Acta Ophthalmologica 2019

Case Series

Macular perfusion velocities in the ocular

ischaemic syndrome
Oliver Niels Klefter,1,2 Peter Kristian Kofoed,1 Inger Christine Munch2,3 and Michael Larsen1,2
1
Department of Ophthalmology, Rigshospitalet, Glostrup, Denmark
2
Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark
3
Department of Ophthalmology, Zealand University Hospital, Roskilde, Denmark

ABSTRACT. symptoms, ophthalmoscopic, angio-

Purpose: To assess retinal perfusion in eyes with unilateral ocular ischaemic graphic and paraclinical findings
syndrome (OIS) and to compare with control subjects. (Marx et al. 2004; Kofoed et al.
Methods: Retrospective case series. Linear blood flow velocities in macular 2010), which in our clinical practice
vessels were estimated using motion-contrast fundus photography in eight includes ocular pneumoplethysmogra-
patients with unilateral OIS (eight OIS eyes, seven fellow eyes) and 12 control phy (OPG) to assess ocular systolic
subjects. The diagnosis of OIS was supported by carotid artery Doppler perfusion pressure (Gee 1982) and
ultrasonography and pneumoplethysmographic measurement of ocular systolic motion-contrast fundus photography
perfusion pressure. to assess macular perfusion velocities
(Izhaky et al. 2009). The present study
Results: Macular arterial blood flow velocity (median, range) was 1.8 (1.4–2.7)
retrospectively compared macular per-
mm/s in OIS eyes, 4.0 (2.9–5.3) mm/s in fellow eyes (p = 0.016) and 3.8 (2.3–
fusion measurements in eyes with OIS,
5.1) mm/s in control eyes (p = 0.0004 and p = 0.67 versus OIS and fellow eyes, non-OIS fellow eyes and healthy
respectively). Macular venous blood flow velocity was 1.5 (1.0–2.1) mm/s in OIS control eyes.
eyes, 2.6 (2.0–2.9) mm/s in fellow eyes (p = 0.016) and 2.7 (1.8–3.5) mm/s in
control eyes (p = 0.0007 and p = 0.64). Arterial velocities were below or equal to
the lowest value observed in control subjects (≤2.3 mm/s) in seven of eight eyes Materials and Methods
with OIS. Visual acuity 0.7 or worse was found in two OIS eyes with arterial
velocities below 1.7 mm/s and venous velocities below 1.3 mm/s and together Subjects
with neovascular glaucoma or polycythemia vera (one eye each).
This retrospective case series included
Conclusion: Motion-contrast imaging revealed markedly reduced macular
eight patients with unilateral ocular
perfusion velocities in OIS eyes compared with unaffected fellow eyes and ischaemic syndrome (three men, five
healthy control eyes. The method appears to provide a clinically meaningful women, median age 71.5 years, range
quantitative measure of macular hypoperfusion. 64–77 years; Table 1) seen at our insti-
tution between June 2011 and October
Key words: ocular ischaemic syndrome – perfusion velocities – retina – retinal blood flow 2014. All patients had unilateral OIS.
The clinical diagnosis was based on
Acta Ophthalmol. 2019: 97: 113–117 previously published criteria (Kofoed
ª 2018 Acta Ophthalmologica Scandinavica Foundation. Published by John Wiley & Sons Ltd et al. 2010) including ultrasonographic
or radiological demonstration of clini-
doi: 10.1111/aos.13950
cally significant ipsilateral carotid
hypertension, diabetes mellitus, and artery stenosis and OPG findings of
Introduction increased risk of ischaemic heart subnormal ocular perfusion pressure
The ocular ischaemic syndrome (OIS) disease and cerebrovascular events (Gee 1982).
is a relatively rare vision-threatening (Sivalingam et al. 1989). Although All patients had a history of cardio-
disease commonly caused by severe long-standing ocular hypoperfusion vascular disease, five patients had addi-
carotid artery stenosis (Brown & can lead to various pathological man- tional systemic comorbidities (Table 2),
Magargal 1988; Sivalingam et al. ifestations in the eyes (Mendrinos et al. and all patients received pharmacother-
1991; Mendrinos et al. 2010). Patients 2010) no uniform diagnostic criteria apy for various medical conditions.
often have systemic comorbidi- exist for OIS. The diagnosis thus relies One OIS eye underwent peripheral reti-
ties, notably atherosclerosis, arterial on the evaluation of a spectrum of nal photocoagulation because of iris

