Commentary: FDA draft guidance on in-vitro DDI studies

How will it impact your drug discovery program?

Dr. Ron Laethem of BioIVT
Register for free to listen to this article
Listen with Speechify
0:00
5:00
The U.S. Food and Drug Administration (FDA) published a new in vitro drug-drug interaction (DDI) guidance on October 25, 2017, entitled “In Vitro Metabolism and Transporter Mediated Drug-Drug Interaction Studies: Guidance for Industry.” This was the first new DDI guidance on the FDA’s current thinking since February 2012 when it released “Drug Interaction Studies—Study Design, Data Analysis, Implications for Dosing, and Labeling Recommendations.” It was long-anticipated and didn’t disappoint those who like change. There is a lot to take in from this latest guidance, and some significant changes were presented. As with all guidances in this area, the FDA does not establish legally enforceable responsibilities, but rather presents recommendations that reflect the mindset of those who will be reviewing your regulatory filings. Ignore this valuable insight at your peril.
 
Along with the many tactical changes suggested in the guidance, there were a couple of strategic changes as well. One overarching theme was the further acceptance and promotion of modeling and simulation to more fully assess DDI risk before entering the clinic. When in-vitro DDI studies are carried out appropriately, the data can be used to make in-vivo predictions of potentially clinically relevant DDIs. This insight can be invaluable for designing appropriate clinical studies to prevent missteps and potentially avoid having to repeat trials.
 
Another major strategic shift was the endorsement of moving in-vitro metabolic studies to earlier in the drug development continuum, prior to first-in-human (FIH) studies. This concept isn’t new, and the intense pressure for pharmaceutical companies to reduce the high attrition rates in discovery has led to the deployment of drug metabolism and pharmacokinetics (DMPK) assays earlier in the development process. Assays such as metabolic stability, cytochrome P450 (CYP) inhibition and nuclear receptor reporter assays have been steadily moving closer to the discovery arena. This shift has resulted in better-quality drug candidates from an absorption, distribution, metabolism, and elimination (ADME) standpoint, but attrition rates are still very high in the pharmaceutical industry.
 
The major driver for FDA desiring to shift the in-vitro assays earlier in development is so that the data can be used to better design clinical trials and prevent the unnecessary exclusion of subjects. This rationale was stated in the FDA’s clinical DDI guidance that was released in tandem with the in-vitro DDI guidance (Clinical Drug Interaction Studies – Study Design, Data Analysis, and Clinical Implications). Patients in randomized controlled trials are, by definition, random; however, this can lead to studies where the volunteers are unrepresentative of the reference population for which the drug is to be prescribed.
 
By studying DDIs for a drug candidate more fully, it should be possible to better recruit for a trial and not exclude representative participants based solely on the fact that they are taking one or more other medications. Reducing this selection bias should enable improved studies where the data more accurately represent the target population for whom the intervention is intended. This paradigm, in theory, should also provide the FDA with data that will allow it to better assess the risk and benefits of an investigational new drug to the intended patient population.
 
While this approach should lead to better outcomes for the pharmaceutical industry and patients, it does require some changes to be made with respect to the drug development process. Currently, definitive in-vitro metabolic studies aren’t carried out until later in development, often after FIH clinical studies. This is because DDIs aren’t normally a concern for FIH studies when healthy volunteers with no comedications are recruited. The advantage of this paradigm is that the clinical Cmax for the experimental drug is known and can be used to guide the design of the definitive in-vitro metabolic studies and ensure that the results are relevant to the in-vivo situation. The FDA in-vitro DDI guidance suggests that when running in-vitro studies in lieu of the clinical Cmax value, sponsors design studies using the highest achievable concentration of the test article, if necessary up to the limit of solubility. When using high concentrations of test article in cellular systems, such as hepatocytes for CYP induction or mammalian cell lines for transporter studies, cytotoxicity can become limiting. If the highest achievable concentration of test article is outside the realm of physiological relevance, the cytotoxicity can confound results and raise concerns that may not be relevant in vivo.
 
When designing the in-vitro metabolic studies, the sponsor will have animal PK data to guide the concentration ranges used in vitro, and indeed this is a major part of determining the doses for FIH studies. However, the quality of the human dose prediction varies significantly for different potential drugs and having the clinical Cmax value is of great help in designing in-vitro studies that give predictive data. It would be taking a step backward for sponsors to run definitive in-vitro studies using nonclinical data to guide in-vitro study design, only to have to repeat those studies after clinical FIH data is available that suggests those original concentrations are not clinically relevant. Using the highest soluble concentration of test article in vitro addresses the clinical relevance issue for assays that don’t involve cellular systems, such as inhibition studies using pooled human liver microsomes. If the concentration range goes low enough from the highest soluble concentration, the data should cover all the possible Cmax concentrations that will be determined in the FIH studies. However, as noted above, for the cellular systems this could be problematic.
 
Using high concentrations of test article with the in-vitro cellular systems, such as human hepatocytes, could take an approach similar to the microsomal work. Rather than employing the limit of solubility as the benchmark, acceptable cytotoxicity would dictate the highest concentration used. However, the interplay between the cytotoxicity and the assay readout must be defined. For instance, in a CYP induction study using cultured human hepatocytes, how do we know if the cytotoxicity is relevant to the in-vivo situation? One approach is to determine the highest tolerated dose with the cultures and run the induction study to that concentration. This assumes that the dose-limiting cytotoxicity is reflective of what happens in vivo. If the cytotoxicity is an artifact of the in-vitro system, the study may not be able to reach concentrations that are found in the FIH studies to be clinically relevant. This would necessitate having to do a clinical study to look at the inductive potential.
 
If we implement in-vitro metabolic assays prior to FIH studies, greater care must be given to the concentrations of test article used. This is particularly important for cell-based assays where cytotoxicity can confound results. For in-vitro assays using cells, particularly human hepatocytes, we must ensure that the test system is accurately recapitulating all of the biological processes relevant to the in-vivo situation. For hepatocytes, this means that in addition to metabolic capacity and properly functioning nuclear receptor signaling pathways, the cultures should also maintain robust uptake and efflux transporter function. Setting up and deploying appropriate test systems to handle higher, and potentially non-physiological, concentrations of test article will be key to avoiding doing in-vitro assays both before and after FIH studies.
 
Sandwich-culture systems using hepatocytes that have physiologically-relevant uptake and efflux function, appropriate regulatory function and are metabolically-competent have been shown to have better in vitro/in vivo extrapolation (IVIVE) correlations than monolayer models. The improved predictability of the sandwich-culture system is likely due to the model being better able to manage high concentrations of test article and recapitulate how the liver is able to respond to a drug.

Ron Laethem, Ph.D., is lead, in vitro research at BioIVT, Hepatic Research Services

Dr. Ron Laethem of BioIVT

Published In:


Subscribe to Newsletter
Subscribe to our eNewsletters

Stay connected with all of the latest from Drug Discovery News.

March 2024 Issue Front Cover

Latest Issue  

• Volume 20 • Issue 2 • March 2024

March 2024

March 2024 Issue