The Preoperative Lymphocyte to Monocyte Ratio Predicts Clinical Outcome In Resected Patients With T2-4N0-3M0 Siewert Type II/III Adenocarcinoma Of The Esophagogastric Junction

Background Inammation has a critical role in the pathogenesis and progression of cancer. The lymphocyte to monocyte ratio (LMR) could be a new biomarker in various tumors. Aims We analyzed the LMR with clinicopathological parameters and outcome in resected patients with T 2-4 N 0-3 M 0 Siewert type II/III of advanced adenocarcinoma of the esophagogastric junction (AEG) . Methods A total of ve hundred and seventy-one patients with Siewert type II/III of AEG between Jan 2006 and Jun 2012 were included in this retrospective study. Kaplan-Meier curves were used to calculate the cancer-specic survival (CSS). Univariate and multivariate Cox-regression analyses were performed to evaluate the prognostic factors. Results We set 3.64 as the cut-off level based on the receiver operating characteristic curve. The preoperative absolute lymphocyte count tended to decrease in ‘LMR ≤ 3.64’ group, and the preoperative absolute monocyte count tended to decrease in ‘LMR (cid:0) 3.64’ group. The decreased preoperative LMR was signicantly associated with decreased CSS in univariate (HR:2.311, 95%CI:1.639-3.254, P=0.008) and multivariate analysis (HR:1.642, 95%CI:1.242-2.171, P=0.027 ). According to the treatment regimen(surgery alone versus surgery and adjuvant chemotherapy), no signicant difference in the 5-year CSS was identied in ‘high-risk’ group (LMR ≤ 3.64) (HR: 1.121, 95%CI: 0.733-1.725, P=0.605). Conclusions The LMR might be an independent prognostic marker for CSS in resected patients with T 2-4 N 0-3 M 0 Siewert type II/III of advanced AEG. When the patients with LMR ≤ 3.64 were analyzed, no benet chemotherapy could be found.


Introduction
In the recent decades the incidence of adenocarcinoma of the esophagogastric junction (AEG) has been increased in the worldwide [1] . Especially in China, due to the lack of routine screening, AEG is often diagnosed at the advanced staged(T2+) with or without lymph nodes metastasis and hematogenous dissemination [2] , so the mortality caused by AEG is higher than the other gastrointestinal tumors. According to the distance between the tumor center and the anatomic esophagogastric junction(EGJ), AEG was divided into three subgroups by Siewert's classi cation [3] . Type I (adenocarcinoma of the distal esophagus with the tumor center located within 1 cm to 5 cm above the EGJ) is the most prevalent type in Western countries. It is often treated as distal esophageal cancer [4] . Type II (true adenocarcinoma of the cardiac with the tumor center located within 1 cm above and 2 cm below the EGJ) and the type III (subcardial adenocarcinoma that in ltrates the esophagogastric junction with the tumor center located within 2 cm and 5 cm below the EGJ) are more common than type and are mostly treated as proximal gastric cancer in China. Because of the different characteristics, the surgical technique and treatment strategies are different in three subgroups. Although the great development of radical surgery and adjuvant chemotherapy, long time survival rates for non-metastatic AEG range from 18 to 50 percent after curative surgery [5] . In some trials, chemotherapy based on 5-uorouracil with oxaliplatin or docetaxel may improve patient outcomes [6] . However, there are still a large number of advanced AEG patients after receiving radical surgery can not bene t from 5-uorouracil-based chemotherapy [7] . The high mortality is primarily related to complications of tumor dissemination. As we all know, lots of studies evaluated the association of various biomarkers with clinical outcome in gastrointestinal tumor. Owing of high costs, lack of standardization, and limited availability, the biomarkers still can not come into routine clinical practice [8.9] . Some evidence revealed that tumor-associated in ammation could play an important role in cancer progression and prognosis. Previous studies have suggested that the preoperative peripheral blood cells can re ect the systemic in ammation, such as lymphocyte to monocyte ratio(LMR). Some studies show that survival bene t is associated with an increased pretreatment LMR [10,11] . The LMR might be a good re ection of host's immune response and tumor burden. However, the prognostic value of the LMR in advanced AEG has not been reported. The aim of this study was to investigate the clinical signi cance and the prognostic value of the preoperative LMR for resected patients with T 2 − 4 N 0 − 3 M 0 Siewert type II/III of advanced AEG.

