Integrating safety pharmacology endpoints into toxicology studies makes sense. But uncertainty remains.
Fundamentally, safety pharmacology and toxicology are different disciplines. The former uses preclinical models to assess acute and potentially damaging effects of drug candidates on vital functions—namely the cardiovascular, central nervous, and respiratory systems—that could be harmful to study volunteers during Phase I clinical studies. Toxicology studies, on the other hand, are designed to detect more general—and possibly chronic—adverse effects in animals from repeat dosing. Measuring the effects help researchers establish safe starting doses and drug limits in clinical trials.
But these two disciplines can be compatible in an increasing number of contexts. So compatible, in fact, that when safety pharmacology endpoints get incorporated into a general toxicology study, the results can actually be a robust predictor of risk in late-stage clinical trials—when exposure is repeated and occurs over a much longer duration.
So when is it acceptable to integrate safety pharmacology endpoints into toxicology studies? Contract research organizations and pharmaceutical manufacturers largely agree that the approach reduces animal use and generates more robust safety data for both acute and chronic dosing. Ultimately, the US Food and Drug Administration, which is traditionally conservative, embraced the approach in its 2010 industry guidelines for nonclinical studies of cancer drugs because of the combination of extreme clinical need in advanced-stage cancer patients and sensitivity of available pre-clinical technologies to predict risk in humans to the vital organ systems.
Along with cancer biotherapeutics, the initial and logical compounds evaluated in integration studies have been biologic-derived products—primarily monoclonal antibodies. This makes sense. Due to their chemistry (specificity to target and steric inaccessibility to cardiac ion channels), these newer-generation therapeutics rarely interfere with classic safety endpoints. They also are desperately needed on the front lines of medicine, and like many pioneering therapies can assume a higher risk until alternative treatments are found or more robust testing can be performed.
With that said, momentum surrounding integration has diminished as the field wrestles with how (and how far) to take this paradigm. These questions were front and center during a lively panel discussion at last fall’s Safety Pharmacology Meeting in Rotterdam, where voices from the CRO, pharmaceutical, academic, and regulatory sectors weighed in on the pros and cons, and the issue continues to generate discussion and debate today.
Not a fail-safe system
The issues are not black and white. If integrated designs are sufficient to characterize safety of a class of compounds, why not more classes of compounds? How much data is needed in a generic design to meet regulatory requirements? Is snapshot ECG and blood pressure enough, or are jacketed/fully instrumented animals necessary? If we are to use these tools, how sensitive must they be to detect an indication of risk?
The answer from regulators (speaking on their own behalf at the Rotterdam meeting) was an unavoidable “it depends,” but there was progress made. The rush to integrate studies was replaced with a more balanced approach. Instead of keying on a certain limit of blood pressure change, an evidence-based approach was suggested that includes classical toxicology endpoints, such as clinical pathology and histology. The overwhelming sentiment, especially from regulators, was to have safety pharmacology data generated in both stand-alone and integrated designs to address both acute and chronic safety signals.
Of course executing both designs requires more resources than most companies have available to individual programs, and is likely overkill for the majority of prospective pharmaceuticals. So when considering the possibility of integrating safety pharmacology into a toxicology study, developers should consider expected pharmacology, known drug class effects, target population, dose levels, and any in vitro data that might be available, then decide if a combined approach makes sense. Although there is a perceived risk that some toxicology laboratories house studies in environments not suited for safety pharmacology assessments, choosing an experienced laboratory will help ensure that the preclinical work is sensitive enough to detect meaningful changes in safety endpoints.