内镜下特征联合Ki67对早期结直肠癌黏膜下浸润的评估价值

Endoscopic findings combined with Ki67 for evaluating submucosal invasion of early colorectal cancer

目的 评估内镜下特征联合Ki67判断早期结直肠癌(early colorectal cancer, ECC)黏膜下浸润深度的准确性。方法 收集2019年09月—2022年01月上海第十人民医院内镜中心接受内镜治疗并经术后病理诊断为非浸润性高级别瘤变(high-grade neoplasia, HGN)、黏膜内癌(intramucosal carcinoma, MC)和黏膜下浸润癌(submucosal invasive carcinoma, SMIC)的132例患者的临床病理资料,通过分析ECC的白光内镜下特征、基于NBI内镜下的JNET分型、免疫组化方法检测到Ki67的表达,探讨内镜下特征联合病理免疫组化判断ECC浸润深度的准确性。结果 MC组病灶直径>30 mm占比最高(P<0.05),SMIC组病灶表面存在中央凹陷的占比显著高于HGN组、MC组(P<0.05)。有序多因素Logistic回归分析显示:位于左半结肠及直肠、表面存在中央凹陷、黏膜颜色较周围正常黏膜颜色红和Ki67阳性细胞比率>25%是结直肠病灶发生黏膜下浸润的独立危险因素。ROC曲线分析显示JNET分型联合Ki67阳性细胞比率评估黏膜下深层浸润具有非常高诊断价值(AUC=0.952,95%CI:88.32%~102.16%)。结论 本组ECC在白光内镜下的特征难以准确判断SMIC的浸润深度,而JNET分型联合Ki67阳性细胞比率可提高评估黏膜下浸润深度的准确性。

Objective To assess the application of endoscopic findings combined with Ki67 detection in evaluating the depth of submucosal invasion in early colorectal cancer(ECC). Methods The clinicopathological data of 132 patients with noninvasive high-grade neoplasia(HGN), intramucosal carcinoma(MC) or submucosal invasive carcinoma(SMIC) from September 2019 to January 2022 were retrospectively analyzed. Endoscopic features of early colorectal cancer were analyzed and The Japan NBI Expert Team(JNET) classification was used;the application of JNET classification combined with Ki67 positive cell ratio for evaluating invasive depth of ECC was assessed. Results The proportion of lesions with diameter >30 mm in MC group was the highest(P<0.05), and the proportion of central depression on the surface of lesions in SMIC group was significantly higher than that in HGN group and MC group(P<0.05). Multivariate Logistic regression analysis showed that lesion in the left colon and rectum, central depression on the surface, deeper red mucosal color, and Ki67 positive cells>25% were independent risk factors for submucosal invasion. The area under ROC curve(AUC) of JNET classification combined rate of Ki67 positive cell for predicting risk of deep submucosal invasion was 0.952(95%CI: 88.32%-102.16%). Conclusion The JNET classification combined with the Ki67 positive cell ratio can improve the accuracy in evaluating invasive depth of early colorectal cancer.