Applied Radiology

Dr. Henkin is a Professor of Radiology and Vice-chair of the Department of Radiology, Loyola Stritch School of Medicine, Maywood, IL. He is also a member of the Editorial Board of this journal.

Hepatobiliary imaging in nuclear medicine with 99m-technetium­labeled agents is about to celebrate its 20th anniversary. ¹ It is appropriate to take a look at where this technique fits into the current practice of nuclear medicine and the care of patients with suspected hepatobiliary disease. In addition, over the last year, some difficulties have been encountered with one of the major pharmaceuticals used as an ancillary tool in hepatobiliary imaging. Alternatives have been developed by a number of different individuals to help in the continuation of this valuable technique. In this review, we will attempt to discuss the physiologic basis of hepatobiliary imaging, review its major indications, and, in general, assess its place in 2002.

Hepatobiliary imaging in nuclear medicine is actually not a new technique. Early in the history of nuclear medicine, iodine-131 (I-131)­labeled rose bengal was used as a tracer for outlining the biliary tract. Actually, rose bengal was a very good agent for imaging the liver but because of the I-131 label, radiation dosimetry was unfavorable and the technique was used infrequently.

Beginning in the 1980s, a series of compounds were introduced into clinical nuclear medicine that employed technetium as the isotopic label. The compounds are all related and fall in the iminodiacetic acid (IDA) family. The original compound, hepatic 2,6-dimethyliminodiacetic acid (HIDA), had significant limitations, mainly in that it could only be used to image when the patient's bilirubin level was below approximately 5 mg/100 dL. While this agent was useful for imaging the gallbladder, the technique could not be performed with accuracy in patients with elevated bilirubin levels. Subsequent to the introduction of HIDA, disofenin and mebrofenin were introduced for clinical imaging. Both of these compounds are substituted IDA molecules and offer the advantage of imaging at elevated serum bilirubin levels. Imaging up to bilirubin levels of approximately 20 mg/100 dL have been reported. ² With the introduction of these latter two compounds, nuclear medicine was offered the opportunity to distinguish between medical and surgical jaundice as well as to evaluate the gallbladder in a number of conditions. The IDA family of compounds is excreted via the biliary pathways. However, unlike bilirubin, these compounds are not conjugated.

The normal biliary tract study

The sequence of events after the injection of technetium-labeled IDA compounds provides information about hepatic function as well as the patency of various portions of the ductal system. It is anticipated that the tracer will clear the circulation by 5 minutes postinjection. Therefore, visualizing significant cardiac blood pool activity on a hepatobiliary study on the first or second frames of a study suggests that there is impaired clearance ability of the hepatocytes, which represents an indirect measure of hepatic parenchymal function.

Typically, the gallbladder and common duct system are identified by 15 to 30 minutes postinjection. Visualization after 1 hour postinjection is still considered to be normal. The gastrointestinal tract should be visualized by 1 hour (Figure 1).

Failure to visualize the gallbladder by 4 hours postinjection raises questions of cholecystitis, both acute and chronic. Failure to visualize the ductal system suggests bile-duct obstruction of that system. At times, the proximal ductal system can be visualized without the distal duct system being seen. This offers a clue as to the level of obstruction. The kidney represents an alternate route of excretion for these compounds. While low-level renal and bladder activity are commonly identified on scans, the presence of significant renal and bladder activity suggests either biliary tract obstruction or hepatic parenchymal disease.

General indications for hepatobiliary imaging

One of the most common indications for hepatobiliary imaging is suspected acute cholecystitis. While diagnostic ultrasound is commonly employed as the first diagnostic test when this entity is clinically suspected, the ultrasound appearance of sludge, the absence of stones, or significant edema of the gallbladder may lead to further evaluation by hepatobiliary imaging.

