Hepatobiliary Imaging
Helical CT Cholangiography with Three-Dimensional Volume Rendering Using an Oral Biliary Contrast Agent
Feasibility of a Novel Technique
Audio Available | 
Share
OBJECTIVE. We evaluated the feasibility and image quality of a new noninvasive biliary imaging technique: helical CT cholangiography with three-dimensional volume rendering using an oral biliary contrast agent.
SUBJECTS AND METHODS. Nineteen subjects including five healthy volunteers and 14 patients underwent helical CT cholangiography. Subjects ingested 6.0 g of iopanoic acid 6-10 hr before undergoing imaging. Axial data were used to construct three-dimensional volume-rendered cholangiograms. Two radiologists, an endoscopist, and a laparoscopic surgeon reviewed the images and evaluated overall image quality. In the 14 patients, findings from CT cholangiography were compared with those from ERCP, surgery, and intraoperative cholangiography.
RESULTS. All segments of the biliary tree were opacified in all volunteers except one, in whom the intrahepatic ducts were not opacified. Image quality was good to excellent in all volunteers. Anomalous cystic duct insertions were seen in two volunteers. Opacification of the biliary tree was rated as acceptable to excellent in nine patients and suboptimal in five. In five patients with good or excellent opacification, the biliary anatomy correlated with findings on intraoperative cholangiography or ERCP. CT cholangiography revealed additional conditions (gallbladder varices and acute pancreatitis) and variant anatomy in three patients.
CONCLUSION. Results of this pilot project suggest that obtaining CT cholangiograms using an oral biliary contrast agent is a feasible, noninvasive method for revealing biliary anatomy. However, visualization of the biliary tree was suboptimal in 36% of the patients, which represents a limitation of this technique.
| Introduction |   |
|---|
The gold standard for biliary imaging is direct contrast opacification by either ERCP or percutaneous transhepatic cholangiography (PTC). Although these techniques provide excellent delineation of biliary anatomy and pathology, both are physician-intensive and costly. Furthermore, because these procedures are invasive, they are associated with risks and complications. PTC has a complication rate of 3.4% [1, 2], whereas ERCP has a complication rate that ranges from 0.5% to 5% [2, 3].
Currently, MR cholangiography is a popular noninvasive alternative to ERCP or PTC in many practices and academic centers. This technique has been shown to be both sensitive and specific for visualization of various conditions of the biliary tract [4, 5, 6]. Although MR cholangiography is safe, this technique is contraindicated in patients with aneurysm clips or cardiac pacemakers. In addition, MR cholangiography may not be suitable for patients with claustrophobia or those with multiple metallic clips, which may cause artifacts, in the porta hepatis. For the patients who cannot undergo MR cholangiography, there are few alternatives to the invasive techniques, thus emphasizing the need for additional noninvasive techniques.
Helical CT allows imaging of a volume of tissue during a single breath-hold. Axial CT data may be reconstructed into two-dimensional multiplanar or three-dimensional (3D) volume-rendered images using workstations and image-rendering software. This CT technology combined with the administration of IV cholangiographic contrast agents produces diagnostic images of the biliary tract, particularly of the extrahepatic ducts. The main limitation of these contrast agents is that the rate of allergic reactions and of renal or hepatic toxicity (or both) is relatively high [3].
Oral cholangiographic contrast agents are a potential alternative to IV contrast agents. Indeed, before the widespread use of sonography to image the biliary tree, oral—contrast-enhanced cholecystography was widely used to evaluate the gallbladder. Unlike IV contrast agents, these contrast agents have few side effects. To our knowledge, there have been few studies or reports in which oral biliary contrast agents have been used. In the first study in which an oral contrast agent was used, the visualization of the common bile duct within the pancreatic head was evaluated [7]. CT cholangiography using oral contrast agents to assess choledochal variants in pediatric patients has also been described [8, 9]. Recently, this technique was used to evaluate the visualization of choledocholithiasis [10]. It has not yet been established whether these oral contrast agents can reliably opacify the biliary system using helical CT. The purpose of this study is to investigate the feasibility of obtaining clinically useful CT cholangiograms with an oral biliary contrast agent.
| Subjects and Methods | |
|---|
The study population consisted of five healthy volunteers and 14 patients in whom hepatobiliary or pancreatic disorders were suspected. The control subjects were three women and two men who ranged in age from 28 to 42 years (mean, 35 years). None had a history of hepatic or biliary disease.
