Purpose: We have reported loss of SMAD4 promotes expression of CCL15 from colorectal cancer to recruit CCR1+ myeloid cells through the CCL15-CCR1 axis, which contributes to invasion and liver metastasis. However, the molecular mechanism of lung metastasis is yet to be elucidated. Our purpose is to determine whether similar mechanism is involved in the lung metastasis of colorectal cancer.

Experimental Design: In a mouse model, we examined whether SMAD4 could affect the metastatic activity of colorectal cancer cells to the lung through the CCL15-CCR1 axis. We immunohistochemically analyzed expression of SMAD4, CCL15, and CCR1 with 107 clinical specimens of colorectal cancer lung metastases. We also characterized the CCR1+ myeloid cells using several cell-type–specific markers.

Results: In a mouse model, CCL15 secreted from SMAD4-deficient colorectal cancer cells recruited CCR1+ cells, promoting their metastatic activities to the lung. Immunohistochemical analysis of lung metastases from colorectal cancer patients revealed that CCL15 expression was significantly correlated with loss of SMAD4, and that CCL15-positive metastases recruited approximately 1.9 times more numbers of CCR1+ cells than CCL15-negative metastases. Importantly, patients with CCL15-positive metastases showed a significantly shorter relapse-free survival (RFS) than those with CCL15-negative metastases, and multivariate analysis indicated that CCL15 expression was an independent predictor of shorter RFS. Immunofluorescent staining showed that most CCR1+ cells around lung metastases were tumor-associated neutrophil, although a minor fraction was granulocytic myeloid-derived suppressor cell.

Conclusions: CCL15-CCR1 axis may be a therapeutic target to prevent colorectal cancer lung metastasis. CCL15 can be a biomarker indicating poor prognosis of colorectal cancer patients with lung metastases. Clin Cancer Res; 23(3); 833–44. ©2016 AACR.

Translational Relevance

Tumor microenvironment contains several types of cells including cancer cells and normal host cells. The use of various mouse models has shown that bone marrow–derived myeloid cells, such as tumor-associated macrophages, myeloid-derived suppressor cells (MDSCs), and tumor-associated neutrophils (TANs), have important roles in tumor progression. However, it remains to be determined whether similar mechanisms are involved in humans. We have recently reported loss of SMAD4 stimulates expression of chemokine CCL15 from colorectal cancer to recruit CCR1+ MDSCs, which promotes tumor invasion of primary colorectal cancer. This is the first clinical study showing that loss of SMAD4 promotes lung metastasis of colorectal cancer by accumulation of CCR1+ TANs through the CCL15-CCR1 axis, and that CCL15 expression of lung metastatic lesions is an independent predictor of worse prognosis in colorectal cancer patients. These results may justify the treatment with CCR1 inhibitors to prevent lung metastasis of colorectal cancer after pulmonary resection of metastatic colorectal cancer.

Colorectal cancer, one of the most common cancers worldwide, frequently metastasizes to the liver and lung. The lung is one of the most frequent sites of metastases from colorectal cancer, with an incidence of approximately 5% to 15% (1, 2, 3). Although the development of chemotherapy for metastatic colorectal cancer has been reported, surgical resection of lung metastases appears the most optimal treatment, if possible (4). It was reported that complete pulmonary resection can improve the 5-year survival rate up to more than 45% (5–9). However, there are few unified criteria for surgical resection for lung metastases.

To achieve comprehensive therapeutic effects, it is imperative to understand the molecular mechanisms behind individual cancers. Studies in various mouse models have shown that bone marrow–derived myeloid cells, such as tumor-associated macrophages (TAM), myeloid-derived suppressor cells (MDSC), and tumor-associated neutrophils (TAN), play important roles in tumor progression (10–14). For example, Lyden and colleagues reported that hematopoietic progenitor cells expressing myeloid cell marker (CD11b) and vascular endothelial growth factor receptor 1 (VEGFR1) accumulate at the premetastatic niche in the lung to form cellular clusters and promote tumor metastasis through the VEGF/VEGFR1 and CXCL12/CXCR4 pathways (10). Moses and colleagues reported that a mutant form of TGF-β receptor recruits MDSCs (CD11b+ Gr-1+) at the tumor invasion front via CXCL12/CXCR4 and CXCL5/CXCR2 axes, contributing to lung metastasis of breast cancer (11). However, most of these findings were obtained from mouse models, and it remains to be determined whether such mechanisms are actually involved in humans.

SMAD4 is a key mediator of the TGF-β superfamily signaling and acts as a tumor suppressor in colorectal cancer. Loss of SMAD4 protein expression is found in 20% to 40% of colorectal cancer cases, and strongly correlated with the prognosis of colorectal cancer patients (15–18). We previously showed that in cis-Apc+/Δ716 Smad4+/– (Apc/Smad4) mice that develop invasive intestinal adenocarcinomas, mouse CCL9 (mCCL9) is secreted from the cancer epithelium, which recruits myeloid cells expressing its receptor CCR1 (19, 20). Using a mouse model of colorectal cancer liver metastasis, we also demonstrated that mCCL9-expressing colorectal cancer cell lines recruit CCR1+ myeloid cells to expand metastatic foci in the liver (21), and that four distinct types of myeloid cells are recruited to the metastatic foci: CCR1+ neutrophils, eosinophils, monocytes, and fibrocytes (22). In addition to these mouse models, we recently reported that SMAD4 binds directly to the promoter region of human CCL15 gene (a human ortholog of mouse CCL9) and negatively regulates its expression (23). Using human clinical specimens, we also showed that loss of SMAD4 promotes CCL15 expression to recruit CCR1+ myeloid cells and facilitates primary tumor invasion and liver metastasis of colorectal cancer (23, 24). Most CCR1+ cells accumulating at the invasion front of primary colorectal cancer were of the MDSC phenotype (CD11b+, CD33+, and HLA-DR; ref. 24). Importantly, patients with CCL15-expressing liver metastases showed significantly shorter relapse-free survival (RFS) than those with CCL15-negative ones (P = 0.04; ref. 23). Stage II/III patients with CCL15-positive primary colorectal cancer tended to have shorter RFS than those with CCL15-negative colorectal cancer (P = 0.15), although not a significant difference (24). However, it has not been investigated whether the CCL15-CCR1 chemokine axis is involved in lung metastasis of colorectal cancer.

Here, we report that in human clinical specimens, SMAD4-deficient lung metastases of colorectal cancer showed a significant correlation with CCL15 expression, which was associated with CCR1+ cell accumulation. Most of the CCR1+ cells accumulating around the lung metastases were of the TAN phenotype (CD11b+, CD33, HLA-DR, CD15+, and CD16+), although some were of the granulocytic-MDSC phenotype (CD11b+, CD33+, HLA-DR, and CD15+). Importantly, colorectal cancer patients with CCL15-positive lung metastases exhibited a significantly shorter RFS than those with CCL15-negative ones (P = 0.022). We also found that CCL15 expression of lung metastases could be a significant predictor of shorter RFS, as determined by multivariate Cox proportional hazards model analysis. In a mouse model of lung metastasis, SMAD4-deficient colorectal cancer cells showed high capacity of metastasis to the lung, which was suppressed by the treatment of CCR1 inhibitor. These results suggest that blocking the CCL15-CCR1 axis may provide a novel therapeutic strategy for lung metastases of colorectal cancer, and that CCL15 expression may be a novel predictive biomarker of colorectal cancer patients with lung metastases.

Patients' population

A total of 107 lung metastases were obtained from 81 colorectal cancer patients undergoing pulmonary resection at Kyoto University Hospital between 2006 and 2013, and their tissue samples were retrospectively analyzed. The diagnosis of colorectal cancer was confirmed by pathologic examination. The study protocol was approved by the Institutional Review Board of Kyoto University, and the patients provided their consents for data analysis.

