Multimodality approach brings results in pancreas

Diagnosing pancreatic disease generally requires many different imaging procedures. Ultrasound and CT are most commonly used to evaluate the pancreatic ducts, parenchyma, and adjacent soft tissues. CT-based assessment of pancreatic pathology has been greatly aided by the advent of multislice technology.

Improvements to MRI technology, enabling faster imaging with improved contrast resolution, have likewise had a positive impact on imaging-led diagnoses of pancreatic disease. MRI may now be used to assess pancreatic abnormalities that have equivocal features on multislice CT. MR cholangiopancreatography has been accepted as an accurate method of imaging the pancreatic duct and is now regarded as the primary imaging technique for diagnosing chronic pancreatitis.

Endoscopic ultrasound and PET have also emerged as important complementary modalities for imaging pancreatic abnormalities.

ACUTE PANCREATITIS

The role of ultrasound is rather limited in the early diagnosis and staging of acute pancreatitis, due to the presence of bowel gas (paralytic ileus). It is, however, particularly effective at identifying and eliminating causal factors, such as gallbladder and/or common bile duct stones. Ultrasound is the diagnostic procedure of choice when stones are suspected. It may be helpful in selecting those patients likely to benefit from endoscopic retrograde cholangiopancreatography and sphincterotomy.

Ultrasound is excellent for short term follow-up, particularly in patients who are too ill to undergo CT. Color Doppler ultrasound enables detection of vascular complications such as pseudoaneurysms.

The role of endoscopic ultrasound has been investigated for the detection of choledocholithiasis in patients with biliary pancreatitis.1 This technique can reliably identify cholelithiasis and is more sensitive than transabdominal ultrasound. Endoscopic ultrasound can distinguish edematous and necrotizing pancreatitis, identify peripancreatic fluid collections, and predict disease severity. The presence of infection can be confirmed from ultrasound-guided fine-needle aspiration.

CT is the modality of choice for the diagnosis and staging of acute pancreatitis. It is highly successful in detecting associated complications such as necrosis, which is the hallmark of severe acute pancreatitis. The positive predictive value of CT for detecting necrosis is in the range of 90%. A number of authors have assessed the value of CT for early prediction of morbidity and mortality.2,3

MRI is at least as sensitive as CT for the depiction of necrosis and peripancreatic fluid collections. It also offers a superior assessment of a fluid's internal consistency and ability to drain. MRI can replace CT in patients with renal failure. Adding MR cholangiopancreatography enables the presence of choledocholithiasis to be assessed. The role of MRI at the onset of an acute episode of pancreatitis is limited, though. Patients with severe pancreatitis are generally too ill to cooperate with an MRI examination.

CHRONIC PANCREATITIS

Chronic pancreatitis is a progressive inflammatory disease resulting in permanent structural damage to the pancreas. Although early changes of uncomplicated chronic pancreatitis are difficult to detect with ultrasound, this modality is especially accurate at detecting parenchymal and ductal changes in advanced or complicated chronic pancreatitis.

Endoscopic ultrasound is the test of choice for detection of cystic changes in the diagnosis of chronic pancreatitis. It can localize and characterize pseudocysts in detail and identify the best candidates for drainage. Endoscopic ultrasound may reveal abnormalities in patients with early stages of chronic pancreatitis when other imaging tests, including endoscopic retrograde cholangiopancreatography, are normal. Differentiation between benign and malignant ductal stenoses or focal masses is still difficult to establish with endoscopic ultrasound alone. It is consequently followed by fine-needle aspiration, CT, or MRI when differentiating small pancreatic cancers from focal chronic pancreatitis.

CT outclasses all imaging modalities-ultrasound included-when it comes to detecting calcifications, a specific sign of advanced chronic pancreatitis.4 CT can also detect most complications associated with chronic pancreatitis, such as pseudocysts, calculi in the pancreatic duct, inflammatory masses, or pseudoaneurysms (Figure 1). Its sensitivity to early chronic pancreatitis is poor, though. Calcifications and parenchymal atrophy occur late in the course of the disease or in patients who already have severe disease.

MRI is less sensitive than CT for the detection of calcifications associated with chronic pancreatitis. It is more sensitive, however, for the detection of early chronic pancreatitis, prior to the development of calcifications. Fat-suppressed T1-weighted MRI demonstrates abnormally low signal intensity. Immediate and delayed imaging, following administration of gadolinium-based contrast, reveals decreased enhancement to the inflamed parenchyma of the pancreas.

