ORIGINAL ARTICLE

Serum D-Dimer Levels to Evaluate the Riskfor Arterial Thromboembolism After Intravitreal

Injection of Bevacizumab and Ranibizumab

Donghyun Jee,1,2 Masahiro Zako,3 and Tae Yoon La1,2

Abstract

Purpose: There are concerns about arterial thromboembolic event after intravitreal injection of bevacizumab orranibizumab. Motivated by the fact that D-dimer was a sensitive biomarker for thromboembolism, we evaluatedserum D-dimer levels in patients with age-related macular degeneration (AMD) after intravitreal injection ofbevacizumab and ranibizumab.Methods: In this prospective, nonrandomized, uncontrolled study, 122 patients (122 eyes) with AMD wereenrolled. Sixty-two eyes received intravitreal injections of bevacizumab and 60 eyes received intravitrealinjections of ranibizumab monthly for 3 months. Serum D-dimer levels were measured in patients beforeintravitreal injection and 1 day, 1 week, 1 month, and 3 months thereafter.Results: Serum D-dimer levels were not significantly altered following injection of either bevacizumab orranibizumab. Subgroup analysis for patients at risk for thromboembolic events revealed that serum D-dimerlevels showed no significant change after injection of ranibizumab. However, D-dimer levels significantlyincreased at 1 day (P = 0.041) and 1 week (P = 0.022) after injection of bevacizumab.Conclusions: Serum D-dimer levels were not changed after injection with either bevacizumab or ranibizumab.In subgroup analysis, bevacizumab injection in patients at risk of thromboembolism increased serumD-dimer levels.

Introduction

Inhibitors of vascular endothelial growth factor(VEGF) are widely used to treat patients with wet, age-

related macular degeneration (AMD).1,2 Neutralization ofall isoforms of VEGF-A significantly improves the ability tominimize vision loss, and restore sight in patients withAMD.3,4 Commonly used antibody-derived drugs for treat-ing AMD are bevacizumab (Avastin�; Hoffmann La Roche,Basel, Switzerland) and ranibizumab (Lucentis�; Genetech,Inc., San Francisco, CA). Bevacizumab is a full-length an-tibody that was developed as a potential therapeutic agentfor use in oncology.5 Ranibizumab is an antigen-bindingfragment (Fab) of an antibody, and was developed as part ofan anti-VEGF program for AMD treatment.6

Although anti-VEGF antibodies provide remarkableclinical benefits for AMD patients, concerns remain about thepotential systemic off-target effects of anti-VEGF inhibition.Intravitreally injected drugs enter the systemic circulation,

which blocks physiologic VEGF.7 This blockage of physio-logic VEGF increases the risk of a thromboembolic event orcardiovascular accident.7,8 Because VEGF not only stimulatesendothelial cell proliferation, but also maintains vascular in-tegrity, VEGF inhibition could cause endothelial dysfunctionleading to thrombosis.9 A recent meta-analysis of prospectiveclinical trials using anti-VEGF antibodies indicated an asso-ciation between intravitreal injection of anti-VEGF antibodiesand the subsequent incidence of cardiovascular accidents.10

Another study reported that the rate of arterial thromboembolicevents was marginally higher in treatment arms receivinghigher doses of anti-VEGF antibodies, although this trend wasnot significant.11

Thromboembolism is a major cause of death in the UnitedStates and a leading cause of morbidity, with an annualincidence of about 1 case per 1,000 individuals.12,13 Al-though a thromboembolic event can be fatal, few studieshave investigated thromboembolism after intravitreal anti-VEGF therapy. One possible reason is that thromboembolic

1Department of Ophthalmology and Visual Science, St. Vincent Hospital, Suwon, Korea.2College of Medicine, Catholic University of Korea, Seoul, Korea.3Department of Ophthalmology, Aichi Medical University, Aichi, Japan.

