Preclinical characterisation of antibody-drug conjugates - ADME perspective
23 August 2021
Antibody drug conjugate (ADC) comprise of a monoclonal antibody and cytotoxic payload conjugated by a linker. The antibody is purposed to deliver the payload to the target located on the cancer cell. The payload is typically a highly potent small molecule, which should be released from the carrier antibody inside the tumour cell. The number of the released but differently conjugated small molecules is affected by the used conjugation method and strategy.
Evaluating the ADME characteristics of ADCs forms an important part of their research and development and all the components of the ADC affect the ADME properties. With this specific modality of therapeutics, its essential to have scientific understanding of both aspects – biologics and small molecules. In this blog post, Dr Vesa Ruotsalainen and Dr Ari Tolonen will share their considerations about ADC characterisation.
What would be the typical first assays to be conducted for ADCs from the ADME perspective?
Stability of the ADC in plasma is essential and good to consider early enough in various species, because payload release should occur at the target site and instability in circulation may lead to increased systemic payload exposure and harmful off target effects.
Another point to consider is the disposition of intact molecule and its components; target mediated and catabolic clearance of mAb, release and distribution, metabolism and excretion of the released payload, i.e. PK of the ADC and the released drug, typically with mice or rats. The complexity of the ADC structure and possibility to go through various transformations will set additional challenges for the bioanalytical team in comparison to a standard PK assay with small molecules.
A possible phenomenon for biologics is anti-drug-antibody formation (ADA). However, apart from the antibody portion of the ADC, ADAs can potentially be directed towards the linker and linker-payload structure as well. Hence ADA formation should be evaluated, as that might affect the PK, efficacy, and safety profiles.
ADCs are essentially prodrugs. The antibody part of the molecule is eliminated by catabolism, but the payload and linker-payload are eliminated either directly or metabolised. Systemic levels of active payload are typically low but possible, for example due to off target release via non-specific catabolism in the liver. Therefore, one should also conduct the typical small molecule ADME evaluation for the payload only, especially if the compound is an NCE.
Can you apply similar experimental in vitro settings for ADCs than for small molecules when investigating their in vitro metabolism?
Important aspects to be studied include evaluating ADC payload release and release mechanism and identifying payload metabolites. However, there are points, which should be bear in mind when planning the experimental designs. The most critical being selection of the optimal metabolic system.
Lysosomal preparations mimic the ADC processing inside the target cell and can be used to study the payload release from ADC. However, they lack the typical drug-metabolizing enzymes (e.g., CYPs and UGTs) and cannot therefore be used for identification of further metabolites that are formed from the released payload. Liver S9 fractions instead contain all major metabolic enzymes, and their use is not dependent on active or passive transport like in hepatocytes. Furthermore, by selecting the pH appropriately, S9 fractions can be used to study both payload release and metabolism. Acidified liver S9 fractions mimic the acidic environment of lysosomes, that are the site of degradation for ADCs. At neutral pH, S9 fractions can be used to study the metabolism of the payload and linker-payload. S9 fractions can therefore be applied to intact ADC, linker-payload and payload molecule.
What are the analytical challenges related to ADCs and how those can be addressed?
Typical ADC is a heterogenous mixture of antibody, to which multiple payload molecules are conjugated with different drug-antibody ratio (DAR). Antibodies are very large (MW 150 000) biological drugs produced in cell cultures and are already heterogenous due to their biological origin. Antibodies have a vast number of lysine residues and relatively random conjugation of payload to the lysine sidechain epsilon amino group adds to the complexity of the molecule. Analytically, the ADC bioanalysis can be divided to techniques typically applied for biological drugs (LBA methods, LC-MS utilizing surrogate peptides) and small molecule drugs (LC-MS), as multiple analytes need to be measured; total antibody (unconjugated and conjugated antibody), ADC (conjugated and partially conjugated antibody) and payload (conjugated and unconjugated). Determination of DAR is an example of a new analytical method developed specifically for ADC molecules.
How do you determine the drug-antibody ratio (DAR) and why it is important?
As DAR may have effect on the efficacy and PK it is important to control and measure the DAR of ADC. Typically, incorporation of the payload affects the hydrophilicity/hydrophobicity of the ADC. This is utilized in hydrophobic interaction chromatography where more hydrophobic molecules are retained more strongly and number of ADCs with different amounts of payloads can be determined from the UV-chromatogram. In a more straightforward approach, the DAR of ADC can be determined using LC-HR/MS. In this approach, the intact protein is often deglycosylated prior LC-HR/MS analysis. After data processing, the number of payloads in the antibody can be identified and average DAR calculated utilising the deconvoluted mass spectra.
When conducting preclinical PK studies with ADCs, are there any specific points to consider?
There are few points that should be kept in mind. First, the mechanisms how the payload reaches its target are dependent on the ADC receptor internalisation. Cross-reactivity of mAb with the target in the preclinical species can affect the PK and distribution. In the case the antibody is not cross-reactive in the preclinical species, then the PK may not be reflecting the PK in species expressing the target. As said also earlier, DAR can affect the PK profile and should be characterised. As an example, high amount of payload may increase the hydrophobicity of the ADC altering the PK properties. It would be good to consider also total antibody, released payload and ADC over the experiment.
What is your take home message?
If you are not sure how to approach the ADME profiling of your ADC, it might be good idea to set a call with an expert CRO on the field to discuss about the project status and what would be the reasonable next steps.
Written by Vesa Ruotsalainen, Ari Tolonen and Miia Kovalainen