IntroductionRecent improvements in radiology have occurred, both in computerized tomography (CT) and magnetic resonance imaging (MRI), which greatly facilitate the diagnosis of pancreatic and biliary disease. Helical (spiral) CT with intravenous contrast allows a rapid analysis of the pancreatic contour, the arteries around the pancreas, and the portal venous system. New MRI systems, called magnetic resonance cholangiopancreatography (MRCP), also can provide parenchymal and vascular imaging, and in addition can rapidly produce detailed images of the pancreatic and biliary ductal systems.
Gastroenterologists usually assume a central role in the diagnosis of pancreaticobiliary disease, and have the tools of endoscopic retrograde cholangiopancreatography (ERCP) and endoscopic ultrasonography (EUS). Choosing among these tests and procedures in the evaluation and management of patients with benign and malignant diseases of the pancreas and bile ducts remains difficult and controversial, and was the subject of several presentations at the 7th United European Gastroenterology Week, held November 13-17, 1999, in Rome, Italy.
Radiology and Endoscopy: Competitive or Complementary?The role of EUS in the diagnosis of pancreaticobiliary disease in the face of improved CT, MRI, and MRCP was the focus of a roundtable symposium chaired by T. Rösch on Tuesday, November 16, 1999.
Dual-Phase Spiral CT
Dr. B. Marincek, of Zurich, Switzerland, led off the speakers with a strong emphasis on "go with CT alone" for pancreatic cancer staging. He declared that dual-phase spiral CT has become the "gold standard" and is the initial diagnostic test of choice for the clinical suspicion of pancreatic cancer. This method provides the ability to detect small primary pancreatic lesions, and concomitantly obtain staging information on local cancer extension to arterial and portal venous systems, assess regional lymph nodes, and diagnose distant metastases in the liver.
MRI/MRCP "All in One"
There have been recent advances in MRI, described by Dr. Marincek, that may result in this imaging modality eventually replacing CT as the primary method for examining the pancreas and biliary tree. Ultrafast MR can provide a noninvasive "all-in-one" imaging approach. This consists of breath-hold cross-sectional body images (both T1- and T2-weighted), MRCP for views of the extrahepatic biliary system and pancreatic ducts, and three-dimensional MR angiography (3D-MRA) for detailed images of the abdominal arterial and portal venous systems.
The role for endoscopic ultrasound in the evaluation of pancreaticobiliary disease was discussed at length. Dr. M. Giovannini of Marseilles, France, presented results strongly supporting the value of endosonography in patients with a solid pancreatic tumor. He described his results in 174 patients who had endoscopic ultrasound-guided biopsy (EUSGB). The ability of EUSGB to detect and obtain tissue from small pancreatic masses was a consistent advantage. In 39 patients, the pancreatic mass was < 20 mm in diameter. Sensitivity of EUSGB was 87.2%, specificity 100%, and accuracy 87.9%. There were no complications. Most cases were adenocarcinomas, but other diagnoses included neuroendocrine tumors, metastatic cancers, lymphoma, chronic pancreatitis, and a pancreatic abscess. Management was altered in 63.2% of patients on the basis of the EUSGB findings.
Dr. M. Büchler, of Bern, Switzerland, argued that because of the theoretical risk of tumor seeding, he preferred not to perform EUSGB on operable patients. Dr. Giovannini countered that EUSGB tumor seeding has not been described in the literature, and that EUSGB was thought to be of increasing value with the improvement in radiochemotherapy for pancreatic adenocarcinomas.
Endosonography for Tumor Staging
Dr. L. Buscail, of Toulouse, France, emphasized the importance of endosonography as a method for the staging of cancer throughout the gastrointestinal tract. The overall accuracy of endosonography for staging the depth of invasion of primary gastrointestinal cancers (T) was 78%-85%, and for staging regional lymph node metastases (N) was 73%-79%. These results were consistently better than those obtained by CT.
The development of high-frequency (12 and 20 MHz) miniprobes has greatly improved the capacity of endosonography staging of small, early cancer and the staging of stenotic tumors and tumors of the bile duct and ampulla of Vater. The staging information obtained has been of great value in deciding on initial management of cancers in the esophagus, stomach, and rectum, and has also been of value in gastric lymphoma and in planning the management of submucosal tumors.
Best Test for Bile Duct Stones
The final segments of the roundtable on pancreaticobiliary imaging were in the form of a debate on the best diagnostic approach to the patient with possible bile duct stones. Transabdominal ultrasonography (US) was the first diagnostic test propounded by Dr. L. Gandolfi, of Bologna, Italy, who argued further that in many cases the value of US was underestimated. Dr. J. Devière, of Brussels, Belgium, on the other hand, considered endosonography to be an indispensable test in tumor staging and in screening selected patients with suspected pancreaticobiliary disease.
