This article is mentioned in:

Introduction

Breast cancer is the most commonly diagnosed malignancy among women, with 2,088,849 new cases of invasive breast cancer and 626,679 deaths reported in 2018(1). Metastatic disease is typically an incurable condition associated with significant morbidity and mortality in breast cancer patients (2). The metastatic cascade is characterized as a multi-step process that includes escape of cancer cells from the primary tumor to a distant location, where they can potentially form new tumor colonies (3). To successfully complete the metastatic cascade, epithelial tumor cells detached from the primary tumor must penetrate into the peripheral circulation and ultimately undergo extravasation at the distant site and establish a new tumor focus. Previous findings have demonstrated that cancer cells undergoing epithelial-to-mesenchymal transition (EMT) acquire the characteristics necessary to go through the multiple steps of metastasis (4).

Circulating tumor cells (CTCs) essentially contribute to tumor dissemination and progression and are a crucial factor in the metastatic cascade (3,5). The prognostic and predictive value of CTCs was consistently established by numerous trials, not only in metastatic, but also in primary breast cancer (PBC) (6-11). CTCs are closely associated with several biological processes, involving EMT, ‘self-seeding’, which is characterized as re-infiltration of the primary tumor by CTCs, as well as establishment of metastasis by more aggressive CTCs. Therefore, CTCs have been suggested to represent an indicator of treatment failure and disease progression (12,13). Available experimental and translational research data indicate that the generation of CTCs is a continuous process spanning from one end of the spectrum (epithelial phenotype) to the other end (mesenchymal phenotype), and involves those with a partial EMT phenotype (14,15). Accordingly, CTCs may be considered a heterogeneous population of cells, including CTCs with partial or complete EMT phenotype, and these subpopulations have different clinical and biological properties (5). In addition, CTCs display dynamic changes in epithelial and mesenchymal composition (15).

The most common somatic mutations in breast cancer are TP53, PIK3CA and GATA3, which are present in >10% of all breast cancers (16). In a study including 216 patients, 12 genes (TP53, PIK3CA, GATA3, ESR1, MAP3K1, CDH1, AKT1, MAP2K4, RB1, PTEN, CBFB and CDKN2A) were established as significantly mutated in metastatic breast cancer (mBC), while 8 genes (ESR1, FSIP2, FRAS1, OSBPL3, EDC4, PALB2, IGFN1 and AGRN) were more frequently mutated in mBC as compared to early breast cancer (17). Several published studies have investigated the mutational status of CTCs (18-21); however, data on the association between gene mutations in primary tumor tissue and the presence of CTCs in the peripheral blood are lacking.

The aim of the present study was to identify the breast cancer-specific mutation status in PBC in association with the presence of various subpopulations of CTCs using targeted resequencing.

Patients and methods

Study patients

This translational study (Protocol TRU-SK 002; Chair: M. Mego) included 78 PBC patients with stages I-III after definitive surgery. Patients enrolled in this study were selected from a cohort of 427 PBC patients analyzed in a previous study detecting CTCs with EMT phenotype in 77 (18.0%) patients with early breast cancer (22). Fresh frozen tumor tissue and status of CTCs in peripheral blood were available for all enrolled participants. To exclude the presence of distant metastasis, each patient underwent a full diagnostic evaluation. Patients with concurrent malignancies other than non-melanoma skin cancer in the previous 5 years were excluded. Patients' data regarding age, tumor stage, histological type, regional lymph node involvement, hormone receptor status, and human epidermal growth factor receptor 2 (HER2) status were also tabulated and statistically analyzed.

The study was performed between March 2012 and February 2015. Healthy donors (n=60) were age-matched women without breast cancer who were recruited according to the Institutional Review Board (IRB)-approved protocol. The present study was approved by the Institutional Review Board (IRB) of the National Cancer Institute of Slovakia (Bratislava, Slovak Republic). Written informed consent was obtained from each participant prior to study enrollment.

Detection of CTCs in the peripheral blood

The peripheral blood samples used for CTC detection were collected into Vacutainer® Blood Collection Tubes EDTA (BD Biosciences) in the morning on the day of surgery. Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) assay using CD45+ cell depletion for CTC enrichment was employed for CTC identification in the peripheral blood, as described previously (23,24).

