Purpose
To evaluate the incidence of and risk factors for subretinal hemorrhages in age-related macular degeneration (AMD) patients on anticoagulation or antiplatelet therapy.
Design
Retrospective, observational case series.
Methods
We retrospectively reviewed the medical and photographic records of 71 consecutive patients who sought treatment at our institution with acute subretinal hemorrhages complicating age-related macular degeneration. The size of the subretinal hemorrhage was measured in standardized Macular Photocoagulation Study disc areas. Data on the use of medications and medical indications for anticoagulation and antiplatelet therapy were obtained.
Results
Overall, patients receiving antithrombotic therapy had a significantly larger subretinal hemorrhage size (mean, 9.71 disc areas) than patients not receiving anticoagulant or antiplatelet therapy (mean, 2.99 disc areas). Subgroup analysis revealed that both antiplatelet ( P < .0001) and anticoagulant therapy ( P = .003) were associated with a significantly larger bleeding size. Moreover, subgroup analysis among patients with arterial hypertension revealed that individuals receiving antithrombotic therapy had a statistically significantly larger hemorrhage size than hypertensive patients who did not receive anticoagulants or antiplatelet agents ( P < .0001).
Conclusions
Our results indicate that anticoagulants and antiplatelet agents are strongly associated with the development of large subretinal hemorrhages in AMD patients. Moreover, arterial hypertension is a strong risk factor for large subretinal hemorrhages in AMD patients receiving anticoagulants or antiplatelet agents. Physicians should be aware of an increased risk of extensive subretinal hemorrhage in AMD patients when deciding on the initiation and duration of anticoagulant and antiplatelet therapy.
Exudative age-related macular degeneration (AMD) is among the conditions that have been shown to be associated with a risk of subretinal hemorrhages. Furthermore, patients with thick submacular hemorrhage complicating AMD typically have a poor visual prognosis. Antiplatelet therapy with aspirin, clopidogrel, or ticlopidine has established benefits in the secondary prevention of fatal and nonfatal coronary and cerebrovascular events. Anticoagulation frequently is used in this elder age group for a variety of other comorbidities, including prosthetic heart valves, atrial fibrillation, ischemic heart disease, cerebrovascular disease, and venous thromboembolism. However, it is well established that the longer patients remain on anticoagulant therapy, the higher their cumulative risk of bleeding.
Over the past years, there has been rapidly growing literature concerning the risk of hemorrhagic ocular complications with ophthalmic surgery in patients receiving anticoagulation therapy. By contrast, there are still few data on the relation between anticoagulation or antiplatelet therapy and spontaneous ocular hemorrhages. However, only few reports have focused on patients with age-related macular degeneration. To date, there are no studies quantifying severity of subretinal hemorrhages associated with anticoagulants or antiplatelets in ophthalmic patients.
In the present study, we sought to investigate whether there is a relationship between various anticoagulant or antiplatelet agents and the development of subretinal hemorrhages in patients with age-related macular degeneration. Moreover, in light of the substantial variability in hemorrhage size, we attempted to quantify the severity of these bleeding complications.
Methods
We retrospectively reviewed the medical and photographic records of 71 consecutive patients who sought treatment at our institution with subretinal hemorrhages complicating AMD. All patients had acute hemorrhages (< 1 week old) and reading vision in the affected eye before the onset of hemorrhage. If both eyes of the same patient met the inclusion criteria, only the first affected eye was selected for further analysis to avoid possible bias. Patients were excluded whenever pathologic myopia, ocular trauma, or infection was present.
Each patient underwent a complete bilateral ophthalmologic examination, including best-corrected visual acuity, intraocular pressure, and anterior segment slit-lamp examination. Stereoscopic color fundus photographs from the initial presentation were available for all patients. Patients with massive vitreoretinal hemorrhage underwent pars plana vitrectomy. In these cases, color fundus photographs were obtained during surgery. The initial hemorrhage size was measured in standardized Macular Photocoagulation Study (MPS) disc areas to indicate specific absolute areas, not an area relative to an individual’s area. A clear piece of acetate with a millimeter rule and MPS disc area circles of different sizes were placed over film images to measure the greatest disc areas of hemorrhages. Each subject’s weight and height were recorded, and resting blood pressure was measured. Data on demographic and lifestyle variables; medical history, including history of diabetes mellitus, arterial hypertension, hyperlipidemia, smoking habits, and alcohol consumption; the use of medications, including acetylsalicylic acid, clopidogrel, ticlopidine, or warfarin; and medical indications for antithrombotic agents were obtained.
Testing for any relationship between anticoagulation or antiplatelet agents and the size of subretinal hemorrhages was performed using the Mann–Whitney U test. All tests were 2-tailed, and acceptable significance was recorded when P values were less than .05. Analyses were performed with the SAS statistical software package (SAS Institute Inc, Cary, North Carolina, USA). Potential risk factors including age, gender, arterial hypertension, smoking, diabetes mellitus, and anticoagulants or antiplatelet agents were entered in a multivariate logistic regression analysis. We used a cutoff P value of .1 for the risk factors in the multivariate model. The severity of subretinal hemorrhage was graded according to the following scheme: moderate subretinal hemorrhage (< 4.5 MPS disc areas, i.e., less than 3-fold median bleeding size among our patients) and severe hemorrhage (≥ 4.5 MPS disc areas, i.e., 3-fold median bleeding size or more). Multivariate logistic regression analysis was performed with the BIAS statistical software package (version 8.0.5; Epsilon, Darmstadt Hochheim, Germany).