Cell lines and reagents

HT29 and HCT116 cells were supplied from the American Type Culture Collection (ATCC) in the year 2011 during study initiation and were maintained in low glucose DMEM with 10% FBS and 1% penicillin/streptomycin mixture. All cell lines were authenticated by TAKARA Bio Inc. using DNA profiling of short tandem repeat markers and further verified by morphology and/or flow cytometry on July 2014. They were routinely tested negative for mycoplasma. Stable transductants of HT29 cells for firefly luciferase and cMyc-tagged human SMAD4 were established, as previously described (23, 24). Stable transductants of HCT116 cells for firefly luciferase and shRNA against SMAD4 (shSMAD4) were established, as previously described (23). A detailed list of the antibodies used was reported previously (23, 24). Anti-CD16 mouse antibody was obtained from Leica Biosystems and anti-Luciferase goat antibody from Promega.

Quantitative reverse transcription PCR

Total RNA was extracted using the RNeasy Mini Kit (Qiagen) according to the manufacturer's protocol. Complementary DNA generated by reverse transcription was quantified using ABI PRISM 7700 (Applied Biosystems). The primer probe sets for CCL15 and ACTB were from TaqMan Gene Expression assays (assay ID, Hs00361122_m1 and Hs99999903_m1, respectively). The mRNA level for CCL15 was normalized to that for ACTB by a ΔCt method.

Western blotting

Cells were lysed in NP-40 lysis buffer (20 mmol/L Tris-HCl, pH 7.5, 150 mmol/L NaCl, 10% glycerol, 1% NP-40, 10 mmol/L sodium fluoride, 1 mmol/L sodium pyrophosphate, 1 mmol/L sodium orthovanadate, and protease inhibitors). Cell lysates were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis, then immunoblotted with the respective primary antibodies followed by horseradish peroxidase–conjugated secondary antibodies, and analyzed. A detailed list of the antibodies used was reported previously (23).

Immunohistochemistry

Formalin-fixed, paraffin-embedded sections were stained with the respective antibodies by an avidin–biotin immunoperoxidase method. Expression of SMAD4 was evaluated as a nuclear staining, and the percentage of positively stained cells was scored as previously described (15, 24).Expression of CCL15 was interpreted as positive when > 10% of the tumor cells were stained, as previously described (23, 24). We quantified the densities of CCR1+ cells at the margin of lung metastases [5–9 fields (0.1mm2) analyzed per one sample], as previously described (23, 24). Three researchers (T. Yamamoto, Y. Itatani, and S. Inamoto) independently evaluated all immunohistochemistry samples without prior knowledge of other data. The slides with different evaluations among them were reinterpreted at a conference to reach the consensus.

Immunofluorescence analysis

After deparaffinization of 4-μm–sliced sections, heat-induced antigen retrieval was performed. The first antibody was applied and incubated overnight at 4°C. After fluorescence-labeled second antibody was applied (Alexa Fluor 488 anti-mouse/goat or 594 anti-rabbit), slides were mounted with a mounting medium including DAPI. Representative photos were taken by a fluorescence microscope.

Cell proliferation assay

Cells were incubated in 6-well plates (1 × 105 cells/well) and then cell numbers were counted at 1, 3, and 5 days later using a Countess Automated Cell Counter (Invitrogen).

In vivo xenograft studies

For experimental lung metastasis model, cancer cells (1.5 × 106 cells for HT29 or 3 × 106 cells for HCT116) were injected via tail veins of 8-week-old female nude mice. For in vivo bioluminescence imaging, 3 mg of D-luciferin (VivoGlo luciferin; Promega) was injected intraperitoneally into anesthetized tumor-bearing mice 10 minutes before imaging. Bioluminescence from the luciferase-expressing tumor cells was monitored at days 7, 14, 21, 28, and 35 after injection, using a Xenogen IVIS system (Xenogen Corporation). In the CCR1 antagonist–treated metastasis experiment, 3 × 106 cancer cells (Luc-HT29 cMyc-tag cells) were injected via tail veins of 8-week-old female SCID mice. As a CCR1 antagonist, J-113863 (10 mg/kg; TOCRIS Bioscience) was intraperitoneally administered from 1 day before tumor injection to the end of analyses (25). The animal experiment was approved by the Animal Care and Use Committee of Kyoto University.

Statistical analysis

Analyzed values are expressed as mean ± SD. Categorical data were determined with the Fisher exact test. Continuous variables were determined with the Student t test or Mann–Whitney U test. For factors associated with CCL15 expression, multivariate logistic regression analysis was used, and factors with a P value of < 0.15 were included in the model. Probability of survival was analyzed by the Kaplan–Meier method using the date of pulmonary resection as the starting point. The significance of differences between subgroups was calculated using the log-rank test. Multivariate analysis of the prognostic factors was carried out using the Cox's proportional hazard regression model. Relationships between variables were determined by Pearson's correlation coefficients. All analyses were two-sided, and a P value with < 0.05 was considered as statistically significant. Statistical analyses were performed with the JMP Pro software, version 11.0.0 (SAS Institute Inc).

Loss of SMAD4 in colorectal cancer lung metastases accumulates CCR1+ cells through CCL15-CCR1 chemokine axis in a mouse model

We previously reported that loss of SMAD4 causes recruitment of CCR1+ myeloid cells through CCL15-CCR1 axis to promote tumor invasion and liver metastasis of colorectal cancer (23, 24). Therefore, we hypothesized that CCL15 expression may also promote lung metastasis of colorectal cancer through the CCL15-CCR1 axis.

To test this hypothesis, we first examined the role of SMAD4 in lung metastasis of colorectal cancer using a mouse model. We injected luciferase (Luc)-expressing human colorectal cancer cells into the tail veins of nude mice, and then monitored the metastasized tumor cells within the lungs by bioluminescence. A SMAD4-deficient colorectal cancer cell line, HT29, endogenously expressed high levels of CCL15, whereas overexpression of cMyc-tagged SMAD4 in these cells (Luc-HT29 cMyc-SMAD4) dramatically reduced CCL15 expression without affecting cell viability (Supplementary Fig. S1A and S1B; refs. 23, 24). When Luc-HT29 cMyc-SMAD4 cells were injected into nude mice through the tail veins, their lung luciferase activities were reduced compared with the controls (Fig. 1A). We dissected lungs 35 days after injection and examined metastatic foci by bioluminescence. Although the SMAD4-deficient cells (Luc-HT29 cMyc-tag) formed foci in 38.9% of their lungs (14 of 36 lungs), the SMAD4-expressing cells (Luc-HT29 cMyc-SMAD4) metastasized to lungs in only 13.2% (5 of 38 lungs; Fig. 1B; P = 0.02). We histologically examined the mouse lungs containing metastatic colorectal cancer cells. As anticipated, lung tumors of Luc-HT29 cMyc-tag cells expressed CCL15, whereas those with Luc-HT29 cMyc-SMAD4 cells did not (Fig. 1C). Consistently, CCR1+ cells were accumulated around the lung metastases of Luc-HT29 cMyc-tag cells, whereas few CCR1+ cells accumulated around those of Luc-HT29 cMyc-SMAD4 cells (Fig. 1C). Next, we used another colorectal cancer cell line, HCT116, in which SMAD4 was endogenously expressed (Supplementary Fig. S1C–S1E). In HCT116 cells engineered with stable SMAD4 knockdown (Luc-HCT116 shSMAD4), shSMAD4 construct decreased SMAD4 expression and increased CCL15 expression, as previously reported (23). When injected into nude mice through the tail veins, Luc-HCT116 shSMAD4 cells formed foci in 50% of their lungs (9 of 18 lungs), while the control cells (Luc-HCT116 scramble) metastasized to lungs in only 25% (6 of 24 lungs; Fig. 1D and E; P = 0.12). Histologic analysis indicated that lung tumors of Luc-HCT116 shSMAD4 cells expressed CCL15, whereas those with Luc-HCT116 scramble cells did not (Fig. 1F). Consistently, CCR1+ cells were accumulated around the lung metastases of Luc-HCT116 shSMAD4 cells, whereas few CCR1+ cells accumulated around those of Luc-HCT116 scramble cells (Fig. 1F). Finally, we evaluated a CCR1 antagonist (J-113863) for its suppressive effect in the mouse lung metastasis model. J-113863 did not affect the proliferation rate of Luc-HT29 cMyc-tag cells in vitro (Supplementary Fig. S1F). We administered J-113863 (25) or the vehicle to SCID mice that were injected with Luc-HT29 cMyc-tag cells (Supplementary Fig. S1G). Luc-HT29 cMyc-tag cells do not necessarily metastasize to the lungs in nude mice (Fig. 1B), but we found they could inevitably metastasize to the lungs in SCID mice (Fig. 1G). Treatment of the host mice with 10 mg/kg of J-113863 tended to reduce the luminescence levels compared with vehicle-treated mice at day 21 (17.11 ± 18.54 vs. 5.54 ± 6.71, respectively; P = 0.12), although there was no apparent difference in the luminescence levels between the two groups at day 14 (Fig. 1H; Supplementary Fig. S1H). As anticipated, we verified by IHC analysis that J-113863 reduced the accumulation of CCR1+ cells around the lung metastases of Luc-HT29 cMyc-tag cells (Fig. 1I). These results indicate that loss of SMAD4 in colorectal cancer cells promotes CCL15 expression to recruit CCR1+ cells, which causes metastatic colonization of tumor cells disseminated to the lungs in this mouse model.

