Targeted therapies have been very successful in cancer treatment. Approved for a range of targets across many cancer types, they help prolong the life of cancer patients, hopefully with fewer side effects then treating with systemic chemotherapy.
However, it’s well known that the major drawback of targeted agents is the development of resistance, often quite quickly after treatment initiation. There are many known mechanisms of resistance to targeted agents, with the main 3 classes:
- Alterations to the targeted driver oncogene such as gene mutations, so that targeted agents can no longer bind and function
- Bypass of the inhibited target – activating the protein’s signaling pathway by a different route either downstream of the inhibitor or in parallel
- Activation of a different signaling pathway resulting in cell survival.
To make targeted agents more useful, we need to be able to develop drugs to delay or circumvent this resistance, which can often prove challenging.
Many Resistant Genotypes to Just One Drug
Focusing on point 1 above, one of the difficulties in developing drugs to overcome resistance is the complex and wide range of alterations that can occur in just one gene following just one treatment, and for which different follow-up or combination therapies may be needed.
A good example of this is in gastrointestinal stromal tumors (GIST) with the KIT gene. GIST affects around 4,000-5,000 people in the US each year and had a very poor prognosis before treatment was revolutionized by characterization of the genomic and molecular events driving tumor development.
It was found that about 80% of GIST patients harbor a KIT mutation within a range of exons, which confers increased sensitivity to KIT inhibitors. Imatinib mesylate (Gleevec™) was then introduced as a treatment option to inhibit KIT, and response and survival rates soared.
But, acquired resistance occurs quickly – around half of patients see resistance to imatinib within just 20 months of therapy. Resistance is linked to a range of secondary single nucleotide KIT mutations in exons including 13, 14, 17, and 18. These differing acquired mutation genotypes are known to respond differently to current second and third line agents, e.g. secondary mutation in exon 17 confers resistance to sunitinib, while exon 13 and 14 secondary mutations show sensitivity to sunitinib in vitro.
Preclinical Models are Needed for a Range of Resistance Genotypes
We need to be able to develop drugs, or find the best combination treatments, for a whole range of resistance genotypes and mechanisms. While the genetics of resistance to an agent like imatinib are well understood, there is then a lack of models for preclinical drug development which fully replicate all the disease backgrounds seen in the clinic. This is true for many targeted agents, and for all the resistance mechanisms covered above.
Conventional xenografts just aren’t available to cover all of the mutation genotypes available or alternate pathways activated, and with any immortalized cell line derived xenograft used there’s a chance of drift from original disease and resistance mechanism.
Patient-Derived Xenografts – Recapitulating Human Resistance Genotypes
PDX models are the perfect tool for developing drugs to overcome acquired resistance. Derived directly from primary tumor tissue they fully recapitulate the histopathological and genetic profiles of original patient tumors, including the complex interplay of mutations (both primary and secondary) seen in the clinical population, as well the other genetic and signaling changes driving resistance.
Developing a PDX panel from a range of patients allows the clinical diversity to be captured, not only for drug resistant models, but also those still responsive to agents for comparison, or one model from the same patient pre- or post-treatment resistance for genetic analysis.
Models can be fully RNA sequenced to reveal all genetic alterations, and how these correlate to response or resistance. Novel agents to overcome resistance can then be tested on individual models with highly specific clinically relevant genotypes of interest to a researcher, or across panels of models to see a population response and tease out how this relates to a range of mutations.
As PDX are highly predictive of patient response to treatment, efficacy data can be used to suggest exactly which treatments or combinations should be used for which set of patients, and potentially guide clinical trial design and stratification.
Get Ahead of Resistance - Develop your Own Model of Resistance during Preclinical Drug Development
Another way to use PDX in acquired resistance research is to develop your own model of resistance during preclinical drug development. Take a model which responds to your drug due to a specific genetic feature. You can then treat with consecutive cycles of your targeted agent, and see when and how resistance occurs.
Sequencing your model to see what new mutations, overexpression etc have happened can help identify the mechanisms of acquired resistance, and you can get ahead by testing other agents or initial/resulting combination regimens which can re-elicit a response. Again due to the highly predictive nature of PDX, mechanisms and results should be clinically relevant.
For anyone working in the targeted agent field, hopefully this blog shows that resistance is futile when you use PDX.