ArticlePDF Available

Retinal migraine: Caught in the act



Content may be subject to copyright.
Heparin therapy in giant cell
Giant cell arteritis (GCA) is a systemic
vasculitis that affects large and medium sized
arteries. Visual loss is one of the most
devastating complications of GCA and
usually occurs from occlusion of the posterior
ciliary arteries (PCA) leading to anterior
ischaemic optic neuropathy (AION). Visual
loss can also occur from occlusion of other
arteries that supply the visual pathway, such
as the central retinal and ophthalmic arteries.
Corticosteroid therapy, given orally or
intravenously, is the standard treatment for
GCA associated visual loss.
The optimal
route of administration and dosage to pre-
vent further visual loss are not known;
however, most clinicians advocate higher
doses in patients who already have experi-
enced visual loss. Treatment with corticoster-
oids usually results in stabilisation of visual
loss and some patients may have some degree
of visual recovery.
However, despite treat-
ment with high dose intravenous corticoster-
oids, visual loss may progress.
The reported
use of adjunctive agents under these circum-
stances has been limited. We report a patient
who had progressive visual loss while on high
dose intravenous corticosteroids and who
markedly improved after treatment with
Case report
An 85 year old man presented to his opto-
metrist for a routine eye examination. His
visual acuity was 20/40 both eyes and his
optic discs were normal. Three weeks later
(day 1), he lost vision in his right eye. His
visual acuity was now 20/100 right eye and
20/40 left eye. On visual field testing with
Figure 1 Photomicrograph of temporal artery
biopsy histological section shows a marked
proliferation of smooth muscle cells that can be
seen as dark reddish/brown concentric rings in
the thickened medial and intimal layers (stain,
actin which is selective for smooth muscle cells,
original magnification, 6250).
Figure 2 Serial automated static perimetry, (A) Day 5: mean deviation (MD) = 219.63 dB left eye, no light perception (NLP) right eye. (B) Day 6:
MD = 217.99 dB LE, hand motion (HM) right eye. (C) Day 7: MD = 212.30 dB left eye, MD = 224.60 dB right eye. (D) Day 8: MD = 25.58 dB left
eye, MD = 215.66 dB right eye. (E) Day 9: MD = 217.63 dB left eye, MD = 219.66 dB right eye. (F) Week 4: MD = 212.80 dB left eye,
MD = 226.87 dB right eye.
If you have a burning desire to respond to a
paper published in BJO, why not make use
of our ‘‘rapid response’’ option?
Log on to our web site (www.bjophthalmol.
com), find the paper that interests you, and
send your response via email by clicking on
the ‘‘eLetters’’ option in the box at the top right
hand corner.
Providing it isntlibellousorobscene,it
will be posted within seven days. You can
retrieve it by clicking on ‘‘read eLetters’’ on
our homepage.
The editors will decide as before whether to
also publish it in a future paper issue.
Br J Ophthalmol 2004;88:298315 on July 15, 2011 - Published by bjo.bmj.comDownloaded from
static perimetry he had an inferior altitudinal
defect in the right eye and his visual field was
normal in the left. His right optic disc was
pale and oedematous and his left optic disc
was normal. He was suspected of having GCA
and was admitted for treatment with intra-
venous methylprednisolone (IVMP)
250 mg every 6 hours. On day 2, the visual
field defect moved upward with involvement
of fixation and his visual acuity was counting
fingers. On day 3, the visual field defect
increased further with a small rim of vision
remaining superiorly. On day 4, his visual
acuity was light perception (LP) and he was
transferred to our institution.
On our initial examination (day 4), his
visual acuity was LP right eye and 20/50 left
eye. He had a right relative afferent pupillary
defect (RAPD), pale optic disc oedema on the
right, and a normal appearing optic disc on
the left. The Westergren erythrocyte sedi-
mentation rate (ESR) was 74 mm in the
first hour. The IVMP was continued as before,
and a temporal artery biopsy (TAB) was
performed which was positive for GCA
(fig 1).
On day 5, his visual acuity decreased to no
light perception (NLP) right eye and 20/80 left
eye. His right pupil was amaurotic and his left
pupil reacted sluggishly. Static perimetry now
showed an inferior altitudinal defect in the
left eye (fig 2A). The right optic disc
appearance was unchanged but the superior
portion of the left optic disc was now pale and
oedematous. Posterior ciliary artery (PCA)
blood flow could not be detected in either eye
with orbital colour Doppler imaging (CDI)
(table 1). Heparin (5000 units intravenous
bolus) was started, and titrated to maintain
the partial thromboplastin time (PTT) within
a therapeutic range (4670 seconds).
On day 6, his PTT was 74 seconds and his
visual acuity improved to hand motion (HM)
right eye and 20/70 left eye. His right pupil
reacted sluggishly and the left pupil reacted
more briskly with a right RAPD. Static
perimetry was essentially unchanged (fig 2B)
and there was no appreciable change in the
optic disc appearance. The ocular blood flow
characteristics with orbital CDI (fig 3A) were
unchanged (table 1).
On day 7, while continuing IVMP and
heparin, his visual acuity improved to count-
ing fingers (CF) at 1 foot right eye and 20/50
left eye. The pupillary examination was
unchanged. Static perimetry was improved
(fig 2C). Posterior ciliary artery blood flow
was now detectable with orbital CDI (table 1).
Warfarin (2.5 mg by mouth four times daily)
was added and titrated to a therapeutic level.
By day 8, his visual acuity had improved to
20/400 right eye and 20/50 left eye. Static
perimetry was further improved (fig 2D). The
ocular blood flow characteristics on orbital
CDI likewise improved (table 1). The pulsa-
tility index of the central retinal artery (CRA)
was markedly decreased (fig 4). Oral corti-
costeroid therapy (prednisone 100 mg by
mouth daily) was started and IVMP therapy
was stopped.
On day 9, the international ratio (INR) was
therapeutic (between 2 and 3) and heparin
was stopped. His visual acuity was 20/200
right eye and 20/70 left eye. Static perimetry
was slightly worse in the right eye and the
inferior altitudinal defect in the left eye was
denser and closer to fixation (fig 2E). The
ocular blood flow characteristics on orbital
CDI (fig 3B) remained stable (table 1). The
patient was discharged from the hospital
with oral corticosteroid therapy (prednisone
100 mg by mouth daily) and warfarin.
At the 1 week follow up visit (day 15), his
visual acuity improved to 20/80 right eye and
20/40 left eye. Static perimetry was
unchanged. The ocular blood flow character-
istics on orbital CDI (fig 3C) remained
unchanged (table 1).
At the 4 week follow up visit while taking
100 mg of prednisone daily and warfarin, his
visual acuity deteriorated to 20/400 right eye
and 20/80 left eye. Static perimetry worsened
in the right eye with diffuse loss and the
inferior altitudinal defect left eye was less
dense (fig 2F). The ocular blood flow
Figure 3 Serial orbital colour Doppler imaging. (A) Day 6: There is diminished blood flow through
the ophthalmic artery (OA) in the right eye (OD) and central retinal artery (CRA) in both eyes. The
blood flow through the posterior ciliary artery circulation is undetectable in both eyes. Blood flow
through the central retinal vein (CRV) can be seen in the left eye (OS). (B) Day 9: 4 days after
starting heparin, there is blood flow through the nasal posterior ciliary artery (NPCA) OD and
through the temporal posterior ciliary arteries (TPCA) both eyes, with improved blood flow through
the OA and CRA in both eyes. Blood flow through the CRV can be seen in both eyes. (C) Day 15:
6 days after discontinuation of heparin, the ocular blood flow remained stable.
Table 1 Spectral analysis of orbital colour Doppler imaging
Day 5 Day 6 Day 7 Day 8 Day 9 Day 15
CRA systolic (cm/s) 10.1+/21.9 4.1 2.4 5.4 4.7 5.3 4.5 5.9 4.4 8.2 5.6 9.7 5.7
CRA diastolic (cm/s) 2.6+/21.2 0.2 0.4 0.4 0.6 1.3 1.4 1.6 0.7 2.4 1.6 3.0 1.7
CRA pulsatility index 1.52 20.0 6.3 49.0 8.2 42.0 32.0 3.9 4.2 1.0 2.0 2.8 1.87
PCA systolic (cm/s) 12.4+/24.2
Nasal NR NR NR NR 4.2 NR 5.1 NR 7.7 NR 8.0 NR
Temporal NR NR NR NR 3.1 7.2 4.8 9.1 7.4 9.4 7.4 5.0
PCA diastolic (cm/s) 4.3+/22.2
Nasal NR NR NR NR 0.8 NR 0.4 NR 2.6 NR 2.3 NR
Temporal NR NR NR NR 0.5 1.3 1.2 0.2 2.3 0.2 2.7 1.7
PCA pulsatility index 1.09
Nasal NR NR NR NR 1.8 NR 1.0 NR 1.3 NR 2.0 NR
Temporal NR NR NR NR 7.7 3.0 3.1 3.9 1.7 2.2 1.9 3.0
OA systolic (cm/s) 31.3+/24.2 58.0 133.4 65.1 118.9 65.7 118.6 59.9 94.9 58.0 71.5 58.0 50.3
OA diastolic (cm/s) 8.3+/23.9 19.3 31.9 16.9 37.7 21.3 16.9 19.3 23.7 23.2 11.6 20.3 10.3
OA pulsatility index 1.55 1.2 1.4 1.2 1.2 1.1 1.5 1.2 1.2 1.0 1.4 1.1 2.1
CRA = central retinal artery, PCA = posterior ciliary artery, OA = ophthalmic artery, NR = not recordable.
PostScript 299 on July 15, 2011 - Published by bjo.bmj.comDownloaded from
characteristics on orbital CDI remained
unchanged. At the 3 month follow up visit
while still on the same medications, his
visual acuity deteriorated to LP right eye
and improved to 20/60 left eye. Static
perimetry of the left eye was unchanged
and the ocular blood flow characteristics on
orbital CDI again remained stable bilaterally.
His prednisone dosage was slowly tapered
and the warfarin was discontinued.
Once visual loss occurs in GCA, the goals of
treatment with corticosteroid therapy are to
prevent further progression and to reverse the
visual loss if possible. Table 2 reviews the
treatment results of GCA with either oral or
intravenous corticosteroid therapy in five
large series of patients. After initiation of
corticosteroid therapy, most patients experi-
enced visual stabilisation, a small number
had improvement of vision, and even fewer
patients had a progression of visual loss.
The reason for progression of visual loss
despite treatment with corticosteroid therapy
is unknown. One possible reason is that the
dose of corticosteroids was inadequate. Both
the dose and route of administration of
corticosteroids varied in the studies in
table 2. Therefore, it is difficult to assess
whether the treatment dose was adequate or
if intravenous administration is superior to
oral administration.
Most clinicians support the use of high
dose intravenous corticosteroids for patients
who have experienced visual loss from GCA.
However, progression of visual loss has been
described despite high dose IVMP.
Another possible reason for the progression
of visual loss despite treatment is that
corticosteroid therapy alone is an inadequate
treatment for some patients. A review of our
patients TAB showed near total luminal
stenosis by intimal hyperplasia with prolif-
eration of actin stained smooth muscle cells
in the arterial media and intimal layers
(fig 1). Complete arterial occlusion can occur
with progression of the hyperplastic process
or may be the result of superimposed arterial
thrombosis. Perhaps with advanced luminal
stenosis, corticosteroids are not as effective as
they are early in the inflammatory cascade
and may not be able to prevent thrombosis.
We found one report of the use of heparin
for visual loss in GCA.
In a review of 174
patients with GCA, Liozon et al reported the
use of heparin in most patients with perma-
nent visual loss and some patients with
threatening symptoms.
However, the specific
results of this treatment were not reported.
The use of heparin has also been described
for manifestations of GCA other than visual
loss. Staunton et al described a 64 year old
man with TAB proved GCA and clinical signs
of an evolving vertebrobasilar stroke and
ischaemic cerebellar lesion on MRI while on
oral corticosteroid therapy.
His symptoms
improved following treatment with high dose
intravenous dexamethasone and heparin. It
has been postulated that systemic anti-
coagulation may be beneficial during the
initial phase of steroid treatment because of
recent evidence that anti-cardiolipin antibo-
dies are present in a higher frequency in
patients with GCA.