MRI can also help distinguish inflammatory masses from tumors. Up to 30% of patients with chronic pancreatitis will have a focal inflammatory mass, which typically involves the pancreatic head. Differentiation is often difficult because the mass may simulate pancreatic malignancy. The presence of tumor can be ruled out if MRI shows positive duct penetration (dilated side branches in the mass).

PANCREATIC ADENOCARCINOMA

Significant progress has been made in the diagnosis of pancreatic cancer. The major aims of imaging pancreatic malignancies are to make an early diagnosis, differentiate between cancer and inflammation, stage the disease correctly, and identify patients whose disease is not amenable to resection.

Transabdominal ultrasound is commonly used as the primary method of detecting tumor. Reports of its sensitivity range from 60% to 90%.5,6 These differences in sensitivity are related to patient-dependent factors, the operator's degree of training, and the tumor's size and location. While detecting large tumors is usually easy, small tumors and tumors located in the body and tail of the pancreas may be difficult to visualize because of interference from bowel gas.

Advances in equipment and the introduction of contrast agents mean that despite its limitations, ultrasound can play an important role in diagnosing pancreatic neoplasms in most patients. It is inferior to CT for the staging of pancreatic neoplasms and is employed less often in this role. Color Doppler ultrasound of the superior mesenteric artery or celiac trunk is of little importance because lumen narrowing occurs only in advanced disease.

Endoscopic ultrasound has become an important way to assess pancreatic tumors, especially small tumors. Endoscopic ultrasound-guided fine-needle aspiration biopsy can increase diagnostic accuracy, while imaging alone can localize lymph node metastases or vascular tumor infiltration with high sensitivity. The technique is not without limitations. It is operator-dependent, requires patients to be sedated, and, perhaps most important, is unable to examine the entire liver or detect peritoneal metastases. Many patients with pancreatic cancer will present with liver and/or peritoneal metastases, so this is a significant drawback.

CT is now considered to be the imaging modality of choice for the detection and presurgical staging of pancreatic cancer. Its performance is less operator- and patient-dependent than that of ultrasound. CT's sensitivity in diagnosing pancreatic neoplasms and its accuracy in tumor staging improved significantly following the introduction of multislice technology.7

The accuracy of CT for detecting pancreatic tumors ranges from 80% to 91%, according to the literature. Reported positive predictive values for surgical unresectability range from 89% to 95%. CT is less good, however, at judging tumor resectability. Surgical results show that 60% to 91% of tumors that CT shows to be resectable can be excised.8 This discrepancy is largely due to noncontrast CT's inability to detect vascular involvement by the tumor.

One group of researchers reported an improvement in pancreatic cancer staging by using MSCT and interactive multiplanar reconstructions.9 The negative predictive value of resectable tumors was 96% when CT angiography was added, compared with 70% for axial images alone. Features indicating vascular involvement include tumor involvement for one half of the vessel's circumference, dilatation of peripancreatic veins, and focal narrowing of the vessel (Figure 2).

Accurate staging implies evaluation of locoregional and distant metastases, as well as the assessment of vascular involvement. Identification and characterization of small hepatic lesions and detection of small metastatic peritoneal and omental metastases remain problematic.7 CT may provide a false-negative result if peritoneal metastatic disease is not accompanied by ascites. CT is also unable to differentiate accurately between benign and malignant causes of lymph node enlargement.

The role of PET/CT in most patients with pancreatic cancer is unclear. This combined modality allows detection of unsuspected metastases, helps evaluate masses with equivocal diagnoses on CT and MRI, and aids evaluation of patients with suspected recurrent pancreatic carcinoma.

Technical advances in MRI, such as higher field magnet strengths, phased-array coils, and ultrafast imaging, have yielded excellent results in the detection and staging of pancreatic cancer. MRI performed with a high gradient system and phased-array coils can produce a sensitivity of 95% for the diagnosis of pancreatic tumors. MRI has a positive predictive value of 90% and a negative predictive value of 83% for nonresectability. Criteria for unresectability include involvement of the superior mesenteric artery, involvement of the celiac trunk or the superior mesenteric vein-portal vein confluence, and observed peritoneal or hepatic metastases.10

The primary role of MRI lies in the detection of small lesions that are potentially curable and avoidance of possible understaging with CT. MRI is also helpful in detecting or excluding pancreatic cancer in patients whose CT and/or ultrasound images showed a prominent head or uncinate process. One group of investigators showed mangafodipir trisodium-enhanced MRI and spiral CT to be equivocal for local staging of pancreatic cancer but found that contrast-enhanced MRI offered advantages in the detection of small pancreatic malignancies and liver metastases.11 Underdetection of small peripancreatic lymph nodes and peritoneal implants remains a limitation for MRI.

