to .05.
No adjustment for multiplicity was performed.
Safetyanalysesincludedallsubjectsreceivingatleast1dose
ofthestudydrug(pegaptanibsodiumorsham).Adverseevents
weresummarizedusingMedicalDictionaryforRegulatoryAc-
tivities5.1terms.Summarystatisticswerecalculatedforsafety
end points.
The sample size calculation was based on the following as-
sumptions: the test of significance should be 2-sided, with a sig-
nificance level of P=.05; and the percentage of subjects gaining
15 letters or more of vision at week 30 was expected to be 30%
in the sham group, increasing to 65% in groups treated with 0.3
mg or 1 mg of pegaptanib sodium. Using these assumptions, it
was determined that a sample of at least 30 subjects per treat-
ment group was required for an overall power of 80%.
RESULTS
ThestudywasconductedbetweenAugust2004andSep-
tember 2006. A total of 98 subjects entered the study,
with 33 subjects in both the 0.3-mg and 1-mg pegap-
tanib sodium groups and 32 in the sham group. At base-
line, subject characteristics were comparable and visual
acuitywaswellbalancedacrossstudyarms(Table).Over-
all,themeanagewas62.6yearsandthemeanvisualacu-
ity in the study eye at baseline was 48.1 letters (20/100
Snellen lines).
Fewsubjects(7%overall)withdrewfromthestudybe-
tweenbaselineandweek30.Threesubjectsinthe0.3-mg
grouprequestedtobediscontinuedfromthestudy,asdid
1 subject in the 1-mg group and 2 subjects in the sham
group.Theonlyotherdiscontinuationduringthestudyoc-
curred in 1 subject in the sham cohort as a result of non-
compliance. Most subjects adhered to the study regimen,
receivingall5plannedinjectionsfrombaselinetoweek30:
81%ofsubjectsinthe0.3-mggroup,90%inthe1-mggroup,
and 88% of subjects in the sham group.
In the primary analysis of this phase 2 study, al-
thoughnotstatisticallysignificant,12of33(36%)0.3-mg
pegaptanib sodium and 13 of 33 (39%) 1-mg subjects
gained 15 or more letters from baseline to week 30 vs 9
of 32 (28%) subjects in the sham group (P=.48; relative
risk [RR],1.29; 95% confidence interval [CI],0.63-2.64
andP=.35;RR,1.40;95%CI,0.70-2.81vssham,respec-
tively).Secondaryanalysesconsistentlyshowedbettervi-
sual results in subjects treated with pegaptanib sodium.
Both doses of pegaptanib sodium had an early and sus-
tained effect on mean visual acuity (Figure 1). At week
Table. Baseline Demographic and Ocular Characteristics
Characteristic
Pegaptanib
Sodium,
0.3 mg
(n=33)
Pegaptanib
Sodium,
1 mg
(n=33)
Sham
(n=32)
Mean age, y
Sex, female:male
Mean systolic blood pressure, mm Hg
Mean diastolic blood pressure, mm Hg
Mean time between occlusive event and study entry, da
Mean baseline visual acuity, letters (approximate Snellen lines)
Subjects with visual acuity ?35 and ?65 letters (approximately 20/200-20/50 Snellen lines)
at baseline, No. (%)
Subjects with visual acuity ?65 letters (approximately 20/50 Snellen lines) at baseline, No. (%)
Mean baseline center point thickness, µm
Mean baseline central subfield thickness, µm
64 6459
18:15
133
75
81
15:18
129
77
82
13:19
127
80
77
47.6 (20/100)
21 (64)
48.4 (20/100)
19 (58)
48.5 (20/100)
23 (72)
4 (11)
688
675
6 (18)
632
619
3 (9)
674
656
aData available for 25, 25, and 27 subjects in 0.3-mg pegaptanib sodium, 1-mg pegaptanib sodium, and sham groups, respectively.
