胃癌診断過程におけるpanendoscopyの位置づけ

Role of Endoscopy in the Diagnosis of Cancer of the Stomach

要旨　癌研付属病院における最近の内視鏡診断の実態と，使用器種別の能力とを明らかにするために1981，1982年の12,029件の上部消化管内視鏡検査の成績を，一部コンピューターを利用して解析した．被検者への侵襲，1回で食道から十二指腸まで検査しうる点でpanendoscopeは先行するX線検査などの情報が十分でない場合は第一選択とすべきだが，記録性に劣ること，病変の多い胃体部胃角の後壁が観察しづらいなど，万能とは言い切れないものがあり，内視鏡検査に当たっては各器種の特徴をよく知ったうえで最も対象に適した器種を選択すべきであると考えられた．

　Role of endoscopy in the diagnosis of cancer of the stomach was analyzed, based on 12,029 examinations of endoscopy which had been done at the Cancer Institute Hospital between 1981 and 1982. The data for 1982 were analyzed by office computer (IBM System/34). The following results were obtained.

　(1) GTF was used in 7,051 examinations (58.6％), GF-B3 in 2,938 examinations (24.4％) and GIF-P in 1,856 examinations (15.5％). GIF was most frequently used for the stomach (57.8％), second frequently for the esophagus (24.9％), and third frequently for the douodenal bulb (17.3％).

　(2) The first endoscopic diagnosis resulted in false negative diagnosis in 121 cases which were finally diagnosed as cancer. Of these, the first diagnosis of peptic ulcer was made in 35.5％, that of ulcer scar in 22.3％ and that of polyp in 12.4％. In relation to location of cancer, the incisura angularis was most frequent site of the false negative diagnosis. In relation to the type of endoscope used in this region, the false negative diagnosis was most frequently made by GIF-P (40％), second frequently by GF-B3, and third frequently by GTF (10.7％).

　(3) The false negative diagnosis by the first biopsy was done in 4％ by GIF-P and 2.4％ by GF-B3. It was most frequent at the gastric body by GIF-P and was most frequent at the gastric antrum by GF-B3.

　(4) Endoscopic diagnosis of cancer was greatly influenced by the information of preceding x-ray diagnosis. It was an interesting fact that 46％ of the cases which were not diagnosed as cancer by x-ray were not also diagnosed as cancer by endoscopy.

　(5) Sensitivity to the diagnosis of cancer was 91.5％ by GTF, 87.5％ by GIF-P and 86.8％ by GF-B3, but specificity was lowest by GTF (85.5％) because of over-reading. In respect to location of cancer, sensitivity by GIF-P was low (80.5％) from the incisura angularis to gastric body, and that by GF-B3 was low (83.3％) at the pylorus.