Biomarkers

Biomarkers are the bridge between basic and clinical research. It enables us to speed up the process of drug development (Winterer, G. The Handbook of Neuropsychiatric Biomarkers. Springer 2009). An NIH (National Institute of Health) study group committed to the following definition of biomarkers: A characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention (Biomarkers Definitions Working Group. Clin Pharmacol Ther 2001). This biomarker definition is relatively broad encompassing both physiological indicators (functional biomarkers) such as blood pressure, heart rate and brain function measures as well as molecular measures (molecular biomarkers) like gene expression, proteins, immunological markers etc.

Why biomarkers?

Biomarkers are particularly useful when translating pre-clinical to clinical findings (bench-to-bedsite). Biomarkers can be helpful:

• when difficulties arise in the decision-making process with regard to investigational new drug (IND) applications, i.e, when trying to extrapolate results from preclinical animal studies to the use in humans.
• for proof-of-principle and proof-of-concept studies. In the field of drug research, a company will often undertake research initially, to prove that the core ideas are workable (proof-of-principle) and feasible, before going further (first-in man studies, first-in-patients studies). The use of “biomarker-enhanced” proof-of-concept studies helps to establish viability, technical issues, and overall direction.
• For stratification of patients.During phase-I, early phase-II and in the subsequent process during late phase-II and phase-III (or even phase-IV studies), biomarkers can be helpful, i.e., to stratify proband and patient groups („personalized medicine“).

How does it work?

Pharmaimage takes care of the required infrastructure for biomarker investigations wherever your patients or probands are located in Europe. This includes:

• planning and set-up of the (biomarker) study
• training of staff for the collection (biological samples, imaging and electrophysiological data etc.)
• storage and transport of imaging/electrophysiological data and biological samples
• documentation of sample/data acquisition
• data and sample analyses are mostly conducted in Berlin

Normative Reference Databank

Any valid biomarker study requires normative reference data for comparison. Reference data vary based on several factors, including the sociodemographic background of the healthy population from which imaging data or biological samples were obtained and the specific methods and/or instruments used to collect the data/samples.

Why are reference data important?

• Providing validated reference data/material (population level data) would be too expensive for many academic (or private) institutions.
• Sample stratification: when selecting your normal control or patient group, e.g. for a drug trial, it is important to know beforehand what is “normal” and what is “on the edge”.

In this regard, Pharmaimage is in a worldwide unique position. Cross-sectional and longitudinal population-based normative reference data (3Tesla neuroimaging, electrophysiology, cognitive performance) and biological samples (blood, plasma) from > 3000 adult subjects across all age ranges with an emphasis on elderly subjects are available. The reference databank is continuously expanded. We expect that within the framework of the current BioCog project (funded by the European Union), the number will rise to N = 4000 healthy (elderly) subjects (www.biocog.eu) (multimodal imaging data, molecular data). All data sets included undergo regular QC. Sociodemographic data are available for all subjects and all subjects have undergone extensive cognitive testing (mostly using the Cantab® test battery). Biomarker data can be combined with demographic and cognitive endpoints. Sample stratification can be conducted on the basis of predefined cut-offs for any imaging, electrophysiological or molecular measure as well as combination of measurements. Alternatively, using the normative reference post hoc evaluations may facilitate a better understanding of the obtained study results – positive or negative.

Imaging & Electrophysiology Biomarkers

Pharmaimage offers a wide range of in-house neuroimaging and electrophysiology capabilities. Generally, imaging and electrophysiological experiments are conducted close to the patients/probands wherever they are (CROs, hospitals in the Berlin area or anywhere in Europe). Imaging facilities are contracted on-site and/or electrophysiological equipment is provided by Pharmaimage for the duration of the study. Pharmaimage also takes care of on-site staff training if required which includes regular monitoring. Data are transferred via cloud service and analyzed by Pharmaimage staff in Berlin (Germany). Our imaging protocols are closely adapted to the protocols of ADNI (Alzheimer´s Disease Neuroimaging Initiative). Based on our population-based reference databank of electrophysiological and multimodal imaging data (cross-sectional, and longitudinal data from > 3000 healthy adult subjects across all age ranges which includes test-retest data), pre-selection (stratification) of study subjects (healthy, patients) can be conducted before volunteers/patients enter the study (e.g. drug trial). Finally, study results including statistical analyses and reports are delivered to our customers.