JOURNAL OF OCULAR PHARMACOLOGY AND THERAPEUTICSVolume 00, Number 00, 2014ª Mary Ann Liebert, Inc.DOI: 10.1089/jop.2013.0234

1

events after intravitreal injection of anti-VEGF occur atextremely low rates. To identify low rates of events toevaluate safety, highly sensitive methods are needed.

D-dimer is the most reliable and sensitive biomarker forthromboembolic events.14 It is regarded as a valuable bio-marker for diagnosing and managing a vast array ofthrombosis-related clinical conditions.15 D-dimer is a spe-cific degradation product of fibrin clots. This small proteinfragment, which is present in blood clots degraded by fi-brinolysis, is known as ‘‘D-dimer’’ because it contains 2crosslinked D-fragments of fibrinogen. D-dimer levels areelevated when the coagulation system is activated, such asin cases of thrombosis. Measurement of D-dimer levels is agold standard for diagnosing thrombosis and identifyingincreased risk of thrombotic events.15–19 Elevated serum D-dimer levels indicate a high risk of thrombosis.20 Manylong-term prospective studies have demonstrated an asso-ciation between D-dimer and thromboembolism or coronaryheart disease.21

In this study, we evaluated serum D-dimer levels afterintravitreal injection of bevacizumab or ranibizumab inpatients with AMD.

Methods

This prospective, nonrandomized study examined patientswho received 3 monthly intravitreal injections of bev-acizumab or ranibizumab for wet AMD between March andNovember 2010. The decision about which drug to use wasmade by the physician (D. Jee), taking into account thepatient’s wish, economic status, and insurance status. Weexcluded patients with ocular neovascularization or rubeosisiridis, or who had undergone intraocular surgery within thelast 6 months. Patient with underlying thromboembolicdisease was defined as those with present or past history forstroke, coronary heart disease, and venous thromboembolicdisease. The study design followed the tenets of the De-claration of Helsinki for biomedical research and was ap-proved by the institutional review board of the CatholicUniversity of Korea in Seoul, Korea. Informed consent wasobtained from the patients. To evaluate the change of D-dimer levels after injection, we assume the large effect size(d = 0.8) as a clinically important difference. Using 2-sidedsignificance level 0.05 and power 80%, we obtained *60eyes per each group.

All patients received an intravitreal injection monthly for 3months. We collected blood samples from patients to exam-ine serum D-dimer levels before the initial injection and 1day, 1 week, 1 month, and 3 months after initial injection(Fig. 1). D-dimer levels were measured by a medical tech-nologist who is masked to the clinical status of patients orinjected drugs in a clinical laboratory using VIDAS D-dimerELISA method (VIDAS, BioMerieux, craponne, France).17

All patients were instilled with proparacaine (Alcaine;Alcon, Fort Worth, TX) and had skin draping that contained10% povidone. After inserting a speculum, a 5% povidonesolution was applied to the conjunctival sac. After steriledrape placement, the conjunctiva was displaced and bev-acizumab (1.25 mg in 0.05 mL) or ranibizumab (0.05 mg in0.05 mL) was injected perpendicular to the sclera at the in-ferotemporal position of the pars plana (3.5 mm posterior tothe surgical limbus) by using a 30-gauge needle attached toa 1-mL tuberculin syringe. Using an operating microscope,

the location of the needle tip was identified as being in thecenter of the vitreous cavity.

All patients underwent clinical examinations, includingbest-corrected visual acuity measurement, slit-lamp biomi-croscopy, intraocular pressure measurement, and fundusexamination.

We assessed differences between the serum D-dimer levelsbefore and after intravitreal injection within each group byrepeated measured analysis of variance (ANOVA) by usingcommercial software (SPSS version 17.0; SPSS, Inc., Chicago,IL). We also evaluated differences in D-dimer levels betweenthe groups receiving bevacizumab and ranibizumab by re-peated measured ANOVA. P < 0.05 was considered statisti-cally significant.