Figure 1.

Mouse xenograft model of colorectal cancer lung metastasis. A,In vivo bioluminescence images of nude mice injected with Luc-HT29 cells (left) and their lungs at 35 days after injection (right). B, Metastasis frequencies of HT29 clones to lungs at 35 days after injection. Metastasis frequencies based on the numbers of lungs or the number of mice were shown (Fisher exact test; P = 0.02 and 0.09, respectively). Note that each mouse has right and left lungs. C, Immunohistochemistry analyses of mouse lung with metastatic foci. Note that SMAD4 of cMyc-SMAD4 cells is expressed as a nuclear staining. Scale bar, 100 μm. D, Histologic analyses of mouse lung with metastatic foci. Scale bar, 100 μm. E, Metastasis frequencies of HCT116 clones to lungs at 28 days after injection. Metastasis frequencies based on the numbers of lungs or the number of mice were shown (Fisher exact test; P = 0.12 and 0.39, respectively). F, Immunohistochemistry analyses of mouse lung with metastatic foci. Note that SMAD4 of scramble cells is expressed as a nuclear staining. Scale bar, 100 μm. G,In vivo bioluminescence images of SCID mice injected with Luc-HT29 cells (left) and their lungs at 21 days after injection (right). H, Relative quantification of the lung metastatic lesions (photon counts) in SCID mice treated with vehicle or J-113863 at 10 mg/kg. Each triangle represents an individual mouse. Each photon count at days 14 and 21 was normalized to that at day 7. Horizontal lines show the means of representative groups (Student t test; P = 0.12 at day 21). n = 6–8 mice in each group. I, Immunohistochemistry analyses of lung metastases of mice treated with vehicle or J-113863. T, Tumor; scale bar, 100 μm.

Figure 1.

Mouse xenograft model of colorectal cancer lung metastasis. A,In vivo bioluminescence images of nude mice injected with Luc-HT29 cells (left) and their lungs at 35 days after injection (right). B, Metastasis frequencies of HT29 clones to lungs at 35 days after injection. Metastasis frequencies based on the numbers of lungs or the number of mice were shown (Fisher exact test; P = 0.02 and 0.09, respectively). Note that each mouse has right and left lungs. C, Immunohistochemistry analyses of mouse lung with metastatic foci. Note that SMAD4 of cMyc-SMAD4 cells is expressed as a nuclear staining. Scale bar, 100 μm. D, Histologic analyses of mouse lung with metastatic foci. Scale bar, 100 μm. E, Metastasis frequencies of HCT116 clones to lungs at 28 days after injection. Metastasis frequencies based on the numbers of lungs or the number of mice were shown (Fisher exact test; P = 0.12 and 0.39, respectively). F, Immunohistochemistry analyses of mouse lung with metastatic foci. Note that SMAD4 of scramble cells is expressed as a nuclear staining. Scale bar, 100 μm. G,In vivo bioluminescence images of SCID mice injected with Luc-HT29 cells (left) and their lungs at 21 days after injection (right). H, Relative quantification of the lung metastatic lesions (photon counts) in SCID mice treated with vehicle or J-113863 at 10 mg/kg. Each triangle represents an individual mouse. Each photon count at days 14 and 21 was normalized to that at day 7. Horizontal lines show the means of representative groups (Student t test; P = 0.12 at day 21). n = 6–8 mice in each group. I, Immunohistochemistry analyses of lung metastases of mice treated with vehicle or J-113863. T, Tumor; scale bar, 100 μm.

Close modal

Loss of SMAD4 is associated with CCL15 expression and CCR1+ cell accumulation, which can result in lung metastasis

In order to confirm the clinical relevance of the results from the mouse model, we next investigated expressions of SMAD4, CCL15, and CCR1 using clinical specimens of colorectal cancer lung metastases resected between 2006 and 2013 (Table 1). Of 107 lung metastases obtained from 81 patients, we found that 58% (62 of 107) was negative for SMAD4 expression, whereas 42% (45 of 107) was positive (Table 1). Immunohistochemistry indicated that CCL15 was expressed mainly at the marginal parts rather than the center of the lung metastases (Fig. 2A), and that CCL15 expression was positive in 76% (81 of 107; Table 1), which is similar to the frequencies in primary tumors and liver metastases of colorectal cancer (23, 24). Especially, CCL15 expression was positive in 85% (53 of 62) of the SMAD4-negative metastases, whereas it was found in 62% (28 of 45) of the SMAD4-positive metastases, with a significant inverse correlation between SMAD4 and CCL15 (odds ratio = 0.22; P < 0.01, Table 1).

Table 1.

Univariate analysis of factors associated with CCL15 expression

FactorsCCL15 negative (n = 26)CCL15 positive (n = 81)P value
Tumor size (mm) 17.4 ± 15.2 14.5 ± 10.7 0.68 
SMAD4 expression 
 Negative 53  
 Positive 17 28 <0.01 
CCR1-positive cell count 12.0 ± 6.8 22.6 ± 16.0 <0.01 
FactorsCCL15 negative (n = 26)CCL15 positive (n = 81)P value
Tumor size (mm) 17.4 ± 15.2 14.5 ± 10.7 0.68 
SMAD4 expression 
 Negative 53  
 Positive 17 28 <0.01 
CCR1-positive cell count 12.0 ± 6.8 22.6 ± 16.0 <0.01 
Figure 2.

