Orthotopic PDX Models with Quantitative Pathology End Points in Translational Oncology Research

Orthotopic Patient-Derived Xenografts (PDX)

Despite the rise of personalized approaches to the treatment of cancer, translating these therapies into a clinical setting still poses an exceptional challenge. This is highlighted in drug development where research into targeted therapies requires highly characterized tools that mimic human tumors as closely as possible. Patient-derived Xenografts (PDXs) are one such tool whereby tumor cells from cancer patients are transplanted directly into immunodeficient mice where they grow into tumors resembling their origin (1). PDX models have become invaluable in pharmaceutical research and cancer treatment as they mirror the human tumor histology, clinical chemosensitivity, growth kinetics, and local disease progression of a patient’s tumor in vivo (2).

PDX models are used routinely in research; however, the implantation route of a tumor into immunodeficient mice can affect the tumor’s growth. When a tumor is transplanted into a mouse subcutaneously, or at a site other than that of the primary tumor, it often does not grow and metastasize as expected in a patient (1-3). These limitations have led to increasing applications of orthotopic PDX models created by transplanting the tumor into anatomical sites that correspond to the location of the primary human tumor (Figure 1). Certis Oncology Solutions has developed orthotopic PDX models that rapidly and accurately mimic human tumor growth, metastasis and response to drug therapies, providing invaluable insight to patient clinical outcomes (4).

Figure 1. Orthotopic PDX from Certis Oncology Solutions. Cells from a primary human pancreatic tumor are transplanted into an immunodeficient mouse at the corresponding anatomical location of the pancreas. This approach closely mimics human tumor growth, metastasis, and response to drug therapies. Image obtained from: https://certisoncology.com/about-certis/our-technique/

Orthotopic PDX models from Certis Oncology Solutions present numerous opportunities to create reproducible studies targeting human cancer development and treatment. Tumor growth or a reduction in tumor size can be precisely studied in real-time with advanced imaging techniques, generating quantitative measurements at high resolution. Orthotopic PDX models provide the opportunity to test the effects of chemotherapy or targeted drug therapy on the host tumor in an in vivo environment. For patients, this testing can help shield them from toxic and ineffective therapies, increasing overall clinical and disease-free patient survival. For drug companies, this approach provides characterized in vivo models for rapidly and reproducibly testing compounds at scale. Orthotopic PDX models replicate human tumors better than traditional PDX models and can be utilized to study drug-induced changes in tumor histology including the expression of specific protein and RNA biomarkers (2).

Quantitative Pathology End Points

Drug-induced changes in biomarker expression can be studied on a cell-by-cell basis using quantitative immunohistochemistry (IHC) or in situ hybridization (ISH). Quantitative IHC and ISH offer the advantage of being able to assess biomarker expression in the context of an intact tumor, revealing additional insights into tumor heterogeneity, inflammatory cell infiltration, and biomarker expression gradients. IHC utilizes antibodies for the detection of specific proteins (antigens) in tissue sections, whereas ISH employs specific probes to detect and visualize mRNA. Certis Oncology Solutions has partnered with Reveal Biosciences, a leader in computational pathology, to characterize tumors and provide high quality quantitative IHC and ISH data to their pharmaceutical clients.

Translational oncology researchers commonly seek quantitative data to assess cell proliferation, apoptosis, and vascularization in orthotopic PDX models. This data is provided by Reveal on a cell-by-cell basis using IHC for Ki67 (proliferation), cleaved caspase-3 (apoptosis) and CD31 (vascularization). A broad range of inflammatory cell markers is also available. Data is expressed in terms of the percentage of cells positive for each biomarker, together with a marker intensity measurement (Figure 2). Other data parameters including positive cells per square mm, subcellular compartment data (nuclear, cytoplasm, membrane), colocalization, proximity, and distance are also available, and high resolution whole slide images are provided. Reveal offers over 250 IHC biomarker protocols plus the ability to optimize new antibodies on request. These assays can be combined into panels targeting specific signaling pathways or drug target families to help elucidate the mechanism of action of drugs in vivo.

Figure 2. IHC image with image analysis mask overlay.

Tumor Profiling using Artificial Intelligence (AI)

Like self-driving cars and Amazon’s Alexa, the recent dramatic computation performance improvements have been game changing for digital pathology.  The industry is trending from supervised feature engineering driven by experts toward unsupervised feature discovery driven by large data sets on deeper networks.  Reveal Biosciences is leading the application of AI in digital pathology to generate a pipeline of decision support pathology diagnostics for specific disease verticals to benefit patients and healthcare providers. AI can also be applied to digital pathology in research for clinical trials, companion diagnostics, toxicology, and the analysis of tissue samples such as orthotopic PDX tumors. Reveal has developed AI-based tumor profiling models that accurately segment tumors into tumor, stroma, necrosis, and inflammatory cells based on a Hematoxylin and Eosin (H&E) or IHC images. Tumor profiling in this way can be combined with traditional IHC or ISH to evaluate biomarker expression patterns with respect to each segmented region, and analyze in-depth relationships between biomarker expression and drug treatment.

Summary

In summary, orthotopic PDX models replicate human tumors better than traditional PDX models and can be utilized to study drug-induced changes in tumor histology including the expression of specific protein and RNA biomarkers. This approach provides a new standard for translational oncology research.

References:

  1. Hidalgo, Manuel, et al. “Patient-derived xenograft models: an emerging platform for translational cancer research.” Cancer discovery (2014).
  2. Russell, Tara A., et al. “Clinical Factors That Affect the Establishment of Soft Tissue Sarcoma Patient-Derived Orthotopic Xenografts: A University of California, Los Angeles, Sarcoma Program Prospective Clinical Trial.” JCO Precision Oncology 1 (2017): 1-13.
  3. Hoffman, Robert M. “Patient-derived orthotopic xenografts: better mimic of metastasis than subcutaneous xenografts.” Nature Reviews Cancer 15.8 (2015): 451.
  4. “Our Technique.” Certis Oncology Solutions, 2018, https://certisoncology.com/about-certis/our-technique/
  5. “Perkel, Jeffrey M. “Immunohistochemistry for the 21st Century.” Science 351.6277 (2016): 1098-1100.

Pathology Revealed

Sign up to receive updates on the latest AI-powered pathology breakthroughs, access digital pathology resources, and more.