Radiation Modality Influences Risk of Osteoradionecrosis in Head and Neck Cancer

Key Clinical Summary:

• Design/Population: A retrospective cohort study evaluated 1,564 patients with oropharyngeal cancers treated with radiation therapy between 2013 and 2023, comparing the risk of osteoradionecrosis after proton therapy versus intensity-modulated radiation therapy using time-to-event Cox modeling.
• Key Outcomes: Osteoradionecrosis incidence at 3 years was 6.5% with proton therapy vs 4% with intensity-modulated radiation therapy. In multivariable analysis, proton therapy, concurrent chemotherapy, and smoking history each remained independently associated with approximately twofold increased osteoradionecrosis risk. Severe osteoradionecrosis requiring surgery (grade ≥3) was rare and similar between modalities.
• Clinical Relevance: Although proton therapy offers dosimetric advantages, it was associated with higher rates of lower-grade osteoradionecrosis compared with intensity-modulated radiation therapy. These findings highlight the need for optimized patient selection, radiation planning, and preventive strategies to minimize mandibular toxicity while maintaining oncologic outcomes in head and neck cancer.

Edward Christopher Dee, MD, Memorial Sloan Kettering Cancer Center, New York, New York, discusses results from a retrospective analysis assessing osteoradionecrosis risk among patients with oropharyngeal cancer following either proton or intensity-modulated radiation therapy. 

Results demonstrated that proton therapy was associated with a higher rate of developing osteoradionecrosis within 3 years, with risk also increased by concurrent chemotherapy and smoking history, though severe grade 3 osteoradionecrosis requiring surgery remained rare and similar between modalities. 

Transcript:

Hi there, good afternoon everybody. Thank you for the opportunity to talk about our work. I'm looking at osteoradionecrosis amongst patients who receive radiation for cancers of the head and neck. 

My name is Dr Edward Christopher Dee, I'm a chief resident at Memorial Sloan Kettering Cancer Center, and I am going to be staying on as faculty treating patients with cancers of the head and neck as well as of the central nervous system. I'm excited to be able to present some of our group's recent work today. I want to give a shout out to one of the lead authors of the paper, Dr Fan Yang who's now treating patients at Mayo Clinic in Arizona.

This study really comes out of a broad question of that of osteoradionecrosis. What is osteoradionecrosis? Osteoradionecrosis is a syndrome or a symptom of exposed bone that can occur downstream of radiation for patients with cancers of the head and neck. As you would imagine, some of the late toxicities of radiation can take time, months, even years, to manifest amongst patients who receive this kind of radiation, even after the cancer is gone, after the cancer is cured. One of the side effects is osteoradionecrosis, which is a complicated and broad spectrum of toxicities. 

Osteoradionecrosis was first diagnosed and first described in 1922, and over the past 100 years, a lot of work has gone into understanding its etiologies and understanding what could be done to mitigate this risk. In brief, our understanding of osteoradionecrosis today is related to the anatomy of the mandible– that's the bone here that is just essentially the jawbone. After radiation for head and neck cancers, oftentimes cancers of the tongue or of the floor of mouth or of the oropharynx like the tonsil, the blood supply to the mandible can get damaged or can get affected by that radiation. What happens subsequent to that is that patients are at increased risk of osteoradionecrosis from decreased blood flow to some of the tissues and the cells that make up the bone. 

We know that factors related to osteoradionecrosis include patients who have a history of smoking, the concurrent use of chemotherapy, which is also part of the package of treatment for many of these cancers, as well as dental procedures, for example, dental extractions that happen after patients receive radiation. That's why our standard of care is to have patients see their dentist before they undergo head and neck radiation to decrease that risk so any procedure that needs to get done can be done afterwards.

One of the big questions that we try to ask now that we know that there are other advances in radiation as we've gone from 2D to 3D, to intensity modulated radiation (IMRT) and now proton therapy, [is] do we know whether the rates of osteoradionecrosis (ORN) are any different when we look at patients who are receiving proton radiation for head and neck cancers? 

Proton radiation is one of the kind of newer modalities for treatment and is characterized by the use of a mass bearing particle that is the proton that can deliver radiation to targets but also with a far less spill overdose posterior to the target. The idea is that maybe there’s a possibility that the jawbone could be getting less radiation or could be less in the treatment field. 

