Trial modernisation roadmap: the randomised trial → platforms and adaptive designs → decentralised, pragmatic and always-on

The NCI is the US government's cancer research agency, spending ~$7B a year and running the Cancer Centers Program, TCGA, and Rosenberg's cell therapy lab.

SWOG is one of the US National Cancer Institute's cooperative groups: thousands of community and academic sites running practice-changing trials, including the 2026 first-line Hodgkin lymphoma result.

US cooperative group that pairs a therapy group with an imaging group; its E1910 trial moved blinatumomab into frontline leukaemia treatment.

US cooperative group formed from CALGB, ACOSOG, and NCCTG; running OptimICE-pCR, which asks whether TNBC patients with a complete response still need a year of immunotherapy.

The US cooperative group formed from the radiation, gynaecologic, and breast trial groups; it co-led OlympiA, the trial that put a PARP inhibitor after surgery for BRCA carriers.

Canada's national academic trials group, which ran the CHALLENGE exercise trial and co-led the SABR-COMET oligometastasis trial.

The Children's Oncology Group is the world's largest paediatric cancer trials organisation; most children with cancer in North America are treated on or according to a COG protocol.

ClinicalTrials.gov is the master list of clinical trials. Every trial on this site links to its record.

STAMPEDE is the longest-running platform trial in oncology, and showed that both docetaxel and abiraterone extend life when started at first diagnosis of metastatic disease.

Pediatric MATCH was the first nationwide precision-medicine trial for children: every child with a relapsed solid tumour could have their tumour sequenced and, if a matching drug existed, join a trial arm for it. It proved the plumbing works, even though most single drugs given alone did little.

A tumour-agnostic approval lets a drug be used for any cancer carrying a specific molecular feature, regardless of where it started.

FDA approval based on early evidence (like tumour shrinkage) on condition that a confirmatory trial follows.

A quicker, easier measurement used as a stand-in for what really matters. Tumour shrinkage or delayed growth stands in for living longer, on the assumption, not always true, that one leads to the other.

Drugs approved early on promising results should lose that approval automatically if the company fails to finish the follow-up trial by the agreed date.

To get an early approval, a company would set aside the money for the follow-up trial up front, so the trial cannot be quietly abandoned.

Data from routine care rather than clinical trials, used to check whether results hold outside the trial population.

Apps and sensors that let patients report symptoms between visits, which in trials improved survival and cut emergency visits.

Video visits, remote second opinions, and slides reviewed from afar, which let rural and low-resource patients reach specialists.

Pioneer of the 'metasite' decentralised trial model, acquired by eMed in 2024.

Rave, the most used electronic data capture system in clinical trials, plus synthetic control arm services.

Federated real-world data network across health systems, used for trial feasibility and observational studies.

Oncology EHR (OncoEMR) and the largest curated real-world oncology database, acquired by Roche for $1.9B in 2018.

Real-world evidence at scale: what happened to patients with a given genomic profile on a given treatment.

Trials should let people in unless there is a scientific or safety reason to keep them out. Every exclusion rule would need a written reason, reviewed like the rest of the protocol.

Much of trial screening is paperwork, questions and reviewing scans that already exist. Doing this by video and electronic consent before any travel would let patients decide without a wasted trip.

Companies now have to file a plan for enrolling a representative mix of patients. If the trial misses the plan, the label should say so and the company should be required to fill the gap after approval.

Sometimes a trial cannot randomise, so the new drug is compared with past patients' records. Clear published rules on when that is allowed, and how it must be done, would replace case-by-case guesswork.

An ultra-sensitive blood test after surgery that detects leftover cancer months before a scan would.

A blood test that reads fragments of DNA shed by the tumour, so you can genotype or monitor cancer without a needle in the tumour.

Showed that a blood test can safely halve the number of colon cancer patients given chemotherapy after surgery.

Japan's national programme tests whether a blood test after surgery should decide who gets chemotherapy, and whether treating a positive test early helps.

