Combination of Preoperative Circulating Tumor Cell Count and Neutrophil-Lymphocyte Ratio for Prognostic Prediction in Hepatocellular Carcinoma Patients after Curative Hepatectomy

Background Both the preoperative neutrophil-lymphocyte ratio (NLR) and circulating tumor cell count (CTC) are associated with poor prognosis in hepatocellular carcinoma (HCC). The purpose of this study was to explore the prognostic value of these two indices (CTC-NLR) in HCC. Methods We retrospectively collected demographic and clinical data, including NLR and CTC, from 97 patients with HCC who underwent curative hepatectomy at our institution from March 2014 to May 2017. X-Tile software was used to confirm the optimal cut-off value of NLR and CTC for predicting overall survival (OS) in this study. OS were also analyzed using Kaplan-Meier and Cox regression methods. Based on preoperative CTC and NLR, patients were divided into three groups: CTC-NLR (0), CTC-NLR (1), and CTC-NLR (2). Relationships of CTC-NLR with clinicopathological factors and survival were evaluated. Results Preoperatively, CTC positively correlated with NLR. Patients with NLR and CTC higher than the cut-offs had shorter OS than patients with low NLR and CTC. Kaplan-Meier analysis, and log-rank tests revealed significantly lower OS among patients with CTC-NLR scores of 0, 1, and 2. Uni- and multivariate analyses showed that CTC-NLR (hazard ratio 2.050, P = 0.005), CTC (hazard ratio 2.285, P = 0.032), and NLR (hazard ratio 1.902, P = 0.048) were independent predictor of OS. A time-dependent ROC curve indicated that the prognostic efficacy of the CTC-NLR at 1 year (0.714) was better than that of NLR (0.687) and CTC (0.590); the prognostic efficacy of the CTC-NLR at 2 years (0.746) was better than that of NLR (0.711) and CTC (0.601); the prognostic efficacy of the CTC-NLR at 3 years (0.742) was better than that of NLR (0.694) and CTC (0.629). Conclusions HCC patients with higher NLR and CTC tend to show shorter OS. Preoperative CTC-NLR may be associated with poor survival and might be a reliable prognostic predictor in HCC after curative hepatectomy.


Introduction
Hepatocellular carcinoma (HCC) is one of the most prevalent malignancies in China, and it is associated with a high mortality rate [1]. Although HCC diagnosis and treatment have improved substantially, long-term survival remains unfavorable because of high rates of recurrence and mortality. Conventional prognostic markers for HCC include alpha-fetal protein (AFP), tumor staging, and Barcelona Clinic Liver Cancer (BCLC) staging, but their performance is inconsistent [2]. Therefore, a sensitive and easy-tomeasure indicator is urgently required to help predict prognosis in HCC, which may improve patient management and therefore survival.
Circulating tumor cells (CTC) that shed from the primary tumor mass circulate through the bloodstream and travel to different organs, in what is considered a precursor step to metastasis [3][4][5]. Detecting such cells can serve as a "liquid biopsy" that provides insight into metastasis [4]. Indeed, these cells show potential as a biomarker of HCC progression [6][7][8]. Similarly, an elevated neutrophillymphocyte ratio (NLR) can independently predict overall survival (OS) in patients with HCC after curative hepatectomy [9,10].
In peripheral blood, CTC interact with inflammatory cells to induce systemic inflammation and recruit neutrophils to premetastatic organs [11], in what is considered a precursor step to distant metastasis. For example, circulating tumor cells can interact with non-malignant cells such as white blood cells to promote metastasis [12] and worsen prognosis [13]. This raises the possibility of predicting prognosis based on counts of CTC. In fact, taking NLR into account in addition to CTC can improve the risk stratification and optimal management of high-risk prostate cancer patients resistant to metastatic castration [14]. Combining NLR with CTC can also improve prognostic prediction in patients with gastrointestinal cancer [15].
This led us to ask whether combining NLR and CTC can improve prognostic stratification of HCC patients. To this end, we reviewed prospective data from HCC patients at our institution and analyzed the association of preoperative CTC and NLR with survival. (3) no prior anticancer treatment, such as transarterial chemoembolization or radiation; and (4) R0 resection, defined as complete macroscopic removal of the tumor, negative resection margins, and no detectable intra-or extrahepatic metastatic lesions. Relevant clinical and demographic data were obtained for each patient from medical records.

Materials and Methods
The study was conducted in accordance with the Declaration of Helsinki guidelines, and the study protocol was approved by the Ethics Committee of the Affiliated Cancer Hospital of Guangxi Medical University. On admission, all patients provided written consent for their anonymized   respectively, based on the recommendations of the measuring kit our institute adopted. NLR was calculated as the ratio of the total peripheral lymphocyte count to the total peripheral neutrophil count.