Normally, heparin occurs complexed to
histamine as a macromolecule in mast cells
and its physiological role is unknown.
Heparin has an immediate anticoagulation
effect after intravenous administration. The
coagulation process generates thrombin by
two interrelated pathways, the extrinsic and
intrinsic. Both pathways involve a cascade of
enzymatic reactions that ultimately form
thrombin. Thrombin catalyses the conversion
of fibrinogen to fibrin that forms the matrix
of a thrombus. Thrombin also activates
clotting factor XIII that is necessary for
stabilising the cross links of the fibrin
molecules. If thrombin is not produced or
production is impeded then coagulation is
inhibited. Antithrombin III is an a-globulin
that inhibits thrombin. Heparin indirectly
binds to antithrombin III and forms a
complex that more rapidly inhibits thrombin
formation, thereby preventing coagulation
and clot formation.
Heparin has also been shown to have other
biochemical activities such as regulation of
lipid metabolism, control of blood fluidity at
the endothelial surface, control of cell attach-
ment to various proteins in the extracellular
matrix, binding with acidic and basic fibro-
blast growth factors, binding to interleukin 3
and granulocyte-macrophage colony stimu-
lating factor, and inhibition of serotonin
induced pulmonary artery smooth muscle
cell hypertrophy.
The mechanism by which
heparin led to improvement of vision in our
patient is not known and perhaps its ther-
apeutic effect was unrelated to anticoagula-
tion. None the less, serial orbital CDI showed
improvement in ocular blood flow (table 1;
fig 3AC). Colour Doppler imaging is an
ultrasonic imaging modality that combines
B-mode ultrasonography with Doppler ultra-
sonography. When applied to the orbit, this
imaging modality allows for the assessment
of ocular blood flow and has been described
in detail elsewhere.
11 12
Blood flow towards
the transducer is displayed as red and
represents arterial flow and blood flow away
from the transducer is displayed as blue and
represents venous flow. Pulsed Doppler with
spectral analysis can be used in conjunction
with the CDI to accurately quantify the
systolic and diastolic flow characteristics.
The pulsatility index can be calculated and
represents an assessment of vascular resis-
tance to blood flow. A high pulsatility index
indicates a high resistance and therefore
reduced blood flow.
Initially, the systolic and diastolic pressures
of the CRA were well below normal and the
pulsatility index was markedly increased. The
flow parameters of the PCA were undetect-
able. After heparinisation (day 5), flow was
restored to the PCA circulation (day 7,
table 1). The pulsatility index of the CRA
also decreased (days 68, fig 4). The decrease
in pulsatility index represents a decrease in
blood flow resistance. The decrease in pulsa-
tility index and restoration of blood flow to
the PCA circulation both correspond to an
improvement in visual acuity and visual field.
The sequence of rapid improvement of orbital
CDI haemodynamics, visual acuity, and
visual field strongly suggest that the admin-
istration of heparin was responsible. Since
this occurred rapidly, some of the biochem-
ical activities of heparin, such as inhibition of
smooth muscle cell hypertrophy and the
others that act on a more chronic time line
could not have caused improvement. A more
likely candidate might be the control of blood
fluidity at the endothelial surface.
In summary, the use of corticosteroid
therapy alone for the treatment of GCA may
have limited therapeutic success. The patho-
genic mechanism of luminal stenosis in GCA
is undergoing revision,
and the emerging
model offers possibilities for novel therapeu-
tic intervention. Our patient had a remark-
able improvement in visual acuity within
24 hours after starting heparin. Over the
ensuing 2 days his visual field and ocular
blood flow improved. The improvement with
the addition of heparin may have been
coincidental, and his vision may have recov-
ered with IVMP alone. However, the con-
tinued decline in visual function, despite
3 days of IVMP, and his dramatic improve-
ment after starting heparin, strongly suggests
that heparin had a pivotal role in his
At the 3 month follow up visit, his vision in
the right eye deteriorated to LP, even though
Table 2 Review of five large series of visual outcome in GCA treated with either oral or intravenous corticosteroid therapy
Reference Year
Total number of
patients with GCA
Number of patients
with visual loss
Number of patients
(%) with visual
Number of patients (%)
with visual recovery
Number of patients (%)
with progression of
visual loss
Chan et al
2001 73 73 43 (59) 21 (29) 9 (12)
Liozon et al
2000 147 23 17 (74) 5 (22) 1 (4)
Gonzalez-Gay et al
1998 239 34 22 (65) 8 (23) 4 (12)
Liu et al
1993 45 41 20 (49) 14 (34) 7 (17)
Aiello et al
1992 245 34 24 (70) 5 (15) 5 (15)
Figure 4 Graph of arterial pulsatility index.
Blood flow through the posterior ciliary artery
(PCA) was detectable on day 7 and central
retinal artery (CRA) pulsatility index was nearly
normal by day 8.
300 PostScript on July 15, 2011 - Published by bjo.bmj.comDownloaded from
his ocular blood flow characteristics were
stable. We are uncertain about the reason for
his decline in visual function, since at the
time of the decline, he was maintained on
oral corticosteroid and his serial ESR values
were normal. None the less, we suggest that
in patients who experience a progression of
visual loss from GCA, despite IVMP therapy,
may benefit acutely from the addition of
heparin therapy.
L M Buono
Neuro-Ophthalmology Service, Duke University Eye
Center, Erwin Road, Box 3802, Durham, NC 27710,
R Foroozan
The Cullen Eye Institute, Baylor College of Medicine,
Department of Ophthalmology, 6565 Fannin Street,
NC-205, Houston, TX 77030, USA
M de Virgiliis
Department of Ophthalmology, Hospital Pedro
Lagleyze, Buenos Aires, Argentina
P J Savino
Neuro-Ophthalmology Service, Wills Eye Hospital,
Thomas Jefferson Medical College, 840 Walnut
Street, Philadelphia, PA 19107, USA
Correspondence to: Dr Lawrence M Buono, Duke
University Eye Center, Erwin Road, Box 3802,
Durham, NC 27710, USA;
Accepted for publication 19 March 2003
1 Cornblath WT, Eggenberger ER. Progressive
visual loss from giant cell arteritis despite high-
dose intravenous methylprednisolone.
Ophthalmology 1997;104:8548.
2 Liu GT, Glaser JS, Schatz NJ, et al. Visual
morbidity in giant cell arteritis. Clinical
characteristics and prognosis for vision.
Ophthalmology 1994;101:177985.
3 Aiello PD, Trautmann JC, McPhee TJ, et al. Visual
prognosis in giant cell arteritis. Ophthalmology
4 Matzkin DC, Slamovits TL, Sachs R, et al. Visual
recovery in two patients after intravenous
methylprednisolone treatment of central retinal
artery occlusion secondary to giant-cell arteritis.
Ophthalmology 1992;99:6871.
5 Weinberg DA, Savino PJ, Sergott RC, et al. Giant
cell arteritis. Corticosteroids, temporal artery
biopsy, and blindness. Arch Fam Med
6 Slavin ML, Margolis AJ. Progressive anterior
ischemic optic neuropathy due to giant cell
arteritis despite high-dose intravenous
corticosteroids. Arch Ophthalmol
7 Liozon E, Herrmann F, Ly K, et al. Risk factors for
visual loss in giant cell (temporal) arteritis: a
prospective study of 174 patients. Am J Med
8 Staunton H, Stafford F, Leader M, et al.
Deterioration of giant cell arteritis with
corticosteroid therapy. Arch Neurol
9 Levesque H, Cailleux N, Courtois H. On the
necessity of anticoagulant therapy during the
initial phase of steroid treatment in giant cell
arteritis. Eur J Med 1993;2:1867.
10 Garg HG, Thompson BT, Hales CA. Structural
determinants of antiproliferative activity of
heparin on pulmonary artery smooth muscle cells.
Am J Physiol Lung Cell Mol Physiol
11 Aburn NS, Sergott RC. Orbital colour Doppler
imaging. Eye 1993;7(Pt 5):63947.
12 Lieb WE, Flaharty PM, Ho A, et al. Color Doppler
imaging of the eye and orbit. A synopsis of a 400
case experience. Acta Ophthalmol Suppl
13 Weyand CM. The Dunlop-Dottridge Lecture: The
pathogenesis of giant cell arteritis. J Rheumatol
14 Chan CC, Paine M, ODay J. Steroid
management in giant cell arteritis. Br J Ophthalmol
15 Gonzalez-Gay MA, Blanco R, Rodriguez-
Valverde V, et al. Permanent visual loss and
cerebrovascular accidents in giant cell arteritis:
predictors and response to treatment. Arthritis
Rheum 1998;41:1497504.
Retinal migraine: caught in the
A 22 year old male dancer presented via his
optometrist following an episode of transient
visual loss in his left eye. He described a slow
blurring and darkening of the vision of the
left eye with a similarly gradual return to
normal, the whole episode lasting 10 min-
utes. He described similar episodes every 2
3 months for the previous 3 years with no
associated migrainous aura or headache, and
exercise was not a trigger. Figure 1 shows
images taken before, during, and 10 minutes
after his presenting episode (see figure legend
for description).
Visual acuity was no perception of light
when tested during the episode and had
recovered to 6/9 in the affected left eye when
seen 2 hours later at our ophthalmic emer-
gency department. Aside from congenital
protanomaly there were no other ocular or
systemic abnormalities. On subsequent
review his visual acuity had returned to 6/6
bilaterally and his fundus appearance
remains normal. There was no visual field
abnormality (Goldmann I2e, I4e) and hae-
matological (including thrombophilia
screen), carotid, and cardiac investigations
were normal.
Retinal migraine (otherwise known as
ophthalmic migraines, anterior visual path-
way migraines, or ocular migraines) causes
monocular visual loss for 1020 minutes
which can be associated with diffuse or
unilateral headache.
People experiencing
ocular migraines often have a history of one
of the more conventional forms of migraine,
and exercise may precipitate the attacks.
Vasospasm of the retinal circulation or
ophthalmic artery is thought to be the cause
of the amaurosis of ocular migraine.
Ischaemic optic neuropathy
and permanent
arcuate scotomas
may occur after ocular
migraine and retinal vascular occlusions have
been reported in conjunction with cerebral
and ocular migraine.
Retinal vasospasm may be associated with
underlying systemic diseases such as SLE
and antiphospholipid syndrome.
In older
patients it may be associated with giant cell
arteritis, polyarteritis nodosa, and eosinophi-
lic vasculitis.
Other associated haematologi-
cal abnormalities include low protein C and S
levels and positive antinuclear antibodies.
A relative afferent pupillary defect can be
demonstrated during episodes.
arterioles have been reported to constrict,
and the fovea may become more distinct
with surrounding macular retinal pallor
optic discs may be pale
early and hyper-
aemic later.
Early isolated constriction of
veins has been reported,
which may be
as well as simultaneous constric-
tion of arterioles and veins.
Later engorge-
ment of the retinal veins has been observed
some hours after an attack.
Nerve fibre
bundle defects can sometimes be a late
Our case demonstrates photographically
the arterial vasoconstriction of retinal
migraine (fig 1), which occurred in the
absence of a precipitating cause such as
Figure 1 (A) Left eye taken at routine
optometrist appointment 1 month before
presenting episode; normal appearance of
disc, vessels, and macula. (B) During the
episode of visual loss, constriction of both
arteries and veins is seen with macular pallor
and a central cherry red spot and slight disc
pallor. (C) 10 minutes after the previous picture
when vision according to the patient had
returned to normal. Some pallor remains in the
macula with a central cherry red spot but the
disc and vasculature have returned to normal.
Originally presented in part at the 28th Annual North
American Neuro-Ophthalmology Society (NANOS)
Meeting in Copper Mountain Resort, CO, USA, 914
February, 2002.
PostScript 301 on July 15, 2011 - Published by bjo.bmj.comDownloaded from
exercise and in the absence of a migrainous
aura or headache.
Doppler studies have revealed cessation of
retinal arterial flow during an exercise
induced retinal migraine episode.
A 48 year
old cluster headache sufferer who underwent
fluorescein angiography during an ocular
migraine attack demonstrated narrowing of
the retinal veins and delayed retinal artery
filling during the episode with normal chor-
oidal filling.
