Cancer research is constantly evolving, seeking new methods to better understand and treat this complex disease. Among the most groundbreaking developments in this field are patient-derived xenograft (PDX) models. These models have revolutionized our approach to studying tumors by preserving the heterogeneity and molecular characteristics of a patient's tumor. This blog post aims to provide an in-depth look at the heterogeneity of PDX models and why it’s important to have diversity in your preclinical in vivo studies.

The Importance of PDX Models

Unlike traditional cell line-based models, PDX models retain the genetic diversity and molecular complexity of the original tumor. PDX models offer researchers a closer approximation of how human tumors behave in vivo, with different cell populations responding differently to therapies. They provide a dynamic and clinically relevant platform for studying tumor growth, metastasis, and response to treatment, making them a powerful tool in the fight against cancer. (Read our blog "The Ultimate Guide to Designing a Mouse Clinical Trial and Data Analysis")

Models pretreated with targeted therapies are ideal to test next-generation agents. For example, when developing a drug that aims to overcome resistance to the standard of care therapy in a given tumor type, choosing models pretreated with that standard of care agent for our efficacy study will reveal true therapeutic effects in the target patient population. (Read our blog "Accelerating Innovation & Drug Development with Pre-treated PDX Models")

Tumor Heterogeneity of PDX models is important for translational oncology studies

One of the most significant advantages of PDX models is their ability to preserve tumor heterogeneity. Tumors are not uniform; they consist of various cell types with different genetic mutations and behaviors. On the contrary, traditional cell lines are developed in vitro by clonal selection, leading to a less accurate representation of the human disease.[1]

PDX models also retain the molecular characteristics of the original tumor. This includes genetic mutations, epigenetic modifications, and gene expression patterns. By maintaining these features, PDX models provide a more accurate representation of the human tumor genetic background, which is crucial for developing effective therapies.[1]

Advancing Cancer Research with PDX Models

The ability of PDX models to preserve tumor heterogeneity and molecular characteristics has significantly advanced cancer research. Researchers can study the behavior of different tumor subpopulations, identify potential targets for therapy, and evaluate the efficacy of new treatments in a more realistic setting.

PDX models have heterogeneous biomarker expression. The figure below, for example, shows the expression of KRAS in Champions’ PDX models varying across and within tumor types. Selecting PDX models with a higher KRAS expression will be paramount to successfully test KRAS inhibitors and could provide more clinically relevant results.

Nonetheless, a PDX model with elevated KRAS expression will still present with heterogeneous expression across the different cell clones within the tumor, a setting that realistically reproduces the human tumor heterogeneous cell composition. In this setting, a KRAS inhibitor might only be partially effective at killing the tumor because cell clones with lower KRAS expression within the tumor might not be affected, resulting in lower overall therapeutic efficacy. Similarly, in the clinic, we could observe a partial response followed by tumor progression, due to the selection of the low KRAS-expressing cell clone within the tumor. For this reason, studies that use PDX models instead of cell line-derived models will produce more reliable and clinically relevant results.

KRAS gene expression across Champions' PDX models of different tumor types.

Patient-derived xenograft models have revolutionized cancer research by providing a more accurate representation of human tumors. Their ability to preserve tumor heterogeneity and molecular characteristics makes them an invaluable tool for studying cancer biology, testing new therapies, and personalizing treatment plans. By continuing to develop and refine these PDX models, researchers can accelerate the discovery of new treatments and bring them to patients faster.

Champions Oncology has over 1500 clinically relevant PDX models that can accelerate your oncology research program. Our models have broad clinical annotation and have been deeply characterized using NGS (WES & RNA-seq), proteomic, and phospho-proteomic datasets to interrogate the heterogeneity of the tumor. To learn more about our PDX tumor models, access Lumin Analytics.

In the complex landscape of oncology research and drug development, precision is paramount. The ability to select appropriate models can significantly impact the success and validity of your research. This is where metadata steps in as your reliable guide for model selection.

Understanding Metadata in Research

Metadata, essentially data about data, plays a crucial role in organizing, describing, and finding datasets. In the context of oncology research, metadata encompasses all the details about experiments, datasets, and results. This includes information on sample types, experimental conditions, data collection methods, and software used. By providing a structured summary of data, metadata helps streamline the process of data retrieval and utilization, for model selection and complex data analysis.

Metadata serves as a roadmap, guiding researchers to the datasets they need, while also providing context and ensuring that data is used appropriately. It’s an essential tool for maintaining consistency and transparency in research. The better structured and more detailed the metadata, the easier it is to manage and interpret large datasets, which are common in oncology.