The patients consisted of 12 women and two men who ranged in age from 21 to 57 years (mean, 41 years). Twelve patients had symptomatic cholelithiasis, one had chronic pancreatitis, and one had a biliary cystadenocarcinoma. All patients were scheduled to undergo ERCP or laparoscopic surgery with intraoperative cholangiography within a mean of 3 weeks of CT cholangiography (range, 1-20 weeks).
Written informed consent was obtained after explanation of the procedure. The study was performed according to clinical practice guidelines and with the approval of our institutional review board. Exclusion criteria included age of less than 18 years, pregnancy or lactation, known allergy to iodinated contrast material, renal insufficiency (creatinine level, >2.0 mg/dl), and severe hepatic dysfunction and hyperbilirubinemia (creatinine level, >4.0 mg/dl). The control subjects were examined before the patients. Each subject ingested 6.0 g of iopanoic acid (Telepaque; Winthrop Pharmaceuticals, New York, NY) after a low-fat meal the night before CT. The CT scans were obtained 6-10 hr (mean, 9 hr) after ingestion. All subjects underwent helical CT of the biliary tree on a CT/i unit (General Electric Medical Systems, Milwaukee, WI) using the following parameters: 3-mm collimation, 1:1 pitch, 180° interpolation, 140 kVp, 170-220 mA, 512 × 512 matrix, and an 18- to 20-cm field of view centered at the porta hepatis. Images were acquired during an exposure of 20-40 sec for a single breath-hold from the bases of the lung to the tip of the right hepatic lobe.
After scan acquisition, a cholecystokinin analogue, sincalide (Kinevac; Squibb, New Brunswick, NJ), was administered intravenously to each control subject. Sincalide causes contraction of the gallbladder and stimulates biliary flow into the duodenum while relaxing Oddi's sphincter. Sincalide was administered over 45 min for a total dose of 20 μg/kg in an attempt to improve biliary opacification. A second CT scan was then obtained through the biliary tree. Images obtained before and after administration of sincalide were reconstructed at 1-mm intervals and were transferred to a remote workstation (Advantage Windows; General Electric Medical Systems).
Postprocessing included performing curved multiplanar reformations and 3D reconstructions that consisted of maximal intensity projections and 3D volume-rendering displays on a workstation (Vitrea; Vital Images, Minneapolis, MN). Curved multiplanar reformations were created using a manual tracing algorithm. Markers placed along the cystic duct and common bile duct created a projection based on connectivity. Maximal intensity projections and 3D volume renderings were produced. Unwanted structures were electronically eliminated by trimming and altering the window and level parameters. The maximum-intensity-projection and 3D volume-rendered images preferentially showed the voxels with the highest CT attenuation numbers. Both reconstructions depicted the outer surfaces of the biliary tree and gallbladder and could be rotated in all directions. Postprocessing time for all displays ranged from 10 to 20 min. Postprocessing was performed by two abdominal radiologists, one of whom participated in the final review with a third abdominal radiologist.
From February to October 1998, CT cholangiograms were obtained in 14 consecutive patients who were scheduled to undergo ERCP or laparoscopic surgery with intraoperative cholangiography. The patients underwent the same procedure as the control subjects except that sincalide was not administered because preliminary analysis of images of the control subjects showed no improvement in biliary opacification after sincalide administration. During previous sonographic examinations, cholelithiasis without evidence of acute cholecystitis had been documented in 12 of the 14 patients. Choledocholithiasis was not revealed on any sonographic examination. One patient with a biliary cystadenocarcinoma had undergone MR cholangiography but it had been unsuccessful. No other patient had undergone prior MR cholangiography. One patient with chronic pancreatitis had never undergone imaging.
