Age-related macular degeneration (AMD) affects approximately 20% of people over the age of 65 years in the United States; this proportion increases with age.1 While only about 10% of all patients with AMD have the neovascular, or “wet,” form of the disease, most of the visual impairment and blindness attributable to AMD is due to neovascular AMD.1 Of the patients who become legally blind from AMD, 80% to 90% of them are affected by neovascular AMD.1,2 Intraocular injections of anti–vascular endothelial growth factor (anti-VEGF) agents, including pegaptanib (Macugen; Valeant Pharmaceuticals, West Laval, Canada), ranibizumab (Lucentis; Genentech, South San Francisco, CA), and bevacizumab (Avastin; Genentech, South San Francisco, CA), have revolutionized the treatment of neovascular AMD.3–8 Ocular complications of intravitreal anti-VEGF injections have been reported as infrequent but can be severe and may include intraocular hemorrhage, endophthalmitis, lens injury, retinal detachment, intraocular inflammation, intraocular pressure (IOP) elevation, cataract progression, central retinal artery occlusion, and retinal pigment epithelium (RPE) tears.9,10 It has been reported that patients with neovascular AMD who are on daily antiplatelet or anticoagulant therapy have significantly more intraocular hemorrhages than patients with neovascular AMD who are not on antiplatelet or anticoagulant therapy.11
There is little information in the literature regarding the association of systemic anticoagulation with the risk of intraocular hemorrhage in patients being treated with anti-VEGF agents. Active use of anticoagulant or antiplatelet therapy is not currently a contraindication to intraocular injection. Mason et al reported no intraocular hemorrhages among 520 patients who received a total of 3,106 bevacizumab or ranibizumab injections while taking warfarin (548 injections), clopidogrel (523 injections), aspirin (1,254 injections), warfarin plus clopidogrel (33 injections), or no anticoagulants (2,002 injections).10 The patients had a variety of indications for anti-VEGF treatment including AMD, diabetes mellitus, cystoid macular edema, diabetic macular edema, branch retinal vein occlusion, and central retinal vein occlusion. There were no injection-related or immediate post-injection-related hemorrhagic complications in a small retrospective study of 31 patients treated with anti-VEGF therapy for AMD while taking warfarin.12 Loukopoulos et al evaluated 149 injections performed in 50 patients who received intravitreal injections of ranibizumab and were instructed to continue taking their warfarin as prescribed; all patients had an international normalized ratio (INR) measured on the day before or the day of injection.13 The average INR of study participants was 2.32. No hemorrhagic events were reported.
The current study investigates the association between systemic anticoagulant medication usage at the time of intravitreal anti-VEGF injection and the rate of post-injection intraocular hemorrhage among patients with AMD.
Patients and Methods
The study protocol was approved by the Pennsylvania State College of Medicine institutional review board. A computerized database search was performed to identify all patients treated with intravitreal pegaptanib, ranibizumab, or bevacizumab for neovascular AMD (diagnosis code: 362.52) by retina specialists at the Penn State Hershey Eye Center between January 1, 2004, and December 31, 2010.
The medical records of these patients were reviewed, and data collected included patient age, gender, smoking status, and comorbidities (including diabetes mellitus, hypertension, hyperlipidemia, cardiovascular disease, and any vascular occlusive event such as deep venous thrombosis and cerebrovascular accident). It was noted whether the patient had unilateral or bilateral neovascular AMD, previous intraocular hemorrhage, history of ocular trauma, or prior posterior segment surgery.
For each patient, the type of injection received (pegaptanib, ranibizumab, or bevacizumab) was recorded. If a patient received more than one injection, data were recorded for each injection. The use of aspirin, clopidogrel, or warfarin, as well as other prescription antithrombotics such as cilostazol, dipyridamole, prasugrel, and ticlopidine, was recorded. A list of herbal anticoagulants was obtained from the Natural Medicines Comprehensive Database, and patient usage of these supplements was recorded. Supplements such as multivitamins or AREDS vitamins14 were not included in this database. For patients on warfarin therapy, the value of the most recent INR relative to the time of injection was recorded. In all patients, the IOP and best corrected visual acuity, recorded as the logarithm of the minimum angle of resolution (logMAR), on the day of injection (pre-injection) and the first post-injection follow-up visit, were recorded. In general, patients were seen for follow-up within 1 week of an initial injection and 4 to 12 weeks following a repeat injection. All intraocular hemorrhage events were recorded. The number of days between intraocular injection and hemorrhagic event was also recorded.
