The application of human tissue for drug discovery and development

Abstract

Human tissues are invaluable resources for pharmaceutical research. They provide information about disease pathophysiology – and equally importantly, healthy function; confirmation (or refutation) of potential drug targets; validation (or otherwise) of other models employed; and functional models for assessing drugs' effects, whether beneficial or undesirable, in the most appropriate environment that can be replicated outside the human body. While human tissues have long been prized by pathologists in furthering our understanding of disease processes, there is a growing appreciation of their value at the late pre-clinical stage of drug discovery. Human tissues' potential to contribute to earlier phases of the process, before significant resources have been expended, is also now gaining recognition. Mounting concern over high rates of clinical stage drug failures mandates exploration of avenues for improving efficiency. Human tissue-based assays could play a key role in improving the translation process, as well as in moving towards stratified or personalised medicines. This editorial highlights some of the potential benefits of introducing human biosamples at each stage of the research process as a drug moves from concept to clinic. Some of the challenges with respect to obtaining tissues, minimising variability and gaining acceptance are also discussed.

Keywords::

• biomarker identification
• efficacy
• human tissues
• target identification
• target validation
• toxicity
• translational research

1. Introduction

Human tissues, whether surgically resected and surplus to diagnostic requirements, donated post-mortem or made available by the donation of biopsies or blood samples, are available in a variety of formats, each of which is suited to different applications. Every tissue type is available in some form, if you are prepared to be flexible in your requirements or creative in your use of the available material. Their clinical relevance is clear, and they can play a vital role in translating laboratory insights into sufficiently safe and effective treatments and contributing to the generation of diagnostics [1,2].

Human tissue acquisition is governed by differing national legislation, with restrictions dependent on factors such as whether the donor was living or deceased at the time of donation, the purpose for tissue removal (therapeutic or additional to therapeutic and diagnostic requirements), the donor's mental capacity at the time of tissue removal and whether medical records will be required and if so, whether they will be anonymised. The availability of medical records, and follow-up data, greatly increases human tissues' translational value.

Tissues may be utilised fresh, for functional assays, or frozen or fixed in a variety of ways, for analytical purposes or the production of tissue microarrays (TMAs). TMAs are particularly valuable for rapidly screening large numbers of tissues while using small quantities of those tissues [3]. Where specimens are to be derived from libraries of samples surplus to diagnostic requirements, they are usually formalin-fixed-paraffin-embedded (FFPE). The most suitable format depends on the application for which they are sought; for example, whether you wish to evaluate a tissue's response to the addition of a drug (fresh functional) or to identify biomarkers for disease progression (FFPE could be used).

2. A tissue for every stage

Human tissues can usefully be applied to basic research into disease pathophysiology, target validation, hit-to-lead identification, safety and efficacy assessment, biomarker discovery and companion diagnostic development. A brief explanation of their role at each stage, and examples of their implementation, follows.

2.1 Basic research into disease pathophysiology

In the case of human tissues derived from patients with the diseases and conditions being studied, they are of direct relevance and applicability.

For example, a significant proportion of patients with epilepsy have pharmacologically refractory seizures [4], many of whom are potentially eligible for surgery [5]. Tissues resected in order to remove suspected epileptic foci represent a valuable opportunity to study the processes at work in the brains of epileptic patients in a way that has immediate clinical relevance. Using such tissue, Roopun et al. showed that the mechanism underlying an electrical signature associated with epileptic networks was mediated by non-synaptic pyramidal cell activity [6]. This offered a genuinely new insight into the pathophysiology of this distressing disease and underscored the potential of such tissues in assessing therapeutic interventions targeted at patients with pharmacologically refractory seizures.

In order to understand disease, it's helpful to understand the healthy state. Another organ that has gained from the study of human tissues is the pancreas. Researchers discovered that, contrary to expectation, the structure of human pancreatic islets diverges significantly from that of rodents, which are common research models [7]. Subsequently, researchers have identified other differences in structural and operational features [8,9], which it is believed have implications, not only for studies of diabetes, but also for assessing the quality of islets before transplant, and optimising growth of the relevant cells in culture.

2.2 Target validation

Human tissues can be used to increase (or decrease) confidence that modulating the function of particular drug targets will have therapeutic benefits [10]. It is critical to ascertain whether a potential drug target is present throughout the course of a disease, or only at certain stages. It is also important to find out whether a drug target is present in all instances or the proportion of cases where it is found. These factors help to determine the viability of that target as an approach to tackling the disease in question.

