Alpha Seeds First, Then the Checkpoint Drug Claiming the Four Week Window Between Two Cancer Treatments

This application claims a method of treating metastatic cancer by putting alpha emitting radiation sources directly inside a tumour and then giving the patient a drug that releases the brakes on the immune system, with the two treatments separated by no more than four weeks. It was filed by Alpha Tau Medical Ltd., an Israeli oncology company listed on Nasdaq, and names ten inventors including Professors Yona Keisari and Itzhak Kelson of Tel Aviv University. The application is a divisional of Australian application 2022299492. The claimed subject matter is the timing relationship between the two treatments, not either treatment on its own.

The Problem

Immune checkpoint inhibitors are one of the genuine successes of modern oncology, and also one of its bigger disappointments. The specification puts the problem in a single blunt sentence in its background section: response rates to checkpoint inhibitor treatment are relatively low, about 20 per cent. Patients who receive the treatment mostly do not respond, yet still develop serious adverse effects. The document describes an extensive effort to find treatments that raise those response rates, so far without pronounced success.

The reason for the low response rate is mechanical rather than mysterious. A checkpoint inhibitor does not attack cancer. It removes a molecular brake that stops killer T cells from finishing a job they have already started. As the specification explains, a T cell can only recognise a tumour cell if it was first activated by an antigen presenting cell that showed it a specific tumour antigen. If no such T cell clone exists in the patient, releasing the brake achieves nothing, because there is nothing to release.

The obvious fix is to kill some tumour cells first, spilling antigens and danger signals that antigen presenting cells can pick up and carry to T cells. Radiation is the usual way to do that, and the combination has been tried. The specification cites a randomised phase II trial by Sean McBride and colleagues in the Journal of Clinical Oncology which added stereotactic body radiotherapy to nivolumab in metastatic head and neck squamous cell carcinoma and found no improvement.

The specification argues that conventional radiation carries its own penalty. It can damage the immune system it is supposed to be recruiting, by harming immune organs and by expanding immune suppressive cell populations such as myeloid derived suppressor cells, which counteracts the checkpoint drug. The invention is a bet that a different kind of radiation, delivered inside the tumour and reaching only a very short distance, avoids that penalty.

What This Invention Does

Claim 1 is short enough to state in full effect. It is a method of treating a patient with metastatic cancer, comprising treating a tumour in the patient with intra-tumoral alpha-emitter radiotherapy, and administering to the patient a substance which regulates immune checkpoints within four weeks of beginning the treating of the tumour with intra-tumoral alpha-emitter radiotherapy.

That is the whole of it. Read carefully, claim 1 recites no radionuclide, no activity, no dose, no seed geometry and no particular drug. It does not even fix the order: the phrase is “within four weeks of beginning”, and the body text confirms the intended reading, stating that the regulators are administered within four weeks or even within two weeks, before or after, the beginning of the alpha-emitter radiation treatment. The only hard requirements are that the cancer is metastatic, that the radiation is an alpha emitter placed inside the tumour, and that the checkpoint drug falls inside a four week window.

Everything specific sits in the dependent claims. Claim 2 narrows the substance to an immune checkpoint inhibitor and claim 3 to a bi-specific antibody. Claims 4 to 8, 10 and 15 pick out particular targets one at a time: LAG3, TIM-3, PD-1, PDL-1, CTLA4, TIGIT and VISTA. Claims 9 and 11 add an anti-PD-1 or anti-PDL1 agent on top of those. Claims 12 to 14 cover small molecule inhibitors and costimulatory molecules including OX40 and CD40, claim 17 covers an iMID, claim 18 names sorafenib or sunitinib specifically, and claim 19 covers a SLAM stimulator or a Cox-2 inhibitor. Claims 26 to 31 add optional extras: immunostimulators, TLR agonists, RIG-1 or STING agonists, cytokines, ATR or TGFb inhibitors and histone deacetylase inhibitors.

The timing claims are where the invention’s actual argument lives. Claim 21 requires the first session of the drug at least 72 hours after the radiation begins. Claim 22 requires it less than 144 hours after. Together those two define a window of roughly three to six days. The body text explains why. The gap is a deliberate buffer, long enough for the alpha radiation to upregulate MHC1 on the tumour cell surface, trigger cytokine and damage signal release and activate antigen presenting cells, but short enough that T cell infiltration takes effect before the tumour recovers and starts reproducing at scale. The specification lists a ladder of alternative buffers, from at least 6 hours up to at least 120 hours, and upper bounds from under one month down to under 48 hours.

