Companion Diagnostics: From Biomarker Identification to Market Entry

Abstracts

Day 1 — Monday, April 28, 2014

Session I: Paving the Way to Personalized Medicine

Genomics Medicine
Raju Kucherlapati, PhD, Departments of Genetics and Medicine, Harvard Medical School, Boston, Massachusetts, United States

The use of genetic and genomic information for diagnosis, prognosis, and helping treatment decisions in many areas of medicine is undergoing a revolution. Some of these advances are made possible through increasingly rapid and relatively inexpensive DNA and RNA sequencing technologies, and others are fueled by an increase in the knowledge about the etiology of human disease. Among the areas that are impacted heavily are risk assessment, prenatal diagnosis, genetic basis for several childhood disorders, as well as prognosis and treatment decisions in cancer. The ability to obtain fetal DNA from maternal circulation is allowing a simple and non-invasive method to detect fetal abnormalities. The availability of whole exome sequencing is now permitting identification of the genetic basis for several childhood disorders whose etiology is not well understood. In the case of cancer, examination of DNA from tumor samples permits stratification of the patients who can be treated with a therapeutic approach that is tailored to the tumor’s genetic profile. In some cases, the patients can be treated with an approved drug; in other cases, the patient can be enrolled in an appropriate clinical trial or be provided with standard chemotherapeutic approaches. The identification and translation of biomarkers is altering the way we diagnose and treat human disease.
 

Functional Proteomics, Companion Diagnostics, and Precision Medicine
Peter J. Parker, PhD, King’s College London and London Research Institute of Cancer Research UK, London, United Kingdom

Pharmacodynamic and stratification biomarkers are increasingly important in the development and deployment of targeted therapeutics. In cancer, many targets are self selected through somatic or epigenetic changes and mature genomic technologies provide instructive biomarkers in these contexts both to include and to exclude patients. In other situations, targets may be hyper- or hypo-functional in the absence of mutation or altered expression; in this context, we need to understand the functional state of targets. Despite this need, our ability to robustly assess such functional states in patient biopsies is immature, with limited specificity and sensitivity. One of the prominent classes of targets pertinent to many therapeutic areas is protein kinases. The status of a target kinase is a tractable problem but limited by the technologies of detection in patient samples. Here, a generic methodology will be discussed that can be exploited in kinase biomarker development, alongside an assessment of technologies applicable to delivering functional biomarkers in patients.
 

Insights for Personalized Medicine from the Gut Microbiome
S. Dusko Ehrlich, PhD^1,2

Gut microbial biomarkers can be used for diagnostics, patient follow-up, and prediction of risk and response to a treatment. We have found that one out four people has lost a substantial portion of gut microbial richness (Le Chatelier et al, Nature, 500, p.541-546, 2013). The loss is associated with altered metabolic parameters, which place microbe-poor individuals are at a higher risk to develop metabolic syndrome associated pathologies, such as diabetes or cardiovascular diseases. Loss of gut microbial richness can be detected very accurately, by analysis of only 6 bacterial species. A simple diagnostic test to detect people at risk is currently being developed. Importantly, the loss of richness can be corrected, at least in part, by a nutritional intervention, in parallel with improvement of the metabolic parameters (Cotillard et al, Nature, 500, p585-588, 2013). However, the improvement of metabolic parameters is less effective in microbe-poor individuals, illustrating the potential of responder/non-responder stratification by gut microbial biomarkers.
 
¹INRA, Metagenopolis, Jouy en Josas, France
²Human Microbiome Centre, King’s College, London, United Kingdom
 

Session II: Case Studies — Lessons from the Past and Pioneers of the Future

Case Study: Xalkori Companion Diagnostic
Hakan Sakul, PhD, Pfizer Inc

Industry Spotlight — Development of Selzentry (Maraviroc) and the Companion Diagnostic HIV Tropism Assay
Charles Knirsch, MD, MPH, Pfizer Inc

Development of Selzentry (Maraviroc) and the Companion Diagnostic HIV Tropism Assay
Jayvant Heera, MD and Charles Knirsch, MD, MPH, Pfizer Inc, Pearl River, New York, United States

