Table Of Content

MOLECULAR DIAGNOSTICS The Key Driver in Personalized Cancer Medicine V024tp.indd 1 1/26/10 2:37:04 PM Copyright © 2010 by Pan Stanford Publishing Pte. Ltd. The Key Driver in Personalized Cancer Medicine MOLECULAR DIAGNOSTICS Editors Jan Trøst Jørgensen Dx-Rx Institute, Denmark Henrik Winther Dako, Denmark V024tp.indd 2 1/26/10 2:37:07 PM Copyright © 2010 by Pan Stanford Publishing Pte. Ltd. Published by Pan Stanford Publishing Pte. Ltd. Penthouse Level, Suntec Tower 3 8 Temasek Boulevard Singapore 038988 Email: INTRODUCTION It is now 10 years ago since Robert Langreth and Michael Waldholz announced the “New Era of Personalized Medicine” in the Wall Street Journal. This article introduced the concept of personalized medicine to a broader audience, but in fact, the first step towards a more individualized drug therapy was taken already several decades before this manifesto. With the discovery of the estrogen receptor in the 1960’s and the introduction of the anti-estrogen tamoxifen in the 1970’s the road was prepared for a more individualized treatment of cancer patients – in this case patients with breast cancer. The assay for estrogen receptor (ER) status became an important stratification factor for anti-estrogen treatment. In the 1990’s another targeted drug was introduced which was also aimed at a selected group of cancer patients: women whose breast cancer tumors over-expressed the human epidermal growth factor receptor 2 (HER2). The drug was the monoclonal antibody trastuzumab ® (Herceptin , Genentech, CA, USA), which was specifically targeted towards the HER2 protein of the tumor cells. Also, for trastuzumab a pharmacodiagnostic test played an important role in relation to treatment stratification. This was the immunohistochemical assay for the HER2 ™ protein (HercepTest , Dako, Glostrup, Denmark). Since the turn of the century other targeted cancer drugs have been introduced which are guided by a pharmacodiagnostic test or companion diagnostics in order to identify the patients who are most likely to respond to treatment. The development of pharmacodiagnostic tests are, however, not restricted to new drugs only. As our knowledge about the cancer pathophysiology at the molecular level increases and the mechanisms of action of the drugs are explained, it has also become possible to develop pharmacodiagnostic tests for drugs that are already used in the clinic. One recent example is the predictive fluorescence v Copyright © 2010 by Pan Stanford Publishing Pte. Ltd. vi J. T. Jørgensen & H. Winther in situ hybridization (FISH) assay for anthracycline treatment of patients ™ with primary breast cancer (TOP2A FISH pharmDx , Dako Glostrup, Denmark). In fact, the cover of this book is embellished with a picture of breast cancer cells showing amplification of the TOP2A gene assessed by this assay. The advancement within molecular diagnostics, especially genomics, is a very important driver of this development. In a recent report from the US; “Priorities for Personalized Medicine”, published by the President’s Council of Advisors on Science and Technology (PCAST), it is stated that molecular diagnostics is identified as a key drivers of personalized medicine. In this respect pharmacodiagnostics or companion diagnostics are not only important tools used in the development of new targeted anticancer drugs, but they are just as important with respect to improvement of the quality of patient care. The PCAST report points at the potential of personalized medicine in relation to two important trends - the increasing costs of healthcare and the decreasing rate of new medical products being developed. For the quality of patient care the use of diagnostics will enable the treating physician to distinguish in advance those patients who will benefit from a given treatment and those who are likely to suffer the adverse effects only. The use of pharmacodiagnostics or companion diagnostics will lead to a more rational treatment of cancer that could also result in cost savings for the healthcare system. Moreover, with respect to drug development, diagnostics can be used to stratify patients according to their likelihood of response to the drug under development. Such approach could result in reduced size, duration, and costs of clinical trials, thus facilitating the development of more new drugs. Drug-diagnostic co-development will be the future development model in stratified and personalized cancer medicine. We will see more and more drugs and diagnostics being developed in