The OMARstudy
The OMARstudy
The OMARstudy
Purpose: To compare the risks and benefits of adding either intravitreal dexamethasone
implant (DEX) or preservative-free triamcinolone acetonide (TA) to bevacizumab mono-
therapy in refractory cystoid macular edema due to retinal vein occlusion.
Methods: This is a multicenter, comparative, interventional, retrospective study that
included 74 patients who were initially treated with intravitreal bevacizumab and later
received either DEX or TA for the treatment of recalcitrant cystoid macular edema due to
retinal vein occlusion. Main outcomes were best-corrected visual acuity, central macular
thickness, cost of therapy, frequency of intravitreal injections, and side effects.
Results: Thirty-nine patients received TA and 35 patients received DEX injections.
Groups were similar in age and gender distribution. Although the mean central macular
thickness improved significantly for all groups (P , 0.0001), logMAR best-corrected visual
acuity did not change significantly after steroid introduction (P = 0.06). Frequency of any
intravitreal injection decreased significantly from 0.66 ± 0.18 to 0.26 ± 0.08 injections per
month after initiation of steroids (P , 0.0001). This effect was greater in the DEX groups
(P , 0.0001). Monthly cost decreased with TA but increased with DEX.
Conclusion: Adding steroids improved anatomical outcome but did not affect final
vision. Injection frequency decreased significantly after adding steroids, more so with DEX.
There was no difference between TA and DEX regarding anatomical or functional outcomes
or the incidence of side effects.
RETINA 35:1393–1400, 2015
1393
1394 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 7
patients for bevacizumab, ranibizumab, dexamethasone groups further were divided into four groups depending
implant (DEX), and triamcinolone acetonide (TCA) were on whether they received DEX or TA injection (BRVO-
+15.5, +18.3, +6, and +5.7, respectively. The mean TA, BRVO-DEX, CRVO-TA, CRVO-DEX groups).
changes in BCVA in letters at Month 12 in patients with Data collected at each visit included BCVA, CMT
central retinal vein occlusion (CRVO) for bevacizumab, measured by SD-OCT, and the type of intravitreal drug
ranibizumab, aflibercept, DEX, and TCA were +16.1, injected, if any.
+13.9, +16.2, +2, and −1.2, respectively. Change in BCVA during the follow-up period was
Although intravitreal anti-VEGFs are mainstay of plotted over time to evaluate the functional effect of
the treatment for CME due to RVO, there are treatment. Anatomical effects of treatment were assessed
a considerable number of cases who do not respond based on the change in CMT demonstrated by SD-OCT.
or become recalcitrant with time to anti-VEGF The frequency of any intravitreal injection and the cost
treatment.11,12 In these cases, intravitreal steroids are of treatment were calculated and compared between the
commonly used, as it has been shown that they have groups before and after initiating intravitreal steroids.
the ability to reduce tachyphylaxis to anti-VEGF Values of Snellen BCVA were converted to logarithm
agents.12 Two of the available injectable intravitreal of minimum angle of resolution equivalent units
steroids are DEX and TCA. This study was designed (logMAR) for the purpose of statistical calculations.13
to compare the effects of adding DEX or triamcinolone The cost calculations included the office visit, the
to bevacizumab (Avastin) in the treatment of recalci- SD-OCT scan, the intravitreal injection fee, and the
trant CME secondary to RVO with a primary focus on cost of the drug injected.
the frequency of intravitreal injections, changes in cen-
tral macular thickness (CMT) and visual acuity, and Statistical Analysis
the cost of therapy.
Statistical evaluations were performed using Med-
Calc (version 12.3.0.0; MedCalc software, Mariakerte,
Belgium) and Microsoft Excel 2013 (Microsoft,
Materials and Methods
Redmond, WA). The chi-square test or Fisher’s exact
test was used to compare categorical data. Two-tailed
This was a retrospective multicenter study per-
Student’s t-test or analysis of variance was used for the
formed by collaboration of investigators who belong
analysis of the continuous variables. A P value less
to different institutions. The study was reviewed and
than 0.05 was considered statistically significant.
