In vitro DDI approaches for Advanced Modalities

The Challenge of Adapting the Small Molecule DDI Toolkit to New Modalities

In May 2024, the International Council for Harmonisation (ICH) released the M12 guidelines for drug interaction studies¹, aiming to harmonize international requirements for assessing drug-drug interactions (DDIs). It was adopted worldwide by the end of 2024, replacing various regional guidelines (e.g., in the United States, EU, Japan, and China).

As confirmed by the Expert Working Group that authored the guidelines², ICH M12¹ primarily focuses on DDI risk assessment for small molecules and includes only a handful of new modality-related recommendations. Other modalities may require different, specifically adapted strategies for DDI risk assessment, as their physicochemical and ADME properties may be very distinct from those of small molecules.

With the ICH M12¹ recommendation for testing enzyme- and transporter-mediated DDI in vitro as a starting point, we provide strategic insight and study design considerations for the following new modalities:

• Peptide Drugs
• Antibody-Drug Conjugates
• Targeted Protein-Degraders
• Oligonucleotide Therapeutics

Consult an Expert

Peptide Drugs

Peptides represent a diverse group of drug molecules in terms of size, physicochemical properties, and mechanism of action. According to the FDA, as described in their 2023 Drug-Drug Interaction Assessment for Therapeutic Proteins guidelines³, therapeutic peptides include “purified monoclonal antibodies, cytokines, enzymes, and other novel proteins for in vivo use”, but not ”proteins intended to act as vaccines or allergenic products, cellular and gene therapy products, and/or human cells, tissues, and cellular- and tissue-based products.” Major regulatory guidelines, including the ICH M12¹ and the 2023 FDA Peptide DDI guideance³, generally do not provide specific DDI risk assessment considerations driven for most different types of peptide drugs.

A major case of exception highlighted in both documents is the proinflammatory cytokine-related mechanism. Proinflammatory cytokine drugs and proinflammatory cytokine-modulating peptides can increase cytokine levels, potentially leading to alterations in cytochrome P450 (CYP) enzyme levels and activity. CYP450 changes affect the clearance of their substrate drugs, leading to increased or decreased exposure levels; therefore, assessment of CYP-mediated DDIs should be considered in such cases.

For DDI mechanisms unrelated to proinflammatory cytokines, no specific recommendations are provided by either guidance (the 2023 FDA document³ refers back to the 2020 FDA DDI guidelines⁴ for further details, which has been replaced by the ICH M12 in the meantime). Industry-wide survey results published by the Peptide DDI working group⁵ concluded that many peptide drug developers feel uncertain about planning their DDI risk assessment studies.

The survey also revealed that many peptide drug developers conduct ”traditional” in vitro DDI assessment studies with their compound(s) as the potential precipitant (perpetrator) of DDIs. This includes CYP inhibition and induction, as well as transporter inhibition testing, which respondents most often conducted prior to IND submission. Interestingly, fewer than 10% employed different strategies for different peptides in these assays.

The size of the peptide molecule was also found to be a good indicator of DDI potential. Molecules larger than 2 kDa were rarely involved in small molecule-like interactions, while smaller peptides represented significantly higher metabolism-based and/or transporter-based DDI risk. As a result, it is recommended that larger peptide projects follow a risk-based approach to assess the risk for DDI. Peptides smaller than 2 kDa, on the other hand, should be approached like small molecules, especially if they contain nonnatural amino acids or nonpeptide structural motifs. Our in-house experiences studying peptide drug DDI in vitro have also corroborated these trends.

Both regulatory bodies and the peptide DDI working groups have agreed, however, that the field is rapidly evolving, and intensified research and further data are necessary for refining DDI risk assessment strategies for diverse peptide types.

Antibody-Drug Conjugates

Another non-small molecule modality explicitly addressed in the ICH M12 DDI guidelines¹ are antibody-drug conjugates (ADCs) as part of the therapeutic protein DDIs segment (4.2) of the document. Currently, there is no specific stand-alone guidance providing recommendations for these molecules.

