Early Hyperbaric Oxygen Treatment for Nonarteritic Central Retinal Artery Obstruction


To compare hyperbaric oxygen treatment combined with hemodilution with hemodilution only in central retinal artery obstruction.


Retrospective, nonrandomized case series.


We reviewed records of all our patients diagnosed with central retinal artery obstruction between 1997 and 2010. In these patients, hyperbaric oxygen and hemodilution therapy had been administered routinely (oxygen group). Where hyperbaric oxygenation could not be performed, patients were underwent hemodilution only (control group). Patients with presenting visual acuity (VA) of up to 20/200 within 12 hours of onset were included in our analysis. Exclusion criteria included cilioretinal vessels or arteritic occlusion.


The oxygen group comprised 51 patients, and the control group comprised 29 patients. Mean baseline VA was counting fingers (oxygen group) and 20/1000 (control group; P = .1). Most other potential confounders, including duration of symptoms, also did not differ significantly at baseline. In the oxygen group, mean VA improvement was 3 lines ( P < .0001). This was sustained over a follow-up of 3 months ( P = .01). In the control group, mean improvement was 1 line ( P = .23 at discharge, P = .17 at follow-up). Differences between both groups were not significant ( P = .07 at discharge, P = .26 at follow-up). The number of patients gaining 3 lines or more was 38.0% versus 17.9% at discharge ( P = .06) and 35.7% versus 30.8% at follow-up ( P = .76).


We saw significant VA improvement after the combined treatment, but not when using hemodilution only. Confirming superiority of the combination treatment requires a randomized, prospective trial. A high number of nonresponders highlights the need to improve our understanding and treatment of hypoxia-related metabolic insults after central retinal artery obstruction.

To fulfill its distinct functions, the most prominent of which are light detection and stimulus processing in the framework of the visual system, the retinal tissue shows an extremely high level of oxygen consumption. The central retinal artery is responsible for the blood supply to the inner two thirds of the retina. Being a functional end artery, an occlusion or obstruction of this vessel leads to a sudden, painless visual loss. Commonly, this visual loss is permanent because of irreversible damage to the retinal tissue during the ischemic period before recanalization occurs.

Among the causes for retinal artery obstruction are thrombosis, embolism, and arteritis. On fundus examination, common findings are thinning of the retinal vessels and a whitish appearance of the posterior pole resulting from an edematous nerve fiber and ganglion cell layer. A cherry red spot at the fovea, as described by von Graefe more than 150 years ago, may indicate intact background choroidal perfusion. Fluorescein angiography confirms the filling defect of obstructed retinal vessels.

A variety of measures have been advocated in an attempt to treat retinal artery obstruction. These include ocular massage, anterior chamber paracentesis, or intraocular pressure-lowering medication with the intention of dislodging the embolus downstream and increasing retinal perfusion pressure. Vasodilatation of the retinal vessels is the underlying idea of sublingual isosorbide dinitrate and carbon dioxide or carbogen inhalation. Rheologic measures such as hemodilution or application of pentoxifylline aim at improving perfusion so that retinal tissue can be preserved until spontaneous recanalization occurs. Other more recent concepts are intra-arterial fibrinolysis, surgical embolectomy, or neodyium:yttrium–aluminum–garnet laser embolysis. This list is not exhaustive. The sheer number of different treatment methods that have been investigated over the years indicates that, thus far, convincing evidence demonstrating efficacy of any particular intervention is not available. Hayreh and Zimmerman therefore stated that currently no treatment is available for central retinal artery occlusion (CRAO).

Under physiologic conditions, hemoglobin saturation is close to 100% when breathing room air at ambient pressure. However, breathing 100% oxygen under hyperbaric conditions can increase the amount of oxygen physically dissolved in the plasma to 6 vol%, up from 0.3 vol% under normal circumstances. This amount of plasma-dissolved oxygen on its own is able to meet the oxygen requirements of the body’s tissues. It has been shown that with increased partial pressure of oxygen achieved via hyperoxia, the choroidal vasculature is able to deliver an increased amount of oxygen to the ocular tissues, including vitreous body and inner retina, via diffusion. This constitutes the rationale behind the administration of hyperbaric oxygen via medical compression chambers in patients with a retinal vessel obstruction to support tissue survival until reperfusion is established.

