A patient-derived orthotopic xenograft model enabling human high-grade urothelial cell carcinoma of the bladder tumor implantation, growth, angiogenesis, and metastasis

Jessie Gills,Maureen Shuh,Li Li,Jakob Reiser,Grace Maresh,Xin Zhang,Linh Hellmers,Michael P Marino,Stephen Bardot,M’Liss Hudson,Marc Matrana,Eric Laborde,Ravan Moret,Shams Halat,Daniel Canter,John Nelson,Maria Latsis,Sunil Talwar

doi:10.18632/oncotarget.26024

Abstract

High-grade urothelial cell carcinoma of the bladder has a poor prognosis when lymph nodes are involved. Despite curative therapy for clinically-localized disease, over half of the muscle-invasive urothelial cell carcinoma patients will develop metastases and die within 5 years. There are currently no described xenograft models that consistently mimic urothelial cell carcinoma metastasis. To develop a patient-derived orthotopic xenograft model to mimic clinical urothelial cell carcinoma progression to metastatic disease, the urothelial cell carcinoma cell line UM-UC-3 and two urothelial cell carcinoma patient specimens were doubly tagged with Luciferase/RFP and were intra-vesically (IB) instilled into NOD/SCID mice with or without lymph node stromal cells (HK cells). Mice were monitored weekly with bioluminescence imaging to assess tumor growth and metastasis. Primary tumors and organs were harvested for bioluminescence imaging, weight, and formalin-fixed for hematoxylin and eosin and immunohistochemistry staining. In this patient-derived orthotopic xenograft model, xenograft tumors showed better implantation rates than currently reported using other models. Xenograft tumors histologically resembled pre-implanted primary specimens from patients, presenting muscle-invasive growth patterns. In the presence of HK cells, tumor formation, tumor angiogenesis, and distant organ metastasis were significantly enhanced in both UM-UC-3 cells and patient-derived specimens. Thus, we established a unique, reproducible patient-derived orthotopic xenograft model using human high-grade urothelial cell carcinoma cells and lymph node stromal cells. It allows for investigating the mechanism involved in tumor formation and metastasis, and therefore it is useful for future testing the optimal sequence of conventional drugs or the efficacy of novel therapeutic drugs.

Highlights

In the United States in 2018, there will be an estimated 81,190 new patients and 17,240 cancer specific deaths attributable to urothelial cell carcinoma (UCC) of the bladder [1, 2]
To develop a patient-derived orthotopic xenograft model to mimic clinical urothelial cell carcinoma progression to metastatic disease, the urothelial cell carcinoma cell line UM-UC-3 and two urothelial cell carcinoma patient specimens were doubly tagged with Luciferase/ RFP and were intra-vesically (IB) instilled into NOD/SCID mice with or without lymph node stromal cells (HK cells)
We describe the development of a novel patient-derived orthotopic xenograft model (PDOX) model in NOD/SCID mice that reproduces metastatic UCC under the influence of Lymph node stromal cell (LNSC)

Summary

Introduction

In the United States in 2018, there will be an estimated 81,190 new patients and 17,240 cancer specific deaths attributable to urothelial cell carcinoma (UCC) of the bladder [1, 2]. 50% of patients with MIUCC of the bladder, despite attempted curative therapy, e.g., radical cystectomy (RC) with or without systemic chemotherapy, for clinically localized disease, will still develop metastases and die within 5 years [1]. Five-year survival after RC alone for node positive disease is less than 35% at best; making nodal involvement an important negative predictor of patient survival. Apart from adjuvant chemotherapy, there exist few therapeutic options in the advanced and/or metastatic setting, creating a need for novel therapies with better patient-specific treatment options. Clinical trials of immunotherapies in UCC of the bladder have become available for selected patients [12]. Immunotherapies are limited by a low response rate and the challenge of targeted patient selection [13]

Methods

Results

Conclusion