UHM 2004, Vol. 31, No.1 - HBO 2 for delayed radiation injuries
Hyperbaric oxygen for delayed radiation injuries.
Toledo Radiation Oncology and the Medical College of Ohio, Toledo, Ohio
Based on distinctions in their pathophysiologies, radiation complications should be
divided into acute and delayed. Acute complications are generally cellular and mucosal and are
caused by direct damage to cellular DNA with resultant cell death. These are usually self-limited
and generally not dose limiting. Late radiation damage is primarily vascular and stromal.1 Late
complications are the most dreaded and determine the limits of tolerable radiation doses. The
tolerance to radiation varies considerably from organ system to organ system. For years radiation
oncologists and pathologists have felt that the mechanisms of early and late damage are
completely unrelated. Although there are distinctions and identifiable differences in the nature
and site of injury, we are beginning to appreciate that late radiation damage actually begins at the
initiation of treatment. An increase in various biochemical substances including fibrogenetic
cytokines are identifiable from the onset of treatment.2 These have been associated with late
radiation damage and their serial assays may allow for prophylactic interventions. An elevation
of those cytokines, which lead to damage or conversely a depression of protective cytokines may
identify a group of patients at high risk for late radiation damage. This early identification of
those at risk prior to manifestations of frank radiation damage may permit prophylactic
intervention. An exciting potential area for further research will include the application of
hyperbaric oxygen during this latent period after radiation but before expression of the clinically
expressed damage. Certainly, if such a group of reliable predictive assays can be identified, other
therapeutic strategies to prevent radiation damage will also undergo investigation.
To date hyperbaric oxygen (HBO2) has been a very successful therapeutic modality for
delayed radiation injury. It has also been shown to be useful in a prophylactic intervention in
several circumstances. A recent systematic review by Feldmeier and Hampson, summarizes 74
publications that report the results of hyperbaric oxygen for a wide range of soft tissue and bony
necrosis. This review was conducted in an evidence-based fashion. All but seven papers are
positive in terms of their therapeutic effect. Most of the negative reports are in neurologic injury
where, especially in the CNS, injuries have been refractory to virtually all interventions. The
application of HBO2 has had its earliest and best-studied application in mandibular
osteonecrosis. HBO2 is an effective treatment modality for radiation necrosis because, at least in
part, the pathophysiology of this process is vascular and stromal secondary to obliterative
endarteritis. HBO2 has been shown to induce neovascularization in this hypoxic milieu and to
reduce fibrosis in irradiated tissues.
Molecular oxygen available simultaneously with ionizing radiation is the most potent
radiation sensitizer known. The first interactions of radiation and HBO2 were based on this
observation. Mostly during the 1960's and early 1970's, HBO2 was used as a radiosensitizer for
Copyright © 2004 Undersea and Hyperbaric Medical Society, Inc. 133
UHM 2004, Vol. 31, No. 1 - HBO 2 for delayed radiation injuries
external beam radiation, and numerous clinical trials were conducted whereby patients with
malignancies were irradiated while at pressure in monoplace HBO2 chambers. Although many of
these trials showed improved local control of cancers, especially in head and neck cancers and
cervical cancer, this improvement did not generally translate into improved cure rates, and the
practice of HBO2 radio-sensitization for external beam irradiation has been largely abandoned.3
The Etiology of Radiation Necrosis or Late Radiation Injury and the Rationale for
Late radiation changes are characterized by fibrosis and endarteritis. If nutrient blood
vessels are significantly narrowed and if inadequate O2 is available to meet metabolic demands
frank aseptic necrosis occurs.1
The development of radiation-induced endarteritis requires time to become established.
Clinically we see a latent period of months to years before these changes result in necrosis. The
process can be set off by trauma to the tissues such as dental extraction or surgical incision.
