Page 17 - Read Online

be successfully employed to select responsive patients epigenetic alterations. These alterations can take place
and predict response and resistance. [34-36] at multiple points during tumor initiation, progression
and treatment, and they can be preexisting mutations,
Patient-derived Xenograft Models More acquired mutations, or changes in downstream genes
Accurately Refl ect Human Cancer and pathways. For example, resistance to EGFR
tyrosine kinase inhibitors can be attributed to multiple
Accumulating evidence has indicated PDX models are [47-49]
superior to traditional cell line xenograft models because mechanisms, such as gatekeeper mutation (T790M),
[50]
they maintain more similarities to the tumors found in c-Met amplifi cation, activation of alternative pathways
[48,51]
[14]
actual patients. For example, a detailed cytogenetic such as insulin-like growth factor receptor and AXL,
[52]
analysis of PDX models revealed strong preservation trans-differentiation to mesenchymal cells or small cell
[53]
of the chromosomal architecture observed in patients. features; and (4) Tumor microenvironment. Emerging
[23]
Furthermore, other studies have shown strong fi delity data has indicated tumor microenvironment as a key
[54]
in histology, [37,38] transcriptome, polymorphism mediator of drug resistance. For example, several
[39]
[40]
and copy number variations. In some cases, certain potential mechanisms of resistance to anti-angiogenic
[41]
oncogenic gene amplifi cation can be found in cell drugs are microenvironment-derived, including up
[55,56]
lines at levels that are several-multitude higher than regulation of alternative pro-angiogenic signals,
recruitment of bone marrow progenitors, and increased
[57]
in patient rumors, a cell culture-derived artifact that pericyte coverage. Another example can be found in
[58]
may lead to over-predict drug response in the clinic pancreatic ductal adenocarcinoma, in which gemcitabine
(unpublished data). On the other hand, emerging data resistance has been attributed to ineffi cient drug delivery
started to show that PDX models may be more accurately due to poorly perfused tumors. [59]
refl ect clinical response when treated with therapeutic
agents at clinically relevant doses (CRDs). [21] There are obvious advantages of using PDX models to
study drug resistance mechanism and to characterize
Modeling Drug Resistance therapeutic agents for effi cacy. As discussed earlier,
Despite the continuously growing arsenal of new and PDX models are heterogeneous in nature, and more
[60]
improved anti-cancer drugs, for most cancer patients closely refl ective of tumors in actual patients, and
with advanced diseases, treatment failure remains a more appropriate system for understanding acquired
an inevitable outcome. To a given treatment, only a and de novo drug resistance through enrichment of
fraction of the patients would respond the regimen preexisting changes in subsets of cells. [61,62] A large
favorably (responders), which stresses the importance collection of PDX models can best represent a broad
of selecting patients with the appropriate molecular and patient population with various preexisting mutations
pathological characteristics for maximal therapeutic and susceptibility to generate additional mutations, which
benefi t. On the other hand, even when a particular cannot be achieved by other models including cell line
treatment is initially effi cacious in selected patients, xenografts. In addition, PDX models contain TICs/CSCs,
drug resistance will develop overtime. Therefore, drug and proper tumor stroma (albeit controversial) that can
resistance is a fundamental cause of therapeutic failure potentially contribute to resistance as well. Furthermore,
in cancer therapy. Numerous studies have attempted it has become possible to establish PDX models with
to unravel the mechanisms of drug resistance to tumors that had already been treated and later became
traditional chemotherapeutic agents and to recently refractory. This is an important point because in clinic,
developed targeted, small molecule and antibody based most patients entering clinical trials have been treated
drugs. Briefl y, the mechanisms of resistance can be with standard of cares previously and have relapsed with
roughly mapped to four categories: (1) Multi-drug refractory disease. Compared to cell line xenografts, PDX
resistance (MDR). MDR is caused by expression and/or models should better recapitulate patients with refractory
[63]
induction of effl ux proteins, which are members of the and metastatic cancer.
ABC transporter superfamily involved in the transport A number of studies have taken the advantages of PDX
of both hydrophobic and hydrophilic compounds. This models to study drug resistance. Krumbach et al.
[42]
[60]
mechanism is relatively more common for cytotoxic investigated response to cetuximab in 79 PDX models
drugs and payload of antibody-drug conjugates than generated from colon, gastric, head and neck, lung
[42]
targeted agents; (2) Tumor initiating cells/cancer stem and mammary cancer. After an in-depth analysis of
cells (TICs/CSCs). As discussed earlier, these cells different molecular characteristics of the tumors, they
have the capability of self-renewal and differentiation, identifi ed c-MET activation as a key mechanism for
remain relatively quiescent, and can tolerate higher drug resistance, especially in NSCLC adenocarcinomas.
level of DNA damaging agents and oxidative stress. In another study: using PDX models of NSCLC, Dong
These characteristics are important for TICs to survive et al. identifi ed foci of resistance cells after cisplatin
[64]
chemotherapy and radiation and ignite tumor re-growth treatment as a single agent or in combination with
when the condition permits; [43-46] (3) Tumor genetic and vinorelbine, docetaxel, or gemcitabine. The authors

10 Journal of Cancer Metastasis and Treatment ¦ Volume 1 ¦ Issue 1 ¦ April 15, 2015 ¦