Endoscopic Ultrasonography for Common Duct Stones
Dr. L. Aabakken, of Oslo, Norway, picked up on this theme, presenting his analysis of why endoscopic ultrasonography was useful in the armamentarium of tests available for the diagnosis of common bile duct stones. He noted that the issue is of particular importance in the evaluation of patients before laparoscopic cholecystectomy.
Common duct stones are a major source of morbidity in patients with symptomatic gallstone disease. Laparoscopic surgeons are developing methods for common bile duct clearance during the surgical procedure, but the requisite expertise is not widely available. Preoperative diagnosis of common duct stone is highly desirable, but among the many methods available, none has emerged as clearly superior. Factors involved in making the decision on which tests to use include the degree of suspicion, the availability of the modality, and the local expertise.
For the diagnosis of common bile duct stones, endoscopic ultrasonography has a sensitivity and specificity well above 90%, comparing favorably with MRCP and even ERCP. Endoscopic ultrasound can pick up small stones missed on ERCP, and has a much lower complication rate than ERCP. The rate of clinical pancreatitis following ERCP is in the range of 3%. Unlike ERCP, however, endoscopic ultrasonography has no therapeutic capability for removal of common duct stones. Therefore, endoscopic ultrasonography is used as a screening test for ERCP in patients with a clinically medium risk of duct stones, while high-risk patients are generally assigned directly to ERCP. Defining the risk of common duct stones depends on clinical presentation, duct diameter on transabdominal US, and blood tests of liver function. Endoscopic ultrasound appears to be cost-effective when the risk of common duct stones is below 55%.
Magnetic Resonance Cholangiography for Common Duct Stones
Dr. C. Matos, of Brussels, Belgium, considered that MR cholangiography (MRC), the newest modality for evaluation of the biliary tree, was the best test to detect common bile duct stones prior to laparoscopic cholecystectomy. This noninvasive test does, however, require state-of-the-art MR equipment, including high field strength, high power gradient system, and phased-array coils. Furthermore, the system is, like endoscopic ultrasound, operator-dependent to a variable degree. Reported sensitivity, specificity, and positive and negative predictive values have been in the range of 95%-100% -- comparable to endoscopic ultrasonography and ERCP, but not requiring endoscopy.
Other advantages of MRC before laparoscopic cholecystectomy include the high accuracy of MRC in demonstrating anatomic variants of the biliary tree as a guide to surgery, and that the morphological information is widely understandable. Dr. Matos advised the use of MRC in patients with low to moderate risk of common bile duct stones.
In a related paper, A. Tanner and colleagues, of Stockton-on-Tees, United Kingdom, analyzed 1078 consecutive ERCP examinations to calculate the effect of MRC on the future ERCP workload. If MRC was used in patients with gallstones with abdominal pain but without jaundice (with or without abnormal liver function tests) as well as in those with present or past acute pancreatitis, there would be a 33% reduction in ERCP workload, with 20 serious complications avoided.
There were several poster presentations evaluating the role of MRCP and ERCP in various clinical settings. Dr. T. Rösch and colleagues reported on a pilot trial of the European Society of Gastrointestinal Endoscopy (ESGE) prospectively comparing MRCP and ERCP in pancreaticobiliary disorders. In this initial small experience with 59 patients, the positive predictive value for MRCP was good, but of interest, the negative predictive value was better for MRP (magnetic resonance pancreatography) than for MRC. Larger clinical trials seemed to be indicated for confirmation.
Dr. R. Dani and colleagues, of Belo Horizonte, Brazil, reported their prospective evaluation of 24 patients with cystic lesions of the pancreas using MRCP. They found MRCP to be a useful, reliable diagnostic modality in all patients. MR imaging allowed the analysis of many variables, including cysts, ducts, parenchyma, vessels, and cyst relation to neighboring organs.
The use of secretin stimulation of the pancreas in association with MRP for dynamic MRP was assessed by Dr. R. Manfredi, of Rome, Italy, in patients with chronic pancreatitis. In 84 patients with suspected chronic pancreatitis and 34 with known disease, he found that using MRP before and after secretin aided in defining pancreatic ductal anatomy and defining exocrine function. In 7 patients, unsuspected pancreas divisum was detected. In another study using secretin stimulation for dynamic MRP, Dr. C. Matos could not demonstrate an abnormal functional response (persistent ductal dilation and decreased duodenal filling) in pancreas divisum patients compared with controls.
