Recurrent Pyogenic Cholangitis Imaging

Recurrent pyogenic cholangitis (RPC) is a disease that is defined by multiple instances of bacterial biliary tract infection, multiple strictures of the biliary tree, intrahepatic and extrahepatic biliary pigment stone formation, hepatic abscesses, dilation and stricturing of the intrahepatic and extrahepatic bile ducts, and recurrent bouts of cholangitis. ^{[1, 2, 3]}

Specific statistics on worldwide rates of RPC are scarce. It is rare in the United States and in the West (prevalence < 2%), and incidence is highest among immigrant populations from East and Southeast Asia. Disease prevalence is reported to be as high as 30% in East Asian and Southeast Asian regions with endemic helminthic infection. This has likely decreased in the 21st century with improved access to health care, improved hygiene standards, and westernization of diets. This disease affects women more than men, with increasing age beyond 50 years a risk factor. ^[4]

Recurrent pyogenic cholangitis is characterized primarily by hepatolithiasis within the intrahepatic and extrahepatic bile ducts. These calculi, formed for the most part by calcium bilirubinate, do not necessarily cause biliary obstruction (unlike cholesterol stones implicated in biliary colic) but initiate a cycle of inflammation and biliary stricture formation within a dilated biliary tree. ^[4]

Recurrent pyogenic cholangitis results from repeated infections of the biliary tree (cholangitis). RPC is endemic to areas of Asia but is being recognized with increasing frequency in Western nations due in part to increased population mobility and international travel. However, even in these locations, it remains predominantly a disease of Asian immigrants. ^[4]

Affected patients are often plagued by recurrent bouts of cholangitis and commonly suffer from complications such as abscess formation, biliary strictures, pancreatitis, acute kidney injury, and chronic liver disease. In severe cases, cirrhosis with portal hypertension may develop. The exact etiology of RPC is unknown, but parasitic infections such as Clonorchis sinensis and Ascaris lumbricoides, ascending bacterial infection with gut flora (Escherichia coli), and low socioeconomic status have been associated strongly with it. ^[5]

Recurrent pyogenic cholangitis is a chronic progressive disease that is frequently accompanied by cholangiocarcinoma (CCA). Patients with RPC frequently develop acute cholangitis, liver abscess, cirrhotic complications, and CCA. Both-sided liver atrophy was reported by You and associates to be a significant risk factor for CCA. ^[6]

Acute cholangitis (AC) is a common emergency with significant mortality risk. The Tokyo Guidelines of 2018 (TG18) provide recommendations for diagnosis, severity stratification, and management of AC. However, validation of the TG18 remains poor. A retrospective audit identified type 2 diabetes mellitus, systolic blood pressure less than 100 mm Hg, Glasgow Coma Scale score less than 15, and malignant etiology as predictors of in-hospital mortality among patients with AC. ^[7]

Recurrent pyogenic cholangitis, although rare, can present acutely with symptoms in keeping with ascending AC. Patients with severe RPC can present in a life-threatening condition. Specific guidelines for management of acute RPC do not exist per se, but guidelines for initial management of ascending AC remain relevant, given the clinical overlap. Management bundles such as those provided by TG18 facilitate a standard approach to the disease across sites, encourage optimized local protocols, and provide a framework within which a clinician can provide excellent patient care. One must be aware that advanced management of RPC does differ from that of AC, so these guidelines should be used with care and by appropriately trained clinicians. ^[4]

A preplanned analysis of a prospective observational multicenter audit that captured patients undergoing emergency admission for complicated biliary calculous disease compared self-reported institutional adherence to TG18 versus “real-world” contemporary practice across Europe and found that although awareness of TG18 recommendations was high, clinicians showed low compliance with TG18 when treating patients with complicated acute biliary calculous disease. ^[8]