approved by the Institutional Review Board at the
University of Louisville. Exclusion criteria included
macular ischemia on fluorescein angiography, history Results
of focal/grid laser, and follow-up duration less than
12 weeks both before and after initiation of steroid Thirty-nine patients received TA (BRVO-TA 20;
injection. Medical records of 484 patients with RVO CRVO-TA 19) and 35 patients received DEX (BRVO-
were reviewed and 74 patients, who were initially DEX 18; CRVO-DEX 17) injections. The mean ages
treated with intravitreal injections of bevacizumab and gender distribution were not statistically different
(1.25 mg/0.1 mL) and later on received 0.7 mg DEX between the groups (Table 1). The mean total follow-up
or 4 mg/0.1 mL intravitreal preservative-free tri- durations in BRVO patients were 81.4 ± 30.4 weeks and
amcinolone acetonide (compounded from Westmore- 82.1 ± 40.3 weeks in TA and DEX groups, respectively
land Pharmacy, Jeffersonville, IN) for CME between (P = 0.951). The mean total follow-up durations in
November 2009 and December 2013, were identified CRVO patients were 87.1 ± 23.7 weeks and 74.1 ±
and included in the study. 29.8 weeks in TA and DEX groups, respectively (P =
All patients were diagnosed with perfused RVO, 0.172) (Table 1). Patients had a mean of 5.16 ± 1.85
confirmed by fluorescein angiography. Patients were bevacizumab injections before initiation of intravitreal
initially treated with at least three injections of steroids.
bevacizumab monotherapy and later received one of Bevacizumab was discontinued after initiation of
the intravitreal steroids because of refractory CME. steroids in 7 (35%) patients in the BRVO-TA group,
Reinjection of either drug was based on persistence or and 10 (55%) patients in the BRVO-DEX group (P =
recurrence of CME demonstrated by spectral domain 0.327). The mean number of monthly injections
optical coherence tomography (SD-OCT; Spectralis, decreased from 0.60 ± 0.20 to 0.29 ± 0.07 (P ,
Heidelberg, Germany). 0.0001) after initiating steroids in the BRVO-TA
Patients were divided into two groups depending on group and from 0.65 ± 0.18 to 0.21 ± 0.07 (P ,
whether they had BRVO or CRVO, and these two 0.0001) in the BRVO-DEX group (Table 2; Figure 1).
OZURDEX VERSUS TRIAMCINOLONE ACETONIDE IN RVO OZKOK ET AL 1395
Bevacizumab was discontinued after initiation of group and the 12-week mark for the remaining groups
steroids in 6 (31%) patients in the CRVO-TA group (P , 0.05). As the follow-up duration was not the
and 11 (65%) patients in the CRVO-DEX group same in all subjects, we used the final BCVA value
(0.093). The mean number of monthly injections in Tables 2 and 3; however, Figure 2 shows the avail-
decreased from 0.69 ± 0.11 to 0.30 ± 0.06 (P , able results of the subjects at the corresponding week.
0.0001) and from 0.70 ± 0.20 to 0.23 ± 0.06 (P , This is similar for CMT—final CMTs are indicated in
0.0001) after steroid injections in CRVO-TA and the tables, whereas Figure 3 represents fluctuations in
CRVO-DEX groups, respectively (Table 3; Figure 1). CMT throughout the follow-up period. The final visual
In BRVO cases, the mean numbers of steroid acuity was not statistically different from the visual
injections were 2.1 ± 0.7 and 1.4 ± 0.5 in the acuity at the time of steroid initiation.
BRVO-TA and BRVO-DEX groups, respectively For the BRVO subgroups, the number of patients
(P = 0.0011). On average, patients received a steroid who gained 3 or more Snellen lines were 1 (5%) and 3
injection every 19.9 ± 4.3 weeks in the BRVO-TA (16.7%) in the BRVO-TA and BRVO-DEX groups,
group and every 27.8 ± 11.9 weeks in the BRVO- respectively (P = 0.522); the number of patients who
DEX group (P = 0.009). In CRVO cases, the mean lost 3 or more lines were 1 (5%) and 1 (5.5%) in the
numbers of steroid injections were 2.6 ± 0.6 and 1.6 ± BRVO-TA and BRVO-DEX groups, respectively (P =
0.6 in the CRVO-TA and CRVO-DEX groups, respec- 0.939). For the CRVO subgroups, the number of pa-
tively (P , 0.0001). On average, patients received tients who gained 3 or more Snellen lines were 3
a steroid injection every 18.2 ± 4.8 weeks in the (15.8%) and 1 (5.9%) in the CRVO-TA and the
CRVO-TA and every 25.0 ± 12.5 weeks in the CRVO-DEX groups, respectively (P = 0.637); the num-
CRVO-DEX groups (P = 0.034). ber of patients who lost 3 or more lines were 2 (10.5%)
Visual acuity measurement before and after initia- and 2 (11.8%) in the CRVO-TA and the CRVO-DEX
tion of steroid injections are shown in Figure 2. The groups, respectively (P = 0.954; Table 3).