Most previously published
photographs have shown venous, retinal,
and disc changes late in the attack, perhaps
reflecting that arterial vasoconstriction occurs
early during attacks and is not often photo-
2 4 10 11
Alternatively, there may be a
spectrum of severity of retinal migraine
manifestations, which in severe cases may
result in a transient pale macular area with
cherry red spot.
When considered necessary, effective treat-
ments include propranolol,
Prophylactic aspirin or nifedi-
pine may be tried to prevent exercise induced
and inhaled amyl nitrate can be
used early in an attack to try to induce
We thank William Templeton, Eyesite Opticians,
Brighton who provided photographs.
E Doyle, B J Vote, A G Casswell
Sussex Eye Hospital, Eastern Road, Brighton BN2 5BF,
Correspondence to: Mr Eddie Doyle, Sussex Eye
Hospital, Eastern Road, Brighton BN2 5BF, UK;
Accepted for publication 21 March 2003
1 Burger SK, Saul RF, Selhorst JB, et al. Transient
monocular blindness caused by vasospasm.
N Engl J Med 1991;325:8703.
2 Jehn A, Dettwiler B, Fleischhauer J, et al. Exercise-
induced vasospastic amaurosis fugax. Arch
Ophthalmol 2002;120:2202.
3 Glaser JS. Topical diagnosis: prechiasmal visual
pathways. retinal artery occlusions. Duane’s
ophthalmology 2002;CD-ROM ed. Philadelphia:
Lippincott Williams and Wilkins, 2002.
4 Wolter JR, Birchfield WJ. Ocular migraine in a
young man resulting in unilateral blindness and
retinal oedema. J Pediatric Ophthalmol
5 Winterkorn JMS, Kupersmith MJ, Wirtschafter JD,
et al. Brief report: treatment of vasospastic
amaurosis fugax with calcium channel blockers.
N Engl J Med 1993;329:3969.
6 Walsh FB, Hoyt AB. Clinical Neuro-
ophthalmology. Baltimore: Williams and Wilkins,
7 Cassen JH, Tomsak RL, DeLuise VP. Mixed
arteriovenous occlusive disease of the fundus.
J Clin Neuro-ophthalmol 1985;5:648.
8 Gronwall H. On changes in the fundus oculi and
persisting injuries to the eye in ocular migraine.
Acta Ophthalmologica 1938;16:60211.
9 Levine SR, Deegan MJ, Futrell M, et al.
Cerebrovascular and neurologic disease
associated with antiphospholipid antibodies:48
cases. Neurology 1990;40:11819.
10 Kline LB, Kelley CL. Ocular migraine in a patient
with cluster headache. Headache
11 Heyck H. Die neurologischen
Begleiterscheinungen der Migraine and dos
Problems des ‘‘angiospastischen Hirninsults.’’
Long term follow up in a case of
successfully treated idiopathic
retinal vasculitis, aneurysms, and
neuroretinitis (IRVAN)
The idiopathic retinal vasculitis, aneurysm,
and neuroretinitis (IRVAN) syndrome is a
rare clinical entity characterised by peripheral
retinal vascular occlusion, retinal vasculitis,
and multiple posterior retinal arterial aneur-
In most reported cases, visual acuity
was aggravated.
We report a case of
successfully treated IRVAN syndrome in
which good long term visual acuity has been
Case report
A 36 year old woman visited the Kansai
Medical University Hospital on January 1999,
complaining of flies flying in her right eye for
the past year. She also had visual disturbance
and metamorphopsia for 2 months in her
right eye. Her best corrected visual acuity was
20/40 in the right eye and 20/16 in the left
eye. Inflammatory cells were noted in both
the anterior chamber and vitreous cavity.
There was mild rubeosis iridis in the left eye.
In the right eye, there was a fibrovascular
membrane in the epipapillary area accompa-
nying tractional retinal detachment, thick
hard exudates on nasal side of the fovea, and
vitreous haemorrhage (Fig 1A). Her optic disc
was mildly hyperaemic, and multiple aneur-
ysms surrounded by perivascular exudation
were observed at the posterior retinal artery
in the left eye (Fig 1B). Extensive arterial and
venous vascular occlusion and adjacent
anomalous arteriovenous anastomosis were
observed at the peripheral retina in both eyes
(Fig 2).
Systemic examination and laboratory find-
ings did not suggest systemic abnormality,
and the family history was not contributory.
Oral prednisolone was initiated at 50 mg/day
and decreased gradually. In an attempt to
prevent peripheral retinal neovascularisation,
panretinal photocoagulation (PRP) was car-
Figure 1 Vitreous haemorrhage, neovascularisation of the optic disc, and hard exudates were
observed in the right eye (A), and hard exudates surrounding the optic disc and aneurysms shown
by fundus fluorescein angiogram (FA) was observed in the left (B) on January 1999. The right eye
(C), and aneurysms shown by FA in the left (D) were improved by February 2003.
Figure 2 Panoramic FA of the right eye on January 1999 (A) and February 2003 (B).
The authors have no commercial interests in any
products mentioned in this article.
302 PostScript on July 15, 2011 - Published by bjo.bmj.comDownloaded from
ried out to treat the peripheral retinal non-
perfusion area. Since vitreous haemorrhage
in the right eye did not improve, right eye
vitrectomy was performed on March 1999,
and tractional retinal detachment and vitr-
eous haemorrhage subsequently improved.
The aneurysms disappeared by 5 months and
the hard exudates by 12 months. Her cor-
rected visual acuity improved to 20/20 in the
right eye and 20/16 in the left eye by February
2003 (Fig 1C and D).
The disappearance of retinal aneurysm in
IRVAN has been reported in two cases.
both of these cases, scattered retinal photo-
coagulation was performed on the peripheral
retinal non-perfusion area.
In one case in
which scattered retinal photocoagulation was
not performed, it has been reported that
some aneurysms enlarged, others became
small, and new aneurysms appeared.
In our
case, all retinal arterial aneurysms vanished
after PRP (Fig 1B and D). These results
strongly suggest that PRP applied to the
retinal non-perfusion area is useful and
should be performed during the early phase.
On the other hand, it has been reported that
retinal lesions are unresponsive to oral
steroids in most cases of IRVAN.
In our
case, systemic oral steroid was given to treat
the retinal vasculitis. The effect of oral steroid
in IRVAN will need further examination. The
results in our case and past cases strongly
suggest that PRP applied to the non-perfu-
sion area is the most important treatment for
IRVAN. We think vitrectomy must precede
PRP, because PRP cannot be performed when
there is vitreous haemorrhage and retinal
M Tomita, T Matsubara, H Yamada,
K Takahashi, T Nishimura, K Sho, M Uyama,
M Matsumura
Department of Ophthalmology, Kansai Medical
University, Osaka, Japan
Correspondence to: Minoru Tomita, Department of
Ophthalmology, Kansai Medical University, 10-15
Fumizono-cho, Moriguchi City, Osaka, 570-8506,
Accepted for publication 4 April 2003
1 Kincaid J, Schatz H. Bilateral retinal arteritis with
multiple aneurysmal dilatations. Retina
2 Chang TS, Aylward GW, Davis JL, et al.
Idiopathic retinal vasculitis, aneurisms, and
neuro-retinitis. Ophthalmology 1995;102:1089.
3 Owens SL, Gregor ZJ. Vanishing retinal arterial
aneurysms: a case report. Br J Ophthalmology
4 Sashihara H, Hayashi H, Oshima K. Regression
of retinal arterial aneurysms in a case of
idiopathic retinal vasculitis, aneurysms, and
neuroretinitis. Retina 1999;19:250.
5 Yeshurun I, Recillas-Gispect C, Navarro-Lopez, et
al. Extensive dynamics in location, sharp, and size
of aneurysm in a patient with idiopathic retinal
vasculitis, aneurysms, and neuroretinitis (IRVAN)
syndrome. Am J Ophthalmol 2003;135:118.
HLA typing is not predictive of
proliferative diabetic retinopathy
in patients with younger onset
type 2 diabetes mellitus
Chronic hyperglycaemia and the duration of
diabetes are the most important factors in
retinopathy. However, retinopathy progresses
in some patients despite good glycaemia
control. Also, poor glycaemia control does
not always lead to retinopathy in younger
onset patients, while still others develop
severe retinopathy that is resistant to retinal
These facts suggest that the risk factors for
diabetes and retinopathy are not necessarily
the same, and that the development of severe
retinopathy may be influenced by genetic
Human leucocyte antigen (HLA) status has
a significant role in immune responses and
immunological tolerance and is a factor in
the onset of type 2 diabetes.
DR4, DR8,
DR9, and several antigens of the DQ region
are related to retinopathy in patients with
type 1 diabetes.
In addition, it was reported
that HLA-DR was expressed in proliferative
Little is known, however,
about the relation between retinopathy with
type 2 diabetes and the HLA antigen.
Furthermore, most previous studies have
not taken into consideration the background
of glycaemic control or the duration of the
diabetes. A group of younger onset type 2
diabetes patients with PDR, and a group who
had no signs of retinopathy despite a long
duration of diabetes were compared. Younger
patients were studied to reduce the influence
of adult diseases such as hyperlipidaemia and
hypertension. In addition, clinical back-
ground factors were considered when study-
ing the frequency of HLA types.
Following the informed consent of each of
the subjects, blood samples were collected.
The study was approved by the human
studies review board of Tokyo Womens
Medical University and was performed in
accordance with the Helsinki Declaration of
1975 and its 1983 revision. The diagnosis of
type 2 diabetes was made based on 1985
World Health Organization criteria.
excluded subjects who were GAD antibody
positive. The patients had been diagnosed as
having type 2 diabetes aged under 30 years
(range 1221 years) and type 2 diabetes
duration for more than 10 years.
Additionally, the patients whose average
level over 10 years was from 6% to
10%, were selected in this study. All were
receiving treatment at Tokyo Womens
Medical University Diabetes Center.
Further, two groups were selected from
above patients, as follows. The PDR group
consisted of 44 patients, who had undergone
vitreous surgery under the age of 40 (mean
28.8 (SD 4.4) years) where surgery had been
carried out at the department of ophthalmol-
ogy, diabetes centre (as above) during the
period 19939. The non-DR group consisted
of 45 patients who had no signs of retino-
pathy despite having diabetes for more than
10 years. Consequently, the sex, the diabetes
duration, and the blood glucose control have
been matched between the two groups on the
basis of the selection criteria above (table 1).
The control group selected for comparison,
consisted of 50 healthy patients. The HbA
level was determined with resin microcolumn
technique (HPLC, Kyoto Chemical) (normal
range 4.35.8%).
HLA-A, B, C, DR, and DQ typing of blood
samples was conducted on all three groups
using standard microcytotoxicity methods.
For comparisons between the groups, we
applied the x
test of independence or
Fishers exact probability test. The unpaired
t test was used for comparing mean values.
The level of significance was set at p,0.05.
All analyses were performed using the Stat
View statistical software package (Abacus
Concepts, Berkeley, CA, USA).
The frequencies of HLA-A, B, and Cw anti-
gens in the control group, the non-DR group,
and the PDR group are shown in table 2, and
those of DR, and DQ antigens in table 3,
There was no significant difference among
the three groups in HLA-A, B, and DQ
antigens. The non-DR group showed higher
frequencies of HLA Cw4 (x
= 4.027,
Table 1 Clinical profile of the control group, non-DR group, and PDR group
Control group Non-DR group PDR group p Value
Number of patients 50 45 44 2
Age (years)
Male/female 25/25 20/25 20/24 2
Age at the diagnosis
of diabetes (years)
Age at the diagnosis of
diabetic retinopathy
Age at the time of vitreous
surgery (years)
(%) 2
Hypertension (%) 0 (0.0%) 3 (6.7%) 4 (8.3%) NS**
Renal disease (%) 0 (0.0%) 6 (13.3%) 15 (34.1%) .0021**
neuropathy (%)
0 (0.0%) 9 (20.0%) 18 (40.9%)
Positive family history (%) 0 (0.0%) 15 (33.3%) 19 (43.1%) NS**
Number of patients or mean (SD).
NS = not significant.
The age was compared between each group.
*Comparisons between the non-DR group and the PDR group by unpaired t test.
**Comparisons between the non-DR group and the PDR group x
test of independence or Fishers exact
probability test.
Age at the time of this study.
Average value over 10 years.