The first trial to change treatment because of a blood test rather than a scan: switching to camizestrant when an ESR1 mutation appeared in the blood delayed progression by seven months.

The first trial to use a blood test for leftover cancer to decide who gets immunotherapy, and it worked.

If a blood test reliably shows whether cancer will come back after surgery, trials could use it instead of waiting years for relapse. Regulators have a process to bless such a test; oncology should use it.

In NICHE-2, four weeks of immunotherapy before surgery wiped out most mismatch-repair-deficient colon cancers, and nobody had relapsed three years later.

Using an early PET scan to spot responders let a third of women be cured of HER2-positive breast cancer without any chemotherapy.

The trial that added immunotherapy to pre-surgery chemotherapy for triple-negative breast cancer and, uniquely, improved survival.

Giving immunotherapy for a few weeks before surgery produces a tumour sample that shows exactly what the drug did. That is the fastest way to learn who responds.

Between diagnosis and surgery there are usually a few weeks. Giving a new drug in that window and comparing the tumour before and after surgery shows whether it hits its target in real people, quickly and cheaply.

Most drug doses were chosen as the highest a person can tolerate, which is often more than they need.

The FDA now asks companies to find the right dose of a cancer drug before approval. If every regulator asked the same way, companies would do it once and doses would match worldwide.

Cancer drugs are usually tested at the highest dose patients can stand, and that dose sticks for life. Comparing two or more doses head-to-head before the big trial would find doses that work as well with fewer side effects.

Six months of trastuzumab was almost as good as twelve, with half the heart toxicity, though twelve remains the global standard.

Independent radiologists who re-read every scan in a trial the same way, so response rates mean the same thing across hospitals.

The test that decides whether a specific drug is right for you, approved together with the drug.

Software, increasingly LLM-based, that reads a patient's record and finds trials or guideline options they qualify for.

Trial Library is software that helps everyday cancer clinics spot which patients might qualify for a clinical trial, refer them, and remove practical barriers like transport so more people, especially in under served communities, can join trials.

Massive Bio uses artificial intelligence to read a cancer patient's medical records and match them to clinical trials they may be eligible for, anywhere in the world, with doctors checking the results.

Genomic testing plus one of the largest multimodal clinical datasets, used for AI models and trial matching.

When an oncologist opens the order screen to prescribe a new line of treatment, the record would show the trials this patient may fit, with the nearest open site and a one-click referral.

Before a trial is finalised, run its entry rules against records of real patients with that cancer and report what fraction would qualify. If it is under half, explain why.

Measuring tumours on scans for trials is slow, expensive and inconsistent between readers. Software that measures lesions and flags changes, checked by a radiologist, could make trial endpoints cheaper and more reliable.

Before spending millions on a randomised trial, analyse existing patient records as if the trial had already happened. If the answer is obvious or the question is unanswerable, skip or redesign the trial.

Pragmatic trials want to use the progression dates recorded in ordinary clinic notes instead of expensive protocol scans. Checking how well those routine records match formal trial measurements would show when that shortcut is safe.

Using a model of what would have happened to a patient on standard treatment, so fewer people have to be randomised to it.

Rather than building a new trial from scratch for every drug, keep one permanent trial open per cancer where new treatments can be slotted in and dropped out, sharing the same patients, control group and infrastructure.

Instead of starting a new trial for every drug pair, keep one always-open trial per cancer that new arms can join and leave, sharing the same control group.

Adding a new arm to an international platform trial currently needs approval in every country again. A single, pre-agreed process would let arms open in weeks.

As results come in, the trial sends more new patients to the arms that are working and fewer to those that are not, so more people benefit and bad arms die faster.

Blood tests can now find leftover cancer months before scans, but most patients who test positive have nothing to enrol in. One standing trial per country would fix that.

Instead of a new trial for each resistance mechanism, one continuous trial could sort patients into arms based on the reason their last treatment failed.

