Pancreatic MR defines ducts, pinpoints disease
Chemical shift imaging can characterize an adrenal mass as an incidental adenoma and exclude metastasis
By Evan S. Siegelman, M.D.
Patients with genetic hemochromatosis can develop diabetes from damage to the insulin-secreting beta cells of the pancreas secondary to iron overload.9 The iron overload in both the liver and pancreas can be diagnosed in some patients by recognizing a loss of signal as the echo time increases from 2.1 to 4.2 msec. at 1.5 tesla. Even though the in-phase (TE = 4.2 msec) gradient-echo sequence is considered "T1-weighted," the T2* shortening effects of the iron-overloaded pancreas may be great enough to result in signal loss.
MRI is a useful tool for detecting and characterizing pancreatic disorders. Fat-saturated T1-weighted images are very sensitive at finding focal pancreatic disease and MR cholangiopancreatography (MRCP) can define both normal and abnormal pancreatic ducts. MRI also aids in establishing a diagnosis of unresectable adenocarcinoma and in characterizing neuroendocrine tumors and cystic neoplasms of the pancreas.
Figure 1. In a case of suspected pancreatic adenocarcinoma, an axial image obtained from a volume-acquired 3-D fat-suppressed gradient-echo sequence shows normal pancreatic head and neck (arrow). The remainder of the pancreas is of abnormal low signal intensity (*) and difficult to differentiate from the suppressed peripancreatic fat.|
On T1-weighted images, the normal pancreas has higher signal intensity than any other abdominal organ. Its short T1 relaxation time has been attributed to the abundant protein and rough endoplasmic reticulum contained within it. Fat-saturated T1-weighted sequences are useful for distinguishing normal from abnormal pancreatic parenchyma.1 When the signal from fat, which is the only other normal abdominal soft tissue that is routinely of higher signal intensity than pancreas, is suppressed, the dynamic range of the resultant image narrows, making it easier to differentiate abnormal from normal pancreatic signal intensity (Figures 1 and 5A). In patients who can hold their breath, fat-saturated gradient-echo sequences can be done instead of fat-saturated spin-echo.2
Breath-hold, in-phase, and opposed-phase T1-weighted gradient-echo images can be substituted for the longer breathing-averaged spin-echo sequences. Gradient-echo sequences reveal similar anatomic information concerning the pancreas as spin-echo sequences and also reveal chemical shift information.
Chemical shift imaging of the pancreas and peripancreatic structures has many potential uses. When evaluating pancreatic carcinoma, it can characterize an adrenal mass as an incidental adenoma and exclude metastasis.3,4 In addition, chemical shift imaging can detect three uncommon pancreatic conditions:
- Focal fatty change of the pancreas can mimic a mass on sono-graphy or contrast-enhanced CT.5
Chemical shift imaging reveals the lipid content of the affected segment of pancreas and thus excludes a diagnosis of neoplasm.6
- Patients with renal cell carcinoma (RCC) can have metastatic disease to the pancreas. Clear cell RCC, which is the most common subtype of renal cancer, may contain intracellular lipid that can be detected with chemical shift imaging.7 Some pancreatic metastases of clear cell RCC will also reveal loss of signal with chemical shift imaging, thus excluding a diagnosis of pancreatic adenocarcinoma.8
Figure 2. Two corresponding axial images from a 2-D in-phase T1-weighted gradient-echo sequence show fat as high signal intensity to delineate the boundaries of the abnormal segment of pancreas.
A: Flow void (arrow) is present within the splenic artery (**), which is surrounded by low-signal-intensity tissue that is suspicious for encasement.
B: The presence of normal fat around the superior mesenteric artery (*) is helpful for excluding encasement. A border deforming mass of the posterior pancreatic body (white arrow) is present in the expected location of the central splenic vein. Cross-sectional depiction of a tubular structure (black arrow) around the stomach represents dilated gastroepiploic veins, which become enlarged in splenic vein thrombosis.
Figure 3. Image from a volume-acquired 3-D fat-suppressed gradient-echo sequence obtained during the portal venous phase of contrast enhancement shows hypovascular adenocarcinoma (arrow) that encases the splenic artery. The distal pancreatic parenchyma was difficult to distinguish from the adenocarcinoma on both this precontrast T1-weighted image and on precontrast T2-weighted images. After contrast administration, the uninvolved distal pancreas (curved arrow) shows near-normal enhancement and is readily distinguished from the adjacent carcinoma. A prominent enhancing gastroepiploic vein cut in cross section is again seen (small arrows).
