The Chamber Applies Pressure. Whether the Tissue Gets Oxygen Is a Separate Question

Abstract cross-section of tissue rendered as a colour-mapped perfusion grid in cyan and indigo, with a hyperbaric chamber porthole receding behind it

A hyperbaric prescription looks deceptively precise. Two atmospheres absolute, 100 percent oxygen, 90 minutes, five days a week, 30 to 40 sessions. Every number is specified except the one that matters most: how much oxygen actually arrived where the damage is. For most of the therapy’s history, that question has been answered indirectly—by waiting several weeks and looking at whether a wound closed.

Two very different groups of researchers are now trying to close that gap, and their findings share an uncomfortable theme. The dose inside a chamber is not uniform, and the tissue that receives it is not uniform either.

Watching perfusion change between two sessions

The more clinically grounded of the two efforts comes from Hospices Civils de Lyon, where an anaesthesia and intensive-care team attached laser speckle contrast imaging—a non-invasive optical technique that maps blood flow in the skin—to patients receiving hyperbaric oxygen therapy for chronic wounds. The team, led by Dr. Sébastien Abrard, reported its findings in International Wound Journal in October 2026.

Six patients with wounds that had resisted conventional management were imaged before and after sessions at three locations: healthy reference skin, the wound margin, and the wound bed itself. Perfusion was expressed in laser speckle perfusion units.

Baseline readings already told a story that a single “chronic wound” label obscures. The wound margin was the most perfused region by a wide margin—a median of 109 units, ranging from 70 to 172—against 58 units in the wound bed and 45 units in healthy skin. The differences between regions were statistically significant.

After treatment, perfusion rose in all three territories: roughly 11 units in reference skin, 10 units in the wound bed, and 30 units at the wound margin. The wound-margin change was the most pronounced and the most consistent. Over the course of a treatment course, pre-session wound-bed perfusion tended to drift downward in most patients—with substantial variation between individuals.

The authors are careful about what a six-patient case series can support. The study’s stated purpose was feasibility, and its conclusion is that laser speckle imaging appears feasible for monitoring perfusion during hyperbaric sessions. Whether those perfusion trajectories or thresholds track healing outcomes, and whether they could inform who gets treated at all, requires prospective studies with standardized measurement times. What the paper does establish is that the therapy’s acute circulatory effect is measurable, real, and regionally uneven.

A chamber built to fit inside a scanner

If the Lyon work measures blood flow, a pair of studies from the Geisel School of Medicine at Dartmouth measures oxygen itself—inside tissue, in real time, under pressure. Both appeared in the same October 2026 issue of Magnetic Resonance in Medicine, from a group led by Dr. Periannan Kuppusamy.

The first built a small-animal hyperbaric chamber compatible with in vivo electron paramagnetic resonance oximetry, using implanted oxygen detectors to track tissue oxygen tension continuously in mice. The system ran stably to 2 atmospheres absolute for up to 90 minutes. The results were strikingly tissue-dependent. In one melanoma model, tumour oxygen tension climbed from about 9 millimetres of mercury at baseline to roughly 26 at 100 percent oxygen and normal pressure, and to about 50 at 2 atmospheres. A different tumour line barely responded at all, staying near 9 to 10 millimetres of mercury. Skeletal muscle showed large, pressure-dependent gains.

The companion study took the opposite approach: instead of building a chamber around a scanner, the team installed an MRI-compatible chamber inside a clinical 3-tesla scanner and ran whole-body oxygen-enhanced MRI on tumour-bearing mice. Normal tissue—brain and heart—behaved predictably, with oxygenation rising in step with both pressure and inspired oxygen fraction. Tumours did not. Even at 2 atmospheres on pure oxygen, the tumours were spatially patchy: many regions responded strongly while others showed little or no change. Raising the pressure increased the proportion of responsive regions but never eliminated the refractory ones.

