Early diagnosis of pancreatic necrosis based on perfusion CT to predict the severity of acute pancreatitis

• Published in: Journal of gastroenterology 2017-10, Vol.52 (10), p.1130-1139
• Main Authors: Tsuji, Yoshihisa, Takahashi, Naoki, Isoda, Hiroyoshi, Koizumi, Koji, Koyasu, Sho, Sekimoto, Miho, Imanaka, Yuichi, Yazumi, Shujiro, Asada, Masanori, Nishikawa, Yoshihiro, Yamamoto, Hiroshi, Kikuchi, Osamu, Yoshida, Tsukasa, Inokuma, Tetsuro, Katsushima, Shinji, Esaka, Naoki, Okano, Akihiro, Kawanami, Chiharu, Kakiuchi, Nobuyuki, Shiokawa, Masahiro, Kodama, Yuzo, Moriyama, Ichiro, Kajitani, Takafumi, Kinoshita, Yoshikazu, Chiba, Tsutomu
• Format: Article
• Language: English
• Subjects: Abdominal Surgery; Adult; Aged; Biliary Tract; Cohort Studies; Colorectal Surgery; CT imaging; Diagnosis; Early Diagnosis; Female; Gangrene; Gastroenterology; Hepatology; Humans; Male; Medical colleges; Medical diagnosis; Medicine; Medicine & Public Health; Middle Aged; Necrosis; Original Article—Liver; Pancreas; Pancreatitis; Pancreatitis, Acute Necrotizing - diagnostic imaging; Pancreatitis, Acute Necrotizing - physiopathology; Perfusion; Perfusion Imaging - methods; Predictive Value of Tests; Prospective Studies; ROC Curve; Severity of Illness Index; Surgical Oncology; Tomography, X-Ray Computed - methods
• ISSN: 0944-1174; 1435-5922
• DOI: 10.1007/s00535-017-1330-5

Background Perfusion CT can diagnose pancreatic necrosis in early stage of severe acute pancreatitis, accurately. However, no study to date has examined whether early diagnosis of pancreatic necrosis is useful in predicting persistent organ failure (POF). Methods We performed a multi-center prospective observational cohort study to investigate whether perfusion CT can predict the development of POF in the early stage of AP, based on early diagnosis of the development of pancreatic necrosis (PN). From 2009 to 2012, we examined patients showing potential early signs of severe AP ( n  = 78) on admission. Diagnoses for the development of PN were made prospectively by on-site physicians on the admission based on perfusion CT (diagnosis 1). Blinded retrospective reviews were performed by radiologists A and B, having 8 and 13 years of experience as radiologists (diagnosis 2 and 3), respectively. Positive diagnosis for the development of PN were assumed equivalent to positive predictions for the development of POF. We then calculated the area under the curve (AUC) of the receiver operating characteristic for POF predictions. Results Fourteen (17.9%) and 23 patients (29.5%) developed PN and POF, respectively. For diagnoses 1, 2, and 3, AUCs for POF predictions were 74, 68, and 73, respectively. Conclusions Perfusion CT diagnoses pancreatic necrosis and on that basis predicts the development of POF; http://www.umin.ac.jp/ctr/index-j.htm,UMIN000001926 .