When considering cancer therapies, two classes of compounds are often discussed: monoclonal antibodies (mAbs) and potent cytotoxic drugs. Both classes of medicines have their own advantages but can also present diverse side effects.
mAbs are highly specific, typically working by blocking growth of cancer cells, and have the added advantage of boosting the patient’s overall immune system. They are not, however, considered as effective at destroying large tumours as the highly potent cytotoxic drugs. These potent drugs are considered to be very effective against large tumours, but are non-selective and simultaneously destroy healthy cells, causing unacceptable collateral damage to the patient.
The premise of an antibody drug conjugate (ADC) is based on combining the specificity of a mAb with the potency of the anticancer drug. This produces a therapeutic agent capable of delivery of the anticancer drug, specifically to the physiological target of concern, without causing any damage to normal cells, thereby lowering the potential for toxic side-effects and increasing the therapeutic window.
ADCs offer great potential, but despite their obvious advantages, there have only been three such approved compounds to date. One of these – Pfizer’s Mylotarg® – has since been voluntarily withdrawn leaving only two ADCs currently on the market: brentuximab vedotin (Adcetris®) and ado-trastuzumab emtansine (Kadcyla™) (1).
Over recent years, advances in protein engineering and conjugation chemistry have facilitated a marked increase in the development of ADCs, with the number of Investigational New Drug submissions in the last five years approximately double those in the preceding 15 years (2).
Conjugation
Conjugation chemistry is aimed at producing an ADC via the use of a chemical linker to join the potent drug to the mAb. This linker facilitates the covalent attachment of the drug to the mAb with linkage occurring through either conjugation with the protein – typically made through lysine or interchain cysteine – or through the carbohydrate structure of a glycoprotein. The linker does not induce any activity in the ADC, but its design and mode of operation is critical (3).
Characterisation
When considering physiochemical characterisation, the ADC is regarded as three distinct species: the mAb, the cytotoxic agent and the linker. Given the increase in structural complexity of the conjugate, the mAb is characterised pre-conjugation in accordance with appropriate regulatory guidelines (4, 5).
ADCs contain all of the ‘normal’ heterogeneity features associated with a mAb, such as aggregates, deamidation products, glycosylation variants and oxidised amino acid side chains, as well as amino and carboxyl terminal amino acid additions. There is, however, the added complication of variability introduced by the numeric distribution of drug linked forms (6).
Consequently, in terms of complete ADC-specific analytics, there is a requirement for reliable analytical methods to evaluate the distribution of drug linked forms, expressed as the drug antibody ratio (DAR), as well as the payload, which is the average number of drug molecules attached to the mAb. In addition, assessment of the actual attachment positions, assessment of stability of the ADC and quantification of the amounts of free mAb and drug (in addition to ‘normal’ mAb process-related impurities) are of high importance.
DAR Analysis
When considering determination of the DAR, the choice of method is mainly dependent on the type on the linkage involved, but typically the approach is based on mass spectrometry (MS) (matrix-assisted laser desorption/ionisation-time of flight and/or electrospray ionisation), chromatography, or ultraviolet (UV) spectroscopy methodologies.
UV absorption methods are simple, but the results are complicated due to similarities within the extinction coefficients of the antibody and the drug molecule (7).
MS methods present alternative issues in that when deducing distribution, the calculation is based on the assumption of equivalence in ionisation of each of the possible conjugated forms. This may not however be a safe assumption if, for example, conjugation involves positively charged amines, and hence the potential for changes in charge and hydrophobicity, which can affect the ease of ionisation and distorting the apparent distribution (7).
As with many methods designed to assess the physiochemical characterisation of biomolecules in determining DAR or payload, it is perhaps best to consider orthogonal methods of analysis and combine MS or UV methods with chromatographic-based approaches, such as size exclusion chromatography (SEC) incorporating refractive index and UV detection, or hydrophobic interaction chromatography (HIC) (8).
Site Attachment
Perhaps one of the most difficult aspects of an ADC to evaluate is in relation to the confirmation of the site or sites of attachment. For example, consider a situation where the DAR analysis indicates six linker/ drug molecules per mAb on a mAb which has a potential of 12 sites of possible attachment. In assessing the position of linkage, it is therefore possible that there are multiple distribution forms, and in extreme cases a drug linker molecule, to some extent, can present on all 12 possible sites.
