作者：廖伟林1，刁德昌2，汪佳豪1，林佳鑫1，柯乙森1，李洪明2，易小江2，卢新泉2，冯晓创2，陈昭宇2

单位：1.广州中医药大学第二临床医学院，广东 广州 510405；2.广东省中医院胃肠肿瘤中心 结直肠外科，广东 广州 510120

Authors: Liao Weilin1，Diao Dechang2，Wang Jiahao1，Lin Jiaxin1，Ke Yisen1，Li Hongming2，Yi Xiaojiang2，Lu Xinquan2，Feng Xiaochuang2，Chen Zhaoyu2

Unit: 1.The Second School of Clinical Medical Sciences, Guangzhou University of Chinese Medicine, Guangzhou 510405, China；2.Department of Colorectal Surgery, Gastrointestinal Cancer Center, Guangdong Provincial Hospital of Chinese Medicine, Guangzhou 510120, China

摘要：

目的 探究T1期结直肠癌黏膜下浸润深度不同评估标准的应用特点和临床价值。方法 回顾性收集2013年1月至2021年12月广东省中医院胃肠肿瘤中心行根治性手术切除的T1期结直肠癌患者的临床资料，统计黏膜下浸润深度测量值、Kikuchi分级和Haggitt分级3种不同评估标准下的诊断结果 ，对3种评估方式的准确性、主要差异和影响因素进行分析。结果 共纳入46例pT1期结直肠癌患者，其中男性17例，女性29例，中位年龄65(40～85)岁。病变位于升结肠9例、横结肠2例、降结肠3例、乙状结肠11例、直肠21例；带蒂息肉26例(56.5%)，无蒂息肉20例(43.5%)。中位黏膜下浸润深度3.5(0.2～11.0)mm，黏膜下浸润深度<1 mm者10例(21.7%)，黏膜下浸润深度≥1 mm 36例(78.3%)。Haggitt分级下，评估为2级22例(47.8%)、3级4例(8.7%)、4级20例(43.5%)。Kikuchi分级下，评估为1级24例(52.2%)、2级14例(30.4%)、3级8例(17.4%)。量化测量、Kikuchi分级、Haggitt分级3种评估方式下，判断为黏膜下深度浸润的比例分别为78.3%(36/46)、47.8%(22/46)、43.5%(20/46)；预测淋巴结转移的敏感性分别为100%、100%、50%，特异性分别为23.8%、57.1%、57.1%。量化测量和Kikuchi分级结果 一致性最高(69.6%)，两者呈正相关关系(r=0.657, P<0.001)；量化测量和Haggitt分级结果 一致性最低(21.7%)。量化测量值受黏膜肌层毁损和息肉类型的影响(均P<0.001);Kikuchi分级受黏膜肌层毁损的影响(P=0.003)。结论 3种评估方式中，Kikuchi分级预测淋巴结转移的特异性和敏感性最高，且不受息肉类的影响。黏膜下浸润测量值因受黏膜肌层状态和息肉类型的影响而存在较大的测量误差。Haggitt分级敏感性最低且不适用于无蒂息肉。但3种评估标准下均存在高估淋巴结转移的风险，导致过度手术治疗的弊端。

关键词：早期结直肠癌，黏膜下浸润深度，Kikuchi分级，Haggitt分级

Abstract：

Objective To evaluate the diagnostic characteristics and predictive value of different criteria for submucosal invasion depth in T1 colorectal cancer. Method Clinical data of the patients with T1 colorectal cancer who underwent radical surgery at Gastrointestinal Tumor Center of Guangdong Provincial Hospital of Chinese Medicine from January 2013 to December 2021 were retrospectively collected. Pathological results of submucosal invasion depth assessed by quantitative measurement, Kikuchi classification and Haggitt classification were analyzed. The accuracy and influence factors of the three criteria were analyzed and compared. Result A total of 46 included patients with pT1 colorectal cancer consisted of 17 men and 29 women, in which the median age was 65 years （range 40-85 years）. The tumors located in ascending colon, transverse colon, descending colon, sigmoid and rectum in 9, 2, 3, 11 and 21 cases respectively. There were 26 patients with pedunculated polyps and 20 patients with sessile polyps. The median depth of submucosal invasion （DSI） was 3.5 mm （range 0.2-11.0 mm） in all of the 46 patients including 36 cases（78.3%） with DSI > 1 mm and 10 cases （21.7%） with DSI < 1 mm. According to Haggitt classification, 22 （47.8%）, 4（8.7%）and 20（43.5% ） cases were classified as grade 2, 3 and 4 respectively. In accordance with Kikuchi classification, there were 24 （52.2%） cases with sm1, 14 （30.4%） with sm2 and 8 （17.4%） with sm3. There were 78.3% （36/46）, 47.8% （22/46） and 43.5% （20/46） cases assessed as deep submucosal invasion by quantitative measurement, Kikuchi classification and Haggitt classification respectively, with the sensitivity of predicting LNM of 100%, 100% and 50%, and the specificity of 23.8%, 57.1% and 57.1%, respectively. The assessment results of quantitative measurement and Kikuchi classification showed the highest consistency（69.6%）, and there was a significant positive correlation between them （r=0.657, P<0.001）. The consistency between quantitative measurement and Haggitt classification was lowest （21.7%）. The quantitative measure[1]ments were significantly affected by the recognition of mucosal muscularis and polyp morphology （P<0.001）, while Kikuchi classification was significantly affected by the recognition of mucosal muscularis （P=0.003）. Conclusion Among the three evaluation criteria, Kikuchi classification showed the highest specificity and sensitivity in predicting LNM, and was not significantly affected by polyp morphology. The quantitative measurement of submucosal invasion was affected by the recognition of mucosal muscularis and polyp morphology. Haggitt classification showed the lowest sensitivity and was not suitable for sessile polyps. The risk of LNM was overestimated by the three evaluation criteria, leading to excessive surgical treatment.

Key Words: Early colorectal cancer；Depth of submucosal invasion；Kikuchi classification；Haggitt classification