Can Hyperbaric Oxygen Therapy Treat Cancer? What the Evidence Actually Shows

No. Hyperbaric oxygen therapy (HBOT) cannot treat cancer as a primary therapy, and no major oncology body — ASCO, ESMO, or the NCCN — lists it as a cancer treatment. Breathing pure oxygen at 2 to 3 atmospheres raises the oxygen tension in tumors and may shrink hypoxic zones, but it does not reliably kill malignant cells, does not substitute for surgery, chemotherapy, or radiation, and remains formally investigational as an anti-tumor modality. Its legitimate cancer-adjacent use is narrower: supporting healing of radiation-damaged tissue and salvaging complications, where the evidence is stronger and guideline-endorsed.

How HBOT Works Inside a Tumor

Hyperbaric oxygen therapy places a patient in a chamber at 2.0 to 3.0 atmospheres absolute (ATA) while they breathe 100% oxygen. At 3 ATA, arterial oxygen tension can climb above 2,000 mmHg, versus roughly 100 mmHg at sea level. That dissolved-oxygen load reaches tissue that ordinary blood flow cannot oxygenate well.

The cancer rationale rests on tumor hypoxia. Solid tumors outgrow their blood supply and develop cores with oxygen tension below 10 mmHg, compared with 40 to 60 mmHg in healthy tissue. Hypoxic cells resist radiation because radiation kills cells by generating oxygen free radicals — low oxygen means fewer radicals and roughly 2.5 to 3 times more radiation dose needed for the same kill. Hypoxia also selects for aggressive, metastasis-prone clones.

Reoxygenating those cells, in theory, should sensitize them to radiation and chemotherapy. That mechanism is real and measurable. What it does not do is directly kill tumor cells at pressure; oxygen at these doses is not cytotoxic to cancer the way it is to some anaerobic bacteria. In plain terms: HBOT changes the tumor's environment, and a better environment helps other treatments — it is not itself a weapon.

The Quantified Evidence: What Trials Show

The clinical evidence splits sharply by question. For HBOT as a radiosensitizer, the historical data show benefit in specific settings but nothing that has entered standard practice.

A landmark Danish trial (Horsman and colleagues, published in 1998) randomized over 300 women with advanced cervical cancer to radiation alone or radiation plus hyperbaric oxygen. The HBOT arm showed improved local control and a trend toward survival benefit — a genuine positive signal. Yet the trial used radiation equipment now obsolete, and the logistics of pairing HBOT with every radiation fraction (typically 25 to 35 sessions) were prohibitive. It was never replicated with modern radiotherapy and never adopted.

For chemotherapy sensitization, the picture is weaker. Small phase 2 studies of HBOT combined with chemotherapy in head and neck and bladder cancers reported response rates but lacked control arms. A 2019 systematic review of HBOT as an adjunct in cancer treatment found no randomized evidence supporting improved survival from adding HBOT to chemotherapy.

Here is the comparison that matters most:

The distinction within that table is the crux: HBOT's reputation in oncology comes almost entirely from treating the damage that cancer treatments cause, not from treating cancer itself.

Radiation Injury: Where HBOT Has Real Standing

The strongest cancer-adjacent evidence involves late radiation tissue injury. Radiation damages small blood vessels, producing a hypoxic, hypocellular, hypovascular tissue bed that heals poorly. HBOT at 2.0 to 2.4 ATA stimulates angiogenesis and fibroblast activity, and the effect is dose- and pressure-dependent.

Osteoradionecrosis of the jaw is the best-established indication. The Undersea and Hyperbaric Medical Society (UHMS) lists it as a primary indication, and the classic Marx protocol uses 20 preoperative dives at 2.4 ATA, then surgery, then 10 postoperative dives — typically 30 sessions of about 90 minutes each. Cure rates in cohort series run 70 to 85 percent, versus much lower rates for surgery alone in irradiated tissue.

For radiation proctitis and cystitis, randomized trials show modest but real improvement in bleeding and symptom scores, with response rates around 60 to 70 percent in some series. For radiation-induced lymphedema and soft-tissue necrosis, the evidence is thinner but supportive. These uses do not treat the cancer; they treat the collateral damage that makes survivorship difficult.

Head and Neck Cancer: The Most Studied Adjunct

Head and neck cancer is where HBOT and oncology intersect most. Two questions dominate: does it sensitize tumors to radiation, and does it prevent complications of treatment?

