Clival chordoma research
Home › Diet, supplements and claims

Hydrogen and hyperbaric oxygen around radiotherapy

Agent report for Kya, 13 September 2026, written after her partner asked in em-research whether it would be reasonable to use hydrogen up until two weeks before proton beam therapy and then hyperbaric oxygen afterwards to help healing. Working frame: probable clival chordoma in a 34-year-old woman, 51 kg, histology pending, no radiotherapy scheduled yet. Every PMID verified through NCBI E-utilities on 13 September 2026. No survival figures, by rule. Clinical reasoning is labelled.

Evidence levels. L1 randomised trial, meta-analysis or guideline. L2 prospective non-randomised trial or large cohort. L3 retrospective series. L4 case reports or very small series. L5 laboratory, animal or cell-line work, and expert reasoning.

The hydrogen background is in mark-shared/README.md (11 and 12 September entries) and is not repeated here. This thread covers the timing question and the hyperbaric oxygen half, which had no file anywhere in the repo before today.

Tier A. The question, and the short answer

Her partner's proposal, in his words: hydrogen "up until 2 weeks before proton beam and then use HBOT after to help with healing".

Three findings.

  1. Stopping hydrogen before radiotherapy is correct, and the reasoning is stronger than the hydrogen literature alone suggests, because there is a randomised human trial of a different antioxidant given during radiotherapy that came out badly.
  2. The plan as written has a gap. A proton course for a clival chordoma is roughly seven to eight weeks of daily fractions. "Stop two weeks before" leaves hydrogen running through the course itself, which is the window that matters most.
  3. The hyperbaric half is the weaker half, not the stronger one. It is evidenced for established late injury to bone and soft tissue, not for routine healing, it was negative in a UK randomised prevention trial, and the 2023 Cochrane update explicitly found no important effect on neurological tissue, which is the tissue the patient would most want protected. Separately, her particular anatomy raises a safety question that does not arise for most patients.

Tier B. Hydrogen during radiotherapy

The mechanism objection. Photon radiotherapy kills cells substantially by indirect action: the beam splits water, and the resulting hydroxyl radical does the DNA damage. Molecular hydrogen's claimed selectivity is precisely that it scavenges the hydroxyl radical and peroxynitrite while sparing hydrogen peroxide, superoxide and nitric oxide (Ohsawa 2007, Nat Med, PMID 17486089, L5). An agent whose selling point is hydroxyl scavenging, taken during a treatment that works through hydroxyl radicals, is not a neutral addition (clinical reasoning, L5).

The proton nuance, which cuts slightly in her partner's favour. Protons deposit most of their energy in the Bragg peak, where linear energy transfer is high and a greater share of the DNA damage is direct and clustered rather than radical-mediated. In the entrance plateau, LET is low and the physics resembles photons. A spread-out Bragg peak covering a 3.8 cm clival target spans a range of LET values. So the scavenging concern is plausibly smaller with protons than with photons, but it is not absent, and no study has measured whether an antioxidant blunts proton efficacy in a patient (clinical reasoning, L5). This is a reason to be slightly less worried, not a reason to continue.

The direct human evidence on hydrogen, which is thin and points the reassuring way. Kang 2011 (Med Gas Res, PMID 22146004, L2): 49 patients having radiotherapy for liver tumours randomised to hydrogen-rich water or placebo. Quality of life better, oxidative stress markers down, and no difference in tumour response. That is the only trial that bears directly on the question and it did not show tumour protection. It is also one small trial in a different tumour with a different beam.

The human evidence on antioxidants generally during radiotherapy, which points the other way and is much stronger. Bairati 2005 (J Natl Cancer Inst, PMID 15812073, L1): 540 patients with stage I or II head and neck cancer, double-blind, placebo-controlled, alpha-tocopherol 400 IU daily (initially with beta-carotene, stopped early on ethical grounds) beginning on the first day of radiotherapy and continuing three years. During the supplementation period, patients on alpha-tocopherol had a higher rate of second primary cancers (HR 2.88, 95% CI 1.56 to 5.31) and a higher rate of recurrence or second primary (HR 1.86, 95% CI 1.27 to 2.72). After supplementation stopped the direction reversed (HR 0.41 and 0.71, both crossing 1). The companion paper found the supplement did reduce acute radiotherapy toxicity (Bairati 2005, J Clin Oncol, PMID 16027437, L1), and a later analysis tied the local recurrence signal particularly to smokers (Meyer 2008, Int J Cancer, PMID 18059031, L1 secondary).

That trial is not about hydrogen and the mechanism is not identical. But it is the only randomised human test of the general proposition "take an antioxidant through radiotherapy and you will heal better", and the answer was that acute toxicity improved while cancer outcomes got worse during the exposure. That asymmetry is the argument. The benefit sought is symptomatic and the risk taken is oncological (clinical reasoning from L1 evidence).

Timing, on the pharmacology. Molecular hydrogen has no depot. Dissolved H2 is cleared within minutes to a couple of hours, largely by exhalation; the tablets her partner is looking at generate it in the glass and it is gone the same day (Ohsawa 2007, PMID 17486089, L5; product chemistry discussed in mark-shared/README.md, 11 September). Purely on the gas, stopping 24 to 48 hours before the first fraction would suffice.

The reason for a longer margin is secondary: hydrogen's durable effects are attributed to induction of the cell's own antioxidant programme, Nrf2 and downstream HO-1, catalase and superoxide dismutase, which takes days to a week to return to baseline after the stimulus is removed (L5 mechanistic literature). Nrf2 activation is an established radioresistance mechanism in tumours (L5). So a one to two week washout is conservative and costs nothing. Two weeks is a reasonable choice and there is no reason to argue him down from it.

