Using biomarkers in early stage development


Session type:

Josep Tabernero
Vall d'Hebron University Hospital, Barcelona, Spain


Current attrition rate of new oncology drugs is very high, this leading not only to an unsustainable budget in the pharmaceutical industry but a sense of great disappointment in the oncology community level. To curtail this rate of attrition, investigation leaders need to make confident decisions as early as possible during the drug development process and to ensure that only those drugs with an optimal safety/efficacy profile move to phase III development and only patients most likely to benefit from the drug are enrolled into the pivotal regulatory trials.

A greater understanding of the biology of cancer coupled with major advances in biotechnology has resulted in the identification of rationally-designed targeted agents. Proof of principle and robust antitumour activity may be most efficiently demonstrated in phase II studies involving patients bearing tumours that are principally driven by aberrations of the specific target or dysregulation of related signal transduction pathways. The hope is that by identifying tumours that are dependent upon the targeted pathway and by demonstrating that the drug modulates the pathway (either abrogating or promoting the signal) it will be possible to identify the right population of patients for the pivotal trials and hence significantly increase the probability of success.

Once a biomarker has been identified, the assay validated and its intended use defined, the challenge becomes how to incorporate the biomarker assay into the drug development program and design of the phase II and phase III development. At this time point several decisions have to be made regarding whether to evaluate the biomarker prospectively or retrospectively, whether to collect tumour or a surrogate tissue, whether to retrieve archival tumour samples from pathology laboratories or obtain recent fresh biopsies, how best to design and power the trial to test both the effect of the drug and the possible predictive value of the biomarker. Surrogate samples, such as blood and serum, while being much more accessible carry more risk because the tissue may be representative of tumour exposure but not of the tumour biology. In this sense, some biomarkers are present from the very early tumorigenic process (like K-Ras mutations in colon cancer) but some others are closely related in time to late-stage changes (like PI3K and p53 mutations and PTEN deletions in colon cancer, c-MET mutations in NSCLC and secondary c-KIT mutations in GIST). The latter examples would definitely favour the acquisition of a recent tumour sample in order to guarantee that the possible findings in the tumour correlate with the current dysregulated situation of the disease. On the other hand, retrospective biomarker analysis has the advantage that patient enrolment is not compromised, the assay does not need to be ready prior to study commencement, and multiple biomarkers can be evaluated. Additionally the quality and volume of tissue in the archival sample may limit the success of a selected biomarker analyses. Therefore, the decision on whether archival tissue –usually coming from the primary tumour resection- or recently fresh biopsied tissue is needed for the consecution and success of the biomarker should be made taking in consideration all the previously mentioned aspects.

The development of biomarkers in early clinical development holds great promise but also creates new challenges. Further actions are needed in order to implement the development of tumour biomarkers in this setting. These include among others the need for greater information among patients, patients’ coalitions and advocate groups, institutional review boards, local Health Administrations, Regulatory Agencies, clinicians, pathologists and other physicians involved in the acquisition of good quality tumour samples. The ultimate goal of this biomarker development process from tumour biopsies will be to facilitate oncology drugs development and to identify which patients are most likely to benefit.