The emergence of a new infectious disease associated with a new organism often leads to co-development of diagnostic tests with a search for new therapeutics and eventually their subsequent discovery and development. In 1996, the observation was made that resistance to HIV-1 infection occurred in individuals bearing mutant alleles of the CCR-5 chemokine receptor gene. Structure-based design of a CCR5 inhibitor led to the discovery and subsequent development of maraviroc. At the same time, a personalized medicine approach to development of a companion diagnostic was employed to assure that patients would be selected who benefited most — those with CCR5 tropic virus. Over time, the performance characteristics of phenotypic diagnostic tests were enhanced and impacted the clinical development of maraviroc. In addition, genotypic tests were developed that have the potential to allow for local use and rapid turnaround times as compared to the earlier diagnostic tests.
 

Development of Kalydeco as a Genotype Specific Drug for the Treatment of Cystic Fibrosis
Federico Goodsaid, PhD, Vertex Pharmaceuticals, Washington, District of Columbia, United States

There are multiple failure mechanisms for the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR). One of these mechanisms is gating failure due to CFTR gene mutations that prevent the CFTR protein product from opening or working (gating) properly on the cell surface. The most common example of a gating mutation is G551D. Kalydeco was developed to potentiate the gating activity of CFTR in gating mutation and other CF patient populations. There are 30,000 CF patients in the US. A fraction of these will respond to Kalydeco therapy. The scientific, technological, commercial, and regulatory issues associated with the approval of Kalydeco and the identification of CF patients who will benefit from this drug is an iconic example for personalized healthcare product development. In this presentation, we will look at how these issues have been successfully addressed for Kalydeco.
 

Pathways of Disease Progression in Heart Failure: What Can Galectin-3 Tell Us?
Aram S. Adourian, PhD, BG Medicine Inc., Waltham, Massachusetts, United States

We will review the development and recent regulatory clearance path of the galectin-3 blood test, the first novel In vitro diagnostics (IVD) assay for use in chronic heart failure patients since the clinical introduction of Brain natriuretic peptide (BNP) and the N-terminal portion of proBNP (NT-proBNP). Galectin-3 is an evolutionarily highly conserved protein that appears to be a key regulatory molecule at the nexus of acute inflammation, chronic inflammation, and tissue fibrogenesis. As such, in addition to its utility as a diagnostic biomarker in cardiovascular disease, galectin-3 and its associated biochemical pathways are also currently being explored as potential therapeutic targets, both in heart failure and in other fibrotic disorders. In this presentation, we will review both the diagnostic and the potential therapeutic implications of circulating galectin-3 levels for disease and patient stratification.
 
This presentation will include discussion of unlabeled uses of approved/cleared products.
 

Session III: Concurrent Workshops — The Next Frontier(s) of Therapeutic Target Areas

Concurrent Workshop A: Infection, Inflammation, and Airway Disease

Biomarkers for Precision Medicine in Airway Diseases
Tariq Sethi, BSc, MA, PhD, FRCP, King’s College London, London, United Kingdom

Chronic obstructive pulmonary disease (COPD) is mainly defined by airflow limitation that is not fully reversible. Exacerbations and overlap in pathophysiology and symptoms with other chronic airways diseases leads to challenges in differential diagnosis and prediction of outcome. Validated biomarkers in COPD are essential for risk assessment, phenotype identification, and diagnosis of exacerbations. Lung function alone cannot identify the multiple phenotypes of COPD. In targeted populations, fibrinogen and C-reactive protein and other markers of systemic inflammation have been associated with a future risk of developing COPD, risk of accelerated lung function decline, and acute exacerbations and hospitalisation. Fibrinogen is likely to be the first COPD biomarker presented to the FDA for qualification in the drug approval process. Innate defence molecules in serum, which are generally secreted by the epithelium or innate immune cells such as neutrophils, were amongst the first biomarkers of COPD identified; examples include the alpha- and beta-defensins, Pulmonary and Activation Regulated Chemokine, Clara Cells secretory protein-16, and surfactant proteins A and D. Sputum neutrophil counts may be a biomarker of therapeutic response in COPD. Integrative indices such as the BODE index may predict prognosis and response to therapy. Computed Tomography (CT) scan can determine the relative extent of small airways disease or emphysema which may predict prognosis and treatment responses. Biomarker identification in COPD is a developing field and no single biomarker in COPD exists reflecting the challenges of biomarker development in a complex disease. There is urgent need for future biomarker studies in large scale, longitudinal, well phenotyped cohorts.
 

