OR WAIT null SECS
One can use modeling methods to generate data and shorten development timelines in preclinical studies.
Modeling approaches have been gaining in popularity in the biopharma industry with companies seeking ways to maintain drug development costs. Preclinical studies are one area where modeling techniques are applied, but what is the feasibility of going to purely modeling techniques for preclinical studies? Will the industry opt to combine a modeling approach with traditional animal studies?
Animal studies have been the go-to approach in the biopharmaceutical industry for testing new biotherapeutics in the preclinical phase. According to Lorna Ewart, PhD, chief scientific officer at Emulate, a US-based provider of next-generation in-vitro models, the advantage of using animals in preclinical drug development is that they provide a dynamic environment to assess drug response across multiple organ systems. Scientists can, therefore, study the relationship between pharmacodynamics (PD) and pharmacokinetics (PK) required to understand efficacy and safety. Many modern biologic drugs, however, are designed against humanized targets, Ewart notes which may reduce the usefulness of the animal model as the pharmacology won’t be assessed.
“As genetic and physiological differences exist across species, animal models are not likely to give an accurate picture of a drug’s pharmacological and pharmacokinetic properties,” Ewart states. In addition, animal models of disease are regularly cited as being poor surrogates of the human condition and thus generate pharmacological data that may not accurately translate to human outcomes, Ewart explains—pointing out that Alzheimer’s models are a good example of inaccurately translated human outcomes from animal studies.
Sheng Guo, vice-president, data science and bioinformatics, at Crown Bioscience, a contract research organization and a JSR Life Sciences company, says that by providing an effective human substitute, murine animal models offer distinct advantages in preclinical drug development. The similarity in human and animal model in-vivo microenvironments allows for PK/PD studies to produce results closely representative of those expected in humans. Murine models also offer comparability between treated and untreated subjects and are often repeatable, which Guo says is important during preclinical drug development.
“However, these studies can be costly and can present many challenges when trying to screen large cohorts, especially for multi-factor screening. Ensuring studies modeled in vivo are completed within often tight timelines can also be difficult, as some animal models may need to be revived, which can take up to three months,” Guo explains.
Moreover, the differing growth rates of tumors mean that some may take many weeks to reach the size needed for these preclinical studies.
“These potential cons need to be carefully considered and balanced against the pros when deciding whether to commence with a study to be modeled in vivo,” Guo states.
Using a modeling approach in preclinical studies can be beneficial where traditional animal studies may be challenging.
Such models will often be based on historical data from in-vivo models that have been accumulated over many years, emphasizes Guo. These models, therefore, have the potential to save money and time, as the in-vivo studies have already been conducted, Guo points out.
“The use of carefully curated large data sets is essential when using an in-silico model approach to gain the greatest insight into preclinical studies. Implementing well-established mathematical modeling methods that are readily applicable can further shorten time frames when conducting these in-silico studies,” Guo says.
Meanwhile, Ewart states that a modeling approach using organ-on-a-chip technology enables scientists to overcome the major limitation of the animal model because the organ-on-a-chip model can be created using human cells. As a result, human-relevant pharmacology can be explored.
“Because the organ-on-a-chip models are also dynamic, the relationship between pharmacology and pharmacokinetics can be explored over time. In addition, the organ-on-a-chip technology has shown that cells remain functional and viable for longer periods of time, enabling researchers to study longitudinal drug responses,” Ewart says.
As models such as this become more mature, the endpoints that scientists can measure are becoming more relevant to those endpoints performed in the clinic, thus enhancing the translational value of the model, Ewart adds.
Particular modeling strategies would make more sense for use in the preclinical stage of biologic drug development. For instance, models built using human cells would be preferable, according to Ewart.
For biologic drugs in the preclinical stage of development, strategies should combine empirical data analysis, statistical modeling, and computational simulations in the design and analysis of murine clinical trials to produce studies that are more rational and powerful than in-vivo or in-silico models alone, adds Guo. According to Guo, these modeling strategies should use a comprehensive and diverse data collection that could include:
“Ideally, modeling strategies should be based on methods with prior successful applications where there is an established framework and guidelines on the design, analysis, and application of the resulting murine clinical trial. By applying these modeling strategies carefully, insight can be gained into the mechanism of action of biologic drugs, and their efficacy can be predicted,” states Guo.
It is unlikely that modeling will completely substitute in-vivo studies, says Guo. However, modeling is a powerful approach that can be used to complement preclinical murine models to help them achieve their full potential.
“Modeling can be used to improve murine model selection, experimental design, data analysis, result interpretation, and biomarker discovery. As a result, more effective clinical studies that are frequently biomarker-guided in nature can be achieved,” Guo states.
Using modeling methods has made it possible to establish empirical quantitative relationships between mouse number and measurement accuracy for categorical and continuous efficacy endpoints, Guo adds.
“Studying a drug’s efficacy is not currently subject to regulatory guidelines, and it is feasible that models could be used exclusively,” states Ewart. “However, assessment of a drug’s safety is a regulatory requirement with strict guidelines in place. The regulatory community still advocates for animal data to make a risk assessment and is less likely to support drug progression on modeling alone.”
Ewart also points out that, until more models are subject to rigorous evaluation processes and amass data across a broad range of therapeutic modalities, the scenario that will yield better preclinical results is one that incorporates data sets from both animal studies and modeling methods.
Guo concludes that modeling, if applied correctly, can be a useful approach to understanding biologic drugs at the preclinical stage; however, the combination of modeling and murine studies together can be much more powerful.
“As well as the improvements in experimental design and analysis of data, this approach can offer faster biologic drug development, allowing critical drugs to be brought to patients in shorter time frames,” Guo says.
Feliza Mirasol is the science editor for BioPharm International.