Search

Patient-derived xenografts (PDXs) model human intra-tumoral heterogeneity in the context of the intact tissue of immunocompromised mice. Histological imaging via hematoxylin and eosin (H&E) staining is performed on PDX samples for routine assessment and, in principle, captures the complex interplay between tumor and stromal cells. Deep learning (DL)-based analysis of large human H&E image repositories has extracted inter-cellular and morphological signals correlated with disease phenotype and therapeutic response. Here, we present an extensive, pan-cancer repository of nearly 1,000 PDX and paired human progenitor H&E images. These images, curated from the PDXNet consortium, are associated with genomic and transcriptomic data, clinical metadata, pathological assessment of cell composition, and, in several cases, detailed pathological annotation of tumor, stroma, and necrotic regions. We demonstrate that DL can be applied to these images to classify tumor regions with an accuracy of 0.87. Further, we show that DL can predict xenograft-transplant lymphoproliferative disorder, the unintended outgrowth of human lymphocytes at the transplantation site, with an accuracy of 0.97. This repository enables PDX-specific investigations of cancer biology through histopathological analysis and contributes important model system data that expand on existing human histology repositories. We expect the PDXNet Image Repository to be valuable for controlled digital pathology analysis, both for the evaluation of technical issues such as stain normalization and for development of novel computational methods based on spatial behaviors within cancer tissues. Citation Format: Brian S. White, Xing Yi Woo, Soner Koc, Todd Sheridan, Steven B. Neuhauser, Shidan Wang, Yvonne A. Evrard, John David Landua, R Jay Mashl, Sherri R. Davies, Bingliang Fang, Maria Gabriela Raso, Kurt W. Evans, Matthew H. Bailey, Yeqing Chen, Min Xiao, Jill Rubinstein, Ali Foroughi pour, Lacey Elizabeth Dobrolecki, Maihi Fujita, Junya Fujimoto, Guanghua Xiao, Ryan C. Fields, Jacqueline L. Mudd, Xiaowei Xu, Melinda G. Hollingshead, Shahanawaz Jiwani, PDXNet consortium, Tiffany A. Wallace, Jeffrey A. Moscow, James H. Doroshow, Nicholas Mitsiades, Salma Kaochar, Chong-xian Pan, Moon S. Chen, Luis G. Carvajal-Carmona, Alana L. Welm, Bryan E. Welm, Ramaswamy Govindan, Shunqiang Li, Michael A. Davies, Jack A. Roth, Funda Meric-Bernstam, Yang Xie, Meenhard Herlyn, Li Ding, Michael T. Lewis, Carol J. Bolt, Dennis A. Dean, Jeffrey H. Chuang. A pan-cancer PDX histology image repository with genomic and pathological annotations for deep learning analysis. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5407.

Patient-derived xenografts (PDX) are valuable, clinically-relevant models of cancer. Their close genomic, phenotypic, and temporal association with patient tumors makes them well-suited for pre-clinical and co-clinical studies that assess the potential of new therapeutics. However, PDX models are not amenable to large-scale drug sensitivity studies due to their high cost and low throughput capacity. To address this, we established and characterized long-term PDX organoid (PDxO) cultures from breast cancer (BC) PDXs and evaluated their utility in therapeutic studies. To establish PDxO culture conditions, we extensively tested medium supplements, including growth factors, kinase inhibitors, conditioned medium, and antioxidants, along with extracellular matrix composition in 3D gels. Using optimized culture conditions for each BC subtype, we established PDxOs from 16 PDX lines in the HCI series, with a PDX to PDxO success rate of 85%. Around 20% of PDxOs contained aggressive mouse stroma, which had to be eliminated by FACS for long-term culture success. PDxOs were maintained for over 1 year, during which we tracked viability, doubling time, organoid size, genomics, epithelial character, and tumorigenicity. Doubling times stabilized after 60 days of initial culture, with a mean of 6.4±1.7 days for triple negative (TN) lines and 7.2±1 days for ER+ lines. Mean organoid size remained stable, with ER+ PDxOs significantly smaller than TN PDxOs, at 91 and 262 cells/organoid respectively (p=0.0012). PDxOs histologically resembled their PDX counterparts when stained for H&E, Vimentin, and CAM 5.2. RNA-Seq and copy-number variation analyses showed that PDxOs clustered with their PDXs and patient tumors. To assess if culturing affected tumorigenicity, we re-implanted PDxOs into mice following early and late passaging. 5/5 early passage and 5/6 late passage PDxOs engrafted and formed tumors. Resulting tumors showed similar gene expression profiles as their original PDXs by RNA-Seq. These data suggest that PDxOs generally recapitulate the molecular and genomic features of their originating PDXs and patient tumors. Having established PDxO cultures, we evaluated their utility as a therapeutic screening platform. We developed a screen to differentiate compounds with cytotoxic and cytostatic effects. NCI CTEP and clinically-approved BC therapies (n=40) were screened in quadruplicate 8-point dose response curves on all 16 PDxOs across three biological replicates. Drug response profiles were stable across biological replicates, spanning up to 1 year in culture. PDxOs exhibited diverse responses to therapies targeting cIAP1, PI3K, and CHK1/2. Initial work to evaluate concordance between PDxO and in vivo PDX responses returned trending linear correlations between PDxO dose response data and change in PDX growth rate following treatment (R2 = 0.52-0.77; p = 0.045-0.12, across 3 PDXs and 9 compounds). Ongoing work aims to confirm concordance between PDX and PDxO models. Our work demonstrates that PDxO models are cost-efficient, easy to maintain, and grow indefinitely - making them renewable and accessible cancer models. PDxOs are a powerful parallel resource to PDX models, especially useful for efficient determination of PDX drug response. We are currently expanding our PDxO bank to include 100 models which will be deposited with the NCI as part of the PDXNet effort. Citation Format: Katrin P. Guillen, Sandra D. Scherer, Yi Qiao, Satya S. Pathi, James M. Graham, Maihi Fujita, Yoko S. DeRose, Jason Gertz, Gabor T. Marth, Katherine E. Varley, Alana L. Welm, Bryan E. Welm. Predicting breast cancer therapy response using a patient-derived xenograft organoid screening platform [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr LB-038.