Results

We enrolled 122 eyes from 122 patients who receivedintravitreal injections of bevacizumab or ranibizumab fromMarch to November 2010. All eyes were divided into 2groups: bevacizumab (62 eyes) and ranibizumab (60 eyes).The bevacizumab and ranibizumab groups were well-bal-anced overall for age, gender, and underlying arterialthromboembolic disease (Table 1).

In both groups, D-dimer levels showed no significantchange before and after injection at any point (Fig. 2). Nosignificant difference was seen between the bevacizumaband ranibizumab groups (Table 2).

FIG. 1. Timing of intravitreal injection of anti–vascularendothelial growth factor and blood sampling to examine se-rum D-dimer levels. Intravitreal injection was performedmonthly for 3 months, and blood sampling was conductedbefore 1st injection, and 1 day, 1 week, 1 month, and 3 monthsafter initial injection.

Table 1. Clinical Characteristics

of Subjects in the Study

Bevacizumab(n = 62)

Ranibizumab(n = 60) P

Age (years) 68.33 – 6.27 69.36 – 8.99 0.463Gender (male/female) 38/24 40/20 0.5364Diabetes 11 9 0.682Hypertension 28 33 0.277Smoking 18 21 0.479Underlying

thromboembolicdiseases

12 (19.3%) 11 (18.3%) 0.785

Stroke 4 6 0.475Coronary heart

disease10 8 0.663

Venousthromboembolicdisease

1 2 0.549

2 JEE, ZAKO, AND LA

Subgroup analysis for patients with risk of thromboembolicdiseases was performed. D-dimer levels showed no significantchange after ranibizumab injection, but D-dimer levels at 1 dayand 1 week after bevacizumab injection were significantlyhigher than that before injection (P = 0.041 and P = 0.022, re-spectively; Fig. 3). D-dimer levels were significantly higher at1 day (P = 0.038) and 1 week (P = 0.017, Table 3) after bev-acizumab injection than after ranibizumab injection.

During the follow-up period, none of the patients developedany severe local adverse events, such as endophthalmitis,uveitis, or retinal tears, or systemic side effects, such as cere-brovascular attacks or myocardial infarction.

Discussion

Our study demonstrated that D-dimer levels were un-changed after intravitreal injections of either bevacizumab orranibizumab. In patients with thromboembolic risk factors,intravitreal injections of bevacizumab resulted in transientincreases in D-dimer levels. Although the exact mechanismof D-dimer elevation by bevacizumab remains to be eluci-dated, we have several hypotheses.

One possible reason is the entry of different amounts ofanti-VEGF agents into systemic circulation. Intravitreallyadministered bevacizumab might be transferred across theblood-retinal barrier into the systemic vascular system, whileranibizumab might remain localized in the injected eye. In arecent study in which the pharmacokinetics of intravitreallyadministered ranibizumab and bevacizumab were comparedin rabbits, ranibizumab was not detected in serum, while

small amounts of bevacizumab were detected.22,23 In anotherstudy, it was reported that intravitreal bevacizumab appearedin plasma and in the fellow eye via systemic circulation.24

The pharmacokinetic difference in these 2 anti-VEGFagents could be due to differences in molecular structure.Bevacizumab is a full-length antibody, while ranibizumab isa Fab fragment. Ranibizumab does not possess the fragmentcrystallizable (Fc) region, which is important in transferringintravitreally injected full-length immunoglobulin G (IgG)across the blood-retinal barrier into the systemic blood sys-tem.25 In a study using Fc-receptor-knockout mice, IgGwithout Fc receptor did not penetrate the blood-retinal barrier.Thus, bevacizumab may enter the systemic circulation moreeasily. Systemic bevacizumab might disrupt physiologicalVEGF function, which could lead to increased D-dimer lev-els. A recent report demonstrated significantly reduced VEGFplasma levels after intravitreal bevacizumab injection; thesewere not seen after intravitreal ranibizumab injection.26

Another possible reason is that the half-life of ranibizu-mab is shorter than of that of bevacizumab when exposed tosystemic circulation. Antigen-binding fragments, such asranibizumab, tend to have a shorter half-life than full-lengthantibodies, such as bevacizumab. The systemic half-life of aFab is a few hours, whereas the half-life of full-length IgG is*3 weeks.27 The longer half-life of bevacizumab might beinvolved in the transient increase in serum D-dimer levels inthis study.

