Introduction: Biomarker validation and adaptive clinical trial design


Session type:

Caroline Dive1
1Paterson Institute for Cancer Research, Manchester, UK


It is now the majority view that drug and biomarker development must proceed in tandem, from the point of drug target identification, through preclinical evaluation to clinical trials. Ideally, there would be a candidate predictive biomarker and a good idea of how to measure proof of mechanism and concept in patients before early clinical trials are initiated. The key steps to translate biomarkers from the bench to real clinical utility however remain a significant challenge and are the subject of this session.

How do we make a biomarker really ‘fit for purpose’ in the oncology clinic? Clearly a highly robust and comprehensively validated biomarker assay is a pre-requisite. But will the assay cope with a busy clinical environment, which clinical sample is routinely available, do we know enough baseline variation and signal to noise and is the dynamic range in the patient population to be studied known? What evidence must we have to convince the principle investigator and trial sponsor that this biomarker is worth it? Professor Howard Scher, D. Wayne Calloway Chair in Urologic Oncology, Memorial Sloane Kettering Cancer Center, New York will talk to the these points, describing his extensive experience and highlighting his work on circulating tumour cells in a presentation entitled ‘Fit-for-purpose biomarker validation’.

The implementation of pharmacodynamic (PD) biomarkers in Phase I trials has been/is controversial, with skeptics pointing to the small patient numbers, the ‘all comers’ 3 by 3 design and the likelihood of surrogate tissues rather than serial tumour biopsies in which to assess biomarkers all preventing acquisition of statistically robust data. Certainly this is a challenge and whilst the predominant objective of a Phase I trial must remain definition of a safe dose, the optimists amongst us seek biomarker evidence of a biologically active dose. A Phase I trial design that incorporates judicious enrichment of patient types tailored to the drug in question and expansion of patient cohorts as the dose is escalated that allow PD biomarkers ‘to speak’ as a segue to Phase II trials perhaps provides the key to this ambitious goal. These challenges will be addressed by Professor Josep Tabernero Head, Medical Oncology Department, Vall d’ Hebron University Hospital, Barcelona who will talk on ‘Biomarkers in early drug development’.

As we strive for biomarker directed clinical trials and test the hypothesis that this will expedite our mission of the right drug for the right patient at the right dose, and with validated predictive biomarkers to hand that are fit for purpose, we need to marry expertise in biomarker statistics with expertise in clinical trial design. Our UK expert Lucinda Billingham, Professor of Biostatistics & Director of the Medical Research Council Midland Hub for Trials Methodology Research, Birmingham University, will summarise the key elements to ensure success in her presentation entitled ‘Adaptive trial design that incorporates biomarkers’.

The future is personalised medicine for cancer patients and thus the future is both good drugs and good biomarkers deployed well.