Su1369 The Multidetector CT (MDCT) With Contrast Performed in Community Hospital Setting for Suspected Pancreatic Cancer

Abstract

The Multidetector CT (MDCT) With Contrast Performed in Community Hospital Setting for Suspected Pancreatic Cancer Savitha Rao, Pavan Tummala, Banke Agarwal Division of Gastroenterology and Hepatology, Saint Louis University School of Medicine, Saint Louis, MO Background: MDCT with contrast is the most commonly used test for initial evaluation for suspected pancreatic adenocarcinoma (PaCa). The reported sensitivity of MDCT is 76-92% when performed in academic centers and interpreted by dedicated radiologists. However, most MDCT scans for this purpose are performed in community hospitals. There is no published data on its sensitivity and characteristics of PaCa missed by MDCT in this setting.Patients and methods: This is a retrospective analysis from our database of patients who underwent Endoscopic Ultrasound (EUS) from 2002 to 2009. 367 patients who were finally diagnosed to have PaCa were identified; patients were excluded if 1) the initial evaluation was done with MDCT without contrast (n 13) or MDCT with pancreatic protocol (n 45), 2) they were 90 yrs of age (n 5), 3) EUS was performed 4 weeks after MDCT scan (n 41), 4) the pancreatic lesion was cystic (n 40), or was a metastasis to pancreas (n 6). MDCT findings were based on radiology reports obtained from referring hospitals. Final diagnosis of PaCa was based on definitive cytology, surgical pathology or clinical follow-up 1 year. Results: The mean age of 217 patients (116 males) included for analysis was 66.5 12.2 years. 126 patients presented with obstructive jaundice. A definite mass on MDCT was identified in 108 patients, a probable mass in 58 patients and no mass was noted in 51 patients. A dilated pancreatic duct (PD) was noted in 198 patients. Patients with normal sized PD (including 6 patients without identifiable mass lesion) were finally diagnosed to have intrapancreatic cholangiocarcinoma (n 5), or adenocarcinoma in the uncinate process (n 8) and tail (n 6). The sensitivity of MDCT was 49.7% (95% CI 43.1,56.4) and 76.4% (CI 70.3,81.7) when the “probable mass” lesions were counted as false negative and true positive respectively. The median tumor size (range) in patients with definite mass, probable mass and no mass on MDCT was 30 mm(15-100mm), 30 mm(18-70mm) and 25 mm(10-69mm) respectively. For tumors 10-20mm, 21-30 mm, 31-40 mm and 40 mm in size, the sensitivity of MDCT was 33.3%, 49%, 47.1% and 71.8% respectively when ‘probable mass’ was counted as false negative and 55.5%, 71.1%, 90.5% and 87.5% respectively when counted as true positive. Conclusions: In community hospital setting, the sensitivity of MDCT scan with contrast for PaCa is on the lower end of published range of sensitivity from academic centers and is rather low for tumors 20 mm. Even though smaller PaCa is more likely to be missed on MDCT, large tumors that are isodense and hence missed by MDCT are quite common. Presence of PD dilation should warrant further investigation for PaCa even if no mass is noted. However, absence of a mass lesion and non dilated PD on MDCT does not reliably exclude malignancy in patients with strong clinical suspicion of PaCa.