Login to Access Video or Poster Abstract: MP73-06
Sources of Funding: AMOSO, CUASF

Introduction

Next generation preclinical models of renal cell carcinoma (RCC) now offer the ability to pre-determine de novo drug resistance in fresh patient tumor samples prior to targeted therapy. Implantation of tumor specimens into the chorioallantoic membrane (CAM) of the chicken embryo results in high engraftment efficiencies within two days, permitting large scale &[Prime]tumor avatar&[Prime] studies. Functional tumor heterogeneity studies, which can be performed in context of drug resistance within two weeks, can guide the selection of drugs and predict drug resistance outcomes for RCC patients. This ultrafast PDX model is mirrored by high-frequency ultrasound imaging that permits quantification of tumor volume and tumor vascularity in a high-throughput manner. _x000D_

Methods

Several core biopsies were extracted from primary tumors and metastases from clear cell RCC, chromophobe RCC, and type 1/2 papillary RCC patients and submitted to xenografting into the CAM of chick embryos. At least 6 regions of the primary tumor were xenografted and 3 metastases were xenografted. At least N>36 per region was submitted to xenografting with half of these treated with sunitinib or vehicle (DMSO). At T=10 days post-implantation, high frequency ultrasound imaging was used to quantitate tumor vascularity, tumor volume, tumor blood flow and and tumor blood volume. After imaging, total exome sequencing was performed to identify any genetic mutations for correlation to drug resistance.

Results

Using this &[Prime]tumor avatar&[Prime] model paired with a prospective RCC patient cohort, we observe intratumoral functional heterogeneity in the context of sunitinib treatment, as determined by high-frequency ultrasound imaging, highlighting its potential interventional role in the clinic. Exome sequencing and gene copy number variation analysis did not reveal DNA mutation signatures that were associated with resistance to sunitinib treatment within this intratumoral set of PDX biopsies.

Conclusions

These findings suggest that genetic tumor heterogeneity exists, but evidence for a direct relationship to the drug resistant phenotype was not manifest in DNA mutations. Therefore, these results support a phenotype based readout to predict drug resistance within 10 days as opposed to a genotype signature, and that drug resistance to targeted therapy is heterogeneous across the primary tumor.

Funding

AMOSO, CUASF