PostScript 303 on July 15, 2011 - Published by bjo.bmj.comDownloaded from
Table 2 HLA-A, B, and C antigen frequencies (%) in each group
Control group
Control v non-DR Control v PDR non-DR v PDR
Non-DR group PDR group x
p Value Odds ratios x
p Value Odds ratios x
p Value Odds ratios
A1 0 (0.0) 0 (0.0) 0 (0.0) 0.000 NS 22NS 2 0.000 NS 2
A2 19 (38.0) 18 (40.0) 15 (34.1) 0.040 NS 1.088 0.155 NS 0.844 0.333 NS 0.776
A11 10 (20.0) 8 (17.8) 10 (22.7) 0.076 NS 0.865 1.000 NS 0.471 0.338 NS 1.360
A24 (9) 29 (58.0) 28 (62.2) 29 (65.9) 0.176 NS 1.193 0.620 NS 1.400 0.131 NS 1.174
A26 (10) 11 (22.0) 10 (22.2) 11 (25.0) 0.001 NS 1.013 0.118 NS 1.182 0.095 NS 1.167
A28 0 (0.0) 0 (0.0) 0 (0.0) 0.000 NS 22NS 2 0.000 NS 2
A30 (19) 1 (2.0) 0 (0.0) 0 (0.0) 0.910 NS 0.000 0.889 NS 0.000 0.000 NS 2
A31 (19) 10 (20.0) 9 (20.0) 8 (18.2) 0.000 NS 1.000 0.264 NS 0.757 0.252 NS 0.757
A33 (19) 8 (16.0) 7 (15.6) 7 (15.9) 0.101 NS 0.839 0.079 NS 1.167 0.109 NS 1.206
B7 5 (10.0) 4 (8.9) 4 (9.1) 0.034 NS 0.878 0.022 NS 0.900 0.001 NS 1.025
B13 1 (2.0) 1 (2.2) 0 (0.0) 0.006 NS 1.114 0.889 NS 0.000 0.989 NS 0.000
B17 1 (2.0) 2 (4.4) 1 (2.3) 0.463 NS 2.279 0.008 NS 1.140 0.322 NS 0.500
B27 0 (0.0) 0 (0.0) 0 (0.0) 0.000 NS 22NS 22NS 2
B35 8 (16.0) 9 (20.0) 7 (15.9) 0.258 NS 1.313 0.000 NS 0.993 0.252 NS 0.757
B37 1 (2.0) 0 (0.0) 4 (9.1) 0.910 NS 0.000 2.337 NS 4.900 4.283 NS 2
B39 4 (8.0) 3 (6.7) 5 (11.4) 0.062 NS 0.821 0.306 NS 1.474 0.600 NS 1.795
B40 1 (2.0) 0 (0.0) 0 (0.0) 0.910 NS 0.000 0.889 NS 0.000 2 NS 2
B44 (12) 8 (16.0) 10 (22.2) 6 (13.6) 0.597 NS 1.500 0.103 NS 0.829 1.112 NS 0.553
B46 5 (10.0) 5 (11.1) 2 (4.5) 0.031 NS 1.125 1.010 NS 0.429 1.323 NS 0.381
B48 3 (6.0) 2 (4.4) 4 (9.1) 0.115 NS 0.729 0.324 NS 1.567 0.764 NS 2.150
B51 (5) 9 (18.0) 8 (17.8) 6 (13.6) 0.001 NS 0.985 0.332 NS 0.719 0.288 NS 0.730
B52 (5) 11 (22.0) 10 (22.2) 13 (29.5) 0.001 NS 1.013 0.701 NS 1.487 0.623 NS 1.468
B54 (22) 9 (18.0) 8 (17.8) 6 (13.6) 0.001 NS 0.985 0.332 NS 0.719 0.288 NS 0.730
B55 (22) 2 (4.0) 1 (2.2) 1 (2.3) 0.245 NS 0.545 0.226 NS 0.558 0.000 NS 1.023
B56 (22) 2 (4.0) 2 (4.4) 0 (0.0) 0.012 NS 1.116 1.798 NS 0.000 2.001 NS 0.000
B59 2 (4.0) 2 (4.4) 6 (13.6) 0.012 NS 1.116 2.791 NS 3.789 2.298 NS 3.395
B60 (40) 5 (10.0) 3 (6.7) 6 (13.6) 0.341 NS 0.643 0.300 NS 1.421 1.189 NS 2.211
B61 (40) 12 (24.0) 10 (22.2) 14 (31.8) 0.042 NS 0.905 0.715 NS 1.478 1.040 NS 1.633
B62 (15) 8 (16.0) 8 (17.8) 6 (13.6) 0.053 NS 1.135 0.103 NS 0.829 0.288 NS 0.730
B67 1 (2.0) 1 (2.2) 0 (0.0) 0.006 NS 1.114 0.889 NS 0.000 0.989 NS 0.000
B70 1 (2.0) 1 (2.2) 0 (0.0) 0.006 NS 1.114 0.889 NS 0.000 0.989 NS 0.000
B75 1 (2.0) 0 (0.0) 0 (0.0) 0.910 NS 0.000 0.889 NS 0.000 2 NS 2
Cw1 16 (32.0) 12 (26.7) 12 (27.3) 0.324 NS 0.773 0.250 NS 0.797 0.004 NS 1.031
Cw3 23 (46.0) 20 (44.4) 22 (50.0) 0.023 NS 0.939 0.150 NS 1.174 0.276 NS 1.250
Cw4 7 (14.0) 14 (31.1) 9 (20.5) 4.027 0.045 2.774 0.690 NS 1.580 1.318 NS 0.569
Cw5 1 (2.0) 0 (0.0) 0 (0.0) 0.910 NS 0.000 0.889 NS 0.000 2 NS 2
Cw6 1 (2.0) 1 (2.2) 4 (9.1) 0.006 NS 1.114 2.337 NS 4.900 1.979 NS 4.400
Cw7 11 (22.0) 7 (15.6) 7 (15.9) 0.640 NS 0.653 0.561 NS 0.671 0.002 NS 1.027
Cw8 0 (0.0) 0 (0.0) 0 (0.0) 0.000 NS 22NS 22NS 2
Cw9 6 (12.0) 2 (4.4) 1 (2.3) 1.753 NS 0.341 3.213 NS 0.171 0.322 NS 0.500
Cw10 7 (14.0) 2 (4.4) 2 (4.5) 2.522 NS 0.286 2.416 NS 0.293 0.001 NS 1.024
Numbers in parentheses are percentages. NS = not significant. All p values were not significant other than Cw4 with p = 0.045 in the control v non DR test.
304 PostScript on July 15, 2011 - Published by bjo.bmj.comDownloaded from
p = 0.045) and DR4 (x
= 4.398, p = 0.036)
than the control group (tables 2 and 3).
While there was no significant difference
between non-DR group and PDR group in
any of the HLA antigens. The PDR
group showed higher frequencies of HLA
DR4 than the control group (x
= 5.937,
p = 0.014).
Type 1 diabetes is aetiologically different
from type 2 diabetes. Type 1 diabetes is
caused by a failure in the autoimmune
system, is clearly associated with specific
HLA antigens.
Type 2 diabetes is not
autoimmune and has less association or
linkage with genes in the HLA region than
type 1.
The type 2 diabetic patients in this study
showed a typical HLA pattern. Other research
has reported that DR3, DR4, and Cw4
increased in patients with type 2 diabetes
These findings are consistent with
our present results. Additionally an increase
in HLA-DR4, which is in linkage disequili-
brium with the DQB1*0302 allele has pre-
viously been reported in patients with type 2
This increase was mainly reported
to be restricted to patients with relative
insulin deficiency or antibodies to islet cells
or to glutamic acid decarboxylase.
DR4 was detected in 59.1% of the PDR
group, but this was not significantly diffe-
rent from the frequency in the non-DR group
(51.1%). DR4 may, therefore, be related to
the onset of type 2 diabetes, but not to
the development of retinopathy. The HLA-
DR4 levels reflected the antibody levels in
the pancreatic Langerhans island but not
the parameter of diabetic change in the
In summary, our research suggests that
HLA antigen investigations may be useful
for predicting the prognosis of younger onset
type 2 diabetes, but not for retinopathy in
these patients. Finally, we must precisely
define the alleles or combination of
alleles which cause increased susceptibility
to PDR.
The authors gratefully acknowledge the assistance
of Ms Jayne Simons for critically reviewing the
T Mimura, S Amano, S Kato, M Araie
Department of Ophthalmology, University of Tokyo
Graduate School of Medicine, Tokyo, Japan
T Mimura, H Funatsu, S Kitano, E Shimizu
Department of Ophthalmology, Diabetes Center Tokyo
Womens Medical University, Tokyo, Japan
H Noma
Department of Ophthalmology, Hiroshima University
School of Medicine, Hiroshima, Japan
O Yoshino
Department of Obstetrics and Gynecology, University
of Tokyo school of Medicine, Tokyo, Japan
S Hori
Department of Ophthalmology, Tokyo Womens
Medical University, Tokyo, Japan
Correspondence to: Tatsuya Mimura, MD,
Department of Ophthalmology, University of Tokyo
Graduate School of Medicine, 7-3-1 Hongo, Bunkyo-
ku, Tokyo, 113-8655 Japan;
Accepted for publication 21 April 2003
1 Funatsu H, Hori S, Ohashi Y, et al. Risk factors for
occurrence and progression of diabetic
retinopathy. Nippon Ganka Gakkai Zasshi
2 Rich SS, French LR, Sprafka JM, et al. HLA-
associated susceptibility to type 2 (non-insulin-
dependent) diabetes mellitus: the Wadena City
Health Study. Diabetologia 1991;36:2348.
3 Tuomilehto-Wolf E, Tuomilehto J, Hitman GA, et
al. Genetic susceptibility to non-insulin dependent
diabetes mellitus and glucose intolerance are
located in HLA region. BMJ 1993;307:1559.
4 Horton V, Stratton I, Bottazzo GF, et al. Genetic
heterogeneity of autoimmune diabetes: age of
presentation in adults is influenced by HLA DRB1
and DQB1 genotypes (UKPDS 43). UK
Prospective Diabetes Study (UKPDS) Group.
Diabetologia 1999;42:60816.
5 Agardh D, Gaur LK, Agardh E, et al. HLA-
DQB1*0201/0302 is associated with severe
retinopathy in patients with IDDM. Diabetologia
6 Cruickshanks KJ, Vadheim CM, Moss SE, et al.
Genetic marker associations with proliferative
retinopaty in persons diagnosed with diabetes
before 30 yr of age. Diabetes 1992;41:87985.
7 Scheiffarth OF, Tang S, Kampik A. Macrophages
and HLA-DR expression in proliferative
vitreoretinopaty. Fortschritte der Ophthalmologie
8 Mimura T, Funatsu H, Uchigata Y, et al.
Relationship between human leukocyte antigen
status and proliferative diabetic retinopathy in
patients with younger-onset type 1 diabetes
mellitus. Am J Ophthalmol 2003;135:8448.
9 World Health Organization. Diabetes mellitus:
report of WHO Study Group. Geneva: WHO,
Tech Rep Ser 1985;727:1113.
10 Nerup J, Platz P, Andersen OO, et al. HL-A
antigens and diabetes mellitus. Lancet
Effect of docosahexaenoic acid
supplementation on retinal
function in a patient with
autosomal dominant Stargardt-
like retinal dystrophy
The gene, ELOVL4, is mutated in Stargardt-
like macular dystrophy, a juvenile onset
disorder. ELOVL4 is homologous to a fatty
acid elongase presumably involved in the
biosynthesis of docosahexaenoic acid, DHA.