Evaluation

Given its significant overlap in clinical presentation with other causes of cholangitis, diagnosis of RPC is challenging. It requires a multidisciplinary approach involving the emergency physician, gastroenterologist, surgeon, interventional radiologist, and histopathologist, among other specialists. ^[4]  A range of routine and specialized tests should be utilized to aid clinicians in making a diagnosis of RPC. These findings should be supported by careful history taking and comprehensive patient examination. Bloodwork should include complete blood count, renal and liver function testing, coagulation studies, measurement of serum lipase, and peripheral blood cultures. ^[4]  Imaging (see Imaging modalities, below) has a major role in evaluation of RPC. ^[4]

Imaging modalities

Plain abdominal radiographs may be obtained in the workup of patients presenting acutely with abdominal pain or jaundice. Results are nonspecific and add little to no evidence to support a diagnosis of RPC. ^[9]  If RPC is suspected, abdominal ultrasonography or computed tomography (CT) scanning must be performed as the next step. Because radiography is insensitive and nonspecific, false-positive or false-negative findings are not well described.

The initial screening radiologic study of choice is abdominal ultrasonography. It is highly sensitive for ductal stones and ductal dilation and may be helpful in localizing hepatic abscesses or other complications. All patients should undergo abdominal CT scanning with contrast enhancement. CT scanning allows full assessment of the extent of disease. ^[1]

Magnetic resonance cholangiopancreatography (MRCP) is the gold-standard noninvasive imaging modality for the diagnosis of RPC. MRCP is superior to CT and ultrasonography for visualization of intraductal calculi and bile duct strictures characteristic of pyogenic cholangitis. The “arrowhead sign” refers to the finding of reduced arborization of the peripheral biliary tree. A subtle cholangiocarcinoma could also be appreciated. ^[4]

Radionuclide studies are not typically used in the diagnostic algorithm for RPC. Results of these studies are nonspecific and do not yield enough information upon which to base treatment strategy. Although these studies may demonstrate some extrahepatic ductal abnormalities, they are a poor choice for delineating the location and extent of intrahepatic disease.

Treatment

Patients presenting acutely with RPC should be managed through a resuscitative approach involving oxygen therapy, fluid resuscitation, correction of potential coagulopathy, and initiation of broad-spectrum antibiotics. Definitive treatment of RPC for those who do not respond to medical management comes in the form of biliary decompression. ^[5]

A retrospective review, by Tan et al, of 157 patients with RPC treated between January 1990 and December 2013 found that surgical treatment is associated with decreased risk of recurrence in RPC but with significant postoperative morbidity. The disease recurrence rate was 43.9% in the overall study cohort throughout follow-up. Surgical treatment was an independent prognostic factor for decreased disease recurrence risk (hazard ratio [HR] 0.40; 95% CI 0.18-0.87; P=0.021). Stratified analysis revealed that liver resection was prognostic for lower risk of disease recurrence among patients with parenchymal disease (PD) (HR 0.38; 95% CI 0.15-0.94; P=0.036), and biliary bypass was prognostic for lower risk of disease recurrence among patients without PD (HR 0.30; 95% CI 0.15-0.61; P=0.001). ^[10]

In the Tan et al study, the overall postoperative complication rate among surgically treated patients was 31.1%, and the presence of bilobar stones was found to be independently associated with higher odds of postoperative complications (odds ratio 3.51; 95% CI 1.26-9.81; P=0.017). The authors concluded that when surgery is deemed appropriate, patients with RPC with and without PD are likely to benefit from liver resection and biliary bypass, respectively. ^[10]

Another retrospective study of 42 patients with RPC over a 20-year period showed reduced disease recurrence among patients who had undergone surgery (hepatectomy or bile duct exploration with or without access loop formation) when compared with nonoperative patients (15% vs 36%, respectively). ^[11]

Major abdominal surgery brings high risk, and decisions about whether to operate should be made via an interprofessional approach. In very rare cases, liver transplant could be considered in a case of diffuse bilateral disease. ^[12]

Endoscopic retrograde cholangiopancreatography (ERCP) is an invasive procedure that is useful for both diagnosis and management of RPC.

