PDX model

In cancer research and drug discovery, accurately replicating the clinical reality is a crucial challenge. PDX models are gaining attention in this regard. This article explains the role of PDX, covering its basics, latest trends, and strategic applications.

Why the PDX model now?

In recent years, with the advancement of cancer research and drug discovery, the importance of models that more closely resemble the clinical reality has been increasing. While CDX models, which have been widely used traditionally, offer excellent reproducibility due to being based on homogeneous cell lines, they have faced challenges in sufficiently reflecting the tumor heterogeneity and tumor microenvironment inherent in patient tumors. In contrast, PDX models directly transplant patient-derived tumor tissue,Evaluation is possible while maintaining the diversity and microenvironment of cancer.Therefore, it is attracting attention as a more reliable model for elucidating drug response and resistance mechanisms, and for its application in personalized medicine.

Standard Review: Foundational Literature that Established the PDX Paradigm

The utility of PDX models has been systematically demonstrated through extensive basic research, making them a crucial foundation for current drug discovery and cancer research. Here, we will briefly summarize the significance and findings of representative papers that have greatly contributed to the establishment and development of the PDX concept.

[Standard 1] Standardization of PDX Models and Clinical Correlation

• Title: Patient-derived tumor xenografts as models for oncology drug development.
• Author/Journal: Tentler JJ, et al. / Nature Reviews Clinical Oncology

This review by Tentler et al. isA monumental paper positioning the PDX model as a practical foundation for oncological drug discoveryThis study systematically organized the extent to which drug sensitivity and correlation with biomarkers in clinical settings could be reproduced using patient-derived tumors, thereby clearly demonstrating the usefulness of PDX models as predictive models. Furthermore, by comparing them with conventional CDX models, the study emphasized their advantages, such as maintaining intratumoral heterogeneity and molecular biological characteristics. By addressing crucial experimental design factors, including the effects of implantation site and passage, and presenting guidelines for the standardization of PDX research, it has become a foundational literature widely referenced to this day.

[Standard 2] Insights from large-scale PDX consortia (e.g., EurOPDX)

• High-throughput screening using patient-derived tumor xenografts to predict clinical trial drug response.
• Author/Journal: Gao H, et al. / Nature Medicine

Gao et al.'s paperA representative study demonstrating the effectiveness of high-throughput drug screening using a large-scale PDX collectionWe systematically analyzed PDXs derived from diverse cancer types and comprehensively obtained drug response patterns, demonstrating the potential to accurately predict response rates and their correlation with biomarkers in clinical trials. Significantly, we overcame the limitations of single models in capturing tumor heterogeneity by treating them as a large-scale panel, thereby reproducing the variability in patient responses. This research validates the significance of standardization and shared platforms by international consortia such as EurOPDX and presents a new paradigm for preclinical studies using PDXs.

Latest Review: Trends Accelerating Next-Generation Drug Discovery (2023-2025)

Recent reviews have redefined PDX models as more than just preclinical models, positioning them as foundational technologies supporting precision medicine and next-generation drug discovery. Especially since 2023, advancements in integration with organoids and AI analysis, along with the construction of large-scale data platforms, have clarified a new trend of more accurately reproducing and utilizing tumor complexity.

[Trend 1] Humanized PDX (Patient-Derived Xenograft)

• Key point: Mice with reconstituted human immune systems are essential for evaluating immunotherapies such as immune checkpoint inhibitors.
• Recommended Literature: Recent advances in humanized mouse models for cancer immunotherapy. (Review from around 2024)

In recent reviews,Humanized PDX models are an indispensable foundation for immunotherapy researchit is positioned as. Conventional immunodeficient mice could not replicate human immune responses, and had limitations, especially in evaluating immune checkpoint inhibitors. However, models with reconstituted human immune systems have significantly improved this challenge. By transplanting human hematopoietic stem cells or peripheral blood cells, the dynamics of various immune cells, including T cells and NK cells, can be reproduced *in vivo*, enabling the analysis of immune interactions within the tumor microenvironment.

Furthermore, in recent years, the development of genetically engineered mice expressing cytokines such as human IL-15 has enhanced the maturation and maintenance of NK cell function, improving the reproducibility of physiological immune responses. These models are also applied to predict immune-related adverse events and biomarkers, significantly contributing to the improved accuracy of preclinical evaluation in the development of immunotherapies.

[Trend 2] PDX x Spatial Omics x AI

• Key point: Analysis of the vast amount of data obtained from PDX.
• Recommended literature: Integrating patient-derived models with spatial transcriptomics and AI for precision oncology.

In recent reviews, to comprehensively interpret the massive and high-dimensional data obtained from PDX models,The fusion of spatial omics and AI is an important trend.It is positioned as. In particular, spatial transcriptomics can obtain gene expression along with positional information within tissues, making it possible to visualize tumor heterogeneity and the structure of the tumor microenvironment in detail. While such data is extremely complex, the use of AI and deep learning is advancing the extraction of patterns related to cell-cell interactions and drug responses, the search for biomarkers, and the construction of treatment prediction models.

The integration of new AI methods, such as graph neural networks and foundation models, is accelerating the application of precision medicine that unifies spatial and molecular information. PDX plays a crucial role as an experimental system to demonstrate these analyses.

Disease Area Specific Major Review List

PDX models are highly valuable for disease-specific research because they can reflect the distinct biological characteristics and treatment challenges of each cancer type. Here, we outline key discussion points for review and perspectives on utilizing PDX for drug discovery and pathological analysis for each major cancer area.

From a strategic perspective

In utilizing PDX models, the high clinical reproducibility comes at the cost of significant constraints in terms of expense and the time required for establishment and evaluation. Therefore, in recent years,A hybrid strategy using patient-derived organoids (PDXOs) that allow for rapid and high-throughput evaluation is becoming mainstream.This is becoming a reality. By efficiently screening candidate drugs with PDXO and verifying only promising conditions in PDX, we can optimize the balance between accuracy and efficiency. This integrated approach is gaining attention as a realistic strategy to accelerate drug discovery decisions while achieving clinically relevant results.

Summary

PDX models are gaining importance due to their ability to retain tumor heterogeneity and the tumor microenvironment, which are not fully captured by conventional CDX models, and their capacity to reproduce clinically relevant responses. Furthermore, the advancement of humanized PDX, spatial omics, and AI analysis is furthering our understanding of immune responses and tumor structures. On the other hand, co-utilization with PDXO is considered effective for addressing cost and time constraints, and their strategic application is therefore called for. PDX models are positioned not merely as models to evaluate "efficacy," but as foundational tools to elucidate "why a drug works/doesn't work" and to translate these findings back into clinical practice.