ADCs are comprised of a pharmacologically active small molecule payload attached to a monoclonal antibody carrier via a specific linker. After administration, the linker is cleaved and the unconjugated payload is released; therefore, it is advised to consider the DDI potential of both the antibody and the small molecule drug component. The ICH M12 recommends addressing the payload’s DDI potential in a manner similar to that of small molecules¹. This includes CYP induction and inhibition studies, transporter substrate and inhibition studies, and identification of metabolic enzymes involved in the payload’s metabolism. Different possible metabolites of the payload also need to be identified and considered for ADME characterization and risk assessment. This may also mean forms of the payload with parts of the linker attached to it, unless the linker has a single, clearly defined cleavage site.

The ICH M12 guidelines¹ acknowledge that the systemic concentration of the small-molecule payload component may not be high enough to pose a clinically significant threat of DDI risk (as precipitant, or 'perpetrator'). Even at lower concentrations, however, understanding their potential to be substrates (previously referred to as ‘victims’) of DDI may be necessary, as ADCs often have low therapeutic margins, and elevated drug concentrations systemically or in specific tissues may lead to toxicity if clearance mechanisms are inhibited⁶. Applying the appropriate bioanalytical strategy is also crucial in determining ADC concentrations in the systemic circulation and tissues of interest. These data can also inform clinical DDI studies and risk-based DDI interaction strategies, as proposed, for example, in the specific case of oncology drugs⁷.

Targeted Protein-Degraders

Targeted protein degraders (TPDs), including proteolysis-targeting chimeras (PROTACs), consist of a ligand for a protein of interest tethered to a ligand for an E3 ligase via a chemical linker. Upon TPD binding to a protein of interest (POI) and an E3 ligase, it brings them into close proximity, allowing the E3 ligase complex to ubiquitinate the POI, tagging it for proteasomal degradation.

TPDs are typically high molecular weight compounds (usually 700-1,000 kDa) with physicochemical characteristics beyond Lipinski’s rule of 5. They show limited solubility and cellular permeability, and their high non-specific and plasma protein binding can make in vitro assessment challenging.

In a survey conducted by the IQ Consortium’s Degrader DMPK/ADME Working Group⁸, respondents working on orally administered TPDs reported permeability as the most important ADME/Physchem challenge faced in their project (followed by solubility), while for TPDs intended for IV dosage, the main concern was solubility and, therefore, formulation. Plasma protein binding was only listed as an important concern by survey respondents.

Overall, the same in vitro assays can be applied for binding and DDI risk assessment of TPDs as are usually applied for small molecules. However, the assay systems may need to be optimized or adapted on a case-by-case basis to improve data quality and predictive value. Optimization possibilities include applying an extended bidirectional permeability setup for better assessment of permeability and the use of low-binding or low-retention tools to minimize non-specific binding. In addition to puffer-based solubility assessments, solubility testing in simulated gastric or intestinal fluids may also be needed for TPDs intended for oral administration.

Metabolic enzyme and transporter interactions of TPDs and their component have been described to be similar to those of small molecules. The linker between the POI and the E3 ligase ligand is usually cleaved, resulting in the release of the small molecule-like ligands. Depending on the linker cleavage sites and mechanisms, a mix of ligands with different linker-related moieties may be observed, which may undergo further metabolism as well.

Over 90% of survey participants reported CYP450 enzyme involvement in the metabolism of their TPDs, and other enzymes, including amidases, esterases, and UGTs, were also mentioned by over 40% of respondents. Clearance was mainly reported as biliary, and some cases of renal clearance and intestinal secretion were also mentioned. In terms of transporter interactions, P-gp-mediated efflux (mainly intestinal) and OATP-mediated uptake (mainly hepatic) were listed as clearance mechanisms.