CT cholangiograms were reviewed at a workstation and scored by consensus decision of two abdominal radiologists. At the time of review, findings from previous imaging studies were not available. On the workstation, reviewers could quickly view the data set using various rendering techniques. The axial source data, multiplanar reformations, maximum intensity projections, and 3D volume-rendered images were viewed during a single session. The workstation allowed reviewers to rapidly alternate from one rendering technique to another. The reviewers used a grading scale to assess the conspicuity of the biliary tree segments and the overall quality of the images: grade 0, duct not visualized; grade 1, duct barely visualized; grade 2, marginal opacification; grade 3, good opacification; and grade 4, excellent opacification. The following structures were reviewed: left and right intrahepatic ducts, cystic duct, common hepatic duct, common bile duct, and gallbladder. Anatomic variants of the biliary tree were recorded with an emphasis on variations that would place patients at risk of injury during laparoscopic cholecystectomy. The presence of gallstones and common bile duct stones was also noted.
For each patient, representative multiplanar reformations and 3D volume-rendered images that best represented the anatomy were selected. These images were filmed to hard copy using a paper printer (Codonics; Imagemax Media, Middleburg Heights, OH). The hard copy images were reviewed with an endoscopist interested in cholangiography and a laparoscopic surgeon.
Overall image quality and the presence of abnormalities or anatomic variants were assessed. In the 12 patients who underwent cholecystectomy, the anatomy of the extrahepatic and cystic ducts was compared with findings at surgery and intraoperative cholangiography. In particular, the cystic duct insertion relative to the common hepatic duct was assessed.
For the two other patients, results were compared with findings on ERCP. Specifically, the endoscopist compared findings from CT cholangiography with those from ERCP for the patients who underwent endoscopy. Similarly, the surgeon compared the CT cholangiograms with the intraoperative findings at laparoscopic cholecystectomy and the intraoperative cholangiograms.
| Results | |
|---|
The oral contrast agent and imaging protocol were well tolerated. No subjects experienced a major or minor reaction to the oral iopanoic acid. With appropriate instruction, most subjects (18/19) were able to maintain a single breath-hold without difficulty.
In all the volunteers, visualization of the cystic duct, common hepatic duct, common bile duct, and gallbladder was good (grade 3) to excellent (grade 4). The central intrahepatic biliary tree was not visualized in one volunteer. After the administration of sincalide, biliary visualization was unchanged in three of the five volunteers and was diminished in two; in addition, significant contraction of the gallbladder was seen in two patients. CT cholangiography revealed a few anatomic variations of the cystic duct in the volunteer group: a low insertion of the cystic duct (Fig. 1) and a posterior insertion of the cystic duct (Fig. 2). The insertion was considered low when the cystic duct joined the lower third of the total length of the common hepatic and common bile ducts. The insertion was considered posterior when the cystic duct joined the posterior aspect of the common hepatic duct.
 View larger version (125K) | Fig. 1. —32-year-old healthy male volunteer. Three-dimensional volume-rendered CT cholangiogram shows low cystic duct (arrows) insertion. Image is in left posterior oblique orientation. |
 View larger version (129K) | Fig. 2. —42-year-old healthy female volunteer. Lateral projection of three-dimensional volume-rendered CT cholangiogram shows posterior cystic duct insertion (arrow). A=anterior. |
In 11 of the 14 patients, visualization of the common hepatic duct and common bile duct was acceptable (n = 3), good (n = 7), or excellent (n = 1). CT cholangiograms showed the central intrahepatic ducts in nine patients. In nine patients, opacification of the gallbladder was acceptable (grade 2) to excellent (grade 4). Two of the patients had undergone previous cholecystectomy. Visualization of the cystic duct was acceptable (grade 2) to excellent (grade 4) in only six of 14 patients. Variant cystic duct anatomy was diagnosed using CT cholangiograms and was confirmed at surgery in three patients: in two patients, posterior insertion of the cystic duct was detected and in one patient, a helical cystic duct was found. A helical cystic duct is a cystic duct that joins the medial border (left side) of the common duct (Fig. 3A).