Correlation of SMAD4 loss and expression of CCL15 and CCR1 in clinical specimens of colorectal cancer lung metastases. A, Hematoxylin and eosin (H&E) staining and immunohistochemical detection of SMAD4, CCL15, and CCR1 in lung metastases. Top and bottom panels show serial sections of representative SMAD4-deficient and SMAD4-expressing colorectal cancer, respectively. Note that SMAD4 of cancer cells is expressed as a nuclear staining. Scale bar, 100 μm. B, Quantification of the CCR1+ cell density in colorectal cancer lung metastases with and without CCL15 expression (n = 26 and 81, respectively). *, P < 0.01; Mann–Whitney U test; horizontal bands show the means. C, Scatter plot of metastatic tumor size and CCR1+ cell accumulation. The relationship was evaluated by the Pearson's correlation coefficient (r, correlation coefficient). D, Effects of CCL15 expression on RFS in CRC patients who underwent curative pulmonary resection (Kaplan–Meier estimates).

Figure 2.

Correlation of SMAD4 loss and expression of CCL15 and CCR1 in clinical specimens of colorectal cancer lung metastases. A, Hematoxylin and eosin (H&E) staining and immunohistochemical detection of SMAD4, CCL15, and CCR1 in lung metastases. Top and bottom panels show serial sections of representative SMAD4-deficient and SMAD4-expressing colorectal cancer, respectively. Note that SMAD4 of cancer cells is expressed as a nuclear staining. Scale bar, 100 μm. B, Quantification of the CCR1+ cell density in colorectal cancer lung metastases with and without CCL15 expression (n = 26 and 81, respectively). *, P < 0.01; Mann–Whitney U test; horizontal bands show the means. C, Scatter plot of metastatic tumor size and CCR1+ cell accumulation. The relationship was evaluated by the Pearson's correlation coefficient (r, correlation coefficient). D, Effects of CCL15 expression on RFS in CRC patients who underwent curative pulmonary resection (Kaplan–Meier estimates).

Close modal

Univariate analysis of each clinicopathologic factor indicated that CCL15 expression was significantly correlated with SMAD4 expression (P = 0.02), and tended to be correlated with tumor location of the primary colorectal cancer (P = 0.13) and T factor of the primary colorectal cancer (P = 0.14; Table 2). In the multivariate analysis including factors with a P value of < 0.15, only SMAD4 expression remained significantly correlated with CCL15 expression (Table 2; P = 0.043), showing a significant inverse correlation between CCL15 and SMAD4.

Table 2.

Univariate and multivariate analyses of clinicopathologic factors for CCL15 expression

Univariate analysisMultivariate analysis
VariablesCCL15 negativeCCL15 positiveP valueOR (95% CI)P value
Age 
 <60 15    
 60 ≤ 12 46 0.25   
Sex 
 Male 12 36    
 Female 25 1.00   
Location 
 Colon 34    
 Rectum 13 27 0.13 1.73 (0.56–5.56) 0.34 
Histology 
 tub1/tub2 14 42    
 por/sig/muci 1.00   
T factor 
 1/2    
 3/4 15 56 0.14 0.47 (0.07–3.09) 0.42 
N factor 
 Negative 22    
 Positive 12 38 1.00   
M factor 
 Negative 12 40    
 Positive 20 0.79   
H factor 
 Negative 15 49    
 Positive 11 0.75   
Venous invasion 
 Negative 11    
 Positive 39 0.31   
Lymphatic invasion 
 Negative 29    
 Positive 22 0.38   
Prior chemotherapy 
 Negative 34    
 Positive 11 27 0.45   
Neoadjuvant treatment 
 Negative 13 43    
 Positive 18 0.78   
Timing 
 Synchronous 10    
 Metachronous 17 51 1.00   
Other metastases 
 Negative 25    
 Positive 11 35 0.80   
Tumor size 
 <20 mm 15 41    
 20 mm≤ 17 0.78   
Number of metastases 
 1/2 19 57    
 3≤ 1.00   
Hilar/mediastinal LN 
 Negative 20 57    
 Positive 0.57   
Disease-free interval 
 <2years 21    
 2 years ≤ 12 40 0.79   
CEA (ng/mL) 
 <5.0 15 45    
 5.0 ≤ 15 1.00   
CA19–9 (U/mL) 
 < 37 31    
 37 ≤ 1.00   
NLR 
 <3.0 16 43    
 3.0 ≤ 18 0.56   
SMAD4 expression 
 Negative 40    
 Positive 13 21 0.02 3.26 (1.04–10.97) 0.043 
Univariate analysisMultivariate analysis
VariablesCCL15 negativeCCL15 positiveP valueOR (95% CI)P value
Age 
 <60 15    
 60 ≤ 12 46 0.25   
Sex 
 Male 12 36    
 Female 25 1.00   
Location 
 Colon 34    
 Rectum 13 27 0.13 1.73 (0.56–5.56) 0.34 
Histology 
 tub1/tub2 14 42    
 por/sig/muci 1.00   
T factor 
 1/2    
 3/4 15 56 0.14 0.47 (0.07–3.09) 0.42 
N factor 
 Negative 22    
 Positive 12 38 1.00   
M factor 
 Negative 12 40    
 Positive 20 0.79   
H factor 
 Negative 15 49    
 Positive 11 0.75   
Venous invasion 
 Negative 11    
 Positive 39 0.31   
Lymphatic invasion 
 Negative 29    
 Positive 22 0.38   
Prior chemotherapy 
 Negative 34    
 Positive 11 27 0.45   
Neoadjuvant treatment 
 Negative 13 43    
 Positive 18 0.78   
Timing 
 Synchronous 10    
 Metachronous 17 51 1.00   
Other metastases 
 Negative 25    
 Positive 11 35 0.80   
Tumor size 
 <20 mm 15 41    
 20 mm≤ 17 0.78   
Number of metastases 
 1/2 19 57    
 3≤ 1.00   
Hilar/mediastinal LN 
 Negative 20 57    
 Positive 0.57   
Disease-free interval 
 <2years 21    
 2 years ≤ 12 40 0.79   
CEA (ng/mL) 
 <5.0 15 45    
 5.0 ≤ 15 1.00   
CA19–9 (U/mL) 
 < 37 31    
 37 ≤ 1.00   
NLR 
 <3.0 16 43    
 3.0 ≤ 18 0.56   
SMAD4 expression 
 Negative 40    
 Positive 13 21 0.02 3.26 (1.04–10.97) 0.043 

Regarding the relationship between CCL15 expression and CCR1+ cell accumulation, we found that numerous CCR1+ cells accumulated around the CCL15-positive colorectal cancer cells in the lung, but few around the CCL15-negative ones (Fig. 2A). We quantified the density of CCR1+ cells at the periphery of lung metastases, and found that approximately 1.9 times more CCR1+ cells around the CCL15-positive metastases than around CCL15-negative ones (12.0 ± 6.8 vs. 22.6 ± 16.0; P < 0.01, Table 1; Fig. 2B). These results are consistent with the colorectal cancer data for the primary tumors and liver metastases (23, 24). We also determined whether CCR1+ cell density was correlated with the metastatic tumor size, and found little correlation between them (P = 0.249; Pearson's correlation coefficients; Fig. 2C), which is different from the data of the liver metastases of colorectal cancer (23).

Expression of CCL15 in lung metastases predicts poor survival after pulmonary resection of metastatic colorectal cancer

To evaluate the clinical outcome of CCL15 expression, we analyzed RFS of 67 colorectal cancer patients who underwent curative resection of lung metastases. The median follow-up was 32 months, and the estimated RFS rate at 3- and 5-year was 38.8% and 33.1%, respectively (Supplementary Fig. S2A–S2D). Statistical analyses showed that the colorectal cancer patients with CCL15-positive metastases exhibited a significantly shorter RFS than those with CCL15-negative metastases(P = 0.02, Fig. 2D). The cumulative 3-year RFS rate of colorectal cancer patients with CCL15-positive metastases was 33.0%, whereas that of those with CCL15-negative metastases was 66.0%. We previously reported that CCL15 expression on primary tumors or liver metastases of colorectal cancer is correlated with patients' RFS (23, 24). The present results are consistent with our previous studies, supporting that SMAD4-deficient colorectal cancer can worsen patients' survival through the CCL15-CCR1 chemokine axis to accumulate CCR1+ cells in the tumor microenvironment. We also analyzed overall survival (OS) of 81 patients who underwent pulmonary resection (Supplementary Fig. S3A), and found that CCL15 expression tended to correlate with shorter OS although without a significant difference (P = 0.14; Supplementary Fig. S3B).