RNA extraction and cell lines

A RossetteSep™ kit (Stemcell Technologies, Inc.) was used for CD45 depletion in peripheral blood samples, according to the manufacturer's instructions. CD45-depleted cells were mixed with 500 µl TRIzol LS reagent (Invitrogen; Thermo Fisher Scientific, Inc.) and stored at -80˚C until RNA was extracted according to the manufacturer's instructions. The precipitated pellet containing RNA was dissolved in 50 µl nuclease-free water. Absorbance readings at 260 nm (median, 5.95 ng/µl; range, 1.7-38.3 ng/µl) were used to determine RNA concentration. RNA extracted from HeLa, HCT116, MCF-7 and MDA-MB-231 cells served as the positive control.

Identification of gene transcripts in CD45-depleted subsets

EMT-inducing TF gene transcripts (TWIST, SNAIL1, SLUG and ZEB1) and epithelial antigen (CK19) were identified by RT-qPCR. In brief, the reaction was carried out in 25 µl reaction volume containing 12.5 µl QuantiFast Probe RT-PCR kit Master Mix, 0.25 µl QuantiFast RT mix, 8.5 µl water and 1.25 µl primers (Qiagen GmbH). Isolated RNA was subsequently added (in a volume of 2.5 µl) to the reaction mix. The following TaqMan assays were obtained from Thermo Fisher Scientific, Inc.: TWIST1: Hs00361186_m1; SNAIL1: Hs00195591_m1; SLUG: Hs00161904_m1; ZEB1: Hs01566408_m1; GAPDH Hs99999905_m1; and CK19 Hs00761767_s1. Amplicons or probes spanned intron-exon boundaries, with the exception of CK19. Amplification was performed on a Roche LightCycler 480 II Real-Time PCR system (Roche Diagnostics GmbH) using the following cycling program: 95˚C for 10 min, followed by 40 cycles at 95˚C for 15 sec and at 60˚C for 60 sec. All the samples were analyzed in triplicate. Calibrator samples were run with every plate to ensure consistency of the PCR. For all fluorescence-based RT-PCR, fluorescence was detected between 0 and 40 cycles for the control and marker genes in single-plex reactions, which allowed for the deduction of the cycle quantification (Cq) value for each product. Establishment of GAPDH housekeeping gene expression was used as the expression control in the identification of the genes of interest. Target cDNA was measured by the ΔCq method with the formula: 1=2 Cq (target-GAPDH) (25).

CTC definition

Patient samples with elevated CK19 gene transcripts compared with healthy donors were categorized as epithelial CTC-positive (CTC_EP), while patient samples with increased EMT-TF (TWIST1, SNAIL1, SLUG and ZEB1) gene transcripts compared with those of healthy donors were categorized as CTC_EMT-positive. To categorize a patient sample as CTC-positive, expression of at least one of the markers (either epithelial or mesenchymal) at levels above the defined cut-off was required. The highest expression levels of the CK19 and EMT-inducing TF gene transcripts relative to that of GAPDH were 3.4x10-3 (median, 2.8x10-6; range, 0-3.4x10-3) for CK19, 7.5x10-4 (median, 0; range, 0-7.5x10-4) for TWIST1, 3.8x10-2 (median, 0.003135; range, 5.0x10-4-3.8x10-2) for SNAIL1 and 1.7x10-1 (median, 1.4x10-2; range, 2.2x10-3-1.7x10-1) for ZEB1, while SLUG transcripts were not detected in any of the samples from healthy donors. The cut-off value determining CTC positivity was established according to the highest expression values in healthy donors (14).

DNA isolation from the tumor

Tumor tissues were homogenized in liquid nitrogen and DNA was isolated from disintegrated cells using a DNeasy blood and tissue kit (Qiagen GmbH) according to the original protocol. The most representative central part of the primary tumor was used for extraction of tumor DNAs.