(See the images below.)

Prognosis

Retrospective studies suggest that disease recurrence is common despite attempts at definitive treatment. Some of the best outcomes for RPC are seen among patients who underwent hepatectomy as part of their treatment. One study reported stone recurrence on long-term follow-up in 31% of patients and cholangiocarcinoma in 5%. Twenty percent of patients may eventually die of RPC and its associated complications, some of which are often terminal. ^[4]

Frameworks such as the Tokyo Guidelines can improve patient prognosis if used appropriately. Still, it is important to be aware that advanced management of RPC does differ from that of ascending acute cholangitis, so these guidelines should be used with care and by appropriately trained clinicians. ^[4]

Patients with severe RPC can present in a life-threatening condition. Specific guidelines for management of acute RPC do not exist per se, but guidelines for initial management of ascending acute cholangitis remain relevant, given the clinical overlap. Management bundles such as these facilitate a standard approach to the disease across sites, encourage optimized local protocols, and provide a framework within which a clinician can provide excellent patient care. ^[4]

A retrospective analysis with propensity score matching that compared patients with intrahepatic cholangiocarcinoma (ICC) with and without RPC found that patients with RPC had significantly worse median disease-free survival (10 vs 23 months; P=0.0020) and overall survival (15 vs 45 months; P=0.004) than patients without RPC, suggesting that RPC represents a poor prognostic factor affecting outcomes after hepatectomy for patients with ICC. ^[13]

Abdominal CT scanning is the most important noninvasive radiologic study in the diagnosis of RPC. It provides a wealth of information and is an invaluable adjunct in therapeutic planning. ^[14]  

All patients should undergo abdominal CT scanning with contrast enhancement. This allows full assessment of the extent of disease. ^[1]

(See the image below.)

Chan and colleagues used CT to examine 50 patients with RPC. Among these patients, 22 had undergone previous surgical procedures, including sphincteroplasty or choledochoenterostomy. ^[15] A wide range of pathologic features was observed, including intrahepatic ductal dilation (100%), intrahepatic ductal calculi (74%), common bile duct dilation (68%), pneumobilia (52%), hepatic segmental atrophy (36%), common bile duct calculi (30%), bile duct strictures (22%), and splenomegaly (14%). Unilobar hepatic disease was seen in 28%. The left lateral segment was most commonly involved.

Chan et al concluded that CT is sensitive and allows complete radiologic evaluation of RPC. Furthermore, they stated that CT was excellent for confirming ultrasonographic findings, evaluating associated masses, and planning therapeutic procedures (eg, hepatic resection, duct manipulation, stone extraction). Findings consistent with portal hypertension, such as splenomegaly and varices, can be demonstrated by CT scan. ^[15]

Spiral CT data can be reconstructed in a 3-dimensional fashion following infusion of cholangiographic contrast agents. This reconstruction provides a noninvasive way of specifically imaging the biliary tree. The technique is called CT cholangiography. ^{[16, 17]} Although demonstration of stones and strictures is about as good as with endoscopic retrograde cholangiopancreatography (ERCP), CT cholangiography does not yet have the overall accuracy to serve as a front-line replacement for studies such as ERCP or percutaneous transhepatic cholangiography (PTC).

Degree of confidence

Abdominal CT scanning is currently the most reliable imaging study for ascertaining the complete status of RPC. It effectively detects disease in the biliary tree and within the hepatic parenchyma with a high degree of sensitivity and overall accuracy.

Current spiral CT scanners are sensitive for detecting extrahepatic and intrahepatic calculi. More than 90% of intrahepatic stones appear hyperattenuating compared with the nonenhancing hepatic parenchyma.