Functional Magnetic Resonance Imaging

Functional magnetic resonance imaging (fMRI) is increasingly used in CNS drug research. As a functional biomarker, fMRI is frequently applied when it is the aim to detect local drug effects, particularly in subcortical circuits of the brain.

Perfusion imaging (Arterial Spin Labeling) also is of tremendous value, in particular to assess brain perfusion in elderly patients with (early) vascular dementia and to evaluated whether and how a particular drug affects brain perfusion. We offer:

• Study set-up and fMRI data acquisition incl. QC/monitoring (and SOP for multicenter data acquisition)
• fMRI task conditions: resting (default mode network), N-back, verbal memory, oddball, reward etc.
• Data analysis: whole brain voxelwise, region-of-interest (ROI), seed-based functional connectivity and independent component analyses using the FSL/SPM8 software packages
• Arterial Spin labeling (ASL): Pulsed ASL sequence with echoplanar readout and online calculation of CBF (cerebral blood flow) maps
• Pharmaco-fMRI: established subanesthetic ketamine challenge, nicotine challenge
• Reference data from N = 100 clinically stable schizophrenia patients and > N = 100 subjects with MCI (Mild Cognitive Impairment) are available (in addition to N = 1000 healthy control subjects)
• Special offer: Simultaneous fMRI/EEG. e.g. for EEG-informed functional connectivity analyses without the need to define a priori an anatomical seed. For instance, EEG information (e.g. vigilance staging during the imaging session can be used for (partial) regression of the BOLD signal. Advantages: EEG together with fMRI increases sensitivity (e.g. to detect drug effects). EEG is a highly reliable measure, it stabilizes the obtained BOLD response measures (e.g. in default mode network analyses)

PharmfMRI: Comparative evaluation of subanesthetic ketamine effects on resting-state functional networks

Study design: double-blind, randomized, placebo-controlled cross-over subanesthetic S-ketamine application in N=17 healthy male subjects.

Simultaneous fMRI/EEG: EEG-informed fMRI Analysis increases Sensitivity to detect Drug Effects

A 1-mg nasal nicotine spray (0.5 mg each nostril) or placebo (pepper) spray was administered in a double-blind, placebo-controlled, within-subject, randomized, cross-over design. Simultaneous EEG-fMRI and behavioral data were recorded from 19 current smokers in response to an oddball-type visual choice RT task. Conventional general linear model analysis and single trial P300 (P3) amplitude informed general linear model analysis of the fMRI data were performed.

Upper left: (A) The conventional analysis where a constant impulse (same height and duration) is positioned at stimulus onset. (B) A variable impulse model where the height of the basis function is modified by single-trial P3 amplitude. When P3 amplitude was included in the model, this resulted in the height of the basis function being higher when P3 amplitude was larger. This reflects differences in the intensity of responses, as reflected in P3 amplitude. For both models, the input functions (onset time, event duration, and event intensity) were convolved with the gamma HRF, which blurs and delays the waveform to match the difference between the input function and the measured hemodynamic response.
Upper right: Grand-averaged ERPs for the electrode pool (Cz, Pz, CP1, CP2) in the placebo (black line) and nicotine (red line) conditions for the target (left side) and infrequent stimuli (right side). The central part of the figure shows stacked plots of single-trial responses to target stimuli for two representative participants, illustrating a consistent positivity around 350–400 msec poststimulus (P3).
Lower left: BOLD activation (target > baseline, red) and deactivation (target < baseline, blue) for the conventional analysis (second-level mixed-effects FLAME, n = 19, cluster-corrected threshold Z = 2.3, p = .05) for the placebo and nicotine conditions.
Lower right: Nicotine effects (nicotine > placebo) on the BOLD response to target stimuli (target > baseline) for both the conventional and P300 amplitude informed models (second-level mixed-effects FLAME, n = 19, cluster-corrected threshold Z = 2.3, p = .05).
Note:  Whereas in the conventional model (classic fMRI investigation) only little nicotine effects are observed, strong nicotine effects appear in the modern state-of-the-art EEG-informed fMRI analysis. This again supports the notion that the combination of EEG (high sensitivity to drug effects) combined with fMRI (superior spatial accuracy and resolution) can greatly improve  functional imaging in the context of drug trials.

