1. Describe why non-invasive biomarkers are important in the pharmaceutical industry and what defines a “good” pharmaceutical biomarker

2. Understand how, why and where the pharmaceutical industry relies on molecular MR imaging and related biomarkers

3. Describe several recently successful and emerging MR molecular imaging applications in the pharmaceutical industry

4. Describe how MR imaging agents are facilitating MR molecular imaging and related specific biomarkers that can be used in pharmaceutical assessment

5. Describe current challenges and opportunities for MR molecular imaging in the pharmaceutical industry

Non-invasive medical imaging technologies came to prevalence in the 70s and 80s, seeing widespread use as diagnostic tools. MRI in particular became a standard for patient care across all the major human diseases, and more recently has been adapted for quantifying pharmaceutical response in most of the major disease areas. This adaptation has occurred in parallel with the shift toward personalized and translational medicine, a shift facilitated by the advances that now routinely enable largely the same MRI protocols to be used in laboratory rodents, 1000 times smaller than their human counterparts. Since rodent models have become a standard for early testing of new pharmaceuticals (“drug discovery”) over the last 1-2 decades, technological advances have greatly increased the interest of the pharmaceutical industry in the use of MRI based translational biomarkers, and most recently MR based molecular imaging applications [2-5].

The rapid shift into targeted and personalized therapeutics by the pharmaceutical industry has led to increasing need for specific and predictive biomarkers of therapeutic response. “Biomarker” is defined as a measurable parameter used as an indicator of a normal biological process, a pathogenic process or a response to a therapeutic intervention. An MRI biomarker is therefore an MRI measurable parameter that is proven to correlate with biologic, pathogenic or therapeutic response process. The ‘ideal goal’ for a biomarker is for it to progress to be a ‘surrogate marker’ or, that is, a true substitute or equivalent for a clinical standard end point.

A ‘sweetspot’ for MRI has been anatomical and volumetric biomarkers, due to the excellent soft tissue contrast and relatively high resolution of the imaging modality, compared with others. This has led to one area where MRI is used in a surrogate-like application, that being imaging of tumor volume, as a surrogate for the clinical standard end point, lifespan, in cancer. Some of the other more successful and widespread clinical MRI applications have tended to rely on detection of tissue and lesion boundaries, and related volumetrics. It has been a far greater challenge to see non-anatomical MRI quantifiable parameters become biomarkers and see broad use. Some of the challenges are related to the relatively low sensitivity of MRI (compared with other technologies). Despite this however, there are many examples of MR based biomarkers that are true measures of disease mechanism, or physiologic/molecular signatures of the pathology. Testament to this is the pharmaceutical industry’s routine use of MR molecular imaging protocols in discovery and clinical trials. However, it remains that most of this use involves non-specific biomarkers that readout complicated aspects of structure, tissue physiology and the tissue microenvironment including diffusion MRI, dynamic contrast MRI and functional MRI. As has been previously stated, molecular imaging is defined as the visualization, characterization, and measurement of biological processes at the molecular and cellular levels in humans and other living systems. To elaborate, “Molecular imaging typically includes two-or three-dimensional imaging as well as quantification over time. The techniques used include radiotracer imaging/nuclear medicine, MRI, MRS, optical imaging, ultrasound and others.” True molecular imaging biomarkers are therefore sought after by the pharmaceutical industry as they potentially enable much greater specificity to state of the art molecular and cellular therapeutic targets than traditional imaging biomarkers.

Beyond tumor anatomical measurement however, there are few examples where MRI based biomarkers have become validated and standard enough to approach use as a surrogate marker. Figure 1 represents the basic path an image quantifiable parameter would take to be used as a reliable biomarker or surrogate in drug discovery or clinical development. Unfortunately the surrogate validation requires many years of work across dozens of sites, and standardization of imaging scan protocols, software and analysis. The surrogate goal has therefore more and more become an ideal goal, not a necessity for imaging biomarkers to play a significant role in pharmaceutical research decision making. The biomarker validation though is critical to any kind of informative use of the parameter/biomarker.

The pharmaceutical industry aspires to the use of biomarkers that can increase the precision or speed of quantifying disease progression and response. The major driver for this is the need to reduce costs and increase the speed and success of clinical trials. Clinical trial success is of course related to the power and accuracy of data produced, and subsequent assessment of patient therapeutic response, and reliable biomarkers can improve these. Other important properties that maximize biomarker value include the following properties:

(i) Non-invasive: maximizes patient compliance and better enables repeat measurements

(ii) Translational: enables use, validation and optimization of the biomarker in preclinical models prior to clinical use and/or learnings from clinical testing to be applied to new preclinical hypotheses, in an iterative feedback mechanism