Hepatic segmental atrophy, compensatory hypertrophy, capsular distortion, and complications (eg, abscesses, bilomas, cholangiocarcinomas) are reliably imaged with CT. Other parenchymal infectious complications such as acute focal pyogenic hepatitis are easily delineated and appear as persistent parenchymal enhancement in affected segments.

Computed tomography is good for detecting peripheral cholangiocarcinomas, as reported by Honda and coworkers, ^[18] but is less reliable (40%) for hilar lesions, as described by Soyer and coworkers. ^[19] Most central lesions less than 1.5 cm in diameter are missed by CT scanning. Features of cholangiocarcinoma on CT include hypoattenuation, lack of encapsulation, central scarring, contrast pooling, and capsular retraction.

False positives/negatives

Hyperattenuating stones become more difficult to image after the hepatic parenchyma is enhanced through administration of contrast material. Pneumobilia in patients presenting with acute exacerbations or in those who have undergone a biliary-enteric anastomosis for drainage can decrease the sensitivity of the examination. However, this can be remedied by adjusting window width.

Upper gastrointestinal preparation with oral contrast agents should be avoided. This approach may cause contrast material to reflux into the common bile duct, thus confounding study findings.

Examiners should be aware that chronic disease can distort the hepatic parenchyma, with atrophy of affected segments or lobes and compensatory hypertrophy of other areas. This can cause rotation of the liver and alteration of normal anatomic relationships of portal triad structures. Hepatic segments that most commonly display atrophic changes include the lateral segment of the left lobe and the posterior segment of the right lobe.

Abdominal CT provides information similar to that provided by ultrasonography, with the added benefits of less user dependence and clearer visualization of other abdominal structures and associated potential pathologies. Computed tomography findings include intrahepatic and extrahepatic biliary tree dilation (central > peripheral zones), hyperdense intraductal calculi (especially on unenhanced scans), biliary wall thickening, biliary stricture, and, in rare cases, pneumobilia. Patients with later-stage disease may demonstrate hepatic parenchymal atrophy, most commonly in the left lateral segment. ^[4]

Magnetic resonance imaging (MRI) is being used more frequently for assessment of RPC. It can reveal all salient features of the disease, including stones, ductal dilation and strictures, and hepatic parenchymal disease. Hepatic segmental or lobar atrophy and hypertrophy are easily delineated. Pyogenic hepatitis appears as hepatic parenchymal enhancement on T1-weighted images. Hepatic abscesses and cholangiocarcinomas are focally hypointense and hyperintense masses on T1- and T2-weighted images, respectively. ^{[20, 21, 22]}

(See the images below.)

T2-weighted images are well suited for depicting ductal pathology because bile-filled structures produce particularly high signal intensities. According to Brinks and Borello, stones are easily detected as intraductal filling defects, irrespective of their chemical composition. ^[23] In comparison, T1-weighted images are better for demonstrating acute suppurative cholangitis, which appears as ductal wall enhancement.

Magnetic resonance cholangiopancreatography (MRCP) is the gold-standard noninvasive imaging modality for diagnosis of RPC. ^{[24, 25, 26]}

Available MRCP techniques currently include single-shot rapid acquisition and relaxation enhancement (RARE) and half Fourier acquisition single-shot turbo-spin echo (HASTE) in a single breath-hold with negative oral contrast to decrease the MRCP signal from fluid in the gastrointestinal tract. ^[20]

Park and colleagues compared MRCP with direct conventional cholangiography. ^[27] A total of 24 patients underwent preoperative MRCP before surgery for RPC. Eighteen patients also underwent preoperative conventional cholangiography. Results were verified by anatomic findings at surgery. Magnetic resonance cholangiopancreatography was able to depict 100% of hepatic segments with ductal dilation, 98% with ductal calculi, and 96% with focal strictures. Conventional cholangiography showed 47% of segments with ductal dilation, 45% with ductal calculi, and 44% with focal ductal strictures. The authors concluded that MRCP was superior to conventional cholangiography for anatomic evaluation of RPC because it can depict the entire biliary tree, even in the presence of ductal obstruction or stenosis.