In colorectal cancer patients who underwent complete pulmonary resection in a large-scale retrospective analysis, it has been recently reported that the number of pulmonary metastatic lesions, their distribution, the tumor size, the serum carcinoembryonic antigen (CEA) level prior to pulmonary resection, the disease-free interval, and the lymph node metastases of primary colorectal cancer are independent prognostic factors for prognosis (5–8). Therefore, we also analyzed the effect of CCL15 expression in conjunction with these factors in this study. According to a univariate analysis, T stage of primary colorectal cancer, N stage of primary colorectal cancer, CEA level prior to pulmonary resection, and CCL15 expression in lung metastases were significant predictors of shorter RFS (Table 3; Supplementary Fig. S2B–S2D). In the multivariate analysis including these variables, CCL15 expression in lung metastases was the only parameter that remained significant for predicting shorter RFS [Table 3; HR, 2.53; 95% confidence interval (CI), 1.11–6.67; P = 0.026].

Table 3.

Univariate and multivariate analyses of clinicopathologic factors for relapse-free survival

Univariate analysisMultivariate analysis
VariablesNHR (95% CI)P valueHR (95% CI)P value
Age 
 <60 20     
 60≤ 47 1.03 (0.54–2.10) 0.93   
Sex 
 Male 37     
 Female 30 0.66 (0.34–1.24) 0.20   
Location 
 Colon 34     
 Rectum 33 0.82 (0.44–1.52) 0.52   
Histology 
 tub1/tub2 46     
 por/sig/muci 0.71 (0.11–2.35) 0.62   
T factor 
 1/2     
 3/4 58 3.30 (1.00–20.3) 0.049 2.18 (0.59–14.36) 0.27 
N factor 
 Negative 23     
 Positive 43 2.03 (1.04–4.25) 0.038 0.81 (0.26–3.52) 0.75 
M factor 
 Negative 44     
 Positive 22 1.83 (0.96–3.43) 0.07   
H factor 
 Negative 52     
 Positive 14 1.54 (0.73–3.01) 0.24   
Venous invasion 
 Negative 11     
 Positive 43 0.98 (0.45–2.45) 0.96   
Lymphatic invasion 
 Negative 28     
 Positive 27 1.31 (0.67–2.59) 0.44   
Prior chemotherapy 
 Negative 37     
 Positive 30 0.98 (0.52–1.81) 0.95   
Neoadjuvant treatment 
 Negative 50     
 Positive 17 1.49 (0.73–2.85) 0.26   
Timing 
 Synchronous 12     
 Metachronous 55 0.70 (0.35–1.50) 0.34   
Other metastases 
 Negative 31     
 Positive 36 1.53 (0.83–2.91) 0.18   
Tumor size 
 <20 mm 48     
 20 mm ≤ 17 0.95 (0.44–1.89) 0.90   
Number of metastases 
 1/2 63     
 3 ≤ 0.83 (0.14–2.72) 0.79   
Distribution 
 Unilateral 57     
 Bilateral 10 0.94 (0.32–2.20) 0.90   
Hilar/mediastinal LN 
 Negative 63     
 Positive 1.77 (0.43–4.91) 0.38   
Disease-free interval 
 <2 years 25     
 2 years ≤ 42 0.94 (0.51–1.78) 0.84   
CEA (ng/mL) 
 <5.0 50     
 5.0 ≤ 16 2.07 (1.03–3.94) 0.041 1.97 (0.90–4.14) 0.09 
CA19–9 (U/mL) 
 <37 32     
 37 ≤ 3.17 (0.87–9.47) 0.08   
NLR 
 <3.0 48     
 3.0 ≤ 19 1.08 (0.53–2.08) 0.82   
CCL15 
 Negative 15     
 Positive 52 2.58 (1.19–6.47) 0.015 2.54 (1.11–6.67) 0.026 
SMAD4 
 Negative 37     
 Positive 30 0.63 (0.33–1.17) 0.15   
Univariate analysisMultivariate analysis
VariablesNHR (95% CI)P valueHR (95% CI)P value
Age 
 <60 20     
 60≤ 47 1.03 (0.54–2.10) 0.93   
Sex 
 Male 37     
 Female 30 0.66 (0.34–1.24) 0.20   
Location 
 Colon 34     
 Rectum 33 0.82 (0.44–1.52) 0.52   
Histology 
 tub1/tub2 46     
 por/sig/muci 0.71 (0.11–2.35) 0.62   
T factor 
 1/2     
 3/4 58 3.30 (1.00–20.3) 0.049 2.18 (0.59–14.36) 0.27 
N factor 
 Negative 23     
 Positive 43 2.03 (1.04–4.25) 0.038 0.81 (0.26–3.52) 0.75 
M factor 
 Negative 44     
 Positive 22 1.83 (0.96–3.43) 0.07   
H factor 
 Negative 52     
 Positive 14 1.54 (0.73–3.01) 0.24   
Venous invasion 
 Negative 11     
 Positive 43 0.98 (0.45–2.45) 0.96   
Lymphatic invasion 
 Negative 28     
 Positive 27 1.31 (0.67–2.59) 0.44   
Prior chemotherapy 
 Negative 37     
 Positive 30 0.98 (0.52–1.81) 0.95   
Neoadjuvant treatment 
 Negative 50     
 Positive 17 1.49 (0.73–2.85) 0.26   
Timing 
 Synchronous 12     
 Metachronous 55 0.70 (0.35–1.50) 0.34   
Other metastases 
 Negative 31     
 Positive 36 1.53 (0.83–2.91) 0.18   
Tumor size 
 <20 mm 48     
 20 mm ≤ 17 0.95 (0.44–1.89) 0.90   
Number of metastases 
 1/2 63     
 3 ≤ 0.83 (0.14–2.72) 0.79   
Distribution 
 Unilateral 57     
 Bilateral 10 0.94 (0.32–2.20) 0.90   
Hilar/mediastinal LN 
 Negative 63     
 Positive 1.77 (0.43–4.91) 0.38   
Disease-free interval 
 <2 years 25     
 2 years ≤ 42 0.94 (0.51–1.78) 0.84   
CEA (ng/mL) 
 <5.0 50     
 5.0 ≤ 16 2.07 (1.03–3.94) 0.041 1.97 (0.90–4.14) 0.09 
CA19–9 (U/mL) 
 <37 32     
 37 ≤ 3.17 (0.87–9.47) 0.08   
NLR 
 <3.0 48     
 3.0 ≤ 19 1.08 (0.53–2.08) 0.82   
CCL15 
 Negative 15     
 Positive 52 2.58 (1.19–6.47) 0.015 2.54 (1.11–6.67) 0.026 
SMAD4 
 Negative 37     
 Positive 30 0.63 (0.33–1.17) 0.15   

Characterization of CCR1+ cells accumulated around lung metastases of colorectal cancer

We previously reported that CCR1+ cells accumulating around the invasion front of primary colorectal cancer were of the MDSC phenotype, and that most of these CCR1+ MDSCs were of the granulocytic-MDSC lineage (24). In order to determine whether the CCR1+ cells accumulating around lung metastases of colorectal cancer were of the same phenotype as those in the primary colorectal cancer, we performed double immunofluorescence staining with several cell-type–specific markers. The majority of these CCR1+ cells were positive for MPO, CD11b, CD15, and CD16, whereas they were negative for HLA-DR, CD33, CD3, CD8, CD14, CD68, and α-SMA (Fig. 3A–F; Supplementary Fig. S4A–S4E), suggesting that they were of the TAN phenotype (CD11b+, CD33, HLA-DR, CD15+, and CD16+). TANs are characterized by having immunosuppressive (by producing ARG1) and angiogenic (by producing MMP) functions (12, 13, 14). Consistently, we confirmed that these CCR1+ cells expressed ARG1 in addition to MMP9 (Fig. 3G and H), and that a subpopulation of these CCR1+ cells was positive for CXCR2 (Supplementary Fig. S4F), a major chemokine receptor expressed on TAN (12). Moreover, we also observed that only a minor fraction of CCR1+ cells were positive for CD11b, CD33, and CD15 with negative HLA-DR (Supplementary Fig. S5A–S5D), suggesting that they were of the granulocytic-MDSC lineage.

