Introduction

Adjuvant chemotherapy using fluorouracil (5-FU) and folinic acid for stage III colon cancer has been widely accepted for reducing recurrence since 1990 [1]. However, whether stage II colon cancer patients should receive postoperative adjuvant chemotherapy remains controversial. Furthermore, the decrease in recurrence with adjuvant chemotherapy should be large enough to offset treatment costs, drug toxicity, and patient inconvenience. In the Quick and Simple and Reliable clinical trial [2], stage II colon cancer patients were randomized to either 5-FU-based therapy or observation groups. An approximately 3% improvement in outcome was observed when 5-FU was administered to this patient subset; however, 97% of patients who received chemotherapy received no benefit. Two other studies with a large series of stage II colon cancer patients found that only 5% and 2% of patients were benefited by the adjuvant chemotherapy [3, 4]. Hence, routine use of adjuvant chemotherapy is not recommended for stage II colon cancer [5]. Although stage II colon cancer patients have better prognosis than stage III colon cancer patients, however, the 5-year survival rates for stage II colon cancer patients varies between 60% and 80%, and up to 40% will experience disease recurrence [6, 7]. Selection of appropriate stage II colon cancer patients for adjuvant chemotherapy is critical for improving survival outcome and treatment efficacy in terms of costs and drug side effects.

National Comprehensive Cancer Network (NCCN) guidelines are widely used to help clinicians determine treatment plans. In the latest (2010) edition of the NCCN Practice Guidelines in Oncology [8], several features were identified as high risk factors of stage II colon cancer for systemic recurrence, including grade III/IV cancer, lymphatic/vascular invasion; bowel obstruction; <12 lymph nodes examined; and T4 or T3 tumors with localized perforation and close, indeterminate, or positive margins. Current recommendations state that patients with higher risk stage II colon cancer should be offered chemotherapy [9].

In addition to high risk factors of tumor and clinical characteristics, the host immune environment also plays an important role in invasion or metastases of colon cancer. The roles of the immune system and inflammatory response in disease progression have been examined before [10]. Immunocompetent white blood cell populations (i.e., lymphocytes and neutrophils) play a crucial role in the systemic inflammatory response to severe infection, injury, and shock. The neutrophil–lymphocyte ratio (NLR) has been used not only as a marker of inflammation [11, 12] but also as a prognostic index for several types of malignancies [13–16].

With the aim of finding more high risk factors for stage II colon cancer, we reviewed our surgical outcome in the context of a preoperative normal or elevated NLR. To reduce the complexity of multiple factors involved, we attempted to determine if an elevated NLR is a prognostic and predictive marker for stage II colon cancer patients who do not receive adjuvant chemotherapy only.

Materials and methods

Detailed information regarding patient and tumor-related variables were retrieved from the Colorectal Section Tumor Registry in Chang Gung Memorial Hospital between January 1995 and December 2005. Of the 7,905 consecutive patients undergoing surgery for colorectal adenocarcinoma, 1,308 patients had stage II colon cancer. All patients were recently diagnosed. A total of 258 patients who received adjuvant chemotherapy treatment were excluded from our study and an additional 10 were lost to follow-up. A total of 1,040 patients were enrolled in the study. Information about the variables of age, sex, tumor obstruction and/or perforation, NLR, postoperative adjuvant chemotherapy, carcinoembryonic antigen (CEA) levels, tumor morphology, examined lymph node numbers, tumor size (maximal diameter), histological type, histological grade, and cancer T-stage was collected from the Colorectal Section Tumor Registry at Chang Gung Memorial Hospital. We chose an NLR cutoff of 5 as used in other studies [13–16].

The tumor locations were as follows: 243 (23.4%) in the cecum and ascending colon, 212 (20.4%) in the transverse colon, 138 (13.3%) in the descending colon, and 447 (43.0%) in the sigmoid colon. All patients underwent colectomy using an open or laparoscopic method, and all procedures were performed by the same surgical team of 10 attending surgeons. Standard curative surgery included surgical resection of the primary tumor and mesocolon (330 right colectomy, 72 left colectomy, 113 subtotal colectomy, 521 anterior resection, and 56 transverse colectomy). All patients were confirmed by pathology to have had a successful R0 resection. The tumor stage and T categories were assigned based on the American Joint Committee on Cancer, seventh edition, guidelines for CRC staging [17]. T4 cancers were stratified by tumors that perforate the visceral peritoneum (T4a) versus tumors that invade or adhere to adjacent organs or structures (T4b). Leukocyte differential counts were recorded before the operation. NLR was defined as the absolute neutrophil count divided by the absolute lymphocyte count.