A second major indication is for an unusual entity that is sometimes entitled "biliary dyskinesia." The cause of this clinical syndrome is not completely documented, but it is thought to represent an end organ defect with regard to response to cholecystokinin (CCK) secondary to an abnormal or inhomogeneous distribution of CCK or neurotransmitter cells in the gallbladder wall. The clinical picture in these patients is one of acute cholecystitis. There is postprandial right upper quadrant pain. This pain may be intense, and clinical symptomatology suggests the presence of gallstones. However, on ultrasound examination, the gallbladder is usually normal and usually no stones are detected. Below, we will discuss protocols for imaging these patients. Historically, removal of the gallbladder results in relief in only about half the patients. The remainder continue to have symptoms for reasons that are not well understood.

Another common indication for hepatobiliary imaging is suspected bile leakage. The majority of these patients have either had recent surgery or mechanical trauma to the abdomen. In this setting, extravasation of tracer into the abdomen is usually obvious and a quick diagnosis of bile leakage can be made (Figure 2). These patients are commonly explored surgically. Persistent postoperative bile drainage from surgical drains or signs of peritoneal irritation may indicate continued bile leakage. In this setting, hepatobiliary studies may be of particular use in confirming and defining the origin of the leak. Sequential imaging during the first hour after injection may disclose whether the leak is occurring from the liver itself or from the duct system. Delayed imaging may be required to determine the drainage pattern of the leak.

The pediatric population presents another group in whom hepatobiliary imaging is often requested. Here the issue is usually not one of gallbladder disease, but rather a differentiation of neonatal hepatitis from biliary atresia. Once again, some special considerations are required in this patient population, and these will be discussed below.

Patient preparation

No matter what the indication is for biliary tract imaging, in general the patients have to be fasting for a period of 2 to 4 hours prior to imaging. Since a recent meal will empty the gallbladder, and it may not refill, imaging soon after feeding or during the nonfasting state may result in false-positive examinations. ³ In cases of suspected acute cholecystitis or chronic cholecystitis, the patient usually has not eaten for a number of hours. However, one must be careful with regard to how long a fast has been carried on. Fasting beyond 24 hours, or the introduction of nasogastric suction for >24 hours tends to put the bowel and the gallbladder at rest. In this setting, the gallbladder may not visualize even though it is normal. Caution should be exercised in imaging patients who have been on prolonged nasogastric suction and in whom evaluation of the gallbladder is considered essential. If the gallbladder visualizes in this setting, the test may be considered normal. Failure of the gallbladder to visualize is inconclusive.

The protocol for imaging biliary leaks is similar to that of acute cholecystitis. However, in general, a dynamic protocol has to be established for these patients since leaks may be transient and activity needs to be followed to see where bile pools in the abdomen. Therefore, a slow dynamic study is usually acquired for a period of at least 1 hour. Replaying the study in cine mode is often beneficial in terms of determining the site of the leak and where the bile is traveling to after leakage.

The pediatric patient presents a different set of problems for imaging. In the neonatal period, liver transport enzymes necessary for excretion of IDA compounds are not normally developed. Injection of hepatobiliary tracers without preparation of the patient may result in a false-positive study. Hepatic parenchymal stasis is common (the retention of activity in the liver with no ductal system visualized) in this population and failure to visualize the ductal system and gallbladder is not uncommon. In order to induce the liver enzymes necessary for transport of the nonconjugated hepatobiliary agent, patients must be pretreated with 5 to 7 days of phenobarbital.

Interventions

Patients undergoing testing for biliary dyskinesia require a further intervention. When imaged fasting, these patients generally have normal studies. However, when the gallbladder fills, a challenge with a CCK analog or fatty meal is required to make the diagnosis. It is important to note the response of the gallbladder to the CCK analog and whether or not clinical symptoms are reproduced. In those patients with both reproduction of clinical symptoms and failure of the gallbladder to respond normally to CCK, biliary dyskinesia is suspected (Figure 3).

