食道表在癌の深達度診断―通常内視鏡の立場から

Endoscopic Estimation of Cancer Invasioninto the Esophageal Wall

要旨　2006～2009年までに内視鏡治療を行った食道表在癌247例296病変を対象に深達度診断の診断精度について検討した．正診率はT1a-EP・LPM癌では95％，T1a-MM・SM1癌は66％，SM2・SM3癌は61％であり，全体で90％（296病変中265病変）であった．病型は，T1a-EP・LPM癌はIIc 61％，IIb 31％，T1a-MM・SM1癌はIIc 89％，SM2以深癌はIIc 67％，0-I 28％であり，IIcが全体の65％を占めていた．食道表在癌の深達度診断においてIIcが最も重要な病型であった．0-IIc（EP・LPM）癌の正診率は97％と良好であった．0-IIc（MM・SM1）癌の正診率は68％で，浅読み例の71％が微小浸潤例であった．また，0-IIc（SM2以深）癌の正診率は50％で，誤診例の半数が微小浸潤例であり，SM1との鑑別が困難であった．現在の深達度診断は，通常観察で予測をし，拡大観察で確診を得るという過程で行われていた．通常観察の診断の限界は浸潤幅1.8mm程度の微小浸潤であり，それらは拡大観察を行っても血管変化がみられない病変では診断が困難であった．

　In cases with squamous cell carcinoma of the esophagus, depth of cancer invasion into the esophageal wall shows a close relation to incidence of microvascular permeation and lymph node metastasis which should be considered in selection of treatment for superficial esophageal cancer. Accuracy of endoscopic estimation of depth of cancer invasion was studied in 247 patients with 296 lesions of superficial esophageal cancer treated by endoscopic resection from 2006 to 2009 at the Metropolitan Komagome Hospital Tokyo, Japan. Pathological studies on resected specimens revealed 243 lesions with invasion of EP and LPM（type 0-IIc occupied 61％ and type 0-IIb 31％）,35 lesions with MM and SM1（type 0-IIc occupied 89％）and 18 with invasion of SM2 or more（type 0-IIc 67％ and type 0-I 28％）. Endoscopic estimation was correctly carried out in 90％ of all cases（95％ in cases with T1a-EP and LPM,66％ T1a-MM and SM1, and 61％ SM2 and SM3）. Type 0-IIc cancer lesions were most frequent among superficial esophageal cancers（65％ of all lesions）. In case of type 0-IIc with invasion of EP or LPM, depth of invasion was correctly estimated by endoscopy in 97％ of all lesions,MM and SM1 68％ and 50％ SM2 or more. In case of type 0-IIc with MM or SM1 invasion, endoscopic estimation of invasion failed in 32％ of all lesions. Pathological studies on resected specimens revealed narrow invasion less than 1.8mm in width into the muscularis mucosae in 71% of all type 0-IIc with underestimation of depth of invasion. Fifty percent of all type 0-IIc and SM2 cancer cases were correctly estimated by endoscopy, while 25％ of all SM2 cases were difficult to differentiation from SM1, because of deeper invasion in a very narrow area. At present, we apply conventional endoscopy for esophageal cancer cases first, and follow it up by magnified endoscopy for further evaluation of any findings suggestive of deeper invasion. A narrow and deeper invasion of less than 1.8mm in width probably presents abnormality to neither conventional endoscopy nor magnifying endoscopic observation.