FIG. 2. D-dimer levels after intravitreal injection of bev-acizumab (B) or ranibizumab (�). No significant changewas detected at any time after intravitreal injection in eitherbevacizumab or ranibizumab groups. Repeated measuredanalysis of variance (ANOVA).

Table 2. D-Dimer (mg/L) Levels in Patients

After Bevacizumab or Ranibizumab Injection

Bevacizumab(n = 62)

Ranibizumab(n = 60) P

Preoperative 182 – 206 206 – 127 0.345Postoperative 1 day 215 – 325 192 – 194 0.1061 week 231 – 280 211 – 233 0.6721 month 210 – 167 185 – 241 0.0893 months 192 – 227 208 – 316 0.715

Continuous variables are expressed as n or mean – standarddeviation. Repeated measured analysis of variance (ANOVA).

FIG. 3. D-dimer levels after intravitreal injection of bev-acizumab (B) or ranibizumab (�) in patients with under-lying arterial thromboembolic disease. D-dimer levels werenot significantly changed after ranibizumab injection, but D-dimer levels at 1 day and 1 week after bevacizumab injec-tion significantly increased compared with before injection.*P < 0.05, repeated measured ANOVA.

Table 3. D-Dimer (mg/L) Levels in Patients

with Underlying Arterial Thromboembolic

Disease After Bevacizumab or Ranibizumab Injection

Bevacizumab(n = 12)

Ranibizumab(n = 11) P

Preoperative 215 – 242 221 – 264 0.654Postoperative 1 day 356 – 225 238 – 165 0.038a

1 week 412 – 268 217 – 192 0.017a

1 month 327 – 274 263 – 271 0.1473 months 281 – 266 241 – 304 0.526

Continuous variables are expressed as n or mean – standarddeviation.

aP < 005, repeated measured ANOVA.

D-DIMER AFTER ANTI-VEGF INJECTION 3

The exact mechanism of how anti-VEGF agents inducethromboembolism is unclear. One possible mechanism isthat suppression of physiological VEGF causes damage tovascular endothelial cells, especially to the balanced coa-gulative and fibrinolytic systems. This develops intothromboembolism, resulting in elevated D-dimer levels. Ourhypothesis is supported by the finding that D-dimer levelsare positively associated with VEGF levels.28,29

AMD is associated with strokes.30,31 The AtherosclerosisRisk in Communities Study showed an almost 2-fold risk ofstroke in AMD patients over 10 years (hazard ratio, 1.87;95% confidence interval [CI], 1.21–2.88).32 Hu et al. re-ported that neovascular AMD is associated with a higherrisk of stroke over 5 years (hazard ratio, 2.42; 95% CI, 1.47–3.98).30 Thus, the suppression of physiological VEGF byintravitreal injection of anti-VEGF for AMD treatmentcould induce thromboembolic events, because patients withAMD is more susceptible to stroke.

D-dimer levels were not elevated at 3 months after in-travitreal injection in this study. Our study had a schedule of3 monthly injections as a loading dose. Cumulative dosesfrom 3 monthly injections might have an adverse effect onD-dimer levels. However, we found no cumulative effectfrom repeated injections. We presume that D-dimer levelsmight be normalized after repeated and regular inhibition ofVEGF by a complex feedback mechanism. In a recent study,pro re nata (PRN; as needed) anti-VEGF therapy was as-sociated with more thromboembolic events than a monthlyregimen.33 This implies that abrupt inhibition of physiologicVEGF is associated with higher risk of thromboembolismthan repeated inhibition at regular intervals.