Each intravitreal injection was performed by first placing topical proparacaine in the eye to be injected, followed by a lid speculum, additional topical proparacaine, topical antibiotic, and then 5% povidone-iodine. The decision to use post-injection antibiotics was at the discretion of the treating retina specialist. Pegaptanib (0.3 mg/0.09 mL), ranibizumab (0.5 mg/0.05 mL), or bevacizumab (1.25 mg/0.05 mL) was administered intravitreally via a 30-gauge needle in the inferotemporal quadrant between 3.5 and 4 mm posterior to the limbus. Patients were scheduled for a follow-up appointment, at which time an ocular examination was performed, including Snellen visual acuity, IOP measurement, and slit lamp and dilated funduscopic examinations. All post-injection complications were recorded regardless of the length of time since the injection.
At the time of intravitreal anti-VEGF injection, a patient was considered to be on systemic anticoagulant therapy if the patient was taking any of the following medications: warfarin, clopidogrel, aspirin, other non-herbal anticoagulants (ie, other NSAID or dipyramidole), herbal medication with anticoagulant properties (ie, coenzyme Q10, garlic, ginkgo, and vitamin E), or fish oil. Generalized estimating equations (GEE) with a logit link15 were used to assess the dichotomous outcome of intraocular hemorrhage (defined as a patient experiencing choroidal, sub-RPE, subretinal, intraretinal, or vitreous hemorrhage) after intravitreal anti-VEGF injection. GEE with a logit link is an extension of ordinary logistic regression that accounts for the clustering of multiple injections per subject. The effect sizes for the GEE analysis were quantified using the odds ratio (OR) and 95% confidence interval. First, we conducted bivariate analyses to evaluate the association of systemic anticoagulant medication usage, as well as potential confounding variables such as diabetes and coronary artery disease, with intraocular hemorrhage occurring after intravitreal anti-VEGF injection. Next, again using GEE with a logit link, a multivariate model was constructed consisting of the primary exposure variable of interest, systemic anticoagulant medication usage, and any potential confounder that was found to have a P value less than 0.10 from the bivariate analysis. We were concerned about overfitting the multivariate model due to few intraocular hemorrhage events after injection; therefore, we wanted to adjust for at most two covariates in addition to the primary exposure of systemic anticoagulation medication usage. All hypothesis tests were two-sided, and all analyses were performed using SAS software, version 9.3 (SAS Institute, Cary, NC).
A total of 1,710 anti-VEGF injections were performed in 228 eyes of 191 patients during the study period. The mean age of the patients was 80 years (range: 49 to 98 years). Of the 191 patients, 66 (35%) were men, 37 (19%) had diabetes mellitus, 154 (81%) had hypertension, 28 (15%) had coronary artery disease, 81 (42%) had bilateral neovascular AMD, and 121 (63%) had either a choroidal, sub-RPE, subretinal, intraretinal, or vitreous hemorrhage prior to anti-VEGF therapy. Of the 1,710 injections, 1,042 (61%) were performed in patients who were taking systemic anticoagulants, including 116 (7%) in patients taking warfarin, 120 (7%) in patients taking clopidogrel, 702 (41%) in patients taking aspirin, 156 (9%) in patients taking other anticoagulants (ie, other NSAID or dipyramidole), 249 (15%) in patients taking fish oil, and 105 (6%) in patients taking herbal medication with anticoagulant properties (ie, coenzyme Q10, garlic, ginkgo, and vitamin E). The remaining 668 injections (39%) were performed in patients not taking systemic anticoagulants. Mean pre-injection Snellen visual acuity in logMAR scale was 0.69 (range: 0 to 2.6). Mean post-injection Snellen visual acuity in logMAR scale was 0.67 (range: 0 to 2.9). Mean pre-injection IOP was 15.5 mm Hg (range: 7 to 26 mm Hg). Mean post-injection IOP was 15.3 mm Hg (range: 5 to 32 mm Hg).