 View larger version (132K) | Fig. 3A. —48-year-old woman with symptomatic cholelithiasis. Three-dimensional volume-rendered CT cholangiogram shows helical or left cystic duct insertion (arrows). Cystic duct courses posteriorly before joining left side of common hepatic duct. |
Of the five volunteers, neither cholelithiasis nor choledocholithiasis was diagnosed using CT cholangiography and no additional examinations were performed to confirm or disprove these findings. Of the 12 patients who underwent laparoscopic surgery, cholelithiasis was present in all patients. CT cholangiograms revealed cholelithiasis (Fig. 3B) in all 12 patients. Choledocholithiasis was present at surgery in three patients and was documented in the procedure note. The size of the stones was not recorded intraoperatively. This diagnosis was not made using CT cholangiography in any of the cases. There were two unexpected diagnoses revealed by CT cholangiography: gallbladder varices of unknown etiology (Fig. 4) and acute pancreatitis. These diagnoses were made using the axial source images and multiplanar reformations. Because the presence of gallbladder varices was detected, a laparoscopic procedure was converted into an open cholecystectomy.
 View larger version (130K) | Fig. 3B. —48-year-old woman with symptomatic cholelithiasis. Three-dimensional volume-rendered CT cholangiogram reveals gallstones. |
 View larger version (195K) | Fig. 4. —48-year-old woman with symptomatic cholelithiasis. Source image shows varices (arrows) at gallbladder neck and dependent gallstones. |
The mean score for each segment of the biliary tree was calculated for the volunteer and patient groups. The mean scores for the control subjects (range, 2.8-3.8) were consistently higher than the those for the patients (range, 1.6-2.4). In addition, the mean score for complete biliary visualization was determined. A mean score of 2.0 was considered a diagnostic study. All five control subjects had diagnostic studies (range, 3.0-4.0), whereas nine of the 14 patients had diagnostic studies (range, 2.0-3.8) and five had nondiagnostic studies.
In two patients, the biliary tree was not opacified; both of these studies were considered failures. In both patients, high-attenuation contrast material was present in the bowel, indicating the contrast agents had been ingested but were either not absorbed by the enterohepatic circulation or not excreted into the biliary tree. One of these patients had chronic pancreatitis and was subsequently imaged using ERCP. The other patient had a biliary cystadenocarcinoma and underwent ERCP and then percutaneous cholangiography. Both of the patients in whom CT cholangiography failed had previously undergone cholecystectomy.
CT cholangiography revealed limited detail of the anatomy of the intrahepatic biliary tree. Visualization of the third- or fourth-order intrahepatic biliary radicles was not possible in any case. The course of the common hepatic and common bile ducts was well depicted in most patients (11/14); however, the distal course of the common bile duct as it entered the duodenum was not clearly seen with diagnostic CT cholangiograms in two patients. In most patients (8/14) opacification of the cystic duct was unacceptable (score, <2). In three of these patients, surgical findings revealed a hydropic or severely scarred cystic duct through which an intraoperative cholangiogram catheter could not be passed. In two other patients, CT cholangiographic examinations failed. Nonvisualization or poor visualization of the gallbladder was associated with acute cholecystitis in only one control subject (1/5). CT cholangiography failed in two other subjects who had a previous cholecystectomy (2/5). Other imaging limitations included pseudostrictures (Fig. 5), which were notable during examinations of two volunteers without known biliary disease; thus, these pseudostrictures likely represented an artifact of the technique. Presumably, the pseudostrictures were caused by underdistention of the biliary tree or incomplete opacification of the bile. In addition, the inability to suspend respiration or motion caused segmentation artifact of the gallbladder and biliary tree in one patient (Fig. 6).