The risk of thromboembolism after intravitreal injectionof anti-VEGF has been described in previous reports.7,34,35

Comparison of AMD Treatments Trials (CATT) study—arandomized, controlled clinical trial—reported that throm-boembolic events were more frequent in a bevacizumabgroup than in a ranibizumab group.33 Our results correspondwith these previous results.

The D-dimer assay has a high negative predictive value. Itis highly sensitive but not specific to thromboembolic dis-ease.36 D-dimer levels can also be elevated in malignancy,pregnancy, recent trauma or surgery, bleeding, and sepsis,because all these diseases have a common underlying pro-cess of thromboembolism.16 Therefore, D-dimer elevationby bevacizumab in patients with underlying thromboem-bolic disease does not indicate an elevated postoperativethromboembolism risk. However, intravitreally injectedbevacizumab and ranibizumab seem to be safe fromthromboembolic complications considering the high nega-tive predictive value of the D-dimer test. D-dimer assays canbe used to exclude a diagnosis of thromboembolic eventsafter intravitreal injection of anti-VEGF.

One limitation of our study is that actual thromboembolicevents were not evaluated. However, a number of previousstudies demonstrated that D-dimer levels are associated withthromboembolism risk.14–18,20 We are continuing to followthe patients in this study to evaluate thromboembolic events.Another limitation is the study design that means a non-randomized design, no clinical trial, and a small number ofcases.

To our knowledge, this is the first study that measured D-dimer levels after intravitreal injection of anti-VEGF agentsin patients with AMD. Intravitreal injection of bevacizumab

and ranibizumab did not increase D-dimer levels overall,whereas D-dimer levels were transiently increased in pa-tients with thromboembolic risk factors after intravitrealinjection of bevacizumab. Intravitreal injection of ranibi-zumab is safe with respect to the risk of thromboemboliccomplications. The results from this study contribute to theunderstanding of systemic safety concerns after intravitrealinjection of anti-VEGF agents for the treatment of AMD.

Author Disclosure Statement

None of the authors have any financial or proprietaryinterest in any material or method mentioned.

References

1. Rosenfeld, P.J., Brown, D.M., Heier, J.S., Boyer, D.S.,Kaiser, P.K., Chung, C.Y., and Kim, R.Y. Ranibizumab forneovascular age-related macular degeneration. N. Engl. J.Med. 355:1419–1431, 2006.

2. Schmucker, C., Ehlken, C., Hansen, L.L., Antes, G.,Agostini, H.T., and Lelgemann, M. Intravitreal bev-acizumab (Avastin) vs. ranibizumab (Lucentis) for thetreatment of age-related macular degeneration: a systematicreview. Curr. Opin. Ophthalmol. 21:218–226, 2010.

3. Brown, D.M., Kaiser, P.K., Michels, M., Soubrane, G., Heier,J.S., Kim, R.Y., Sy, J.P., and Schneider, S. Ranibizumabversus verteporfin for neovascular age-related macular de-generation. N. Engl. J. Med. 355:1432–1444, 2006.

4. Schouten, J.S., La Heij, E.C., Webers, C.A., Lundqvist, I.J.,and Hendrikse, F. A systematic review on the effect ofbevacizumab in exudative age-related macular degenera-tion. Graefes Arch. Clin. Exp. Ophthalmol. 247:1–11, 2009.

5. Ferrara, N., Hillan, K.J., Gerber, H.P., and Novotny, W.Discovery and development of bevacizumab, an anti-VEGFantibody for treating cancer. Nat. Rev. Drug Discov. 3:391–400, 2004.

6. Ferrara, N., Damico, L., Shams, N., Lowman, H., and Kim,R. Development of ranibizumab, an anti-vascular endo-thelial growth factor antigen binding fragment, as therapyfor neovascular age-related macular degeneration. Retina.26:859–870, 2006.