113
Acta Ophthalmologica 2019

Table 1. Clinical characteristics of patients with unilateral ocular ischaemic syndrome and control subjects with healthy eyes.

OIS patients Control subjects

OIS eyes Fellow eyes p-Value* Healthy eyes p-Value†

Gender (men/women) 3/5 2/10 0.35

Smokers (never/active or previous) 1/7 6/6 0.16
Age (years) 72 (64–77) 60 (52–69) 0.0032
Mean arterial blood pressure (mmHg) 105 (97–120) 102 (86–119) 0.56
Best-corrected visual acuity (Snellen) 0.85 (0.03–1.25) 1.0 (0.1–1.25)‡ 0.67 1.25 (1.0–1.25) 0.0025; 0.015
Intraocular pressure (mmHg) 13 (8–26) 14 (10–15) 1.00 16 (11–19) 0.11; 0.032
Ocular systolic pressure (mmHg) 78 (48–93) 100 (95–110) 0.016 NA NA
Arterial perfusion velocity (mm/s) 1.8 (1.4–2.7) 4.0 (2.9–5.3) 0.016 3.8 (2.3–5.1) 0.0004; 0.67
Venous perfusion velocity (mm/s) 1.5 (1.0–2.1) 2.6 (2.0–2.9) 0.016 2.7 (1.8–3.5) 0.0007; 0.64

Gender and smoking status are presented as numbers while all other variables are presented as median (range). Ocular systolic pressure was measured
at the time of diagnosis. All other results represent the measurements closest in time to the macular perfusion assessment.
OIS, ocular ischaemic syndrome; NA, not available.
* Two-sided p-value for paired comparison between eyes with ocular ischaemic syndrome and fellow eyes (Wilcoxon signed-rank test).
†
Two-sided p-value for unpaired comparison between control subjects and patients with ocular ischaemic syndrome (Wilcoxon two-sample test). If
two p-values are presented, the first p-value represents a comparison with eyes with ocular ischaemic syndrome and the second p-value represents a
comparison with asymptomatic fellow eyes.
‡
Includes one amblyopic eye.