The gallbladder response to the CCK analog is variable. In general, most individuals consider a gallbladder ejection fraction >= 50% to be normal. In addition, most individuals will agree that a gallbladder ejection fraction <30% is abnormal. A gray zone lies between 40% and 50% ejection fraction. Other clinical factors have to be considered in this patient population before cholecystectomy is recommended.

While a number of different protocols have been proposed for the use of CCK analogs, the one we have found most effective in our laboratory involves the injection of 0.02 µg/kg of the CCK analog followed by 15 minutes of imaging. If the patient does not experience severe abdominal symptoms with the 0.02 µg injection, then a second injection of 0.04 µg/kg is performed at the conclusion of the 15-minute imaging cycle. Another 15 minutes of imaging is then undertaken. Care must be taken with the injection of CCK analog. Too rapid an intravenous injection of this product will produce spasm of the gallbladder neck and prevent the gallbladder from emptying normally. The tracer is generally injected over a period of 3 to 5 minutes. The patient is observed for replication of symptoms, such as abdominal pain. Specifically, right upper quadrant pain should be noted. ⁴

Patients with biliary dyskinesia may have significant psychological overlay associated with their disease. Since the evaluation of this patient population in part requires the subjective estimation of abdominal discomfort, we precede the first injection of CCK with an injection of normal saline. The patient is then questioned about abdominal discomfort. If significant abdominal discomfort is reported on the normal saline injection, then we are forced to discount any symptoms subjectively reported by the patient after CCK injection. Minor symptoms such as nausea and vague abdominal discomfort are common after CCK analog injection and have no clinical significance. The replication of right upper quadrant pain, however, should be considered to be clinically significant.

Recently, a commercial product representing the terminal octopeptide of CCK that has been in use for a number of years has become temporarily unavailable due to a change in manufacturer. It is anticipated that the product will be on the market again in late 2002. Some pharmacies offer a compounded material that represents the terminal octopeptide of CCK and, depending upon location in the country, this product may still be available. However, in those regions where the product is not available, a substitute procedure is required.

In the era before hepatobiliary imaging, radiographic imaging of the gallbladder with a fatty meal was not uncommon. Commercial preparations of this fatty meal agent were routinely available. Today, with the relative absence of radiographic gallbladder imaging, it is more difficult to find such products. We have had success with a commercial dairy product normally used as an additive for coffee. This product, half & half, contains about 20 g of fat per 8 oz serving. In the vast majority of patients, this is an acceptable substitute for CCK and produces significant gallbladder emptying after feeding. While some patients may have difficulty with drinking 8 ounces of half & half, the majority of patients can tolerate the product.

Another indication for the use of CCK is in patients who have been fasting for prolonged periods of time. As noted previously, in such patients the gallbladder may be filled with sludge and intraluminal gallbladder pressure may rise. These patients may be pretreated with CCK 30 minutes prior to injection of tracer. This causes the gallbladder to contract, thus ejecting the residual material reducing intraluminal pressure. ⁵

In a situation in which the gallbladder does not visualize initially during the hepatobiliary study, two options exist for further evaluation. The first is continuous imaging out to about 4 hours postinjection. A gallbladder that does not visualize by 4 hours is considered to be abnormal. However, for the emergency department patient and in many other clinical settings the 4-hour delay is unacceptable. Therefore, an intervention exists that permits more rapid imaging of the gallbladder.

This intervention is to administer 2 mg of morphine sulfate intravenously to the patient. Morphine sulfate causes constriction of the sphincter of Oddi and increases the back pressure in the biliary duct system. This increase in back pressure diverts bile into the gallbladder when it might normally drain into the gastrointestinal (GI) tract (Figure 4). In order for this intervention to be successful, there must be significant tracer remaining in the liver to be excreted. If the liver has emptied completely, injection of additional tracer compound is required before the morphine injection. Typically, visualization of the gallbladder occurs within 30 minutes of the injection of morphine sulfate. Failure to visualize the gallbladder after morphine sulfate injection is consistent with an obstructed cystic duct. ⁶

There is an entity that has been described postcholecystectomy in which there is a paradoxical response of sphincter of Oddi to CCK. When this condition is present, the scan shows a delay in the bowel to biliary transit time and the common duct may appear dilated on scan (Figure 5). To demonstrate this pattern on the scan, the patient is pretreated with CCK 15 minutes prior to the injection tracer. A slow dynamic acquisition is acquired on a computer using an interval of 1 minute per frame for 60 minutes. Using electronic regions of interest over the liver and common bile duct, dynamic curves are generated from the acquired data.

