初回検査における胃癌の見逃しとその対策

Failure to Detect Gastric at the Initial Examination and Measures aginst It

要旨　最近3年間に唐津胃研究所，福岡成人病センター，福岡大学第1内科，福岡大学筑紫病院で診断され，診断過程の明らかな614個の胃癌につき検討した．614病変中，紹介例，集検例を除いたルーチン検査例は366病変であった．X線先行例中，少なくとも癌の存在がチェックされたものを対照群(C群)とし，見逃し群(A群)と比較した．同様に内視鏡先行例を見逃し群(a群)と対照群(b群)に分けて比較した．A群，a群ともに対照群に比較して早期癌，多発癌，20mm以下の癌が有意に多かった．X線見逃し群(A群)ではC領域が有意に多かった．見逃し率はX線検査先行例で10.2％，早期癌だけを対照とすると19.0％であった．内視鏡の見逃し例は1例0.8％であった．しかし，2年以内さかのぼった期間に見逃されていた“見逃し既往例”がX線先行例に15例，内視鏡先行例に19例みられた．臨床的には胃癌の確診が内視鏡と生検で確かめられた症例のみに行われることを考えると，内視鏡検査でもX線検査と同じくらいの見逃しが起こっていることが示唆された．また，C群の多発病変や小胃癌の頻度，病変部位の分布などはpanendoscopyで発見されたものの報告と同じであった．以上の事実から，初回X線検査の見逃しはpanendoscopyの見逃しと差はないと考えられ，その対策は自信を持ってX線検査に従事し，質の高い検査を目指すことにあると考えた．

　Six hundred and fourteen lesions of gastric cancer were reviewd regarding their diagnostic process. Three hundred and sixty-six lesions were detected by routine examination and the remaining lesions by either referral or mass screening.

　Among the cases in which roentgenologic examination was done prior to enodoscopy were divided into the two groups: group A consistes of patients whose gastric cancer was not detected, and control group C, gastric cancer was pointed out on roentgenologic examination. In the same way, subject cases in whom endoscopic examination was done prior to roentgenologic examination were divided into two groups: a in which cancer was not detected, and control group b in which cancer was detected on endoscopy.

　Compared with control groups, the proportions of early cancer, multiple cancers, and small cancer with diameter less than 20 mm were high in both group A and a. Area C was more frequently involved in Group A than in control group, which was statistically significant. Failure to detect gastric cancer occurred in 10.2% of the cases in which roentgenologic examination was done first, and that rate was 19.0% when limited to early cancer. Among the cases in which endoscopy was done first, the rate of cancer detection failure only 0.8% (1 case). Retrospective review covering immediate past 2 years, however, "past detection failure" occurred in 15 cases among roentgenologic examination-initiated group and 19 cases among endoscopy-initiated group. Given the fact that final diagnosis of gastric cancer was made by endoscopy with biopsy, it was suggested that rate of cancer detection failure of endoscopy is equivalent to that of roentgenologic examination. Proportions of multiple lesions and minute cancer as well as distribution of sites of cancer lesions in group C were not different from those in whom panendoscopy detected gastric cancer.

　Based on these findings, i.e., no difference in the rate of failure to detect gastric cancer between roentgenologic examination and endoscopy, we are encouraged to perform roentgenologic examination with confidence, making an effort to enhance the quality of the examination.