That being said, it's not known, it's not known whether the rates of osteoradionecrosis are any different when you look at patients who receive proton versus photon. There's been a lot of work over the years that have examined this question, but a lot of these papers published in some of the top journals in the fields from our colleagues at MD Anderson, our colleagues in Canada, have enrolled relatively heterogeneous cohorts of patients. On the backend of radiation oncology, we know that patients who have whether it's a nasopharynx cancer or oral cavity cancer or tonsil cancer, get pretty different fields of radiation. Radiation is not just one kind of overall bucket, but a whole 3-dimensional and dose-related process that is quite nuanced on the backend. So, we said, we need to study patients in a more homogeneous cohort. We as one of the large academic centers, looked back over the past 10 years from 2013 to 2023 and collected the 1,564 patients who underwent radiation for cancers of the oropharynx. Those are people with cancers of the tonsil, the base of tongue, and several other relevant sites. We felt that bucketing this common set of conditions together allows for a bit more of a homogenous cohort to better understand the question of treatment modality and precipitating osteoradionecrosis risk. What we did was we conducted a retrospective review, another shout out to another team member who played an important role is Yingzhi Wu, who's a nurse practitioner in our service who also helped a lot with the data collection and of course, Dr Singh, who's going to talk a bit later today, so all of us got together and collected all the data and we looked, especially for this rare toxicity.

From a statistical standpoint, we didn't just divide by the denominator of patients, we actually used a time varying formulation, which is a Cox model that looks at the hazard of developing this rare toxicity, osteoradionecrosis. Given that proton especially is a relatively newer technology, one could argue is the follow-up time any different, is there a difference in the follow-up time for patients who undergo proton therapy versus IMRT such that there's more time in the past to develop that rare toxicity for patients, so what we did again was we employed that Cox model, which actually accounts for differences in follow-up time.

We found in a univariable analysis that quite surprisingly, the rates of osteoradionecrosis of any grade were actually higher for patients who had treatment with protons as opposed to IMRT. At 3 years, we found that the rate amongst proton patients was about 6.5% whereas the rate amongst IMRT or photon patients was about 4%. We found that the univariable cox hazard ratio for that was 2.62, and that was statistically significant. Of note, we also found that receiving concurrent chemotherapy, which in many cases is associated with improved overall survival and disease control, is also associated with an increased risk of osteoradionecrosis. Lastly, smoking history, patients who have a history of smoking, is also associated with an increased risk of osteoradionecrosis.

We said, okay, how do these things interact so we built a multi-variable model that includes proton therapy, concurrent chemotherapy, and smoking as co-variables in the same model. We found that all 3 were still statistically significant with a pretty similar hazard ratio of 2 and change for each of these factors. We can say with some confidence that the risk of ORN is essentially twice in the first 3 years for patients receiving fully proton-based therapy compared with patients who are receiving fully IMRT-based therapy. We also know that that risk is the increase when patients are treated with concurrent chemotherapy and when patients have a history of smoking.

Of note, we also looked at patients who had grade 3 osteoradionecrosis, that's the very rare extreme subset of patients who need some sort of surgery for osteoradionecrosis. We found that fortunately there is no difference between the 2 modalities and that the risk of grade ≥3 osteoradionecrosis in either cohort is only 1%, slightly less actually.

We know that this is safe, this is an important toxicity, and thankfully, the higher grade toxicity is of lower frequency and no different in the 2 treatment modalities, but this has stimulated a lot of work in our group to understand better how do we mitigate this risk, how do we lower the chance that a patient ends up with osteoradionecrosis [and] I think I want to put the caveat there that another important thing is the vast majority of these patients were cured of their cancer, but in treating the cancer, we also want to treat the patient. To do so, we want to minimize the risk of long-term toxicities, we want these people to go out into the world and live functional, happy, normal lives in the decades to come after this cancer diagnosis and this cancer treatment and part of that includes optimizing the late toxicities. A lot of our work in the coming years, we'll go into figuring out how do we decrease this risk while optimizing the benefits of proton. This is not to say that proton is all bad– proton has immense advantages as well. We're trying to see what combination of proton and photon might be better, which patients might be better served by 1 modality or the other, and how can we optimize our dosimetry, our modeling of proton therapy to mitigate those to the mandible and therefore decrease the risk of osteoradionecrosis.

I want to thank our entire team for all their work that went into this immense effort. Dr Fan Yang, who served as our co-first author of this paperwork really led so much of the intellectual effort here. Dr Singh is an oral oncology and dental expert whose insight is invaluable to this study and is probably the world expert in osteoradionecrosis, along with her mentor and my mentor, Dr Cherry Estilo, as well as Dr Nancy Lee, who's the head of head and neck radiation oncology. Dr Estillo and Dr Lee served as principal investigators for this important study. 

Thank you.

Source:

Yang F, Dee EC, Singh A, et al. Osteoradionecrosis after intensity-modulated radiation therapy or proton therapy in oropharyngeal carcinoma. JAMA Otolaryngol Head Neck Surg. Published online November 26, 2025. doi:10.1001/jamaoto.2025.4179