Isolation of Circulating Tumor
Cells. The CanPatrol™ system was used to isolate CTCs as previously described [16,17]. Blood samples were collected one or two days before surgery [16,18,19]. Peripheral blood samples (5 mL, anticoagulated with ethylenediaminetetraacetic acid) 3 BioMed Research International were collected after discarding the first 2 mL to avoid contamination with skin cells. Red blood cell lysis buffer (Sur-Exam, Guangzhou, China) was used to remove erythrocytes, and the cells were resuspended for 5 min in phosphate-buffered saline (PBS) with 4% formaldehyde (Sigma, St. Louis, MO, USA). Next, the blood was filtered using a system including a filtration tube with an 8 μm pore membrane (Sur Exam), a manifold vacuum plate with valve settings (Sur Exam), an E-Z96 vacuum manifold (Omega, Norcross, GA, USA), and a vacuum pump (Auto Science, Tianjin, China). The pumping pressure was 0.08 MPa [20].

Surveillance and
Follow-Up. The 97 patients were followed up every 1-2 months for the first year and every three months thereafter, with a final follow-up date of June 30, 2021. Postoperative follow-up consisted of one or more of the following tests: serum AFP measurement, ultrasonography, dynamic computed tomography, and magnetic resonance imaging. Recurrence was classified as intrahepatic recurrence or extrahepatic metastasis and was based on comprehensive evidence from serum AFP assays, imaging (computed tomography, magnetic resonance imaging, digital subtraction angiography, positron emission tomography-computed tomography), and other examinations. OS was measured by comparing the data of surgery to the data of death.

Statistical Analysis.
Statistical analyses were performed with SPSS 25.0 (IBM, Chicago, IL, USA), R version 4.1.2 (http://www.r-project.org/) and GraphPad Prism 8.0 (GraphPad, San Diego, CA, USA). OS was defined as the time from the date of surgery to the date of death for any cause or the last recorded follow-up visit. Patient characteristics were analyzed using descriptive statistics. Intergroup differences were assessed for significance using the chisquared test. Correlation between NLR and CTC was assessed using Spearman's correlation coefficient. Kaplan-Meier survival curves were calculated and compared using the log-rank test. The optimal cut-off points of CTC and NLR for the OS were calculated using the X-Tile statistical package (version 3.6.1, Yale University, New Haven, CT, USA). X-tile software package was used to generate the optimum cutoff point for continuous variables according to the highest x 2 value defined by the Kaplan-Meier survival analysis and the log-rank test [21]. X-tile plot shows the presence of significant HCC subpopulations, and a two-dimensional projection of each possible subpopulation was used to show the robustness of the relationship between an outcome and a biomarker. Univariate and multivariate Cox proportional hazard analyses were applied to explore associations between patient characteristics and OS. Differences associated with P < 0:05 were considered statistically significant.  Table 1. Preoperatively, NLR showed a weak positive correlation with CTC (r = 0:264, P = 0:008; Figure 1).

Results
3.2. CTC and NLR. X-Tile was used for traversing expression of the biomarkers and serological indicator value as the cutoff point for dividing the patients and estimating the magnitude CTC and NLR benefits against control in the high-or low-level groups, which were 20 and 2.15, respectively. According to these cut-off values, the high cut-off value was calculated as 1, and the low cut-off value was calculated as 0, CTC-NLR score was calculated as follows: For OS, patients with CTC ≥ 20 and NLR ≥ 2:15 were assigned a score of 2; patients with only CTC ≥ 20 or NLR ≥ 2:15, a score of 1; and patients with CTC < 20 and NLR < 2:15, a score of 0.  Kaplan-Meier survival curves showed that patients with CTC < 20 had longer OS than patients with higher CTC (Figure 2(a)), similar to those with NLR < 2:15 compared with those with high NLR (Figure 2(b)).

BioMed Research International
Our data showed that preoperative CTC-NLR score correlated with AFP, tumor size, tumor number, BCLC stage, and PVTT.
Kaplan-Meier curves demonstrated that patients with CTC-NLR (0) were associated with the best OS, whereas patients in CTC-NLR (2) presented the worst OS (Figure 3).
Factors showing significance by univariate analysis were integrated into multivariate Cox proportional hazard analysis. In this study, NLR, CTC, and CTC-NLR were highly correlated. Therefore, two separate multivariate models were generated to avoid the multicollinearity among the above three variables. In the multivariate analysis, NLR, CTC, and CTC-NLR were independent prognostic factors for shorter OS (Table 5).