Stratification of Rheumatoid Arthritis
Iain McInnes, FRCP, PhD, FRSE, University of Glasgow, United Kingdom

Rheumatoid arthritis is a chronic inflammatory disease of increasingly clear autoimmune pathogenesis. The last decade has seen remarkable advances in the available therapeutics and also parallel evidence of benefit when appropriate therapeutic strategies are employed, regardless of chosen therapeutic agent. Unmet clinical need remains. Remission that is consistent and stable over time is rarely achieved. Even with the best therapeutics, only proportionate response rates are observed across the range of patient populations treated. Moreover, there are significant toxic events still reported in a proportion of patients. As such, there are now studies ongoing that seek to delineate a molecular taxonomy for the disease which will allow appropriate patient selection. Thus, a variety of approaches, including genomic, transcriptomic, proteomic, and latterly metabolomic methodologies have been applied to create subsets of disease which may be clinically meaningful. In this presentation, I shall overview the key strategies being developed and highlight exemplar successes thus far in the field.
 

mTOR Signaling: a Central Pathway to the Pathogenesis of Multiple Inflammatory and Immune System Disorders — Implications for Personalized Medicine
Andras Perl, MD, PhD, State University of New York, Division of Rheumatology, Departments of Medicine and Microbiology and Immunology, Biochemistry and Molecular Biology, State University of New York, Upstate Medical University, College of Medicine, Syracuse, New York, United States

The mechanistic or mammalian target of rapamycin (mTOR) is a large serine/threonine kinase that has been recently recognized as a central sensor and cell type-specific regulator of metabolic pathways. mTOR is part of two protein complexes termed mTOR complex 1 (mTORC1) and 2 (mTORC2). Within the immune system, mTORC1 and mTORC2 differentially regulate lineage specification within the CD4+ T cell compartment, such as T helper (Th) subsets that promotes immune responses to distinct infectious organism and regulatory T cells (Tregs) that in turn prevent autoimmune diseases. mTORC1 generally promotes the development of Th1 and Th17 cells, which produce interleukin-2, interferon gamma, or tumor necrosis factor alpha and interleukin-17, respectively. In turn, mTORC2 supports the differentiation of Th2 cells, which promote the production of interleukin-4 and interleukin-10. In addition to regulating immune responses to infectious organisms, mTOR activation has been documented in autoimmune diseases, such as lupus and multiple sclerosis. Blockade of the mTOR pathway at various levels has shown clinical efficacy in lupus and multiple sclerosis. Activation of mTORC1 also predisposes to inherited forms of hyperproliferative diseases, such as Lymphangioleiomyomatosis and tuberous sclerosis (TSC). Both of these diseases are caused by mutations in the tuberous sclerosis genes TSC1 or TSC2, which negatively regulate mTOR. Moreover, mTOR activation has emerged as key survival factor in cancer cells. While rapamycin was originally developed to block transplant rejection via inhibiting CD4+T-cells, new analogues of rapamycin (rapalogues) have been purposefully designed for treatment of cancer. Interestingly, rapamycin also expands lifespan both in unicellular and mammalian organisms. Therefore, selective targeting of the mTOR pathway may be uniquely beneficial in safely controlling major pathways of disease pathogenesis.
 

Companion Diagnostics in Autoimmune Disorders: Improving Outcomes in RA and IBD
Mark E. Curran PhD, Janssen Research and Development, LLC, Springhouse, PA USA

Rheumatoid Arthritis and Inflammatory Bowel Disease are severe immune diseases with significantly reduced quality of life for patients. Despite advances in treatment for moderate and severely affected patients through evolution of antibody and recombinant protein based therapeutics, there remains a significant unmet clinical need for new therapies, companion diagnostics and integrated treatment solutions. We are focused on transforming therapy in these diseases by applying Systems Pharmacology, Precision Medicine principles, integrated solutions and developing Companion Diagnostics to create novel and more effective treatment paradigms. Our objective is to providing for higher response rates, deeper remission, early interception and eventually prevention of these life altering diseases.  Recent progress toward meeting these objectives will be discussed.
 

