A Conversation with Prof. Aron Popovtzer, MD
Prof. Aron Popovtzer, MD Chair, Sharett Institute of Oncology, and Co-Chair, Hadassah Cancer Research Institute Hadassah University Medical Center, Jerusalem
With Gideon D. Matthews PhD, Commercial VP, Alpha Tau Medical
Introduction
Every year, roughly 890,000 people around the world are told they have a cancer of the mouth, the throat, the tongue or the voice box. Doctors group these cancers together as head and neck cancer, and more than nine out of ten of them are a type called squamous cell carcinoma. In the United States alone there are more than 60,000 new cases a year.
Most of these patients are treated with the intention of curing them, usually with some combination of surgery, radiotherapy and chemotherapy. For about half of them, that works. For the other half, the cancer comes back, and when it does, the situation changes completely: The tissue in the neck has already been operated on and irradiated, so surgeons often cannot operate again, and giving a second full course of radiation risks serious injury. At that point, treatment that acts on the whole body, injected through a vein, becomes the only remaining option.
For many years that meant chemotherapy, which worked for some patients but was hard to tolerate, particularly for the older and frailer people who make up much of this group. Then came immunotherapy, and it changed the picture.
Immunotherapy works differently from chemotherapy. It does not attack the cancer directly. Instead, it removes a brake that the tumor has placed on the immune system, so that the patient’s own immune cells are reactivated and able to recognize the cancer and attack it. The drug that established this approach in head and neck cancer is called pembrolizumab, and it became the standard first treatment for advanced disease in 2019, on the strength of a large international study named KEYNOTE-048. Trials in the pembrolizumab program are all numbered this way, so KEYNOTE-048 is simply the name of that study, which compared the new drug against the older chemotherapy regimen in 882 patients across 37 countries.
That study was a genuine advance, and the drug remains the standard of care today. But it also revealed the limits of what immunotherapy can do on its own. Given as a single drug, pembrolizumab shrinks the tumor in fewer than one in five patients. The obvious way to improve on that is to add chemotherapy back in, and while that does help, it can also multiply the severe side effects roughly four-fold, which puts it out of reach for many of the patients who need it most.
So, the question occupying this field is no longer how to find a better drug. It is how to make the drug we already have work for more people.
The study presented at the American Head and Neck Society’s International Conference on Head and Neck Cancer in Boston this July approached that question from an unusual direction: Rather than add a second systemic drug, the investigators added Alpha DaRT®, an intratumoral radiotherapeutic, placed directly inside a single tumor while the immunotherapy continued to act throughout the body. The hope behind the combination is that treating one tumor in this way may help the immune system to recognize the same cancer elsewhere, so that the immunotherapy works better everywhere.
This was the first study ever done to combine Alpha DaRT with immunotherapy in patients. I sat down with the study’s lead Principal Investigator, Aron Popovtzer, MD, to talk through what was presented, what it may mean, and what still needs to be proven.
FIVE TAKEAWAYS FROM AHNS 2026
TAKEAWAY 1: The central question in this field has shifted from finding new drugs to making the existing drug work for more patients.
This came through strongly across the meeting. Would you agree?
Prof. Popovtzer: I would, and I think it represents a healthy maturing of the field.
For several years, the dominant conversation was about drugs. Which checkpoint inhibitor, which second immune target, which combination of antibodies. A great deal of effort and investment went in that direction, and while some of it produced useful knowledge, it has not fundamentally moved the response rate in this population.
What I heard in Boston was a different framing. We have an active drug with genuine biology behind it. The question is why it succeeds in only a minority of tumors, and what can be done locally, within the tumor itself, to change that.
That is a shift away from pharmacology and toward what I would describe as modifying the environment inside and around the tumor. It also brings radiation oncologists and surgeons back into a conversation that had become almost exclusively a medical oncology conversation. Personally, I welcome that. In head and neck cancer the local problem and the whole-body problem have never really been separable, and I do not believe we should have been treating them as though they were.
TAKEAWAY 2: Radiation combined with immunotherapy has failed before in this disease, and understanding why is essential.
A well-informed reader might point out that combining radiation with immunotherapy has already been tried in head and neck cancer, without success. How do you respond to that?
Prof. Popovtzer: They would be entirely right to raise it, and in my view, anyone presenting a combination such as ours has an obligation to address those trials directly rather than hope the audience has forgotten them.