Targeted resequencing Panel of analyzed genes

The gene panel contained well-known genes associated with clinically relevant characteristics in breast cancer patients: ACVR1B, AKT1, ARID1A, ATM, BARD1, BRCA1, BRCA2, BRIP1, CASP8, CBFB, CDH1, CDKN1B, CHEK2, CTCF, CUL4, EP300, EPCAM, ERBB2, ERBB3, FANCC, FRFR2, FOXA1, GATA3, GNPTAB, HIST1H3B, KMT2A, KRAS, MAP2K4, MAP3K1, MED23, MLH1, MLL3, MSH2, MSH6, MYB, NBN, NCOR1, NF1, PALB2, PIK3CA, PIK3R1, PMS2, PTEN, RAB40A, RAD51C, RAD51D, RB1, RUNX1, SF3B1, SPEN, STAG2, STK11, TBL1XR1, TBX3, TP53 and XRCC2. The customized SureSelect XT (Agilent Technologies, Inc.) probe-based enrichment strategy was used.

Genomic library preparation

DNA libraries were prepared by SureSelect QXT Target Enrichment Kit (Agilent Technologies, Inc.) according to the protocol for Illumina Multiplex Sequencing. Final DNA libraries were diluted to 13-20 pmol/l pools with 10-11 samples analyzed in parallel per one MiSeq sequencing run using a MiSeq sequencing reagent kit v3 150 cycles (Illumina, Inc.), set to run in the 2x100 mode.

Genomic data analysis

Following standard BaseSpace-based sequencing data analysis, read mapping and variant calling sequencing data were annotated and filtered using Ingenuity Variant Analysis (Qiagen GmbH), which was set up to search for pathogenic variants associated with breast cancer. The identified pathogenic and likely pathogenic variants were verified by visualization in IGV and subsequently confirmed by Sanger sequencing.

Statistical analysis

The patients' characteristics were tabulated and summarized as the median (range) for continuous variables and frequency (percentage) for categorical variables. Categorical data were tested by Fisher's exact test or Chi-squared test. All P-values were two-sided, and P≤0.05 was considered to indicate statistically significant differences. Statistical analyses were performed using NCSS 2007 statistical software (NCSS LLC) (26).

Results

Patient characteristics

The analyzed cohort included 78 patients with PBC. The median age of the enrolled patients was 57 years (range, 33-83 years). The clinicopathological characteristics of the patients are summarized in Table I. The majority of the patients had invasive ductal (89.7%), hormone receptor-positive (84.6%) tumors and T1 stage disease (66.7%). HER2/neu amplification was observed in 20.5% of the tumors.

Table I Patient characteristics.

Table I

Patient characteristics.

Variable	N	%
All	78	100.0
T-stage
1	52	66.7
>1	26	33.3
N-stage
0	48	61.5
>1	30	38.5
Grade
1 and 2	50	64.1
3	28	35.9
TNM classification
Stage I	34	43.6
Stage II	24	30.8
Stage III	14	17.9
Histology
Invasive ductal carcinoma	70	89.7
Other	8	10.3
Hormone receptor status
Negative for both	12	15.4
Positive for either	66	84.6
HER2 status
Positive	16	20.5
Negative	62	79.5
Ki 67 (cut-off 20%)
Low	48	61.5
High	30	38.5
Epithelial CTC
Present	20	25.6
Absent	58	74.4
CTC_ EMT
Present	28	35.9
Absent	50	64.1
Any CTC
Present	48	61.5
Absent	30	38.5

In the present analysis, CTCs in PBC were detected in 44 (61.1%) patient tumor samples, while no CTCs were detected in 28 (38.9%) of the analyzed tumors. In order to improve clarity of published results only CTCs with epithelial or EMT phenotype were taken into consideration. In subgroup of CTCs-positive patients, 38.6% (17/44) samples were characterized as epithelial CTCs (EP_CTC)-positive, while CTCs EMT (CTC_EMT)-positive phenotype was actually identified in 61.4% (27/44) of samples Table II.

Table II Association between mutation status and clinicopathological characteristics.

Table II

Association between mutation status and clinicopathological characteristics.