Table 3 HLA-DQ and -DR antigen frequencies (%) in each group
Control v non-DR Control v PDR Non-DR v PDR
group PDR group x
p Value
ratios x
p Value
ratios x
Value Odds ratios
DR1 4 (8.0) 3 (6.7) 4 (9.1) 0.062 NS 0.821 0.036 NS 1.150 0.180 NS 1.400
DR2 16 (32.0) 11 (24.4) 13 (29.5) 0.665 NS 0.688 0.066 NS 0.891 0.294 NS 1.296
DR3 0 (0.0) 0 (0.0) 0 (0.0) 2 NS 22NS 22NS 2
DR4 15 (30.0) 23 (51.1) 26 (59.1) 4.398 0.036 2.439 5.937 0.014 2.804 0.932 NS 1.510
DR6 14 (28.0) 8 (17.8) 6 (13.6) 1.391 NS 0.556 2.883 NS 0.406 0.288 NS 0.730
DR7 1 (2.0) 0 (0.0) 0 (0.0) 0.910 NS 0.000 0.889 NS 0.000 2 NS 2
DR8 11 (22.0) 7 (15.6) 7 (15.9) 0.640 NS 0.653 0.561 NS 0.671 0.002 NS 1.027
DR9 24 (48.0) 14 (31.1) 13 (29.5) 2.815 NS 0.489 3.340 NS 0.454 0.026 NS 0.928
DR10 1 (2.0) 0 (0.0) 0 (0.0) 0.910 NS 0.000 0.889 NS 0.000 2 NS 2
DR11 1 (2.0) 1 (2.2) 4 (9.1) 0.006 NS 1.114 2.337 NS 4.900 1.979 NS 4.400
DR12 2 (4.0) 1 (2.2) 0 0.0 0.245 NS 0.545 1.798 NS 0.000 0.989 NS 0.000
DR13 8 (16.0) 3 (6.7) 2 (4.5) 2.015 NS 0.375 3.230 NS 0.250 0.189 NS 0.667
DR14 6 (12.0) 4 (8.9) 3 (6.8) 0.243 NS 0.715 0.726 NS 0.537 0.132 NS 0.750
DRHR5 1 (2.0) 1 (2.2) 0 (0.0) 0.006 NS 1.114 0.889 NS 0.000 0.989 NS 0.000
DRHR6 3 (6.0) 2 (4.4) 1 (2.3) 0.115 NS 0.729 0.798 NS 0.364 0.322 NS 0.500
DR52 19 (38.0) 13 (28.9) 9 (20.5) 0.880 NS 0.663 3.445 NS 0.420 0.850 NS 0.633
DR53 33 (66.0) 29 (64.4) 31 (70.5) 0.025 NS 0.934 0.214 NS 1.228 0.366 NS 1.316
DQ1 32 (64.0) 28 (62.2) 20 (45.5) 0.032 NS 0.926 3.257 NS 0.469 2.518 NS 0.506
DQ2 1 (2.0) 0 (0.0) 0 (0.0) 0.910 NS 0.000 0.889 NS 0.000 2 NS 2
DQ3 28 (56.0) 23 (51.1) 15 (34.1) 0.228 NS 0.821 3.625 NS 0.449 1.965 NS 0.547
DQ4 17 (34.0) 12 (26.7) 20 (45.5) 0.601 NS 0.706 1.287 NS 1.618 3.410 NS 2.292
6 (12.0) 8 (17.8) 13 (29.5) 0.629 NS 1.586 4.468 NS 3.075 1.709 NS 1.940
DQ7 6 (12.0) 1 (2.2) 4 (9.1) 3.317 NS 0.167 0.208 NS 0.733 1.979 NS 4.400
Numbers in parentheses are percentages. NS = not significant. All p values were not significant other than DR4 with p = 0.036 in the control v non DR test and
0.014 in the control v PDR test.
The authors have no commercial or proprietary
interest in the product or company described in this
PostScript 305 on July 15, 2011 - Published by bjo.bmj.comDownloaded from
Humans can synthesise DHA from precursors
in small amounts, which may not be suffi-
cient for normal retinal function if intake is
Case report
A 15 year old girl with Stargardt-like macular
dystrophy and a mutation in ELOVL4 was
given a dietary supplement of DHA on two
occasions. A 7 day dietary history, prior to
starting DHA supplementation, estimated her
daily intake of DHA as 20 mg, in a diet of
1602 kcal. Beginning in April 2001, she was
supplemented with 20 mg/kg body weight
per day of DHA in capsules (Martek
BioSciences, Columbia, MD, USA). Her func-
tional response was determined by a VF-14
visual acuity testing, multi-
focal electroretinography (mfERG), and
plasma phospholipid analysis.
At baseline, fundus photography showed
minimal macular changes (fig 1); however,
the mfERG revealed significant macular
dysfunction. A full field ERG performed in
March 2000 revealed a reduction of approxi-
mately 3040% in rod and cone responses,
with the predominant effect on the b-wave
amplitude in comparison with normative
data (DTL electrode, dilated pupils, recorded
according to ISCEV standards). No anomaly
of latency was observed in either the mfERG
or full field ERG recordings.
Peripheral blood samples were taken in a
non-fasting state at the beginning of the
study. Her baseline plasma DHA level was at
the low end of the median for a North
American diet, 24 mg/ml; whereas her omega-
6 arachidonic acid was 76 mg/ml. Her plasma
DHA level increased after 2 months of sup-
plementation to 86 mg/ml (fig 2). She
reported progressive subjective improvement
in vision on the VF-14 questionnaire.
Analysis of the mfERG tracings revealed that
an improvement had occurred in amplitudes
from the foveal and parafoveal regions of the
macula (fig 2). For this study, we accepted
the group average response of rings 12(0
), and rings 35(525
), derived from a 63
hexagon stimulus (Veris, San Mateo, CA,
USA), as simulating the foveal and parafo-
veal responses respectively. Visual acuity
improved from 20/200 to 20/100 only in her
left eye with the DHA supplementation.
Unfortunately, she lost interest in the
study, became non-compliant despite our
encouragement, and was lost to follow up
from August 2001 to the end of March 2002.
She returned after perceiving that her vision
had declined since halting the DHA supple-
mentation. Repeat plasma DHA testing con-
firmed that her plasma DHA had dropped to
levels approaching those from the beginning
of the study (35 mg/ml, fig 2).
Supplementation was once again initiated.
After 4 months (July 2002), she showed the
same improvements in visual function both
subjectively (VF-14 score) and objectively
(mfERG amplitudes, fig 2), coincident with
an increase in plasma DHA.
The patients visual acuity again improved
only in the left eye from 20/200 to 20/100.
Another full field ERG was conducted in July
2002 to compare with the diagnostic series
from March 2000. The amplitudes of the
photopic b-wave and the scotopic b-wave
both improved by 30%, averaging both eyes.
The scotopic mixed rod-cone response
showed increased amplitude of the a-wave
by 85%, and b-wave by 46%, when responses
from both eyes were averaged. These results
could not be explained by intervisit variability
and suggest the effect of DHA supplementa-
tion occurs beyond the macula, over the
entire retina.
The patients young age, low dietary intake of
DHA, and a mutation in ELOVL4 may have
particularly predisposed her to an early
macular degeneration. A longer experience
with DHA supplementation in young patients
with Stargardt-like macular dystrophy could
determine if DHA supplementation can alter
the natural history.
The technical assistance of Peggy Kaminski, OMT,
YK Goh, PhD, and Ezekiel Weis, MD, is gratefully
I M MacDonald, M He´bert, R J Yau, S Flynn,
J Jumpsen, M Suh, M T Clandinin
Departments of Ophthalmology and Agricultural,
Food and Nutritional Science, University of Alberta,
Edmonton, Alberta, Canada; and the Department of
Ophthalmology (MH), Medical Research Centre,
CHUQ, Laval University, Que´bec City, Que´bec,
Correspondence to: Dr Ian M MacDonald,
Department of Ophthalmology, Royal Alexandra
Hospital, 10240 Kingsway Ave., Edmonton, Alberta,
Canada T5H 3V9;
Accepted for publication 23 April 2003
1 Zhang K, Kniazeva M, Han M, et al. A 5-bp
deletion in ELOVL4 is associated with two related
forms of autosomal dominant macular dystrophy.
Nat Genet 2001;27:8993.
2 Salem N Jr, Wegher B, et al. Arachidonic and
docosahexaenoic acids are biosynthesized from
their 18-carbon precursors in human infants. Proc
Natl Acad Sci USA 1996;93:4954.
3 Mackenzie PJ, Chang TS, Scott IU, et al.
Assessment of vision-related function in patients
with age-related macular degeneration.
Ophthalmology 2002;109:7209.
4 Hargreaves KM, Clandinin MT.
Phosphatidylethanolamine methyltransferase:
evidence for influence of diet fat on selectivity of
substrate for methylation in rat brain synaptic
plasma membranes. Biochim Biophys Acta
Severe retinopathy of
prematurity (ROP) in a premature
baby treated with sildenafil
acetate (Viagra) for pulmonary
Sildenafil is used as a selective pulmonary
vasodilator in children with primary pulmon-
ary hypertension and severe lung fibrosis.
improves gas exchange, increasing life expec-
tancy and exercise tolerance. Recent animal
models of neonatal pulmonary hypertension
have also shown that sildenafil reduces
vascular resistance. This has encouraged its
use in units treating premature infants.
report a case of severe retinopathy of pre-
maturity in a preterm infant who was treated
with intravenous sildenafil for severe respira-
tory failure.
Case report
The patient was born at 26 weeks gestation
weighing 525 g. He was ventilated from birth
for respiratory insufficiency, secondary to
respiratory distress syndrome. He required
high flow oxygen ventilation and received
surfactant at delivery and 16 hours later. His
oxygen requirements then stabilised at 30
At 29 weeks his oxygen requirements
increased to 100% due to coagulase negative
Staphylococcus aureus and candida sepsis. He
was treated with ambisome and 5-flucyto-
sine. However, he was only able to maintain
oxygen saturations of 7080% while receiving
positive pressure ventilation on 8090% oxy-
Figure 1 Fundus photograph of 15 year old
patient at baseline examination.
Figure 2 Effect of DHA on retinal function in patient with ELOVL4 mutation.
Grant support: MSI Foundation, Alberta, Canada;
University of Alberta Hospitals Foundation, Olive
Young Fund; Royal Alexandra Hospital Foundation;
Canadian Institutes for Health Research (MH
Fellowship support).
Ethics: The health research ethics board of the
University of Alberta approved this study and written
consent was obtained before participation in the study
Conflict of interest: There are no financial agreements
between any of the authors and Martek Biosciences,
Columbia, MD, USA
306 PostScript on July 15, 2011 - Published by bjo.bmj.comDownloaded from
gen. At 31 weeks +2 days post-conception he
was commenced on inhaled nitric oxide (NO)
540 ppm. There was no demonstrable
improvement in oxygenation with the NO,
and therefore it was decided to start sildenafil
3 days later in addition to the NO. During the
period of treatment with sildenafil, there was
an improvement in his clinical condition,
with his oxygen requirements falling to 30
40% by 33 weeks. A subphrenic collection
was discovered at 33 weeks, and ciproflox-
acin and metronidazole were started. The
sildenafil was discontinued at 34 weeks of
life, after 16 days of treatment, because
of a rising alanine aminotransferase level.
There was no rebound hypoxia observed.
Other than intrauterine growth retard-
ation, no further risk factors for ROP such
as intraventricular haemorrhage were
The patient was examined by an ophthal-
mologist weekly from 31 weeks. No ROP was
seen up to and including the 33 week check.
At 34 weeks he was found to dilate poorly, to
have bilateral iris neovascularisation, hazy
media, and dilated and tortuous fundal
vessels. Peripherally he had developed 7
contiguous clock hours of stage III ROP in
zone II in the right eye, and 5 clock hours in
the left. He was treated the following day
with bilateral peripheral laser photocoagula-
tion to ischaemic retina. The ROP regressed
over the ensuing weeks and he remains
under review.
Sildenafil relaxes arteriolar smooth muscle in
the presence of nitric oxide by inhibiting
phosphodiesterase type 5 (PDE5). PDE5 is
found in high concentrations in the smooth
muscle of the corpus cavernosum, and in
lower concentrations in other tissues such as
vascular smooth muscle. NO may control
retinal blood flow. Measurement of ocular
blood flow changes following sildenafil have
shown conflicting results, with some groups
showing a significant increase and others
showing no increase in choroidal blood
In ROP, the initial constriction of retinal
vessels by the high levels of oxygen induces a
neovascular drive through the release of
growth factors such as vascular endothelial
growth factor. Subsequent retinal hyperper-
fusion has been linked to progression of other
neovascular disease such as diabetic retino-
by the local release of growth factors
and free radical production. In addition, NO
and cGMP accumulation caused by PDE5
inhibition has been proposed to exert a
proliferative effect on retinal post-capillary
Although sildenafil seems to have a unique
place in the treatment of preterm infants in
respiratory failure, this case may link its use
to the development of aggressive ROP. We
have observed a recent increase in treatable
ROP in our unit, coinciding with the use of
sildenafil. Further work on the retinal effects
of sildenafil may be of use in determining
whether it truly is a risk factor in the
pathogenesis of ROP.