2.3 Hit-to-lead identification

While conventional tissue slices may be impractical for high-throughput screening (HTS), primary cells or reconstituted tissues on standardised 3-D scaffolds offer a good compromise solution that enables more rapid screening while facilitating greater mechanistic understanding and multi-cell type analysis.

Research to generate transplantable organs or organ parts can also generate useful models for testing new drugs. Accurate liver models are especially in demand as drug-induced liver injury is the most commonly cited reason for withdrawal of an approved drug [11]. Fortunately, primary hepatocytes can be cultured as 2-D or 3-D models of liver function, and unlike many tumour- or animal-derived versions, they have been shown to either maintain expression of key metabolic enzymes or retain the ability to induce their expression on exposure to xenobiotics [12,13].

Respiratory tissue is also challenging to model as it is necessary to examine effects administered via a gaseous phase at the air–cell–liquid interface as well as via the liquid–blood phase. However, 3-D models of the relevant processes have been developed using biomimetic scaffolds in conjunction with suitable chambers for aerosol administration of drugs [14]. Another approach to reconstituting lung functions in a more complex and realistic yet reproducible and more convenient format is the ‘lung on a chip’ [15]. The organ on a chip format, which utilises microfluidic technologies to link phenotypically relevant cells, has also been extended to the gut [16], and both organ models and system and even multi-system chips, which use microfluidic technologies to link different 'organs' on a chip in physiologically proportionate and relevant relationships, have been undergoing continuous development and improvement for some time [17-19].

2.4 Safety and efficacy assessment

Complex organ and system reactions to a drug can be examined in isolated tissues. For instance, cytokine release syndrome (such as resulted when Phase I volunteers were administered the experimental immunomodulator TGN 1412 in 2006) can now be modelled using assays on cells extracted from donated blood [20,21]. Assays based on cells extracted from blood are being used to model tissue responses since fresh lymphoid tissue samples are not readily available. This process may include maturation of blood cells, co-culture with stromal cells and 3D cell culture to emulate tissue responses in vitro.

Functional assays may examine both efficacy and safety in fresh tissues. Some tissues are more difficult to maintain ex vivo than others and lose viability within hours. However, other organ and tissue piece or slice types can be maintained for days or even weeks. To illustrate: intact segments of human artery have been cultured for up to 56 days in a model of Human Cytomegalovirus infection [22]; lymphoid tissue fragments from tonsillectomies have been used to study Human Herpes Virus 6 infection for 15 days [23] and a range of normal and tumour slices have been maintained for 5 days [24]. Moreover, imaginative use of one tissue can sometimes substitute for one that is more difficult to obtain. For example, intact gut tissue expresses many of the same receptors as CNS tissue and so can, to some extent, provide a surrogate model of the CNS receptors [25]. Limitations remain; for example, this model would not facilitate the study of migraine, psychiatric disease, thermoregulation, memory or blood–brain barrier function.

Another example of the use of human tissues to provide unique insights prior to clinical trials comes from cardiology. Based on mouse data that indicated differential ion channel expression dependent on whether the atrium or ventricle was examined [26], two drugs were considered for the treatment of arrhythmias. However, when they were tested using human heart tissue, the results were quite different [27]. These drugs would have failed in these indications if tested in patients, probably at considerable expense, and possibly putting patients at risk.

Human tissues can also be used to identify whether a target of interest is also expressed in non-target organs, and at what level. TMAs can enable rapid detection of targets across a wide range of donors or tissues. Where the therapeutic is a monoclonal antibody (mAb), early detection of potential cross-reactivities or of target expression on non-target tissues facilitates analysis and prediction of likely side effects, and screening against a panel of human tissues is a US Food and Drug Administration (FDA) requirement [28]. A decision can then be made whether to halt development or to include a strategy for monitoring and mitigating likely side effects.

2.5 Biomarker discovery and companion diagnostic development

Human tissue-based research techniques are relatively well-established in oncology. This is partly facilitated by the wide use of surgery to resect tumours and the ease of access to samples of blood-borne cancers. Therefore, they are widely used to identify molecular signatures, known as biomarkers, of particular disease types to enable better classification and hence treatment selection and prognosis [29]. Such developments go hand in hand with companion diagnostic development. As pressures on healthcare budgets mount, coupled with the rising costs of new therapeutics, public healthcare providers and private insurers are coming to expect that with increasing price tags will come better means of identifying which patients are most likely to benefit from a particular therapy, and at what dose, moving towards personalised or stratified medicine.

