Relapsed head and neck cancer gets first new FDA drug in 10 years

The US Food and Drug Administration has approved the first new drug in 10 years for head and neck cancer that comes back or spreads after chemotherapy and immunotherapy. On 5 November 2026 the agency granted accelerated approval to Rybrevant Faspro, an under-the-skin injection of the antibody amivantamab made by Johnson & Johnson's Janssen Biotech. In the trial behind the decision, tumours shrank or disappeared in 42% of 102 patients.

That figure needs context. Older second-line options such as cetuximab or paclitaxel shrink tumours in roughly one patient in five once immunotherapy has failed. Nobody yet knows whether amivantamab helps people live longer, because the trial had no comparison group. The FDA has made continued approval depend on a larger confirmatory study.

What exactly did the FDA approve?

The approval covers adults with recurrent or metastatic head and neck squamous cell carcinoma (HNSCC) whose cancer has grown on or after two earlier types of treatment: platinum-based chemotherapy and an immunotherapy drug that blocks PD-1 or PD-L1, such as pembrolizumab or nivolumab. Recurrent means the cancer has come back after first treatment. Metastatic means it has spread to other parts of the body.

Rybrevant Faspro is given on its own, as a single drug. The FDA's approval notice says the dose is based on body weight and given every three weeks in the trial, and Johnson & Johnson says the injection can be given every three or four weeks for this use. The drug is co-formulated with hyaluronidase, an enzyme that loosens tissue under the skin so a large dose can be absorbed. J&J describes it as a five-minute injection.

There is one big exclusion. The label says the drug is not for oropharyngeal cancer (cancer of the back of the throat, tonsils or base of the tongue) that is linked to the human papillomavirus (HPV). Doctors usually check this with a p16 test on the tumour sample. HPV-positive throat cancers behave differently and were kept out of the main trial population.

This is the first time the injectable form of amivantamab has been approved for anything outside lung cancer. In the US it was already cleared for several types of non-small cell lung cancer with EGFR mutations. The head and neck approval does not require any EGFR or MET test.

How does amivantamab work?

Amivantamab is a bispecific antibody, meaning one molecule has two different arms. One arm latches onto EGFR (epidermal growth factor receptor) and the other onto MET. Both are proteins on the surface of cancer cells that send growth signals.

In HNSCC that is not caused by HPV, EGFR and MET are overexpressed in 80% to 90% of recurrent or metastatic tumours, according to the Journal of Clinical Oncology paper that reported the main results. Blocking EGFR alone is an old idea. Cetuximab, an EGFR antibody, has been used in head and neck cancer since the 2000s. The problem is that tumours often get round it by switching on MET instead. Hitting both targets at once is meant to close that escape route.

The antibody also has what the company calls immune cell-directing activity. It flags cancer cells for attack by immune cells such as macrophages. How much each of these effects contributes in head and neck cancer has not been pinned down.

What did the OrigAMI-4 trial find?

The approval rests on cohort 1 of OrigAMI-4 (NCT06385080), a phase 1b/2 study that tests amivantamab in several groups of people with head and neck cancer. Results for the full cohort of 102 patients were presented at the ASCO Annual Meeting in Chicago and published in the Journal of Clinical Oncology on 31 May 2026, with Barbara Burtness of Yale Cancer Center as first author.

Everyone in the cohort had cancer not linked to HPV, had already received immunotherapy and platinum chemotherapy, and had never had an EGFR drug such as cetuximab. Nobody could have had more than two earlier lines of treatment for advanced disease. Just over half (52%) had one earlier line and 48% had two.

The median age was 63, and 77% were men. Seven in ten (71%) had a history of smoking. Almost half the tumours (47%) started in the mouth, with 24% in the larynx (voice box), 16% in the hypopharynx (lower throat) and 14% in the oropharynx. About 41% of patients were enrolled in East Asia, 28% in North America and 28% in Europe.

Responses were judged by a blinded independent central review, meaning radiologists who did not know the patients read the scans using the standard RECIST 1.1 rules. At a median follow-up of 11.8 months, the results looked like this:

The investigators' own reading of the scans gave a slightly higher response rate, 47%. Regulators rely on the independent figure. Responses came quickly, with a median time to response of 6.6 weeks. When the data were frozen on 18 March 2026, 27 of the 43 responders were still responding and the median length of response had not been reached.

How does 42% compare with older treatments?

The FDA says currently approved therapies in this setting have response rates of "around 20% or less". The JCO paper points to INTERLINK-1, a randomised trial in which cetuximab alone was given to people whose cancer had already been through immunotherapy and platinum. There, cetuximab shrank tumours in 24%, produced complete responses in 1%, and gave a median progression-free survival of 3.8 months and median overall survival of 8.6 months.

On paper, amivantamab looks better on every one of those measures. These are cross-trial comparisons, though. The patients in OrigAMI-4 were picked for a trial, mostly had a good performance status (all scored 0 or 1 on the ECOG scale, meaning they were fully active or only mildly limited) and had never had an EGFR drug. People in routine care are often frailer. A single-arm trial cannot tell you how the same patients would have done on cetuximab. Our guide to reading trial results explains why a number without a comparison arm is harder to judge.

