Trial design
Adaptive trials are well-suited to experimental treatments for cancer, where changes in tumour size can be observed in a matter of months.
specifically”, Wason says, “some of the adaptive designs – like umbrella trials, for example – have developed because of new biological findings about stratifying cancers and tumour mutations”. As a result, oncology has been one important area that has seen an increase in the use of adaptive trials. As Wason explains: “One of the things that influences how much benefit an adaptive design gives, is how long it takes to observe the patient’s outcome when given treatment. Generally, we have a primary outcome that we’re looking to show is improved by the treatment, which could be mortality, or it could be something more short term.” In oncology, it’s common to use tumour response, which is how the tumour changes in size. The adaptive design benefit comes from when you observe outcomes pretty quickly, says Wason. “So, if you’re using tumour response, for example, that’s observed within a few months of randomising the patient. If you imagine that it takes two years to recruit patients and you’re observing patient outcomes after three months, then it means that halfway into the recruitment of the trial you have a decent amount of information to make the decision about whatever the adaptive design is aiming to do.” Zhou also notes the prevalence of adaptive designs in cancer trials. “In oncology, dose finding, dosing regime and drug combinations are very important,” he says. “Say you want to combine a new targeted therapy with some existing chemotherapy, these kinds of exploratory trials are especially suitable for adaptive designs.” This is due, in part, to the sequential nature of the Bayesian framework, which focuses more on learning than confirming that a drug is effective. “For dose finding, for example, we are trying to learn what would be a good dosage for the patient – and that is an estimation question rather than a hypothesis- testing question,” says Zhou. “If the dose is really high and we have seen a lot of side effects, then in those cases it would be very desirable to change the dose level, which makes the design adaptive, and the Bayesian framework is especially useful [for this].”
36
The advantages of an adaptive approach are wide reaching: among the key benefits, Wason lists improved efficiency, the ability to tailor specific treatments to patients, as well as flexibility and robustness – though, as he notes, “it really depends on the type of adaptive design”. Of course, there are also a number of challenges when it comes to adaptive designs. “With a simple trial, there’s a formula for getting the sample size required,” says Wason. “But there’s no formula that exists for more complex adaptive designs, so you typically need more statistical input to come up with the trial design. You also need other disciplines for the trial, like the data manager and trial managers, so it generally increases the amount of effort needed.” All of that said, however, Wason is confident that “overall, the benefits outweigh the drawbacks and many of the issues that we had when adaptive designs were first being used, have been overcome”. The last ten years have seen an increase in the design and use of adaptive trials – so what do the next ten years hold in store? Zhou is hopeful that with greater technological and biological improvements, and with a better understanding of Bayesian statistics, “we will see more and more adaptive clinical trials in the future”. Wason is optimistic too, if a little more reticent about how widespread the use of adaptive methods will become. “There will be more and more software available, there will be more examples, and there will be more learning from the trials about things that went wrong and [how they] were overcome,” he says. “So, I think there will be fewer barriers to using adaptive designs – but I also think that there will be greater understanding about the fact that they’re not necessarily always appropriate to use.” Wason is right to pull on the reins at this early stage – the hype that surrounds novel approaches can be unhelpful when judging which methods are really the most appropriate – but there’s no doubt about it: adaptive designs are here to stay. ●
Clinical Trials Insight /
www.worldpharmaceuticals.net
CI Photos/
Shutterstock.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53