neovascularization with neovascular Capital Region of Denmark). The tubing is placed posterior to the limbus
glaucoma for which the patient also control subjects were selected on the on the inferotemporal quadrant of the
received topical latanoprost, dorzo- basis of a normal eye examination and eyeball and vacuum suction is applied
lamide and timolol. Another OIS eye absence of systemic cardiovascular and at 300 or 500 mmHg. The vacuum is
had sequels of an extramacular branch metabolic disease. One control subject then gradually released with simulta-
retinal artery occlusion. Both eyes in received prophylactic statin treatment, neous registration of pressure oscilla-
two patients had macular drusen but no one received oestrogen replacement tions by a side port of the tube. The
choroidal neovascularization or geo- therapy and two received thiazide vacuum pressure, at which the first
graphic atrophy. An epiretinal mem- diuretics for lower extremity oedema. pressure oscillation is detected, is used
brane was found in one OIS eye. Two to calculate the ocular systolic pressure
patients had undergone cataract surgery (OSP). The lower normal limit for the
Procedures
in both eyes and one patient had under- OSP is estimated by the formula:
gone cataract surgery in an eye that had All subjects underwent ophthalmolog- OSP = 39 + 0.43 9 brachial systolic
subsequently developed OIS. No patient ical testing including measurement of blood pressure (Gee 1985). OSPs below
had untreated cataract. best-corrected visual acuity (BCVA) this level were considered subnormal.
Seven of the eight OIS patients had using Snellen charts, applanation In one patient, the numerical OSP
100% occlusion of the ipsilateral inter- tonometry, slit lamp examination, values could not be extracted but
nal or common carotid artery and one dilated ophthalmoscopy, transfoveal pressure meter observations character-
patient had 70–80% stenosis. The con- optical coherence tomography and fun- istic of unilateral OIS were made by an
tralateral internal carotid artery had dus photography, including motion- experienced examiner.
50–69% stenosis in three patients and contrast fundus photography (Retinal Macular perfusion velocity was mea-
19% stenosis in one patient. In three Function Imager 3005; Optical Imag- sured at a median interval of
patients, contralateral carotid artery ing Inc., Rehovot, Israel). Blood pres- 21 months (0–48 months) after the ini-
blood flow was described as normal sure was measured and mean arterial tial diagnosis of OIS. A fundus camera
whereas the degree of stenosis was not blood pressure (MAP) was calculated recorded eight red-free 20-degree
reported. In one patient, there was no as MAP = 2/3*diastolic blood pres- fovea-centred images in rapid succes-
information about the contralateral sure + 1/3*systolic blood pressure. At sion during the cardiac diastole. At
carotid artery. No patient had under- the time of diagnosis, patients with least three image series were obtained
gone carotid artery surgery. suspected OIS underwent OPG (OPG per eye. As previously detailed (Klefter
The eyes of OIS patients were com- Gee; Electro-Diagnostic Instruments, et al. 2015), proprietary software (RFI
pared with 12 healthy eyes from a Burbank, CA, USA) and seven patients Browse, version 2.1.00.209, Optical
control group of 12 subjects (two men, underwent fluorescein angiography. Imaging Inc., Rehovot, Israel) was
10 women, median age 60 years, range Results of carotid Doppler ultra- used to calculate the average linear
52–69 years) consisting of the oldest sonography or computed tomography perfusion velocity in small perifoveal
participants in a protocolized cross- angiography were retrieved. arterioles and venules, respectively.
sectional observational study that fol- Ocular pneumoplethysmography is Statistical analyses were performed
lowed the tenets of the Declaration of designed to provide a quantitative esti- using SAS 9.1 (SAS Institute, Cary,
Helsinki and obtained informed con- mate of the systolic blood pressure in NC, USA).
sent (protocol number H-1-2012-048 the ophthalmic artery (Gee 1985). In Wilcoxon signed-rank test was used
approved by the Committee for brief, a suction cup connected to the for within-subject comparisons while
Biomedical Research Ethics in the pneumoplethysmograph via plastic Wilcoxon two-sample test was used for

114
 Acta Ophthalmologica 2019

between-subject comparisons. Rela- with patients’ fellow eyes (p = 0.016; Discussion