The JAVELIN Head and Neck 100 trial added an immunotherapy drug called avelumab to chemotherapy and radiation in locally advanced disease. It was a phase 3 study of 697 patients, and it was closed early because an interim analysis found it was not going to work. A second trial, KEYNOTE-412, added pembrolizumab to chemotherapy and radiation and did not meet its main goal either. And in patients with disease that had already spread, McBride and colleagues at Memorial Sloan Kettering ran a randomized trial of immunotherapy with or without focused, high dose radiation to one tumor, designed specifically to test whether treating one site could produce a response elsewhere. Their published conclusion states plainly that they found no improvement in response, and no evidence of that effect.
So, the honest position is that the general idea, that radiation will expose the cancer to the immune system and the immunotherapy will then finish the job, has been tested in this disease and has not delivered.
Then what makes this approach different?
Prof. Popovtzer: My view, and I want to present it as a hypothesis rather than as established fact, is that the concept did not fail. The delivery method failed.
Lymphocytes, the immune cells upon which this whole strategy depends, are among the most radiation-sensitive cells in the human body. They are killed at doses far below those used to treat a tumor. Conventional radiotherapy delivered from outside the body, even when it is planned with great care, inevitably spreads a low dose across a substantial volume of tissue, including the blood circulating through the treated area, the lymph nodes that drain it, and the bone marrow. In head and neck cancer we routinely irradiate the lymph nodes of the neck, and those nodes happen to be exactly where an anti-tumor immune response is first organized.
So, the patient receives a drug whose entire purpose is to unlock the immune system, but at the same time receives a treatment that depletes the very cells which that drug depends on. Seen in that light, the disappointing results become considerably easier to understand.
Alpha DaRT presents a fundamentally different physical situation. The Alpha DaRT sources sit inside the tumor. Alpha particles travel between 40 and 90 micrometers in tissue, less than the width of a human strand of hair. The atoms released from the sources extend the effect out to a few millimeters, and no further. Circulating immune cells are not exposed. The draining lymph nodes are not exposed. In the conventional sense there is no radiation field at all.
The hypothesis, therefore, is that we may be able to destroy the tumor aggressively while leaving the immune system itself intact, so that the immunotherapy has something meaningful to work with. It is a specific, mechanistic hypothesis, and importantly it explains the earlier failures rather than simply setting them aside.
TAKEAWAY 3: In this early study every evaluable patient responded, but it is also important to state the limitations of the data.
Please take us through the results.
Prof. Popovtzer: Eleven patients were enrolled at Hadassah, all with disease that had either returned and could not be removed surgically, or had spread, and all with a CPS of one or more. The average age was 72, ranging from 52 to 96. Each patient received a first dose of pembrolizumab, followed by insertion of Alpha DaRT sources into a single target tumor. The sources were removed approximately two weeks later, and pembrolizumab continued the normal schedule thereafter.
Two patients died before their response could be assessed. One died before the Alpha DaRT was inserted, and one shortly after the procedure from a cardiovascular event judged unrelated to the treatment. That left nine patients evaluable for response.
All nine responded. There were four complete responses and five partial responses, giving an objective response rate of 100 percent, with a complete response rate of 44 percent. Median overall survival was 18.2 months, and median progression-free survival was 5.4 months, with four patients still alive at the time of this analysis.
I would emphasise one methodological point. Response was assessed across the whole body, in all measurable tumors, both those treated with Alpha DaRT and those not treated. We were not measuring the tumor we had irradiated and reporting that as the result.
How does that compare with the benchmark?
Prof. Popovtzer: Pembrolizumab on its own, in the comparable population of KEYNOTE-048, produced an objective response rate of approximately 19 percent, median overall survival of 12.3 months, and median progression free survival of approximately 3.2 months.
I want to be precise about what that comparison is, and what it is not. It compares two separate studies of very different design, size, population and era, conducted years apart, with no randomization between them. Caution should be exercised in drawing conclusions from comparisons of that kind, and I would not want any reader to treat it as a direct head-to-head result, because it is not one.
And the honest limitations of your own data?
Prof. Popovtzer: They are substantial, and I would rather state them myself than have someone else point them out.
Eleven patients is a very small study. It was conducted at a single center, with a single treatment group and no comparison group. Nine evaluable patients is a small denominator, and in a denominator of that size, a figure of 100 percent should generate curiosity rather than conviction.
The study used what is known as a Simon two stage adaptive design, under which it was permitted to stop early for success once more than six patients had responded. That threshold was reached, and recruitment was closed on that basis. This is a legitimate rule, specified in advance rather than decided afterwards, but the result remains an early signal, not proof of benefit.
