Background
Nasopharyngeal carcinoma (NPC) is a disease with low prevalence in large parts of the world, while it is notable for its high incidence in South China and South Asia. In China, there were an estimated 60,600 new cases of NPC and 34,100 deaths in 2015 [
1‐
3]. NPC is an endemic carcinoma, which is different from other head and neck cancers on account of its epidemiology, histopathology, and clinical characteristics. There are always no symptoms until the cancer has metastasized to other sites of the body such as the neck. More than 70% of newly diagnosed cases of NPC are classified as locoregionally advanced disease [
4]. Radiation therapy (RT) has been the primary treatment modality for NPC based on its inherent anatomic constraints and high sensitivity to irradiation. Previous studies have showed that concurrent chemoradiotherapy (CCRT) was more effective than RT alone for locoregionally advanced NPC, and became the standard treatment for those patients. Recently, induction chemotherapy (IndCT) followed CCRT significantly improved failure-free survival in locoregionally advanced NPC [
5,
6].
Despite the survival rates significantly enhanced from modern treatment techniques, however, treatment failure is still due to distant metastasis, relapse, or both, leading 20 to 30% of patients′ death. Clinicopathologic characteristics such as local invasion (T stage), regional extension (N stage) have been revealed to correlate with the occurrence of relapse and metastasis, and TNM stage is a reliable predictive factor of prognosis in NPC [
7]. Clinically, diagnosis and TNM classification for NPC mainly rely on imaging technique such as magnetic resonance imaging (MRI), in addition, plasma Epstein-Barr virus (EBV) DNA level is useful for monitoring and predicting treatment outcome of NPC. But they are often failed to detect tumor lesions in early stage and assess therapeutic response from NPC in time due to the hysteresis and instability. Therefore, there is an urgent need for a novel biomarker to early diagnosis, predicting therapy efficacy and identifying patients at high risk of treatment failure, who could be treated with an intensive treatment strategy in time to improve survival of NPC [
8,
9].
Circulating tumor cells (CTCs) are tumor cells disseminated from primary or metastatic tumors into blood circulation and act as the origin of distant metastasis [
10]. Many studies have reported that enumeration of CTCs is significant to evaluate a prognosis, predict therapy efficacy, and monitor recurrence or metastasis in various malignancies including breast, colorectal, and lung cancers, etc. Moreover, circulating tumor microemboli (CTM, a cluster of 2 or more CTCs) was also demonstrated to be contributory to tumor metastasis in various cancer types, even more malignant and aggressive than CTCs. CTCs and CTM have attracted increasing attention due to the development of detection technologies. A recent study performed by a strategy of integrated subtraction enrichment and immunostaining fluorescence in situ hybridization (SE-iFISH) had described that CTCs karyotyping upon ploidy of chromosome 8 may provide an approach for predicting chemotherapy response and monitoring chemo-resistance in advanced gastric cancer patients. Ge et al. recently even used this system to detect CTM in patients with glioma [
11‐
17]. However, the role of CTCs and CTM with respect to diagnosis and treatment in NPC are unknown.
SE-iFISH enables non-hemolytic elimination of red blood cells (RBCs) and depletion of white blood cells (WBCs) by anti-multiple WBC antibody, and characterizes CTCs integrated phenotyping and karyotyping of centromere probe 8 (CEP8) regardless of epithelial cell adhesion molecule (EpCAM) and cytokeratins (CKs) expression on the tumor cell surface [
17,
18]. In the present study, we applied SE-iFISH to detect and characterize CTCs and CTM in patients with NPC, and then we explored the relationship between CTCs number and clinical staging. In addition, the value of CTCs number and karyotyping to therapeutic efficacy in NPC was investigated. Furthermore, we explored the correlation between CTCs count and plasma EBV DNA level in NPC.
Discussion
Circulating tumor cells had been established as a risk factor of tumor progressions in various kinds of solid tumor types. It also has been reported that CTCs are promising to detect the harboring small tumors including primary and metastatic lesions in pancreatic cancer, while traditional imaging modalities fail to detect [
22]. Recent studies even had demonstrated that CTCs might be an independent prognostic factor and CTCs presence seemed related to tumor burden. Moreover, changes of CTCs number had also been found to be correlation with treatment response to carcinomas of head and neck [
23‐
27]. Si et al. reported that CTCs were tightly correlated with characteristics of NPC patients and CTCs number was decreased after therapy in NPC [
28]. Nevertheless, the value of CTCs in the fields of diagnosis and treatment in NPC was still vague.
In this study, SE-iFISH automatic testing system was applied to detect CTCs and CTM in NPC. This strategy ensured the depletion of WBCs and removed RBCs effectively, and later, combined DAPE, CEP8, EpCAM and CD45 to identify CTCs. SE-iFISH platform improved the sensitivity of CTCs detection regardless of cancer heterogeneity, down-regulation or absence of CKs and EpCAM. In addition, aneuploidy of chromosome 8 detected by CEP8-FISH had been reported in gastric cancer, bladder cancer and lung cancer [
18]. Through the previous description, the definition of CTCs in the present study were EpCAM+/CD45-/DAPI+/CEP8 ≥ 2, EpCAM -/CD45-/DAPI+/CEP8 > 2. A cluster of 2 or more CTCs defined as CTM. In this study, SE-iFISH yielded CTCs detection rate and CTM detection rate of 92.0 and 12.0% in NPC prior to treatment. Specifically, the CTCs and CTM positive rate of newly diagnosed (M0) patients were 93.1 and 10.3% respectively. While recent data presented by He et al. showed that CTCs/CTM was detected in 66.7 and 6.1% of NPC patients using ISET assay, separately, which were considerably lower than the detection rate in our study [
28,
29].
