Published RIC Radiologists

May 2012 RadioGraphics, 32, 764-766. Invited Commentary
Myrosia T. Mitchell, MD
Department of Radiology, Advocate Christ Medical Center Oak Lawn, Illinois

The preceding article by Dr Yamauchi and colleagues (1) provides an excellent overview of the currently used surgical techniques for the treatment of pancreatic neoplasms and pancreatitis. Postoperative patient evaluation can be difficult, particularly when recurrent disease may have subtle manifestations. A thorough understanding of (a) the normal pancreatic anatomy, (b) the surgical procedure and how it alters that normal anatomy, and (c) the normal postoperative changes that may be expected with these pancreatic surgical procedures is intrinsic to the radiologist’s ability to accurately interpret the postoperative imaging findings. Before the development of multidetector CT, this task was particularly daunting because of the inherent limitations of axial scanning techniques, including thicker sections, long scan duration, and section misregistration artifacts. The speed of multidetector CT allows multiphasic scanning, and the isovolumetric capabilities of multidetector CT provide higher resolution, multiplanar and curved reformatted images, and volume rendering, all of which allow a more detailed and accurate assessment of the complex anatomy in these patients. These techniques facilitate delineation of the vasculature and bowel anatomy that is critical for differentiation of normal from pathologic findings.

Contrast material administration and scan timing remain an evolving topic as our understanding of the pharmacokinetics of contrast materials and the physics of multidetector CT imaging improves (2). Power injection of contrast material at rates of 3 mL/sec and higher provide better enhancement of the anatomy and faster times to peak enhancement, but with briefer periods of peak enhancement for the same volume of contrast material (3). This briefer period of enhancement creates its own set of challenges in protocol design (4,5). Yamauchi et al present a triphasic scanning protocol that they use to optimize evaluation of the postoperative pancreas in all of their patients. The reader should note that the protocol as described in the article specifies scanning delays after peak aortic enhancement time determined with contrast material bolus tracking and not actual injection-to-scan delays. In a patient with a normal circulation time to peak aortic enhancement estimated at 20 seconds, these delays equate to an injection-to-scan delay of about 35 seconds for the late arterial/pancreatic phase and 60 seconds after injection for the portal phase. The use of the bolus tracking technique minimizes the effect of individual variations in circulation time that may affect the time to desired enhancement. This bolus tracking technique becomes increasingly important with scanners that have 256 channels and more that reduce imaging times down to 10 seconds and less, because the potential for the scan to miss the optimal desired phase of enhancement increases with less precise scan timing.

Current concerns with regard to radiation safety have thrown into question the need for multiphasic scanning in every patient with suspected pancreatic disease. To minimize patient radiation exposure, we should consider tailoring imaging protocols to the individual patient and the clinical question. Rather than routinely scanning before the administration of contrast material, precontrast scanning can be reserved for specific indications. Precontrast imaging helps identify the hyperattenuation of acute hemorrhage and may be warranted in this situation (6); however, acute hemorrhage is not a frequent clinical concern. Pancreatic calcifications can often be identified on contrast-enhanced CT images (7). Although precontrast scanning may make it easier to identify these parenchymal calcifications and intraductal calculi, picture archiving and communication system (PACS) workstations allow us to review the pancreas with wider window settings and with bone window settings to improve our detection of calculi and other calcifications on contrast-enhanced images. Sensitivity for intraductal calculi may still be limited for smaller calculi, in the presence of hyperenhancing intrapancreatic vessels, and in cases when positive oral contrast material refluxes back up a Roux-en-Y limb to the pancreatic bed. In these settings, when detection of such calcifications is of clinical concern, the addition of precontrast imaging would be helpful. However, if the clinical question is whether there are fluid collections, portal phase scanning is more than sufficient. For primary pancreatic tumor detection the literature has proponents of multiphasic techniques (5,8) and proponents of the single portal phase technique (9), but in the postoperative setting Yamauchi et al describe the typical appearance of recurrent tumor as an infiltrating soft-tissue mass with perineural invasion and vascular encasement. This manifestation of tumor does not necessarily require a multiphasic technique for detection, but if hypervascular liver metastases or vascular complications are a concern, then a multiphasic technique may be indicated. Again, tailoring the protocol to the individual patient and clinical question can avoid unnecessary radiation.

At this time, multidetector CT remains more effective than ultrasonography (US) or MR imaging in the evaluation of the condition of patients after pancreatic surgery (10). US examination can be useful in the follow-up of fluid collections, mass lesions, and other findings but is limited by bowel gas artifacts and patient body habitus. MR imaging has some limitations with bowel imaging, is more sensitive to motion artifact, and is less sensitive in the detection and differentiation of intraductal calculi and air. Nonenhanced MR imaging can be useful in the evaluation of biliary complications of pancreatic surgery (11) and in patients with renal insufficiency, but MR imaging in general remains more of a problem-solving tool to use when CT imaging is insufficient. The consistency of anatomic delineation afforded by multidetector CT makes it the modality of choice for these postoperative patients.

In conclusion, the evaluation of the postoperative pancreas is a challenging clinical problem. The review by Yamauchi et al of the multidetector CT findings after the various pancreatic surgical procedures and their complications should help radiologists to better understand the sometimes confusing postoperative anatomy and to more confidently evaluate the results of imaging examinations in these patients.