Instead of testing one drug at a time, a novel lung cancer study announced on Thursday will allow British researchers to test up to 14 drugs from AstraZeneca and Pfizer at the same time within one trial.
The aim is to quickly pinpoint medicines that can fight advanced lung cancer by targeting specific rare genetic mutations - and it upends the normal approach of putting a particular drug at the centre of a study.
Harpal Kumar, chief executive of charity Cancer Research UK, which is working on the 25-million-pound ($42-million) project with the two drugmakers, said the new approach would "re-write the rule book on how we do clinical trials".
"We are shifting the emphasis from designing a trial around a specific drug to designing it around selecting from a range of drugs for a specific patient," he told reporters.
The National Lung Matrix trial, which is expected to open in July or August at centres across Britain, is part of a growing trend in cancer research to remodel the way new drugs are tested to keep up with the age of genomic medicine - fine-tuning treatments to the genetic profile of patients.
The U.S. group Friends of Cancer Research said in November it was backing a similar multi-drug, multi-arm clinical trial to evaluate five experimental lung cancer drugs.
Because any single gene mutation behind a particular cancer can be extremely rare, perhaps affecting as few as 1 or 2 percent of patients, studying one drug at a time is extremely challenging.
The new trial offers a way around this by creating a protocol under which scientists can test patients' tumour samples for multiple gene faults and then assign them to an appropriate drug from among those in the trial for testing.
Significantly, it does away with the idea of randomisation, where some patients are given a treatment and others get a "control" substance for comparison - traditionally the gold standard in drug testing.
Increasingly, researchers are questioning whether such randomisation makes sense in patients with incurable cancer if a certain treatment is genetically bound to work much better in some people than others.
Faster, smaller testing could shave years off the usual drug approval process. That also points to lower costs, although firms will be left with small markets from which to recoup investment.
"It's premature to talk about pricing but it will make drug development more cost-effective," said Menelas Pangalos, AstraZeneca's head of innovative medicines.
Mace Rothenberg, senior vice president at Pfizer Oncology, said: "It should be a more efficient process but it is for a more niche, focused group of patients."
The new trial in non-small cell lung cancer will use up to 12 experimental drugs from AstraZeneca, including two compounds designed to boost the immune system that will be given to patients who do not show one of 21 sets of gene faults.
Pfizer, meanwhile, is supplying Xalkori, which is approved for certain rare lung cancers and may also work in some even more unusual types, as well as the experimental drug palbociclib, which has shown promise in breast cancer.
Around 15 to 20 patients will be given each drug initially and promising medicines could be fast-tracked into larger-scale testing. Those that show no benefit will quickly be dropped.
Kumar said new medicines could be added to the existing trial as science advances, and he hopes other drug companies will join the programme in due course.
A successful programme in lung cancer could also lead to the concept being rolled out into testing drugs for other tumour types, he said.