tionships between continuous variables Figs. 1 and 2) and healthy control eyes
were tested by linear regression. In the (arterial p = 0.0004; venous Motion-contrast retinal blood veloci-
case of nonlinearity, the Spearman p = 0.0007). Perfusion velocities did metry demonstrated significantly
rank correlation coefficient was calcu- not differ between the OIS patients’ reduced macular perfusion velocities in
lated. The level of statistical signifi- fellow eyes and healthy control eyes OIS eyes compared with non-OIS fellow
cance was set at p < 0.05. (arterial p = 0.67; venous p = 0.64). eyes and healthy control eyes. Hence,
Arterial velocities were below or the present case series demonstrated in a
equal to the lowest value observed in quantitative manner the slowing of
Results control subjects (≤2.3 mm/s) in seven macular perfusion in carotid artery
At the time of diagnosis, median BCVA of eight OIS eyes, and above this value stenosis. The results are in agreement
(range) was 0.85 (0.03–1.25) in the eight in all fellow eyes. Venous velocities with colour Doppler ultrasonographic
eyes of eight patients with unilateral OIS were below or equal to the lowest value studies that found reduced blood flow
and 1.0 (0.1–1.25, one eye amblyopic) in observed in control subjects (≤1.8 mm/ velocities in the ophthalmic and central
the eight non-OIS fellow eyes. Best- s) in five of eight OIS eyes, immediately retinal arteries of patients with ipsilat-
corrected visual acuity (BCVA) was above this value in two OIS eyes eral carotid artery stenosis (Hu et al.
unchanged at the perfusion velocity (1.9 mm/s) and within the range 1993; Wong et al. 1998; Paivansalo
assessment (Table 1). All eyes of the observed in control subjects in one et al. 1999; Kawaguchi et al. 2012).
control subjects had BCVA 1.0 or bet- OIS eye. In all fellow non-OIS eyes, Furthermore, semi-quantitative analy-
ter. Except for one OIS eye with neo- venous velocities were within the range sis of fluorescein angiograms showed
vascular glaucoma, the intraocular observed in controls. delayed retinal perfusion in relation to
pressure was below 21 mmHg in all Best-corrected visual acuity below carotid artery stenosis (Brown &
eyes. Patients with OIS were older than 0.8 was found in two OIS eyes with Magargal 1988) and prolonged arterio-
the control subjects (p = 0.0032). Mac- arterial velocities below 1.7 mm/s and venous transit times in OIS eyes, but
ular perfusion velocities were obtained venous velocities below 1.3 mm/s and with extensive overlap between OIS and
in eight OIS eyes, seven non-OIS fellow together with neovascular glaucoma or non-OIS eyes (Hansen et al. 2013).
eyes (image quality insufficient in one polycythemia vera (one eye each). Objective quantitative criteria for
fellow eye) and 12 healthy eyes of 12 There was no detectable correlation the diagnosis of OIS are desirable.
control subjects. between macular perfusion velocity Ocular perfusion pressure measure-
Arterial and venous perfusion veloc- and BCVA, ocular systolic pressure, ment and retinal blood velocimetry
ities were lower in OIS eyes compared age or smoking status. come some way towards meeting this
demand. Although setting a clinically
meaningful threshold using a single
parameter is unlikely to be successful,
Table 2. Systemic and ocular comorbidities in patients with unilateral ocular ischaemic syndrome. the demonstration of macular hypop-
erfusion by motion-contrast imaging
Ocular comorbidities may help confirm the diagnosis of OIS.
Therefore, it appears to be a relevant
Patient Systemic comorbidities OIS eye Fellow eye
component of a diagnostic algorithm.
1 Stroke, hypertension, None None The specificity of macular perfusion
hypercholesterolaemia velocity measurements should be
2 Stroke, hypertension, None None tested, however, by comparison with
hypothyroidism, conditions such as diabetic retinopa-
type 2 diabetes thy, retinal vein occlusion, uveitis and
3 Hypertension, Cataract surgery Cataract surgery central serous chorioretinopathy (Wil-
type 2 diabetes, surgery for
liamson & Baxter 1994; Noma et al.
atherosclerosis
in aorta and 2009; Burgansky-Eliash et al. 2010;
iliac arteries, Beutelspacher et al. 2011; Feng et al.
polycythemia vera 2013).
4 Stroke, hypertension Drusen, epiretinal membrane Drusen Visual acuity can remain normal at
5 Hypertension, Cataract surgery, None retinal perfusion rates that are well
hypercholesterolaemia, neovascular glaucoma below the range observed in control
peripheral arterial disease subjects. Notably, this observation was
6 Osteoporosis, Drusen Amblyopia, drusen,
made in a chronic condition with
orthostatic hypotension pigment
epithelium hypoperfusion of insidious onset. The
detachment response of the eye to acute hypoper-
7 Stroke Extramacular branch None fusion may differ.
retinal artery occlusion Potentially contributing and com-
8 Stroke, hypercholesterolaemia, Cataract surgery Cataract surgery peting causes of altered retinal perfu-
type 2 diabetes, sion were present in our patients,
heart failure, chronic including neovascular glaucoma, poly-
obstructive pulmonary disease
cythemia vera, smoking and diabetes

