Here is a summary of our main findings in this study. By comparing G-NET patients with or without lymph node metastasis, we found that older age, preoperative lower albumin level, higher CEA level, higher INR, longer TT, higher Ki67 and CgA positive rate were associated with lymph node metastasis. Logistic regression identified that ALB, CEA, tumor location and Ki67 were risk factors for LNM in patients with G-NET. There has been an increasing incidence of G-NET in recent decades.[7] Surgical resection is the first-line recommendation for G-NET.[8] However, it is largely unknown how to determine the possibility of LNM preoperatively. Here, our study has provided useful information that ALB, CEA and Ki67 as well as location of tumor are associated with the possibility of LNM in patients with G-NET.
As with other digestive NETs, our patients were divided into several groups as specified according to the WHO G grade classification system. Albeit G grade system has exerted the versatile negative prognostic factor in digestive NETs from pancreas and jejunum-ileum, its diagnosis value for determining the prognosis of patients with G-NET didn’t live up to expectation.[9] Likewise, the lack of solid evidence situated on the G grade system effective on G-NET lymph node metastasis aroused our interest. However, resultant data retard the harnessing of G grade system to predict nodal metastasis. We sought to figure out risk factors with forecasted usage value to address this issue in this scenario.
Serum ALB level is an easily accessible laboratory indicator that reflects individual nutritional status. Previous study has demonstrated that albumin is a vital source of energy and amino acids for tumor cells, and it was increasingly absorbed by tumor cells owing to fast grow and active metabolism of tumors.[10] In addition, ALB is considered as an indicator of systemic inflammatory reaction in malignant tumors. G-NET potentially affects digestive and absorptive ability and is associated with systemic inflammatory response. Both of reasons lead to suppressed synthesis of ALB, which was found more severe in G-NET patients with LNM. [11] A study involving 207 patients with malignant tumors reported that patients with lower prognostic nutrition index exhibited higher lymph node metastasis rate. Another retrospective study recruited 136 patients found that lymph node invasion was significantly correlated with ALB level.[12] Moreover, a scoring system named Glasgow prognostic score based on inflammation (CRP and ALB) has been validated as a versatile predicting progress for gastric cancer[13].
CEA is associated with various types of cancer including gastric cancer and correlated with overall survival of patients. [14, 15]A study in China found that increased CEA level were associated with LNM in remnant gastric cancer.[16] Another study discovered that gastroenteropancreatic neuroendocrine neoplasm patients with elevated CEA, CA125 or CA19-9 exhibited worse overall survival.[17] Nevertheless, there were few data about the relationship between CEA and LNM in NET. Our study found that elevated CEA could serve as a predicting factor of LNM in G-NET.
There were few studies discussing the correlation between tumor location and LNM in G-NET. Liang J et al revealed that G-NET mainly located in esophagogastric junction, most of which were aggressive malignant.[18] To our knowledge, our findings are the first investigation towards tumor location and LNM in G-NET. Our study has highlighted the tumor distribution in stomach associated with LNM manifestation in this specific cohort of patients. As for clinicians, LNM is worthy of more concern facing the G-NET patient whose tumor located in cardia and fundus of stomach and body of stomach.
The nuclear antigen Ki67 structurally associated with chromatin helps determine tumor grade and prognosis. [19] Previous studies suggested a significant correlation between Ki-67 level and clinical outcome. BOO, Y. J. et al revealed that higher Ki67 (> 60%) was associated with aggressive G-NET.[20] Another study illustrated that higher Ki67 was not only associated with higher T stage (p = 0.003) but also tended to be associated with LNM (p = 0.071).[21] In consistent with previous reports, our study revealed that higher KI67 could serve as an independent predict factor for LNM in G-NET. In fact, when it comes to neuroendocrine tumors, Ki67 is the major prognostic factor and utilized in the novel grading system.[22]
We are aware of the limitations in our study. First, this is a single-center retrospective study that may lead to potential selection bias. Second, since the short-turn outcomes between two groups was not significant and the follow-up data is not fully available, we did not compare long-term outcomes of G-NET patients in LNM and non-LNM group. Third, the rarity of G-NET and limited sample size hampers subgroup analysis such as distant metastasis compared to adjacent metastasis. In addition, molecular analysis was conducted in very few patients, which leads to failure of comparison of molecular features between LNM and non-LNM groups. Nevertheless, our study has provided a comprehensive exploration towards possible risk factors of lymph node metastasis in patients with G-NET to explore effective prediction for clinicians. (Figure.2) Future prospective studies are expected to provide more possibility in the identification of LNM in G-NET.