An intraocular hemorrhage occurred following four of the 1,710 injections (0.25%). Three were vitreous hemorrhages, all of which occurred in patients taking systemic anticoagulants. One vitreous hemorrhage occurred in a patient taking aspirin and clopidogrel (patient A, Table 1); one vitreous hemorrhage occurred in a patient taking warfarin and an NSAID (patient B, Table 1); and one vitreous hemorrhage occurred in a patient on NSAID monotherapy (patient C, Table 1). Subretinal hemorrhage occurred in one patient who was not on anticoagulant therapy (patient D, Table 1). In a bivariate analysis, the odds of an intraocular hemorrhage occurring after intravitreal anti-VEGF injection are 1.9 times higher for injections performed in patients on systemic anticoagulation compared to injections performed in patients not on systemic anticoagulation; this difference is not statistically significant (OR = 1.9; 95% CI [.2, 18.5]; P = .56) (Table 2). According to the bivariate analyses, of the potential confounders, only diabetes mellitus and male gender appear to be mildly associated with intraocular hemorrhage (Table 2). In the multivariate analysis adjusting for diabetic status and gender, there is not a significant association between systemic anticoagulation usage and intraocular hemorrhage (systemic anticoagulation usage: OR = .8, 95% CI [.1, 12.2], P = .88; male gender: OR = 7.9, 95% CI [1.5, 41.0], P = .01; diabetic: OR = 4.5, 95% CI [.3, 60.5], P = .26).
Table 1: Outcomes of Patients With Intraocular Hemorrhagic Events
Table 2: Bivariate Associations of Systemic Anticoagulant Medication Usage and Potential Confounding Variables With Intraocular Hemorrhage After Intravitreal Anti-VEGF Injection
Data from the VISION, FOCUS, ANCHOR, and MARINA trials indicate that the risk of intraocular hemorrhage, while low, is present with the use of intravitreal ranibizumab and pegaptanib.16 In a recent retrospective, longitudinal case control study by Day et al, 6,154 Medicare patients with a diagnosis of neovascular AMD who received at least one anti-VEGF injection were compared to a matched control group of patients with neovascular AMD who had never received any injections.17 This study showed the risk of vitreous hemorrhage was significantly higher in the injection group (0.75% vs 1.54%; P < .01). This study, however, did not distinguish between patients on anticoagulation therapy and patients not on systemic anticoagulation.
In the current study, two of the three vitreous hemorrhages were small hemorrhages in the inferotemporal periphery observed during routine follow-up examinations (patients A and B, Table 1); the hemorrhages were visually insignificant and resolved spontaneously within 1 month with no recurrence after 22 months of follow-up for patient A and 43 months of follow up for patient B. The one visually significant vitreous hemorrhage occurred 13 days after the seventh injection in that eye (patient C, Table 1). The patient had a submacular hemorrhage prior to initiation of anti-VEGF therapy. Visual acuity at that time was 20/70. Based on B-scan and fluorescein angiography conducted at the time of presentation with vitreous hemorrhage, it appeared that the submacular hemorrhage had enlarged and extended into the vitreous. The patient underwent pars plana vitrectomy for a non-clearing vitreous hemorrhage 3 months after the vitreous hemorrhage occurred. Unfortunately, due to the formation of a large disciform scar, the best vision the patient recovered was 2/200 with 2 years of follow-up after the vitreous hemorrhage. The patient who developed a subretinal hemorrhage (patient D, Table 1) presented 152 days after her last injection with a pigment epithelial detachment and hemorrhage involving the macula; visual acuity was not affected (remained stable at 20/400). The patient subsequently developed no light perception vision in that eye due to progressive pseudoexfoliative glaucoma.
Patients commonly use herbal and nutritional supplements. Various studies of adults in the United States have reported that 14% to 18% of survey respondents had taken an herbal or nutritional supplement during the week prior to the survey.18,19 Among patients with intermediate-stage AMD in the AREDS Study, a supplement containing vitamins A, C, E, zinc, and copper was associated with a 25% reduction in the risk of advanced AMD development and a 25% reduction in the risk of severe vision loss.14 However, some herbal and nutritional supplements are also known to have adverse effects. For example, ginkgo biloba is a supplement marketed to improve memory and cognition. It is also known to inhibit platelet aggregation,20 and serious cases of bleeding have been reported in association with its use.21,22 Other commonly used nutritional supplements believed to have anticoagulant properties are omega 3 fatty acids, ginger, vitamin E, and coenzyme Q10. The current study is the first to assess the relationship between the use of herbal and nutritional supplements with anticoagulant properties and intraocular hemorrhage in AMD patients treated with anti-VEGF therapies. None of the patients who suffered a hemorrhagic complication were taking supplements that had anticoagulant properties. However, it is has been demonstrated that patients do not always report taking these supplements,23 and this may have limited our ability to study this relationship.
Additional study limitations include the retrospective nature of the study and the fact that there were very few intraocular hemorrhage events, which occurred after intravitreal anti-VEGF injection. However, this was designed as an exploratory study to provide valuable information with respect to intraocular hemorrhage in patients with AMD on concomitant anticoagulation therapy to guide future research in this area.