		Any mutation vs. no mutation		TP53		PIK3CA		BRCA1/2		Other than TP53, PIK3CA, BRCA		Double/triple hit vs. zero/single
Variables	No.	No.	%	No.	%	No.	%	No.	%	No.	%	No.	%
T-stage
1	49	38	77.6	13	26.5	12	24.5	4	8.2	28	57.1	12	24.5
>1	23	13	56.5	8	34.8	4	17.4	2	8.7	10	43.5	6	26.1
P-value			0.10		0.58		0.56		1.00		0.32		1.00
N-stage
0	45	33	73.3	13	28.9	11	24.4	4	8.9	26	57.8	10	22.2
>1	27	18	66.7	8	29.6	5	18.5	2	7.4	12	44.4	8	29.6
P-value			0.60		1.00		0.77		1.00		0.33		0.58
Histology
Invasive ductal cancer	69	47	68.1	20	29.0	15	21.7	6	8.7	36	52.2	15	21.7
Other	5	4	80.0	1	20.0	1	20.0	0	0.0	2	40.0	3	60.0
P-value			1.00		1.00		1.00		1.00		0.66		0.10
Grade
1 and 2	42	29	69.0	9	21.4	13	31.0	3	7.1	23	54.8	9	21.4
3	25	18	72.0	11	44.0	2	8.0	3	12.0	13	52.0	6	24.0
P-value			1.00		0.06		0.04		0.66		1.00		1.00
Hormone receptor status
Negative for both	12	11	91.7	9	75.0	1	8.3	2	16.7	11	91.7	4	33.3
Positive for either	60	40	66.7	12	20.0	15	25.0	4	6.7	27	45.0	14	23.3
P-value			0.16		0.0004		0.28		0.26		0.004		1.00
HER2 status
Positive	16	13	81.3	9	56.3	1	6.3	1	6.3	10	62.5	4	25.0
Negative	56	38	67.9	12	21.4	15	26.8	5	8.9	28	50.0	14	25.0
P-value			0.37		0.01		0.10		1.00		0.41		1.00
Ki 67 (cut-off 20%)
Low	42	29	69.0	5	11.9	13	31.0	2	4.8	18	42.9	10	23.8
High	30	22	73.3	16	53.3	3	10.0	4	13.3	20	66.7	8	26.7
P-value			0.80		0.0002		0.05		0.23		0.06		0.79
CTC epithelial
Negative	28	19	67.9	8	28.6	4	14.3	0	0.0	11	39.3	7	25.0
Positive	17	13	76.5	3	17.6	4	23.5	4	23.5	11	64.7	3	17.6
P-value			0.74		0.49		0.45		0.02		0.13		0.72
CTC EMT
Negative	28	19	67.9	8	28.6	4	14.3	0	0.0	11	39.3	7	25.0
Positive	27	19	70.4	10	37.0	8	29.6	2	7.4	16	59.3	8	29.6
P-value			1.00		0.57		0.20		0.24		0.18		0.77
CTC Any	N
Negative	28	19	67.9	8	28.6	4	14.3	0	0.0	11	39.3	7	25.0
Positive	44	32	72.7	13	29.5	12	27.3	6	13.6	27	61.4	11	25.0
P-value			0.79		1.00		0.25		0.08		0.09		1.00

[i] Bold, statistically significant.

Identification of mutations in breast cancer-related genes in the primary tumor and its association with the presence of CTCs in the peripheral blood

Six of the 78 samples (7.7%) were excluded due to poor quality of DNA for analysis. Mutations were detected in 68.1% (49/72) of tumor samples, 36.1% (26/72) were classified as likely pathogenic and 61.1% (44/72) as pathogenic while in 27.8% (20/72) of tumors no mutation was detected. In 43.1% (31/72) of tumor samples the patient's single pathogenic or likely pathogenic mutation was detected, in 18.1% (13/72) patient mutations in two genes were found, while in 4.2% (3/72) of tumor samples patient mutations in 3 genes were detected.

The most commonly affected genes were TP53, mutated in 25.0% (18/72) tumors, followed by PIK3CA mutated in 22.2% (16/72) tumors, BRCA1/2 in 9.7% (7/72) tumors (2 for BRCA1 and 5 for BRCA2), CDH1 and GATA3 in 6.9% (5/72) tumors. RUNX1 and PTEN were mutated in 4.2% (3/72) tumors, NF1, BRIP1 and ATM in 2.8% (2/72) tumors while BARD1, CDKN1B, GNPTAB, KRAS, PIK3R1 and PMS2 were mutated in 1.4% (1/72) tumor (Table III).