Figure 3.

Characterization of CCR1+ cells in the tumor microenvironment of CRC lung metastases. Simultaneous immunofluorescence staining for CCR1 (red) and MPO (A), CD11b (B), CD33 (C), HLA-DR (D), CD15 (E), CD16 (F), ARG1 (G), or MMP9 (H; green). Scale bar, 20 μm.

Figure 3.

Characterization of CCR1+ cells in the tumor microenvironment of CRC lung metastases. Simultaneous immunofluorescence staining for CCR1 (red) and MPO (A), CD11b (B), CD33 (C), HLA-DR (D), CD15 (E), CD16 (F), ARG1 (G), or MMP9 (H; green). Scale bar, 20 μm.

Close modal

It has been shown that various types of myeloid cells promote tumor progression by immune suppression as well as by facilitating cell growth, migration, and angiogenesis. These cells include TAMs, MDSCs, and more recently TANs (12, 13, 14). TAMs, the best-characterized among them, are divided into two populations: M2 macrophages that promote tumorigenesis and M1 macrophages that are antitumorigenic. MDSCs constitute a heterogeneous population of immature myeloid cells that are increased in cancer, inflammation, and infection (26). Human MDSCs comprise two subpopulations; monocytic-MDSCs (HLA-DR, CD11b+, CD33+, and CD14+), and granulocytic-MDSCs HLA-DR, CD11b+, CD33+, and CD15+). On the other hand, the presence and functional characteristics of TANs may have been underestimated and therefore need reappraisal. As defined with M1 and M2 TAMs, TANs also have differential states of activation/differentiation: N1 neutrophils are antitumorigenic, whereas N2 neutrophils display protumorigenic properties (27, 28). It has been reported that, in a mouse lung cancer model, TGF-β is a major cytokine within tumors that defines the TAN phenotype and skews differentiation toward the N2 phenotype (27). We previously reported that CCL15 expression by loss of SMAD4 in cancer cells helps recruitment of CCR1+ MDSCs (CD11b+, CD33+, and HLA-DR) at the invasion front of primary colorectal cancer (24). In this study of the lung metastases of colorectal cancer, we have verified that CCL15 expression was significantly correlated with loss of SMAD4 (Table 1), and that CCL15-positive metastases recruited more numbers of CCR1+ cells than CCL15-negative metastases did (Table 1; Fig. 2B). CCR1+ cells accumulating around lung metastases were positive for MPO, CD11b, CD15, and CD16 (Fig. 3A, B, E, and F), whereas they were negative for HLA-DR, CD3, CD8, and α-SMA (Fig. 3D, and Supplementary Fig S4A, S4B, and S4D), suggesting their myeloid origin rather than lymphoid or fibroblastic origins. Contrary to our anticipation, most of these CCR1+ cells expressed mature myeloid cell markers (positive MPO and negative CD33; Fig. 3A and C), and granulocyte-lineage markers (MPO, CD15, and CD16; Fig. 3A, E, and F), while they lacked monocyte-lineage markers (CD68 and CD14; Supplementary Fig. S4C and S4E). These cells may contribute to the development of tumor microenvironment through production of MMP9 and ARG1 (Fig. 3G and H). We also found that only a minor fraction of CCR1+ cells were positive for CD11b, CD33, and CD15 with negative HLA-DR (Supplementary Fig. S5A–S5D). Taken together, these results indicate that CCR1+ cells accumulating around lung metastases of colorectal cancer are composed mainly of the TAN and only partially of the granulocytic-MDSC, which may suggest these myeloid cell subsets are specific to the tumor and organ. Regarding the CCR1+ cells accumulating around colorectal cancer liver metastases, we previously reported that these CCR1+ cells were positive for both CD11b and MPO (23). In this study, we further characterized these CCR1+ cells using double immunofluorescence staining, and found that most of CCR1+ cells were positive for MPO, CD11b, CD15, CD16, ARG1, and MMP9 while they were negative for HLA-DR and CD33 (Supplementary Fig. S6A–S6H), which suggests that they are composed mainly of TAN. Regarding the relationship between TANs and MDSCs, it has been suggested that MDSCs can enter tumors and differentiate to mature TAMs or TANs (29). It remains to be determined whether the neutrophils within the tumors are derived from granulocytic-MDSCs or whether they are blood-derived neutrophils converted to the N2 phenotype by the tumor environment (e.g., high local concentration of TGF-β). To further characterize the cell lineage and the functions of these myeloid cells, our next step would be to isolate viable tumor-associated CCR1+ myeloid cells from the tumor tissues by fluorescence-activated cell sorting. Several studies show that an elevated neutrophil count in the peripheral blood [i.e., a high neutrophil-to-lymphocyte ratio (NLR)] is associated with poor clinical outcomes in several types of cancer including colorectal cancer (30), emphasizing the importance and relevance of neutrophils in cancer biology.

Chemokine CCL15, a member of the CC chemokine family and constitutively expressed in the gut and liver, was cloned and partially characterized (31). CCL15 plays an important role in the development of inflammation and allergic diseases (32, 33, 34), and binds mainly to CCR1 as a chemoattractant for monocytes, neutrophils, lymphocytes, and dendritic cells (35). Accumulating evidence has suggested that CCL15 has a crucial role in the progression of several types of cancer including colorectal cancer, hepatocellular carcinoma (HCC), and lung cancer (23, 24, 36–39). We previously showed that SMAD4 binds directly to the promoter region of human CCL15 gene to negatively regulate its expression, and that CCL15 expression is inversely correlated with SMAD4 expression in clinical specimens of human primary colorectal cancer and its liver metastasis (23, 24). In the lung metastases of colorectal cancer, we have determined here that there is a significant inverse correlation between CCL15 and SMAD4 levels by immunohistochemical analysis of 107 clinical samples (Tables 1 and 2). Expression of SMAD4 was lost in 58% (62 of 107), whereas that of CCL15 was found in 76% (81 of 107). The difference of approximately 20% between these frequencies suggests that factors other than SMAD4 can also contribute to CCL15 expression. In fact, the promoter of CCL15 contains multiple potential binding sites for transcription factors involved in the immune system and inflammation, although the functionality of these regions are yet to be investigated experimentally (35). It has been reported that CCL15 is the most significant serum markers associated with short survival in early-stage (i.e., Stage I/II) lung cancer (38). In HCC, CCL15 is one of the independent predictors of shorter OS as well as tumor–node–metastasis stage (39). Our previous reports show that CCL15 is associated with a shorter RFS in liver metastasis and Stage I/II primary colorectal cancer (23, 24). In this study, we have also demonstrated that patients with CCL15-positive lung metastases show a significantly shorter RFS than those with CCL15-negative ones (P = 0.02; Fig. 2D). It is worth noting that a multivariate analysis has indicated that CCL15 expression is an independent predictor of shorter RFS (P = 0.026; Table 3), although further studies with a large multi-institutional cohort are required. Taken together, CCL15 expression within tumors may be associated with poor clinical outcomes in several types of cancer.