Patients and specimens
We conducted this study on consecutive patients with Siewert type and of AEG. A total of 708 patients with histologically con rmed AEG were included in this study. All patients were treated at the rst ward of gastrointestinal surgery in The First A liated Hospital of Anhui Medical University from Jan 2006 to Jun 2012. In our hospital, type of AEG were treated at thoracic surgery as the distal esophageal cancer, type and were treated at gastrointestinal surgery. All the patients were in relatively ne overall conditions without severe dysfunction of important organs or systematic un t like dyscrasia, rheumatic disease or hypohepatia. Patients undergoing multivisceral resection or having other gastroenteric diseases were excluded from our study. To avoid possible effects of such treatments on preoperative laboratory pro les those who had received neoadjuvant chemotherapy or received a blood product transfusion within one month before resection were also excluded. They also had not received any radiotherapy or interventional therapy before. Besides, patients were con rmed without severe mental disorders.
All enrolled patients underwent radical surgery. For type II adenocarcinomas invading the distal of esophagus, transhiatal total gastrectomy or proximal gastrectomy combined with mediastinal lymphadenectomy was preferred. Thoracoabdominal incision might be performed for subtotal esophagectomy to guarantee curability, if the frozen section of proximal esophageal cutting edge was positive. For type III AEG, transabdominal TG was performed in our department of gastrointestinal surgery. D2 lymphadenectomy was routinely performed. The inferior mediastinal or extended lymph node dissection was performed for patients with esophageal involvement. Intraoperative frozen section was a routine procedure aiming to secure the tumor cells free from the resection margins. All operations were conducted by the same group of surgeons who routinely operated on more than 50 cases per year and who had surgical practice of 5 or more consecutive years.

Evaluation And Staging
Clinical stage was assessed according to computed tomography(CT) and ultrasound endoscope examinations. In our department, preoperational abdominal CT and ultrasound endoscope are examined routinely. After surgery, the tumor were staged according to the tumor-node-metastasis (TNM) criteria from the 7th edition of the International Union Against Cancer's classi cation.

Laboratory Date
As part of pretreatment evaluation, all patients' peripheral blood samples were collected into tubes containing dipotassium ethylenedinitrotetra-acetic acid (EDTA) at 3 days before surgery and all measurements were performed within 30 min after blood collection. The peripheral LMR was calculated as the ratio of absolute counts between peripheral lymphocytes and monocytes. Analysis of the white blood cell count was performed in the routine laboratory of our hospital (Sysmex XE-2100 Hematology Analyzer).

Follow-up
Enrolled patients were prospectively followed-up until Jun 2017. Follow-up was performed in regular intervals (every 3 months for the rst 2 years after treatment, every 6 months intervals in 3-5 years and every 12 months intervals after 5 years). Follow-up evaluations included clinical check-up, laboratory including complete blood count (CBC), liver function test, and radiological assessment (liver scan and chest X-ray every 6 months). All of the clinical features were retrospectively obtained from the patients' history. Follow-up data were available.

Statistical analysis
As the current study described the prognosis of patients with AEG II/III, therefore, a cancer-speci c survival (CSS) analysis was ascertained. The CSS was de ned as the time from surgery to cancer-related death. The optimal cut off levels for the LMR were by applying receiver operating curve (ROC) analysis.
The association between the clinic pathological features and the LMR with CSS was analyzed by using Kaplan-Meier curves and compared by the log-rank test. The clinic pathological features was correlated with the LMR by X 2 -test. The multivariate analysis was performed using the Cox proportional hazards regression. Hazard ratios (HRs) estimated from the Cox-regression analysis were reported as relative risks with corresponding 95% con dence intervals. All statistical analyses were performed using the Statistical Package for Social Sciences version 19.0(SPSS Inc, Chicago, IL, USA). A two-sides P value of < 0.05 was considered statistically signi cant.

Results
In our study, 708 patients were enrolled. 4 patients were excluded because they died of postoperative complications. 71 patients were belong to pT1 stage. 62 patients were lost to follow-up. At last, 571 patients were included in this study. 113(19.8%) were women and 458(80.2%) were men, with an age range from 28 to 89 years. The median age at time of diagnosis was 61.9 years. The median follow-up time was 63.5 months (range 1 to 115). The basic clinic pathological characteristics of patients are presented in Table 1. Abbreviation: LMR = lymphocyte to monocyte ratio.
Applying ROC curve analysis, the optimal cut-off levels for the LMR was 3.64 for CSS. The ROC curve was shown in Fig. 2. The area under the curve was 64.3% for CSS (P<0.001). Based on the optimum cut-off values of LMR, patients were divided into two groups for further analysis (Patients with a LMR ≤ 3.64 is the high-risk group and with a LMR>3.64 is the low-risk group). In our study cohort, clinic pathologic characters were compared between the 'high-risk' and 'low-risk' groups for LMR, we found signi cant associations between gender(P = 0.021), age(P = 0.004), tumor length(P = 0.005), pT stage(P = 0.008), pN stage(P = 0.039), pTNM stage(P = 0.022) and adjuvant chemotherapy(P < 0.001). (Table 1) The preoperative absolute neutrophil/lymphocyte/monocyte counts were shown in  Table 3, clinic pathological characters for prediction of CSS were further investigated by univariate analysis with Cox regression model. In univariate analysis, tumor length (P < 0.001), pT stage (P < 0.001; Figure 3A), pN stage (P < 0.001; Fig. 3B), vessel invasion (P = 0.027), pTNM stage (P < 0.001; Fig. 3C) and LMR (P = 0.008; Fig. 3D) were signi cantly associated with 5-year CSS (  Fig. 4B). Abbreviation: LMR = lymphocyte to monocyte ratio.