15
10
5
0
–5
–10
6 12
1824 30
Time, wk
Mean Change in Visual Acuity, Letters
Pegaptanib, 0.3 mg (n = 33)
Pegaptanib, 1 mg (n = 33)
Sham (n = 32)
+9.9∗
+7.1
–3.2
Figure 1. Mean change in visual acuity from baseline. Last observation
carried forward. *P?.05.
(REPRINTED) ARCH OPHTHALMOL/VOL 127 (NO. 4), APR 2009 WWW.ARCHOPHTHALMOL.COM
376
©2009 American Medical Association. All rights reserved.
on April 21, 2009 www.archophthalmol.com Downloaded from

Page 4

30,subjectsreceiving0.3mgand1mgofpegaptanibso-
dium had gained an average of 7.1 and 9.9 letters, re-
spectively,whilethosetreatedwithshamhadlostanav-
erageof3.2letters(P=.09;95%CI,−1.3to21.8andP=.02;
95% CI,1.5 to 24.6 for 0.3 mg and 1 mg of pegaptanib
sodium vs sham, respectively). Less than 10% of those
receivingeitherdoseofpegaptanib(9%and6%for0.3-mg
and 1-mg pegaptanib sodium groups, respectively) lost
15 letters or more of visual acuity by week 30 compared
with31%ofsubjectstreatedwithsham(P=.03;RR,0.29;
95% CI, 0.09-0.96; and P=.01; RR, 0.19; 95% CI, 0.05-
0.82for0.3mgand1mgofpegaptanibsodiumvssham,
respectively) (Figure 2). Approximately 90% (59/66)
ofsubjectsreceivingpegaptanibhadvisualacuitiesof20/
200 or better at week 30 compared with 63% (20/32) of
those treated with sham (P=.02 and P=.01 for 0.3 mg
and 1 mg of pegaptanib sodium vs sham, respectively),
buttherewasnodifferenceinthepercentageofsubjects
with visual acuity of 20/50 or better at week 30 (33% for
both pegaptanib doses and 34% for sham).
The mean decrease from baseline in retinal thickness
at both the center point and the center subfield were
greaterinbothpegaptanibsodiumtreatmentgroupsthan
in the subjects receiving sham (Figure 3) at all time
points. The mean reduction in retinal thickness at the
center point and at the center subfield from baseline to
week 30 were 95 µm and 112 µm lower, respectively, in
the0.3-mgpegaptanibsodiumgroupcomparedwithsham
(P=.13; 95% CI,−209 to 26 and P=.06; 95% CI,−170 to
63, respectively; last observation carried forward data).
Significant changes were identified on OCT scans ob-
tained at weeks 1 and 3. Between baseline and week 1,
themeandecreaseincenterpointthicknesswas269µm
inthe0.3-mgpegaptanibsodiumand210µminthe1-mg
pegaptanib sodium group compared with 5 µm in the
sham group. The mean decrease from baseline to week
3was329µminthe0.3-mgpegaptanibsodiumand198
µminthe1-mgpegaptanibsodiumgroupscomparedwith
40 µm in the sham group (P?.001) (Figure 3).
A total of 23, 25, and 26 subjects in the 0.3-mg, 1-mg,
and/or sham groups, respectively, had gradable fluores-
cein angiograms for the presence of capillary nonperfu-
sionatbaseline.Assessingthecenter,inner,andoutersub-
fields(central16discareas)inthosesubjects,themedian
and first quartile of the measurement of capillary nonper-
fusionwaszerodiscareasinallstudyarms.Thethirdquar-
tiles of capillary nonperfusion were 0.01, 0.03, and 0.00
disc areas in the 0.3-mg pegaptanib sodium, 1-mg pegap-
tanib sodium, and sham groups, respectively. There was
noassessmentofperipheralretinalnonperfusion.Thede-
velopmentofretinaloririsneovascularizationwasuncom-
50
40
30
20
10
0
≥3015-2910-14 5-9±4 10-145-9 15-29
≥30
Decreased
Visual Acuity
Increased
Change in Visual Acuity, % of Subjects
Pegaptanib, 0.3 mg (n = 33)
Pegaptanib, 1 mg (n = 33)
Sham (n = 32)
Week 6
50
40
30
20
10
0
≥30 15-2910-14 5-9±4 10-14 5-915-29
≥30
Decreased
Visual Acuity
Increased
Week 30
Figure 2. Distributions of changes in visual acuity from baseline at 6 and 30 weeks.