HBO2 has been shown to induce neovascularization and reduce fibrosis in radiated
hypoxic tissues. Marx4 has shown an increase in vascular density and cellularity in histologic
specimens of tissues before hyperbaric oxygen compared to specimens taken from the same site
after hyperbaric treatments just prior to mandibular reconstruction. Marx5 has also shown an
improvement in sequential transcutaneous oxygen measurements in patients undergoing
hyperbaric oxygen for mandibular necrosis. Feldmeier6-8 and his co-workers have shown
decreased stromal fibrosis by morphometric collagen assay and functional compliance
measurements in a murine model of radiation enteritis.
Site Specific Applications of Hyperbaric Oxygen to Delayed Radiation Injuries
Prior to the 1970’s, HBO2 had been used with mixed success for mandibular radiation
Beginning in the late 1970's Marx and Johnson and their co-workers while at Wilford
Hall USAF Medical Center in collaboration with the USAF Hyperbaric Medicine Center at
Brooks Air Force Base established a formal protocol for treatment and evaluation of mandibular
necrosis.9 The development of a “Staging” system led to a logical application of HBO2 integrated
with surgery. The specific recommendations for therapy followed quite logically from the
assignment to a particular stage.
a. Stage I - Patients receive 30 HBO2 treatments at 2.4 ATA for 90 minutes In these patients
no more than minor debridement in the dental chair was felt to be necessary.
(1) If response was good and exposed bone was covering, the patient continued for 10
b. Stage II - These were non-responders to the first 30 treatments in Stage I but in whom
debridement rather than a discontinuity resection was felt to be appropriate. Debridement
was not accomplished until 30 HBO2 treatments are completed.
(1) Post debridement ten additional treatments were given.
c. Stage III- These were non responders to Stage II treatment or those who present with
cutaneous fistulae, pathologic fracture or resorption of the inferior cortical border of the
UHM 2004, Vol. 31, No. 1 - HBO 2 for delayed radiation injuries.
(1) 30 HBO2 treatments given followed by resection with 10 post-op HBO2 sessions.
(2) 10 weeks after resection reconstruction was accomplished followed by 10 post
reconstructive HBO2 sessions. Reconstruction involved the use of freeze-dried cadaveric
bone used as a carrier tray for the patient’s own corticocancellous bone harvested from the
(3) External jaw fixation was maintained for 8 weeks following the reconstruction.
With this protocol, success in mandibular reconstruction has been unexcelled even when
compared to more surgically complex procedures such as microvascular anastomosis with free
flaps. However, there is nothing to preclude the use of free flaps or myocutaneous flaps along
with hyperbaric oxygen and it is always a good principle to combine optimal surgery with HBO2
as an adjunct in improving the quality of tissues in the recipient bed.
Including the publications by Marx, a total of fourteen publications are available reviewing the
experience of applying HBO2 to mandibular necrosis10-23. One very small randomized controlled trial
by Tobey et al13 is positive. Only twelve patients were studied in this publication. These patients were
randomized to 100% oxygen at 1.2 vs 2.0 ATA. The authors state that those patients treated at 2.0
ATA “experienced significant improvement” compared to the group receiving oxygen at 1.2 ATA.
No details are given regarding randomization or outcome determination. In fact we cannot tell how
many patients were assigned to each group. The study is randomized and doubly blinded in that
neither the patient nor the clinician assessing the patient knew which therapy the patient was receiving
Other than the trial by Tobey, all of the rest of the publications present case series. Of the
fourteen publications, only the report by Maier et al21 fails to show an advantage for hyperbaric
oxygen in the treatment of existing ORN. In this paper, none of the hyperbaric oxygen is given prior
to surgery. Hyperbaric oxygen is part of the overall management only after an attempt is made at
surgical correction. Marx has previously established the importance of pre-operative HBO2 prior to
surgical wounding in irradiated tissues. This principle has been widely accepted by those applying
HBO2 to the treatment of ORN.
If all of the cases are combined (excluding those reported by Tobey and noting that Marx’s
second report includes the fifty-eight patients reported earlier), we have a total of 371 reported cases
of mandibular ORN. Improvement is reported in 310 cases or 83.6%. Resolution would certainly be a
superior endpoint. However, especially in the earlier reports, hyperbaric oxygen was not combined
with aggressive extirpation of necrotic bone or with surgical reconstruction of bony discontinuity.