(iii) Spatially resolved: due to the typically heterogeneous nature of disease, this can provide more comprehensive information than can be obtained though limited blood sampling or biopsies (in human patients) or whole tissue based assays (in animal subjects)

Imaging biomarkers bring these important features to drug testing, and are therefore sought after. The inherent “cost” of biomarker use though, is associated with the required validation efforts and the risk that this will prove the biomarker not completely validated enough, or even invalid for the intended application. Fortunately there are many well established MRI biomarkers that have been and continue to be used in routine clinical and preclinical practice, as will be discussed below. The challenge with these existing biomarkers is choice of where to use or apply them, and maybe more importantly where they should not be used. However, the relative non-specificity of the currently used MR biomarkers underlies the need for the continued move toward more specific MR molecular imaging biomarkers, for example through the expanded use of new or existing MR imaging agents.

For the many potentially valuable biomarkers that are the subject of current or future discovery work, the path to successful use is a difficult one. Figure 1[3] depicts the required timing for completion of required biomarker discovery in order for the biomarker to be available for use by a pharmaceutical company in clinical development. As shown, the biomarker “sweetspot” in in the mid stages of the discovery/development path. Importantly, the biomarker must be ready for use (ie. sufficiently validated) before the first clinical trials commence for a new therapeutic target or approach. This creates large barriers and challenges associated with effort and funding required for parallel discovery and validation of imaging biomarkers.

Figure 2 [7] is a summary of searches (2016) using pubmed that highlight trends in imaging and MRI based pharmaceutical research. While significant growth in the % of pharmaceutical/drug reviews that referenced imaging and MRI occurred between 1995 and 2005, there has been no growth in this proportion since then. This may be related to many factors both economic and governmental but shows that any kind of ‘explosion’ in MRI (and imaging in general) for pharmaceutical discovery and development has not yet occurred. In general, the data suggests that MRI (and imaging in general) currently plays a relatively small, but nonetheless significant role in pharmaceutical and drug research. Futher evidence of the current state is highlighted by examining registered actively recruiting clinical trials using the clinicaltrials.gov database. Figure 3a [8] summarizes clinical trial imaging prevalence across several major imaging modalities and Figure 3b [8] summarizes the most active areas of MRI use in clinical trials (by protocol/biomarker) including:

(i) Diffusion weighted MRI

(ii) Dynamic contrast-enhanced (DCE) MRI

(iii) Magnetic resonance spectroscopy (MRS)

(iv) Arterial spin labeling (ASL)

(v) Diffusion tensor imaging (DTI)

(vi) Functional MRI (fMRI)

The data suggests reasonable prevalence of MRI in clinical trials, but also shows that it is largely associated with biomarkers that were first used or “discovered” 10-20 years ago, such as tumor Ktrans and apparent diffusion coefficient, and 1H MRS based metabolite biomarkers. Furthermore, except perhaps MRS, the most prevalent MRI biomarkers in clinical use today are relatively non-specific in terms of association with molecular targets and pathways and not necessarily considered true “molecular imaging” biomarkers. They are more appropriately termed physiological or functional biomarkers that limit to precision and specificity with which pharmaceutical target modulation can be quantified. Cancer and neurodegenerative/CNS diseases account for more than 50% of the clinical trials with MRI. For the biomarker categories examined, there is >65% greater prevalence in neurodegenerative/CNS diseases, compared with cancer (where predominant use of MRI is in anatomical determination of volume only, and not in using more sophisticated biomarkers).

The above referenced data clearly highlights the need for improvement in our ability to discover, validate, fund and appropriately use MR based biomarkers and MR imaging agents, especially molecular imaging biomarkers, and the relatively modest use of these biomarkers (except tumor volume) by the pharmaceutical industry currently. While the same can be said for other prominent translational imaging modalities including PET, CT and ultrasound, MRI presents the broadest set of applications, image contrast drivers, biomarkers and potential value adds for the pharmaceutical industry. Tomorrow’s MR imaging biomarkers therefore will require rapid adaptation to current trends in disease and therapeutic paradigms to ensure needed sensitivity, specificity and speed of underlying validation work. This will be needed to enable meaningful use of MR imaging in time for new clinical trials in a given disease and therapeutic area. A critical aspect of this challenge will be discovery and rapid optimization and validation of MR detectable imaging agents [6]. These imaging agents will drive MR molecular imaging biomarkers with required specificity and sensitivity for assessing modulation of pharmaceutical targets and pathways. The remainder of this paper will outline several of the most promising emerging MR Molecular Imaging applications.