BCVA increased slightly after starting steroid injec- Central macular thickness changes for the dura-
tions in all groups, but the increase was statistically tion of the study are shown in Figure 3. The mean
significant only at the 6-week mark for the CRVO-TA CMT decreased from the initiation of steroid
therapy through all follow-up visits (P , 0.05 in all than 6 mmHg. This difference was not statistically
groups). significant (P = 0.764). Fifteen patients (38%) in
Details of steroid-related cataract and intraocular TA groups and 11 patients (31%) in DEX groups
pressure (IOP) increase are summarized in Table 4. had IOP . 21 mmHg after initiation of steroids (P =
Of the 12 phakic patients in the TA groups, 2 0.462). In all except 1 of the cases of ocular hyper-
(16.6%) needed cataract surgery, whereas of the 8 tension, IOP was controlled with topical drops
phakic patients in DEX groups, 1 (12.5%) needed alone. One patient in the TA-CRVO group required
cataract surgery. This difference was not statisti- trabeculectomy.
cally significant (P = 0.701). Of the 39 patients in The cost analysis data are summarized in Table 5.
the TA groups, 10 (25.6%) had an IOP increase The mean monthly cost of therapy increased from
more than 6 mmHg, whereas of the 35 patients in $222 ± 87 to $313 ± 152 in BRVO (P = 0.03) and
DEX groups, 7 (20%) had an IOP increase of more from $239 ± 83 to $351 ± 125 in CRVO (P = 0.01)
after dexamethasone injections were added to bevaci- to bevacizumab. Frequency of injections decreased after
zumab therapy. The mean monthly cost of therapy initiation of steroids, and the decrease was more
decreased from $213 ± 63 to $80 ± 17 in BRVO prominent in DEX groups. This finding emphasizes the
(P , 0.0001) and from $219 ± 46 to $92 ± 14 in longer duration of effect of DEX compared with TA.
CRVO (P , 0.0001) after TA injections were added Although CMT improved in all steroid groups, there was
to bevacizumab therapy (Table 5). no statistically significant difference between the DEX
and TA groups in terms of BCVA and CMT changes.
To the best of our knowledge, this is the first study to
Discussion compare the two available intravitreal steroids in this
particular patient group.
We compared the effectiveness of addition of either We also compared the rates of steroid-related ocular
TA or DEX in patients with RVO-related CME resistant side effects including cataract and IOP between the
Table 4. Comparison of Steroid Related Ocular Complications Between TA and DEX Groups
TA DEX P
Number of phakic cases 12 8
Number of cases that needed cataract surgery 2 (16.6%) 1 (12.5%) 0.701
Number of cases with IOP increase .6 mmHg or needed IOP 10 (25.6%) 7 (20%) 0.764
decreasing drops
Number of cases with IOP .21 mmHg after steroids 15 (38%) 11 (31%) 0.462
Number of cases that needed glaucoma surgery 1 (2.6%) 0 (0%) 0.340
Both BRVO and CRVO cases included.
TA, intravitreal triamcinolone acetonide.
two steroids. There was no statistically significant preferring anti-VEGF agents, bevacizumab being the
difference between TA and DEX groups in terms of most popular.14
steroid related side effects in this study. To the best of Intravitreal steroids constitute an important alternative
our knowledge, this is the first study to directly in cases with RVO-related CME that is bevacizumab
compare the side effect profiles of these two injectable resistant. Sharareh et al15 reported improved BCVA and
steroids. CMT with a short follow-up interval after initiation of
In this study, the percentages of cases that needed DEX in RVO-related CME that is resistant to bevaci-
IOP lowering medications were in agreement with the zumab. This study showed statistically significant im-
SCORE study, which investigated the use of 4 mg provements in BCVA and CMT in the short term after
TA,1,2 and with the GENEVA study, which studied initiation of DEX. The improvement in CMT was sta-
0.7 mg DEX implant.9 ble, but the improvement in BCVA was temporary,
There was no statistically significant difference in losing significance after the first 12 weeks. Although
the proportion of patients requiring cataract surgery CMT improved in both steroid groups, there was no
between the TA group (16.6%) and the DEX group statistically significant difference between the steroid
(12.5). However, both groups showed a higher pro- groups in terms of BCVA and CMT improvements.