In the field of prostate cancer, research using donated tissues has revealed that patients whose tumours express two particular biomarkers require aggressive intervention, meaning that other patients can enter a monitoring programme rather than undergoing surgery, with the complications and side effects this can introduce [30,31].

3. Challenges

3.1 Availability

The time sensitivity of the processes to be detected impacts upon sample availability; tissues that deteriorate rapidly, but are usually available only post mortem, such as CNS, can be challenging to obtain in the state required and may require creative or flexible use of materials. Likewise, quantitative detection of highly labile biomolecules such as mRNA may require that strict time frames are observed or particular preservatives used, with any deviations to noted so that they can be factored into the analysis. Fresh tissue use, where it is not available to the researcher locally, requires an adequate infrastructure to ensure its timely arrival at the laboratory.

3.2 Variability

Tissues may exhibit significant donor variability [32], complicating statistical analysis. However, this is surely more representative of patient populations than studies conducted using inbred animals and may simply require that larger sample numbers are used. A key consideration is to keep variability due to extrinsic factors such as handling and processing to a minimum by employing consistent standards, and noting any deviations [33-36].

3.3 Acceptability

As with any newer technique, the likely acceptability by regulatory bodies of the data generated can be a concern, as can the issue of whether and to what extent such assays should be validated, and if so, how and by whom. It makes sense to validate models of human biology against clinical results, rather than data from different species. Unfortunately, limited availability of human data can hamper this [37-39], and the impact of preclinical human tissue data generated by private research companies for commercial clients is difficult to ascertain as it usually goes unpublished. However, as industrial researchers increasingly appreciate the ability of human tissues to generate the correct answers to some of their most pressing questions, regulatory bodies will presumably see more of this kind of data and it is to be hoped that a virtuous circle will ensue, with evidence of acceptability of data leading to increased use of human tissues, in turn leading to greater familiarity with the assays, the data generated and any limitations. As previously mentioned, cross-reactivity testing using an extensive panel of human tissues has long been recommended by the FDA for mAb therapies [28]. Meanwhile, recognition that there are problems using animal models to identify mAb whose administration leads to cytokine release syndrome logically mandates the adoption of certain in vitro assays using human fresh blood, which may generate more accurate data [40,41].

4. Conclusion

Human tissues demonstrate considerable versatility in the range of applications and formats in which they can be exploited. Their value lies in providing insights that can be gained only by examining patients, healthy volunteers or their tissues. Although there are challenges to using human tissues, in obtaining sufficient tissues of adequate quality and quantity, with the associated clinical data, and in the desired format, and ensuring the results from those tissues are accepted, the difficulties that may be encountered are worth overcoming.

5. Expert opinion

While human tissue-based research is already contributing significant insights at every stage of drug discovery, there remains considerable room for improvement and expansion of their application, particularly to earlier stages of the process. The continued development and characterisation of induced pluripotent stem cell lines (iPS), created using skin biopsies from patients with particular diseases to generate cells typical of the less accessible organs affected by their disease or condition, such as nervous tissue, holds great promise for contributing further supplies of cells for medical research and pharmaceutical testing [42]. Meanwhile, despite the challenges, the benefits of their inclusion outweigh the costs. Their relevance to the clinic has never been more pertinent: as 92% of novel drug candidates fail in clinical trials [43] and the cost of bringing a drug to market was most recently estimated as $3.7 – 11.8 billion [44], so it is time to examine whether the application of human tissues could reduce these figures. If clinical attrition could be reduced by heavier investment in human tissue models, leading to the selection of candidates with more favourable safety and efficacy profiles, costs and development times could be reduced.

Where collaboration agreements with local clinicians are absent, networks of institutes and publicly funded biobanks, as well as commercial sourcing networks, are proving their worth. Organs from heart-beating donors, where those organs are unsuitable for transplant, provide unparalleled opportunities for studying healthy whole organs intact and in superb condition and every effort should be made to encourage their donation to research, as well as to develop infrastructure for their dissemination [45].

Declaration of interest

M Clotworthy is a director of Human Focused Testing, a company which sources human tissues for researchers.