50
0
–50
–100
–150
–200
–250
–300
–350
–400
13 12
6 18 24 30
Time, wk
Mean Change in Thickness at the Center Point, µm
Pegaptanib, 0.3 mg (n = 33)
Pegaptanib, 1 mg (n = 33)
Sham (n = 32)
–148
–179
–243
∗
∗
∗
∗
Figure 3. Mean change from baseline in center point thickness. Last
observation carried forward. *P?.001.
(REPRINTED) ARCH OPHTHALMOL/VOL 127 (NO. 4), APR 2009WWW.ARCHOPHTHALMOL.COM
377
©2009 American Medical Association. All rights reserved.
on April 21, 2009 www.archophthalmol.com Downloaded from

Page 5

mon,butoccurredmorefrequentlyintheshamarm(3sub-
jects[9%])thanineitherofthepegaptanibtreatmentarms
(1subject[3%]inbothpegaptanibarms;P=.29).All5sub-
jects who developed ocular neovascularization were sub-
mitted to panretinal photocoagulation.
No serious ocular adverse events were reported, and
no subject developed endophthalmitis, traumatic cata-
ract, or retinal detachment. No evidence of a sustained
effectonintraocularpressurewasnoted.Noevidenceof
an increased risk of systemic adverse events related to
VEGF inhibition was detected.
COMMENT
Inthisphase2,randomized,controlled,30-weektrialthat
evaluatedintravitreouspegaptanibsodiumgivenat6-week
intervals for subjects with macular edema secondary to
CRVO, analysis of the primary efficacy end point of the
percentage of subjects gaining 15 letters or more mea-
sured at the 30-week follow-up visit revealed a trend fa-
voring eyes treated with pegaptanib sodium, although
withoutstatisticalsignificance.Secondaryanalyses,how-
ever, showed a significant reduction in the percentage
of subjects treated with pegaptanib sodium losing 15 or
more letters and a significant increase in the percentage
of subjects treated with pegaptanib sodium remaining
above the threshold for legal blindness. The combina-
tionoftheresultsofthese2binaryanalyses—onlyaposi-
tive trend for doubling with a significant difference for
avoidinghalvingthevisualangle—explainsthesubstan-
tial(2-3Snellenlines)shiftinthemeanvisualacuitydis-
tributionfavoringthegroupstreatedwithpegaptanibso-
dium. These results illustrate one of the disadvantages
of binary vs continuous variable analyses for visual acu-
ity. Binary variables (eg, proportions of subjects gaining
orlosingatleast15letters)areoftenselectedasprimary
end points for clinical trials in retina. However, as re-
cently discussed by Beck et al,29despite some regulatory
advantages, there is an intrinsic loss of information in
binaryvariableanalysescomparedwithcontinuousvari-
ableanalysessuchasmeanvisualacuityscore.Whenusing
a binary outcome, a cutoff point is chosen (eg, 15-letter
gain, 15-letter loss). All results on the same side of the
cutoff point contribute equal weight to the results, ie, a
subject with a 16-letter gain and a subject with a 25-
letter gain contribute equally. In contrast, in a continu-
ousvariableanalysissuchasmeanvisualacuity,eachsub-
ject’s visual acuity score contributes its specific value to
theanalysissodistributionshiftsbetweengroupsarere-
flected without loss of information. Although the differ-
ence in the proportion of subjects who gained 15 letters
in this trial was not statistically significant, the differ-
ences in the proportion of those who lost 15 letters and
the mean visual acuity were substantial, providing sup-
port for a potential benefit of pegaptanib sodium in vi-
sual acuity of patients with macular edema secondary to
CRVO.