Chemokines have overlapping actions, and their activities can be altered by proteolytic degradation. Therefore, effective targeting of chemokine pathways can require high levels of receptor occupancy throughout treatment. In mouse models, tumor invasion and metastasis of colorectal cancer were suppressed by inhibiting CCL15-CCR1 signaling with a CCR1 inhibitor, genetic knockdown of CCL15 gene, or CCR1 knockout mice (20, 21, 23, 24, 40). In addition, it has been reported recently that activation of CCL2/CCR2 axis prompts TAMs to secrete another chemokine CCL3, which in turn activates its receptor CCR1 in TAMs and promotes lung metastatic seeding of breast cancer in a mouse model (41). These results suggest that inhibition of CCR1 may have a therapeutic impact on metastatic progression of certain types of cancer, including colorectal cancer and breast cancer. A number of CCR1 inhibitors have already been used in phase I/II clinical trials for rheumatoid arthritis, chronic obstructive pulmonary disease, and multiple sclerosis (42, 43). Thus, it would be worthwhile to test CCR1 inhibitors to prevent cancer metastasis, because they have been already cleared for safety concerns.

No potential conflicts of interest were disclosed.

Conception and design: K. Kawada, Y. Itatani, T.F. Chen-Yoshikawa, Y. Sakai

Development of methodology: K. Kawada, Y. Itatani

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T. Yamamoto, S. Inamoto, M. Iwamoto, T.F. Chen-Yoshikawa, H. Date

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T. Yamamoto, K. Kawada, S. Inamoto, R. Okamura, H. Date

Writing, review, and/or revision of the manuscript: T. Yamamoto, K. Kawada, Y. Itatani, S. Hasegawa, H. Date, M.M. Taketo

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Iwamoto, E. Miyamoto, H. Hirai, H. Date