Table III Panel of specific mutations identified in PBC patients (n=79).

Table III

Panel of specific mutations identified in PBC patients (n=79).

Patient no.	Category	Gene	Mutation effect
1	No mutation	No mutation	No mutation
3	Pathogenic	CDH1	Stop gain
4	No mutation	No mutation	No mutation
5	Likely Pathogenic	BARD1	Frameshift
	Likely Pathogenic	TP53	Frameshift
6	NA	NA	NA
7	Likely Pathogenic	GATA3	Frameshift
8	Likely Pathogenic	RUNX1	Frameshift
9	Likely Pathogenic	PIK3R1	Frameshift
10	Pathogenic	PIK3CA	Missense
	Likely Pathogenic	TP53	Frameshift
11	Likely Pathogenic	TP53	Frameshift
12	Likely Pathogenic	CDH1	Frameshift
	Pathogenic	TP53	Missense
13	Pathogenic	NF1	Stop gain
	Pathogenic	TP53	Missense
14	Pathogenic	TP53	Missense
15	Pathogenic	ATM	Frameshift
16	No mutation	No mutation	No mutation
17	Pathogenic	PIK3CA	Missense
18	No mutation	No mutation	No mutation
19	Likely Pathogenic	BRIP1	Missense
	Likely Pathogenic	TP53	Frameshift
20	Pathogenic Likely	PIK3CA	Missense
	Pathogenic	GATA3	Frameshift
	Pathogenic	ATM	Stop gain
21	Pathogenic	TP53	Missense
22	No mutation	No mutation	No mutation
23	No mutation	No mutation	No mutation
24	No mutation	No mutation	No mutation
25	NA	NA	NA
26	Pathogenic	GATA3	Frameshift
27	Pathogenic	PIK3CA	Missense
28	Pathogenic	PIK3CA	Missense
29	No mutation	No mutation	No mutation
30	No mutation	No mutation	No mutation
31	Pathogenic	PIK3CA	Missense
33	Pathogenic	PIK3CA	Missense
	Pathogenic	TP53	Missense
35	Pathogenic	CDH1	Stop gain
36	Likely Pathogenic	GATA3	Frameshift
37	Pathogenic	TP53	Missense
38	Pathogenic	TP53	Missense
40	Pathogenic	BRCA1	Frameshift
	Pathogenic	TP53	Missense
	Pathogenic	RUNX1	Frameshift
43	Pathogenic	PIK3CA	Missense
44	No mutation	No mutation	No mutation
45	Pathogenic	PIK3CA	Missense
46	Likely Pathogenic	TP53	Frameshift
47	Pathogenic	BRCA2	Frameshift
	Pathogenic	BRCA2	Stop gain
48	Pathogenic	TP53	Missense
49	No mutation	No mutation	No mutation
50	No mutation	No mutation	No mutation
51	Pathogenic	TP53	Missense
52	No mutation	No mutation	No mutation
53	Likely Pathogenic	GATA3	Frameshift
54	Likely Pathogenic	BRCA2	Missense
55	Likely Pathogenic	NF1	Frameshift
	Likely Pathogenic	TP53	Frameshift
56	Likely Pathogenic	TP53	Frameshift
57	Pathogenic	GNPTAB	Frameshift
	Pathogenic	BRCA2	Frameshift
58	NA	NA	NA
59	No mutation	No mutation	No mutation
60	Pathogenic	PIK3CA	Missense
61	Pathogenic	PIK3CA	Missense
	Pathogenic	BRIP1	Frameshift
62	No mutation	No mutation	No mutation
63	Pathogenic	PIK3CA	Missense
	Pathogenic	TP53	Stop gain
	Likely Pathogenic	PMS2	Frameshift
64	No mutation	No mutation	No mutation
65	No mutation	No mutation	No mutation
66	Pathogenic	PIK3CA	Missense
	Likely Pathogenic	BRCA2	Missense
67	Likely Pathogenic	TP53	Missense
68	Likely Pathogenic	CDH1	Frameshift
	Likely Pathogenic	RUNX1	Frameshift
69	Pathogenic	KRAS	Missense
70	No mutation	No mutation	No mutation
71	Pathogenic	PIK3CA	Missense
72	No mutation	No mutation	No mutation
73	No mutation	No mutation	No mutation
74	Likely Pathogenic	PTEN	Frameshift
	Likely Pathogenic	CDKN1B	Frameshift
75	Pathogenic	PTEN	Stop gain
	Pathogenic	PTEN	Frameshift
76	Pathogenic	BRCA1	Frameshift
77	Pathogenic	PIK3CA	Missense
	Likely Pathogenic	CDH1	Frameshift
78	Pathogenic	PIK3CA	Missense

