Full text of DOI:10.1007/PL00007887

Graefe’s Arch Clin Exp Ophthalmol (2000)
238:482–485 © Springer-Verlag 2000
C L I N I C A L I N V E S T I G AT I O N
Maurits A.D. Tilanus
Willem Vaandrager
Marinus H.M. Cuypers
Adrianus M. Verbeek
Carel B. Hoyng
Relationship between
anticoagulant medication
and massive intraocular hemorrhage
in age-related macular degeneration
Abstract Background: A massive
intraocular hemorrhage in the course massive intraocular hemorrhage and
of age-related macular degeneration
(AMD) is a devastating event. We
set out to determine the role of anti-
was not significant. A patient with a
Received: 1 June 1999
Revised: 25 November 1999
Accepted: 25 November 1999
AMD is 11.6 times more likely to
use anticoagulant medication. It ap-
peared that more than 50% of the pa-
coagulant therapy prescribed for vas- tients in the massive hemorrhage
cular or cardiac indications in the de- group were allowed to stop the anti-
velopment of a massive hemorrhage. coagulant medication.
Methods: A retrospective case-con-
trolled study was conducted of 50
cases of age-related macular degen-
Conclusion: Anticoagulant medica-
tion poses a significant risk in the
development of a massive intraocu-
eration complicated by massive sub- lar hemorrhage in patients with exu-
retinal and vitreous hemorrhage. The dative AMD. Antiplatelet medication
M.A.D. Tilanus ( ) · W. Vaandrager
✉
M.H.M. Cuypers · A.M. Verbeek
C.B. Hoyng
control group consisted of 50 cases
of AMD with small subretinal hem-
poses a less significant risk. Physi-
cians prescribing anticoagulant med-
M.A.D.Tilanus,
orrhage. Results: There was a signif- ication should be informed about the
Institute of Ophthalmology,
University Hospital Nijmegen,
P.O. Box 9101, 6500 HB Nijmegen,
The Netherlands
icant difference in the use of antico-
agulant medication (warfarin sodi-
um) between the groups. The differ-
ence in the use of antiplatelet medi-
cation (aspirin) between the groups
macular status of the patient. the In
case of neovascular AMD, anticoag-
ulant medication should be pre-
scribed only for absolute systemic
indications.
e-mail: m.tilanus@ohk.azn.nl,
Tel: +31–24–3614448,
Fax: +31–24–35440522
with a dense vitreous hemorrhage. As a result of the
higher incidence of cardiovascular problems in elderly
patients, anticoagulant medication is encountered more
Introduction
Age-related macular degeneration (AMD) occurs in 20% frequently in this age group than in the general popula-
of individuals over 65 years of age, and the prevalence tion.
increases to 35% in subjects older than 74 years [2]. The
Anticoagulants appear to be a risk factor for hemor-
majority of patients present with the “dry” or geographic rhagic complications in patients with AMD [4]. We
form. The “serous” or neovascular form accounts for ap- therefore examined the clinical data of 50 patients with
proximately 10%–20% of patients with AMD but makes AMD complicated by a massive subretinal hemorrhage
up 80%–90% of patients who eventually become legally and compared them with an age-matched group of 50 pa-
blind as a result of AMD [2, 6]. Eyes with a massive in- tients with AMD with a small subretinal hemorrhage
traocular hemorrhage in the course of AMD end up (less than one disc diameter). We also reviewed the use
much worse, often with light perception only [1, 4]. A of anticoagulant medication and the indications for all
massive hemorrhage is the result of rupture of vessels patients in both groups.
within an extrachoroidal neovascular membrane, leading
to a dome-shaped subretinal coagulate often combined

483
lant medication was significantly higher in the group of
patients with a massive hemorrhage (P=0.0001); there
was no significant difference in the use of antiplatelet
medication between the two groups.
Patients and methods
We performed a retrospective study of the clinical records of 50
consecutive patients with a massive subretinal hemorrhage com-
bined with a vitreous hemorrhage as a complication of AMD who
had previously undergone diagnostic ultrasonography. The inclu-
sion criterion for this study was a massive subretinal hemorrhage
in combination with a vitreous hemorrhage dense enough to pre-
vent funduscopy in patients known to have AMD. We compared
these patients with a control group of 50 AMD patients with limit-
ed, less than one disc diameter, subretinal hemorrhages.
Diabetic patients and patients with myopia of more than 8
diopters were excluded from this study. All patients in both groups
were diagnosed with neovascular AMD by their referring ophthal-
mologist before the onset of the hemorrhage.