Here is a way to read the headline numbers that you can repeat yourself. Picture 100 people starting treatment. About 42 would see their tumours shrink by at least 30%, and about 15 of those would have no visible cancer on scans. The share whose response lasted six months or more was 24 of 102 in the trial, which works out at roughly 24 in every 100 people treated. So about one person in four got a response that lasted at least half a year. Of the rest, about 35 in 100 had cancer that stopped growing for a while without shrinking enough to count as a response.

What are the side effects?

Amivantamab hits EGFR, which is also active in healthy skin, nails and the lining of the mouth. Side effects in those places are expected.

According to the prescribing information summarised by J&J, the most common side effects in OrigAMI-4 were rash (79% of patients), tiredness (39%), nail problems (35%), mouth sores (32%) and swelling (28%). The JCO paper also reported low blood albumin, a protein made by the liver, in 50% of patients, though only 5% had a severe drop.

Serious adverse reactions affected 39% of patients. The most frequent serious one was pneumonia, in 8%. Six patients (6%) died of adverse events during the trial. The investigators judged one of these, from inflammation of the lungs (pneumonitis), to be caused by the drug and the other five to be unrelated. Interstitial lung disease, a type of lung scarring and inflammation, is one of the warnings on the label.

Reactions to the injection itself were mild in the trial. They affected 15% of patients and were all grade 1 or 2, and the label says most happened with the first dose. Patients are given antihistamines, fever reducers and steroids beforehand, and are watched for two to four hours after the first injection. Only 8% stopped treatment because of side effects the investigators linked to the drug, though 23% needed a dose reduction for any adverse event.

The label also tells patients to limit sun exposure during treatment and for two months afterwards, to use broad-spectrum sunscreen, and to report new eye symptoms quickly. The drug can harm an unborn baby, so effective contraception is needed during treatment and for three months after the last dose.

Why is this an accelerated approval?

Accelerated approval lets the FDA clear a drug for a serious disease on the strength of an early measure, here tumour response and how long it lasts, before there is proof that people live longer or better. The company then has to run a confirmatory trial. If that trial fails, the FDA can withdraw the approval. We explain the wider process in our guide to how new medicines get approved.

The FDA has used this route often this autumn. It gave accelerated approval to dostarlimab for a type of rectal cancer in October, and to Voceryq for recurrent respiratory papillomatosis, an HPV-driven airway disease.

Amivantamab got breakthrough therapy designation for head and neck cancer in February 2026, based on the first 38 patients in the cohort, and priority review in July. The review ran under Project Orbis, an FDA scheme for reviewing cancer drugs alongside other regulators. The partners named for this review were Brazil's ANVISA, Health Canada, Israel's Ministry of Health and Singapore's Health Sciences Authority.

The confirmatory study is OrigAMI-5 (NCT07276399). It is a randomised phase 3 trial with an estimated 600 participants, and it tests a different situation from the one just approved. People with newly diagnosed recurrent or metastatic disease get either amivantamab plus pembrolizumab and carboplatin, or the current standard of pembrolizumab, 5-fluorouracil and a platinum drug. Its two primary endpoints are overall survival and response rate. The registry entry gives an estimated primary completion date of June 2029.

That gap matters for anyone weighing the drug. If OrigAMI-5 shows a survival benefit in first-line treatment, it should support the later-line use too. If it does not, the head and neck approval could be reviewed.

Is it available in the UK or Europe?

Not for head and neck cancer, as far as we can find. Johnson & Johnson's announcement does not mention a filing with the MHRA or the European Medicines Agency for this use, and we found no public record of one. In the UK and EU, amivantamab is licensed only for lung cancer.

UK hospitals did take part in the main study. The list of cohort 1 investigators in the JCO paper includes the Royal Marsden Hospital in London, where Kevin Harrington is the senior author, along with University College London Hospitals, Addenbrooke's Hospital in Cambridge and The Christie in Manchester. OrigAMI-5 also has a UK ethics approval, given by the London Dulwich Research Ethics Committee on 7 November 2025, according to the Health Research Authority.

For people in the UK, a trial may be the most realistic way to get amivantamab for head and neck cancer in the near term. Our guide on how to find and join a clinical trial covers the UK registries and what to ask the research team.

What should patients ask their cancer team?

If your head and neck cancer has come back after immunotherapy and chemotherapy, these questions can help frame a conversation with your oncologist:

• Is my cancer HPV-positive, and was a p16 test done on my tumour?
• Have I had cetuximab or another EGFR drug before? People who had were excluded from OrigAMI-4.
• How many lines of treatment have I had since the cancer came back or spread?
• Are there open trials of amivantamab or similar drugs, such as OrigAMI-5 or studies of other EGFR-targeting antibodies, that I could join?
• How would skin, nail and mouth side effects be managed, and who do I call if a rash gets worse?

This article is for general information and is not medical advice. Decisions about treatment should be made with your own doctors, who know your history.

Other drugs are chasing the same group of patients. The JCO paper notes that petosemtamab, an antibody against EGFR and LGR5, had a response rate of 36% in an earlier trial, and ficlatuzumab plus cetuximab reached 38% in a randomised phase 2 study. Each of those numbers came from a different trial population, so none should be read as a direct ranking. The next firm answer will come from OrigAMI-5.