Originally published 2021-10-12.
In October 2021, researchers in the UK reported results from CheckMate 651, a large phase III trial that tested two immunotherapy drugs, nivolumab and ipilimumab, as a first treatment for head and neck cancer that had come back or spread. The short answer to what it found: some patients did remarkably well, a few for years, and the drugs caused far fewer severe side effects than the chemotherapy-based standard. But across the whole trial the combination did not prove it helped people live longer, so it was not a cure and did not replace existing treatment.
Headlines at the time spoke of a treatment that "destroys" tumours in terminally ill patients. That was true for some individuals and misleading as a description of the trial. Below is what the study actually measured, why the result was described as both disappointing and encouraging, and what it means for patients and families looking for options today.
What the CheckMate 651 trial tested
CheckMate 651 enrolled 947 adults with squamous cell carcinoma of the head and neck, the most common type of cancer in the mouth, throat and voice box. Every participant had disease that had returned after earlier treatment or had spread to other parts of the body, and none had yet received drug treatment for this advanced stage.
Patients were randomly split into two roughly equal groups. About half received nivolumab plus ipilimumab. The other half received the "EXTREME" regimen, which was then a standard first-line option: the targeted antibody cetuximab combined with platinum chemotherapy and the drug 5-fluorouracil.
The trial was led internationally by researchers including Professor Kevin Harrington of the Institute of Cancer Research (ICR) in London and the Royal Marsden NHS Foundation Trust. It was funded by Bristol Myers Squibb, which makes both immunotherapy drugs. The results were presented at the European Society for Medical Oncology (ESMO) virtual congress in September 2021, and the ICR publicised them on 11 October 2021.
How nivolumab and ipilimumab work
Both drugs are immune checkpoint inhibitors. Our immune system has built-in "brakes" that stop T cells from attacking healthy tissue. Many tumours exploit those brakes to hide. Checkpoint inhibitors block the brakes so that T cells can recognise and attack cancer cells again.
- Nivolumab blocks PD-1, a receptor on T cells. When tumour cells display a partner molecule called PD-L1, PD-1 switches the T cell off. Blocking it keeps the T cell active.
- Ipilimumab blocks CTLA-4, a different brake that acts earlier, when T cells are first being primed to respond.
Using both at once is meant to release two separate brakes. The approach had already worked in melanoma and some other cancers, so researchers hoped it would help in head and neck cancer too.
One common misunderstanding, repeated in some coverage at the time, is that these drugs are built individually for each patient. They are not. Nivolumab and ipilimumab are standard, mass-produced antibodies given by intravenous infusion. What varies is how each person's immune system and tumour respond.
What the results showed, in plain numbers
The trial had two main goals: to show longer overall survival in all patients, and in a subgroup whose tumours had high levels of PD-L1 (a combined positive score of 20 or more). It met neither goal by the strict statistical standard set in advance.
- All patients: median overall survival was 13.9 months with immunotherapy and 13.5 months with EXTREME, a difference too small to count as a real effect.
- High PD-L1 group: median survival with immunotherapy was 17.6 months, which the ICR described as about three months longer than with EXTREME and the longest median survival reported in this patient group at the time. The trend favoured immunotherapy but still fell short of statistical significance.
- Serious side effects: severe (grade 3 or 4) treatment-related side effects affected about 28% of patients on the immunotherapy pair, compared with about 71% on EXTREME, according to the final published analysis.
Part of the explanation, the researchers said, was that patients in the comparison group did better than earlier data predicted, which made a difference harder to show. Professor Harrington summed up the tension in a line widely quoted at the time: "Despite the lack of statistical significance, these results are clinically meaningful."
That phrase is worth unpacking. "Not statistically significant" means the trial could not rule out that the difference happened by chance. "Clinically meaningful" is the investigators' judgement that the size of the benefit in some patients, plus the lower toxicity, matters to people living with the disease. Both statements can be true, but only the first one decides whether regulators approve a new use.
Barry Ambrose: an exceptional responder
The story that made headlines was that of Barry Ambrose, a 77-year-old from Bury St Edmunds in Suffolk. He had throat cancer that had already spread to his lungs and was told he might have months to live. He joined the trial at the Royal Marsden in August 2017.
Within about two months of starting immunotherapy, the tumour in his throat had disappeared on scans, although disease remained in his lungs. He later had further treatment, including chemotherapy and surgery, and by 2021 the ICR reported that he had no evidence of disease. He said the trial "turned out to be a lifeline" and described being able to keep sailing, cycling and spending time with family during treatment.
Doctors call patients like Mr Ambrose exceptional or long-term responders. Immunotherapy often produces a "long tail" on survival curves: a minority of patients respond strongly and for a long time, while many others gain little. That pattern is exactly why a dramatic individual story and a neutral trial average can sit side by side. One patient's outcome cannot tell anyone how likely the same result is for them.
Why side effects mattered so much
For people with advanced head and neck cancer, quality of life is a central concern. The cancer itself can affect swallowing, speech, breathing and appearance. Chemotherapy combinations such as EXTREME frequently add nausea, tiredness, mouth sores, loss of appetite, infections and skin reactions.
Immunotherapy has its own risks. Because it releases immune brakes, it can cause the immune system to attack healthy organs, leading to inflammation of the bowel, lungs, liver, skin or hormone glands. These problems are usually manageable when caught early but can occasionally be serious. Even so, in CheckMate 651 the overall burden of severe side effects was clearly lower with the immunotherapy pair, which is why researchers argued it could be an alternative for some patients who cannot tolerate chemotherapy.
Where this fits in head and neck cancer treatment
CheckMate 651 did not arrive in a vacuum. Since 2019, another PD-1 inhibitor, pembrolizumab, has been approved in many countries as a first-line option for recurrent or metastatic head and neck cancer, either alone for tumours expressing PD-L1 or combined with chemotherapy. That earlier success is part of why CheckMate 651's mixed result did not change practice: an immunotherapy option already existed, and adding a second checkpoint inhibitor did not prove better overall.
The trial still added useful knowledge. It reinforced that PD-L1 levels help predict who benefits from checkpoint inhibitors, it showed a lower-toxicity profile for the combination, and it gave researchers more data on long-term responders. Work continues on biomarkers that could pick out the patients most likely to respond, and on combining immunotherapy with surgery and radiotherapy at earlier stages.
Common myths and practical takeaways
- Myth: immunotherapy cures head and neck cancer. It can lead to long remissions in some people, but most patients with advanced disease do not respond that strongly.
- Myth: a negative trial means a drug is useless. A trial that misses its main goal can still reveal subgroups who benefit and safety advantages worth studying further.
- Myth: immunotherapy has no side effects. It is often gentler than chemotherapy, but immune-related side effects can affect almost any organ.
If you or someone close to you has head and neck cancer, talk to a doctor or the oncology team about which treatments fit the specific diagnosis, including whether PD-L1 testing has been done and whether any clinical trials are open nearby. Cancer charities and national trial registries list current studies and can help families prepare questions for appointments.
The honest summary of CheckMate 651 is that it offered real hope to a minority, a gentler side-effect profile to many, and no proven overall survival advantage. Stories like Barry Ambrose's are genuine and inspiring, and they are also a reminder of why careful trials, rather than single cases, guide cancer care.
Sources: The Institute of Cancer Research: Immunotherapy combination could be alternative to 'Extreme' chemotherapy in some head and neck cancers (11 October 2021); Oncology Nursing News: CheckMate 651 misses primary end points; ClinicalTrials.gov: NCT02741570 (CheckMate 651); original reporting by The Guardian.

COMMENTS