Often referred to as “magic bullets” or “biological missiles”, antibody-drug conjugates (ADCs) are one of the most promising advances in targeted cancer therapy, as they allow potent anti-cancer agents to be selectively delivered to cancer cells, while minimizing damage to healthy cells. This unique class of hybrid drugs typically contain a monoclonal antibody (mAbs) and a cytotoxic drug, attached via a chemical linker, combining “both the advantages of highly specific targeting ability and highly potent killing effect to achieve accurate and efficient elimination of cancer cells”. Despite their potential, ADC development faces significant challenges, including the complexity of selecting targets, optimizing linker chemistry, and predicting clinical efficacy.
In light of these challenges, patient-derived xenograft (PDX) models have emerged as invaluable tools in the preclinical ADC development process. PDX models are created by implanting tissue from patient tumors into immunodeficient mice. This means they retain the genotype and phenotype of the original tumor, including the complex architecture and cellular diversity. Modern PDX models have been called “one of the most faithful modeling systems in oncology” and they provide a clinically relevant platform for evaluating the efficacy and mechanisms of action of ADCs.
Latest advances in PDX models
In recent years, a number of advances in PDX modeling have further enhanced their utility and translatability and overcome some of the drawbacks associated with this type of model.
The development of PDX biobanks allows researchers to test drug candidates without the years of preparatory work of having to obtain, propagate, establish and store these models which was previously required. These libraries enable researchers to assess ADC efficacy across multiple cancer types and test multiple ADC candidates in parallel. Integrating these platforms with advanced imaging technologies allows researchers to monitor tumor responses in real-time and better understand ADC biodistribution in vivo.
Genomic characterization of PDX models has also evolved substantially alongside the rapid growth of PDX libraries. Major efforts to catalogue models, harmonize metadata, and organize repositories are ongoing and PDX Minimal Information standard guidelines have also been created. Comprehensive profiling is now becoming standard practice, including whole-genome sequencing, RNA sequencing, and proteomics analyses. This characterization is driving deeper insight into drug response or resistance to ADCs, the ways biomarkers can predict ADC efficacy, and the best strategies for patient stratification for clinical trials.
Additionally, the development of humanized PDX models has provided an even more relevant in vivo platform for immunotherapy evaluation. These humanized models are created by transplanting tumor tissue into mice with human immune systems (HIS). Their unique make-up means they can better recapitulate human tumor microenvironment (TME) and immune responses, so they are well-suited for ADC evaluation. As work continues to generate more comprehensive and functional immune systems within humanized mice, humanized PDX models are set to become “an unprecedented research platform in cancer immunology and personalized medicine”.
Using PDXs to model drug resistance
As ADCs combine an antibody with a cytotoxic payload, “multifactorial modes of resistance are emerging” that are inherent to their structure and function. Research has revealed many reasons for this resistance, including antigen downregulation, drug transporter over-expression, defects in ADC trafficking pathways and alterations in receptor, apoptotic, or other signaling pathways. As ADCs represent such an exciting new avenue in drug development, it’s essential that researchers identify and utilize the most-appropriate model at every preclinical stage to help overcome drug resistance. Crown Bioscience is a leader in the development of drug-resistant PDX models, providing these research teams with the specialized tools needed for this critical work.
PDX models are particularly useful for researchers trying to understand how patients acquire resistance and pinpoint exactly when resistance may develop in patients. To do this, researchers can create drug-resistant PDX models by treating PDXs that display specific genomic features with consecutive drug cycles. Once resistance has been acquired, teams can use these models to map the evolution of resistance mechanisms in real-time and utilize genetic sequencing to reveal pre- and post-treatment features. Subsequent testing of single or combination drugs can also be completed to determine which ones re-elicit a response.
Alternatively, PDX models can be developed from patients who have developed resistance to ADCs. These models offer highly clinically relevant insights into resistance mechanisms as they recapitulate the original tumor’s genetic and signaling alterations and histopathological and genetic profiles, including the complex associations seen between primary and secondary mutations. Models generated from the same patient, before and after they developed resistance, can be used for genomic and other -omics profiling technologies.
Where PDX models are most valuable in ADC development
The full pre-clinical ADC development pipeline relies on a range of model types, including cell lines, organoids, and PDXs. Research teams must adopt a strategic, holistic approach if they are to leverage the strengths of different models and mitigate their limitations. For example, during the early phases of ADC development, organoids can be established quickly, unlocking high-throughput screening with results within 2-3 months of propagation. Their three-dimensional cellular structures retain the form and functions of patient tumors, allowing researchers to understand more about the role of the tumor microenvironment, how ADCs behave within tumors, how they bind with target antigens, and how they might impact normal cells.
As PDX models are derived directly from patient tissue, they closely recapitulate this disease, making them particularly useful for hypothesis testing and supporting clinical development. In the later stages of preclinical development, PDX models prove most valuable when evaluating novel therapies and validating in vitro discoveries. Their unique make-up also offers insights into target distribution and accessibility within tumor microenvironments, efficacy across diverse patient samples, and off-target toxicity. It is PDX models that can bridge the gap between preclinical findings and clinical outcomes, guiding patient selection strategies and informing the design of clinical trials.
By making use of a suite of models, and choosing the right model at the right time, developers can prioritize drug candidates with potential and build confidence before clinical trials, helping to increase success rates.
Conclusion
In January 2025, Datroway (datopotamab deruxtecan) was approved by the FDA, bringing the total number of approved ADCs to 15. The development pipeline continues to grow; additional approvals are anticipated in 2025 and there are currently more than 350 ADCs in development. The potential of ADCs is wide-ranging as there is an almost unlimited number of possible combinations of target antigen, antibody, linker, and payload that can be utilized.
Advanced preclinical models will be key to the development of this next generation of ADCs. PDX models, already considered the new gold standard of preclinical therapeutic development, will be at the center of this innovation. Within the pre-clinical ADC development pipeline, PDX models are already playing an essential role, allowing researchers to model drug resistance, test hypotheses developed in vitro, and bridge the gap between preclinical findings and successful clinical outcomes. While recent advances, including the emergence of PDX biobanks, humanized PDX models, and extensive genomic profiling, are further enhancing the utility and translatability of these powerful tools.