Rare cancers together are a fifth of all cancers, but each is too small for its own trial. One permanent trial with many arms would give all of them a route.

The UK's RECOVERY trial tested many cheap COVID drugs quickly by randomising thousands of ordinary hospital patients. Cancer needs the same machine for old drugs.

Cheap generic versions of targeted cancer drugs like imatinib exist, but patients with rare tumours carrying the matching mutation often cannot get them. A structured programme would treat them and collect the evidence.

Give combinations before surgery and look at how much tumour is left when it is removed. That answer comes in months, so many pairs can be tested quickly.

National cancer registries already collect the outcome data. Adding a randomisation button lets doctors compare two standard treatments across thousands of patients at a fraction of the usual cost.

Add-Aspirin is an 11,000-patient trial across four common cancers in the UK, Ireland and India asking whether cheap daily aspirin after curative treatment stops cancer coming back.

The first randomised trial to show that a coached exercise programme after cancer treatment reduces recurrence and death, in colon cancer.

Doctors record only part of what patients suffer. Make it compulsory that patients report their own side-effects in every trial used to approve a drug, and that those data are published next to the doctor-graded ones.

Every patient on the new drug is compared with every patient on the old one: who lived longer, and if equal, who had fewer serious side effects, and if still equal, who felt better. The share of 'wins' becomes the result.

Trials report side effects as a table of percentages that hides how long they lasted, how bad they felt and whether people stopped treatment. Tolerability should be measured with defined endpoints and a decision rule, like efficacy.

Most people with cancer are over 65, but trials mostly enrol younger, fitter people. Requiring a group of older patients, assessed for frailty, in every big trial would show whether the drug works and is safe for those most likely to receive it.

Drugs are approved on trials of fit, younger patients and then given to everyone. A required follow-on trial in older, sicker and more diverse patients would show whether the benefit holds in real life.

Most of the world's cancer patients live in countries that host almost no registrational trials. Including sites there, and paying to build them up, would make results apply globally and speed local access.

Trial visits take hours and cost people wages. Paying a fair hourly rate for time spent beyond normal care would make trials possible for those who cannot afford unpaid days off.

Discussing and enrolling a patient in a trial takes an oncologist far longer than prescribing the usual treatment, and they are not paid for it. Paying for that time would remove a quiet disincentive.

Trials often measure a stand-in for survival, such as time until the cancer grows on scans. An independent body would test, for each cancer and treatment type, whether the stand-in actually predicts survival, and publish the answer.

Seven in ten cancer deaths happen in countries with almost no cancer care, and even rich systems cannot afford every new drug. The roadmap is the set of levers that already work, from generics and biosimilars to trials that cut the dose, and the ones being built: regulators trusting each other's reviews, pooled purchasing, and drugs and cell therapies made where patients live.

Fewer than one in ten adults with cancer joins a trial. Trials close for lack of patients, not lack of ideas.

A phase 3 trial takes years and hundreds of millions of dollars, and often answers a question that has already moved on.

Older, Black, Hispanic, Asian, rural, poor and multimorbid patients are under-represented, so results may not apply to them.

We do not reliably know what happens to patients after approval, so we cannot tell which drugs deliver in practice.

Most people with cancer are over 65 but most trial patients are younger and fitter. We guess how to treat the majority.

Negative trials, failed drugs and abandoned programmes are rarely published, so the same mistakes are repeated.

When a public company announces a trial failure, it should be required to give the actual numbers, as it must for a success.

Keep a running, well-documented list of cancer practices and approvals that were reversed by later evidence, so the pattern of mistakes is visible and teachable.

Regulatory divergence means a drug approved in one country can take years to reach another, or never arrive.

Project Orbis is a scheme where the FDA and partner regulators (Australia, Canada, UK, Switzerland, Singapore, Brazil, Israel) review a cancer drug at the same time.

Regulators in several countries already look at the same cancer drug dossier at the same time. Let them split the work and write one report instead of six.