Radium-224 appears nowhere in the claims. It appears throughout the description, which is where the applicant’s Alpha DaRT technology is set out: seeds carrying radium-224 or radium-223 atoms bound to the surface, such that not more than 20 per cent of the radium atoms leave the support in 24 hours without decay, but at least 5 per cent of the daughter radionuclides leave upon decay. The daughters diffuse outward through the tumour, so the radioactivity spreads gradually rather than all at once, with activity always highest nearest the seed. Seed diameters of less than 0.7, 0.5, 0.4 or even 0.3 millimetres are described, on a support at least 1 millimetre long.

Why alpha rather than the beta and gamma radiation used in conventional brachytherapy? An alpha particle is a helium nucleus, heavy and doubly charged. It deposits all its energy over a track of well under a tenth of a millimetre in tissue, a few cell diameters, and the DNA damage it leaves behind is complex and hard to repair. Beta particles travel millimetres and gamma photons travel centimetres. The specification leans on exactly this difference: because the effect is short ranged, lymph nodes, bone marrow and tertiary lymphoid structures near the tumour stay intact and can support the immune response the treatment is trying to provoke.

The experimental results are all in mice, and they are modest and carefully described. In the squamous cell carcinoma model, Balb/c mice received a 6.5 mm seed loaded with 75 kBq of radium-224 on day 0, with mouse anti-PD-1 given intraperitoneally at 10 mg/kg on days 2, 6, 9 and 13. Anti-PD-1 alone had no effect on tumour development compared with control. The seed alone significantly reduced tumour development. The combination inhibited tumour development compared with both control and seed-only groups. In a pancreatic ductal adenocarcinoma model in C57BL/6 mice, using an 80 kBq seed and anti-PD-1 on days 1, 4, 7, 10 and 14, the specification reports that on day 14 only the combination group showed a significant reduction in tumour volume compared with control, and adds honestly that the experiment was terminated on day 20, which it calls a relatively early timepoint.

The mechanistic experiments are the more interesting half. Flow cytometry at day 7 found more activated dendritic cells in seeded tumours than in tumours with an inert source. Immunohistochemistry at day 16 found higher CD3, CD8 and granzyme B densities in the combination group than with anti-PD-1 alone. Spleen analysis at day 16 found fewer polymorphonuclear myeloid derived suppressor cells in the combination group than with anti-PD-1 alone. The body text also records an odd two stage pattern: at seven days there was mass destruction of cells in the tumour including T cells, and cells performing negative regulation moved in, but by days 14 to 16 there was an unexpected increase in T cells and in their function, which did not occur with checkpoint regulators alone.

Key Features

• A four week window as the claimed invention. Claim 1 requires only that the checkpoint drug be given within four weeks of the start of intra-tumoral alpha radiotherapy. No radionuclide, activity, dose or drug is recited in the independent claim.
• A three to six day preferred gap. Claims 21 and 22 narrow the window to at least 72 hours and less than 144 hours after the radiation begins, matching the body text’s finding that dendritic cells activate in the tumour within about a week of alpha irradiation.
• Order is not fixed. The specification states the regulators may be given before or after the radiation starts, and describes a class of embodiments where the drug is given up to a week beforehand so that it operates throughout the radiotherapy, for particularly aggressive tumours.
• Seeds that keep the parent and release the daughters. The described sources hold radium-224 so that not more than 20 per cent leaves without decay in 24 hours, while at least 5 per cent of daughter radionuclides leave on decay and diffuse outward, producing a gradual rather than immediate destruction of the tumour.
• A very long menu of checkpoint targets. The dependent claims and description name PD-1, PDL-1, CTLA4, LAG3, TIM-3, TIGIT and VISTA inhibitors, bi-specific antibodies, CD47 and SIRP alpha agents, CSF-1R agents, costimulatory molecules such as OX40 and CD40, and named drugs including nivolumab, pembrolizumab, ipilimumab, atezolizumab and relatlimab.
• Waiting before surgery. The specification notes that while a cancerous tumour is normally removed as soon as possible, the applicant found it better to wait at least a week, or even 14 days, after the combined treatment before surgery, so the treatment has time to work.