Why unevenness cuts in two directions

These are preclinical and feasibility findings, and they should not be stretched into claims about patient outcomes. Their value is conceptual, and it is double-edged.

For the therapy’s critics, heterogeneity is ammunition. If substantial parts of a tumour or a wound bed simply do not take up extra oxygen no matter how high the pressure goes, then “more oxygen” is not a synonym for “more healing,” and blanket enthusiasm for the modality—particularly in wellness and longevity marketing—rests on an assumption that the physics does not guarantee. The Dartmouth authors describe their work as a tool for identifying oxygen-refractory zones, which is a useful phrase: some tissue is refractory before treatment even begins.

For the therapy’s defenders, the same data point toward a better argument than the one usually made. The most defensible use of an oxygenation measurement is not to justify treating everyone but to identify who is likely to respond—and, equally, who is not. The Lyon authors say as much when they suggest their perfusion data might eventually “inform HBOT decision-making.” Imaging that fails to predict response is a technology searching for a use; imaging that reliably separates responders from non-responders is a way to stop spending months of sessions on patients who will not benefit.

What regulators actually cleared

That selection problem is not academic, because the coverage rules already assume it. Medicare’s national coverage determination for hyperbaric oxygen therapy covers diabetic wounds of the lower extremities only when three conditions are met: the patient has type I or type II diabetes with a diabetes-related lower-extremity wound, the wound is classified Wagner grade III or higher, and the patient has failed an adequate course of standard wound therapy. Coverage applies as adjunctive therapy only, after no measurable signs of healing for at least 30 days of standard treatment—and only in addition to that treatment.

Note what that rule does. Lacking a reliable way to measure whether oxygen reaches the tissue, the system defines probable non-response by elapsed time: thirty days of watching a wound not heal. It is a reasonable proxy. It is also an admission that the measurement does not exist at the point of care.

The evidence base behind the cleared indication has its own limits. The Cochrane review of hyperbaric oxygen therapy for chronic wounds, last updated in 2015 by Kranke and colleagues, pooled 12 trials and 577 participants. It found that hyperbaric oxygen improved diabetic foot ulcer healing at six weeks—a risk ratio of 2.35—but the benefit was not sustained at one year, and the effect on amputation was not statistically significant. No trials addressed arterial or pressure ulcers. Chronic wounds are a legitimate, FDA-cleared use of the therapy; they are also a use where the outcome data are more modest than the marketing suggests, which is precisely why a response biomarker would matter.

The device side carries its own cautions. The FDA’s August 25, 2025 letter to health care providers reminds clinicians that hyperbaric oxygen therapy devices are Class II devices cleared through the 510(k) pathway under Product Code CBF, and that the agency is aware of fires with these devices that caused serious injuries and deaths. The letter asks providers to follow manufacturer instructions, maintain fire-prevention measures, train and monitor staff, and control prohibited items including electrical and static-generating devices. None of that changes with better imaging; it is the baseline condition for using the therapy at all.

The distance between a signal and a decision

It is worth being precise about how far these papers are from changing practice. The Lyon study enrolled six patients, had no control group, and measured a surrogate—perfusion—rather than healing. The Dartmouth work is in mice, requires implanted detectors or a bespoke scanner-compatible chamber, and says nothing about whether oxygenating a tumour improves a patient’s outcome. A perfusion or oxygenation number is not a clinical endpoint.

What the three papers collectively do is reframe an old question. Hyperbaric oxygen therapy has always been prescribed by pressure and session count because those are the variables a chamber can control. Tissue oxygen tension and microcirculatory perfusion are the variables the body experiences, and they vary by region, by tissue type, by tumour line, and by individual. Instruments that read those variables directly are the difference between treating a diagnosis and treating a patient.

Whether that translates into better selection, shorter futile courses and a more honest evidence base is not yet known. But the era in which the inside of a chamber was a black box is, at least in the laboratory, ending—and what the instruments show is not a uniform wash of oxygen. It is a map.

Sources

← Back to blog