 View larger version (105K) | Fig. 5. —36-year-old healthy male volunteer. Three-dimensional volume-rendered CT cholangiogram shows pseudostrictures (arrows) involving common hepatic and common bile ducts. Pseudostrictures are probably caused by underdistention or incomplete opacification of bile ducts. |
 View larger version (104K) | Fig. 6. —48-year-old woman with symptomatic cholelithiasis. Three-dimensional volume-rendered CT cholangiogram shows segmentation of biliary tract and gallbladder; segmented appearance results from respiratory motion. Irregular appearance of gallbladder fundus is caused by motion artifacts. Gallstones cannot be seen on this image. |
| Discussion | |
|---|
Currently, there are several clinical scenarios for which helical CT cholangiography is applicable. Laparoscopic cholecystectomy has become the procedure of choice for removal of the gallbladder because this procedure is associated with a shorter hospital stay, lower hospital costs, and improved patient cosmesis [11, 12, 13].
Because initial visualization of Calot's triangle at laparoscopy is limited, knowledge of the precise relationship of the cystic duct to the hepatic and common bile ducts may decrease laparoscopic surgical dissection time. Preoperative knowledge of the biliary anatomy may help decrease the rate of complications [14, 15, 16, 17] because ductal anomalies occur in as many as 21.9% of patients undergoing cholecystectomy [18]. CT cholangiography was beneficial in our study, revealing variant cystic duct anatomy, which was confirmed at surgery, in three patients. In addition, CT cholangiograms revealed unexpected findings, such as gallbladder varices, that changed a laparoscopic procedure to an open approach.
Helical CT cholangiography may be useful in patients with a nonobstructive abnormality and thus not requiring therapeutic intervention by ERCP. ERCP is not without risks—including pancreatitis, cholangitis, and pancreatic sepsis and perforation; the rate of complications is reported to range from 0.6% to 5.0% [19]. There are a few scenarios in which ERCP may not be feasible including an ampulla located within a duodenal diverticulum or prior choledochojejunostomy. In such situations, CT cholangiography may be a useful alternative to ERCP. PTC is another invasive technique for opacifying the biliary tree. The overall complication rate is 3.5% and risks include sepsis, bile leak, and hemorrhage [1]. CT cholangiography may be useful in patients with contraindications to PTC including those with bleeding disorders or ascites.
MR cholangiography, another noninvasive imaging technique, is an effective and accurate technique but has a few limitations including cost, operator dependence, and patient tolerance. Although early experience with MR cholangiography revealed that normal-caliber ducts may be poorly visualized in as many as 50% of patients [20], more recent series have shown excellent visualization in most patients [4, 5]. Despite improved MR imaging technology, several contraindications including aneurysm clips, pacemakers, and claustrophobia to MR imaging persist. In this small subset of patients unable to undergo MR imaging, CT cholangiography may be a reasonable noninvasive alternative.
CT cholangiography was first described in 1982 by Greenberg et al. [7] who succeeded in opacifying the distal common bile duct using an oral hepatotropic contrast agent. Greenberg et al. found it useful to opacify the distal common bile duct to differentiate the pancreatic head from the duodenum. With the advent of helical CT, several investigators experimented with the innovative approach of creating 3D CT cholangiograms [21, 22, 23, 24, 25, 26, 27, 28]. Helical CT cholangiography with IV biliary contrast agents has been shown to be a sensitive and accurate method for depiction of the biliary tract. In a study by Van Beers et al. [29], depiction of the common bile duct and the confluence of the hepatic ducts using IV cholangiographic agents was accurate in 100% of patients (26/26), with visualization of higher order branches in 81% of patients (21/26). However, the use of IV cholangiographic agents is limited because of reactions to iodipamide meglumine and its derivatives. The incidence of minor reactions varies from 4% to 24%. Major reactions are rare and include hepatic and renal toxicity. In many practices, these contrast agents have been abandoned because they have the highest potential for mortality among radiographic contrast agents [3]. Recently, in a study in which helical CT cholangiography was performed with an oral contrast agent and showed good sensitivity and specificity for choledocholithiasis [10]; helical CT cholangiography using iopanoic acid was performed in 31 patients. Only four patients had normal-caliber ducts found on ERCP. Fourteen patients had choledocholithiasis. Sensitivity for detection of choledocholithiasis was 93% and 86%, and specificity was 100% for both reviewers. Our study was not designed to specifically address the detection of common bile duct stones but was designed to evaluate the overall clinical ability to produce cholangiograms. To our knowledge, the overall quality of this new technique for imaging the entire biliary tree has not been evaluated.
