Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

CSM statisticians published their experiences in overcoming the challenges of using adaptive designs for early phase clinical trials. These trial designs are more likely to find the right answer than traditional designs.

From left: Jane Holmes, Eleni Frangou, and Sharon Love, CSM coauthors of a recent paper in BMC Trials on the challenges in implementing model-based phase I designs in a grant-funded clinical trials unit

CSM statisticians are part of a movement to modernise the designs used in early-phase clinical trials for the benefit of patients. As part of this movement, they have published a paper to encourage other clinical trialists to use adaptive, model-based trial designs. They describe the challenges they faced and solutions they used in setting up an adaptive trial called CHARIOT using the time-to-event continual reassessment method.

Phase I trials aim to find the dose of a drug that minimises side effects and maximises efficacy. Many grant-funded clinical trial units in the UK are still using algorithm-based trial designs for their phase I trials, which decide before the trial begins the rules for deciding which trial dose each recruited patient gets. Model-based designs instead use the information gathered from every patient during the trial to decide the dose that is best for the next patient. These new designs are better for patients and are statistically more advanced, but trialists have been reluctant to use them.

CSM medical statisticians Eleni Frangou, Jane Holmes, and Sharon Love recently designed CSM’s first adaptive phase I trial, CHARIOT. They have published their experiences in using a model-based design to encourage others. As Eleni Frangou says, the challenges in using these designs are worth the benefits to patients. ”We faced many challenges when setting up CHARIOT but now have a trial that is more likely to find the correct answer and push forward research for those diagnosed with oesophageal cancer.”

The statistical team found that using a model-based design required more manpower earlier in the trial design than the usual algorithm-based design, particularly before the grant application was made. They needed to write code to implement the design, as their trial unit had not covered this design before, and to run extensive simulations. The trial’s planned management also differs from usual designs, as information will be collected from participants and fed back to the statistical team much more efficiently and quickly than usual. The team effectively overcame these challenges and share their learning in their new paper.

CHARIOT is a single-arm open-label, phase I dose escalation trial and is the first trial examining the combination of radiotherapy, chemotherapy, and the ataxia telangiectasia mutated Rad3-related (ATR) inhibitor M6620 (previously known as VX-970) in patients with oesophageal cancer. The aim is to identify the maximum tolerated schedule associated with no more than a predefined target toxicity level. The model-based design used is the continual reassessment method, referring to the fact that all of the information known about every patient already in the trial is used to decide the dose for the next patient recruited. The design also uses a modification called time-to-event to take into account toxicity from radiotherapy occurring for many weeks after the treatment has finished.

For more information on the benefits of model-based designs for phase I trials, see the first paper in this series and The NIHR Statisticians Research Group on Early Phase Clinical Trials.