IC‐PL3: Pharma as a consumer: Biomarker applications for developing new Alzheimer's disease treatments

SATURDAY, JULY 26, 2008 ALZHEIMER’S IMAGING CONSORTIUM PRESENTATIONS PLENARY IC-PL IC-PL1 INTEGRATING NEUROIMAGING, GENOMICS AND COGNITION IN ALZHEIMER’S DISEASE RESEARCH Andrew J. Saykin, Indiana University School of Medicine, Indianapolis, IN, USA. Contact e-mail: asaykin@iupui.edu Background: New transdisciplinary conceptual and analytical approaches are needed to realize the unprecedented opportunities inherent in combining neuroimaging and genomics data in the study of Alzheimer’s disease (AD). Methods: Advances in structural, functional and molecular brain imaging using MRI and PET are providing novel insights into the risk for AD, as well as disease pathophysiology and progression. Parallel advances in molecular genetics permit analysis of up to a million single nucleotide polymorphisms (SNPs) on a microarray “chip” that can assess an estimated 90% of the meaningful variance in the human genome. Results: To date, most research combining neuroimaging and genetics to study AD has concentrated on the role of a single important candidate gene, APOE. Recent studies in healthy individuals and other clinical populations have examined several candidate genes for cognition such as COMT and BDNF. Genome-wide approaches to relating brain structure and function and genetic variation are now possible. An exciting development is that DNA from the ADNI cohort is being analyzed using an array with over 550K SNPs in preparation for examining cross-sectional and longitudinal MRI and PET endophenotypes and an update will be provided. Morphometric and fMRI findings, from a cohort including patients with MCI and older adults with cognitive complaints but no deficits, using a hybrid approach based on targeted candidate pathways will also be presented. An array was designed by a collaborative multidisciplinary team that incorporates molecular pathways related to neurocognition and mechanisms implicated in CNS disorders (e.g., neurodegeneration, neuroinflammation, neurotransmission and receptors, plasticity/repair and growth factors, and candidate genes associated with cognition and memory/LTP). Conclusions: Despite numerous computational challenges such as doubly complex data dimensionality, gene/gene and gene/environment interactions, modest statistical power, risk of false discovery, and the need for better informatics tools to facilitate interpretation of results in biological systems and pathways, the thoughtful integration of imaging and genomics holds great promise for an improved mechanistic understanding of AD and its progression that can help drive the development of better diagnostic and treatment approaches. IC-PL2 AMYLOID IMAGING: DEVELOPMENTS, CHALLENGES, FUTURE DIRECTION Chester A. Mathis, University of Pittsburgh, Pittsburgh, PA, USA. Contact e-mail: mathisca@upmc.edu Background: The first in vivo positron emission tomography (PET) imaging studies of amyloid-beta (A ) in human subjects were reported by the UCLA group in 2002 using a fluorine-18 labeled membrane dye derivative, [F-18]FDDNP, that bound preferentially to both A plaques and neurofibrillary tangles (NFT) in the brains of Alzheimer’s disease (AD) patients relative to elderly control subjects. Groups from Pittsburgh and Toronto followed in 2004 with literature reports of PET imaging studies in AD and control subjects using A -selective carbon-11 labeled thioflavin-T ([C-11]PIB) and stilbene ([C-11]SB-11) derivatives, respectively. Methods: Researchers have sought to employ these PET agents in cross-sectional and longitudinal studies of A deposition in a variety of non-invasive imaging studies in amyloidbased diseases. In addition, investigators have sought to expand the choice of A -specific radioligands to include other F-18 and C-11 labeled compounds for PET imaging applications, as well as I-123 labeled agents for single photon emission computed tomography (SPECT) imaging studies. Results: Over the past 4 years, literature reports of A imaging in human subjects using PET have exploded, with more than 70 publications in a variety of subject populations. The vast majority of these studies have utilized [C-11]PIB, but the relatively short half-life of carbon-11 (20 min) has limited its distribution to about 40 specialized PET radiochemistry facilities throughout the world. The development of a more convenient, longer-lived (110 min) F-18 labeled radioligand with in vivo properties comparable to those of [C-11]PIB has been an active area of research over the past several years with some recent encouraging results. In contrast, efforts to develop a SPECT A imaging agent have been largely unsuccessful. Conclusions: PET A imaging studies have provided some important insights into the pathophysiology of amyloid deposition, particularly in mild cognitive impairment (MCI), healthy elderly controls, and familial AD subjects. A PET methodology is proving valuable in helping to assess the amyloid cascade