[i] PBC, primary breast cancer.

There were no differences in the number of tumors with pathogenic or likely pathogenic mutations between participants with detectable CTCs in the peripheral blood compared with patients with non-detectable CTCs (67.9 vs. 72.7%, respectively; P=0.79) (Table II). This difference remained unchanged when likely pathogenic mutations were excluded and no mutations vs. pathogenic mutations were compared (46.4 vs. 52.3%, respectively; P=0.78) (data not shown). Similarly, there were no differences in CTC status regarding mutations of TP53 and PIK3CA and/or between tumors with single mutations vs. those with double/triple mutations. However, no BRCA1/2 mutations were detected in CTC-negative tumors compared with 9.7% of BRCA1/2 mutations (P=0.08) in CTCs-positive tumors. Moreover, 4 (23.5%) patients with epithelial CTCs in peripheral blood had BRCA1/2 mutations compared to 0 (0%) patients without BRCA1/2 mutations (P=0.02), while there was no significant correlation between mutation in the specific gene and presence of CTC_EMT-positive cells. Similarly, there was a trend for an increased mutation rate of genes other than TP53, PIK3CA and BRCA1/2 in CTC-positive compared with CTC-negative patients (61.4 vs. 39.3%, respectively; P=0.09) (Table II).

Examination of specific mutations revealed a significant association between TP53 mutation and HER2-positive status, hormone receptor negativity, high grade and increased tumor cell proliferation, as determined by the expression of Ki67. PIK3CA mutations were associated with lower grade and low proliferation rate, as determined by Ki67 (Table II).

Discussion

To the best of our knowledge, the present study was the first to reveal the association between the presence of epithelial CTCs in the peripheral blood and mutations of the BRCA1/2 genes in primary tumor tissue. We observed a numerically higher mutation rate in genes other than TP53 and PIK3CA and BRCA1/2 in patients with CTC-positive compared with CTC-negative breast tumors; however, the differences did not reach statistical significance. The most commonly mutated genes in our patient cohort included TP53, PIK3CA, BRCA1/2, CDH1 and GATA3, corresponding to the observed incidence in published datasets (16,17). Similarly, Stephens et al detected somatic driver substitutions and small insertions/deletions (indels) in cancer genes previously implicated in breast cancer development, including AKT1, BRCA1, CDH1, GATA3, PIK3CA, PTEN, RB1 and TP53 (27). Cancer genes TP53, PIK3CA, ERBB2, MYC, FGFR1/ZNF703, GATA3 and CCND1 were mutated in more than 10% of the analyzed cohort (27). CDKN1B (also known as p27 or KIP1) normally inhibits the activation of cyclin E/CDK2 and cyclin D/CDK4 complexes, thus preventing cell cycle progression at phase G1(28). Our analysis revealed one frameshift mutation in CDKN1B. These data confirm previously published results reporting CDKN1B gene as a cancer gene (27). Stephens et al also identified 73 different combinations of mutated cancer genes. In our analyzed group we found 18 different combinations of mutated cancer genes. Based on these results, we determined that most breast cancers differed from all others (27). Significant correlations between TP53 mutational status and certain clinicopathological characteristics of the primary tumor were in concordance with literature data (29). BRCA1/2 are tumor suppressor genes that are involved in DNA repair pathways associated with hereditary breast and ovarian cancer syndrome (30). A study by Erturk et al demonstrated that in BRCA-mutated triple-negative breast cancer, increased invasion and metastasis of cancer cells was mediated through the BRCA mutation-associated decreased expression of miR-200c (31). Therefore, our results may be partially explained by BRCA1/2 mutation-mediated increased invasion of breast cancer cells. Germline mutations of BRCA are not associated with a higher rate of lymphovascular invasion (LVI) compared with non-mutated tumors (32); however, previous studies showed a correlation between LVI and the presence of CTCs (33,34).