7. Csaky, K., and Do, D.V. Safety implications of vascularendothelial growth factor blockade for subjects receivingintravitreal anti-vascular endothelial growth factor thera-pies. Am. J. Ophthalmol. 148:647–656, 2009.

8. Schmidt-Erfurth, U. Clinical safety of ranibizumab in age-related macular degeneration. Expert. Opin. Drug Saf. 9:149–165, 2010.

9. Kilickap, S., Abali, H., and Celik, I. Bevacizumab, bleed-ing, thrombosis, and warfarin. J. Clin. Oncol. 21:3542,2003; author reply 3543.

10. Ueta, T., Yanagi, Y., Tamaki, Y., and Yamaguchi, T.Cerebrovascular accidents in ranibizumab. Ophthalmology.116:362, 2009.

11. Alexander, S.L., Linde-Zwirble, W.T., Werther, W., Dep-perschmidt, E.E., Wilson, L.J., Palanki, R., Saroj, N., But-terworth, S.L., and Ianchulev, T. Annual rates of arterialthromboembolic events in medicare neovascular age-relatedmacular degeneration patients. Ophthalmology. 114:2174–2178, 2007.

12. Mannucci, P. Venous thrombosis: the history of knowledge.Pathophysiol. Haemost. Thromb. 32:209–212, 2002.

13. Thorneycroft, I.H., and Goldzieher, J.W. Venous throm-boembolism: a review. J. Reprod. Med. 48(11 Suppl):911–920, 2003.

4 JEE, ZAKO, AND LA

14. Lippi, G., Cervellin, G., Franchini, M., and Favaloro, E.J.Biochemical markers for the diagnosis of venous throm-boembolism: the past, present and future. J. Thromb.Thrombolysis. 30:459–471, 2010.

15. Hargett, C.W., and Tapson, V.F. Clinical probability andD-dimer testing: how should we use them in clinicalpractice? Semin. Respir. Crit. Care Med. 29:15–24, 2008.

16. Tripodi, A. D-dimer testing in laboratory practice. Clin.Chem. 57:1256–1262, 2011.

17. Cosmi, B., and Palareti, G. Update on the predictive valueof D-dimer in patients with idiopathic venous thrombo-embolism. Thromb. Res. 125(Suppl 2):S62–S65, 2010.

18. Verhovsek, M., Douketis, J.D., Yi, Q., Shrivastava, S., Tait,R.C., Baglin, T., Poli, D., and Lim, W. Systematic review:D-dimer to predict recurrent disease after stopping antico-agulant therapy for unprovoked venous thromboembolism.Ann. Intern. Med. 149:481–490, W494, 2008.

19. Brill-Edwards, P., and Lee, A. D-dimer testing in the di-agnosis of acute venous thromboembolism. Thromb. Hae-most. 82:688–694, 1999.

20. Wada, H., Kobayashi, T., Abe, Y., Hatada, T., Yamada, N.,Sudo, A., Uchida, A., and Nobori, T. Elevated levels ofsoluble fibrin or D-dimer indicate high risk of thrombosis.J. Thromb. Haemost. 4:1253–1258, 2006.

21. Lowe, G.D. Can haematological tests predict cardiovascu-lar risk? The 2005 Kettle Lecture. Br. J. Haematol. 133:232–250, 2006.

22. Bakri, S.J., Snyder, M.R., Reid, J.M., Pulido, J.S., Ezzat,M.K., and Singh, R.J. Pharmacokinetics of intravitreal rani-bizumab (Lucentis). Ophthalmology. 114:2179–2182, 2007.

23. Bakri, S.J., Snyder, M.R., Reid, J.M., Pulido, J.S., andSingh, R.J. Pharmacokinetics of intravitreal bevacizumab(Avastin). Ophthalmology. 114:855–859, 2007.