TPDs and their released ligands have often been seen as transporter inhibitors, especially of P-gp and BCRP. They have been described as more inert in terms of SLC transporter interactions. As precipitants involved in DDI, they may also act as inhibitors or inducers of CYP enzymes; therefore, to assess DDI risk of TPDs and their metabolites (including the ligands released upon linker cleavage), it is generally recommended to follow a small molecule strategy and perform in vitro experiments as described in the ICH M12¹.

Oligonucleotide drugs

Oligonucleotide therapeutics “include a wide variety of synthetically modified RNA or RNA/DNA hybrids that are specifically designed to bind to a target RNA sequence to alter RNA expression and/or downstream protein expression,” as described by the FDA in their Clinical Pharmacology Considerations for the Development of Oligonucleotide Therapeutics, released in 2024⁹. Admittedly, this group contains a large range of different molecules that may differ in their structure, size, backbone, sequence, mechanism of action, delivery method, and so on. The FDA’s recommendations apply to “oligonucleotide therapeutics that target RNA by Watson-Crick base pairing,” without recommendations regarding specific characteristics such as backbone structure or conjugation.

While diverse, oligonucleotide therapeutics are usually of high molecular weight, ranging from antisense oligonucleotides (ASOs) at 6,000-7,000 Da, to double-stranded oligos weighing above 12,000 Da. Usually, they are not prone to non-specific binding, aside from a few examples, and their cellular permeability is also considered low or non-measurable in cell cultures or isolated cells. Due to these characteristics, oligonucleotide drugs are usually administered directly into circulation. For in vitro assays, the relevant concentration range is determined based on 50 x unbound C_max.

Oligonucleotide therapeutics are usually metabolized by endonucleases and exonucleases. The FDA guidance⁹ also confirms that they are not expected to undergo CYP450-mediated metabolism or be processed by other enzymes involved in small molecule metabolism. They are also not expected to be transporter substrates, including main efflux transporters (P-gp) or BCRP, or SLC (uptake) transporters¹⁰.

Transporter inhibition has, however, been described for certain oligonucleotide drugs¹¹. The FDA 2024 Oligonucleotide guidelines⁹, therefore, recommend conducting in vitro assessment of transporter-mediated DDI as described in the ICH M12 guidelines¹ (the 2024 document refers to the 2020 FDA DDI guidelines⁴ that have since been replaced by the ICH M121).

According to the FDA, precipitant interactions (inhibition, induction) with CYP450 enzymes are less common for oligonucleotides, however, omission of these assays from the DDI risk assessment study needs to be justified. Indeed, CYP450 enzyme induction by oligonucleotides has been described in several cases. Assessment of transporter induction may also be relevant.

Oligonucleotides often demonstrate high (> 95%) plasma protein binding (PPB)¹², which may affect their cellular uptake. Oligonucleotide PPB can be tested using the electrophoretic mobility shift assay (EMSA), where the drug molecules are co-incubated with the protein of interest and then run on a polyacrylamide gel. With appropriate staining, this allows separation of the protein-bound and unbound drug, and quantification via densitometry. The impact of PPB on DDI risk is likely limited, and the correction of assay results with measured unbound fraction data may not be necessary, the FDA still recommends generating and submitting PPB data as part of the DDI dataset⁹.

A crucial consideration for conducting in vitro tests with oligonucleotides is the identification of an appropriate and sufficiently sensitive bioanalytical method for their quantification. For the smallest molecules, such as ASOs, hybridization-ligation ELISA is often recommended. For larger oligonucleotides, like siRNA or modified RNA drugs, labelling the molecule with a complementary peptide nucleic acid labelled and subsequent fluorescence detection, as well as HPLC coupled MS/MS (ion pairing or reverse phase) is often a viable approach. Bioanalytical methods of all kinds are, however, evolving rapidly, and the selection of the best system for each project should to be considered on a case-by-case basis to match the needed sensitivity with the construct and relevant concentrations.

Consult an Expert