Data regarding the association between mutations in the primary tumor and the presence of CTCs in the peripheral blood are limited. While there are some studies that investigate the correlations between mutations in specific genes in CTCs and corresponding tumor tissue, trials addressing the association between mutational status of the primary tumor and the presence of CTCs are currently lacking. For example, in a study by Bredemeier et al, no correlation was observed between PI3KCA mutations in cancerous tissue and the presence of CTCs in peripheral blood (19). In another study, there was a high match rate between specific mutations in CTCs and corresponding tumor tissue, but no data regarding the association between a specific mutation and the number of CTCs were reported (18). The cell-cell adhesion molecule CDH1 (E-cadherin) as well as CTNNB1 (beta-catenin), are involved in the process of cancer cell detachment from the primary tumor. Their association with EMT was revealed through several studies (35-37). The mechanism of EMT is predominantly activated by epigenetic events; however, data explaining the role of genetic are still limited. The study of Busch et al suggested that somatic mutations in CDH1 and CTNNB1 genes do not represent a major contributor to cancer cell detachment, and therefore presumably play a limited role in the etiology of tumor metastasis as well as in EMT (38). The association between CTC_EMT status and RFS (relapse-free survival) was evaluated in another study (22). This study including 427 PBC patients revealed the prognostic value of CTCs with EMT phenotype in all analyzed subgroups of patients. Patients enrolled into the present study were selected from the abovementioned cohort of PBC patients. In addition, CTCs with epithelial phenotype had no prognostic role in the abovementioned cohort (data not shown). It may be suggested that CTCs reflect specific biological characteristics of the tumor, similar to previous studies that showed no association between CTCs and common clinicopathological characteristics (6,9,23). Due to intratumoral heterogeneity, we cannot exclude an association between mutational status in specific parts of the tumor (such as the tumor edge) and/or involvement of genes other than the ones examined in CTC generation.

There were certain limitations to the present study, such as the limited sample size with decreased statistical power of analyses, as well as the fact that we only assessed 56 breast cancer-associated genes. It was not possible to assess the impact of total mutational burden on CTC positivity. In addition, the study population was a homogenous cohort of patients, treatment-naive, without metastatic disease, in order to avoid the effect of the metastatic site heterogeneity factor on analyzed variables. Furthermore, the evaluation of epithelial and mesenchymal phenotype changes via BRCA gene silencing in vitro was not performed in present study.

In conclusion, a correlation between the presence of epithelial CTCs in the peripheral blood and mutations of BRCA1/2 genes in primary tumor tissue was identified, while there was no mutation in specific genes associated with CTC_EMT. The number of mutated breast cancer-associated genes was not associated with the presence of CTCs or the mutation of genes other than BRCA1/2, suggesting that different factors may be involved in the generation and migration of CTCs. These data support the concept that CTCs are of high biological and clinical value in breast cancer.

Acknowledgements

Not applicable.

Funding

The present is the result of the implementation of project no. APVV-16-0010 and APVV-14-0327 funded by the Slovak Research and Development Agency.

Availability of data and materials

All datasets generated and analyzed during the present study are included in this published article.

Authors' contributions

MM, JMar and GM participated in the conception and design of this study. MM performed statistical analysis. GM, TS and KK were involved in CTC detection. GR, LK and GM performed next-generation sequencing. MK, JB and DP were involved in patient accrual and performed breast surgery. GS and JMac performed pathological examination. SJ was involved in patient accrual. MM drafted the article and all authors reviewed it critically for important intellectual content. All the authors participated in the acquisition, analysis and interpretation of data. All the authors have read and approved the final version of the manuscript for publication.

Ethics approval and consent to participate

The present study was approved by the Institutional Review Board (IRB) of the National Cancer Institute of Slovakia (Bratislava, Slovak Republic). Written informed consent was obtained from each participant prior to study enrollment.

Patient consent for publication

Not applicable.

Competing interests

Authors declare they have no competing interests.

References