Figure 4. Axial maximum intensity projection image from the portal venous phase of contrast enhancement reveals the gastroepiploic vein (arrow), which provides collateral flow to the superior mesenteric vein via the gastrocolic trunk. This collateral vessel becomes enlarged in patients with splenic vein obstruction and occlusion. Note the absence of contrast within the expected location of the splenic vein (**), which was occluded by tumor.
Figure 5. Characteristic MRI features are seen in a 1-cm hyperfunctioning insulinoma of the junction of the pancreatic body and tail in a 33-year-old woman.
A: Axial T1 fat-suppressed gradient-echo image reveals normal high signal intensity pancreas (large arrow). There is a 1-cm mass of the anterior pancreas (small arrow). The adjacent focus of even lower signal intensity (**) represents a flow void within a segment of splenic artery.
B: Fast spin-echo T2-weighted image show that the mass has higher signal intensity (arrow) compared to the adjacent pancreas. There is no peripheral duct dilation or parenchymal atrophy.
C: Axial T1 fat-suppressed gradient-echo image obtained during the arterial phase of contrast administration shows marked homogeneous enhancement of the tumor (arrows) to almost the same degree as the adjacent splenic artery (**).
MRCP is one application of a technique generally referred to as MR hydrography.10-15 This term reinforces the concept that an MR system does not selectively "enhance" the bile or pancreatic secretions. Any other fluid-containing structures within the imaging volume will also show high signal intensity.
MRCP has many applications in the evaluation of pancreatic diseases.16,17 It can detect the level of the obstructive mass in patients with suspected pancreatic cancer and show the classic double duct sign in patients with obstructive jaundice.
MRCP can be used to diagnose pancreatic divisum.18 In patients with pancreatic divisum and abdominal pain, the latter has been attributable to a stenosis of the accessory duct of Santorini with obstructive changes of the dorsal pancreas. Secretin stimulation of the pancreas can be used to create "stress" or "functional" MRCP by increasing the production and flow of bicarbonates and fluid within the pancreatic ducts.19
Stress MRCP in patients with pancreatic divisum can reveal dilation of the duct of Santorini and a cystic dilation of the dorsal duct just before its insertion into the minor papilla. The latter finding, termed a "Santorinicele," may result from distal duct obstruction and is analogous to a ureterocele or choledochocele.20
An MRCP sequence does not represent a comprehensive MR examination of the pancreas.22 The depiction of a double duct sign, for example, may reveal the site of the obstructing pancreatic carcinoma in a patient with obstructive jaundice, but it provides no information about the staging of the cancer. The soft-tissue contrast of many MRCP sequences is optimized to show fluid as high signal intensity and other soft tissue as very low signal intensity. Findings that may make a patient unresectable, such as vascular encasement (Figures 2, 3, and 4) or liver metastases, are better detected on non-MRCP pulse sequences.23
Radiologists should familiarize themselves with their surgeons' criteria for unresectable pancreatic adenocarcinoma disease and the imaging findings that satisfy those criteria with high specificity and positive predictive value.24 Establishing a definitive MR diagnosis of unresectable disease prevents a subset of patients with incurable pancreatic cancer from undergoing unnecessary surgery for an attempted curative resection ("peek and shriek"). These patients would benefit more from palliative therapy and stenting or bypass of the bile ducts and duodenum.
A diagnosis of unresectable pancreatic cancer can be established with high specificity and positive predictive value on T1-weighted images alone if they show a pancreatic mass that replaces the normal fat around the proximal segment of the celiac artery or superior mesenteric artery. Even though fat-saturated T1-weighted images are the most sensitive for detection of pancreatic disease, it can be difficult to assess vascular invasion because both tumor and suppressed normal fat may be of similar low signal intensity and thus indistinguishable. While liver metastases are often detected on T1-weighted images, I use T2-weighted images and/or contrast- enhanced images to characterize liver lesions as benign or malignant.
Primary ductal adenocarcinoma of the pancreas tends to have signal intensity on T2-weighted images that is hypointense to isointense to normal pancreas. Liver metastases are usually isointense to spleen, while benign cysts and hemangiomas, as well as intrahepatic duct dilation, follow the signal intensity of cerebrospinal fluid.
On dynamic contrast-enhanced imaging, pancreatic cancers are hypovascular relative to normal pancreas (Figure 3). The arterial or portal phase of a contrast-enhanced study best depicts the tumor margins. Thrombosis or encasement of the splenic, superior mesenteric, and portal veins is best revealed on the venous phase of contrast enhancement (Figure 4).