C S Marsh, B Marden, R Newsom
Southampton Eye Unit, Southampton, UK
Correspondence to: Catherine Marsh, Southampton
Eye Unit, Southampton, UK;
Accepted for publication 28 April 2003
1 Abrams D, Schulze-Neick I, Magee AG.
Sildenafil as a selective pulmonary vasodilator in
childhood primary pulmonary hypertension.
Heart 2000;84:E4.
2 Shekerdemian LS, Ravn HB, Penny DJ.
Intravenous sildenafil lowers pulmonary vascular
resistance in a model of neonatal pulmonary
hypertension. Am J Resp Crit Care Med
3 Paris G, Sponsel WE, Sandoval SS, et al.
Sildenafil increases ocular perfusion. Int
Ophthalmol 2001;23:3558.
4 Grunwald JE, Siu KK, Jacob SS, et al. Effect of
sildenafil citrate on the ocular circulation.
Am J Ophthalmol 2001;131:7515.
5 Burton AJ, Reynolds A, ONeill D. Sildenafil
(Viagra) a cause of proliferative diabetic
retinopathy? Eye 2000;14:7856.
Indocyanine green localisation in
surgically excised choroidal
neovascular membrane in age
related macular degeneration
Clinical indocyanine green (ICG) angiogra-
phy has gained an established role in the
diagnosis of choroidal neovascularisation
(CNV) in age related macular degeneration
It is potentially able to localise CNV
with greater accuracy compared with fluor-
escein angiography because it fluoresces in
the infrared range, allowing imaging of
choroidal pathology through blood and pig-
Additionally ICG angiography may
better define CNV because of its high protein
binding affinity in serum,
resulting in
reduced leak from the CNV. This allows ICG
angiography to clearly delineate the feeding
vascular supply to the CNV allowing direct
ablative thermal laser to precisely target this
Treatments for CNV, including thermal
laser, photodynamic therapy, and macula
translocation are all reliant on the ability of
the clinician to diagnose accurately the
location and nature of CNV. Few clinico-
pathological data exist to correlate the ICG
angiographic appearance of CNV with its
histology. The presence of CNV was shown
pathologically in an eye with a plaque like
hyperfluorescence on ICG angiography
shown years previously.
Lafaut reported the
pathology of surgically removed recurrent
CNV in two eyes that were preoperatively
imaged with ICG angiography.
The hyper-
fluorescence of drusen on ICG angiography in
the monkey model has been correlated with
ICG dye that has been found histologically on
infrared fluorescence microscopy.
In this case report we hope to correlate ICG
angiographic findings in a patient with
exudative AMD with the histological localisa-
tion of ICG to the surgically excised CNV.
This will verify that ICG angiographic find-
ings are an accurate representation of the
location and nature of the CNV.
Declaration: The authors have no financial interest in
the products discussed in this paper
Figure 1 Clinical angiography. (A) Preoperative fluorescein angiography shows that the lesion is
well defined and extends into the fovea. (B) Preoperative mid-phase ICG angiogram shows the
feeding vascular network that is originating from a peripapillary site. (C) Postoperative fluorescein
angiography shows the removal of the CNV. (D) Postoperative ICG angiogram confirms the
removal of the CNV.
PostScript 307 on July 15, 2011 - Published by bjo.bmj.comDownloaded from
Case report
A 71 year old patient with classic subfoveal
CNV was recruited into this study to histolo-
gically localise ICG dye to excised CNV in
AMD. The study was granted institutional
research approval by the Sydney Eye Hospital
Ethics Committee.
Angiographic evaluation
The patient underwent preoperative imaging
with fluorescein and ICG angiography to
define the nature and location of the CNV.
The angiographic studies were performed on
the Heidelberg Scanning Laser
Ophthalmoscope (Heidelberg, Germany).
Diagnostic studies were performed using
5 ml of 10% fluorescein dye and 25 mg of
ICG dye (Pulsion Medical Suppliers,
Mu¨nchen, Germany).
Follow up angiographic studies were per-
formed one month following surgical
removal of the new vessel using standardised
methods as outlined above. This was done to
confirm the removal of the new vessel.
Surgical procedure
The patient underwent surgical removal of
the CNV under assisted peribulbar regional
anaesthesia 72 hours later. A three port trans
pars plana vitrectomy and removal of the
posterior vitreous hyaloid was performed. A
small retinotomy adjacent to the CNV was
made through which a localised retina
detachment was created. The CNV was
extracted with subretinal forceps through
the retinotomy. Haemostasis was achieved
by raising the height of the infusion bottle,
increasing intraocular pressure. A fluid-air
exchange was carried out after checking the
retina periphery for tears.
Histologic localisation of ICG dye
Fifteen minutes before excising the CNV,
50 mg of ICG dye was administered as an
intravenous bolus to allow the ICG to bind to
the CNV.
The excised specimen was taken immedi-
ately to be fixed in 20% dimethyl sulphoxide
(Sigma, NSW, Australia) plus 80% adult
fixative containing 2% paraformaldehyde in
0.1 M phosphate buffered saline (PBS)
(pH 7.4) for 20 minutes.
The tissue was embedded in a small tissue
embedding mold with Tissue-Tek OCT 4583
compound embedding medium (Bayer, NSW,
Australia) and then rapidly frozen in liquid
nitrogen. The frozen tissue block was
mounted onto Cryostat Microtome (Leitz,
Kryostat 1720, Australia) and 20 mm sections
were cut. Sections were collected with gele-
tin-chrom-Alum coated slides and air dried.
Sections were then examined for ICG infra-
red fluorescence or stained with cresyl violet
for histologic examination.
Infrared fluorescence microscopy
The technique of infrared fluorescence micro-
scopy to identify ICG fluorescence has been
previously described.
An 810 nm diode laser
(Iris Technologies) was used with an excita-
tion filter to stimulate the surgical specimen.
A Nikon Eclipse fluorescence microscope
with a custom 860 nm barrier filter
(Chroma Technologies, Vermont, USA) was
used to detect the emitted ICG fluorescence.
Paired bright field images with the corre-
sponding infrared fluorescence image were
used for orientation of the infrared fluores-
cence image and recorded digitally.
For histological studies, tissue sections
were stained by cresyl violet for 30 seconds,
rinsed and dehydrated in graded ethanol,
cleared in xylene, and mounted in DePex
(BDH, Melbourne, Australia) The sections
were examined and photographed under
light microscopy.
Angiographic appearances
Fluorescein angiography shows a well
defined hyperfluorescent lesion which
extends under the fovea consistent with
presumed CNV (Fig 1A). ICG angiography
shows a hyperfluorescent vascular lesion in
the same location (Fig 1B).
Postoperative fluorescein and ICG angio-
graphy (Fig 1C and D) reveal the absence of
the hyperfluorescence confirming that the
subfoveal lesion has been excised.
Histologic localisation of ICG
Figure 2A shows the light microscopic image
of the surgical specimen stained with cresyl
violet. The CNV has a fibrovascular structure.
There are associated patchy islands of intact
retinal pigment epithelium (RPE) monolayer
which are indicated by the arrowheads.
Figure 2B is a bright field microscopic
image of the surgical specimen. Figure 2C is
the corresponding fluorescence microscopic
image which shows ICG localised to the
excised CNV. Note that the small islands of
intact RPE cells indicated by arrows are
intensely fluorescent.
This report provides a direct clinicopatholo-
gical correlation of CNV defined on ICG
angiography, with ICG localised to the
excised surgical membrane by infrared fluor-
escence microscopy. This is unique in that the
ICG itself is detected within the excised CNV.
This report confirms that hyperfluorescence
defined on fluorescein and ICG angiography
is consistent with CNV. The hyperfluorescent
lesion shown preoperatively is absent follow-
ing surgical removal of the specimen.
Indocyanine green was histologically loca-
lised to the fibrovascular structure of the
CNV. It is proposed that ICG enters rapidly
into the neovascular complex through its
feeding vascular network. The ICG within the
feeding vascular network is able to be imaged
on the early phase clinical ICG angiography.
Subsequently, it is likely that the ICG
extravasates through the leaky vessels to
bind to the supportive stromal tissues. In
this later phase, clinical ICG angiography
does not show the feeding vasculature.
Histologically, the CNV was associated
with an incomplete layer of RPE cells which
were removed with the CNV. This suggests
that the CNV in the case studied appeared to
be sub-RPE in location. This is consistent
with the histopathology of specimens from
the Submacular Surgery Trial which would
indicate that the CNV in AMD tends to grow
under the RPE.
This report provides information regarding
the interaction of ICG with RPE. The intact
RPE that were removed with the CNV were
intensely fluorescent on infrared fluorescence
microscopy. The fluorescence of the RPE was
more intense than the fibrovascular tissue.
This suggests that that RPE cells accumulate
ICG. This finding corroborates the previous
histologic localisation studies in the monkey
and human model which showed ICG loca-
lisation to the RPE.
The frozen section technique was employed
because it allows the water soluble ICG to
remain within its in vivo location.
Conventional histological processing would
cause the ICG to leak through the tissue
planes misrepresenting its true position.
Previous histological localisation studies in
the rat model have been performed using
frozen section techniques.
Preservation of
tissue architecture in this study was limited
by crush artefact by the subretinal forceps
and subsequent extraction from the subret-
inal space. Additionally, the frozen section
processing technique results in tissue expan-
sion and disruption.
This clinicopathological correlation provides
the clinician with greater confidence in
diagnosing CNV based on ICG angiography.
The study was granted institutional research
approval by the Sydney Eye Hospital Ethics
A A Chang, M Zhu, F A Billson
Department of Clinical Ophthalmology & Save Sight
Institute, University of Sydney, Sydney, Australia
A A Chang
Sydney Eye Hospital, Sydney, Australia
A A Chang, N L Kumar, P E Beaumont
Sydney Retina Clinic, Sydney, Australia
Figure 2 Light and infrared fluorescence
microscopy. (A) Light microscopy of excised
CNV reveals patchy islands of intact RPE
monolayer as indicated by the arrowheads. (B)
Bright field image shows the surgical specimen
after frozen sectioning. The arrows indicate
intact RPE. (C) Infrared fluorescence microscopy
demonstrating that ICG fluorescence is localised
to the excised CNV. Areas of intact RPE cells
are more intensely fluorescent.
308 PostScript on July 15, 2011 - Published by bjo.bmj.comDownloaded from
Correspondence to: Dr Andrew Chang, Department of
Clinical Ophthalmology & Save Sight Institute, The
University of Sydney and Sydney Eye Hospital, GPO
BOX 4337, Sydney, NSW 2001, Australia;
Accepted for publication 4 May 2003
1 Yannuzzi LA, Slakter JS, Sorenson JA, et al.
Digital indocyanine green videoangiography and
choroidal neovascularization. Retina
2 Lim JI, Sternberg P Jr, Capone A Jr, et al. Selective
use of indocyanine green angiography for occult
choroidal neovascularization. Am J Ophthalmol
3 Kramer M, Mimouni K, Priel E, et al. Comparison
of fluorescein angiography and indocyanine
green angiography for imaging of choroidal
neovascularization in hemorrhagic age-related
macular degeneration. Am J Ophthalmol
4 Cherrick GR, Stein SW, Leevy CM, et al.
Indocyanine green: observations on its physical
properties, plasma decay, and hepatic extraction.
J Clin Invest 1960;39:592.
5 Shiraga F, Ojima Y, Matsuo T, et al. Feeder vessel
photocoagulation of subfoveal choroidal
neovascularization secondary to age-related
macular degeneration. Ophthalmology
6 Chang TS, Freund KB, de la Cruz Z, et al.
Clinicopathologic correlation of choroidal
neovascularization demonstrated by indocyanine
green angiography in a patient with retention of
good vision for almost four years. Retina
7 Lafaut BA, Aisenbrey S, van den Broecke C, et al.
Clinicopathological correlation in exudative age-
related macular degeneration: recurrent
choroidal neovascularization. Graefes Arch Clin
Exp Ophthalmol 2001;239:511.