Study supervision: Y. Itatani, H. Date, M.M. Taketo, Y. Sakai

This study was supported by grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan (grant number 24591975), from Japan Research Foundation for Clinical Pharmacology, and from Mochida Memorial Foundation for Medical and Pharmaceutical Research (to K. Kawada).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Mitry
E
,
Guiu
B
,
Cosconea
S
,
Jooste
V
,
Faivre
J
,
Bouvier
AM
. 
Epidemiology, management and prognosis of colorectal cancer with lung metastases: A 30-year population-based study
.
Gut
2010
;
59
:
1383
8
.
2.
Kobayashi
H
,
Mochizuki
H
,
Sugihara
K
,
Morita
T
,
Kotake
K
,
Teramoto
T
, et al
Characteristics of recurrence and surveillance tools after curative resection for colorectal cancer: A multicenter study
.
Surgery
2007
;
141
:
67
75
.
3.
Watanabe
T
,
Itabashi
M
,
Shimada
Y
,
Tanaka
S
,
Ito
Y
,
Ajioka
Y
, et al
Japanese society for cancer of the colon and rectum (JSCCR) Guidelines 2014 for treatment of colorectal cancer
.
Int J Clin Oncol
2015
;
20
:
207
39
.
4.
Chen
F
,
Sakai
H
,
Miyahara
R
,
Bando
T
,
Okubo
K
,
Date
H
. 
Repeat resection of pulmonary metastasis is beneficial for patients with colorectal carcinoma
.
World J Surg
2010
;
34
:
2373
8
.
5.
Iida
T
,
Nomori
H
,
Shiba
M
,
Nakajima
J
,
Okumura
S
,
Horio
H
, et al
Prognostic factors after pulmonary metastasectomy for colorectal cancer and rationale for determining surgical indications: A retrospective analysis
.
Ann Surg
2013
;
257
:
1059
64
.
6.
Hirosawa
T
,
Itabashi
M
,
Ohnuki
T
,
Yamaguchi
N
,
Sugihara
K
,
Kameoka
S
. 
Japanese Society for Cancer of the Colon and Rectum (JSCCR) study group for pulmonary metastases from colorectal cancer. Prognostic factors in patients undergoing complete resection of pulmonary metastases of colorectal cancer: a multi-institutional cumulative follow-up study
.
Surg Today
2013
;
43
:
494
9
.
7.
Suzuki
H
,
Kiyoshima
M
,
Kitahara
M
,
Asato
Y
,
Amemiya
R
. 
Long-term outcomes after surgical resection of pulmonary metastases from colorectal cancer
.
Ann Thorac Surg
2015
;
99
:
435
40
.
8.
Hirosawa
T
,
Itabashi
M
,
Ohnuki
T
,
Yamaguchi
N
,
Sugihara
K
,
Kameoka
S
. 
Japanese Society for Cancer of the Colon and Rectum (JSCCR) study group for pulmonary metastases from colorectal cancer. Proposal of a new prognostic staging system for pulmonary metastases from colorectal cancer
.
Surg Today
2015
;
45
:
576
84
.
9.
Chen
F
,
Hanaoka
N
,
Sato
K
,
Fujinaga
T
,
Sonobe
M
,
Shoji
T
, et al
Prognostic factors of pulmonary metastasectomy for colorectal carcinomas
.
World J Surg
2009
;
33
:
505
11
.
10.
Kaplan
RN
,
Riba
RD
,
Zacharoulis
S
,
Bramley
AH
,
Vincent
L
,
Costa
C
, et al
VEGFR1-positive haematopoietic bone marrow progenitors initiate the pre-metastatic niche
.
Nature
2005
;
438
:
820
7
.
11.
Yang
L
,
Huang
J
,
Ren
X
,
Gorska
AE
,
Chytil
A
,
Aakre
M
, et al
Abrogation of TGFβ signaling in mammary carcinomas recruits Gr-1+CD11b+ myeloid cells that promote metastasis
.
Cancer Cell
2008
;
13
:
23
35
.
12.
Murdoch
C
,
Muthana
M
,
Coffelt
SB
,
Lewis
CE
. 
The role of myeloid cells in the promotion of tumour angiogenesis
.
Nat Rev Cancer
2008
;
8
:
618
31
.
13.
Gabrilovich
DI
,
Ostrand-Rosenberg
S
,
Bronte
V
. 
Coordinated regulation of myeloid cells by tumours
.
Nat Rev Immunol
2012
;
12
:
253
68
.
14.
Kitamura
T
,
Qian
BZ
,
Pollard
JW
. 
Immune cell promotion of metastasis
.
Nat Rev Immunol
2015
;
15
:
73
86
.
15.
Salovaara
R
,
Roth
S
,
Loukola
A
,
Launonen
V
,
Sistonen
P
,
Avizienyte
E
, et al
Frequent loss of SMAD4/DPC4 protein in colorectal cancers
.
Gut
2002
;
51
:
56
9
.
16.
Roth
AD
,
Delorenzi
M
,
Tejpar
S
,
Yan
P
,
Klingbiel
D
,
Fiocca
R
, et al
Integrated analysis of molecular and clinical prognostic factors in stage II/III colon cancer
.
J Natl Cancer Inst
2012
;
104
:
1635
46
.
17.
Alazzouzi
H
,
Alhopuro
P
,
Salovaara
R
,
Sammalkorpi
H
,
Järvinen
H
,
Mecklin
JP
, et al
SMAD4 as a prognostic marker in colorectal cancer
.
Clin Cancer Res
2005
;
11
:
2606
11
.
18.
Voorneveld
PW
,
Jacobs
RJ
,
Kodach
LL
,
Hardwick
JC
. 
A meta-analysis of SMAD4 immunohistochemistry as a prognostic marker in colorectal cancer
.
Transl Oncol
2015
;
8
:
18
24
.
19.
Takaku
K
,
Oshima
M
,
Miyoshi
H
,
Matsui
M
,
Seldin
MF
,
Taketo
MM
. 
Intestinal tumorigenesis in compound mutant mice of both Dpc4 (Smad4) and Apc genes
.
Cell
1998
;
92
:
645
56
.
20.
Kitamura
T
,
Kometani
K
,
Hashida
H
,
Matsunaga
A
,
Miyoshi
H
,
Hosogi
H
, et al
SMAD4-deficient intestinal tumors recruit CCR1+ myeloid cells that promote invasion
.
Nat Genet
2007
;
39
:
467
75
.
21.
Kitamura
T
,
Fujishita
T
,
Loetscher
P
,
Revesz
L
,
Hashida
H
,
Kizaka-Kondoh
S
, et al
Inactivation of chemokine (C-C motif) receptor 1 (CCR1) suppresses colon cancer liver metastasis by blocking accumulation of immature myeloid cells in a mouse model
.
Proc Natl Acad Sci USA
2010
;
107
:
13063
8
.
22.
Hirai
H
,
Fujishita
T
,
Kurimoto
K
,
Miyachi
H
,
Kitano
S
,
Inamoto
S
, et al
CCR1-mediated accumulation of myeloid cells in the liver microenvironment promoting mouse colon cancer metastasis
.
Clin Exp Metastasis
2014
;
31
:
977
89
.
23.
Itatani
Y
,
Kawada
K
,
Fujishita
T
,
Kakizaki
F
,
Hirai
H
,
Matsumoto
T
, et al
Loss of SMAD4 from colorectal cancer cells promotes CCL15 expression to recruit CCR1+ myeloid cells and facilitate liver metastasis
.
Gastroenterology
2013
;
145
:
1064
75
.
24.
Inamoto
S
,
Itatani
Y
,
Yamamoto
T
,
Minamiguchi
S
,
Hirai
H
,
Iwamoto
M
, et al
Loss of SMAD4 promotes colorectal cancer progression by accumulation of myeloid-derived suppressor cells through the CCL15-CCR1 chemokine axis
.
Clin Cancer Res
2016
;
22
:
492
501
.
25.
Amat
M
,
Benjamim
CF
,
Williams
LM
,
Prats
N
,
Terricabras
E
,
Beleta
J
, et al
Pharmacological blockade of CCR1 ameliorates murine arthritis and alters cytokine networks in vivo
.
Br J Pharmacol
2006
;
149
:
666
75
.
26.
Gabrilovich
DI
,
Nagaraj
S
. 
Myeloid-derived suppressor cells as regulators of the immune system
.
Nat Rev Immunol
2009
;
9
:
162
74
.
27.
Fridlender
ZG
,
Sun
J
,
Kim
S
,
Kapoor
V
,
Cheng
G
,
Ling
L
, et al
Polarization of tumor-associated neutrophil phenotype by TGF-beta: "N1" versus "N2" TAN
.
Cancer Cell
2009
;
16
:
183
94
.
28.
Mantovani
A
,
Cassatella
MA
,
Costantini
C
,
Jaillon
S
. 
Neutrophils in the activation and regulation of innate and adaptive immunity
.
Nat Rev Immunol
2011
;
11
:
519
31
.
29.
Kusmartsev
S
,
Nagaraj
S
,
Gabrilovich
DI
. 
Tumor-associated CD8+ T cell tolerance induced by bone marrow-derived immature myeloid cells
.
J Immunol
2005
;
175
:
4583
92
.
30.
Guthrie
GJ
,
Charles
KA
,
Roxburgh
CS
,
Horgan
PG
,
McMillan
DC
,
Clarke
SJ
. 
The systemic inflammation-based neutrophil-lymphocyte ratio: Experience in patients with cancer
.
Crit Rev Oncol Hematol
2013
;
88
:
218
30
.
31.
Pardigol
A
,
Forssmann
U
,
Zucht
HD
,
Loetscher
P
,
Schulz-Knappe
P
,
Baggiolini
M
, et al
HCC-2, a human chemokine: Gene structure, expression pattern, and biological activity
.
Proc Natl Acad Sci U S A
1998
;
95
:
6308
13
.
32.
Berahovich
RD
,
Miao
Z
,
Wang
Y
,
Premack
B
,
Howard
MC
,
Schall
TJ
. 
Proteolytic activation of alternative CCR1 ligands in inflammation
.
J Immunol
2005
;
174
:
7341
51
.
33.
Joubert
P
,
Lajoie-Kadoch
S
,
Wellemans
V
,
Létuvé
S
,
Tulic
MK
,
Halayko
AJ
, et al
Expression and regulation of CCL15 by human airway smooth muscle cells
.
Clin Exp Allergy
2012
;
42
:
85
94
.
34.
Park
KH
,
Lee
TH
,
Kim
CW
,
Kim
J
. 
Enhancement of CCL15 expression and monocyte adhesion to endothelial cells (ECs) after hypoxia/reoxygenation and induction of ICAM-1 expression by CCL15 via the JAK2/STAT3 pathway in ECs
.
J Immunol
2013
;
190
:
6550
8
.
35.
Forssmann
U
,
Magert
HJ
,
Adermann
K
,
Escher
SE
,
Forssmann
WG
. 
Hemofiltrate CC chemokines with unique biochemical proterties: HCC-1/CCL14a and HCC2/CCL15
.
J Leukoc Biol
2001
;
70
;
357
66
.
36.
Reckamp
KL
,
Gardner
BK
,
Figlin
RA
,
Elashoff
D
,
Krysan
K
,
Dohadwala
M
, et al
Tumor response to combination celecoxib and erlotinib therapy in non-small cell lung cancer is associated with a low baseline matrix metalloproteinase-9 and a decline in serum-soluble E-cadherin
.
J Thorac Oncol
2008
;
3
:
117
24
.
37.
Li
Y
,
Wu
J
,
Zhang
W
,
Zhang
N
,
Guo
H
. 
Identification of serum CCL15 in hepatocellular carcinoma
.
Br J Cancer
2013
;
108
:
99
106
.
38.
Bodelon
C
,
Polley
MY
,
Kemp
TJ
,
Pesatori
AC
,
McShane
LM
,
Caporaso
NE
, et al
Circulating levels of immune and inflammatory markers and long versus short survival in early-stage lung cancer
.
Ann Oncol
2013
;
24
:
2073
9
.
39.
Li
Y
,
Yu
HP
,
Zhang
P
. 
CCL15 overexpression predicts poor prognosis for hepatocellular carcinoma
.
Hepatol Int
2016;
10:488–92.
40.
Rodero
MP
,
Auvynet
C
,
Poupel
L
,
Combadière
B
,
Combadière
C
. 
Control of both myeloid cell infiltration and angiogenesis by CCR1 promotes liver cancer metastasis development in mice
.
Neoplasia
2013
;
15
:
641
8
.
41.
Kitamura
T
,
Qian
BZ
,
Soong
D
,
Cassetta
L
,
Noy
R
,
Sugano
G
, et al
CCL2-induced chemokine cascade promotes breast cancer metastasis by enhancing retention of metastasis-associated macrophages
.
J Exp Med
2015
;
212
:
1043
59
.
42.
Gladue
RP
,
Brown
MF
,
Zwillich
SH
. 
CCR1 antagonists: What have we learned from clinical trials
.
Curr Top Med Chem
2010
;
10
:
1268
77
.
43.
Szekanecz
Z
,
Koch
AE
. 
Successes and failures of chemokine-pathway targeting in rheumatoid arthritis
.
Nat Rev Rheumatol
2016
;
12
:
5
13
.

Supplementary data