A number of case series in which hyperbaric oxygen treatment has been used in CRAO have been reported. Their authors suggest that hyperbaric oxygen treatment shows some beneficial effects on visual acuity (VA) while entailing a relatively low risk profile. Encouraged by such reports, starting in 1997, hyperbaric oxygen treatment was offered routinely to patients diagnosed with CRAO at our center in combination with hemodilution treatment. The purpose of the study described herein was to analyze retrospectively the performance of the combined treatment in these patients and to compare it with that of hemodilution alone.


We analyzed medical records of all patients admitted to our clinic between 1997 and 2010 with visual loss from angiographically confirmed CRAO. Patients seeking treatment within 12 hours of onset of symptoms with VA of 20/200 or worse were included in the analysis. Administering both hyperbaric oxygen and hemodilution treatment had been the goal in all patients. These patients make up the oxygen group. Only where systemic contraindications impeded administration of hyperbaric oxygen, if the patient declined this procedure, or in cases where the medical compression chamber was not available, did patients receive hemodilution only. These patients make up the control group.

Patients found to have a nonobstructed cilioretinal vessel were excluded from analysis, as were those with an established diagnosis of arteritic obstruction. Further exclusion criteria were absence of choroidal perfusion (as determined by funduscopic examination, fluorescein angiography, or both), severe prior visual loss (worse than 20/40, as established by personal history and previous medical recordings, where available), or previous intraocular surgery (except cataract surgery).

Hyperbaric oxygen treatment was conducted according to the Marx protocol, which entails 10 minutes of compression, followed by a hyperbaric phase of 90 minutes and 15 minutes of decompression. Treatment was carried out in a StarMed 2200/5.5 multiseat chamber (HAUX-LIVE-SUPPORT, Karlsbad, Germany). Patients were administered 100% oxygen via a face mask at a maximum ambient pressure of 2.4 atmospheres absolute. The treatment scheme aimed at administering hyperbaric oxygen treatment 5 times within 48 hours, with 3 treatments within the first 24 hours. Where fewer than 5 treatment sessions were achieved, this was because the patient was unavailable because he or she underwent diagnostic procedures to establish a possible causative factor for the vascular event. In cases where marked improvement in VA was noted, but funduscopy, angiography, or both failed to show reperfusion, additional hyperbaric oxygen treatment sessions were performed.

Hemodilution aimed at reaching or maintaining a hematocrit value of less than 40%. This was achieved using a combination of lactated 500 ml Ringer’s solution (Fresenius Kabi AG, Bad Homburg, Germany) and 500 mL 6% hydroxyethyl starch 130/0.4 in 0.9% saline solution (Voluven; Fresenius Kabi AG) administered once daily. To achieve sustained lowering of the hematocrit in cases where this was markedly above the therapeutic target, application of intravenous fluid was flanked by extraction of up to 400 ml blood from a peripheral vein as judged necessary by the attending physician.

All therapeutic interventions were conducted as inpatient procedures. In cases where after discharge the patient was not seen again at our center, follow-up data on VA was obtained from the patient’s referring ophthalmologist. VA was recorded using Snellen optotypes and was converted to the logarithm of the minimal angle of resolution scale for analysis. To compare means or groups, the paired or unpaired t test, chi-square test, and linear regression analysis were performed using SPSS Statistics version 19.0 (IBM, Armonk, New York, USA).


We included 51 patients in the oxygen group; 29 patients entered the control group. Medical reasons for not administering hyperbaric oxygen treatment to patients then entering the control group included chronic obstructive pulmonary disease, inability to perform the Valsalva maneuver after acute upper respiratory tract infection, or presence of a cardiac pacemaker.

Age ranged from 35 to 82 years in the oxygen group and from 47 to 90 years in the control group. Age, sex, cardiovascular risk factors, but particularly duration of visual loss and pretreatment VA did not differ significantly at baseline ( Table 1 ). Differences were found in medication, with 15 (29.4%) of 51 patients in the oxygen group taking acetyl salicylic acid (100 mg/day), compared with 18 (62.1%) of 29 patients in the control group ( P = .01).