portion of cataract progression than previously re- Data regarding the cost of RVO treatment is limited
ported: 4.3% in the SCORE study for 4 mg TA and and is mostly outdated. In a retrospective analysis of
1.3% in the GENEVA study DEX groups.1,2,8,9 secondary insurance payment data-based study, Fekrat
Further studies comparing the side effects of the two et al16 reported 16% to 22% higher 1-year costs and
medicines are needed to figure out the side effect rates 12% to 15% higher 3-year costs in RVO patients com-
in different indications such as uveitis, diabetic CME, pared with patients with hypertension. The annual
and RVO-related CME. direct cost of RVO treatment for Medicare population
Dexamethasone implant has received approval from is estimated to reach 5.8 billion dollars in 2006.17
the Food and Drug Administration for the treatment of Obviously, management of RVO has changed signif-
macular edema secondary to RVO in June 2009. icantly since 2006. Taylor et al18 evaluated the cost-
However, its use in clinical practice remains largely effectiveness of ranibizumab versus observation
overshadowed by anti-VEGF agents, particularly the and/or laser photocoagulation by using the data of
off-label drug bevacizumab. According to Preferences BRAVO, CRUISE, and HORIZON trials, and they
and Trends (PAT) 2013 survey released by the reported ranibizumab as a cost-effective option with
American Society of Retina Specialists, only 0.5% to an incremental cost-effectiveness ratio of £17.103.
2% of member retina specialists recommended the use This study is the first to evaluate the cost of RVO
of DEX or TA as initial therapy for CME secondary to treatment in the era of intravitreal injections. Monthly
RVO compared with more than 90% of members cost increased significantly after the initiation of
steroid therapy in DEX groups. In contrary, the groups but decreased in TA groups. There was no
monthly cost decreased significantly after initiation difference between TA and DEX regarding anatomical
of steroid therapy in TA groups. These differences or functional outcomes or rates of side effects. Starting
were noted despite a larger decrease in the frequency steroid addition to anti-VEGF therapy earlier in the
of intravitreal injections in the DEX groups compared course of the disease may yield better functional results
with the TA groups. and needs to be further investigated.
This study has limitations as it does not have focal/ Key words: refractory cystoid macular edema,
grid laser groups to compare the effects of laser in retinal vein occlusion, cost of therapy, combination
BRVO patients. As the number of cases with focal/grid therapy, bevacizumab, dexamethasone implant,
laser in our cohort was too few to compare and most of triamcinolone acetonide.
them also had steroids, to eliminate the possible effect
of laser treatment, we excluded these cases. Although
References
BCVA was used as a functional measurement, other
benefits of reduced macular edema such as reduction of 1. Ip MS, Scott IU, VanVeldhuisen PC, et al. A randomized trial
central scotoma were not measured in this study. comparing the efficacy and safety of intravitreal triamcinolone
with observation to treat vision loss associated with macular
Although not typically measured in clinical studies, it
edema secondary to central retinal vein occlusion: the Standard
can be of great functional benefit to patients. Patients Care vs Corticosteroid for Retinal Vein Occlusion (SCORE)
receiving TA often have short-lived (hours to days) study report 5. Arch Ophthalmol 2009;127:1101–1114.
“cloudy vision” from dispersed drug particles, which 2. Scott IU, Ip MS, VanVeldhuisen PC, et al. A randomized trial
are not experienced with long-acting DEX implant. comparing the efficacy and safety of intravitreal triamcinolone
with standard care to treat vision loss associated with macular
Transient cloudy vision should be considered in the
Edema secondary to branch retinal vein occlusion: the Standard
decision-making process because it may affect daily Care vs Corticosteroid for Retinal Vein Occlusion (SCORE)
living activities especially in monocular patients. In study report 6. Arch Ophthalmol 2009;127:1115–1128.
addition, there are patients who are unresponsive to 3. Brown DM, Campochiaro PA, Singh RP, et al. Ranibizumab
bevacizumab who do show clinical response to other for macular edema following central retinal vein occlusion: six-
month primary end point results of a phase III study. Ophthal-
anti-VEGF medications. Of note, here are patients who
mology 2010;117:1124–1133; e1121.
respond to combined anti-VEGF therapy with steroids 4. Campochiaro PA, Brown DM, Awh CC, et al. Sustained ben-
who experience exceptional anatomical and functional efits from ranibizumab for macular edema following central
improvement if instituted early before negative prog- retinal vein occlusion: twelve-month outcomes of a phase III
nostic changes. study. Ophthalmology 2011;118:2041–2049.