115
Acta Ophthalmologica 2019

(Hayreh 2007; Yang et al. 2013; Lehr perfusion. Also, the two measure- latter factors reflect the rarity of OIS
2000; Burgansky-Eliash et al. 2010). ments were not always obtained (Mendrinos et al. 2010). The associa-
The heterogeneity may complicate the concomitantly. Although mean ocular tion between the degree of carotid
interpretation of perfusion measure- perfusion pressure can theoretically be artery stenosis and macular perfusion
ments but it makes our group repre- estimated from mean brachial arterial could not be assessed from our data.
sentative of OIS patients in general pressure and the intraocular pressure Although OIS patients were older than
(Sivalingam et al. 1989) and it should (Rishi et al. 2013), OPG appears to the control subjects, age did not influ-
not detract from the potential of provide a more precise measure of the ence the evaluation of perfusion veloc-
motion-contrast velocimetry to identify larger pressure drop across a significant ities.
eyes with poor retinal perfusion. The carotid artery stenosis (Gee 1985). In conclusion, motion-contrast reti-
lack of association between OPG and Our results should be interpreted nal velocimetry demonstrated signifi-
perfusion velocity measurements may with caution because of the retrospec- cantly decreased macular perfusion
be due to the former representing tive design, the limited sample size and velocities in OIS eyes compared with
systolic and the latter diastolic the time it took to accrue the cases. The unaffected fellow eyes and healthy
control eyes. The method appears to
provide a clinically meaningful quanti-
p = 0.0004 tative measure of retinal perfusion.
Prospective studies should determine
p = 0.016 the prognostic value of macular per-
fusion velocity measurements, effects
of carotid revascularization surgery
and thresholds for retinal functional
impairment.

References
Beutelspacher SC, Serbecic N, Barash H,
Burgansky-Eliash Z, Grinvald A & Jonas
JB (2011): Central serous chorioretinopathy
shows reduced retinal flow circulation in
retinal function imaging (RFI). Acta Oph-
thalmol 89: e479–e482.
Brown GC & Magargal LE (1988): The ocular
ischemic syndrome. Clinical, fluorescein
angiographic and carotid angiographic fea-
Fig. 1. Arterial perfusion velocities in ischaemic, unaffected and healthy eyes. Perifoveal arterial tures. Int Ophthalmol 11: 239–251.
perfusion velocities (mm/s) in eight eyes with ocular ischaemic syndrome (OIS), seven non-OIS Burgansky-Eliash Z, Nelson DA, Bar-Tal OP,
fellow eyes and 12 healthy eyes in 12 control subjects. Lowenstein A, Grinvald A & Barak A
(2010): Reduced retinal blood flow velocity
in diabetic retinopathy. Retina 30: 765–773.
Feng X, Kedhar S & Bhoomibunchoo C
p = 0.0007
(2013): Retinal blood flow velocity in
patients with active uveitis using the retinal
function imager. Chin Med J (Engl) 126:
1944–1947.
p = 0.016
Gee W (1982): Carotid physiology with ocular
pneumoplethysmography. Stroke 13: 666–
673.
Gee W (1985): Ocular pneumoplethysmogra-
phy. Surv Ophthalmol 29: 276–292.
Hansen GL, Kofoed PK, Munch IC et al.
(2013): Retinal angiographic blood flowme-
try is reduced in the ocular ischaemic syn-
drome. Dan Med J 60: A4716.
Hayreh SS (2007): Neovascular glaucoma.
Prog Retin Eye Res 26: 1035–1485.
Hu HH, Sheng WY, Yen MY, Lai ST & Teng
MM (1993): Color Doppler imaging of
orbital arteries for detection of carotid
occlusive disease. Stroke 24: 1196–1203.
Izhaky D, Nelson DA, Burgansky-Eliash Z &
Grinvald A (2009): Functional imaging using
Fig. 2. Venous perfusion velocities in ischaemic, unaffected and healthy eyes. Perifoveal venous the retinal function imager: direct imaging of
perfusion velocities (mm/s) in eight eyes with ocular ischaemic syndrome (OIS), seven non-OIS blood velocity, achieving fluorescein angiog-
fellow eyes and 12 healthy eyes in 12 control subjects. raphy-like images without any contrast agent,

