GEP-NEN is the most common type of neuroendocrine tumor. Some researches have found that the prognosis is associated with factors like tumor classification, stage, immunohistochemistry[19, 20]. An accurate and convenient prognostic indicator is needed to assist in clinical decision-making in cancer management. We conducted a meta-analysis by consolidating the published literature to prove the relationship between NLR and the prognosis in patients with GEP-NEN. In the current study, we incorporated 13 studies with 1598 patients to assess the clinical significance of NLR in GEP-NEN. Our research indicated that a high pre-treatment NLR was associated with a poor unfavorable OS and RFS in patients with GEP-NEN.
Several studies have implied that an elevated NLR is associated with the poor survival in several types of cancer, such as esophageal cancer[21], breast cancer[22], colorectal cancer[23], prostate cancer[24] and gynecologic cancers [25]et al. The results of our pooled analysis focus on GEP-NEN agree with results from these abovementioned studies on other cancers.
The relationship between chronic inflammation and cancer has been gradually known in recent years. However, the mechanisms behind the relationship between a high NLR and a worse prognosis in cancer have not been demonstrated. Neutrophils are the most abundant white blood cells in circulation and are the first responders to sites of infection and tissue damage. Tumor-associated neutrophils (TANs) predict poor overall survival in many types of cancer[26]. Persistent inflammation promotes tumor growth, and the chemokines and cytokines, including CXCL8, CXCL5 and CXCL6 generated by tumor cells, and the surrounding microenvironment are involved in neutrophil recruitment[27]. In a broad sense, lymphocytes have anti-tumor activity, so the reduction of lymphocytes is conducive to the maintenance of tumor microenvironment and the growth of tumor cells. These are the possible mechanisms by which high NLR can predict tumor Invasive growth.
We can see that most of the included researches are from China, which may cause selection bias; therefore, a subgroup analysis of the studies based on ethnicity was performed to explore whether race had any effect on the outcome. We considered that a higher NLR is associated with reduced survival time in both Asian and Caucasian. However, the NLR may not be instructive for RFS in Caucasian. In the future, relevant trials are needed to provide further evidence for the prognostic significance of NLR on race. In the sensitive analysis of NLR and OS, we considered that the Fan's study might be the source of high heterogeneity. After the revaluation of its design, sample size, outcome indicators, evaluation criteria, we haven't found apparent defeats. However, it comes from a low-impact journal that can no longer be retrieved by PubMed and is less reliable. As for RFS, after revaluation of Grenader's study that might be the source of heterogeneity, we found that in this study, all patients underwent somatostatin therapy. In contrast, other studies adopted surgical treatments. This indicates that the treatment has a specific effect on the prognosis of GEP-NEN, which may be a confounding factor in NLR prognosis. Overall, the results before sensitive analysis were not robust enough, so after the reassessment, we decided to adopt the pooled HR results after excluding the literature leading to increased heterogeneity.
In our study, there are some limitations. First, 12 of 13 incorporated researches were retrospective studies with small sizes. Second, although most included patients underwent surgical treatment, some patients underwent chemotherapy and somatostatin treatment, which also resulted in a little bit heterogeneity. Third, the study exists a publication bias, as mentioned previously. Fourthly, non-English language literature was not included, and there might be more valuable results that were not included.