24. Nomoto, H., Shiraga, F., Kuno, N., Kimura, E., Fujii, S.,Shinomiya, K., Nugent, A.K., Hirooka, K., and Baba, T.Pharmacokinetics of bevacizumab after topical, sub-conjunctival, and intravitreal administration in rabbits. In-vest. Ophthalmol. Vis. Sci. 50:4807–4813, 2009.

25. Kim, H., Robinson, S.B., and Csaky, K.G. FcRn receptor-mediated pharmacokinetics of therapeutic IgG in the eye.Mol. Vis. 15:2803–2812, 2009.

26. Carneiro, A.M., Costa, R., Falcao, M.S., Barthelmes, D.,Mendonca, L.S., Fonseca, S.L., Goncalves, R., Goncalves, C.,Falcao-Reis, F.M., and Soares, R. Vascular endothelial growthfactor plasma levels before and after treatment of neovascularage-related macular degeneration with bevacizumab or rani-bizumab. Acta Ophthalmol. 90:e25–e30, 2011.

27. Meyer, C.H., and Holz, F.G. Preclinical aspects of anti-VEGF agents for the treatment of wet AMD: ranibizumaband bevacizumab. Eye (Lond). 25:661–672, 2011.

28. Matsuyama, W., Hashiguchi, T., Mizoguchi, A., Iwami, F.,Kawabata, M., Arimura, K., and Osame, M. Serum levels ofvascular endothelial growth factor dependent on the stageprogression of lung cancer. Chest. 118:948–951, 2000.

29. Tseng, C.S., Lo, H.W., Teng, H.C., Lo, W.C., and Ker,C.G. Elevated levels of plasma VEGF in patients withdengue hemorrhagic fever. FEMS Immunol. Med. Micro-biol. 43:99–102, 2005.

30. Hu, C.C., Ho, J.D., and Lin, H.C. Neovascular age-relatedmacular degeneration and the risk of stroke: a 5-year pop-ulation-based follow-up study. Stroke. 41:613–617, 2010.

31. Sun, C., Klein, R., and Wong, T.Y. Age-related maculardegeneration and risk of coronary heart disease and stroke:the Cardiovascular Health Study. Ophthalmology.116:1913–1919, 2009.

32. Wong, T.Y., Klein, R., Sun, C., Mitchell, P., Couper, D.J.,Lai, H., Hubbard, L.D., and Sharrett, A.R. Age-relatedmacular degeneration and risk for stroke. Ann. Intern. Med.145:98–106, 2006.

33. Martin, D.F., Maguire, M.G., Ying, G.S., Grunwald, J.E.,Fine, S.L., and Jaffe, G.J. Ranibizumab and bevacizumabfor neovascular age-related macular degeneration. N. Engl.J. Med. 364:1897–1908, 2011.

34. Tolentino, M. Systemic and ocular safety of intravitrealanti-VEGF therapies for ocular neovascular disease. Surv.Ophthalmol. 56:95–113, 2011.

35. Schmucker, C., Ehlken, C., Agostini, H.T., Antes, G.,Ruecker, G., Lelgemann, M., and Loke, Y.K. A safetyreview and meta-analyses of bevacizumab and ranibizu-mab: off-label versus goldstandard. PloS One. 7:e42701,2012.

36. Wells, P.S., Anderson, D.R., Rodger, M., Forgie, M.,Kearon, C., Dreyer, J., Kovacs, G., Mitchell, M., Le-wandowski, B., and Kovacs, M.J. Evaluation of D-dimer inthe diagnosis of suspected deep-vein thrombosis. N. Engl.J. Med. 349:1227–1235, 2003.

Received: February 10, 2014Accepted: August 10, 2014

Address correspondence to:Prof. Tae Yoon La

College of MedicineCatholic University of Korea

Seoul 137-701Republic of Korea

E-mail: laty@catholic.ac.kr

D-DIMER AFTER ANTI-VEGF INJECTION 5