116
 Acta Ophthalmologica 2019

qualitative oximetry, and functional meta- Noma H, Funatsu H, Sakata K et al. (2009): Wong YM, Clark JB, Faris IB, Styles CB &
bolic signals. Jpn J Ophthalmol 53: 345–351. Macular microcirculation and macular Kiss JA (1998): The effects of carotid
Kawaguchi S, Iida J & Uchiyama Y (2012): oedema in branch retinal vein occlusion. Br endarterectomy on ocular haemodynamics.
Ocular circulation and chronic ocular J Ophthalmol 93: 630–633. Eye (Lond) 12(Pt 3a): 367–373.
ischemic syndrome before and after carotid Paivansalo M, Riihelainen K, Rissanen T, Yang HS, Joe SG, Kim JG, Park SH & Ko HS
artery revascularization surgery. J Ophthal- Suramo I & Laatikainen L (1999): Effect of (2013): Delayed choroidal and retinal blood
mol 2012: 350475. an internal carotid stenosis on orbital blood flow in polycythaemia vera patients with
Klefter ON, Lauritsen AØ & Larsen M (2015): velocity. Acta Radiol 40: 270–275. transient ocular blindness: a preliminary
Retinal hemodynamic oxygen reactivity Rishi P, Rishi E, Mathur G & Raval V (2013): study with fluorescein angiography. Br J
assessed by perfusion velocity, blood oxime- Ocular perfusion pressure and choroidal Haematol 161: 745–747.
try and vessel diameter measurements. Acta thickness in eyes with polypoidal choroidal
Ophthalmol 93: 232–241. vasculopathy, wet-age-related macular
Kofoed PK, Munch IC, Sander B et al. (2010): degeneration, and normals. Eye (Lond) 27:
Received on March 15th, 2018.
Prolonged multifocal electroretinographic 1038–1043.
Accepted on September 19th, 2018.
implicit times in the ocular ischemic syn- Sivalingam A, Brown GC, Magargal LE &
drome. Invest Ophthalmol Vis Sci 51: 1806– Menduke H (1989): The ocular ischemic
Correspondence:
1810. syndrome. II. Mortality and systemic mor-
Oliver Niels Klefter, MD, PhD
Lehr HA (2000): Microcirculatory dysfunction bidity. Int Ophthalmol 13: 187–191.
Department of Ophthalmology
induced by cigarette smoking. Microcircula- Sivalingam A, Brown GC & Magargal LE
Rigshospitalet
tion 7: 367–384. (1991): The ocular ischemic syndrome. III.
Valdemar Hansens Vej 3
Marx JL, Hreib K, Choi IS, Tivnan T & Visual prognosis and the effect of treatment.
DK-2600 Glostrup
Wertz FD (2004): Percutaneous carotid Int Ophthalmol 15: 15–20.
Denmark
artery angioplasty and stenting for ocular Williamson TH & Baxter GM (1994): Central
Tel: +45 38 63 44 46
ischemic syndrome. Ophthalmology 111: retinal vein occlusion, an investigation by
Fax: +45 38 63 46 69
2284–2291. color Doppler imaging. Blood velocity
Email: oliver.niels@gmail.com
Mendrinos E, Machinis TG & Pournaras CJ characteristics and prediction of iris neo-
(2010): Ocular ischemic syndrome. Surv vascularization. Ophthalmology 101: 1362– Supported by the University of Copenhagen and
Ophthalmol 55: 2–34. 1372. Øjenforeningen (Danish Eye Health Society).

117