Modeling of Personalized Treatments in Colon Cancer Based on Preclinical Genomic and Drug Sensitivity Data

Abstract

The current standard therapies for advanced, recurrent or metastatic colon cancer are the 5-fluorouracil and oxaliplatin or irinotecan schedules (FOxFI) +/− targeted drugs cetuximab or bevacizumab. Treatment with the FOxFI cytotoxic chemotherapy regimens causes significant toxicity and might induce secondary cancers. The overall low efficacy of the targeted drugs seen in colon cancer patients still is hindering the substitution of the chemotherapy. The ONCOTRACK project developed a strategy to identify predictive biomarkers based on a systems biology approach, using omics technologies to identify signatures for personalized treatment based on single drug response data. Here, we describe a follow-up project focusing on target-specific drug combinations. Background for this experimental preclinical study was that, by analyzing the tumor growth inhibition in the PDX models by cetuximab treatment, a broad heterogenic response from complete regression to tumor growth stimulation was observed. To provide confirmation of the hypothesis that drug combinations blocking alternatively activated oncogenic pathways may improve therapy outcomes, 25 models out of the well-characterized ONCOTRACK PDX panel were subjected to treatment with a drug combination scheme using four approved, targeted cancer drugs.

A Molecularly Characterized Preclinical Platform of Subcutaneous Renal Cell Carcinoma (RCC) Patient-Derived Xenograft Models to Evaluate Novel Treatment Strategies

“…The ratio of mean TV of treated animals to mean TV of control animals (T/C mean) is illustrated for the four drugs in twenty-two PDX models of the RCC cohort as a waterfall plot for each experiment at the final study day ( Figure 4A ). As the use of clinical RECIST criteria for solid tumors is not optimal for the characterization of preclinical tumor response in mice, we implemented our previously published adapted response criteria, delineating strong (T/C <10%), moderate (T/C <25%), and minor response (T/C <50%) from resistance (T/C >50%) ( 44 ). For better visualization of responding PDX models, we shifted the baseline to 25%.…”

An Editorial View on the Special Issue “Colorectal Cancers: From Present Problems to Future Solutions”

“…Therefore, two articles in this special issue introduce novel, and timely, means of modelling colorectal cancer . A highly interesting article by Jens Hoffmann and his group [ 12 ] introduces powerful patient-derived xenograft (PDX) models that are excellently suited to model personalized treatment in CRC, within defined molecular human patient subgroups, the genomic/molecular characteristics of the tumors being analyzed by systems biology approaches. The article demonstrates how biomarkers, or biomarker panels, can be developed for single drugs, or drug combinations, within these CRC PDX models, or how, for example, alternative therapeutic strategies could be suggested in the individual case, depending on individual oncogenic pathway analysis.…”