The phrase “well begun is half done” (attributed to everyone from Aristotle to Mary Poppins) could be considered true of drug discovery and development. One of the early steps in the process that can make a difference is selecting the right biological model for the job.
Outsourcing-Pharma (OSP) recently discussed the evolution of biological models used in drug discovery with Katrin Hoeck (KH), associate director of marketing and business development for discovery solutions with Lonza.
OSP: Could you please describe how the drug discovery pipeline has evolved (or, possibly, devolved) in recent years? a
KH: Drug discovery is always in a state of continuous evolvement. For example, in the 1980s and 90s, the rise of automation and high-throughput capabilities triggered the large-scale screening of compound libraries to identify new hits.
Though this approach resulted in the development of many commercially available drugs, it also saw researchers quickly exhaust easily druggable targets, causing companies to explore iterations of existing drugs, which is a less commercially rewarding activity. It was this declining commercial profitability, combined with increasing project costs, that led to the industry concentrating more on the newly evolving field of bio-therapeutics.
The rise of new technologies and the ability to produce bio-therapeutics in large scale have since resulted in the development of completely new therapeutic approaches with some outstanding results, particularly in immuno-oncology.
What’s more, in recent years we’ve also seen a general mindset shift in the industry, where instead of companies running (and often failing) many projects at a time, they typically now run one project that looks more likely to succeed. In other words, we’ve moved from focusing on quantity to quality. And quality projects require a far better understanding of the physiological context of a disease and its potential therapies, which is driving the development of many current innovations.
OSP: What are some of the primary obstacles drug developers face in their efforts to bring therapies to trial and then to the marketplace?
KH: Drugs typically fail during development or clinical trials for two main reasons: efficacy or safety, which account for 80–90% of all failures. Preclinical safety usually involves testing the lead candidate in three animal models. However, as the physiology of humans and animals differs, the insight gathered from testing in these model systems does not often reflect how the drug will behave in humans, resulting in many clinical study failures.
As the financial consequences of failure rise disproportionally at the later stages of drug discovery, the aim is to shift attrition rate to the early stages and only move forward with promising “quality” projects.
OSP: You’d said that “the foundation of any successful drug development project is making the right choices as early in the process as possible – particularly by considering the right target, tissue, safety and patient.” Could you please go into detail as to why those things are important, and why hitting those marks is such a challenge?
KH: Identifying and deciding on the right target is key for the future success of any drug development project. While many genes or proteins (i.e. potential drug targets) may be known to play a role in a disease model, not all of them will provide a functional effect if “targeted”. In addition, target expression may vary in patient populations due to genetic or epigenetic reasons, diminishing the success rate of a drug if the patient cohort for the clinical trial is not chosen carefully.
We should also stress that drug-drug interactions may lead to toxic effects, so it’s critical that the target is understood in the context of health and disease to avoid safety failures in the clinical trial phase. So, in short, if you bet on the wrong target (i.e. make the wrong decision), you are doomed to failure right from the start.
In addition to selecting the right target, researchers need to understand the pharmacokinetics (PK) of a candidate drug—how it is absorbed, distributed, metabolized and excreted within the human body—as well as how the body will influence the candidate drug (the pharmacodynamics (PD). Understanding the PK and PD effects are equally important when it comes to identifying the candidate drugs that will have a functional effect on the target at a tolerable dose.
However, the safety of the drug goes beyond the prediction of PK and PD, as the drug may have toxic effects on unrelated organs. Predicting potential toxic effects of the drug candidate early in the project is a key part of shifting attrition to stages prior to preclinical development.
OSP: Could you talk about why selecting the correct biological model system is so crucial?
KH: The major reason for building biological models during the discovery and preclinical phases is to predict activities of and responses to drugs in human patients in the clinic as early as possible in the development pipeline. For a biological model to be predictive, it needs to replicate the same mechanisms and environments found within the human body.
If researchers select the wrong models, they could very well end up delivering a drug into clinical trials that is either not safe, or not effective. Such an outcome is very expensive and could also be harmful.
While models were formerly built with cell lines and often in a 2D cell culture, we are witnessing an ongoing paradigm shift within the industry to working with more physiologically relevant models. Based on our experience, using human primary cells instead of cell lines is certainly the more predictive model, as human primary cells have not undergone genetic modifications, and so they truly represent the biology of the target tissue.
With the rise of more sophisticated 3D technologies like spheroids, organ-on-a-chip and bioprinting to name just a few, the intent is to start modeling organs as they are found in the human body. We find that the option to co-culture various cells, as found in the organ, paired with the more physiological 3D culture of cells, leaves hope that at some point these models may replace animal testing to a large degree, as they may prove to be the more predictive systems.
OSP: Could you please talk about how to best go about building the right biological model system?
KH: To develop the right model for research, it’s crucial to first understand its purpose. Currently, there is no single model for the laboratory that captures all of human biology, so the first step is to know what physiological processes are most important to have in the model in order to establish whether a drug is effective and safe. The next step is to determine how to capture those processes in an addressable model system.
For drugs that are targeting a human disease of a specific tissue type, the cells present in that tissue are an obvious base for developing a predictive model system. For example, to develop a model to predict the effects of a drug for Non-Alcoholic Fatty Liver Disease (NAFLD), one would start with the various human liver cells, including hepatocytes, stellates, Kupffer and endothelial cells.
Researchers would then combine these cells with supporting media and other factors in such a way to replicate a fatty liver. To identify specific genes or pathways involved in NAFLD, researchers may then use a CRISPR screen prior to forming the model.
In such a phenotypic screening approach, you would randomly knock-out genes in the primary liver cells using the CRISPR technology to identify which genes influence the development of NAFLD, thus identifying potential targets.
OSP: How can one determine the proper combination of human primary cells, specialty media and transfection technologies?
KH: We are mindful that determining the right combination of products can seem like a daunting task of trial and error. When exploring their options, researchers should always look for and use reagents, cells and technologies that were developed to work together.
We find that such integrated products can save researchers a lot of time. For example, many Lonza primary cells are supported by specialty media, so called BulletKits, that have been optimized for many key performance indicators.
We also create and support numerous cell type-specific transfection protocols for use with the Nucleofector technology that are optimized for high efficiency and high functionality. Our Nucleofector Technology enables non-viral transfection of hard-to-transfect cell types, including primary cells.
As the technology efficiently co-transfects multiple substrates, it has become the preferred method of choice when transfecting CRISPR substrates into hard-to-transfect cell types. With this, the modification and knock-out of genes in human primary cells is now possible, which was previously achieved using cumbersome viral transduction technologies.
Lonza’s knowledge center also provides researchers with links to hundreds of literature resources where successful models using primary cells have been developed and used.
OSP: Is there anything else you’d like to add—either about this topic, other projects you have in the works, etc.?
KH: Lonza has been a provider of human primary cells for many decades and delivers a unique combination of integrated solutions, including cells, cell-specific culture media and cell modification products. We understand the drug discovery industry’s need for more physiologically relevant in vitro model systems that can increase prediction accuracy and target outcome.
Our broad portfolio of primary cell types and available donors allows researchers to conveniently build models that mimic almost all organs and tissues of the human body. Specific cell types or donor characteristics, which may not be found in our standard catalog offering, can be provided through Lonza’s Cellbio services team.
As customization is key for many in vitro models, we can also offer further characterization of the cells before they are tested internally in a new model system. Our extensive scientific expertise around the isolation and culturing of primary cells has been at the heart of what we do for many decades and we aim to grow this further by closely collaborating with companies that provide 3D technologies so we can provide even more easy-to-use and robust 3D in vitro models in the future.