Discussion
Recently, systemic in ammation has been found to correlate with tumor progression. Leucocytes in the tumor microenvironment promote tumor growth, angiogenesis and metastasis [12] Many peripheral in ammatory markers including PLR, NLR and LMR and their prediction of clinical outcomes in solid tumor entities have been uncovered [13,14] In this study, we examined a large cohort of patients with advanced type / AEG and investigated the prognostic signi cance of preoperative lymphocyte to monocyte ratio as a biomarker for predicting the outcomes after radical surgery.
As we all know, lymphocytes play an essential role in systemic in ammatory response to tumorous disease, including the inhibition of tumor cell proliferation and migration [15] . lymphocytes are the important components of the adaptive and innate immune system and the cellular basis of immunosurveillance. A decreased lymphocyte count is assumed to re ect an insu cient immunologic reaction to the tumor, thus promoting tumor cell apoptosis and metastasis [16] .Some studies have shown a prognostic impact of tumor-in trating lymphocytes (TILs) in a series of cancers including gastric cancer [17,18] . In the study of gastric adenocarcinoma, dense in tration of CD3 + and CD8 + TILs had been associated with an optimistic prognosis [17] . In esophageal adenocarcinoma, the lower level of the FOXP3 + regulatory T cell (Treg) in trate present in the residual tumor or scar correlated with the more pathological response [19] . In resent study, Higher levels of TILs in the pathological specimen were associated with signifcant pathological response to neoadjuvant chemotherapy (NAC), increased levels of CD4 + and CD8 + TILs were associated with signifcant local tumour regression and lymph node downstaging [20] .
The role of monocytes in tumor invasion, proliferation and metastasis is important. Monocytes can differentiate into tumor-associated macrophages(TAMs) in the tumor microenvironment. TAMs are recruited at the tumor site, where they accelerate tumor progression through the angiogenesis and antiimmune responses [21] . More and more evidence show that tumor-associated macrophages(TAMs) suppress the immune system of the host and promote tumor angiogenesis, proliferation and migration [22] . Macrophages are capable of producing osteonectin, which is essential in forming metastasis. This evidence suggests a tumorous potential of monocytes due to formation of different macrophage phenotypes that promote the malignant process. The high absolute monocytes count in the peripheral blood is reported to stand for the formation of TAMs and an elevated tumor burden in patients [23] . Therefore, it is not surprising that peripheral blood monocytosis is an adverse prognostic factor for various tumors. The LMR might be a good re ection of responsiveness of the immune system of the host and a microenvironment marker of tumor burden. The prognostic values of LMR in patients with AEG remain uncertain.
In our study, we set 3.64 as the cut-off levels for the LMR by applying ROC curve analysis, different from some other studies, such as 4.0 in the study by Hirahara N [24] , 2.86 in pancreatic adenocarcinoma by Li's study [25] , a decreased peripheral lymphocyte to monocyte ratio (LMR) has been identifed as a poorer prognostic indicator in various cancers [26,27] .From Table 2, we found that the median of Lymphocyte count was higher in 'LMR 3.64'(1.9) than in 'LMR ≤ 3.64'(1.3). The median of monocyte count was nearly the same in two groups(0.3vs0.4). 5-year CSS with 'LMR 3.64' was 51.8% higher than 43.2% with 'LMR ≤ 3.64' (P = 0.008; Fig. 2D). Sometimes TAMs not only enhance tumor growth and progression, but also modulate the e cacy of various forms of anticancer therapy, such as chemotherapy and radiotherapy. In some circumstances, they also facilitate tumor regrowth, revascularization, and spread after the treatment [28] . In patients with esophageal cancer, in ltration with CD68 + and CD163 + TAMs, especially M2 macrophages, is associated with a poor prognosis for patients undergoing neoadjuvant chemotherapy [29] . Not only the rst-line chemotherapeuitc drug, but also epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) had a poor response in patients with advanced non-small cell lung cancer [30] . As we all know, how to predict whether patients of AEG can bene t from chemotherapy was still a problem which perplexed physicians for many years. Our results suggest that 'high-risk' patients based on LMR ≤ 3.64 do not bene t from 5-uorouracil-based adjuvant chemotherapy. For high-risk patients of AEG, we should avoid to use of adjuvant 5-uorouracil -based chemotherapy until disease progression. Moreover, the LMR provides an easy available and low price biomarker with outcome.
To the best of our knowledge this is the rst study to assess the LMR in advanced type II/III AEG. The strength of this study is the large sample size and the long follow-up period. However, there are some possible limitations associated with this study. Firstly, because of the retrospective design of the study, we cannot fully exclude the selection bias in our cohort. Furthermore, potential confounding factors such as infection, coronary syndrome, and diabete mellitus, that might affect the lymphocyte and monocyte count were not taken into consideration. Finally, although we set 3.64 as the cut-off level by using ROC curve, different cut-off levels may also valuable. So large prospective and multi-centric clinical trails should be performed to con rm our ndings in future.