Concurrent Workshop B: Metabolic and Cardiovascular Diseases

Implications for Personalized Medicine
Julie A. Johnson, PharmD, FCCP, BCPS, University of Florida, United States

The use of genetic information to predict cardiovascular disease risk and drug response has been the focus of much research in the past decade. While there has been success in identifying genetic markers for multiple cardiovascular diseases and related phenotypes, these collectively still explain only a small percent of disease risk (often < 5%) and so translation to practice has been limited. A promising new drug class for treatment of elevated LDL cholesterol was identified through genetic studies. Commonly used cardiovascular drugs, including clopidogrel, warfarin, and statins have genetic markers that are being used to guide drug selection or dose in some clinical settings and research is active in other a areas, e.g., antihypertensive drugs. This presentation will focus on these various advances.
 

Advances in Blood Pressure Genomics
Mark J. Caulfield, MD, FRCP, William Harvey Research Institute, Barts National Institute for Health Research Biomedical Research Unit, Queen Mary University of London, United Kingdom

The identification of genes implicated in Mendelian forms of hypertension demonstrates that rare variants with substantial effects on blood pressure do exist. Often, these genes lie within pathways managing sodium homeostasis. More recently, advances in affordable high-throughput genotyping strategies have enabled detection of multiple common genetic variants with modest effects on blood pressure at the level of the population. In aggregate these common variants explain less than 3% of the variance of blood pressure. These findings may offer new mechanistic insights into the biology of blood pressure. It is timely to reflect on recent advances in genomics, and the utility of new resources such as the 1000 Genomes Project and the Encyclopaedia of DNA Elements to annotate likely causal variants, and their relevance to cardiovascular disease. A key question is whether these gene discoveries might influence cardiovascular risk assessment, help to stratify patient response to medicine, or identify new biologic pathways for novel therapeutic targets.
 

Personalized Diabetes Medicine
Toni I. Pollin, MS, PhD, University of Maryland School of Medicine, Baltimore, MD, USA

Diabetes mellitus affects over 25 million individuals in the United States and is a leading cause of morbidity and mortality. Most of these individuals have type 1 diabetes (T1DM) or type 2 diabetes (T2DM).  Over 40 and 65 genetic loci influencing susceptibility to T1DM and T2DM, respectively, have been identified.  While currently there is no clear clinical utility to testing for variations in these genes because of the small total amount of susceptibility explained, identification of these genes through candidate gene and genome-wide association studies has provided a foundation for elucidating the etiology to pave the way for improved treatment and prevention.
 
In contrast, there is clear clinical utility to genetic testing for monogenic diabetes, which accounts for at least 1% of diabetes and results from mutations in several genes including HNF1A, GCK, HNF4A, KCNJ11. Diagnosing monogenic diabetes enables personalized treatment, resulting in improved glucose control, better prediction of prognosis, and an enhanced familial risk assessment.  However, a recently population-based study revealed that 94% of children with monogenic diabetes were misdiagnosed, mostly with T1DM or T2DM, and accordingly usually receiving the wrong treatment.  To address this gap, we are implementing, evaluating and disseminating a multi-pronged approach which includes provider eduction, efficient screening all diabetic individuals for “red flags” suggestive of monogenic diabetes, molecular diagnosis using targeted next-generation DNA sequencing, and personalized treatment (e.g., sulfonylureas rather than insulin for individuals who have HNF1A mutations) and genetic counseling for affected individuals and their families.
 

A Novel Fasting Blood Test for Insulin Resistance: Application of Metabolomics to Diagnostic Development
Jeff E. Cobb, PhD, Metabolon, Inc.,Durham, NC, US

