Abstract B35: Molecular and phenotypic profiling of pancreatic cancer cachexia in novel murine models and patients

Abstract

Abstract Background: Patients with pancreatic ductal adenocarcinoma (PDAC) suffer the highest rates of cachexia, with ~85% experiencing muscle and fat loss. Weight loss and muscle loss correlate with mortality in PDAC. Patients with cachexia experience fatigue, weakness and reduced quality of life. Cachexia impairs response to surgery, chemotherapy and radiation therapy. Currently the only effective therapy is removal of the tumor. However, pre-clinical studies demonstrate that slowing muscle or fat loss prolongs function and life, even absent effects on tumor growth. Thus anti-cachexia therapies should increase quality of life, clinical response and survival. Most cachexia research has been conducted using two classic non-PDAC mouse models, LLC lung and C26 colon cancer cell lines implanted s.c. Little genomic data are available in patients, with only 3 small gene expression studies in cachexia, none specific to PDAC. Thus whether mouse models faithfully reflect human PDAC cachexia is unknown. Methods: Mice: Cell lines derived from the KrasG12D;Trp53R172H;Pdx1-Cre (KPC) genetically engineered mouse model (GEMM) were injected orthotopically. Elastase-Cre and Pdx1-Cre KPC mice were generated. Mice were followed for tumor growth and metastasis, weight loss, body composition, strength, and muscle and fat wasting versus matched controls. A subset of models was characterized for serum cytokine levels and muscle transcriptomes. Patients: Patients undergoing surgery for PDAC (n=24) or benign conditions (n=12) donated clinical data, blood, muscle, subcutaneous adipose tissue and tumor (PDAC). Body composition was measured from CT scans. Whole blood, muscle and fat gene expression were profiled using the Ion AmpliSeq Transcriptome Panel and analyzed with Partech, Ingenuity, GSEA and NextBio. Results: Mice: We validated 9 novel murine models of PDAC cachexia. These models demonstrated considerable diversity in cachexia severity (latency to cachexia), body composition, tumor histology and metastasis, blood cytokine levels, and muscle gene expression. Orthotopic PDAC cell line and KPC GEMM models were more similar to each other than to LLC or C26 by phenotype and muscle gene expression. IL-6 and Activin, known to be elevated in human PDAC cachexia, were consistently elevated across mouse models. Inhibition of either reduced PDAC cachexia. Patients: PDAC and control groups had similar BMI, although 6-month history of weight loss and BMI-adjusted weight loss category were both greater for the PDAC group (p<0.0001). PDAC patients displayed reduced skeletal muscle and subcutaneous adipose tissue volume (p<0.05). Ampliseq (false discovery rate <0.05) revealed 23 differentially expressed genes in whole blood, 906 in muscle, and 2962 in adipose tissue, with ~250 overlapping genes in muscle and fat. The top predicted biofunctions increased by PDAC in both tissues were organismal death, followed by anomaly of musculoskeletal system, and hypoplasia; the most reduced biofunctions were cell movement, migration, and size of body. The top common canonical pathways increased were EIF2 signaling, and G1/S checkpoint regulation, while the most decreased were signaling from integrins, RhoGTPases, and leukocyte extravasation. Comparing mouse muscle to human muscle, “organismal death” was oppositely regulated and decreased in mouse. However, common pathways included EIF2 signaling and G1/S checkpoint regulation (increased) and cardiac hypertrophy signaling (decreased). At the gene level, substantial differences were observed between mouse and human muscle. Conclusions: Mouse models of PDAC are superior for modeling PDAC cachexia. In patients, the muscle and fat transcriptomes show a cachexia signature even in early stage (resectable) disease. Comparative molecular phenotyping of mouse and human PDAC cachexia reveal considerable overlap in pathways, albeit clear differences in gene level analysis. Thus careful characterization and application of models is warranted. Citation Format: Teresa A. Zimmers, Yanlin Jiang, Jianguo Liu, Eugene Ceppa, Atilla Nakeeb, Michael House, Nicholas Zyromski, C. Max Schmidt, Katharyn Hannaford, Daniel Schloss, Cynthia Brooks, Milan Radovich, Yunlong Lui, Guanglong Jiang, Stephen F. Konieczny, Patrick Schweickert, Brad Hancock, Joshua Kayes, Xiaoling Zhong, Christophe Poirier, Ernie Au, Andrea Bonetto, Safi Shahda, Marc Kohli, Bert O’Neil, Marianne Pons, Leonidas Koniaris.{Authors}. Molecular and phenotypic profiling of pancreatic cancer cachexia in novel murine models and patients. [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Advances in Science and Clinical Care; 2016 May 12-15; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2016;76(24 Suppl):Abstract nr B35.