Neuroendocrine tumors of the pancreas have a better prognosis than ductal adenocarcinoma of the pancreas. Hyperfunctioning tumors such as insulinomas, gastrinomas, and vipomas tend to be smaller and more homogeneous than nonhyperfunctioning neuroendocrine tumors.25
Unlike ductal adenocarcinomas, neuroendocrine tumors of the pancreas (islet cell tumors) generally have solid components that are hyperintense relative to pancreas on T2-weighted images (Figure 5B).26,27 Intratumoral cyst formation and necrosis may be present in larger, nonhyperfunctioning tumors that will be depicted as very high signal intensity on both T2-weighted images and MR hydrograms. While ductal adenocarcinoma is hypovascular after contrast, neuroendocrine tumors often have components that are hypervascular compared with pancreas during dynamic imaging.
Cystic pancreatic neoplasms. The two major subtypes of cystic pancreatic neoplasms are microcystic adenomas and mucinous cystic neoplasms of the pancreas. Microcystic adenomas, which are considered benign, typically contain more than six cysts that individually measure no more than 2 cm.28 The cysts are separated by thin septa that may enhance. Surgery can be avoided when tumors show these classic imaging findings; however, in some cases it may be difficult to differentiate a microcystic adenoma from a mucinous cystic neoplasm.29
Mucinous cystic neoplasms are considered to be malignant or of low malignant potential until proven otherwise.31 Mucinous cystic neoplasms classically have fewer than six cysts, and some of the cysts measure greater than 2 cm. Tumors may have thick septa, nodules, and enhancing solid tissue components.
Mucinous pancreatic neoplasms are notorious for being confused with pancreatic pseudocysts,32 and imaging differentiation can be difficult, if not impossible. I agree with Scott and colleagues who have suggested either a "watch and wait" strategy or surgical cyst removal in equivocal cases.
Pancreatitis. CT is used most often for detecting pancreatitis and its associated complications, but MR may offer some selective advantages. First, in patients with elevated serum creatinine (which may in part be secondary to the pancreatitis) one may want to avoid the nephrotoxic effects of iodinated contrast agents. Gadolinium chelates are not nephrotoxic at the doses used for clinical MR scanning33 and can be used in evaluating patients with pancreatitis, especially for pancreatic necrosis.
Second, it has been suggested that imaging studies cannot detect hemorrhagic pancreatitis and surgical confirmation is necessary. But MR is sensitive in detecting methemoglobin within subacute hemorrhage and may thus noninvasively establish the complication of intrapancreatic or peripancreatic hemorrhage.
Finally, MR offers two potential advantages over CT in the evaluation of pancreatic pseudocysts. MRCP techniques can readily detect pseudocysts and determine whether they communicate with the pancreatic duct. Knowledge of any such communications will help guide therapy because pseudocysts that have direct duct communications may decompress spontaneously. MR can more accurately distinguish between complicated and non-complicated pseudocysts.
Dr. Siegelman is an associate professor of radiology at the University of Pennsylvania in Philadelphia.
- Winston CB, Mitchell DG, Outwater EK, Ehrlich SM. Pancreatic signal intensity on T1-weighted fat saturation MR images: clinical correlation. J Magn Reson Imaging 1995;5:267-271.
- Siegelman ES, Outwater EK, Vinitski S, Mitchell DG. Fat suppression by saturation/opposed-phase hybrid technique: spin echo versus gradient echo imaging. Magn Reson Imaging 1995;13:545-548.
- Mitchell DG, Crovello M, Matteucci T, et al. Benign adrenocortical masses: diagnosis with chemical shift MR imaging (see comments). Radiology 1992;185:345-351.
- Outwater EK, Siegelman ES, Radecki PD, et al. Distinction between benign and malignant adrenal masses: value of T1-weighted chemical-shift MR imaging. AJR 1995;165:579-583.
- Jacobs JE, Coleman BG, Arger PH, Langer JE. Pancreatic sparing of focal fatty infiltration. Radiology 1994;190:437-439.
- Isserow JA, Siegelman ES, Mammone J. Focal fatty infiltration of the pancreas: MR characterization with chemical shift imaging. AJR 1999;173:1263-1265.
- Outwater EK, Bhatia M, Siegelman ES, et al. Lipid in renal clear cell carcinoma: detection on opposed-phase gradient-echo MR images (see comments). Radiology 1997;205:103-107.
- Carucci LR, Siegelman ES, Feldman MD. Pancreatic metastasis from clear cell renal carcinoma: diagnosis with chemical shift MRI. JCAT 1999;23:934-936.
- Lu JP, Hayashi K. Selective iron deposition in pancreatic islet B cells of transfusional iron-overloaded autopsy cases. Pathol Int 1994;44:194-199.