There is an unmet need for a simple measure of insulin resistance (IR). Recently we reported that non-biased global metabolomic profiling had identified a number of metabolites that correlated with insulin resistance as measured by the hyperinsulinemic euglycemic clamp, the gold standard for measuring insulin sensitivity. These metabolites were used to develop a novel test for IR. Data from the RISC Study (a healthy non-diabetic cohort, n=1277) were used in an iterative process of algorithm development to define the best combination of metabolites for predicting the clamp’s M value of insulin sensitivity. The process converged to a model consisting of a multiple linear regression (natural log transformed) on the fasting levels of a-hydroxybutyrate, linoleoylglycerophosphocholine, oleate and insulin used to calculate ln(M). The resulting score, MQ, correlated with M having an r value of 0.66 and identified subjects with insulin resistance with an AUC of 0.79, outperforming other simple measures such as fasting insulin, fasting glucose, HOMA-IR or BMI.   In addition, in the ACT NOW diabetes prevention study, MQ was shown to increase significantly with pioglitazone treatment in IGT subjects and the increase was greatest in subjects improving to NGT status. In summary, this simple IR test, Quantose IRTM, based on a single fasting blood sample, may have clinical utility for identifying subjects with IR and in monitoring therapeutic intervention in insulin resistant subjects.
 

Concurrent Workshop C: Neurodegenerative Disease

The Genetics of Multiple Sclerosis — Implications for Personalized Medicine
Philip Lawrence De Jager, MD, Harvard Medical School

Biomarkers for Parkinson's Disease—Imaging Onset, Progression, and Effect of Treatment
Kenneth Marek, MD, Institute for Neurodegenerative Disorders (IND)

During the past two decades, clinical, genetic, biochemical, and imaging biomarkers have guided and accelerated research to elucidate Parkinson Disease (PD) pathobiology and assess potential PD therapeutics. These biomarkers have enabled research to pose and to inform essential question including: “When does PD begin?”, “Where does PD begin?”, “How does PD progress?”, and “What are rational PD therapeutic targets?”. Subtypes of PD have been identified by molecular genetic biomarker signatures offering the potential of targeted therapies. Imaging biomarkers have emerged as critical tools to further define PD subsets, accurately identify research subjects with early PD pathology, monitor disease progression, and assess biomarker progression prior to the onset of symptoms during a prodromal phase of PD.
 
Both advances in imaging technology and an explosion in the availability of imaging probes have enabled Positron Emission Tomography (PET) and Single-Photon Emission Computerized Tomography (SPECT) to inform PD drug development decisions. Several imaging ligands target the loss of nigrostriatal dopaminergic function in PD. Studies have demonstrated that imaging outcomes may predict disease progression and suggest that a growing toolbox of ligands probing striatal function may identify sub-groups likely to develop PD clinical milestones such as dyskinesia or cognitive impairment and those subjects likely to respond to specific PD therapies.   Imaging tracers targeting the underlying pathology of PD such as alpha-synuclein will ultimately provide further insight into disease onset and progression. Several studies are further exploring PD biomarkers including the Parkinson Progression Marker Initiative (PPMI), an international, multi-center, observational longitudinal study designed to identify and validate PD markers and PD subsets for further therapeutic studies.
 

Identification of Biomarkers for the Diagnosis, Prediction, and Progression of Alzheimer's Disease
Michelle M. Mielke, PhD, Departments of Epidemiology and Neurology, Mayo Clinic, Rochester, Minnesota, United States

Alzheimer’s disease (AD) is the most prevalent form of dementia, currently affecting more than 5 million individuals in the United States. This number will continue to rise with the aging of the population. The hallmark pathologies associated with AD (i.e., amyloid-beta plaques, neurofibrillary tangles, and neurodegeneration) begin decades prior to the emergence of clinical symptoms. This lag time, between initial pathology and symptoms, provides a ‘window of opportunity’ to potentially prevent or delay the clinical onset of the disease. Therefore, as the emphasis in the field shifts to the earlier detection of AD, including the preclinical phase, biomarkers have become central for the diagnosis, prediction, and monitoring of progression of AD. The enhanced developments of in vivo neuroimaging measures, including the newest tau tracers, provide an unprecedented opportunity to further our understanding of disease mechanisms in humans. Instead of utilizing clinical phenotypes as the outcome, the field can now go one level deeper to understand the mechanistic pathways associated with the pathology. An example of blood and CSF sphingolipids will be presented to demonstrate the capabilities of using these pathological outcomes to further our understanding of the pathways involved in the development and progression of AD, and to enable us to identify new biomarkers and treatment targets.
 

Hot Topic
Speaker selected from submitted abstracts