Hyperbaric oxygen therapy (HBOT) is a treatment that administers 100% oxygen in a pressurized environment exceeding 1.0 ATA (atmospheric pressure). The Undersea and Hyperbaric Medical Society (UHMS) defines it as requiring at least 1.4 ATA, and its effects are dose-dependent.
The chambers used are broadly classified into two types.
The UHMS approves HBOT for 14 diseases/conditions. The category relevant to ophthalmology is “arterial insufficiency,” which includes CRAO and refractory ulcers. Other approved indications are as follows.
Use for ocular diseases other than those listed above is off-label.
The history of HBOT is long. In 1662, Henshaw devised the first pressurized chamber called “Domicilium.” Junod in 1834 and Pravaz in 1837 further developed it, and in 1879 Fontaine introduced a pressurized operating room. In 1921, Cunningham built the world’s largest chamber, which was dismantled in 1937. In the 1930s, the U.S. Navy adopted HBOT for decompression sickness treatment, and Boerema demonstrated its usefulness in animal experiments. The predecessor organization of UHMS was established in 1967.
The only ophthalmic indication officially approved by UHMS is central retinal artery occlusion (CRAO). Use for diabetic retinopathy, retinal vein occlusion, optic neuropathy, scleral melting, and ocular infections has also been reported, but these are all off-label and the quality of evidence is limited.
The physiological effects of HBOT are due to a complex mechanism centered on the increase in plasma dissolved oxygen under high pressure.
Enhanced Oxygen Supply
Hemoglobin saturation: Under normal conditions, hemoglobin is almost fully saturated, and additional oxygen binding is limited.
Increased plasma dissolved oxygen: Under high pressure (Henry’s law), the amount of oxygen dissolved in plasma increases significantly, promoting oxygen delivery to tissues.
Bubble reduction: Under high pressure, bubble size decreases, allowing oxygen to reach small blood vessels more easily.
Effects on Blood Vessels and Cells
Vasoconstriction: Elevated oxygen levels reduce NO production, causing vasoconstriction. However, tissue oxygen supply is maintained due to the hyperoxic state. Rapid vasodilation occurs after HBOT ends.
Enhanced leukocyte function: Hyperoxia improves the oxidative bactericidal capacity of leukocytes.
Antibacterial effects: Suppresses the production of Clostridium toxins. Synergistic effects with fluoroquinolones, amphotericin B, and aminoglycosides have also been reported.
It has also been suggested that increased oxidants may function as cellular mediators, contributing to fibroblast proliferation and wound healing. For detailed mechanisms, see Section 6: Detailed Mechanism of Action.
The rationale for HBOT in CRAO is to enhance oxygen supply to the ischemic retina. During HBOT, reversible vasoconstriction occurs in the ocular vasculature, but the increase in plasma oxygen in the choriocapillaris allows the inner retina to maintain adequate oxygenation via oxygen supply from the choroidal side. Animal studies have shown that even with retinal artery occlusion, the retina is sufficiently oxygenated under HBOT. An intact choroidal circulation is a prerequisite for successful HBOT.
Clinical evidence for HBOT in CRAO is presented below.
| Study design | Number of cases | Main results |
|---|---|---|
| Retrospective (2001) | HBOT group 35 vs control group 37 | Visual improvement in 82% of HBOT group vs 29.7% in control |
| Clinical trial (2000) | Started 1 day after symptom onset | No significant difference |
| Cochrane review | Integration of multiple studies | Evidence uncertain, RCT needed |
Retrospective case series suggest that HBOT provides a modest benefit for CRAO, but the Cochrane review points out the uncertainty of the evidence and calls for high-quality RCTs1). As part of the Eye stroke protocol, efforts are underway to reduce the time to HBOT initiation by placing fundus cameras and OCT in emergency departments1).
Prognostic factors in CRAO:
Starting within a few hours of symptom onset is considered best. Irreversible damage is thought to begin 1.5 hours after retinal ischemia, and a clinical trial that started HBOT one day after symptom onset did not show significant efficacy1). The presence or absence of a cherry-red spot may be more useful for prognosis than time.
Based on the hypothesis that a hyperoxic state reduces VEGF expression and improves disruption of the blood-retinal barrier (BRB). In cases of diabetic macular edema, 14 sessions of HBOT over one month improved visual acuity (right 20/125→20/63, left 20/320→20/160). In two clinical trials, 68% improved by 2 or more lines, with an average gain of 3.5 lines. A prospective cohort study also reported that HBOT has a thinning effect on the macula.
Controlled trials have not shown efficacy, and it is not currently standard treatment. However, there is a hypothesis that HBOT provides neuroprotective effects by downregulating apoptosis-related genes, and promising case reports exist2). Evidence is conflicting, and further research is needed.
The following safety conditions are established for performing HBOT.