What I would say is that after many years of treating this disease, I am not aware of another combination study with pembrolizumab that has produced results of this kind in this population. In my opinion that justifies testing the approach properly, in a larger controlled trial. It does not justify telling any patient that anything has been established.
TAKEAWAY 4: The tolerability of the treatment may be the most consequential finding of all.
You have returned to side effects several times. Why do they carry such weight here?
Prof. Popovtzer: Because in this population, tolerability determines whether a treatment is usable at all.
There were no reported serious adverse events related to Alpha DaRT in the study. Across the entire treated group there were exactly two adverse events attributed to Alpha DaRT, and both were grade 1, meaning mild. One of them was pain at the insertion site.
The practical significance of that is considerable. It means patients stayed on their pembrolizumab. We did not have to interrupt, delay or reduce their systemic treatment in order to accommodate the local one. In oncology, combination strategies frequently fail in real world practice not because the underlying biology was wrong, but because patients cannot tolerate both components at the same time. Here, the local component appears to have been added without any measurable cost in toxicity.
The procedure itself is also undemanding. Alpha particles are stopped by the outer layer of the skin, so there is no need for a radiation bunker, lead shielding, or isolation of the patient afterwards. That has real consequences for who can be treated. We treated a 96-year-old woman with a tumor of the jaw who had heart disease, dementia, and a performance status that would have excluded her from virtually any conventional trial. She showed marked regression on PET-CT imaging at approximately nine weeks, and the response was still visible on imaging around eighteen months later.
This is consistent with what we saw in the first published study of Alpha DaRT used on its own, in 2020, which reported a 100 percent objective response rate across 28 evaluable lesions, with 78.6 percent of them showing a complete response, in an elderly, heavily pre-treated population whose tumors had already proven resistant to radiation, and with no severe acute side effects.
TAKEAWAY FIVE: The most scientifically interesting observation concerns the tumors that were not treated.
You mentioned that response was measured in untreated tumors as well. What did you see there?
Prof. Popovtzer: This is the aspect I find scientifically most compelling, and also the one where I would ask for the greatest care in interpretation.
We treat a single tumor. If Alpha DaRT were purely a local treatment, one would expect that tumor to respond, and the remaining tumors to behave exactly as they would have on pembrolizumab alone. In one patient we observed a second tumor, never touched by any source, go into complete response.
That is what is usually described as the abscopal effect.
Prof. Popovtzer: That is the term people use, and I would use it cautiously. A single patient is an observation, not a demonstrated phenomenon. But it is consistent with what the laboratory work would predict.
In mouse models of squamous cell carcinoma, combining Alpha DaRT with anti PD-1 immunotherapy produced the highest level of immune cells infiltrating into the tumor, and greater effect than either treatment on its own. The published conclusion of that work, which appeared in the International Journal of Radiation Oncology, Biology, Physics in 2023, is that Alpha DaRT promotes what researchers call a “hot” tumor environment and strengthens the immune response that PD-1 blockade produces. In other models, treating one tumor produced a statistically significant slowing in the growth of a separate, untreated tumor, and the resulting immunity was shown to be specific to that cancer type and transferable between animals.
The concept is sometimes described as an in-situ vaccine. The tumor is destroyed in a way that releases its own identifying markers, in the presence of an intact immune system, and the immune system may then learn to recognize that same cancer elsewhere in the body. Alpha particles cause a form of DNA damage the cell cannot repair, and they do so regardless of oxygen levels, producing a kind of cell death that is far more visible to the immune system than that caused by conventional radiation.
Is that what happened in our patients? In my opinion it is the most plausible explanation for a whole-body response rate of 100 percent, in a population where the drug alone produces a response in about one in five. But the most plausible explanation is not the same thing as a demonstrated mechanism, and I would not want to blur that distinction.
Closing
The best way to summarize this interview is to let Prof. Popovtzer summarize it himself:
“For twenty years we moved this disease forward a few weeks at a time. Immunotherapy moved it much further, but only for the few, and what we tried adding on top of it either failed or was too toxic to give. Here we added something that apparently sits inside the tumor, does its work, and leaves the immune system alone. Every patient we could evaluate responded. I have not seen that before in this disease, and I would very much like to find out how far it goes.”
Prof. Aron Popovtzer served as lead Principal Investigator for the clinical study described here, which was sponsored by Alpha Tau Medical Ltd. Prof. Popovtzer is also a consultant for Alpha Tau Medical Ltd.