The time-to-peak hepatic uptake and the percent of common duct emptying are calculated. Common duct emptying is computed by determining the peak counts in this area and those remaining at 60 minutes in the same region. Division of the residual counts by the peak counts yields the percent emptying. The following visual assessments are also made: time to biliary visualization; the presence or absence of intrahepatic biliary ducts on scan; and the time of bowel visualization. Using a scoring system it is then possible to produce a semiquantitative estimate of sphincter dysfunction. ⁷

Expected outcomes

Since the introduction of the morphine sulfate challenge, numerous papers indicate that the technique offers a diagnostic accuracy for the diagnosis of acute cholecystitis in excess of 90%, when an appropriate imaging protocol is followed. ⁸ The most common reason for a false-negative study has been acalculous cholecystitis. Clinical experience shows that up to 10% of patients with acalculous cholecystitis have normal studies. A false-positive study for acute cholecystitis more commonly occurs when the imaging protocol with regard to fasting and nasogastric suction has not been followed. As noted above, patients who are fasting for longer than 24 hours or have nasogastric suction in place may indeed have nonvisualized gallbladders. Those patients who have eaten recently may also have nonvisualized gallbladders. Therefore, adherence to protocol is quite important.

In patients with significantly elevated bilirubin levels, visualization of the gallbladder can take a considerable period of time. The normal rules of visualization by 4 hours do not hold, and delayed imaging out to 24 hours may be required (Figure 6). In these patients, the elements of hepatitis that may exist can increase the time course of excretion of the tracer and therefore delay accumulation in the gallbladder.

In evaluating patients for biliary atresia, up to 25% of infants with severe neonatal hepatitis had no evidence of biliary excretion despite pretreatment with phenobarbital. The sensitivity of the technique for the diagnosis of biliary atresia itself is 100% in many series, but the specificity is as low as 74%. ⁹ There are other entities that mimic biliary atresia on hepatobiliary scan. Severe neonatal hepatitis can have a similar appearance on scan, and the reader must be aware of the full clinical picture at the time of scan interpretation. Perhaps the simplest approach to interpretation of these scans is that GI tract visualization essentially excludes the need for surgery. However, the absence of GI tract visualization is not always diagnostic of biliary atresia.

Conclusion

Although cholescintigraphy is highly accurate, diagnostic ultrasound is usually the first test performed when cholecystitis is suspected. In many clinical situations, nuclear hepatobiliary studies are more accurate, but less convenient. In situations of trauma or biliary dyskinesia, the nuclear hepatobiliary study is preferred.

Despite the high accuracy of hepatobiliary imaging and its diverse applications, this technique remains somewhat underutilized. When one considers that patients with right upper quadrant pain, acalculous cholecystitis, obstructed cystic duct, biliary dyskinesia, post-cholecystectomy syndrome, trauma, and unusual presentations of abdominal pain (Figure 7) are candidates for hepatobiliary imaging, we realize that all these groups can benefit from the technique. It odd that more hepatobiliary studies are not performed.

Despite recent shortages of the CCK analog, local compounding of CCK by various pharmacies and the substitution of fatty meals appear to be filling this void. Successful gallbladder ejection fraction studies continue to be available routinely in most medical centers. Perhaps it is a question of out of sight, out of mind. The visibility of the hepatobiliary study must be raised in the community. It is a low-risk, high-yield study that has an important place in the clinical armamentarium of nuclear medicine. AR