UHMS recommends the following stepwise protocol for CRAO.
| Step | Content |
|---|---|
| ① | Pressurize to 2 ATA |
| ② | If vision improves, maintain for 90 minutes |
| ③ | If no improvement within 30 minutes, increase to 2.4 ATA (US Navy Table 6) |
| ④ | If still no improvement, discontinue or continue normobaric oxygen |
HBOT can administer 100% oxygen for a cumulative total of over 9 hours, and small studies have shown its effectiveness for CRAO1).
Systemic Complications
Oxygen convulsions: Effects on the central nervous system due to hyperoxia. Prevented by limiting session duration.
Middle ear barotrauma: Caused by eustachian tube dysfunction during pressurization and depressurization.
Pulmonary rupture: Complication due to lung overinflation during rapid decompression.
Claustrophobia: More common in monoplace chambers.
Temporary pulmonary dysfunction: May occur with long-term or high-frequency sessions.
Ophthalmic Complications
Hyperoxic myopia: Most frequent. Progresses by about 0.25 D per week, causing a change of 1 line or more in visual acuity in 60% of patients. Often resolves within 3–6 weeks after treatment ends.
Cataract formation: Oxidation of lens proteins due to reactive oxygen species (ROS) production. Requires caution with long-term treatment.
Eyelid twitching: Considered the most common sign of oxygen toxicity.
Increased intraocular pressure (in eyes with intraocular gas): Absolute contraindication. Causes severe elevation of intraocular pressure.
Hyperoxic myopia is the most common ocular complication experienced by patients undergoing HBOT.
| Item | Details |
|---|---|
| Frequency | Approximately 60% of patients show a change of 1 line or more. |
| Progression Rate | Approximately 0.25 D per week |
| Range of Myopic Shift | 0.5 to 5.5 D at 2.5 ATA (Lyne, 1978) |
| Recovery Period | Usually 3 to 6 weeks (up to 6 to 12 months) |
| Exception | Does not occur in pseudophakic eyes |
| Device difference | Less myopic shift with oronasal mask |
At 2.4 ATA, the average myopic shift after 30 sessions is reported to be 0.95 D. Changes in lens structural proteins and water distribution are considered mechanisms (see Section 6: Detailed Mechanism of Action).
Hyperoxic myopia can occur, causing a change of one or more lines in visual acuity in about 60% of patients. It usually resolves within 3–6 weeks after treatment ends. However, myopic shift does not occur in pseudophakic eyes (eyes with artificial lenses). Also, long-term treatment can lead to cataract formation, which is irreversible.
According to Henry’s law, the amount of oxygen dissolved in plasma increases with higher pressure. Under normal conditions, plasma-dissolved oxygen is minimal, and oxygen delivery to tissues depends almost entirely on hemoglobin. However, inhaling 100% oxygen under high pressure significantly increases plasma-dissolved oxygen, enabling hemoglobin-independent tissue oxygen supply.
Elevated oxygen levels reduce nitric oxide (NO) production, causing vasoconstriction. Normally, vasoconstriction implies decreased tissue oxygen supply, but under HBOT, plasma-dissolved oxygen is significantly increased, so tissue oxygenation is maintained and enhanced despite vasoconstriction. Rapid vasodilation occurs after HBOT ends.
In CRAO, oxygen supply from the central retinal artery to the inner retina is blocked. Under HBOT, plasma oxygen in the choriocapillaris increases, promoting oxygen diffusion from the outer to the inner retina. This allows oxygen supply to all retinal layers if choroidal circulation is intact. This is why intact choroidal circulation is considered a prerequisite for successful HBOT.
Changes in the structural proteins (crystallins) and water distribution within the lens are thought to be the main cause of myopic shift. The hyperoxic environment accelerates oxidation of lens proteins, altering the refractive power of the lens. In pseudophakic eyes (with artificial lenses), this change does not occur because the natural lens is absent.
HBOT may exert neuroprotective effects by downregulating the expression of apoptosis-related genes2). This mechanism provides the theoretical basis for research into its application in optic neuropathies (e.g., non-arteritic anterior ischemic optic neuropathy).
Currently, evidence for HBOT in CRAO is mainly from retrospective studies and case series; the Cochrane Review points out this uncertainty and calls for RCTs1). With the spread of the Eye Stroke Protocol, systems are being developed to expedite fundus photography and OCT in emergency departments and reduce time to HBOT initiation1).
Small-scale studies are accumulating showing efficacy for diabetic retinopathy and retinal vein occlusion (RVO). In particular, the mechanisms of VEGF expression suppression and BRB protection are attracting attention, and research on combination effects with existing anti-VEGF therapy is expected.
Mechanistic studies are progressing showing that HBOT exerts neuroprotective effects by downregulating apoptosis-related genes 2). Evidence accumulation is expected, especially for NA-AION (non-arteritic anterior ischemic optic neuropathy).