To evaluate site attachment, peptide mapping or sequencing studies are performed, but the data produced is often highly complex and requires a considerable amount of interpretation by highly trained analysts. There are, however, unique features of ADCs which can aid this process. For example, in a lysine conjugated ADC, the addition of the linker drug to the mAb backbone inhibits the action of trypsin. Thus the absence of a tryptic cleavage in the conjugated compared to its unconjugated form can be characteristic of conjugation in the peptides lysine.
Assessing Stability
Assessing the stability of the ADC presents further complexity, but is a key criterion in product selection and should be considered in three ways. Firstly, there should not be any significant structural change in the mAb ‘purity’ throughout the conjugation process. The ADC is still prone to all the stability issues connected with an unconjugated mAb, such as aggregation and fragmentation. Chromatographic (SEC, reversed phase high-performance liquid chromatography and ionexchange) and electrophoretic (sodium dodecyl sulfate, native-polyacrylamide gel electrophoresis and isoelectric focusing) methods used to assess aggregation and fragmentation in the unconjugated mAb should be applied to the ADC. Usually methods that are applied to the unconjugated mAb can be directly applied, or readily adapted. In addition to these structural ‘normal features’, it is also crucial to assess changes in the conformation of the mAb during conjugation, which are possible, if linkage is specifically made through lysine (1).
Circular dichroism is perhaps the most accepted and reliable technique used to evaluate tertiary structure through evaluation of the near UV profile. However, many of the cytotoxic drugs associated with ADCs absorb strongly in this region, which can cause interference and add ambiguity in the data interpretation process.
Stable Properties
Secondly, in relation to stability, the selectivity and target binding properties of the mAb must be maintained to ensure the continued specificity during the conjugation process. To assess the effect on biological activity of the mAb during conjugation, cell-based potency assays for assessment of antigen binding are applied.
The major benefit of the ADC is its ability to deliver a potent drug to the tissue/organ of interest. If the drug is released prior to reaching the target, there is a two-fold risk: the drug could cause unanticipated toxicities; and, as the plasma half-life of mAbs can exceed two weeks, the target tumour could be saturated with unconjugated mAb if the ADC breaks down, which may be ineffective or compete with subsequent ADC binding (6).
For the final stability assessment it is therefore critical to the product’s success to establish that a conjugate is stable once in systemic circulation. As such, in vitro and in vivo tests are required to establish the stability of the drug linker in plasma or serum.
As a final analytical challenge, the levels of free mAb and/or drug in the final ADC should be quantified, typically using chromatographic methods. These residual determinations add to those that are normally associated with the process – for example, antibiotics, stabilising agents and antifoaming agents. In most cases, the methods used for the unconjugated mAb can be adapted.
Summary
To conclude, from an analytical perspective, the characterisation of ADCs presents many unique and complex challenges. Overcoming these issues and being able to support development and registration programmes undoubtedly plays a critical part in the continued growth of interest in this class of potentially beneficial and specific therapeutics, paving the way for this type of personalised medicine.
References
1. Pfizer voluntarily withdraws cancer treatment Mylotarg from US market, 2010. Visit: www.fda.gov/newsevents/newsroom/pressannouncements/2010/ucm216448.htm
2. Marjorie AS and Xiao-Hong C, Regulatory considerations when developing assays for the characterization and quality control of antibody-drug conjugates, American Laboratory, 2012
3. Wakankar A et al, Analytical methods for physiochemcial characterisation of antibody drug conjugates, MABS 3(2): pp161-172, 2011
4. ICH guideline Q11 on development and manufacture of drug substances, chemical entities and biotechnological/ biological entities, 2012
5. Guidelines on development, production, characterisation and specifications for monoclonal antibodies and related products, EMEA/CHMP/BWP/157653/2007, 2008
6. Beck A, Wurch T, Bailly C and Corvaia N, Strategies and challenges for the next generation of therapeutic antibodies, Nat Rev Immunol 10(5): pp345-352, 2010
7. WYATT Technical Note: Antibody Drug Conjugate Analysis (ADC). Visit: www. wyatt.com/files/literature/antibody_ drug_conjugate_(ADC)_analysis.pdf