The use of systemic anticoagulation therapy is widespread. More than one million individuals in the United States are prescribed warfarin annually.24 In the current study, 116 (60.7%) patients were taking aspirin or prescription anticoagulation at some point during the study period. Because anti-VEGF injections are often administered monthly,3,4,25 discontinuing and restarting anticoagulation in this setting would be challenging and may place patients at substantial risk of thromboembolic events. Fortunately, results of the current study indicate that the risk of intraocular hemorrhage following intravitreal injection of anti-VEGF therapy among patients with AMD is low, and there is no significant difference in the risk of intraocular hemorrhage between patients who are taking anticoagulant medication at the time of injection and patients not on anticoagulation.
- Bressler NM, Bressler SB, Fine SL. Age-related macular degeneration. Surv Ophthalmol. 1988;32(6):375–413. doi:10.1016/0039-6257(88)90052-5 [CrossRef]
- Klein R, Klein BE, Linton KL. Prevalence of age-related maculopathy. The Beaver Dam Eye Study. Ophthalmology. 1992;99(6):933–943.
- Brown DM, Kaiser PK, Michels M, et al. Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N Engl J Med. 2006;355(14):1432–1444. doi:10.1056/NEJMoa062655 [CrossRef]
- Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab for neovascular age-related macular degeneration. N Engl J Med. 2006;355(14):1419–1431. doi:10.1056/NEJMoa054481 [CrossRef]
- Stone E. A very effective treatment for neovascular macular degeneration. N Engl J Med. 2006;355:1493–1495. doi:10.1056/NEJMe068191 [CrossRef]
- Rosenfeld PJ, Rich RM, Lalwani GA. Ranibizumab: Phase III clinical trial results. Ophthalmol Clin North Am. 2006;19(3):361–372.
- Spaide R. Ranibizumab according to need: a treatment for age-related macular degeneration. Am J Ophthalmol. 2007;143(4):679–680. doi:10.1016/j.ajo.2007.02.024 [CrossRef]
- Avery RL, Pieramici DJ, Rabena MD, Castellarin AA, Nasir MA, Giust MJ. Intravitreal Bevacizumab (Avastin) for Neovascular Age-Related Macular Degeneration. Ophthalmology. 2006;113(3):363–372.e5 doi:10.1016/j.ophtha.2005.11.019 [CrossRef]
- Fung AE, Rosenfeld PJ, Reichel E. The International Intravitreal Bevacizumab Safety Survey: using the internet to assess drug safety worldwide. Br J Ophthalmol. 2006;90(11):1344–1349. doi:10.1136/bjo.2006.099598 [CrossRef]
- Mason JO, Frederick PA, Neimkin MG, et al. Incidence of hemorrhagic complications after intravitreal bevacizumab (avastin) or ranibizumab (lucentis) injections on systemically anticoagulated patients. Retina. 2010;30(9):1386–1389. doi:10.1097/IAE.0b013e3181e09739 [CrossRef]
- Kiernan DF, Hariprasad SM, Rusu IM, Mehta SV, Mieler WF, Jager RD. Epidemiology of the association between anticoagulants and intraocular hemorrhage in patients with neovascular age-related macular degeneration. Retina. 2010;30(10):1573–1578. doi:10.1097/IAE.0b013e3181e2266d [CrossRef]
- Dayani PN, Siddiqi OK, Holekamp NM. Safety of intravitreal injections in patients receiving warfarin anticoagulation. Am J Ophthalmol. 2007;144(3):451–453. doi:10.1016/j.ajo.2007.05.006 [CrossRef]
- Loukopoulos V, Meier C, Gerding H. Hemorrhagic complications after intravitreal injections of ranibizumab in patients under coumarin-type anticoagulation. Klin Monbl Augenheilkd. 2010;227(4):289–291. doi:10.1055/s-0029-1245225 [CrossRef]
- Age-Related Eye Disease Study Research Group. A randomized, placebo-controlled, clinical trial of high-dose supplementation with vitamins C and E, beta carotene, and zinc for age-related macular degeneration and vision loss: AREDS report no. 8. Arch Ophthalmol. 2001;119(10):1417–136.
- Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73(1):13–22. doi:10.1093/biomet/73.1.13 [CrossRef]
- Tolentino M. Systemic and ocular safety of intravitreal anti-VEGF therapies for ocular neovascular disease. Surv Ophthalmol. 2011;56(2):95–113. doi:10.1016/j.survophthal.2010.08.006 [CrossRef]
- Day S, Acquah K, Mruthyunjaya P, Grossman DS, Lee PP, Sloan FA. Ocular complications after anti-vascular endothelial growth factor therapy in Medicare patients with age-related macular degeneration. Am J Ophthalmol. 2011;152(2):266–272. doi:10.1016/j.ajo.2011.01.053 [CrossRef]
- Kaufman DW, Kelly JP, Rosenberg L, Anderson TE, Mitchell AA. Recent patterns of medication use in the ambulatory adult population of the United States: the Slone survey. JAMA. 2002;287(3):337–344. doi:10.1001/jama.287.3.337 [CrossRef]
- Kelly JP, Kaufman DW, Kelley K, Rosenberg L, Anderson TE, Mitchell AA. Recent trends in use of herbal and other natural products. Arch Intern Med. 2005;165(3):281–286. doi:10.1001/archinte.165.3.281 [CrossRef]
- Cho HJ, Shon YH, Nam KS. Ginkgolide C inhibits platelet aggregation in cAMP- and cGMP-dependent manner by activating MMP-9. Biol Pharm Bull. 2007;30(12):2340–2344. doi:10.1248/bpb.30.2340 [CrossRef]
- Rosenblatt M, Mindel J. Spontaneous hyphema associated with ingestion of Ginkgo biloba extract. N Engl J Med. 1997;336(15):1108. doi:10.1056/NEJM199704103361518 [CrossRef]
- Matthews MK Jr, . Association of ginkgo biloba with intracerebral hemorrhage. Neurology. 1998;50:1933–1934. doi:10.1212/WNL.50.6.1933 [CrossRef]
- Kaye A, Clarke R, Sabar R, Vig S. Herbal medicines: current trends in anesthesiology practice—a hospital survey. J Clin Anesth. 2000;12(6):468–471. doi:10.1016/S0952-8180(00)00195-1 [CrossRef]
- Fihn SD, McDonell M, Martin D, et al. Risk factors for complications of chronic anticoagulation: a multicenter study. Ann Intern Med. 1993;118(7):511–520. doi:10.7326/0003-4819-118-7-199304010-00005 [CrossRef]
- Martin DF, Maguire MG, Ying GS, Grunwald JE, Fine SL, Jaffe GJ. Ranibizumab and Bevacizumab for Neovascular Age-Related Macular Degeneration. N Engl J Med. 2011;364(20):1897–1908. doi:10.1056/NEJMoa1102673 [CrossRef]
Outcomes of Patients With Intraocular Hemorrhagic Events
|Patient||Hemorrhage Type||No. of Pre-Hemorrhage Injections||Days Between Injection, Hemorrhage Discovery||Medications||Outcome|
|A||Vitreous||3||35||aspirin, clopidogrel||Small vitreous hemorrhage in inferotemporal periphery. No effect on visual acuity; resolved spontaneously.|
|B||Vitreous||1||8||warfarin (INR = 2.5), NSAID||Small vitreous hemorrhage in inferotemporal periphery. No effect on visual acuity; resolved spontaneously.|
|C||Vitreous||7||13||NSAID||Extension of submacular hemorrhage into the vitreous. Vision-impairing; treated with pars plana vitrectomy.|
|D||Subretinal||7||152||No anticoagulation||Pigment epithelial detachment with hemorrhage. No effect on visual acuity. Slowly progressed to no light perception vision due to pseudoexfoliation glaucoma.|
Bivariate Associations of Systemic Anticoagulant Medication Usage and Potential Confounding Variables With Intraocular Hemorrhage After Intravitreal Anti-VEGF Injection
|Predictor||Odds Ratio (95% CI) Associated With Intraocular Hemorrhage After Intravitreal Anti-VEGF Injection||P value|
|Anticoagulant therapy||1.9 (0.2, 18.5)||0.56|
|Male gender||7.2 (0.8, 68.4)||0.09|
|Diabetes mellitus||6.4 (1.0, 41.2)||0.05|
|Hypertension||0.9 (0.1, 8.0)||0.89|
|Coronary artery disease||4.5 (0.6, 32.5)||0.14|
|Bilateral neovascular AMD||2.4 (0.3, 23.7)||0.44|
|Choroidal, sub-RPE, subretinal, intraretinal, or vitreous hemorrhage prior to anti-VEGF therapy||1.4 (0.1, 13.6)||0.77|
|Age (5-year increment)||0.9 (0.5, 1.6)||0.68|
|INR (1-unit increment)||1.3 (0.7, 2.5)||0.40|