Introduction
Circulating tumor cells (CTCs) are cancer cells originating from either a primary or metastatic tumor and circulating freely in the peripheral blood [
1]. It has been proposed that the spread of a primary tumor through the bloodstream as CTCs is a critical step in tumor metastasis [
2,
3]. In breast cancer (BC), CTCs can be detected in patients at early stages or late stages of disease with overt metastases [
4-
6]. Many studies have shown that the detection of CTCs may help to predict the outcome in patients with different types of cancers. In particular, the enumeration of CTCs before starting systemic treatment is associated with clinical outcome in both metastatic and non-metastatic BC patients [
4,
6,
7]. Furthermore, CTC count evaluated at different time points during systemic treatment is a reliable surrogate marker of treatment response [
8-
12]. Preliminary studies have suggested that selecting therapies based on molecular characteristics of CTCs may improve treatment outcomes in patients [
13-
15]. Because CTCs are found in circulation as a collectable fraction that is representative of the tumor, they may provide an ideal model to study the biology of the tumor at various intervals before and during treatment [
16].
Interestingly, the presence of CTCs has been found to correlate with the presence of disseminated tumor cells in the bone marrow (BM-DTCs) in BC patients [
17,
18]. Similar to CTCs, BM-DTCs play a crucial role in the metastatic cascade as the earliest detectable form of micrometastatic disease and potential precursors of overt metastases [
19]. Notably, several studies have shown that persistence of BM-DTCs after therapy predicts a higher risk of relapse in BC patients [
20,
21]. Therefore, BM-DTCs represent an additional tool for studying the metastatic process in its initial stage.
Despite the evidence to support the roles of CTCs and BM-DTCs in studying tumor biology and predicting treatment response, routine clinical and preclinical use of these cells is challenging because of multiple factors. First, CTCs are present in small numbers in only 10% to 50% of BC patients [
22-
24]. Therefore, they cannot be isolated in sufficient numbers from a small volume of blood from most patients. Similarly, a longitudinal study of BM-DTCs is impractical because of limited access to bone marrow aspirates and the small number of DTCs that can be enriched from aspirates of standard volume. In addition, commonly used methods to detect CTCs use antibodies against epithelial cell markers and exclude identification of tumor cells with mesenchymal properties. This is especially problematic, as the cells that have undergone epithelial-to-mesenchymal transition (EMT) may play an essential role in the metastatic process [
25].
To address these major challenges, we aimed to determine whether BC patient-derived xenograft (PDX)-bearing mice could provide a constant and renewable source of CTCs and BM-DTCs as a unique system to study the molecular changes responsible for tumor progression and metastases. Here, we report the detection of human CTCs and BM-DTCs in various BC PDX mice models [
26]. To identify the PDX lines with high numbers of CTCs and BM-DTCs, we screened a total of 18 lines representing different molecular subclasses of BC. Further, we evaluated the association of CTC detection with the presence of BM-DTCs and with the lung metastatic potential of these PDX lines. Finally, we determined the predictive value of a genetic profile computed from the primary tumors of various PDX lines that was associated with the presence of CTC clusters and lung metastatic potential.
Discussion
Multiple lines of evidence suggest that CTCs and BM-DTCs can be used to study the metastatic process and be evaluated in “real-time” to monitor the molecular changes in progressing tumors. However, their use has been largely limited because of challenges in their isolation as well as the very low yield of cells detected in human subjects, especially in the early stages of disease. As an alternative strategy, we established the conditions for the detection of human CTCs and BM-DTCs in unique BC PDX models in this study. As previously shown, these preclinical models accurately resemble their parental human tumors in their molecular features and biologic behavior [
26]. Important in this study, we have found that BC PDX models can provide a continuous and renewable source of human CTCs and BM-DTCs. In support of our findings in BC, a recently published study showed that CTCs isolated from pancreatic adenocarcinoma PDX-bearing mice also represent a reliable tool to predict and monitor treatment response [
29].
