In the present study, we found that CTC-WBC cluster was associated with worse progressive-free survival in patients with colorectal cancer after radical or palliative surgery. Furthermore, the multivariate analysis underscored the independent influence of CTC-WBC clusters for PFS as well. Compared to medical image scans, laparoscopy and some serological examination, CTCs have been used as an early-sensitive and convenient “liquid biopsy” approach for the surveillance of tumor in clinical analysis. From the count of single CTC, to the comprehensive acquirement of tumor functional representatives, this weapon has been continuously optimized with multiple detection strategies including a package with WBCs. Our findings hold promise for the CTC-WBC cluster becoming a new therapeutic target through disassembling of this package.
Long before the last decade, the value of CTC for tumor prognosis had been understood by oncology researchers, and a large number of studies using cytometric methods or polymerase chain reaction (PCR)-based approaches had been conducted in colorectal cancer and confirmed the prognostic value of CTC. Later, the CellSearch system (Veridex LLC) was established and became popular[21, 22]. However, this method using anti-EpCAM-coated magnetic beads to capture markers absent from leukocytes has considerable limitations in detecting EMT CTCs, which have been proven to be linked to a worse treatment response in gastric cancer patients by our previous study[13]. In addition, due to the low detection rate of CTC in patients with early staged CRC, the uncertainty of the CTC cut-off value will bewilder the interpretation of the carcinoma prediction by CTC. In a prospective multicenter study of 519 patients with stage III CRC recruited, no significant association was seen between CTC enumeration and the prognosis[23]. Nevertheless, there has been an increased interest regarding the possible use of the association between tumor cells and immune cells. In the present study, we used an improved CanPatrol™ CTC enrichment technique to unbiasedly isolate CTCs and CTC-WBC clusters. The system combined the superiority of CD45 + leukocytes depletion and isolation of CTCs by size, which is more efficient and duplicable for CTC gathering and enumeration[24].
The role of tumor-associated neutrophils (TANs) as a driver of tumor progression have been drilled down deeply by previous studies. Specific examples of transforming growth factor (TGF)-β shaping the pro-tumorigenic (N2) or antitumorigenic (N1) effects of TAN were demonstrated[25]. However, trials exploring the correlation between CTC-WBC clusters and the prognosis in cancer patients have been blank until the inspiration by Szczerba et al[15]. The research revealed the mechanism of interactions between CTCs and WBC-clusters in breast cancer mouse models. They found that the proportion of neutrophils in CTC-WBC clusters ranged from 85.5–91.7% whereas only a minority (8.3–14.5%) were monocytes, which was consistent with studies reporting an association between a higher neutrophil-to-lymphocyte ratio (NLR) and poor clinical outcomes in various cancers[26]. Szczerba et al. also showed that CTCs from CTC–WBC clusters were predisposed to be exuberant in positive regulators of cell cycle and deoxyribonucleic acid (DNA) replication programs compared to CTCs alone and could express genes that encode granulocyte colony-stimulating factor (G-CSF), which facilitated pro-tumor activities like neutrophil recruitment, Bv8 expression and angiogenesis[27–29]. On the contrary, genes involved in EMT were not related with the interaction between CTCs and neutrophils. Accordingly, in our study, although only 33 (10%) patients expressed epithelial solely CTC, no significant association was observed between CTC-WBC positive and mesenchymal phenotype. However, the sample size derived from patients in Szczerba’s study were small with merely 34 heterogenous CTC-detectable patients. In the absence of confounding cause adjustment, the bonus influence of CTC-WBC clusters could be attenuated.
Although TANs were identified as the supporting factors mediating distant CRC metastasis, especially liver metastasis[30], the therapeutic benefit of CTC-WBC clusters in CRC patients has not been studied so far. Our study mainly focused on validating the prediction value of CTC-WBC clusters in CRC patients after surgery. By means of the Kaplan-Meier survival curves, a significant shorter progression-free survival period could be seen compared with the CTC-WBC negative group. The univariate Cox risk proportional regression analysis proposed that the surgery type, tumor stage, preoperative CEA and CTC-WBC/CTC group were all independent predicting factors correlated with poor PFS and the multivariate analysis further suggested that CTC-WBC clusters was an independent predictor for the PFS of post-operative CRC patients after adjustments with all of these factors.
The study has several limitations. The significant association with CTC-WBC clusters was neither exhibited in the overall survival, nor the post-operative liver metastasis. Among the primary culprit for causing is the small number of cases from the single-center. Also, a longer follow-up and a consistent CTC monitor after the surgery are needed. The first two factors induce insufficient end-point events, while the latter indicates the lack of pre- and post-operative comparisons and subsequent long-term surveillance. In addition, the post-operative chemotherapy information of patients was not included due to the discharge of patients, which prevented us from adjusting the influence of adjuvant chemotherapy on prognosis or evaluating the therapy efficacy via CTC-WBC clusters. The failure in predicting liver metastasis may be explained by these reasons. Finally, the study was limited by the inherent flaws of retrospective studies and wanted in an extensive exploration of molecular mechanisms as we mentioned above. More work will need to be done to establish multi-center prospective studies with larger-scale samples and a longer follow-up duration.