- Lomas DJ, Gimson A. Magnetic resonance cholangiopancreatography. Hospital Medicine (London) 2000;61:395-399.
- Jara H, Barish MA. MR cholangiopancreatography techniques. Semin Ultrasound, CT, MR 1999;20:281-293.
- Barish MA, Yucel EK, Ferrucci JT. Magnetic resonance cholangiopancreatography. NEJM 1999;341:258-264.
- Barish MA, Soto JA. MR cholangiopancreatography: techniques and clinical applications. AJR 1997;169:1295-1303.
- Pavone P, Laghi A, Catalano C, et al. MRI of the biliary and pancreatic ducts. Eur Radiol 1999;9:1513-1522.
- Jara H, Barish MA, Yucel EK, et al. MR hydrography: theory and practice of static fluid imaging. AJR 1998;170:873-882.
- Takehara Y. MR pancreatography. Sem Ultrasound, CT, MR 1999;20:324-339.
- Fulcher AS, Turner MA. MR pancreatography: a useful tool for evaluating pancreatic disorders. Radiographics 1999;19:5-24; discussion 41-24; quiz 148-149.
- Bret PM, Reinhold C, Taourel P, et al. Pancreas divisum: evaluation with MR cholangiopancreatography. Radiology 1996;199:99-103.
- Matos C, Nicaise N, Metens T, et al. Secretin-enhanced MR pancreatography. Semin Ultrasound, CT, MR 1999;20:340-351.
- Manfredi R, Costamagna G, Brizi MG, et al. Pancreas divisum and "Santorinicele": diagnosis with dynamic MR cholangiopancreatography with secretin stimulation. Radiology 2000;217:403-408.
- Varghese JC, Liddell RP, Farrell MA, et al. The diagnostic accuracy of magnetic resonance cholangiopancreatography and ultrasound compared with direct cholangiography in the detection of choledocholithiasis (see comments). Clin Radiol 1999;54:604-614.
- Kim MJ, Mitchell DG, Ito K, Outwater EK. Biliary dilatation: differentiation of benign from malignant causes-value of adding conventional MR imaging to MR cholangiopancreatography. Radiology 2000;214:173-181.
- Pavone P, Laghi A, Passariello R. MR cholangiopancreatography in malignant biliary obstruction. Semin Ultrasound, CT, MR 1999;20:317-323.
- Catalano C, Pavone P, Laghi A, et al. Pancreatic adenocarcinoma: combination of MR imaging, MR angiography and MR cholangiopancreatography for the diagnosis and assessment of resectability. Eur Radiol 1998;8:428-434.
- Buetow PC, Parrino TV, Buck JL, et al. Islet cell tumors of the pancreas: pathologic-imaging correlation among size, necrosis and cysts, calcification, malignant behavior, and functional status. AJR 1995;165:1175-1179.
- Thoeni RF, Mueller-Lisse UG, Chan R, et al. Detection of small, functional islet cell tumors in the pancreas: selection of MR imaging sequences for optimal sensitivity. Radiology 2000;214:483-490.
- Semelka RC, Custodio CM, Cem Balci N, Woosley JT. Neuroendocrine tumors of the pancreas: spectrum of appearances on MRI. J Magn Reson Imaging 2000;11:141-148.
- Buck JL, Hayes WS. From the archives of the AFIP. Microcystic adenoma of the pancreas. Radiographics 1990;10:313-322.
- Procacci C, Graziani R, Bicego E, et al. Serous cystadenoma of the pancreas: report of 30 cases with emphasis on the imaging findings. JCAT 1997;21:373-382.
- Iselin CE, Meyer P, Hauser H, et al. Computed tomography and fine-needle aspiration cytology for preoperative evaluation of cystic tumours of the pancreas. Br J Surg 1993;80:1166-1169.
- Buetow PC, Rao P, Thompson LD. From the archives of the AFIP. Mucinous cystic neoplasms of the pancreas: radiologic-pathologic correlation. Radiographics 1998;18:433-449.
- Scott J, Martin I, Redhead D, et al. Mucinous cystic neoplasms of the pancreas: imaging features and diagnostic difficulties. Clin Radiol 2000;55:187-192.
- Rofsky NM, Weinreb JC, Bosniak MA, et al. Renal lesion characterization with gadolinium-enhanced MR imaging: efficacy and safety in patients with renal insufficiency (see comments). Radiology 1991;180:85-89.
- Morgan DE, Baron TH, Smith JK, et al. Pancreatic fluid collections prior to intervention: evaluation with MR imaging compared with CT and US. Radiology 1997;203:773-778.