hypothesis, as a diagnostic agent for brain amyloid deposition, and as a potential surrogate marker of therapeutic efficacy in anti-A drug trials. IC-PL3 PHARMA AS A CONSUMER: BIOMARKER APPLICATIONS FOR DEVELOPING NEW ALZHEIMER’S DISEASE TREATMENTS Eric Siemers, Eli Lilly and Company, Indianapolis, IN, USA. Contact e-mail: esiemers@lilly.com Background: Drug development by the pharmaceutical industry may involve applications of techniques developed prior to their incorporation in a randomized clinical trial. The role of imaging and other biomarkers to the development of disease-modifying treatments for Alzheimer’s disease (AD) is reviewed. These applications apply to Phase 2 and 3 strategies and to the identification of patients who might benefit from disease-modifying treatments even before the onset of clinical symptoms. Methods: Possible strategies for Phase 2 development including qualifications for surrogate markers were reviewed. An economic model was developed that compares the costs of various imaging or biochemical diagnostic tests for AD pathology to possible cost savings assuming a disease-modifying treatment is available. Results: While qualification for biomarkers in general may be “fit for Alzheimer’s & Dementia 4 (Suppl 2) (2008) 1552-5260/08/$ – see front matter © 2008 The Alzheimer’s Association. All rights reserved. purpose”, qualification for a surrogate marker is more stringent and may be difficult to achieve. Nevertheless, imaging and biochemical biomarkers may together provide convincing evidence that a given treatment at a given dose does impact amyloid plaque in brain parenchyma or soluble A in the central compartment, thus providing support for the initiation of long and costly Phase 3 studies. Diagnostic biomarkers including amyloid imaging and cerebrospinal fluid biochemical studies were considered for patients with mild cognitive impairment or for cognitively normal individuals 60 years of age or older. Data from economic models will be presented showing the impact of early diagnosis using these biomarkers on subsequent total costs related to AD, assuming the availability of disease-modifying treatments. Conclusions: Imaging and biochemical biomarkers are techniques that have important roles when applied to the drug development process. In addition to providing important Phase 2 data leading to better decision making for Phase 3 development, imaging and biochemical biomarkers may provide important data to identify which patient populations could benefit most from disease-modifying treatments. While pharma companies may not develop biomarker technologies themselves, close collaboration with academic groups and diagnostic companies can foster these efforts. SATURDAY, JULY 26, 2008 ALZHEIMER’S IMAGING CONSORTIUM IC-O1 IMAGING AND GENETICS IC-01-01 CORTICAL BINDING OF PITTSBURGH COMPOUND B, AN ENDOPHENOTYPE FOR GENETIC STUDIES OF ALZHEIMER’S DISEASE Anthony L. Hinrichs, Mark Mintun, Denise Head, Anne M. Fagan, David M. Holtzman, John C. Morris, Alison Goate, Washington University School of Medicine, St. Louis, MO, USA. Contact e-mail: tony@fire.wustl.edu Background: Little progress has been made in identifying genetic risk factors for Alzheimer’s disease (AD) during the last decade, underscoring the need for novel approaches to this problem. Because all known AD genes have been shown to influence A we hypothesized that quantitative measures of A deposition may be a useful endophenotype. The development of in vivo imaging of A deposition in the human brain using the Positron Emission Tomography (PET) ligand, Pittsburgh compound B (PIB), offers the possibility of using cortical PIB binding as a quantiative endophenotype for genetic studies of AD. Objective: To determine the heritability of cortical PIB binding and to test for association between cortical PIB binding and known risk factors for AD. Methods: Heritability can be estimated by twice the intraclass correlation coefficient (ICC) among siblings. We therefore undertook a study of elderly sibships. We ascertained 25 sibships without regard to disease status. The mean age was 72.3 with a mean of 2.2 individuals per sibship. Since the trait distribution is naturally truncated and many individuals show no deposition, the trait is significantly non-normal (p-value .0001). We therefore tested heritability using two other methods. We examined a sample of unrelated individuals in whom both PIB and CSF biomarker measurements have been made to test how much of the variance in PIB binding is explained by the biomarker levels and by APOE genotype. Results: ICC analysis using the Winer measure shows a heritability of .64. Permutation testing yields an empirical p-value of 0.03 to reject the hypothesis that the trait is non-heritable. Treating deposition as a dichotomous trait of presence (n 17) or absence (n 38) yields a heritability estimate of 0.63. While we observed that cortical PIB binding is correlated with AB42 levels in CSF, dementia severity (CDR), ApoE genotype, and gender, together t