The rate of CTC and DTC detection (83% and 62.5%, respectively) in our studies with PDX models is higher than that reported in the literature for non-metastatic breast cancer patients. We attribute this to the differences in the CTC/DTC detection methods as well as evaluation of larger blood and bone marrow volume relative to the body size of PDX mice than that of patients, as elaborated here. First, most of the commercial techniques to detect CTCs have relied on the presence of a limited number of epithelial markers (that is, epithelial cell adhesion molecule (EpCAM) and/or “epithelial” cytokeratins 8, 18, 19) [
30]. This approach likely omits CTCs with a predominant mesenchymal phenotype and a lack of the epithelial markers [
25,
31-
33]. In our approach for CTC/DTC detection, we used a quantitative immunohistochemistry assay to examine “human” tumor cells for the expression of multiple cytokeratin subtypes. Indeed, the pan-cytokeratin antibodies (AE1/AE3) we used bind to multiple cytokeratins present on both human epithelial and mesenchymal cells [
34,
35]. Second, the higher CTC/DTC detection rate may also result from accessibility to large amounts of peripheral blood and bone marrow, relative to the small size of the mouse body. This is in contrast with only small blood and bone marrow volume used to assess CTC and DTCs in patients. These factors, as well as the immunodeficiency status of the PDX models, may contribute to the higher CTC/DTC detection rates we find in PDX models. In future, it will be of interest to compare CTCs/DTCs from patients side-by-side with those derived from matched PDXs as well as to compare characteristics of CTCs/DTCs within various immunodeficiency models and versus those with reconstituted immune components [
36].
The observation of clusters of CTCs and BM-DTCs in the BC PDX-bearing mice is of great interest and further justifies the use of our PDX lines as clinically representative models to study these cells. Several studies have identified multicellular CTC clusters in BC [
37,
38] and other types of cancer patients [
39-
43]. In a recent study, CTC clusters were found in 26% of patients with small cell lung cancer, and their presence was an independent prognostic factor [
39]. Moreover, CTC clusters isolated from BC patients had high expression of mesenchymal markers and relatively low expression of epithelial markers, suggesting a potential link between the generation of CTC clusters and the EMT process [
33]. However, the role of CTC clusters in cancer metastatic dissemination remains unclear. In our study, we found a significant association between the presence of CTC clusters and lung metastatic potential. Only one other study, to our knowledge, has shown a similar association in patients with clear cell renal cell carcinoma [
41]. The infrequent finding of clusters of CTCs and BM-DTCs in other studies may be related to the isolation and detection techniques. It is possible that our method of processing the blood and bone marrow fractions may facilitate the detection of clusters.
Of note, we found variability in the detection of CTCs and BM-DTCs in different mice within the same PDX line. The same variability was also present in the previously reported LM detection among these lines [
26]. This variability may be attributed to the intratumoral heterogeneity that is commonly seen in patients or some host-specific factors that may influence tumor initiation and progression. However, this observation emphasizes that the future studies to understand the influence of CTCs in distant metastases should include the analysis at a mouse level as well as an overall analysis for the PDX line. In addition, our finding of a significant correlation between the presence of CTCs and BM-DTCs within the same mice is consistent with what is observed in early BC patients [
17]. All the BC PDX lines that had CTCs also had BM-DTCs and/or LM. This high concordance rate suggests that these CTCs and BM-DTCs are early indicators of metastatic potential and as such are important for molecular characterization and tumor biology studies. Because BM-DTCs were not evaluated in sufficient numbers of mice for most lines, correlation analyses as well as gene-expression analyses in the PDX lines were restricted to only the CTCs and LM data in this study. Future studies to characterize molecular profiles of CTCs, BM-DTCs, and lung lesions may uncover important biomarkers and treatment targets to prevent metastases.
The four-gene profile we obtained by overlapping the two genetic signatures of primary PDX tumors associated with CTC clusters and with LM was associated with a significant reduction in distant metastases-free survival in early BC patients. However, the observed association was weak because of limitations such as sample size and confounding variables, such as different subtypes and various treatments. Future studies using additional PDX lines are also necessary to validate this gene profile. In support of the derived gene profile, however, other reports have independently identified some of these genes to be associated with LM in BC. For example,
HLA-DP1A, which encodes a transmembrane protein involved in the antigen-presentation process, was downregulated in both the CTC clusters and LM signatures. Interestingly,
HLA-DP1A was one of the genes downregulated in the LM signatures generated in two independent studies in BC [
44]. The other downregulated gene in our four-gene set was
GJA1, which encodes a major protein in gap junctions.