Results
Seventy-one patients (71 eyes; 25 males and 46 females; mean age, 75.8 years; range, 50 to 94 years) were included in the study. The mean size of subretinal hemorrhages was 5.63 disc areas (range, 0.1 to 32 disc areas; median, 1.5 disc areas). The mean hemorrhage size among patients with severe subretinal hemorrhages (≥ 4.5 disc areas) was 14.1 MPS disc areas (range, 5 to 32 MPS disc areas; median, 10.0 disc areas) compared with a mean hemorrhage size of 1.1 MPS disc areas (range, 0.1 to 4.0 disc areas; median, 0.5 disc areas) among patients with moderate hemorrhages (< 4.5 disc areas). Bilateral subretinal hemorrhages occurred in 3 patients. Massive vitreoretinal hemorrhage was observed in 3 patients. In these cases, measurement revealed large subretinal hemorrhages of 15, 25, and 25 MPS disc areas, respectively. Of all patients enrolled in this study, 40 (56.3%) had visual acuities of less than 20/50 in the partner eye, including 26 (36.6%) eyes with vision of 20/200 or less. Only 2 patients had normal visual acuity (20/20) in the fellow eye.
Twenty-eight patients (39.4%) had been receiving antithrombotic therapy before the development of subretinal hemorrhage, and 43 patients (60.6%) were not receiving antithrombotic therapy. Four (14.3%) of 28 patients were receiving anticoagulant therapy without antiplatelet medication, 24 (85.7%) were receiving antiplatelet agents alone, and none of the patients were receiving both anticoagulants and antiplatelet agents. The medical indications of these patients are summarized in the Table . Patients receiving warfarin therapy received extended anticoagulant treatment for more than 3 months with a recommended target international normalized ratio (INR) of 2.0 to 3.0.
| General Medical Diagnoses | No. of Patients Receiving APT | No. of Patients Receiving ACT | No. of Patients without APT or ACT |
|---|---|---|---|
| Arrhythmia | 1 | 0 | 3 |
| Underwent valve replacement | 1 | 0 | 0 |
| History of cerebral vascular incident | 1 | 0 | 0 |
| History of thromboembolic event | 0 | 1 | 1 |
| Coronary artery disease or history of myocardial infarction | 2 | 2 | 5 |
| Hypertension | 17 | 2 | 19 |
| Diabetes mellitus | 2 | 0 | 5 |
| Hypercholesterolemia | 0 | 0 | 2 |
| Autoimmune disorder | 0 | 0 | 1 |
| Unremarkable medical history | 5 | 1 | 14 |
The overall prevalence of subretinal hemorrhages associated with anticoagulation or antiplatelet therapy was 39.4% (n = 28). We observed a strong association between the initial size of the subretinal hemorrhage and antithrombotic agent intake ( P < .0001). Individuals receiving anticoagulants or antiplatelet medication had a statistically significantly larger size of subretinal hemorrhage (mean, 9.71 disc areas) than patients without anticoagulant or antiplatelet therapy (mean, 2.99 disc areas). Subgroup analysis revealed that antiplatelet agents ( P < .0001) as well as anticoagulants ( P = .003) were associated with a significantly larger bleeding size.
Among all patients, the prevalence of arterial hypertension was 53.5%. Overall, the size of the subretinal hemorrhages was not significantly different between patients with and without arterial hypertension ( P = .73). However, subgroup analysis among the 38 patients with arterial hypertension revealed that individuals receiving anticoagulants or antiplatelet therapy had a statistically significant larger size of subretinal hemorrhage than patients with hypertension who were not receiving anticoagulant or antiplatelet therapy ( P < .0001) ( Figure ). Interestingly, all 3 patients with massive vitreoretinal hemorrhages at presentation were receiving antiplatelet therapy associated with arterial hypertension. Of the 3 patients with bilateral subretinal hemorrhages at presentation, 2 had arterial hypertension and were receiving antithrombotic therapy.
Multivariate analysis using a stepwise logistic regression analysis revealed chronic oral anticoagulation therapy or antiplatelet therapy as independent risk factors for the development of severe subretinal hemorrhage associated with age-related macular degeneration ( P = .0004; odds ratio, 7.99; 95% confidence interval, 2.49 to 25.66).
Results
Seventy-one patients (71 eyes; 25 males and 46 females; mean age, 75.8 years; range, 50 to 94 years) were included in the study. The mean size of subretinal hemorrhages was 5.63 disc areas (range, 0.1 to 32 disc areas; median, 1.5 disc areas). The mean hemorrhage size among patients with severe subretinal hemorrhages (≥ 4.5 disc areas) was 14.1 MPS disc areas (range, 5 to 32 MPS disc areas; median, 10.0 disc areas) compared with a mean hemorrhage size of 1.1 MPS disc areas (range, 0.1 to 4.0 disc areas; median, 0.5 disc areas) among patients with moderate hemorrhages (< 4.5 disc areas). Bilateral subretinal hemorrhages occurred in 3 patients. Massive vitreoretinal hemorrhage was observed in 3 patients. In these cases, measurement revealed large subretinal hemorrhages of 15, 25, and 25 MPS disc areas, respectively. Of all patients enrolled in this study, 40 (56.3%) had visual acuities of less than 20/50 in the partner eye, including 26 (36.6%) eyes with vision of 20/200 or less. Only 2 patients had normal visual acuity (20/20) in the fellow eye.
Twenty-eight patients (39.4%) had been receiving antithrombotic therapy before the development of subretinal hemorrhage, and 43 patients (60.6%) were not receiving antithrombotic therapy. Four (14.3%) of 28 patients were receiving anticoagulant therapy without antiplatelet medication, 24 (85.7%) were receiving antiplatelet agents alone, and none of the patients were receiving both anticoagulants and antiplatelet agents. The medical indications of these patients are summarized in the Table . Patients receiving warfarin therapy received extended anticoagulant treatment for more than 3 months with a recommended target international normalized ratio (INR) of 2.0 to 3.0.