Who Is Behind It

Alpha Tau Medical Ltd. is a Jerusalem based oncology company founded in 2016 and listed on Nasdaq under DRTS. Its own description of its origins matches the inventor list on this application: the Alpha DaRT technology was initially developed by Professor Itzhak Kelson and Professor Yona Keisari of Tel Aviv University, a physicist and an immunologist respectively, and both are named here as inventors. The remaining eight named inventors are company researchers and clinicians.

The company’s clinical pipeline runs across skin, oral cavity, pancreatic, prostate, lung, liver and breast tumours. It has Breakthrough Device Designation from the United States Food and Drug Administration for recurrent cutaneous squamous cell carcinoma, recurrent glioblastoma and recurrent squamous cell carcinoma of the oral cavity, and submitted the first module of a pre-market approval application to the FDA in January 2026. None of that appears in this specification, which is a 2022 document, and none of it is claimed here.

On priority, the Australian specification is silent. There is no priority or convention claim anywhere in the body text. The only relationship the document itself establishes is a cross reference paragraph stating that the application is related to International Application PCT/IB2022/055680, published as WO 2022/269446 and filed on 19 June 2022, plus the title page entry recording this application as a divisional of Australian application 2022299492. From the published family record, rather than from this specification, WO 2022/269446 claims priority from United States provisional application 63/212,671 filed 20 June 2021, which makes the priority country the United States.

Why It Matters

The case for this filing is not that alpha radiation is new or that checkpoint inhibitors are new. It is that the interval between them might be a patentable variable, and the specification gives a coherent biological reason why it should be.

Every combination of radiation and immunotherapy has to solve the same conflict. Radiation kills tumour cells, which is what supplies the antigens the immune system needs. Radiation also kills lymphocytes, which are the cells that have to do the work. A conventional external beam course delivers both effects to a large volume over weeks. The argument here is that an intra-tumoral alpha source inverts the ratio. The killing is intense but confined to a few cell diameters around each diffusing atom, so nearby lymph nodes and bone marrow are spared, and it is spread over days as the radium decays rather than delivered in a single fraction. That, the specification argues, lets antigen presenting cells and T cells coexist with dying tumour cells long enough for a real immune response to form, which is then amplified by the checkpoint drug. The reduction in myeloid derived suppressor cells reported in the mouse spleen data is the part of that story with the most direct evidence behind it.

Readers should be clear about what the document does and does not establish. Every result in this specification is preclinical mouse data. There are no human results, no survival figures, no response rates and no cure counts. Claim 1 is directed at a patient with metastatic cancer, but none of the described experiments measures metastases or distant lesions: they measure the size of the treated tumour, plus immune cell counts in that tumour and in the spleen. The systemic effect that would matter most to a patient with metastatic disease, the shrinkage of untreated lesions sometimes called the abscopal effect, is argued for in the description rather than demonstrated in the examples. Separately from this specification, the applicant reported in July 2026 that a study of Alpha DaRT combined with pembrolizumab in locally advanced or metastatic head and neck cancer met its pre-specified threshold for success, which is company reported data from a small study and not part of what is claimed here.

There is also a drafting oddity worth noting. The abstract printed on the front of this specification describes something different from the claims. It describes a substance which regulates immune checkpoints for use as a medicament, with seeds carrying radium-224 implanted less than two weeks from administering the substance. That is a medicament-for-use claim with a two week window and a named radionuclide, which is the shape of the parent application. Claim 1 of this divisional is a method claim with a four week window and no radionuclide at all. Where the abstract and the claims disagree, the claims are what has been applied for.

Related Concepts

• Targeted alpha-particle therapy – the broader field of using alpha emitting isotopes against cancer, of which intra-tumoral seeding is one approach.
• Radium-224 – the isotope loaded onto the seeds in every worked example, chosen for its short lived alpha emitting decay chain.
• Immune checkpoint – the regulatory pathways that the claimed substance acts on, including PD-1, PDL-1 and CTLA4.
• Abscopal effect – the shrinkage of untreated distant tumours after local radiation, the systemic outcome claim 1 is aimed at.
• Dendritic cell – the antigen presenting cell whose activation within about a week of irradiation sets the preferred timing window.

---

AU 2026202185 was published in the Australian Official Journal of Patents on 23 April 2026 and is open for public inspection. Patent applications represent inventions that are sought to be protected and do not necessarily reflect commercially available products.