In oncology research, metadata might include details such as the type of tissue samples collected, the sequencing methods used, and the specific conditions under which experiments were conducted. This ensures that datasets remain comprehensible and reusable, even years after the initial research was completed.

The Importance of Accurate Metadata in Model Selection

Accurate metadata is the backbone of effective model selection. It provides the necessary context to understand datasets fully, ensuring that models are selected based on relevant and precise information. This is particularly important in oncology research, where the accuracy of findings can have significant implications for patient outcomes.

Champions Oncology’s entire database can be explored accessing Lumin, our proprietary integrated software solution that harnesses the power of precisely curated oncology data, computational science, and bioinformatics to help you make data-driven decisions in model selection and data interpretation. Within Lumin you can:

• visualize proprietary genomic, proteomic, and pharmacological profiles from Champions’ patient-derived preclinical models;
• generate analyses on unique multi-omic datasets assembled from over 12,000 patients, including thousands of clinical treatment responses not available in public databases;
• interrogate experimental data and accelerate your oncology research programs.
  
  

Enhancing Model Selection

With accurate metadata, researchers can filter through vast amounts of data to find the most relevant datasets for their specific needs. For instance, if you're looking for data on a particular cancer type and drug target, well curated metadata can help you quickly identify and access the relevant datasets, saving time and resources.

Scientists planning to study the efficacy of a novel targeted therapy developed to treat tumors of a specific type that are resistant to the standard of care for that tumor type could use metadata to select the best set of PDX models to use in their study. Within Lumin, they could filter our bank of tumor models by tumor type first, and then, by drug response profile.

In the example below, we selected all NSCLC PDX models derived from patients who were pre-treated with osimertinib and that did not respond to osimertinib treatment in vivo. This set of models would well represent the target patient population for our drug.

Additionally, scientists could further select models based on gene mutation and RNA and/or protein expression levels of their agent’s target.

A deeper look into the tumor model metadata, leveraging data analysis and visualization tools, can help fine-tuning model selection further improving the quality of the study results.

Deep characterization and extensive clinical annotation of patient-derived tumor models, allows for a more precise model selection, which translates into a well-designed mouse clinical trial that closely reproduces the clinical setting and patient target population, and ultimately leads to more accurate results and a faster track to clinical experimentation.

Best Practices in Structuring Metadata

To harness the full potential of metadata, it's essential to follow best practices in its structuring and management. Here are some key guidelines for oncology researchers.

Standardized Formats

Using standardized formats for metadata ensures consistency and compatibility across different datasets and platforms. This makes it easier to integrate and compare data from various sources, enhancing the overall quality and reliability of your research.

Detailed Descriptions

Detailed descriptions of the datasets, including information on the origin of the data, the methods used for collection and analysis, and any relevant environmental or experimental conditions are very important. The more detailed your metadata, the more useful it will be for model selection and to bridge with future research.

Regular Updates

Keeping metadata up-to-date to reflect any changes or new findings is paramount. Regular updates ensure that data remains relevant and accurate, supporting the ongoing use and application of any research findings.

Future Trends in Metadata for Research and Biotech

The landscape of metadata is continually evolving, driven by advances in technology and increasing demands for data accuracy and transparency. Here are some future trends that are likely to shape the use of metadata in oncology research.

AI and Machine Learning

Artificial intelligence (AI) and machine learning are set to revolutionize the way we manage and utilize metadata. These technologies can automate the process of metadata generation and analysis, making it faster and more efficient. For example, AI algorithms can analyze large datasets to identify patterns and correlations that might be missed by human researchers.

Integration with Big Data

As the volume of data in oncology research continues to grow, there will be an increasing need for robust metadata systems that can handle big data. Integrating metadata with big data platforms will enhance the ability to manage, analyze, and interpret vast amounts of information, leading to more accurate and comprehensive research outcomes.

Metadata is an invaluable tool for oncology researchers, providing the context and structure needed to manage complex datasets effectively. By following best practices in model selection, you can enhance the accuracy and efficiency of your research, leading to better outcomes and discoveries. The incorporation of metadata into your research processes not only supports better model selection and data integrity but also fosters collaboration and innovation.

Ready to advance your oncology pipeline with faster model selection and AI-powered data analytics & visualizations? Sign up for our free trial of Lumin and discover how we can help you streamline your model selection and accelerate your discoveries.