Iopanoic acid, the contrast agent we studied, has previously been used to opacify the gallbladder in oral cholecystography. Iopanoic acid was well tolerated by the subjects in our study. Another advantage of this oral contrast agent is the ease with which it is used. Patients simply ingest the contrast agent approximately 12 hr before CT. Conversely, IV contrast agents require slow infusion over 30-45 min with a 30-min delay before imaging.
There are several disadvantages to using oral biliary contrast agents. Iopanoic acid must be ingested several hours before the examination, thus making CT cholangiography impractical for urgent procedures. The ability of iopanoic acid to opacify the gallbladder is subject to many conditions that could interfere with ingestion, absorption, conjugation, and excretion of contrast material into the biliary system. At oral cholecystography, opacification of the gallbladder after a single dose (3.0 g) of iopanoic acid fails to occur in approximately 25% of patients. After a second dose (total dose, 6.0 g), gallbladder opacification improves in approximately 50% of patients. In patients who have no gastrointestinal or hepatic disease, if the gallbladder cannot be visualized after two doses, then they are presumed to have acute or chronic cholecystitis [30, 31]. We attempted to optimize imaging by giving subjects a double dose (6.0 g) of iopanoic acid before imaging. All patients with cholelithiasis (12/14) had pathologic evidence of chronic cholecystitis but nine of the 14 patients had acceptable opacification of the gallbladder. In five of the 14 patients with suboptimal opacification of the gallbladder four had nondiagnostic CT cholangiograms. A possible explanation for the poor visualization of the gallbladder in these patients is that peak contrast opacification occurs 14-19 hr after ingestion, whereas our subjects were imaged 6-10 hr after ingestion. Of the nondiagnostic studies, opacification of the small bowel failed in only one patient, suggesting that an inadequate amount of oral biliary contrast agent had been ingested or excreted into the biliary system.
Several important limitations exist for CT cholangiography. The success of this procedure is decreased in patients with biliary disease. In our study, all volunteers had diagnostic studies, whereas five of the 14 patients had nondiagnostic studies. This technique is also limited in its depiction of intrahepatic anatomy and extrahepatic luminal irregularities, such as strictures, and intraluminal defects, such as stones. Interestingly, this finding contradicts the findings of Soto et al. [10] who reported CT cholangiography has a sensitivity of 93% and a specificity of 100% for the diagnosis of choledocholithiasis. In another study, Neitlich et al. [32] evaluated the detection of choledocholithiasis with unenhanced CT. The few false-negatives occurred in patients with normal-caliber ducts with common bile duct stones measuring up to 3 mm. Also, in patients in whom biliary opacification was poor, separating the biliary anatomy from surrounding structures—particularly after 3D volume renderings had been performed—was difficult. The multiplanar and maximum-intensity-projection data and source images were not as affected in these patients.
Helical CT cholangiography using an oral biliary contrast agent is a feasible technique: It is well tolerated by patients, inexpensive, and easily applied to current technology. However, this technique is unlikely to replace current noninvasive imaging techniques such as MR cholangiography. Opacification of the biliary anatomy is subject to many factors, thus technically limiting the reliability of CT cholangiography as an imaging alternative.
Address correspondence to E. K. Paulson.
| References | |
|---|
