Institutional members access full text with Ovid®

Computed Tomography Perfusion Analysis of Pancreatic Carcinoma

Klau, Miriam MD; Stiller, Wolfram PhD; Fritz,, Franziska; Kieser, Meinhard PhD; Werner, Jens MD; Kauczor, Hans-Ulrich MD; Grenacher, Lars MD

Journal of Computer Assisted Tomography: March/April 2012 - Volume 36 - Issue 2 - p 237–242
doi: 10.1097/RCT.0b013e31824a099e
Abdominal Imaging

Objective: The purpose of this study was to evaluate CT perfusion of pancreatic carcinomas using the Patlak model for assessing perfusion, permeability, and blood volume.

Methods: A total of 25 patients with pancreatic carcinoma were examined prospectively with a 64-slice computed tomography (CT) using a dynamic sequence after intravenous injection of 80-mL contrast material (370 mg/mL; flow rate, 5 mL/s). Eighty-kilovolt (peak) perfusion acquisitions were evaluated for estimating perfusion parameters for carcinoma and healthy tissue using a 2-compartment model (Patlak model).

Results: Twenty patients had hypodense tumors; in 5 patients, the tumor could not be delineated in contrast-enhanced CT. All carcinomas could be identified clearly in the color-coded perfusion maps. Perfusion, permeability, and blood volume values were significant lower in pancreatic carcinomas compared to healthy pancreatic tissue (0.27 ± 0.20 vs 0.89 ± 0.19 min−1, P < 0.0001; 0.43 ± 0.20 vs 0.75 ± 0.16 × 0.5 min−1, P < 0.0001; and 38.9 ± 20.7 vs 117.8 ± 46.9 mL/100 mL, P < 0.0001).

Conclusion: Computed tomographic perfusion of the pancreas using a 2-compartment perfusion model is feasible. Color-coded perfusion maps could be a helpful tool to delineate pancreatic carcinomas even if they are not visible in contrast-enhanced CT.

From the *Department of Diagnostic and Interventional Radiology, †Institute of Medical Biometry and Informatics, and ‡Department of Surgery, University of Heidelberg, Heidelberg, Germany.

Received for publication November 16, 2011; accepted January 6, 2012.

Reprints: Miriam Klauß, MD, Im Neuenheimer Feld 110, 69120 Heidelberg, Germany (e-mail: miriam.klauss@med.uni-heidelberg.de).

The authors report no conflicts of interest.

Copyright © 2012 Wolters Kluwer Health, Inc. All rights reserved.