Degree of confidence

Magnetic resonance imaging and MRCP are highly sensitive and accurate techniques. They can depict ductal and parenchymal pathology as well as any other radiologic study can, and they are likely to become the studies of choice for RPC.

Ductal strictures are usually short (< 1 cm) and are easily seen on MRCP in comparison to CT, according to Heffernan and colleagues. ^[1]

MRCP is superior to CT or ultrasonography for visualization of intraductal calculi and bile duct strictures characteristic of RPC. The “arrowhead sign” refers to the finding of reduced arborization of the peripheral biliary tree. A subtle cholangiocarcinoma could also be appreciated. ^[4]

These techniques do not expose the patient to ionizing radiation and do not require injection of contrast agents. They are noninvasive, require no anesthesia, and allow better visualization of ducts proximal to an obstruction or stenosis. ^[20] MRCP does not increase the risk of biliary sepsis. ^[1]

Soyer and colleagues described the presence of a hyperintense central scar on T1-weighted images. ^[28] This finding and contrast agent pooling in the mass on late images are more consistent with a cholangiocarcinoma than with other conditions. Data published by Fan and coworkers suggest that MRI is more sensitive than CT in detecting hilar cholangiocarcinoma. ^[29]

No normal variants that mimic this disease process have been well described. Various pathologic findings associated with this disease process are better visualized on T1-weighted images than on T2-weighted images, and vice versa. These techniques are complementary and should be used together to avoid false-negative results.

The initial screening radiologic study of choice is abdominal ultrasonography. It is highly sensitive for ductal stones and ductal dilation, and it may be helpful in localizing hepatic abscesses or other complications.

Ultrasonography is noninvasive and does not expose the patient to ionizing radiation. It is sensitive for biliary tract pathology. Findings can suggest or confirm the diagnosis in up to 97% of cases of RPC, according to Ohto and coworkers. ^[30]

Ultrasonography easily demonstrates dilation of extrahepatic and proximal intrahepatic ducts. Calculi within the extrahepatic biliary radicles and the central intrahepatic ducts are readily depicted as described by Lim and colleagues. ^[31] Intrahepatic stones are usually hyperechoic as compared to the hepatic parenchyma. Sonograms may depict decreased arborization or truncation of hepatic ducts that typifies RPC.

Distribution of ductal dilation as visualized on ultrasonography tends to be unrelated to the location of calculi. This may be secondary to widespread elasticity of the duct walls, as stated by Heffernan and associates. ^[1]

Features of hepatic parenchymal disease that can be detected on ultrasonography include increased periportal echogenicity, liver abscesses, and bilomas. An associated cholangiocarcinoma appears as a homogeneous, hyperechoic mass, according to Zhang and coinvestigators. ^[32]

Ultrasonography can also be used to help guide interventional procedures, such as aspiration or drainage of bilomas and hepatic abscesses, as well as fine-needle or core biopsy of suspected neoplasms. ^[1]

Ultrasonography is the best noninvasive modality for actual imaging of offending parasites. Flukes are poorly visualized on CT and MRI. With ultrasonography, flukes or clusters of parasites appear as echogenic foci that are nonshadowing within the bile ducts. Morikawa and colleagues were able to capture flukes in motion within peripheral bile ducts on M-mode sonograms. ^[33]

Lim and colleagues note that flukes are easiest to visualize in the gallbladder. ^[34] They sink toward the dependent portion of the organ but can be made to float temporarily with a change in position or transdermal agitation of the gallbladder with the transducer. Flukes in the gallbladder appear fusiform and weakly echogenic and do not produce shadows. ^{[34, 35]}

Although ultrasonography is an excellent screening test, it is operator dependent and may not demonstrate the full extent of disease. For example, intrahepatic ductal strictures and hepatic lobar atrophy can be missed. Computed tomography scanning demonstrates the features of anatomic disease such as extent of ductal involvement, status of hepatic lobes, and presence or absence of associated abscesses. On CT scans, isoattenuating features, such as noncalcified stones and biliary sludge, can be missed.

