膵癌の早期診断法の検討 ―膵集検の試み―

膵癌を治癒に導くためには早期診断，早期切除以外に方法はない。そこで，無症状者のなかから早期膵癌を発見するためのスクリーニング法を検討するためProspective studyを実施した。人間ドック目的の来院患者を主とする三朝分院外来患者776名を対象とし，一次スクリーニング検査として，血清アミラーゼ，エラスターゼⅠ，CA19－9，アルカリフォスファターゼ，γ-GTP，血糖値の測定および腹部超音波検査を施行し，異常者に対して二次スクリーニングとして腹部超音波検査の再精査およびERCPを施行した。その結果，早期膵癌1例および進行膵癌2例を検出し，膵癌発見率は0.39％　と満足のいく結果が得られた。しかし一次スクリーニング検査の要精検率は46％　と高値を示したこと，一次スクリーニング検査の費用は胃集検の約3.3倍，二次スクリーニング検査の費用は胃精査の約2.6倍かかることからcost-benifitの面で問題となった。また見逃し症例の有無をモニターするためfollow upも今後の問題として残った。

There is no other way of curing pancreatic cancer than early diagnosis and resection. However, an effective protocol has not been established for detecting early pancreatic cancer among asymptomatic populations by means　of a mass screening test. Therefore, a prospective study was attempted on 776 patients who came to Medical Clinic of Misasa Branch Hospital, Okayama University Medical School mostly for a routine annual check-up. The following tentative protocol was proposed, based on the retrospective analyses of our 13 patients with early pancreatic cancer: (1) a first step test includes serum amylase, elastase Ⅰ, alkaline phosphatase, r-GTP, carcinoembryonic antigen, fasting blood sugar and routine abdominal ultrasonography (US); (2) a second step test includes thorough abdominal US with various kinds of probes and patients' positioning which was performed on individuals with abnormal findings on the first step test, (3) a third step test includes ERCP which was performed on individuals with abnormal or questionable findings on the second step test. Further detail examinations were performed when ERCP was abnormal or questionable. Consequently 357 patients (46% of the total 776 patients) underwent thorough US and 45 patients (5.8%) underwent ERCP. ERCP findings included pancreatic cancer in two patients, suspect of pancreatic cancer in three, advanced chronic pancreatitis in two, equivocal as for chronic pancreatitis (according to the Cambridge classification) in nine, and normal in 29. Final diagnosis was early pancreatic cancer in one patient, advanced pancreatic cancer in two, and calcified chronic pancreatitis in two; detection rate of pancreatic cancer (0.39%) in this series was slightly higher than that usually reported with a mass screening test for stomach cancer. From a cost-benefit point of view, however, examinations for the pancreas cost 2.6 times as much as those for the stomach. Further studies are needed (1) to refine the protocol, (2) to improve cost-benefit efficiency, (3) and also to confirm by follow-up that no cases of pancreatic cancer have passed through the test undetected.