Best corrected visual acuity was obtained using Snellen charts.
Grading of size of the subretinal hemorrhage in the control group
was based on the color fundus photographs. After information
about the general medical condition was obtained for all patients
in both groups, the use of anticoagulant medication and the indica-
tions were evaluated.
The main indications for anticoagulant medication in
the hemorrhage group were previous coronary bypass
surgery (five patients), atrial fibrillation (three patients)
and mechanical heart valve prosthesis (two patients). In
the control group the indications were femoral bypass
(one patient) graft and cardiac aneurysm (one patient). In
the massive hemorrhage group the antiplatelet medica-
tion was mainly prescribed to prevent the recurrence of
an embolic cerebral vascular accident (five patients).
Three patients in the control group received antiplatelet
medication because of atrial fibrillation, in two other pa-
tients this medication was prescribed because of their
general vascular pathology (Table 1).
Ultrasonography demonstrated a dense vitreous hem-
orrhage and a dome or double dome shaped subretinal
hemorrhage with irregular reflectivity in all patients in
the hemorrhage group (Figs. 1, 2).
Statistical analysis was performed with the SAS statistical
analysis software package (SAS Institute, Cary, N. C.).
The thickness of the hemorrhage ranged from 1 to
6 mm; the diameter of the base of the lesion was vari-
able. All the patients in the control group showed a
Results
Of the 50 patients in the hemorrhage group, 19 (38%) small, less than one disc diameter, submacular hemor-
were female and 31 (62%) were male; in the control rhage on funduscopy (Figure 3).
group, 25 patients were female and 25 male. The mean
The median visual acuity within 3 months before the
age in the hemorrhage group was 77 years, range 63–90 hemorrhagic event was 0.1 in both groups. In the mas-
years, SD 6 years; the mean age in the control group was sive hemorrhage group the visual acuity ranged from
76 years, range 64–84 years, SD 4 years. There was no 0.025 to 0.4, in the control group, from 0.05 to 0.4.
statistically significant difference in age between the There was no significant difference in visual acuity be-
groups (P=0.342, Wilcoxon rank sum test).
fore the hemorrhagic event (P=0.784, Wilcoxon rank
In the hemorrhage group 15 patients were using anti- sum test). After the hemorrhagic event, visual acuity was
coagulant medication (warfarin sodium) and 8 patients significantly worse in the massive hemorrhage group as
were taking antiplatelet medication (aspirin). The pa- compared to the control group (mean follow-up 4
tients taking anticoagulant medication all had an INR months, range 3–6 months). The median visual acuity in
value between 3 and 4; the aspirin dosage varied be- the massive hemorrhage group was 0.005 (range
tween 80 and 100 mg per day. In the control group only 0.001–0.05) versus a median visual acuity of 0.1 (range
two patients were on anticoagulant medication; six pa- 0.025–0.3) in the control group (P=0.0001, Wilcoxon
tients used antiplatelet medication. The use of anticoagu- rank sum test).
Table 1 Indication for antico-
Indication
Warfarin
H
Aspirin
H
agulant medication in the hem-
orrhage group (H, n=50) and
the control group (C, n=50)
C
C
No.
%
No.
1
%
No.
%
No.
1
%
Coronary bypass
5
1
2
(10)
(2)
(4)
(2)
1
1
(2)
(2)
(2)
Aortic bifurcation prosthesis
Mechanic heart valve prosthesis
General vascular pathology
Cerebral vascular accident
Atrial fibrillation
1
5
(2)
(10)
2
3
(4)
(6)
1
3
(2)
(6)
1
2
(2)
(4)
Cardiac aneurysm
1
(2)
Femoral bypass graft
1
(2)
Thrombosis in leg (>1year)
1
(2)
15
(30)
8
(16)
6
(12)

484
Fig. 2 Fundus aspect after massive subretinal hemorrhage; vitre-
ous hemorrhage has cleared
Fig. 1 Ultrasound photographs, from three angles, of a massive
subretinal hemorrhage with a vitreous hemorrhage in a patient
with age-related macular degeneration
In the massive hemorrhage group three eyes eventual-
ly lost light perception and ten eyes ended up with a vi-
sual acuity of light perception only.