On sensitization, small trials and one older randomized study suggested improved local control when HBOT accompanied radiotherapy, but sample sizes were small (often under 50 patients) and the results have not held up in larger modern cohorts. The approach is not recommended outside trials.

On complication prevention, the evidence supports a narrower role. A randomized trial found that HBOT before dental extraction in previously irradiated patients reduced the risk of osteoradionecrosis. That is a genuine preventive use, not a treatment of the cancer.

The non-obvious caveat here matters: HBOT can transiently increase oxygen delivery to residual tumor cells, and some in-vitro work suggests this could theoretically promote tumor growth or metastasis through angiogenesis signaling. Clinical data have not shown increased recurrence in HBOT patients, but the theoretical concern is why broad HBOT use in active cancer is discouraged outside trials. In plain terms: the worry is not that HBOT feeds tumors — it is that the science of exactly who benefits is still unsettled.

Common Misconceptions and Red Flags

Several persistent myths deserve correction.

The idea that "cancer can't survive in oxygen" overstates a kernel of truth. Some cancer cells are hypoxia-tolerant and even thrive in low oxygen; reoxygenation makes them more vulnerable to radiation, not necessarily to oxygen itself. Oxygen is a radiosensitizer, not a toxin.

The claim that HBOT "boosts the immune system" against cancer is unsupported by controlled data at clinical pressures. Immune modulation has been studied, but no trial shows HBOT shrinking tumors via immunity.

The belief that more pressure is better is wrong and dangerous. Pressures above 3 ATA add toxicity risk without added anti-tumor benefit. Central nervous system oxygen toxicity can cause seizures, typically above 3.0 ATA with prolonged exposure.

Red flags to treat as stop signs:

• A clinic offering HBOT as a standalone cancer cure, or advising you to delay surgery, chemotherapy, or radiation for it.
• Pricing that exceeds USD 200 to 400 per session outside of a clinical trial or approved indication.
• Claims that HBOT "reverses" cancer without peer-reviewed survival data.
• Pressure settings above 3.0 ATA for cancer, or sessions exceeding 120 minutes.
• Providers who cannot state the ATA, session length, and number of planned treatments in writing.
• Any offer to replace guideline-directed oncology care.

These are warnings, not gray areas. HBOT has real risks — barotrauma to the ear and sinuses (the most common side effect, affecting roughly 1 in 5 patients mildly), oxygen toxicity, and rare pulmonary complications — so it belongs in supervised, indication-appropriate settings.

What to Actually Do If You Are Considering HBOT

If you have cancer and are weighing HBOT, the constructive path is specific.

Ask your oncologist whether HBOT is indicated for a complication you already have — osteoradionecrosis, radiation cystitis or proctitis, non-healing wounds in irradiated tissue, or pre-extraction prophylaxis. Those are the uses with evidence behind them.

If a provider offers HBOT as cancer treatment, ask for the trial protocol, the ATA, and the peer-reviewed survival data. Legitimate investigational use runs through registered clinical trials, not walk-in clinics.

If you are entering a trial or approved use, confirm the pressure (typically 2.0 to 2.4 ATA for tissue injury), session length (60 to 120 minutes), and total sessions (often 20 to 40) before starting, and verify the facility can manage barotrauma and oxygen toxicity.

Continue all guideline-directed oncology care. HBOT does not replace it, and any advice to pause standard treatment is a signal to seek a second opinion.

Track your own outcomes — wound size, bleeding, pain scores — so the effect can be judged on evidence rather than hope.

Limits of the Evidence and Unresolved Questions

Honesty about what is unknown is part of the answer. No large randomized trial has tested HBOT plus modern chemoradiation with survival as the endpoint, and the trials most often cited are decades old and small. The Horsman cervical cancer trial, while positive, predates contemporary radiotherapy by a generation.

The mechanistic question of whether reoxygenation could ever promote tumor spread remains theoretical but unexcluded. Pressure and dosing are not standardized across cancer-adjacent uses, so results from one protocol cannot be generalized to another; a 2.4 ATA wound-healing protocol is not interchangeable with a 3.0 ATA experimental regimen.

The reconciliation point: the tissue-injury evidence is strong enough to support guideline-endorsed use, while the anti-tumor evidence is too thin to support cancer treatment claims — and those two facts are not in conflict, because they answer different questions at different pressures for different purposes.

For now, the accurate sentence is this: HBOT treats the damage cancer treatment causes, not cancer itself, and anyone claiming otherwise is ahead of the evidence.