AUTHOR=Li Chenyu , Xie Suling , Chen Dan , Zhang Jingwen , Zhang Ning , Mu Jinchao , Gong Aixia TITLE=Clinicopathological characteristics of early gastric cancer with different level of undifferentiated component and nomogram to predict lymph node metastasis JOURNAL=Frontiers in Surgery VOLUME=10 YEAR=2023 URL=https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2023.1097927 DOI=10.3389/fsurg.2023.1097927 ISSN=2296-875X ABSTRACT=Background

Few studies showed that mixed type early gastric cancer (EGC) relates to higher risk of lymph node metastasis. We aimed to explore the clinicopathological feature of GC according to different proportions of undifferentiated components (PUC) and develop a nomogram to predict status of lymph node metastasis (LNM) in EGC lesions.

Methods

Clinicopathological data of the 4,375 patients who underwent surgically resection for gastric cancer in our center were retrospectively evaluated and finally 626 cases were included. We classified mixed type lesions into five groups (M1:0% < PUC ≤ 20%, M2:20%<PUC ≤ 40%, M3:40%<PUC ≤ 60%, M4:60%<PUC ≤ 80%, M5:80%<PUC < 100%). Lesions with 0% PUC were classified as pure differentiated group (PD) and lesions with 100% PUC were classified as pure undifferentiated group (PUD).

Results

Compared with PD, LNM rate was higher in group M4 and group M5 (p < 0.05 after Bonferroni correction). Differences of tumor size, presence of lymphovascular invasion (LVI), perineural invasion and invasion depth also exist between groups. No statistical difference of LNM rate was found in cases who met the absolute endoscopic submucosal dissection (ESD) indications for EGC patients. Multivariate analysis revealed that tumor size over 2 cm, submucosa invasion to SM2, presence of LVI and PUC level M4 significantly predicted LNM in EGC. With the AUC of 0.899(P < 0.05), the nomogram exhibited a good discrimination. Internal validation by Hosmer–Lemeshow test showed a good fitting effect in model (P > 0.05).

Conclusion

PUC level should be considered as one of the predicting risk factors of LNM in EGC. A nomogram that predicts the risk of LNM in EGC was developed.