Adaptive Designs Improving Efficiency of Real Trials

An exciting adaptive trial design developed by statisticians at the MRC Biostatistics Unit Hub for Trials Methodology Research has led to the successful, early completion of a trial testing a new scanning technique in renal cell carcinoma. The trial, funded by Cancer Research UK and the NIHR Cambridge Biomedical Research Centre, used a novel two-stage adaptive design. This design allowed the trial to stop after the first stage due to overwhelming evidence that using 18Fluorine-labelled sodium fluoride improved the detection of bone metastases, compared to two routinely used scanning techniques.

Adaptive designs are a recent development in clinical trial methodology. As patients are generally recruited gradually during a trial, information on how earlier patients perform is available before the trial ends. Adaptive designs allow changes to the design of the trial to be made based on this information. Changes could include stopping treatments that are performing poorly, or stopping the trial early if it is clear one of the treatments is best.

In the trial of 18Fluorine-labelled sodium fluoride, the adaptive design would allow the trial to stop after ten patients had been assessed using all three scanning techniques. This would happen if 18Fluorine-labelled sodium fluoride was overwhelmingly better or worse than the two other techniques. This allows substantial reductions in the average cost and number of patients required by the trial. Dr Ferdia Gallagher, from the University of Cambridge School of Clinical Medicine and the Principal Investigator of the trial said:

“Imaging is a very powerful method for non-invasively assessing the spatial distribution of tumour within patients, but is generally expensive and time consuming to undertake as part of a clinical trial. By working with Dr James Wason and the team in the MRC Biostatistics Unit, we have applied an adaptive trial design to a study which has explored the use of a novel imaging test in renal cancer. Through a two-stage approach, we have been able to derive statistically significant results from a relatively small cohort of patients. The results from this trial will have important implications for the design of our future imaging studies.”

Dr James Wason, Senior Investigator Statistician in the Adaptive Clinical Trials programme, led by Dr Adrian Mander at the MRC Biostatistics Unit, and Trial Statistician also said:

“Adaptive designs are an exciting development that are potentially useful in a wide range of clinical trials. The MRC has shown a great commitment to the development and use of these methods. The Biostatistics Unit has a growing portfolio of adaptive designs in development and in use.

We are always very eager to help clinicians to use them in real trials. I am delighted that in this case the adaptive design allowed the trial to finish early with a statistically convincing conclusion that the new scanning technique was highly effective.”

James Wason article image


The adaptive design has subsequently been mentioned in Nature Reviews Clinical Oncology 'Research Highlight'

Reference to publication:

Gerety EL, Lawrence EM, Wason J, Yan H, Hilborne S, Buscombe J, Cheow HK, Shaw AS, Bird N, Fife K, Heard S, Lomas DJ, Matakidou A, Soloviev D, Eisen T, & Gallagher FA. Prospective study evaluating the relative sensitivity of 18F-NaF PET/CT for detecting skeletal metastases from renal cell carcinoma in comparison to multidetector CT and 99mTc-MDP bone scintigraphy, using an adaptive trial design. Annals of Oncology 00: 1–6, 2015