Reference: Warbrick T, Mobascher A, Brinkmeyer J, Musso F, Stoecker T, Shah NJ, Fink GR, Winterer G. Nicotine effects on brain function during a visual oddball task: a comparison between conventional and EEG-informed fMRI analysis. J Cogn Neurosci. 2012 Aug;24(8):1682-94.

Structural Magnetic Resonance Imaging

In CNS drug research, structural MRI can be useful to define sub-groups of patients (e.g. dementia, MCI) or to monitor long-term (> six months) drug effects.

Special offer: Quantifying volumes of cholinergic nuclei in the brain

We offer:

• Study set-up and MRI data acquisition incl. QC/monitoring (and SOP for multicenter data acquisition)
• T2-weighted incl. high-resolution (e.g. hippocampus volume), T1-weighted, diffusion tensor imaging (DTI)
• Data analysis: VBM, ROI analyses including cholinergic nuclei (DARTEL Toolbox), fractional anisotropy, diffusivity, tractography, surface rendering using the Freesurfer software package or in-house software
• Reference data from N = 100 clinically stable schizophrenia patients and > N = 100 subjects with MCI (Mild Cognitive Impairment) are available (in addition to N = 1000 healthy control subjects)
• Upon request: Access to 7T ultrahigh-field MRI (e.g. for MR-imaging of blood-brain barrier, currently under development)
  
  

Electrophysiology

Among functional biomarkers, electrophysiology (EEG) is the work horse in CNS drug research. EEG is highly sensitive to drug effects and it is a classic "translational" biomarker because it allows a direct comparison between (small) animals and humans. An additional advantage is that test-retest reliability is high and that center measurement effects are negligible in multicenter settings. Most importantly, EEG is highly sensitive to detect drug effects.

• We offer:
• Standard EEG with 32-channel EEG systems from Brain Products (BrainAmp DC®). Upon request, amplifiers with up to 64 channel recordings can be provided as well.
• Study set-up and EEG-data acquisition incl. QC/monitoring (in established multicenter settings)
• Data analysis: clinical EEG evaluation, power spectra, vigilance staging, spectral coherence, source localization (LORETA), event-related potentials (ERPs): P50, N100, MMN, P300 incl. time-frequency analyses
• Pharmaco-EEG: established subanesthetic ketamine challenge, nicotine challenge
• Normal reference database: N > 3000. Patient reference database (Schizophrenia, MCI, ADHD): > 800
• Special offer: Simultaneous EEG with functional magnetic resonance imaging (fMRI) (see Functional Magnetic Resonance Imaging)

Molecular Biomarkers

While the core competence of Pharmaimage is in the field of neuroimaging and electrophysiology, we also support investigations with molecular biomarkers, in particular, we provide a large population-based reference database and take care of the required infrastructure (e.g. storage and transportation).

• The reference database (and biobank) is from the same subjects like the imaging and neurophysiological data. It includes blood and plasma samples (and CSF from a subgroup) for genetic, gene expression (mRNA/miRNA), metabolic and immunological marker investigations, as well as for the investigation of neurodegenerative markers and markers to assess blood-brain function among others. The reference database and state-of-the art biobank is continuously expanded (currently N > 3000 healthy adult subjects across all age ranges.
• For biobanking and molecular investigations, Pharmaimage closely works together with additional partners. Preferred partners are Atlas Biolabs and Immundiagnostik (http://www.immundiagnostik.com/en/home.html) among others. Atlas offers a wide range of Omics platforms, next generation sequencing (NGS) capabilities. Close cooperations also exist with a number of academic research labs across Europe which allows us to include sophisticated biomarkers in the study design.