GJA1, also known as connexin
43 (Cx43), has been shown to suppress mammary tumor metastasis to the lung in a mouse model [
45]. Specifically, missense mutation (G60S) in this gene led to production of an altered Cx43 protein that acts in a dominant-negative fashion to disrupt gap-junction assembly and function. This mutation was associated with higher rates of LM in ErbB2-overexpressing mice. Of the two upregulated genes in our four-gene profile, XIST is a noncoding RNA gene on X chromosome and plays a major role in X inactivation. Although overexpression of XIST has been seen in BRCA-1-associated BCs, which typically metastasizes to lungs [
46,
47], a direct link between XIST and LM in BC has not been established. The other upregulated gene,
PEG3, encodes a C2H2 type zinc finger protein implicated in regulation of body temperature, feeding behavior, and obesity [
48], as well as growth, apoptosis, and maternal nurturing behavior [
49]. The role of PEG3 in BC and LM is not clear and warrants further investigation.
Acknowledgements
This study was supported in part by National Institutes of Health grants U54CA149196 (to R.S. and M.T.); P01 SPORE (to R.S. and R.M.J.); faculty start-up funds (to M.T.); Cancer Prevention and Research Institute of Texas (CPRIT) program RP101499, Baylor College of Medicine Comprehensive Cancer Training Program (to M.G.); Susan G. Komen grant KG120001 (to M.L.); Dan L. Duncan Cancer Center Grant P30CA125123 from the National Cancer Institute, Cytometry and Cell Sorting Core at Baylor College of Medicine with funding from the NIH (P30 AI036211, P30 CA125123, and S10 RR024574), research grants from the Breast Cancer Research Foundation; the Entertainment Industry Foundation/Lee Jeans, and SU2C Breast Cancer Program.
We also acknowledge the following Baylor College of Medicine shared resources: Biostatistics & Informatics; Cytometry and Cell Sorting; and Human Tissue Acquisition and Pathology.
Competing interests
Mike T. Lewis is a member of StemMed Holdings LLC, and a limited partner in StemMed Ltd. The following authors declare no competing interests: Mario Giuliano, Sabrina Herrera, Pavel Christiny, Chad Shaw, Chad J. Creighton, Tamika Mitchell, Raksha Bhat, Xiaomei Zhang, Sufeng Mao, Lacey E. Dobrolecki, Ahmed Al-rawi, Fengju Chen, Bianca M. Veneziani, Xiang H. Zhang, Rinath M. Jeselsohn, Susan G. Hilsenbeck, Alejandro Contreras, Carolina Gutierrez, Mothaffar F. Rimawi, C. Kent Osborne, Rachel Schiff, and Meghana V. Trivedi.
Authors’ contributions
MG collected and processed blood and bone marrow samples, conceived of and designed the study, and drafted the manuscript. SH performed pathology analysis. PC participated in collection and processing of blood and bone marrow samples. CS performed bioinformatics analysis, participated in data interpretation, and helped to draft the manuscript. CJC participated in bioinformatics analysis and data interpretation, and helped to draft the manuscript. TM participated in accomplishment of in vivo studies and collection of blood and bone marrow samples. RB participated in processing of blood and bone marrow samples. XZ coordinated the in vivo studies. SM processed blood and bone marrow samples. LED participated in accomplishment of in vivo studies. AA participated in pathology data analysis. FC participated in bioinformatics analysis and data interpretation. BMV participated in interpretation of data and study design. XHZ participated in interpretation of data and helped to draft the manuscript. RMJ participated in study design and interpretation of data. SGH performed statistical analysis and participated in interpretation of data. AC participated in pathology analysis and data interpretation. CG coordinated pathology analyses. M.F.R. participated in study design and data interpretation. CKO participated in study coordination and helped to draft the manuscript. MTL participated in study design, provided in vivo PDX mouse models, and helped to draft the manuscript. RS conceived the study, participated in its design and coordination, and helped to draft the manuscript. MVT conceived of, designed, and coordinated the study, performed data analysis and interpretation, and drafted the manuscript. All authors read and approved the final manuscript.