CYSTIC LESIONS

There are a wide variety of cystic pancreatic lesions. Pseudocysts are by far the most common of these. Ultrasound is extremely useful in differentiating cystic neoplasms from pseudocysts. Most pseudocysts are single and do not have the solid components, septae, or loculations that are characteristic of cystic tumors.

More than 90% of all pancreatic cystic neoplasms are either mucinous cystadenomas or cystadenocarcinomas, serous cystadenoma, or intraductal papillary mucinous tumors (IPMT). This last class of tumor has a favorable prognosis, unclear nature, and obscure relationship to ductal adenocarcinoma.12 Ultrasound, CT, or MRCP can demonstrate the clustered multicystic dilatation of branch ducts (branch type) or the diffuse dilatation of the main pancreatic duct (main duct type). These modalities can also reveal the presence of intraductal abnormalities due to mucin (Figure 3).

An important finding, for branch-type IPMT, that can be made using cross-sectional imaging is communication of the cystic lesion with the pancreatic duct. Both thin-section spiral CT and MRI can demonstrate this well (Figure 4). Administration of intravenous secretin reportedly improves detection of this phenomenon on MRCP.13 Endoscopic retrograde cholangiopancreatography is the best modality for confirming IPMT when diagnosis of this entity remains uncertain. Endoscopic observation can reveal an enlarged papilla with mucus flowing from a patulous orifice.

Imaging also plays an important role in the differential diagnosis of serous versus mucinous lesions (Figure 5). CT and ultrasound can distinguish mucinous cystic neoplasms from microcystic adenomas in approximately 90% of cases. In a review of 45 pathologically proven pancreatic cystic neoplasms, including 29 mucinous cystic tumors and 16 serous cystadenomas, ultrasound diagnosed 78% of cases correctly, compared with 93% for CT.14

Morphological information from MRI and endoscopic ultrasound can help with definitive diagnoses when typical internal architecture is absent. Image-guided cyst fluid sampling is also possible with these modalities. Endoscopic ultrasound seems especially effective at determining which cystic lesions have malignant potential.

ENDOCRINE TUMORS

Endocrine pancreatic tumors are rare lesions that are frequently malignant, despite their usually slow progression. Many different techniques have been proposed for their diagnosis. These include transabdominal and endoscopic ultrasound, CT, MRI, arteriography, transvenous sampling, and octreotide scanning.

The success of ultrasound for the detection of insulinoma ranges from 25% to 65%.15 This drops to an average of 20% for gastrinoma detection.16 Endoscopic ultrasound is generally reported as being far superior for the detection of endocrine pancreatic tumors when compared with transabdominal ultrasound, CT, and angiography. One group of researchers found its overall diagnostic accuracy to be 87.5%.17

The reported sensitivity of CT in localizing functioning islet cell tumors varies from 71% to 82%. This is because small tumors are missed more often. Small hyperattenuating islet tumors in the pancreatic neck or body can be confused with adjacent vascular structures. Multiplanar reconstruction can help separate the lesion from surrounding vessels, thereby improving diagnostic confidence. A statistically significant correlation may be made between tumor vascularity, as evaluated by histological techniques, and lesion enhancement from CT data.18 Endocrine pancreatic tumors with poor enhancement on pancreatic-phase spiral CT correlate well with poorly differentiated endocrine pancreatic tumors and a decrease in the overall survival rate.

The sensitivity of MRI depends largely on the technique used (Figure 6). Fast, high-resolution imaging is critical. Several studies suggest that state-of-the-art MRI may be more sensitive than CT.19,20

A large number of diagnostic procedures are available for the workup of patients with pancreatic disease. No single correct approach exists. Each modality has its strengths and weaknesses. Nonetheless, local expertise and interest and the availability of high-end technology will direct a chosen strategy for the workup of most patients.