All patients were followed up every 3–6 months post-operation. Each follow-up visit included a physical examination and a tumor marker test. Chest radiography, abdominal ultrasonography or abdominal computer-assisted tomography scanning, and colonoscopy were performed after every operation, and every 1–3 years whenever necessary. The follow-up status was confirmed postoperatively every 12 months by a team of three physicians and five specially trained nurses. The median follow-up period for surviving patients in the regular follow-up program was 74.5 months and ranged between 45.9 and 136.8 months.

All analyses were performed using the Statistical Package for the Social Sciences, release 11.0 (SPSS Inc. Chicago, IL). The χ 2 test was used to analyze differences among groups of patients with high or normal NLR values. Univariate analysis of overall survival (OS; the proportion of cancer patients who survived for a specified time interval after surgery) and disease-free-survival (DFS; the proportion of cancer patients who survived without disease recurrence after surgery) data was performed using the Kaplan–Meier method. Cox regression analysis was then performed in a stepwise manner for multivariate analysis of the clinicopathological factors that significantly impact DFS and OS. Survival plots were constructed using the Kaplan–Meier method. Statistical significance was set at P<0.05.

Results

Table 1 shows the differences in the clinicopathological features of the normal NLR and elevated NLR groups. Of the 1,040 patients, 758 (72.9%) patients had a normal NLR while 255 (27.1%) had an elevated NLR. Patients in the elevated NLR group were more likely to be age ≥65 years (P=0.042) or have T4b cancer (P<0.001), CEA ≥5 ng/mL (P<0.001), or tumor obstruction or perforation (P<0.001). No significant differences were observed between the two groups with respect to sex (P=0.771), histological tumor type (P=0.365), tumor size (P=0.471), histological tumor grade (P=0.422), or tumor morphology (P=0.165). The examined number of lymph nodes ranged from 2 to 168 (mean 22.74, median 19).

Table 1 Characteristics of stage II colon cancer patients with elevated or normal NLR (N = 1,040)
Full size table

Patients with an elevated NLR had significantly worse OS (5-year OS, 63.5%; log-rank test, P<0.001; Fig. 1a) and worse DFS (5-year DFS, 77.6%; log-rank test, P<0.001; Fig. 1b) than patients with a normal NLR (5-year OS, 78.5%; 5-year DFS, 86.1%).

Fig. 1
figure 1

Survival curves of patients with stage II colon cancer. a Overall survival curves according to normal and elevated neutrophil–lymphocyte ratio. b Disease-free survival curves according to normal and elevated NLR

Full size image

Univariate analysis of the factors affecting OS and DFS is shown in Table 2. All factors found to be significant by univariate analysis were then subjected to multivariate regression analysis. Cox regression analysis revealed that elevated NLR is an independent predictor of OS (P=0.012; Table 3) but not of DFS (P=0.255; Table 3). Further, multivariate analysis revealed that T4b cancer, CEA ≥5 ng/mL, <12 examined lymph nodes, and tumor obstruction or perforation are independent predictors of both OS and DFS (Table 3). An elevated NLR and an age ≥65 years are the factors associated with a low OS.

Table 2 Univariate analysis of the risk factors for overall survival and disease-free survival in stage II colon cancer patients
Full size table
Table 3 Multivariate analysis of all significant factors affecting overall survival and disease-free survival in stage II colon cancer patients
Full size table

Discussion

In this study, we clearly demonstrated that elevated preoperative NLR in stage II colon cancer patients who do not receive adjuvant chemotherapy indicated poorer surgical outcome. Univariate analysis revealed that an elevated NLR is a significant prognostic factor for both OS and DFS. Furthermore, an elevated NLR is linked with T4b cancer, CEA ≥5 ng/mL, and tumor obstruction or perforation.

Walsh et al. [13] was the first to report that preoperative elevated NLR was correlated with overall and cancer-specific survival in colorectal cancer. In his study, he included patients at all stages of colon and rectal cancer. Our study showed similar results in the survival analyses; however, we have added some points that are different from those mentioned in Walsh's report. First, we worked with a large sample population with only stage II colon cancer patients who did not receive postoperative adjuvant chemotherapy. Second, we also showed that elevated NLR is significantly associated with T4 cancer, CEA level ≥5 ng/mL, and tumor obstruction or perforation that was not mentioned by Walsh et al. Although we did not fully understand if elevated NLR has a causal link with these phenotypes of colon cancer, elevated NLR may reflect the aggressive potential of tumor among these patients. These findings together may indicate elevated NLR in stage II colon cancer patients, thereby implying decreased immune status that may be associated with a high risk of recurrence. We therefore recommended that postoperative adjuvant chemotherapy should be administered although an additional trial is required to evaluate the benefits.

