Data from A Pan-Cancer Patient-Derived Xenograft Histology Image Repository with Genomic and Pathologic Annotations Enables Deep Learning Analysis

• DOI: 10.1158/0008-5472.c.7311385.v1
• Publication Date: 2024-07-02T07:44:40Z
• Authors: Brian S. White; Xing Yi Woo; Soner Koc; Todd Sheridan; Steven B. Neuhauser; Shidan Wang; Yvonne A. Evrard; Li Chen; Ali Foroughi pour; John D. Landua; R. Jay Mashl; Sherri R. Davies; Bingliang Fang; Maria Gabriela Rosa; Kurt W. Evans; Matthew H. Bailey; Yeqing Chen; Min Xiao; Jill C. Rubinstein; Brian J. Sanderson; Michael W. Lloyd; Sergii Domanskyi; Lacey E. Dobrolecki; Maihi Fujita; Junya Fujimoto; Guanghua Xiao; Ryan C. Fields; Jacqueline L. Mudd; Xiaowei Xu; Melinda G. Hollin...; Shahanawaz Jiwani; Saul Acevedo; Brandi N. Davis-D...; Peter N. Robinson; Jeffrey A. Moscow; James H. Doroshow; Nicholas Mitsiades; Salma Kaochar; Chong-xian Pan; Luis G. Carvajal-...; 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. Bult; Dennis A. Dean; Jeffrey H. Chuang
• Keywords: Histology

<div>Abstract<p>Patient-derived xenografts (PDX) model human intra- and intertumoral heterogeneity in the context of intact tissue immunocompromised mice. Histologic imaging via hematoxylin eosin (H&E) staining is routinely performed on PDX samples, which could be harnessed for computational analysis. Prior studies large clinical H&E image repositories have shown that deep learning analysis can identify intercellular morphologic signals correlated with disease phenotype therapeutic response. In this study, we developed an extensive, pan-cancer repository >1,000 paired parental tumor images. These images, curated from Development Trial Centers Research Network Consortium, had a range associated genomic transcriptomic data, metadata, pathologic assessments cell composition, and, several cases, detailed annotations neoplastic, stromal, necrotic regions. The amenability these images to was highlighted through three applications: (i) development classifier regions; (ii) predictor xenograft-transplant lymphoproliferative disorder; (iii) application published microsatellite instability. Together, provides valuable resource controlled digital pathology analysis, both evaluation technical issues image–based methods make predictions based treatment studies.</p><p><b>Significance:</b> A patient-derived xenograft eosin–stained will facilitate cancer biology investigations histopathologic contributes important system data expand existing histology repositories.</p></div>