parallel in order to increase the effectiveness and safety of new anticancer drugs, and it is our hope that this book will contribute to this development and be an inspiration to those who are engaged in this important work. Copyright © 2010 by Pan Stanford Publishing Pte. Ltd. Introduction vii Working with this book has been a pleasure and has convinced us even more that molecular diagnostics will be a decisive factor with respect to further improvement of cancer treatment. We would like to express our gratitude to all the authors who have contributed so positively to this book. Jan Trøst Jørgensen and Henrik Winther Copenhagen, November 2009 Copyright © 2010 by Pan Stanford Publishing Pte. Ltd. CONTENTS Introduction v Jan Trøst Jørgensen and Henrik Winther Developing Molecular Diagnostics in Cancer 1 Henrik Winther and Jan Trøst Jørgensen The Estrogen and Progesterone Receptors — Setting the Scene for Pharmacodiagnostics 21 Kwok-Leung Cheung The Development of the HercepTest™ — From Bench to Bedside 43 Jan Trøst Jørgensen and Henrik Winther Predictive Biomarkers and Translational Cancer Medicine 61 Anne-Sofie Schrohl, Pernille Bræmer Hertel, Maj-Britt Jensen and Nils Brünner Pharmacodiagnostic Testing — Immunohistology 87 Anthony S.-Y. Leong and Trishe Y.-M. Leong In Situ Hybridisation for Pharmacodiagnostic Testing 119 John M.S. Bartlett and Fiona M. Campbell ® MammaPrint Translating Research into a Diagnostic Test 139 Annuska M. Glas, Leonie Delahaye and Oscar Krijgsman DNA Biomarkers in the Diagnosis and Management of Cancer 165 Reza Rafiolsadat Serizawa and Per Guldberg Validating the Analytical Power and Parameters of an Immunohistochemical Test 185 Søren Nielsen ix .iex Copyright © 2010 by Pan Stanford Publishing Pte. Ltd. x Contents Drug and Pharmacodiagnostic Co-Development: Statistical Considerations 207 Richard Simon Clinical Validation of Biomarkers in Cancer 227 Sumithra J. Mandrekar and Daniel J. Sargent Pharmacogenomic Markers for Cancer Therapy 251 Sharon Marsh Safety and Effective Biomakers in Oncology — A Regulatory Drug and Device Perspective 275 Atiqur Rahman and Francis Kalush Personalizing Treatment Selection for Breast Cancer 297 Lajos Pusztai and Chit Cheng Yeoh Pharmacodiagnostics and Personalized Medicine in Cancer 325 Jan Trøst Jørgensen Copyright © 2010 by Pan Stanford Publishing Pte. Ltd. CHAPTER 1 DEVELOPING MOLECULAR DIAGNOSTICS IN CANCER Henrik Winther ImmunoHistology R&D Business Area, Dako A/S Produktionsvej 42, DK-2600 Glostrup, Denmark E-mail: 2 H. Winther & J. T. Jørgensen housekeeping proteins (internal tissue controls) will indeed increase the standardisation and robustness levels and thereby optimise the IHC assay quality and patient care. 1. Introduction When discussing the development of molecular diagnostics, the obvious starting point would be a clear definition of biomarkers and their indicator classification – e.g. classification into diagnostics, prognostics and predictive biomarkers. A biomarker is to be considered a substance used as an indicator of a biologic state. It is a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic 1 processes, or pharmacologic responses to a therapeutic intervention . Within this chapter, the biomarker substances will be dealt with as antibodies. In literature, there is a confusion of terms with regard to the indicator classification of biomarkers and hence we will briefly take you through our definition of a diagnostic-, prognostic- and predictive biomarker (Figure 1). A diagnostic biomarker will be considered to be an antibody that is presently used in a routine setting in routine diagnostic pathology laboratories. The antibody is used as a confirmatory test and is therefore an add-on to the pathologist in confirming the diagnosis initially based on the patient clinical history and tissue biopsy morphological (macroscopic and microscopic) observations. Classification and sub- typing of tumours, e.g. lymphomas, is typically performed by using a 2 diagnostic biomarker . The majority of diagnostic biomarkers are classified as IVD class I products when considered from a regulatory point of view. IVD class I products can be defined as “confirmatory products”. A prognostic biomarker will be characterised as an antibody being able to provide additional information to the pathologist – e.g. it is not just a confirmatory test, but a test that will guide the pathologist towards a therapeutic rational. In its strict sense a prognostic biomarker could be Copyright © 2010 by Pan Stanford Publishing Pte. Ltd.