Certainly, Marx has reported the best results of any single author. Marx has identified the need for
optimizing surgery and sequencing HBO2 and surgery to include and emphasize the pre-surgical
application of HBO2. Marx reports 100% success, but his successful treatment includes
mandibulectomy and reconstruction in 73% of his patients. Dr. Marx also sets high standards for
successful intervention in those patients requiring mandibulectomy and reconstruction. Marx requires
not only the successful re-establishment of bony continuity but also requires functional success in that
these patients must be able to support a denture for cosmesis and mastication.
HBO2 for Osteoradionecrosis Prophylaxis
A randomized trial by Marx et al.24 compared penicillin to HBO2 prior to dental
extractions as prophylactic strategies to prevent mandibular radiation necrosis in heavily
irradiated mandibles. With thirty-seven patients in each group, ORN occurred in 11/37 (29.9%)
of penicillin group and only 2/37 (5.4%) in HBO2 group.
UHM 2004, Vol. 31, No. 1 - HBO 2 for delayed radiation injuries.
This protocol randomized only patients with 6800 cGy or higher dose. The penicillin
group received 1 million units of penicillin just prior to surgery followed by 500 mg penicillin
p.o. Q.I.D. for 10 days. The HBO2 group underwent 20 HBO2 treatments prior to extractions and
10 HBO2 treatments after extractions.
Two additional clinical series present their results in the prophylactic treatment of fifty-three
additional patients. If we combine the patients from all three reports, we find an incidence of
osteoradionecrosis (ORN) in 4.5% (4 of 90) of the HBO2 prophylaxis group (2 of 37 Marx; 1 of 29
Vudiniabola25; and 1 of 24 David23). In Marx’s control group, the incidence of osteoradionecrosis was
29.9% (11 of 37).
Fortunately, radiation-induced laryngeal necrosis is an uncommon complication. In a
well designed radiation treatment course, its incidence should be less than 1%. Higher doses per
treatment fraction, higher total doses and neutron irradiation increase the risk of laryngeal
The effects of HBO2 on chondroradiation necrosis of the larynx have been reported by three
authors from three separate institutions.26-28 The majority of these patients had severe (Chandler’s
Grade III or IV necrosis). Without hyperbaric oxygen treatment, the usual recommendation is
laryngectomy for those patients in whom laryngeal edema is persistent. The rationale for this
recommendation is that vast majority of patients with persistent edema have tumor, and even if they
do not, salvage is not possible because no effective treatment had been known for laryngeal
chondronecrosis. Symptoms of both laryngeal necrosis and tumor recurrence may include airway
compromise, edema, fetid breath and production of necrotic debris. Biopsy in such patients may be
subject to sampling error or may further exacerbate the necrotic process. Biopsy should be avoided or
minimized if possible. Obviously, biopsy may be ultimately required to demonstrate or rule out tumor
recurrence. If the results from these three trials are combined, only six of thirty-five patients treated
with hyperbaric oxygen required laryngectomy. The rest maintained their larynx with most patients
having good voice quality after HBO2.
Other Soft Tissue Necrosis Injuries of the Head and Neck
Dr. Marx’s chapter in the textbook, Hyperbaric Medicine Practice29 reports his experience in
a prospective controlled but not randomized study applying hyperbaric oxygen to soft tissue
radionecrosis of the head and neck. Those patients who refused hyperbaric oxygen or for whom
treatment was not practical due to having homes distant from a hyperbaric chamber or other financial
reasons were assigned to the control group. These cohorts of patients were treated concurrently and all
other aspects of their treatment were identical. In his report of 160 patients receiving hyperbaric
oxygen in support of surgical resection or flap reconstruction in heavily irradiated patients comparing
wound infection, wound dehiscence and delayed wound healing, Marx reports the incidence of
complications in the HBO2 group versus the control group in the following fashion: 1. Wound
infection: 6% versus 24%; 2. Wound dehiscence: 11% versus 48%; and 3. Delayed wound healing:
11% versus 55%. Applying the Chi square test to these results we obtain P values of 0.004, less than
0.0001 and less than 0.0001 respectively for each of these outcome measures. These patients received
twenty pre-operative HBO2 treatments followed by ten post-operative treatments at 2.4 ATA.