Comparison of the Predictive Value of Independent Prognostic Factors.
A time-dependent ROC curve was used to compare the prognostic efficacy of NLR, CTC, and CTC-NLR at 1 year, 2 years, and 3 years. Our results indicated that the prognostic efficacy of the CTC-NLR at 1 year (0.714) was better than that of NLR (0.687) and CTC (0.590) (Figure 4(a)); the prognostic efficacy of the CTC-NLR at 2 years (0.746) was better than that of NLR (0.711) and CTC (0.601) (Figure 4(b)); the prognostic efficacy of the CTC-NLR at 3 years (0.742) was better than that of NLR (0.694) and CTC (0.629) (Figure 4(c)). In addition, the C-index of the CTC-NLR was 0.677, which was higher than that of the NLR (0.642) and the CTC (0.591).

Discussion
Inflammation in the tumor microenvironment, mediated mainly by neutrophils and lymphocytes, can affect tumor development [22,23]. Neutrophils are the primary source of circulating vascular endothelial growth factor, which is associated with increased risk of recurrence in HCC [24,25]. Lymphocytes can induce tumor cell death and thereby play an important role in immune surveillance against cancer [26]. In fact, lymphocytopenia may independently predict metastasis in breast cancer [27], and severe lymphopenia may be associated with worse OS in HCC patients receiving radiotherapy [28]. Peritransplant lymphopenia can predict recurrence of HCC after liver transplantation [29]. Therefore, to some extent, the NLR reflects a balance between inflammation and antitumor immunity [30], and alteration of this balance promotes an inflammatory response and facilitates tumor progression. We found that patients with lower preoperative NLR had better OS, consistent with several reports linking NLR to prognosis of HCC patients who underwent hepatectomy [31][32][33][34][35].
CTCs are considered the source of tumor metastasis and recurrence [36,37]. Several studies have shown that CTCs are independent risk factors for HCC, and patients with higher CTC counts have a poorer prognosis [38,39]. We also found that patients with lower CTC (<20) had longer OS. CTC may be a useful biomarker of recurrence risk following liver transplantation in HCC patients [40]. These findings likely reflect the role of CTC in metastasis.
We detected a weak positive correlation between preoperative CTC and NLR. This may reflect interactions between neutrophils and circulating tumor cells [41,42]. As carcinomas develop, patients may be in a systemic inflammatory state in which neutrophils are recruited to organs where circulating tumor cells can invade [11]. We evaluated the combination index CTC-NLR; CTC-NLR score stratified patients into those showing low or high OS. Our results suggest the potential of CTC-NLR score, CTC, and NLR as independent prognostic indicators of OS in HCC. The CTC-NLR had the largest time-dependent AUC in the first to third years. This suggests that the combination CTC and NLR is more effective than CTC or NLR alone in predicting the prognosis of patients with HCC. These results suggest a new avenue for improving risk-stratified management of HCC.
With regard to the mechanisms underlying the relationship between NLR and CTC, they are multifaceted and remain unclear. Noteworthily, recent studies have also shown that the release and survival of CTCs is the result of interactions with the host's immune system, and the number of CTCs in the peripheral blood can also indirectly reflect the immune-nutritional status of patients [13,43]. Low lymphocyte counts weaken the body's immune defenses, facilitating tumor proliferation, growth, and metastasis [44]. Based on these findings, we deduced that the high NLR  [45]. In some cases, the fact that CTCs clusters are "the soil with seeds" may help further transfer them to distant organs and continue to grow [13]. This may also be the reason that the preoperative higher CTC-NLR have relationship with multiple tumors and larger tumor size. Meanwhile, the decrease of lymphocyte counts in the peripheral blood makes the CTCs less likely to be attacked by immune cells after enter-ing the bloodstream, so that more CTCs survive and are detected. This also provided a reasonable explanation for the findings of our study on the relationship between high preoperative CTC-NLR and unfavourable clinicopathological features, including BCLC stage and PVTT. Our study presents several limitations. First, it was a retrospective analysis of patients at a single center. Second, the sample was relatively small, which meant we could not divide patients into separate training and validation cohorts. Third, follow-up was relatively short, and not all patients underwent standardized follow-up procedures according to National Comprehensive Cancer Network guidelines. Therefore, our results should be verified and extended in a future work.

Conclusions
The present study provides the first evidence that a novel index based on the combination of preoperative CTC and NLR is closely related to postoperative long-term outcomes in surgically treated HCC patients. Preoperative evaluation of the CTC-NLR score may be useful for risk classification and clinical decision-making for HCC patients.

Data Availability
Data is available upon reasonable request from the authors.

Conflicts of Interest
The authors declare that there are no conflict of interests.