Statistical and practical considerations in designing of immuno-oncology trials.

The novel mechanism of action of immunotherapy agents, in treatment of various types of cancer, poses unique challenges during the designing of clinical trials. It is important to account for possibility of a delayed treatment effect and adjust sample size accordingly. This paper provides an analytical approach for computing sample size in the presence of a delayed effect using a piece-wise proportional hazards model. Failing to account for an anticipated treatment delay may result in considerable loss in power. The overall hazard ratio (HR), which now represents the average HR across the entire treatment period, can remain a meaningful measure of average benefit to patients in the trial. We show that, special consideration needs to be given for the designing of interim analyses related to futility, so as not to increase the probability of incorrectly stopping an effective agent. It is shown that the weighted log-rank test, using the Fleming-Harrington class of weights, can be used as supportive analysis to better reflect the impact of a delayed effect and possible long-term benefit in a subset of the overall population.

Evaluation of dynamic contrast-enhanced MRI biomarkers for stratified cancer medicine: How do permeability and perfusion vary between human tumours?

BACKGROUND: Solid tumours exhibit enhanced vessel permeability and fenestrated endothelium to varying degree, but it is unknown how this varies in patients between and within tumour types. Dynamic contrast-enhanced (DCE) MRI provides a measure of perfusion and permeability, the transfer constant Ktrans, which could be employed for such comparisons in patients. AIM: To test the hypothesis that different tumour types exhibit systematically different Ktrans. MATERIALS AND METHODS: DCE-MRI data were retrieved from 342 solid tumours in 230 patients. These data were from 18 previous studies, each of which had had a different analysis protocol. All data were reanalysed using a standardised workflow using an extended Tofts model. A model of the posterior density of median Ktrans was built assuming a log-normal distribution and fitting a simple Bayesian hierarchical model. RESULTS: 12 histological tumour types were included. In glioma, median Ktrans was 0.016min-1 and for non-glioma tumours, median Ktrans ranged from 0.10 (cervical) to 0.21min-1 (prostate metastatic to bone). The geometric mean (95% CI) across all the non-glioma tumours was 0.15 (0.05, 0.45)min-1. There was insufficient separation between the posterior densities to be able to predict the Ktrans value of a tumour given the tumour type, except that the median Ktrans for gliomas was below 0.05min-1 with 80% probability, and median Ktrans measurements for the remaining tumour types were between 0.05 and 0.4min-1 with 80% probability. CONCLUSION: With the exception of glioma, our hypothesis that different tumour types exhibit different Ktrans was not supported. Studies in which tumour permeability is believed to affect outcome should not simply seek tumour types thought to exhibit high permeability. Instead, Ktrans is an idiopathic parameter, and, where permeability is important, Ktrans should be measured in each tumour to personalise that treatment.