Combining ERCP with catheter size, high-frequency ultrasound probes allow detailed examination of the wall of the bile and pancreatic ducts and immediate surroundings. B. Napoléon and colleagues, of Lyons, France, investigated two new Olympus miniprobes designed primarily for pancreaticobiliary examination. Both probes have 20-MHz transducers that rotate to produce a 360° image. The UM-S20-20R (SP) has a tip size of only 1.7 mm. The UM-G20-29R (SG) has a larger, 2-mm tip, but has a channel for a standard 0.035-in guide wire to facilitate ductal insertion.
Napoléon and colleagues reported on the intraductal ultrasound (IDUS) examination of 15 patients, including 13 biliary and 2 pancreatic insertions. In 3 cases, the probes were passed through 12-F plastic drainage stents, which did not interfere with imaging. One patient with a Klatskin tumor had a subsequent surgical resection, with a confirmation of staging accuracy. The smaller-tip probe was preferable for percutaneous insertion. The wire-guided probe was superior for transpapillary insertion and for bypassing strictures.
J.F. Riemann and colleagues, of Ludwigshafen, Germany, presented data reviewing the analysis of pancreaticobiliary stenoses using IDUS and intraductal endoscopy and cytology. Differentiation of benign from malignant strictures was accomplished with an accuracy of more than 90% if both IDUS and intraductal cytology were used. In malignant strictures, IDUS provided more accurate staging than standard endosonography or MRI.
Another new imaging modality recently applied to the diagnosis and staging of pancreatic adenocarcinoma is the PET (positron emission tomography) scan. H. Mertz and colleagues, of Nashville, Tenn, prospectively compared PET scan (18FDG-PET) versus standard endoscopic ultrasound versus EUS-FNA (fine-needle aspiration) in 35 patients. EUS was the most sensitive for tumor detection, significantly better than helical CT (Table I), and was best for local tumor staging, including vascular invasion. PET scan was also significantly better than CT for cancer detection, and best for detecting unsuspected distant metastases. All three modalities gave a false-positive impression in a patient with an acute pancreatic phlegmon.
Table I. Detection of Pancreatic Adenocarcinoma
Helical CT PET EUS 53% 87% (P < .05) 93% (P < .05)
- Marincek B. Pancreatic cancer staging: Go with CT alone? Gut. 1999;45(Suppl V):A20. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Giovannini M, Monges G. Diagnostic and therapeutic value of the endoscopic ultrasound guided biopsy in patients with a solid pancreatic tumor. Endoscopy. 1999;31(Suppl 1):E10. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Buscail L. Tumor staging in the GI tract--we need EUS. Endoscopy. 1999;31(Suppl 1):E10. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Aabakken L. What is the best test for bile duct stones--EUS? Endoscopy. 1999;31(Suppl 1):E10-11. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Matos C. What is the best test for CBD stones? MRCP. Endoscopy. 1999;31(Suppl 1):E11. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Tanner A, Reeves H, Dwarakanath A, Tait N. Potential impact of high quality MRI of the biliary tree on ERCP workload. Endoscopy. 1999;31(Suppl 1):E23. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Rösch T, Helmberger H, Hellerhoff K, et al. A prospective comparison of MRCP and ERCP in pancreatobiliary disorders--pilot trial of the ESGE research group. Endoscopy. 1999;31(Suppl 1):E57. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Dani R, Cundari AMMV, Noguiera CED, Reis GMF, Silva LD. Magnetic resonance cholangiopancreatography in cystic lesions of the pancreas. Endoscopy. 1999;31(Suppl 1):E73-74. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Manfredi R. Dynamic magnetic resonance pancreatography after secretin stimulation: severe chronic pancreatitis vs suspected pancreatic disease. Endoscopy. 1999;31(Suppl 1):E103-104. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Matos C, Nicaise N, Devière J, Delhaye M, Cremer M. Pancreas divisum: evaluation with secretin enhanced MR pancreatography. Endoscopy. 1999;31(Suppl 1):E108. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Napoléon B, Albis R, Ponchon T, et al. Technical evaluation of the new biliary miniprobes for ultrasonography. Endoscopy. 1999;31(Suppl 1):E6. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Riemann JF, Schilling D, Jakobs R. Benign or malignant? Brushes, forceps, miniendoscopes, and miniprobes in the biliary tract and pancreatic ducts. Endoscopy. 1999;31(Suppl 1):E6. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.
- Mertz H, Sechopolous P, Leach S. PET scan, endoscopic ultrasound, and CT scan for evaluation of pancreatic adenocarcinoma. Endoscopy. 1999;31(Suppl 1):E7. Program and abstracts of the 7th United European Gastroenterology Week; November 13-17, 1999; Rome, Italy.