In addition to having a positive treatment effect on
macular edema, fewer eyes in the groups treated with
pegaptanib sodium showed retinal or iris neovascular-
ization, a known and frequent complication of CRVO.
Thissuggeststhatpegaptanibsodiummayhelplimitthe
development of retinal or iris neovascularization, al-
though a larger trial would need to confirm this finding.
In keeping with the favorable safety record of pegap-
tanib sodium in larger longer-term trials,30these ben-
efits accrued without the generation of any systemic or
ocular safety concerns.
Thisisthefirstrandomized,controlled,30-week,phar-
macologic trial that has shown any treatment benefit for
CRVO compared with sham31and is consistent with the
clinicalbenefitalreadyobservedforintravitreouspegap-
tanib sodium in treating diabetic retinopathy.24,25These
resultsareespeciallyencouraginginthatthereisnoproven
effectivetherapyformacularedemasecondarytoCRVO.
A plethora of treatment options have been examined in
prospectiveorretrospectiveuncontrolledstudiesthatrely
on comparison with historical controls to demonstrate
potential benefits but never in randomized controlled
trials.5
In a small trial of 27 eyes with CRVO randomized to
4 mg of intravitreous triamcinolone or sham injections,
the mean visual acuity was better in the triamcinolone
groupat1month,butboththetreatedandcontrolgroups
hadsimilarimprovementsinvisualacuityatthe4-month
final assessment.32Intravitreous bevacizumab, another
anti-VEGF agent, has been reported to provide short-
term improvements in visual acuity and macular edema
subsequenttoCRVOinseveralrecentuncontrolledsmall
studies.33-37Itisuncertainifallanti-VEGFdrugswillshare
similar safety profiles in the context of retinal vein oc-
clusions. However, preclinical data suggest that, in con-
trast to selective inhibition of VEGF165,nonselective in-
hibition of all VEGF isoforms rendered retinal neurons
more sensitive to apoptosis under ischemic conditions
such as CRVO.38
As recently demonstrated in patients with diabetic
macularedema,therewaspoorcorrelationbetweenmacu-
lar thickness and visual acuity. In general, visual acuity
gains are believed to lag behind OCT-documented cen-
terpointthicknessandcentralsubfieldthicknessreduc-
tions. In this study, the macular thickness decrease was
similarinthe1-mgandshamarmsfrombaselinetoweek
30, yet the mean change in visual acuity for subjects re-
ceiving 1 mg of pegaptanib was superior to subjects re-
ceiving sham. A possible explanation for this finding is
that other mechanisms such as ischemia-mediated neu-
ronal death may influence vision outcomes in subjects
withretinalvasculardisease.38Additionalimportantfac-
torsinfluencingvisualacuityinthisdisorderarethepres-
ence of foveal ischemia, retinal hemorrhages, exudates,
orretinalpigmentepitheliumatrophyatthecenterofthe
macula, any of which may limit the potential for vision
improvement.Further,chronicmacularedemacanlead
tounderlyingretinalpigmentepitheliumatrophyand/or
hyperplasia that prevent visual acuity improvement de-
spite macular detergescence, reinforcing the impor-
tance of early treatment of macular edema in CRVO to
avoid irreversible functional damage.
Comparedwithpreviouspegaptanibstudies,therewere
nonewsafetysignalsreportedinthistrial.However,this
trialwasnotdesignedtodetectsmalldifferencesinsafety
events with statistical significance between treated and
(REPRINTED) ARCH OPHTHALMOL/VOL 127 (NO. 4), APR 2009 WWW.ARCHOPHTHALMOL.COM
378
©2009 American Medical Association. All rights reserved.
on April 21, 2009 www.archophthalmol.comDownloaded from