8 Chang AA, Morse LS, Handa JT, et al. Histologic
localization of indocyanine green dye in aging
primate and human ocular tissues with clinical
angiographic correlation. Ophthalmology
9 Grossniklaus HE, Green WR. Histopathologic
and ultrastructural findings of surgically excised
choroidal neovascularization. Submacular
Surgery Trials Research Group. Arch Ophthalmol
10 Matsubara T. Histological localization of
indocyanine green in the retina and choroid.
Jpn J Clin Ophthalmol 1995;49:2533.
Correction of pseudophakic
anisometropia in a patient with
pseudoexfoliation using an
implantable contact lens
Pseudophakic anisometropia may cause sig-
nificant patient dissatisfaction with marked
visual problems. Correction methods include
contact lenses, intraocular lens (IOL)
exchange, piggybacking a supplementary
IOL, and corneal refractive surgery.
The Staar surgical implantable contact lens
(ICL) was first used to correct myopia and
hyperopia in phakic patients. It is now being
used to correct pseudophakic anisometro-
The following case describes the successful
management of pseudophakic anisometropia
using an ICL in a patient with high myopia
and pseudoexfoliation.
Case report
An 80 year old woman was referred to us for
correction of pseudophakic anisometropia.
She had bilateral pseudoexfoliation and was
highly myopic, her original refraction being
214.00 DS right eye, 213.50/+1.75620 left
In 1989 she underwent left cataract extrac-
tion with posterior chamber IOL, keeping her
highly myopic (213.00/22.006105 left eye).
She subsequently developed cataract in her
right eye so underwent uncomplicated right
phacoemulsification with posterior chamber
lens implantation, leaving her +1.00/
21.00645 in her right eye.
Because of her marked anisometropia she
experienced visual difficulties unresolved
with spectacles or contact lenses, so she
requested permanent correction of this.
Refraction, keratometry, corneal thickness
(using pachymetry), axial length (using
ultrasound), anterior chamber depth, and
horizontal white to white diameter (using
Orbscan), were measured and sent to Staar
Surgical who calculated the ICL power.
A 219.0 Dioptre Staar Collamer ICL was
inserted via an injector into the left sulcus,
anterior to her posterior chamber IOL,
through a corneal tunnel. Two iridotomies
were made. Antibiotic and steroid drops were
used postoperatively.
Postoperative refraction was right eye
+1.00/21.00645 6/12, left eye +3.00/
21.006105 6/12+2. Her symptoms resolved
and she was extremely pleased.
Insertion of a Staar Collamer ICL seems to be
an effective alternative method for correcting
anisometropia in pseudophakic patients.
All surgical options have their risks and
Corneal refractive surgery is irreversible
and complications include flap related pro-
blems, corneal scarring,
variable refractive
outcome, and regression.
Supplementary anterior chamber lenses
risk corneal endothelial cell loss,
abnormalities, and they need larger incision
sizes. Posterior chamber IOLs can be used in
but may risk interface opaci-
IOL exchange can be difficult, especially if
performed some years after the original
surgery where the capsular bag has shrunk
around the IOL,
increasing the risk of
capsular damage with subsequent vitreous
loss, and zonule damage, especially impor-
tant in patients with pseudoexfoliation
whose zonules are already weak.
A main complication using ICLs is pupil-
lary block glaucoma.
It can be avoided by
performing adequate iridectomies peropera-
tively. Other side effects include glare, halos
and lens decentration (minimised if accurate
white to white diameter is measured in order
to get an accurate fit).
Advantages include ICL power calculation
being based on actual refraction so repeat
biometry is not required. There is no ICL/IOL
touch because of vaulting of the ICL and so
perhaps less risk of interface opacities.
operation is minimally invasive with a small
incision, as ICLs are thinner than other IOLs
(60 mm) and more flexible. The small incision
size (3 mm) reduces further astigmatism.
ICL power and diameter calculations made
by Staar are devised for phakic eyes. No
alterations were made for our pseudophakic
patient. This may lead to error, though the six
patients of Hsuan et al
had no major errors.
Our patient was slightly hyperopic, prob-
ably because of the different ciliary sulcus
anatomy in pseudophakic patients, accentu-
ated further by pseudoexfoliation.
It is better to prevent postoperative refrac-
tive surprises, but if they do occur, insertion
of an ICL can be used to correct pseudophakic
anisometropia, especially in patients with
N N Ashraff, B V Kumar, A Das, A P Moriarty
Department of Ophthalmology, Stepping Hill Hospital,
Stockport, UK
Correspondence to: Nina Ashraff, Manchester Royal
Eye Hospital, Oxford Road, Manchester M13 9WH,
Accepted for publication 12 May 2003
1 Hsuan JD, Caesar RH, Rosen PH, et al. Correction
of pseudophakic anisometropia with the Staar
Collamer implantable contact lens. J Cataract
Refract Surg 2002;28:449.
2 Sanders DR, Brown DC, Martin RG, et al.
Implantable contact lens for moderate to high
myopia: phase 1 FDA clinical study with 6-month
follow-up. J Cataract Refract Surg
3 Smith LFF, Stevens JD, Larkin F, et al. Errors
leading to unexpected pseudophakic ametropia.
Eye 2001;15:72832.
4 Baikoff G, Colin J. Damage to the corneal
endothelium using anterior chamber intraocular
lenses for myopia (letter). Refract Corneal Surg
5 Saragoussi J-J, Cotinat J, Renard G, et al.
Damage to the corneal endothelium by minus
power anterior chamber intraocular lenses.
Refract Corneal Surg 1991;7:282285.
6 Mimouni F, Colin J, Koffi, et al. Damage to the
corneal endothelium from anterior chamber
intraocular lenses in phakic myopic eyes. Refract
Corneal Surg 1991;7:27781.
7 Eleftheriadis H, Marcantonio J, Duncan G, et al.
Interlenticular opacification in piggyback AcrySof
intraocular lenses: explantation technique and
laboratory investigations. Br J Ophthalmol
8 Assetto V, Benedetti S, Pesando P. Collamer
intraocular contact lens to correct high myopia.
J Cataract Refract Surg 1996;22:5516.
9 Rosen E, Gore C. Starr Collamer posterior
chamber phakic intraocular lens to correct
myopia and hyperopia. J Cataract Refract Surg
10 Zaldivar R, Davidorf JM, Oscherrow S. Posterior
chamber phakic intraocular lenses for myopia of
28to219 dioptres. J Refract Surg
Dopamine is an indicator but not
an independent risk factor for
grade 3 retinopathy of
prematurity in extreme low
birthweight infants
Retinopathy of prematurity (ROP) is a multi-
factorial disease with numerous risk factors.
Indicators besides birth weight or gestational
age might be used to further discriminate the
relative risk of developing ROP associated
with a given gestational age or birth weight.
Dopamine might be such an indicator since
Mizoguchi et al reported on the association of
dopamine and the risk of developing ROP in
survivors (n = 41) with a birth weight below
1000 g and suggested that this association
might be causal.
To document whether dopamine is an
indicator or an independent risk factor for
developing ROP, a retrospective chart review
in extreme low birthweight (ELBWthat is,
below 1000 g) infants in a single neonatal
intensive care unit during a 2 year period
(20001) was performed. Neonatal character-
istics collected were birth weight, gestational
PostScript 309 on July 15, 2011 - Published by bjo.bmj.comDownloaded from
age, and Apgar score. Respiratory character-
istics were either markers of duration of
respiratory disease (day at first extubation,
last day of respiratory support, last day of
supplemental oxygen) or markers of severity
of respiratory disease in the first 72 hours of
life: maximal mean airway pressure (MAP)
and oxygenation index (MAP 6 100 6
fractional oxygen/arterial oxygen mm Hg) at
maximal MAP. In addition, data on CRIB
(Clinical Risk Index for Babies, a disease
severity scoring system) score were collected.
The CRIB score was originally developed to
document risk for neonatal mortality but
there are reports on its association with
morbidity characteristics.
Finally, prescrip-
tion characteristics of dopamine (prescribed
or not, maximal dose, and total hours of
administration in the first week of life) were
Ophthalmological screening was initiated
at the postnatal age of 4 weeks by indirect
funduscopy after dilation. Screening was
continued until full vascularisation or until
diode laser therapy for threshold disease was
provided. Findings were classified according
to the International Classification of
Retinopathy of Prematurity.
Clinical charac-
teristics in infants who developed grade 3
ROP were compared with infants who did
not (Mann-Whitney U or x
Characteristics of dopamine administration
were entered in a multiple regression model
(MedCalc) with other significant risk factors
to document the potential independent effect
of dopamine on the risk of developing grade 3
Of 90 infants admitted with a birth weight
below 1000 g, 69 survived until discharge.
Retinopathy (any grade) was documented
in 45/69 (65%) survivors. Twenty seven (39%)
survivors developed grade 3 ROP and 15
(22%) of these infants developed threshold
ROP. In survivors (n = 69), gestational age
(p,0.001) and birth weight (p,0.003) were
significantly lower in infants who developed
grade 3 ROP while CRIB (p,0.002), oxyge-
nation index (p,0.02) but not MAP were
significantly higher in grade 3 ROP cases. All
markers of duration of respiratory disease
(ventilation (p,0.02), respiratory support
(p,0.001), supplemental oxygen (p,0.005))
were significantly higher in survivors who
developed grade 3 ROP.
Forty seven (68%) survivors received dopa-
mine. Incidence of any retinopathy, grade 3
ROP, and threshold ROP was higher in
infants who received dopamine during the
first week of life (table 1). Maximal dose
(p,0.05) was significantly higher and dura-
tion (hours) (p,0.03) in the first week of life
was significantly longer in infants who
developed grade 3 ROP. When the strongest
marker of cardiovascular instabilitythat is,
duration (hours) of dopamine administration
in the first week of life was entered in a
multiple regression model with either max-
imal oxygenation index or maximal MAP or
CRIB score, dopamine was no longer an
independent risk factor to develop threshold
In line with Mizoguchi, an association of
dopamine with an increased risk to develop
ROP was found.
In addition, dopamine was
no longer an independent risk factor of
developing grade 3 ROP after correction for
either CRIB score or oxygenation index in a
multiple regression model.
Although it is hypothesised that a syner-
gistic effect of hypoxia and dopamine
might induce retinal vascular constriction in
early neonatal life,
it is more likely that
this risk factor is merely an indicator of
cardiovascular instability and not an aetiolo-
gical factor involved in the pathogenesis of
K Allegaert, V Cossey, G Naulaers, C Vanhole,
H Devlieger
Department of Paediatrics, University Hospitals,
Leuven, Belgium
I Casteels
Department of Ophthalmology, University Hospitals,
Leuven, Belgium
Correspondence to: Karel Allegaert, MD, Department
of Paediatrics, University Hospitals, Gasthuisberg,
Herestraat 49, 3000 Leuven, Belgium; karel.
Accepted for publication 12 May 2003
1 Weathley CM, Dickinson JL, Mackey DA, et al.
Retinopathy of prematurity: recent advances in
our understanding. Br J Ophthalmol
2 Mizoguchi MB, Chu TG, Murphy FM, et al.
Dopamine use is an indicator for the development
of threshold retinopathy of prematurity.
Br J Ophthalmol 1999;83:4258.
3 The International Neonatal Network. The CRIB
(Clinical Risk Index for Babies) score: a tool for
assessing initial neonatal risk and comparing
performance of neonatal intensive care units.
Lancet 1993;342:1938.
4 Buhrer C, Grimmer I, Metze B, et al. The CRIB
score and neurodevelopmental impairment at one
year corrected age in very low birth weight
infants. Intensive Care Med 2000;26:3259,
(ofwel ref biology neonate).
5 International Committee for the Classification of
Retinopathy of Prematurity. An international
classification of retinopathy of prematurity. Arch
Ophthalmol 1984;102:11304.
6 Chen Z, Jia W, Kaufman PL, et al.
Immunohistochemical localization of dopamine-
beta-hydroxylase in human and monkey eyes.
Curr Eye Res 1999;18:3948.
7 Wu DM, Kawamura H, Li Q, et al. Dopamine
activates ATP-sensitive K+ currents in rat retinal
pericytes. Vis Neurosci 2001;18:9354.