Fig. 3 ‘Small’ subretinal hemorrhage in a patient with neovascu-
lar AMD
In our study the odds ratio of anticoagulant medication
was 11.6 (95% confidence interval 3.1–44.3). This means
that a patient with a massive hemorrhage is 11.6 times
more likely to receive anticoagulant medication. The odds
ratio of antiplatelet medication was 2.1 (95% confidence
interval 0.7–6.6). In the latter case the value “1” is includ-
ed in the confidence interval, indicating lesser significance
of the odds of antiplatelet medication. On consulting the
patient’s own internist and/or cardiologist we learned that
in 8 (53%) of 15 patients receiving anticoagulant medica-
tion the indications for this kind of therapy were relative
and, if desired, the prescription could be stopped.

485
scribed by several authors [8], but the visual outcome is
poor due to delays between the onset of the hemorrhage
Discussion
Social blindness in AMD is mainly caused (90%) by and the actual vitrectomy, the amount of damage to the
choroidal neovascularization, the so-called “wet” or neo- photoreceptors and the macular disciform degenerative
vascular type. If the choroidal neovascularization is process itself. In this study we described is a significant
complicated by a massive hemorrhage the functional elevated risk for a massive intraocular hemorrhage in pa-
outcome is even worse, with patients often ending up tients with exudative AMD who take anticoagulant medi-
with light perception only [2]. Not all contributing risk cation. The actual INR value at the time of the massive
factors leading to a massive intraocular hemorrhage are hemorrhage was between 3 and 4 in the anticoagulant
known, but age of the patient and extent of the macular medication group. These values are near the upper border
degeneration are among them. Furthermore, high blood of the therapeutic index of anticoagulant medication,
pressure and cardiovascular diseases contribute to this which may further increase the already substantial risk for
risk. In a study by El Baba et al. [3], 19% of patients a massive intraocular hemorrhage. For antiplatelet medi-
with AMD complicated by a massive hemorrhage were cation there also seems to be an elevated risk, albeit less
taking anticoagulant or antiplatelet medication and 40% significant than in the anticoagulant medication group.
had a positive history of systemic hypertension and car-
A better estimate of the relative risk, however, can
diovascular disease. The risk of a hemorrhage in antico- only be established in a cohort study.
agulated patients is related to the level of intensity and
In conclusion, this study confirms the idea that the
the duration of anticoagulant therapy. Increasing age is use of anticoagulant medication is a major risk factor in
also a factor in increased responsiveness to anticoagu- patients with exudative AMD. We propose that in pa-
lants [7]. Once a hemorrhage occurs, anticoagulant ther- tients with exudative AMD anticoagulant therapy should
apy probably intensifies the degree and extent of the be prescribed only for absolute indications such as me-
bleeding [5, 7].
chanical heart valve prosthesis, when antiplatelet medi-
Since the consequences of a massive intraocular hem- cation does not suffice. We advocate that physicians who
orrhage are severe, and since there are few effective thera- prescribe anticoagulant medication always inform them-
peutic options once a hemorrhage has occurred, every risk selves about the ophthalmological status of elderly pa-
factor should be eliminated if possible. The surgical re- tients and, when in doubt, refer a patient to an ophthal-
moval of a massive subretinal hemorrhage has been de- mologist before initiating the medication.
References
1. Benson WE, Spalter HF (1971) Vitreous 3. El Baba A, Jarrett WH, Harbin TS, Fine 7. Lewis H, Sloan SH, Foos RY (1988)
hemorrhage: a review of experimental
and clinical investigations. Surv
SL, Michels RG, Schachat AP, Green
WR (1986) Massive hemorrhage com-
plicating age-related macular degenera-
tion. Ophthalmology 93:1581–1592
4. Feman SS, Bartlett RE, Roth AM, Foos
RY (1972) Intraocular hemorrhage and
blindness associated with systemic anti-
coagulation. JAMA 220:1354–1355
5. Googe JM, Hirose T, Apple DJ, Melgen
S (1987) Vitreous hemorrhage second-
ary to age-related macular degeneration.
Surv Ophthalmol 32:123–130
Massive intraocular hemorrhage associ-
ated with anticoagulation and age-relat-
ed macular degeneration. Graefe’s Arch
Clin Exp Ophthalmol 226:59–64
8. Wade EC, Flynn HW, Olsen KR,
Blumenkranz MS, Nicolson DH (1990)
Subretinal hemorrhage management by
pars plana vitrectomy and internal
drainage. Arch Ophthalmol
Ophthalmol 15:297–311
2. Bressler NM, Bressler SB, Gragoudas
ES (1988) Age- related macular degen-
eration. Surv Ophthalmol 32:375–413
108:973–978
6. Klein R, Klein BEK, Linton KLP
(1992) Prevalence of age-related degen-
eration. The Beaver Dam Eye Study.
Ophthalmology 109:933–943