Innovative clinical trials for development of personalized cancer medicine

The cancer burden is increasing globally due in large part to the ageing of the population. A major challenge for healthcare systems is an increasing population of cancer patients who are not cured but are living longer with their cancer, making cancer into one of the major chronic diseases and one that carry a high economic burden due to the high cost of many cancer therapies. Innovative approaches to cancer prevention will be able to decrease some of these challenges. Developing better treatment strategies for disseminated disease remains one of the main goals to modify the growing patient population with chronic cancer disease. This means that prolonged survival during treatment must be converted to an improved cure rate.

The rapid development of knowledge in cancer biology continually changes the challenges for therapeutic research. There is a shift from selection of therapies based on the histologic origins of tumours to alterations of the cancer genome as targets for more specific anticancer drugs, molecularly targeted agents, and immunotherapies. The large number of genetic and epigenetic alterations driving tumour cells creates a number of subsets within each histologic tumour type. This results in a large inter‐patient variation, and in the end there may be thousands of tumour types with specific molecular backgrounds requiring unique treatments. Further, genomic instability makes the problem still more complex, producing tumour cell heterogeneity and intra‐patient variation. When considering combinations of anticancer treatments for the heterogeneous population of tumours the number of treatment possibilities is enormous, more than existing cancer patients. New strategies for therapeutic research are therefore urgently needed to enable the most efficient process of drug development.

Personalized or precision cancer medicine aims at “the right treatment for the right patient at the right time”. This means individualized treatment with the optimal anticancer agent(s) delivered with optimal dose and scheduling. To achieve this, biomarker research is fundamental. We need to know who should be treated (early detection of disease), prognosis, and how to predict likely treatment outcomes in terms of both anti‐tumour activity and side‐effects. Biomarker translational research involves biomarker discovery and analytical and clinical validation, as well as assessment of clinical utility. Biological cancer research is supporting clinical trials with large numbers of potentially valuable biomarkers. The connection, however, between discovery and clinical validation of biomarkers is still at an early stage. Currently there are few molecular diagnostic tests recommended by major regulatory and advisory bodies. Reliable methods for assessment of biomarkers for clinical use are discussed in this issue of Molecular Oncology.

Conceptual developments based on increasing knowledge of cancer biology suggest changes in the methodology for clinical trials. This issue aims at highlighting some of the problems and challenges for clinical trials methodology when moving into personalized cancer medicine. The traditional strategy involving phase I–III trials with determination of efficacy based on the results of large comparative clinical studies is no longer possible in many situations. Traditional trials focus on large patient populations, while identification of many subgroups of cancer patients results in the need for smaller and more focused clinical trials with preselected patients based on the biomarkers in their cancers. Histology‐based trials with enrichment for patients with specific molecular alterations represent a step forward. New trial methodologies are being developed such as basket, platform, “N‐of‐1”, and adaptive trials. More informative ways to describe individual patient tumour responses are under development like waterfall, spider and swim plots. Understanding of mechanisms behind inter‐ and intra‐patient heterogeneity, including inherited and acquired drug resistance, is fundamental for future biological research and the design of a new generation of clinical trials.

Drug development needs to be better coordinated with biomarker research. Omics technologies with bioinformatics support must deliver data in real time for treatment. Technologies to stratify tumours and patients must be validated and reporting language must be harmonised, so that data from different centres can be compiled and compared. Biological changes in the tumours during the course of treatment must be followed with repeated biopsies and treatment adapted based on the evolving genomic landscape of the tumour. Technologies for liquid (blood) biopsies will expand. Imaging technologies will increase in importance, not only to describe the extent and heterogeneity of the disease, but also to assess early response after treatment, fluctuations of expression of biomarkers within the tumours, as well as target saturation.

Improving dose finding in early phase clinical trials is a specific problem. Molecularly targeted agents and immunotherapies may have toxicities which often are not dose dependent. Pharmacokinetic and pharmacodynamic outcomes together with functional imaging are being explored to obtain the optimal biological dose and schedule of treatment, i.e., continuous or intermittent. Animal models (genetically engineered mouse models and patient derived xenografts) are important sources of biological information when designing early phase clinical trials. Mouse trials performed in parallel with human trials, co‐clinical studies, are under development in order to identify the relevant biological information needed for early clinical trials. A separate issue on animal models in cancer research was recently published in this journal (Berns and Barbacid, 2013).

Comparative (randomized) clinical trials will be more difficult to conduct and observational (not randomized) studies and adaptive trials will play an increasing role in determining the efficacy of new treatments. Assessments of clinical effectiveness and cost‐effectiveness are today unmet needs, and the gap between clinical efficacy, which is the outcome from a clinical trial, and clinical effectiveness, which is the outcome when a total population of patients is treated, must be bridged. Documentation of antitumour response will need fairly small patient groups, while assessment of side‐effects will need larger groups of patients, more heterogeneous populations and long‐term follow‐up. It is important that clinical research and clinical practice become better integrated. The Patient Generated Health Data initiative offers a possibility to collect “big data” for observational studies at the population‐level. A new generation clinical trials with companion diagnostics will require addressing and modernizing regulatory policies to keep pace with scientific innovation. The need for agreement between academic centres, the pharmaceutical industry and regulatory authorities regarding evidence and assessment of the balance of benefit‐risk must be addressed.

For all of these reasons, clinical trials methodology, when moving into personalized cancer medicine, is a complex and expanding area. This issue of Molecular Oncology addresses some of the important areas under development. We have added a chapter on cancer prevention which is developing towards identification of high risk individuals and treatment of stratified high risk groups, based on benefit‐risk assessments that can present new challenges. Recently there have been separate articles in this journal about targeted drugs and resistance mechanisms (Groenendijk and Bernards, 2014) and histology agnostic clinical trials (Lacombe et al., 2014). These publications are adding further information and reading is recommended.

We hope that this issue will achieve its educational mission and stimulate interest in further development of clinical trials methodology. Success will require new types of collaborations between clinical cancer researchers and the biotechnology/pharmaceutical industry. The next generation of clinical trials will need these collaborations in order to reach the critical mass of patients required for trials on small, biomarker‐selected subpopulations. The initiative for this issue of Molecular Oncology came from the EU‐funded EurocanPlatform project, which initiated collaborations between cancer research centres in translational cancer research. The newly formed Cancer Core Europe, a consortium of six European centres, and the WIN consortium, a worldwide clinical research collaboration, are examples of joint efforts aimed at reaching a critical mass of patients required for innovative clinical trials in personalized cancer medicine.

Mendelsohn John, Ringborg Ulrik, Schilsky Richard, (2015), Innovative clinical trials for development of personalized cancer medicine, Molecular Oncology, 9, doi: 10.1016/j.molonc.2015.02.013.