5. Brown DM, Campochiaro PA, Bhisitkul RB, et al. Sustained
In the era of SD-OCT and SS-OCT, the central
benefits from ranibizumab for macular edema following branch
retinal thickness outcome should be accompanied by retinal vein occlusion: 12-month outcomes of a phase III study.
future studies analyzing extensively the status of the Ophthalmology 2011;118:1594–1602.
inner segment/outer segment (IS/OS) line (ellipsoid 6. Boyer D, Heier J, Brown DM, et al. Vascular endothelial
layer), the cone outer segment tips (COST) line, and growth factor Trap-Eye for macular edema secondary to
central retinal vein occlusion: six-month results of the
also the inner retina degeneration before and after
phase 3 COPERNICUS study. Ophthalmology 2012;119:
treatment. Indeed, one of the major findings of this 1024–1032.
study was that although central retinal thickness 7. Brown DM, Heier JS, Clark WL, et al. Intravitreal aflibercept
improved after steroid initiation, BCVA did not. injection for macular edema secondary to central retinal vein
Nonimprovement of vision has been associated with occlusion: 1-year results from the phase 3 COPERNICUS
disruption of photoreceptors and atrophy of the inner study. Am J Ophthalmol 2013;155:429–437.
8. Haller JA, Bandello F, Belfort R Jr, et al. Dexamethasone
retina. This may explain why BCVA did not improve intravitreal implant in patients with macular edema related to
—probably because long-standing CME unresponsive branch or central retinal vein occlusion twelve-month study
to anti-VEGF therapy leads to photoreceptors atrophy. results. Ophthalmology 2011;118:2453–2460.
Qualitative structural changes in OCT at the time of 9. Haller JA, Bandello F, Belfort R Jr, et al. Randomized, sham-
the decision to add steroids or after addition of steroids controlled trial of dexamethasone intravitreal implant in pa-
tients with macular edema due to retinal vein occlusion.
were not analyzed in this study. Future studies to Ophthalmology 2010;117:1134–1146; e1133.
investigate this hypothesis are warranted. 10. Campochiaro PA, Heier JS, Feiner L, et al. Ranibizumab for
In conclusion, adding steroids to bevacizumab for macular edema following branch retinal vein occlusion: six-
refractory CME in RVO patients improved significantly month primary end point results of a phase III study. Ophthal-
the anatomical outcome, but did not affect visual acuity. mology 2010;117:1102–1112; e1101.
11. Pielen A, Feltgen N, Isserstedt C, et al. Efficacy and safety of
Injection frequency decreased significantly after adding intravitreal therapy in macular edema due to branch and central
steroids, more significantly in DEX groups compared retinal vein occlusion: a systematic review. PLoS One 2013;8:
with TA groups. Cost of treatment increased in DEX e78538.
1400 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 7
12. Schaal S, Kaplan HJ, Tezel TH. Is there tachyphylaxis to intra- intravitreal dexamethasone implant in retinal vein occlusion.
vitreal anti-vascular endothelial growth factor pharmacother- Retina 2013;33:1227–1231.
apy in age-related macular degeneration? Ophthalmology 16. Fekrat S, Shea AM, Hammill BG, et al. Resource use and costs
2008;115:2199–2205. of branch and central retinal vein occlusion in the elderly. Curr
13. Ferris FL 3rd, Kassoff A, Bresnick GH, Bailey I. New visual acuity Med Res Opin 2010;26:223–230.
charts for clinical research. Am J Ophthalmol 1982;94:91–96. 17. Laouri M, Chen E, Looman M, Gallagher M. The burden of
14. Preferences and Trends (PAT) Survey 2013. Available at: https:// disease of retinal vein occlusion: review of the literature. Eye
www.asrs.org/content/documents/_2013asrspatsurveyresults.pdf. (Lond) 2011;25:981–988.
Accessed October 22, 2014. 18. Taylor M, Serbetci E, Ferreira A, et al. A United Kingdom-
15. Sharareh B, Gallemore R, Taban M, et al. Recalcitrant macular based economic evaluation of ranibizumab for patients with
edema after intravitreal bevacizumab is responsive to an retinal vein occlusion (RVO). J Med Econ 2014;17:423–434.