The first few weeks may matter most: new clues on who immunotherapy helps
What happened
On 16 September 2026, the journal Cancer Cell published a study that tries to answer a question that has puzzled cancer doctors for years: why does immunotherapy work brilliantly for some people and do almost nothing for others? Immunotherapy does not attack cancer directly. Instead it takes the brakes off your own immune system, so your body can go after the tumour itself. When it works, it can work very well indeed. But it does not work for everyone, and until now there has been no reliable way to tell in advance. The research was led by Professor Dvir Aran and Dr Zhongyang Lin at the Technion in Israel, working with Professor Jurgen C. Becker of the German Cancer Consortium and others. Instead of only looking at what a tumour looked like before treatment started, the team looked at how it changed during the first weeks of treatment. They pulled together data from about 200 patients across 16 separate groups of studies, covering several cancer types, and generated their own melanoma data as well. The finding was consistent. Tumours that shifted towards an 'inflamed', immune-rich state early on tended to do well. Tumours that shifted the other way, towards a state that shuts the immune system down, tended not to. The team then built something they call a 'transition score'. It uses a sample taken before treatment even starts to estimate how likely that tumour is to move in the helpful direction. They tested it against roughly 1,300 more patients.
Why this matters
Immunotherapy is not a gentle treatment. It can cause real side effects, because an immune system with the brakes off can also attack healthy tissue: the thyroid, the bowel, the skin, the liver. Most people manage these, but some become seriously unwell. So every course of immunotherapy involves a trade. If the treatment is going to work, the trade is clearly worth it. If it is not going to work, that person has taken on the risk and the hospital visits and the waiting, and lost weeks they could have spent on a different treatment. At the moment, doctors mostly find out by trying. Scans at around three months tell you whether the tumour has shrunk. That is a long time to wait. A reliable early signal would change that conversation. It would let doctors switch approach sooner for people who are not responding, and reassure people who are. There is a second, quieter point. The study suggests the useful question is not 'what does this tumour look like?' but 'what is this tumour capable of becoming?' That is a genuinely different way of thinking about cancer treatment.
What the evidence actually says
This is careful work, and it is worth being clear about what it is and is not. Around 200 patients in the main analysis is a reasonable size for this kind of detailed molecular study, but it is not huge. The team did test the transition score against about 1,300 patients, which is a real strength - many studies like this never test their idea on a separate group at all. The 16 cohorts were collected by different teams, in different places, at different times, using different methods. Pooling them is standard practice and it makes the finding more general. It also introduces noise, because the samples were not all handled the same way. Most importantly, the researchers themselves say plainly that these findings are 'not yet ready for the clinic'. That is not false modesty. Prediction tools in cancer have a long history of looking excellent in the data they were built from and then underperforming when used on new patients in the real world. The only way to know is to test it properly, in advance, in a planned trial. The team has launched a collaborative project called DYNAMO to do exactly that kind of validation and expansion. That is the right next step, and it will take time. So: this is not a test your oncologist can order. It is not something to ask for. It is a serious piece of research pointing at a promising idea that now has to prove itself.
Practical advice
If you or someone close to you is having immunotherapy, none of this changes your treatment plan. But a few things are worth holding on to. Ask your team how and when they will know whether it is working. Most people are scanned at set points. Knowing the plan in advance makes the waiting easier to bear, because you know what you are waiting for. Report side effects early, even mild ones. With immunotherapy this genuinely matters. Tiredness, diarrhoea, a rash, feeling unusually cold or unusually hot - these can be early signs that the immune system is attacking healthy tissue, and they are much easier to manage when caught early. Cancer teams would far rather hear from you too soon than too late. Carry your alert card. People on immunotherapy are usually given one. If you end up in A&E or at an out-of-hours service, it tells whoever sees you that your symptoms need treating differently. Ask about clinical trials. If a treatment is not working, a trial is sometimes an option worth exploring. Your oncologist can tell you what is realistic for your situation. And be careful with headlines, including this one. 'Scientists discover why immunotherapy fails' would be an overstatement. This is a clue, not an answer.
What to know
Researchers tracked what happens inside tumours during the first weeks of immunotherapy, across about 200 patients and 16 study groups, and found that early movement towards an immune-rich state predicted success better than the tumour's starting state did. They built a 'transition score' from pre-treatment samples and tested it on around 1,300 more patients. It is a thoughtful study, published in a strong journal, with a proper validation step. It is also, in the authors' own words, not ready for the clinic, and a follow-up project called DYNAMO has been set up to test it further. If you are on immunotherapy: keep to your plan, report side effects early, and carry your alert card. That is the part that helps you this month. Sources: Cancer Cell (Cell Press), study led by Dvir Aran and Zhongyang Lin on early tumour microenvironment transitions and immunotherapy response, 16 September 2026, https://www.cell.com/cancer-cell/home. Medical Xpress, 'Early changes in the tumor environment may explain why immunotherapy works for some patients but not others', 16 September 2026, https://medicalxpress.com/news/2026-09-early-tumor-environment-immunotherapy-patients.html. Cancer Research UK, 'Immunotherapy', https://www.cancerresearchuk.org/about-cancer/treatment/immunotherapy. This article is for general information and does not replace advice from a doctor, pharmacist or other qualified healthcare professional.
Related articles
One baby's heartbeat, the wrong screen: a new MHRA warning on labour monitors
The UK's medicines and devices regulator has issued a safety notice about wireless monitors used to track babies' heartbeats during labour. The Medicines and H…
One play specialist for every 22,000 child hospital admissions
A children's charity has found that hospitals in the UK have far too few health play professionals. Health play professionals are trained staff who help childr…
33 studies out of 38 pointed the same way: women are offered less active treatment
Researchers at the University of St Andrews have published a review of global research into how men and women are treated for the same illnesses. The answer wa…