The inflammatory response generated by a tumor is thought to cause upregulation of cytokines and inflammatory mediators and result in an increased propensity for metastasis [18, 19]. Recently, NLR as one of inflammatory markers that gained increased interest and elevated NLR have also been repeatedly demonstrated as a significantly prognostic factor for some other types of cancer than colorectal cancer such as hepatocellular carcinoma [15], epithelial ovarian cancer [20], and malignant methothelioma [16]. Cho et al. [20] argued that measurement of preoperative NLR in combination with CA125 levels may be useful for the identification of ovarian cancer patients with adverse outcomes. Halanum et al. [15] showed that elevated NLR significantly increases the risk of tumor recurrence after liver transplantation for hepatocelluar carcinoma. Therefore, NLR may act as an option in determine which patients benefit most. These findings together may indicate that elevated NLR may be a common prognostic factor for different cancers that reflect systemic inflammation response or immune status generated by the tumor. There could be other possible reasons for explaining the association between elevated NLR and poor prognosis. Elevated NLR represents a relative lymphocytopenia, which may exhibit a poor lymphocyte-mediated immune response to malignancy because of decreasing T-4 helper/T-8 suppressor ratio. Alternatively, relatively increasing number of circulating neutrophils may increase the levels of inflammatory agents such as vascular endothelial growth factor and hence angiogenesis, thereby promoting tumor growth and even metastases [21].

Some serum biomarkers such as C-reactive protein (CRP) and CEA have been identified to determine a cost-effective treatment for postoperative treatment of colorectal cancer. Elevated CRP levels are also one of the serum markers that reflect nonspecific systemic inflammatory response and have a repeated correlation with colorectal cancer [22–24]. Wong et al. [23] showed that a significant proportion of patients with colorectal liver metastases had elevated CRP levels before the operation. However, serum CRP levels are not routinely examined as a part of the preoperative assessment of colon cancer patients. NLR can be easily determined in patients who have undergone surgery for colon cancer. Therefore, in clinical settings, the combination of these two markers to achieve a better treatment outcome may warrant further investigation. Moreover, we found significantly poor survival among patients with elevated preoperative CEA levels. Preoperative serum CEA has predictive prognostic value for colon cancer; [25] however, whether or not stage II colon cancer patients with an elevated preoperative CEA will benefit from adjuvant chemotherapy remains controversial. The CEA cutoff value will also influence its predictive value. In this study, 28.7% of patients with CEA ≥5 ng/mL (mean 4.23) had elevated NLR values and 19.3% of patients with CEA <5 ng/mL (mean 3.42) had elevated NLR values. This finding is compatible with the greater inflammatory response and potential for metastasis of patients with CEA ≥5 ng/mL.

In this study, multivariate analysis revealed that an elevated NLR is an independent predictor of OS but not of DFS. There may be attributed to: (1) patients with an elevated NLR tended to have an increased risk of death from other causes (15.8% vs. 23.9%; Table 1) and (2) elevated NLR is linked with some stronger risk factors: T4b cancer, CEA ≥5 ng/ml, and tumor obstruction or perforation. As a result, the prognostic impact of elevated NLR for DFS is surpassed. T4b cancer or tumor with obstruction or perforation induce the inflammatory response in the tumor environment and decrease the host's immune system. NLR values were higher in this group (47.1% had NLR >5; mean 6.09) which possibly explains why they are high risk factors of systemic metastasis despite lack of lymph node metastasis. However, when patients with high risk factors (T4b cancer and tumor obstruction or perforation) were excluded, elevated NLR group was still associated with a poor OS (5-year survival rate 72.0% vs. 84.5%) and DFS (5-year survival rate 83.1% vs. 89.4%) than normal NLR group.

In conclusion, our study demonstrated that elevated preoperative NLR is a useful prognostic marker in stage II colon cancer patients who do not receive adjuvant chemotherapy. Preoperative NLR measurement in stage II colon cancer patients may provide a simple method for identifying patients with poorer prognosis who may be included for further adjuvant chemotherapy trials.