The correction that matters. It must also be off for the whole course. Chordoma proton regimens run to roughly 70 to 74 Gy(RBE) in 35 to 39 daily fractions, seven to eight weeks. Restart after the last fraction, not before (clinical reasoning, L5).

Tier C. Hyperbaric oxygen after radiotherapy

What HBOT is for. Late radiation tissue injury is a months-to-years complication in which small vessels are lost and tissue becomes hypoxic, hypovascular and hypocellular. HBOT is given to raise tissue oxygen tension enough to drive angiogenesis and fibroblast function in that tissue. It is a treatment for established injury, not a recovery accelerant (L5 rationale; Dejonckheere 2026, CA Cancer J Clin, PMID 41385271, L5 review).

The current systematic evidence. Lin 2023, the fifth update of the Cochrane review (Cochrane Database Syst Rev, PMID 37585677, L1): 18 trials, 1,071 participants, publications 1985 to 2022, four studies new to this update. Conclusions, in the reviewers' own framing:

The prevention question, which is the one her partner is actually asking, and it is negative. HOPON (Shaw 2019, Int J Radiat Oncol Biol Phys, PMID 30851351, L1): UK randomised controlled trial of HBOT to prevent osteoradionecrosis of the irradiated mandible after dental extraction or implant placement, in patients who had received at least 50 Gy. Osteoradionecrosis at six months was 6.4 percent with HBOT versus 5.7 percent without (OR 1.13, 95% CI 0.14 to 8.92, P = 1). Fewer acute symptoms in the hyperbaric arm, no difference in anything else. This is the closest thing to a test of prophylactic HBOT "to help with healing" and it found nothing.

Brain and nerve specifically. For cerebral radiation necrosis the evidence is case series and small prospective work, with no double-blind controlled trial; a 2023 systematic review of management of symptomatic brain radiation necrosis after stereotactic radiosurgery found only one article relevant to HBOT out of 21 (L3/L4). Corticosteroids and bevacizumab have more support. This matters for the patient because the tissues in her radiation field are clival bone, cranial nerves III, IV, V and VI, the cavernous sinuses and the anterior brainstem. HBOT has a case for the bone. For the nerves, which are what her symptoms are made of, it does not (clinical reasoning from L1 and L3).

Does HBOT feed the tumour? The honest answer is reassuring. Moen and Stuhr 2012 (Target Oncol, PMID 23054400, L5 review) reviewed the long-standing fear that raising oxygen would promote growth or recurrence and found no evidence that HBOT stimulates tumour growth or enhances recurrence, with signals of tumour inhibition in some models; two systematic reviews cited there concluded HBOT in patients with malignancy is safe. There is no chordoma-specific data of any kind. Given that chordomas are demonstrably hypoxic (seven of nine residual skull base chordoma sites took up the hypoxia tracer FMISO, Mammar 2012, PMID 22391104, L4, discussed in thread 21), the theoretical arguments run in both directions and neither has been tested here.

Tier D. The safety question specific to the patient

This is the part a generic answer misses, and it is the most useful thing in this thread.

HBOT is delivered at 2.0 to 2.5 atmospheres absolute. Compression and decompression require every gas-containing space in the head to equalise through a patent opening. Middle ear and sinus barotrauma is the commonest complication of HBOT in ordinary patients.

The patient is not an ordinary patient for this purpose:

Pneumocephalus, an unrepaired skull base defect and CSF leak are treated as contraindications to HBOT, because a dural defect allows continuity between the pressurised environment and the intracranial compartment. Tension pneumocephalus after HBOT is documented, including after a single session (Bitterman 2009, Am J Emerg Med, PMID 19371560, L4; further cases in the rehabilitation literature, L4).

Separately, CNS oxygen toxicity seizures occur in HBOT at roughly 1 in 2,000 to 1 in 10,000 sessions in unselected patients. The patient has had four brief episodes since 5 September that may be temporal lobe seizures (metallic taste, unreality, phantosmia of rotting meat) and has an unresected mass abutting the brainstem. Her baseline risk is not the population baseline (clinical reasoning, L5).

None of this is a prohibition. It is the reason the decision belongs to her skull base surgeon and a hyperbaric physician jointly, after imaging of any repair, and not to a private clinic taking a booking.

Tier E. What this yields in practice

  1. Hydrogen now is a reasonable thing for her to try if she wants to, with the magnesium load counted against the renal caution (51 kg, on an NSAID, creatinine never seen).
  2. Stop it two weeks before the first proton fraction. Keep it off for the whole course. Restart after the last fraction if she wishes.
  3. The same rule should be applied to every other antioxidant she is taking, including high-dose vitamin C. Vitamin D is different: it is a hormone being replaced to a normal level, not an antioxidant, and the argument does not apply to it (clinical reasoning, L5).
  4. HBOT is not a post-radiotherapy healing programme. It is a treatment held in reserve for a specific, established late injury, most credibly osteoradionecrosis of the skull base, and even there the 2023 update softened.
  5. Before any HBOT, the skull base repair and the sinuses must be cleared by the surgeon and by ENT.

Questions for Mr Syrris, 16 September.

  1. Does she need to stop antioxidant supplements during proton therapy, and which ones? Who at the proton centre gives that list?
  2. If she has endonasal surgery, would the skull base repair ever contraindicate hyperbaric oxygen, and for how long?
  3. Who would decide about HBOT afterwards, the proton centre or Southampton?

Sources

Research thread 22 · source file em-research/22-hydrogen-hbot-around-radiotherapy.md · last edited 13 September 2026