1. Hyperpolarized 13-C Magnetic Resonance Spectroscopy

Hyperpolarized (HP) carbon-13 (13-C) technology has rapidly risen to prominent use [9-11], driven by the prevalence of metabolic targets in oncology which saw increasing focus about 5-10 years ago. The widespread use of 18F-FDG PET highlighted success in use of a metabolite tracer and respective biomarker for cell metabolism, but at the same time failures have defined the need for other imaging agents that drive biomarkers for other metabolic pathways. HP 13-C MRI has overcome some of the technological hurdles but remains expensive and relatively inaccessible. Nevertheless, several unique features of HP 13-C MRI incuding high sensitivity, lack of background and the ability to simultaneous detect multiple molecules, including substrates, intermediates, and products (due to the broad chemical shift of 13C nuclei). Selectively labeled 13C probes are available for various metabolic enzymes such as pyruvate, lactic acid dehydrogenase (LDH)-A, alanine transaminase (ALT), glutaminase (GLU) and others. These agents are enabling specific biomarkers of response for metabolically targeted cancer pharmaceuticals and may pave the way for new advances in the pharma industry.

2. Macrophage Detection

Inflammation is now recognized as a major feature of a broad spectrum of diseases, including cancer, and macrophage activation and infiltration is one of the key hallmarks of the inflammatory cascade. Macrophages are a relatively new target for many pharmaceutical companies. The ability to detect and quantify activation of macrophages in a spatially resolved manner is therefore very valuable, and MRI approaches to doing this are under development, including the use of F19 and FeOx containing agents that can be administered systemically and are selectively taken up by activated macrophages [12-17]. Applications where this may be used include in cancer and in diseases that implicate liver inflammation such as NASH (nonalcoholic steatohepatitis) and inflammatory bowel disease (IBD). There currently exist large unmet needs for NASH and IBD therapies, and these are currently large areas of focus for the pharmaceutical industry.

3. 1H Magnetic Resonance Spectroscopy (MRS)

Proton MRS is a mature MR based technique that has been used for many years in a variety of indications including cancer, neurodegenerative and other CNS diseases. Drug companies are currently turning to existing MRS techniques to look at metabolites that can be used as pharmacodynamic biomarkers of response [18, 19]. For example, the relatively new cancer target IDH1/2 in glioblastoma (where there remain no effective treatments) is leading to new interest in 1H MRS where 2-hydroxygluturate and other MRS-quantifiable metabolites may provide specific, reliable biomarkers that can be use to assess new treatments against this target [20].

4. Functional MRI and Pharmacological MRI

Functional MRI (fMRI) and the related pharmacological MRI (pHMRI) based on the BOLD effect is another mature MRI technique that is seeing increased interest for pharmacodynamic measurement of brain activation / increased blood flow/metabolism in response to CNS drug treatments [21-23]. By detecting regions of brain metabolism modulation, fMRI can be used for target validation as well as for enabling reliable biomarkers for drugs such as cognitive enhancers, for example in Alzheimer's, schizophrenia and depression. While this approach relies on a physiological biomarker that may not be considered “molecular imaging” these MR based methods overcome some of the specificity limitations of other available techniques and may provide the precision needed for pharmaceutical discovery and clinical trial decision making.

5. Therapeutic cell detection

New areas of medicine based on cell therapies including stem cell/regenerative medicine and the explosive T cell therapeutics being used in immuno-oncology are driving new interest in, and development of non-invasive cell detection, such as through MRI [24-27]. Current MRI approaches include the in vitro labeling of cells with phagocytic properties (including T cells) using benign 19F and FeOx imaging agents, and subsequent detection using MR methods, after cell inoculation. A “holy grail” for the field is the concept of “in vivo cytometry” based on administration of specific labeling agents that perform endogenous labeling of cells such as immune cells. Current MRI methods do not provide the sensitivity or specificity required to do this, but it remains an area for opportunity if major advances in MR imaging agent discovery and optimization can be achieved.

6. Chemical exchange saturation transfer

Chemical exchange saturation transfer (CEST) is a magnetic resonance imaging (MRI) contrast enhancement technique that enables indirect detection of endogenous metabolites with exchangeable protons. These imaging agents include glycosaminoglycans, glycogen, myo-inositol, glutamate, and creatine [28, 29]. CEST molecular imaging is therefore seeing use in clinical trials and by the pharmaceutical industry for a variety of purposes related to treatment monitoring. One further CEST example has been the developments with CEST based quantification of pH [30]. This holds significant promise, for example in interpreting patient response to pH sensitive therapies, that are an active area of research in the pharmaceutical industry.

Non-anatomical, MRI based Molecular Imaging is playing a significant, though overall modest role in the pharmaceutical industry today. Several areas discussed above and that will be detailed during the presentation demonstrate the potential for MRI based biomarkers to achieve unique specificity that can be leveraged by the pharmaceutical industry. Challenges exist in terms of ensuring the sensitivity needed with affordable cost, in time for the next waves of clinical trials based on today’s most promising therapeutic and molecular targets including in the explosive area of immuno-oncology.