Consent of the blind and visually
impaired: a time to change
It is a general legal and ethical principle that
valid consent must be obtained before start-
ing treatment or physical investigation of a
patient. The validity of consent does not
depend on the form in which it is given
(verbal or written), rather it depends on it
being given voluntarily by an appropriately
informed person who has the capacity to
consent to the intervention in question.
Written consent merely serves as evidence
of consent. There is no English statute setting
out the general principles of consent, rather it
is set by precedent or case law (‘‘common
law’’). Hence, in visually impaired people
(VIPs), instead of written consent should we
not have moved to a better alternative
recorded audio consent (RAC)legally
equivalent as a record of consent but practi-
cally superior as a method? There are many
good reasons why we should.
Firstly, health professionals are required to
take all reasonable steps to facilitate commu-
nication with the patient, using interpreters
or communication aids as appropriate.
Department of Health has issued guidance on
communication with patients who have
sensory disabilities.
For those who cannot
see, RAC would be more in line with these
guidelines and recommendations than writ-
ten consent.
Secondly, the problems with vision are
frequently compounded by poor hand-eye
coordination and manual dexterity especially
as most VIPs (2.4 million in England and
Wales) are over 65 years old.
RAC would
overcome the problems in these circum-
stances. Further, as many procedures VIPs
have are not eye related and the consenting
practitioner non-eye trained (therefore not
used to dealing with VIPs and their difficul-
ties), non-ophthalmic units would also ben-
efit from this technique.
Thirdly, we speak faster than we write and
in any event the whole process needs to be
undertaken verbally in the first instance. RAC
would be a time saving exercise.
Fourthly, ever advancing current technol-
ogies allow for large volumes of audio
recording in a more space efficient manner
than cumbersome written consent. These
devices (digital recorders, dictaphones, and
others) are now available on the mass market
at relatively low cost or already present
within the health services.
Fifthly, section 21 of the Disability Discri-
mination Act
states that service providers are
obliged to take reasonable steps to change
practices, policies, or procedures which make
it impossible or unreasonably difficult for a
disabled person to use a service. As from 2004
steps will have to be taken to remove
difficulties altogether. RAC works strongly
in conjunction with this act.
Finally, a practitioner trying to convince an
opposing party that a blind person had been
able to read the form they signed, thus
providing evidence of informed consent,
Table 1 Dopamine use in survivors (n = 69) and its association with the
incidence and severity of retinopathy in extreme low birthweight infants (p value
calculated using x
Dopamine No dopamine p Value
Number 47 22
No ROP 29% 45% NS
Any ROP 70% 54% NS
Grade 3 ROP 44% 27% NS
Threshold ROP 30% 5% ,0.04
G Naulaers is supported by the Fund fot Scientific
Research, Flanders (Belgium) by the Clinical Doctoral
Grant A6/5 CM. D 11.354
310 PostScript on July 15, 2011 - Published by bjo.bmj.comDownloaded from
would certainly find it more difficult to prove
than if the whole consent process had been
recorded with RAC.
RAC would thus benefit patients and
practitioners alike, facilitating consent and
making it a more robust process in the United
Kingdom and other countries alike.
G M Saleh
Essex County Hospital, Lexden Road, Colchester CO3
3NB, UK;
Accepted for publication 12 May 2003
1 Department of Health. Guidance for consent.
London: DoH (
2 Department of Health. Circulars HSC 1999/093
and HSC 1999/156. London: DoH.
3 The Royal College of Ophthalmology.
Epidemiology of eye disease in the older
population, 2.2.1 Visual Impairment and its
causes. London: RCO.
4 Disability Discrimination Act 1995:c 53, Pt III,
s 21.
Magnetic resonance angiography
source images in carotid
cavernous fistulas
Several investigations, including magnetic
resonance imaging (MRI), computed tomo-
graphy (CT), and orbital ultrasound are used
to non-invasively screen for carotid caver-
nous fistula (CCF), with variable results.
Examination of magnetic resonance angio-
graphy (MRA) source images, in addition to
the conventional MRA reconstructions, is
now also recognised as a useful method of
detecting CCF.
The finding of a hyperintense
signal in the cavernous sinus on MRA source
imaging provides additional, and sometimes
the only, neuroradiographic CCF evidence.
We present this patient to familiarise the
ophthalmologist with the typical CCF appear-
ance on MRA source images.
Case report
A 59 year old woman with a left eye pressure
sensation and a four year history of left ear
pulse-synchronous ‘‘buzzing’’ was found to
have elevated left intraocular pressure
(26 mmHg) on routine examination. She
denied diplopia, visual blurring, or pain.
Visual acuity was 20/25 OD and 20/20 OS.
Colour vision, visual fields, and pupils were
normal. There was left proptosis, ptosis, and
dilated episcleral vessels. Optic discs were
normal and there was mild left retinal venous
engorgement. There was a left supraorbital
pulse-synchronous bruit. Carotid cavernous
fistula was suspected. T1-weighted brain MRI
showed left superior ophthalmic vein (SOV)
enlargement (Fig 1). The cavernous sinuses
were symmetric and normal in size, with
slightly increased flow voids in the left
cavernous sinus. Conventional reconstructed
MRA did not reveal a fistula. However, MRA
source images showed an enlarged, hyper-
intense left cavernous sinus (Fig 2), support-
ing the presence of a CCF. Cerebral
angiography confirmed a left dural CCF.
Following endovascular therapy with trans-
venous and transarterial embolisation of the
fistula, all abnormal findings resolved.
Carotid cavernous fistulas are abnormal
communications between the carotid artery
and the cavernous sinus, either directly or via
intradural branches of the internal or exter-
nal carotid arteries.
Direct fistulas are high
flow, frequently follow trauma, and tend to
have a dramatic clinical presentation. In
contrast, indirect fistulas are low flow, often
spontaneous, and may have a subtle clinical
presentation. Symptoms and signs common
to both types of fistulas include proptosis,
chemosis, diplopia, visual loss, pulse-syn-
chronous tinnitus, orbital bruit, elevated
intraocular pressure, dilated episcleral veins,
and retinal venous congestion. The pattern of
venous drainage, either anterior into the
ophthalmic veins or posterior into the petro-
sal sinuses, often dictates the clinical findings
and radiographic appearance. Anterior drai-
nage typically leads to the most dramatic
ocular findings and enlargement of the
superior orbital vein, the latter often detect-
able with CT or MRI. However, superior
orbital vein enlargement is not specific to
Additional radiographic findings with
variable prevalence include lateral bulging of
the cavernous sinus wall and enlargement of
extraocular muscles on CT or MRI, and
abnormal cavernous sinus flow voids on
Conventional MRA reconstructed
images rarely provide additional useful infor-
mation. Direct visualisation of flow related
hyperintensity on the source images of three
dimensional time-of-flight MRA can be
extremely helpful in CCF detection, with
83% sensitivity and 100% specificity, far
superior to standard MRI.
Examination of
MRA source images is particularly useful in
the absence of anterior drainage and superior
orbital vein enlargement. Gadolinium-
enhanced MRA source images do not
increase diagnostic accuracy.
In our patient,
superior orbital vein enlargement and slightly
increased flow voids in the region of the CCF
were present on MRI, but MRA source
images provided the most striking and direct
evidence of CCF. Similar to other non-
invasive techniques, MRA source images
cannot delineate CCF arterial feeders or
detect cortical venous drainage. Hence, con-
ventional angiography remains necessary for
definitive management. However, non-inva-
sive techniques such as MRA source images
have an important role in pre-angiographic
diagnostic decisions.
This manuscript was supported in part by a
departmental grant (Department of Ophthal-
mology) from Research to Prevent Blindness, Inc,
New York, New York, and by core grant P30-
EY06360 (Department of Ophthalmology) from the
National Institute of Health, Bethesda, Maryland.
Dr Newman is a recipient of a Research to Prevent
Blindness Lew R Wasserman Merit Award.
J C Rucker
Departments of Neurology and Ophthalmology, Case
Western Reserve School of Medicine, Cleveland, OH,
V Biousse
Departments of Ophthalmology and Neurology,
Emory University School of Medicine,
Atlanta, GA, USA
N J Newman
Departments of Ophthalmology, Neurology, and
Neurological Surgery, Emory University School of
Medicine, Atlanta, GA, USA
Correspondence to: Dr Newman, Neuro-ophthalmol-
ogy Unit, Emory Eye Center. 1365-B Clifton Rd, NE
Atlanta, GA 30322, USA;
Accepted for publication 1 June 2003
1 Hirai T, Korogi Y, Hamatake S, et al. Three-
dimensional FISP imaging in the evaluation of
carotid cavernous fistula: comparison with
contrast-enhanced CT and spin-echo MR. AJNR
Am J Neuroradiol 1998;19:2539.
2 Chen J, Tsuruda J, Halbach V. Suspected dural
arteriovenous fistula: results with screening MR
angiography in seven patients. Radiology
3 Keltner JL, Satterfield D, Dublin A, et al. Dural
and carotid cavernous sinus fistulas.
Ophthalmology 1987;94:15851600.
4 Hirabuki N, Miura T, Mitomo M, et al. MR
imaging of dural arteriovenous malformations
with ocular signs. Neuroradiology
5 Komiyama M, Yoshihiko F, Hisatsugu Y, et al. MR
imaging of dural AV fistulas at the cavernous
sinus. J Comput Assist Tomogr
Figure 1 T1-weighted axial magnetic
resonance image. There is dilation of the left
superior ophthalmic vein (arrow).
Figure 2 Magnetic resonance angiography
source image shows an enlarged, hyperintense
left cavernous sinus (arrow).
PostScript 311 on July 15, 2011 - Published by bjo.bmj.comDownloaded from
Response to allegations and some
considerations on interferon
treatment in Behc¸etsdisease
Kotter et al refer to some problems, such as
fabrication of authorship, possibly data, and
ethical transgressions,
in our article published
in the Lancet.
However, they do so without
knowing the current facts about an ongoing
process. They cite accusations that rely on an
unfinalised investigation from the year 2000.
As we are now in 2003, I do feel that I have to
present updated correct information.
Before I do that, however, I would like to
note that the best scientific evidence against
fabrication of the results is its reproducibility
by other groups. The results of Kotter et al
show the beneficial effect of interferon in
Behc¸ets disease just as our results did,
our results are being increasingly reproduced.
The fact that the accusations were made on
the basis of an unfinalised investigation was
clear in the editorial
and letter
written about
our article at the time. The editor of the Lancet
stated that ‘‘... further investigations are in
and the Dean of our medical
school stated that ‘‘... the issue will be
finalised ... in a court of law.’’
I would like
to report on the decisions made by courts of
law during the 3 year period since then.
First of all, during ongoing inquires about
the article,
it was established without doubt
by an investigating commission that all my
coauthors had already known that their
names were included as coauthors before the
article was published. None of the coauthors
had objected to the inclusion of his/her name
at the time. It was not until 2 months after
the publication that some of the coauthors
claimed that they were unaware of inclusion
of their names. It is noteworthy that they did
so only after an ethical inquiry was embarked
upon. The issue of fabricated authorship was
brought to a court, accusing me of forging
signatures. The court unambiguously declared
my innocence at its first session on the matter
on 3 April 2003, concluding that there was no
forging of signatures and that it cannot be
imagined that the coauthors had been una-
ware of the article given the totality of the
circumstances surrounding the issue.
conclusion is supported independently.
attested by those court decisions, there is no
fabrication of authorship.
I was also accused of possible fabrication of
many of the patients.
However, in a declara-
tory action taken by another court, it was
definitely established that all 135 patients
mentioned in the Lancet article
were officially
registered at the Hacettepe University
Medical School.
At the same time, the highest adminis-
trative court, the State Council of Turkey
(Danistay) issued a stay order against any
administrative act due to the claimed ethical
in favour of me. This decision
was further approved at a plenary session
held at the State Council of Turkey with the
participation of members from all adminis-
trative chambers of the court.
Kotter et al are unintentionally perpetuat-
ing incorrect accusations about me in the
I am hoping the impropriety of those
accusations is clear in light of all the
independent court decisions I describe above.
Judicial decisions should be respected by
everyone who believes in upholding the
supremacy of law. It is my natural right to
respond and the readers of the BJO deserve to
be informed by updated correct information.
There is no indication that either author-
ship or the results reported in the Lancet
were fabricated. Besides judicial
decisions, the best scientific evidence that
the data are not fabricated is