Nutritional status greatly affects the outcomes of cancer patients, and thus several nutritional markers have been reported as prognostic factors for various malignancies. The GNRI was reported to be a significantly useful and simple nutritional-related prognostic index that consisted of serum albumin level and patient height and body weight to predict survival outcomes in patients with gastrointestinal tract cancers, including esophageal cancer, gastric cancer, and colorectal cancer [3, 5] [8] [9]. Furthermore, pancreatic cancer patients with a low GNRI had significantly poorer prognostic outcomes than patients with high GNRI [2] [10]. Serum albumin, a protein used in clinical tests to assess nutritional condition, has been used to evaluate the morbidity and mortality of patients indicated for surgical treatment or as a prognostic predictive factor of various diseases including malignancies [11, 12] [13]. However, the American Society for Parenteral and Enteral Nutrition recently reported that serum albumin is an inflammatory marker associated with nutritional risk during nutritional assessments, and therefore, should not be used as a nutritional marker [14]. As an alternative, the GNRI is an inflammatory index, which was reported originally as a nutritional assessment tool. CRP is an acute protein involved in systemic inflammation, which can be also used as an indicator to predict survival in patients with cancer [15, 16]. A negative correlation was reported between the production of CRP and albumin, although both proteins are synthesized by hepatocytes [17, 18]. The mGNRI developed by Kouzu et al. is a novel prognostic index that uses inverse CRP instead of albumin, which is based on the reported negative correlation [6]. Furthermore, they suggested that the mGNRI was an independent prognostic factor for patients with esophageal cancer. However, the relationship between the mGNRI and prognostic outcomes in pancreatic cancer patients remains unclear. Therefore, the present study was conducted to evaluate the prognostic usefulness of the mGNRI compared with the GNRI in patients with pancreatic cancer. We found that patients with a high mGNRI had a significantly better prognosis than those with a low mGNRI and that the mGNRI was a predictive prognostic factor for patients with resected pancreatic cancer. Furthermore, the mGNRI was indicated to be more appropriate than the original GNRI when predicting the prognosis of patients with resected pancreatic cancer.
Interleukin (IL)-6 and other proinflammatory cytokines, including tumor necrosis factor, are released by inflammatory cells in the tumor microenvironment in response to tissue necrosis and the presence of tumor cells [19]. These cytokines, especially IL-6, were also shown to promote tumorigenesis by regulating multiple signaling pathways related to apoptosis, proliferation, angiogenesis, invasiveness, and metastasis [20]. Moreover, the synthesis of CRP is promoted in hepatocytes via activation by IL-6. Subsequently, CRP returns to the tumor microenvironment and promotes the autocrine growth of malignant tumors as an opsonin [19]. Therefore, serum CRP levels in patients with cancer might represent the biological behavior of cancer tissues because of the close relationship between IL-6 and CRP. In contrast, the reduced synthesis of albumin in the liver of patients with cancer is caused by the combined effects of inflammation and non-inflammatory factors such as inadequate protein and caloric intake; therefore, albumin may not reflect cancer progression as accurately as CRP in patients with cancer. Indeed, in renal cell carcinoma, CRP had a higher area under the curve for predicting OS compared with albumin [21]. These findings support the results observed in the current study.
There were several limitations in this study. It was a retrospective study with a small population of patients of East Asian ethnicity, and therefore, the findings might have led to bias and limited their generalizability. The cutoff value of the mGNRI in this study was set at 85.3 by ROC analysis. However, the optimal cutoff value in patients with pancreatic cancer remains unclear. A large prospective study involving individuals with various ethnicities is necessary to confirm our findings.