In the current study, investigation of the association with CTCs positive rate and disease staging indicated that the detection rate of CTCs in newly diagnosed (M0) and relapsed/distant metastatic patients were consistent, and the difference of CTCs positive rates in different TNM stags was not significant. Reasons accounting for the results could be that 93.1% of newly diagnosed (M0) patients in this study were stage III/IV, on the one hand, lacking of patients with stage I. On the other hand, the presence of CTCs revealed the disease status during CTCs detection, and patients with relapse or distant metastasis did not certainly present CTCs in peripheral blood.
Investigation of CTCs counts in NPC was rarely reported. Our research in this paper indicated that CTCs number was related to clinical stage of NPC patients. CTCs number in patients with relapse or distant metastasis was much more comparing with newly diagnosed (M0) patients. Furthermore, the number of CTCs gradually increased in later stage patients of both newly diagnosis (M0) and relapse or distant metastasis. However, the T-classification or N-classification was not obviously related to CTCs number. Results showed that CTCs number could indicate the severity degree of disease and tumor burden in NPC. CTCs number revealed the common results of various factors of TNM classification.
Variations in CTCs number had been demonstrated association with the therapeutic efficacy in breast cancer, while current researches about the relationship between CTCs number and therapy in NPC were deficient [
28,
30]. In this study, we observed that CTCs number was remarkably reduced in PR patients following 2 to 4 cycles of chemotherapy, on the contrary, CTCs number in PD/SD patients was slightly increased, and those patients with PD/SD were relapsed/distant metastatic patients. Obtained results demonstrated that changes of CTCs number were in accordance with responses to treatment of NPC patients, and the increase of CTCs number during treatment may indicate poor response to therapy. It is necessary to further expand the current study in order to evaluate CTCs as an indicator to be utilized to predict therapeutic efficacy and select optimal treatment protocols timely. Moreover, further studies are needed to explore the correlation of CTCs count and progression-free survival (PFS) or overall survival (OS) in NPC.
Aneuploidy of the chromosome exists in various tumor cells, and many studies showed that patients with aneuploid tumor cells had a worse outcome, that inspired us to investigate the aneuploidy of chromosome 8 in NPC [
15,
31,
32]. In this study, analysis of CTCs karyotyping revealed that diploid, triploid, tetraploid, and multiploid CTCs were observed in NPC, indicating that the heterogeneity of chromosome 8 exist in CTCs. Triploid CTCs was the most comparing with other CTCs karyotypes in newly diagnosed patients without metastasis. Otherwise, the ratio of multiploid CTCs was the largest in relapsed/distant metastatic patients, in addition, multiploidy ratio was more in patients who were later-stage at initial diagnosis of NPC. Results indicated that multiploid CTCs might have higher malignant degree.
It has been recently reported that colorectal cancer cell lines with aneuploidy are associated with intrinsic resistance and have significantly worse clinical outcome [
33]. Therefore, we investigated the correlation of CTCs karyotyping to chemotherapy efficacy of NPC patients in this study. Results demonstrated that CTCs with triploid and multiploid chromosome 8 were likely to correlate to the acquired resistance to chemotherapy based on gemcitabine plus cisplatin in relapsed/distant metastatic patients, rather than tetraploid CTCs performing initially sensitive to chemotherapy. It is also reasonable that karyotyping of CTCs could predict chemotherapy efficacy and monitor drug resistance. Similarly, Li et al. reported that triploid CTCs was related to intrinsic drug resistance, while tetraploid and multiploid CTCs were related to acquired drug resistance to paclitaxel or cisplatin-based chemotherapy in gastric cancer [
15]. Reasons accounting for the poor concordance might be that chemo-sensitivity and drug-resistance of CTCs karyotyping in different carcinomas and drugs could have multiple mechanisms, and for cancers with effective treatment such as NPC might rarely display intrinsic drug resistance. Extended studies of karyotypic characterization of CTCs on NPC patients are underway.
Plasma EBV DNA level is a significant biomarker in monitoring relapse or distant metastasis in NPC, and the plasma EBV DNA level prior to treatment is correlated with tumor burden of NPC patients. The latest research declared that CTCs number was found to have positive correlation with EBV DNA load [
29]. Currently, we compared the relationship between CTCs enumeration and plasma EBV DNA level in this study. Findings showed a positive correlation between CTCs number and plasma EBV DNA level in all NPC patients. Furthermore, the detection rate of CTCs in NPC by SE-iFISH was higher than the positive rate of plasma EBV DNA. CTCs could be used as a potential indicator for follow-up of NPC, and monitoring CTCs might be a supplementary method of discovering recurrence or metastasis in NPC.
In summary, NPC patients have a great survival rate for standard therapy, but distant metastasis and local recurrence were the main causes of treatment failure. CTCs are considered as the seeds of distant metastasis of malignant tumors. This study demonstrated blood testing of CTCs may be a prospective way in early diagnosis, treatment and follow-up of NPC. By detecting CTCs number and plasma EBV DNA level in combination, it is preferable to discover relapse and metastasis in the course of disease on NPC.