Many drugs never get to market
Cancer drug research is entering a new era which will mean more successful drugs for patients, says a charity.
However, Cancer Research UK called for pharmaceutical firms and academics to be more open about those which do not make the grade.
Improved knowledge of cancer's biology means 18% of new drugs, compared with 5% previously - will become standard treatments, said the charity's experts.
Their research was published in the journal Nature Reviews Drug Discovery.
The hunt for cancer drugs is carried out on a massive scale, but there is also a massive failure rate, as promising candidates fall by the wayside in clinical trials.
This costs drug firms and charities such as Cancer Research UK many millions, although scientists can learn lessons even from expensive failures.
Some studies have estimated that, in the past, just 5% of cancer drugs in the pipeline actually end up in the clinic being used day to day.
Data on 974 drugs under development, gathered by Cancer Research UK experts, suggests that 18% of them will prove successful in clinical trials.
It is hoped that many of them will be alternatives to conventional chemotherapy, which can have unpleasant and dangerous side-effects, targeting the mechanisms of cancer cells more directly, with less damage to healthy cells.
Dr Ian Walker, the licensing manager at the charity's commercial development arm, said: "This clearly demonstrates the benefits of developing molecularly targeted treatments for cancer - understanding more about the basic biology of cancer is making a real difference to the success rate for new anti-cancer drug development."
The ability to tailor drug choices to the genetic make-up of patients in some cases is also having a benefit.
However, Professor Herbie Newell, also from Cancer Research UK, said that minimising the number of "failures" - and their cost to the industry - would be vital, and this could be helped if researchers and drug companies were more open about what worked, and what did not:
"We strongly believe that both industry and academia must improve the availability of data related to failed as well as successful drug development programmes.
"The sharing of such information can only be beneficial for clinical, scientific and commercial reasons - and will help measure our progress as well as pinpoint areas for improvement."