Colorectal cancer is the second leading cause of cancer death in Japan. Since local recurrence and distant metastasis occur to a large number of patients even after curative surgery, a new focus has been placed on identifying biomarkers that predicting prognosis. While we have investigated tumor-related factors in many previous reports, host-related factors have started to attract attention [1].
The host immune system is an important factor which affects the outcome of cancer [14–16]. NLR has been suggested to be a simple index of systemic inflammatory response. Neutrophilia occurs during systemic inflammation, and lymphopenia is a maker for depressed cell-mediated immunity [17]. That is, cell-mediated immune responses are dependent largely on lymphocytes. A large number of lymphocytes at tumor sites has been shown be associated with a good prognosis, whereas lymphopenia has been reported to be a predictor of poor prognosis [18]. On the contrary, neutrophils suppress lymphocyte-mediated cytolysis and has been reported to be associated with a poor prognosis [19]. The prognostic NLR in malignancy may due to the high tumor angiogenesis activity of tumor-induced neutrophils contributing to tumor progression, lymphocyte count associated with disease severity, and immune escape of tumor cells from tumor infiltrating lymphocyte [20].
Walsh et al. first reported that a preoperatively elevated NLR had a relationship with overall and cancer-specific survival in colon cancer [21]. The predictive potential of NLR in evaluating the prognosis of patients with various types of malignancies, including resectable colon cancer, has been demonstrated in previous studies [8–11]. However, no study has analyzed a large cohort of patients with Stage II-III advanced colon cancer using propensity scores to minimize selection bias. The present study cohort consisted of 375 patients with Stage II-III colon cancer, and no significant differences were observed between the lower NLR and higher NLR groups in terms of OS and RFS before case matching. After case matching according to propensity scores, however, NLR was found to predict prognosis both in terms of OS and RFS with a cut-off value of 3.0.
The colon originates from the midgut and hindgut during embryonic development and differentiates into the right-side colon and left-side colon. Tumors in the cecum, ascending colon, and the proximal part of the transverse colon are defined as midgut tumors, and those in the distal transverse, descending, and sigmoid colon and the rectum are defined as hindgut tumors. Differences have been reported between right and left-sided colon tumors in terms of clinical symptoms, incidence, molecular pathways involved, and oncologic outcomes, as well as embryologic origin [22–26]. However, no reports have examined the predictive potential of NLR in colon cancer with a focus on tumor sidedness. The present study evaluated the prognostic value of NLR by tumor sidedness, and found that both 5-year OS and 5-year RFS rates were significantly lower in patients with left-sided colon cancer who had a higher NLR. In contrast, no significant differences were observed among patients with right-sided colon cancer.
National Comprehensive Cancer Network (NCCN) guidelines recommend adjuvant chemotherapy for high-risk Stage II and Stage III colon cancer [27]. High-risk factors for recurrence include poorly differentiated histology, lymphatic/vascular invasion, bowel obstruction, < 12 lymph nodes examined, perineural invasion, localized perforation, and close, indeterminate, or positive margins. The guidelines, however, also state that there are no data correlating risk features with selection of chemotherapy. In the present study, preoperative NLR was associated with both RFS and OS, suggesting that NLR could be used as a tool to identify patients for whom adjuvant chemotherapy should be performed/avoided, especially for left-sided colon cancer. A prospective study will be needed to further explore this possibility.
This study has some limitations. First, we used a single-center retrospective design. However, to minimize selection bias, we analyzed the data according to propensity scores. Second, our data did not include records of whether patients had hematologic or autoimmune disease, which may have influenced preoperative NLR values. Third, no molecular assessment was performed in this study, e.g., to determine microsatellite instability. A prospective study will be needed to examine confounders and further clarify the prognostic value of NLR.