DR. VANBECKEVOORT is an abdominal radiologist at the University Hospitals Leuven in Belgium.

Assisting in the preparation of this manuscript were Dr. Hilde Vandenhout, Dr. Didier Bielen, Prof. Steven Dymarkowski, Dr. Katya Op de beeck, Dr. Ragna Vanslembrouck, and Mrs. Ingrid Fruyt, all from the department of radiology, University Hospitals Leuven.

References

• Chak A, Hawes RH, Cooper GS, et al. Prospective assessment of the utility of EUS in the evaluation of gallstone pancreatitis. Gastrointest Endosc 1999;49(5):599-604.

• Balthazar EJ, Freeny PC, van Sonnenberg E. Imaging and intervention in acute pancreatitis. Radiology 1994;193(2):297-306.

• Mortele KJ, Wiesner W, Intriere L, et al. A modified CT severity index for evaluating acute pancreatitis: improved correlation with patient outcome. AJR 2004;183(5):1261-1265.

• Glasbrenner B, Kahl S, Malfertheiner P. Modern diagnostics of chronic pancreatitis. Europ J Gastroenterol Hepatol 2002; 14(9):935-941.

• Lynch HT, Brand RE, Lynch JF, et al. Hereditary factors in pancreatic cancer. J Hepatobiliary Pancreat Surg 2002;9(1):12-31.

• Minniti S, Bruno C, Biasiutti C, et al. Sonography versus helical CT in identification and staging of pancreatic ductal adenocarcinoma. J Clin Ultrasound 2003;31(4):175-182.

• Scaglione M, Pinto A, Romane S, et al. Using multidetector row computed tomography to diagnose and stage pancreatic carcinoma: the problems and the possibilities. JOP 2005;6(1):1-5.

• Prokesch RW, Chow LC, Beaulieu CF, et al. Local staging of pancreatic carcinoma with multi-detector row CT: use of curved planar reformations-initial experience. Radiology 2002;225(3):759-765.

• Baum U, Lell M, Nomayr A, et al. [Multiplanar spiral CT in the diagnosis of pancreatic tumors]. Radiologe 1999;39(11):958-964. German.

• Mammone JF, Siegelman ES, Outwater EK. Magnetic resonance imaging of the pancreas and biliary tree. Semin Ultrasound CT MR 1998;19(1):35-52.

• Schima W, Fugger R, Schober E, et al. Diagnosis and staging of pancreatic cancer: comparison of mangafodipir trisodium-enhanced MR imaging and contrast-enhanced helical hydro-CT. AJR 2002;179(3):717-724.

• Taouli B, Vilgrain V, O'Toole D, et al. Intraductal papillary mucinous tumors of the pancreas: features with multimodality imaging. JCAT 2002;26(2):223-231.

• Procacci C, Megibow AJ, Carbognin G, et al. Intraductal papillary mucinous tumor of the pancreas: a pictorial essay. Radiographics 1999;19(6):1447-1463.

• Johnson CD, Stephens DH, Charboneau JW, et al. Cystic pancreatic tumors: CT and sonographic assessment. AJR 1988;151(6):1133-1138.

• Galiber AK, Reading CC, Charboneau JW, et al. Localization of pancreatic insulinoma: comparison of pre- and intraoperative ultrasound with CT and angiography. Radiology 1998;166(2):405-408.

• Frucht H, Doppman JL, Norton JA, et al. Gastrinomas: comparison of MR imaging with CT, angiography and US. Radiology 1989;171(3):713-717.

• Palazzo L, Roseau G, Salmeron M. Endoscopic ultrasonography in the preoperative localization of pancreatic endocrine tumors. Endoscopy 1992;24(Suppl 1):350-353.

• Rodallec M, Vilgrain V, Couvelard A, et al. Endocrine pancreatic tumours and helical CT: Contrast enhancement is correlated with microvascular density, histoprognostic factors and survival. Pancreatology 2006;6(1-2):77-85.

• Pfannenberg AC, Burkart C, Krober SM, et al. Dual-phase multidetector thin-section CT in detecting duodenal gastrinoma. Abdom Imaging 2005;30(5):543-547.

• Semelka RC, Cumming MJ, Shoenut JP. Islet cell tumors: comparison of dynamic contrast-enhanced CT and MR imaging with dynamic gadolinium enhancement and fat suppression. Radiology 1993;186(3):799-802.