Baseline Characteristics of Patients with Nonarteritic Central Retinal Artery Obstruction before Receiving Early Hyperbaric Oxygen Treatment plus Hemodilution (Oxygen Group) or Hemodilution Only (Control Group)

Oxygen Group Control Group P Value
Gender distribution (M/F) 27/24 15/14 1.00
History of hypertension (yes/no) 36/15 25/4 .19
History of diabetes (yes/no) 5/46 5/24 .54
History of vasculopathy (coronary heart disease, stroke, peripheral vascular disease; yes/no) 16/35 16/13 .06
History of regular acetyl salicylic acid intake (yes/no/NA) 15/34/2 18/9/2 .01
Mean age ± SD (year) 69 ± 10.8 74 ± 9.6 .06
Mean VA on presentation ± SD (logMAR) 1.8 ± 0.3 1.7 ± 0.3 .10
Mean duration of visual loss ± SD (hour) 5.3 ± 2.9 5.7 ± 3.5 .61
Mean of highest systolic blood pressure ± SD (mm Hg) 169 ± 21.2 167 ± 26.5 .75
Mean of lowest systolic blood pressure ± SD (mm Hg) 112 ± 16.0 105 ± 15.4 .09
Mean of average systolic blood pressure ± SD (mm Hg) 136 ± 15.6 132 ± 14.9 .34
Mean cholesterol level ± SD (mg/dL) 216 ± 40.5 207 ± 41.7 .40
Mean triglyceride level ± SD (mg/dL) 156 ± 95.8 175 ± 153.2 .52

F = female; logMAR = logarithm of the minimal angle of resolution; M = male; NA = no data recorded; SD = standard deviation; VA = visual acuity.

The oxygen group received a median of 5 treatment sessions (range, 2 to 8), distributed over a median of 3 days (range, 2 to 9 days). Mean hematocrit reduction was similar in both groups with −4.5% (standard deviation, 3.9%) for the oxygen group versus −3.9% (standard deviation, 3.9) for the control group ( P = .64). No adverse ocular or systemic events were documented in any of the patients included in this analysis.

Time elapsed between onset of symptoms and establishment of the diagnosis (which would be followed by treatment within approximately 1 hour) were 5.3 hours in the oxygen group versus 5.7 hours in the control group ( P = .61; Table 1 ). In both groups, individual values ranged between 1 and 12 hours. No statistical correlation was found between time to diagnosis and improvement in VA ( r = 0.06).

In the oxygen group, we saw a mean improvement in VA of 3 lines compared with baseline ( P < .0001). In the control group, mean improvement was 1 line ( P = .23). Follow-up of 3 months showed some further improvement in both groups ( Table 2 ). VA remained significantly different from baseline in the oxygen group, but not in the control group. By direct comparison between both groups, the differences in mean change of VA failed to show statistical significance ( P = .07 on discharge, P = .26 at 3 months; Table 3 ). Mean VA on discharge was 20/260 and 20/800, respectively ( P = .55). At follow-up, this remained largely unchanged ( Table 3 ).


Comparing Treatment Outcomes within Groups of Patients with Nonarteritic Central Retinal Artery Occlusion Who Received Early Hyperbaric Oxygen Treatment plus Hemodilution (Oxygen Group) or Hemodilution Only (Control Group)

Mean Baseline VA ± SD (logMAR) Mean Discharge VA ± SD (logMAR) Mean Follow-up VA ± SD (logMAR) P Value
Oxygen group 1.8 ± 0.3 (n = 51) 1.5 ± 0.5 (n = 51) <.0001 a
1.8 ± 0.3 (n = 51) 1.5 ± 0.6 (n = 28) .01 b
Control group 1.7 ± 0.3 (n = 29) 1.6 ± 0.5 (n = 28) .23 a
1.7 ± 0.3 (n = 29) 1.6 ± 0.5 (n = 13) .17 b

logMAR = logarithm of the minimal angle of resolution; SD = standard deviation; VA = visual acuity.

a P value for Mean Baseline VA and Mean Discharge VA.

b P value compares Mean Baseline VA and Mean Follow-up VA.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jan 12, 2017 | Posted by in OPHTHALMOLOGY | Comments Off on Early Hyperbaric Oxygen Treatment for Nonarteritic Central Retinal Artery Obstruction

Full access? Get Clinical Tree

Get Clinical Tree app for offline access