In addition to the large controlled trial reported by Marx, there are three additional
publications reporting case series in which hyperbaric oxygen has been applied to soft tissue
radiation injuries of the head and neck (other than larynx).30-32 These consistently report a
UHM 2004, Vol. 31, No. 1 - HBO 2 for delayed radiation injuries.
positive outcome in patients treated with HBO2 for soft tissue radionecrosis of the head and neck.
The case series by Davis et al31 reports success in fifteen of sixteen patients treated for soft tissue
radionecrosis of the head and neck. Many of these patients had large chronic soft tissue wounds
as a result of their radiation injury. In 1997 Neovius30 and colleagues reported a series of fifteen
patients treated with hyperbaric oxygen for wound complications within irradiated tissues. They
compared this group to a historical control group from the same institution. Twelve of the fifteen
patients in the hyperbaric group healed completely with improvement in two and no benefit in
one. In the control group only seven of fifteen patients healed. Two patients in the control group
developed life-threatening hemorrhage and one of these did exsanguinate. Any practitioner
experienced in caring for head and neck cancer patients has experienced at least one patient in
his or her career that exsanguinated as the result of a soft tissue necrosis of the neck which
progressed to erode into the carotid artery or other major vessel.
Chest Wall Necrosis
Radiation is a frequent modality applied to lung, breast, or esophageal cancers. Chest
wall radiation necrosis occurs most frequently in breast cancer after mastectomy due to the need
to treat skin and subcutaneous tissues to a relatively high dose since this tumor often recurs in the
skin of the chest wall. HBO2 in this setting has not been extensively reported.
Hart 11 has reported the use of hyperbaric oxygen as an adjunct to skin grafting in six patients
with radiation injury of the chest wall with all patients experiencing graft take. Feldmeier and his
colleagues33 have reported a total of twenty-three cases of radiation necrosis of the chest wall: eight
with soft tissue only necrosis and fifteen with a combination of bone and soft tissue necrosis.
Resolution in those with soft tissue involvement only was 75% while those with a component of bone
necrosis had resolution in 53%, and all of these patients required resection of necrotic bone. If bone is
involved aggressive debridement to include resection of non-viable bone is required for good results
In a case report Carl and Hartmann34 in 1998 published their results in treating a patient with
long-standing symptomatic breast edema following lumpectomy and irradiation. The patient received
fifteen, 90 minute HBO2 treatments at 2.4 ATA. The patient had complete resolution of pain and
Carl and his associates35 in 2001 reported the outcome of 44 patients who suffered
complications following lumpectomy and irradiation for early breast cancers. These patients were
found to have pain, edema, fibrosis and telengectasias. Each patient experienced these complications
in various combinations. The severity of symptoms was assessed a score for each patient based on a
modified LENT-SOMA score. Only patients with at least grade 3 pain (persistent and intense) or a
summed LENT-SOMA score of 8 were studied. Each patient was assessed a score from 1 to 4 in the
severity of symptoms in the categories of pain, edema, fibrosis/ fat necrosis and
telangectasia/erythema. Thirty-two patients agreed to undergo hyperbaric oxygen treatment while
twelve women refused HBO2 and constituted the control group. Hyperbaric oxygen treatments
resulted in a statistically significant reduction in the post-treatment SOMA-LENT scores in women
receiving hyperbaric oxygen compared to those who did not. Fibrosis and telangectasia were not
reduced. Women in the control group continued to demonstrate symptoms at the completion of the
trial with no improvement in pain or edema while seven women in the hyperbaric group had complete
resolution of their symptoms at the end of the trial.