Ultrasonography is useful for evaluating findings that are consistent with more common diagnoses such as a thickened gallbladder wall more consistent with acute cholecystitis, making the diagnosis of RPC less likely. ^[4]

Decompressive procedures may be required multiple times and endoscopic ultrasound (EUS)-guided drainage may be warranted, especially in those with concomitant hepatic abscesses, if surgical resection cannot be performed. ^[4]

Degree of confidence

The diagnosis of RPC can be made for the vast majority of patients with a high degree of confidence. Sonographic findings should be confirmed by another modality such as CT or MRI. These studies are better for defining the overall extent of hepatic parenchymal and intrahepatic ductal disease.

Distal intrahepatic ductal and parenchymal disease can be difficult to image with ultrasonography. Patients with acute exacerbations of RPC can present with pneumobilia, which obscures details of the ductal system. The presence of ductal air along with a dilated duct packed with stones can have the appearance of an abscess in the acute setting. Also, because the stones in RPC are pigment stones, they may be isoechoic and therefore may be missed on sonographic examination. As always, ultrasonography is highly operator dependent.

Before the advent of sophisticated CT and MRI techniques, direct contrast cholangiography was considered essential in the diagnosis of RPC and in assessment of disease extent. It can still provide critical information about location and extent of stones and ductal dilation and strictures.

Classic cholangiographic findings of RPC include ductal dilation, which is frequently most marked in the extrahepatic duct; dilation of the proximal central intrahepatic ducts with associated abrupt truncation of more peripheral ducts; and filling defects in contrast agent–filled ducts, which represent the presence of stones.

Khuroo and coworkers published findings of direct contrast cholangiography in 227 patients with RPC. ^[36] Baseline cholangiographic abnormalities included biliary calculi and sludge, bile duct dilation, and strictures. Only 21 patients had cholangiographic evidence of prior parasite infestation.

A cohort of 55 patients in this study underwent serial cholangiography at intervals of about 18 months. In this subgroup, symptoms and cholangiographic abnormalities were resolved for 25 patients, who underwent successful endoscopic sphincterotomy and extraction of biliary calculi. In remaining patients, RPC reoccurred and cholangiographic findings worsened over time.

Endoscopic retrograde cholangiopancreatography (ERCP) and percutaneous transhepatic cholangiography (PTC) are the most common clinical approaches to cholangiography. Both are invasive procedures with low, but not negligible, complication rates. Sherman and Lehman reported inability to cannulate the common bile duct in 5-10% of attempted ERCP studies. ^[19]

Degree of confidence

Cholangiography reveals ductal disease and the presence of stones, with relatively high sensitivity and specificity. This is especially true of the extrahepatic and more centrally located intrahepatic bile ducts. Weaknesses include inability to detect disease distal to strictures and an inherent inability to image the hepatic parenchyma. Computed tomography scanning is required to provide information about the extent of hepatic parenchymal disease. This is particularly critical if hepatic resection is planned.

The predictive value of a positive test is high. False-positive findings are few. False-negative studies can occur when strictures limit ability to image distal ductal disease.

A temporary endoscopic ultrasonography (EUS)-guided hepaticogastrostomy followed by staged antegrade cholangioscopy and electrohydraulic lithotripsy has been described for the management of RPC in patients with altered foregut anatomy when the conventional technique was initially unsuccessful. ^[37]

The rendezvous procedure resulted in shredding of a small piece of wire that was successfully retrieved during cholangioscopy. Otherwise, no adverse events were noted. Follow-up magnetic resonance cholangiopancreatography (MRCP) 6 months after the index procedure showed decompression of the targeted duct without stones. The patient remained asymptomatic. ^[37]