Hepatobiliary anomalies associated with ABCB4/MDR3 deficiency in adults: a pictorial essay

Intrahepatic lithiasis is considered to be very uncommon in Europe and much more frequent in Asia [10]. Since the description of MDR3 deficiencies, an increasing number of patients are diagnosed with LPAC syndrome-related intrahepatic lithiasis. The diagnosis is frequently performed several years after the beginning of the symptoms due to their lack of specificity. The syndrome should be suspected in case of the association of several elements. The three most important are: (1) biliary symptoms in a young adult (<40 years old); (2) symptom recurrence after cholecystectomy; (3) presence of hyperechoic material in the biliary ducts. Other minor criteria have been described, such as mild chronic cholestasis, at least one episode of cholangitis, acute pancreatitis or biliary colic, efficiency of ursodesoxycholic acid (UDCA) and similar symptoms in first-degree relatives. ABCB4 gene mutations have been described in patients with LPAC syndrome in 25–56 % of cases [11‐13]. At imaging, LPAC syndrome is associated with various MR presentations: normal MR cholangiography, isolated intrahepatic lithiasis and, rarely, bile duct dilatations. No imaging differences can be found between patients with or without ABCB4 mutation and no specific mutation can be associated with the different presentations [12, 14].

The most common presentation of LPAC syndrome is isolated intrahepatic lithiasis. In LPAC patients, most stones are referred to cholesterol “yellow” stones. Cholesterol stones may vary in colour from light-yellow to dark-green or brown. To be classified as such, they must be at least 50 % cholesterol by weight [15] (or 70 %, according to the Japanese classification system) [16, 17]. They may also be calcium bilirubinate “brown” stones (pigment stones) [18]. Those are traditionally favoured by bile infection, frequent in the Asiatic population but much rarer in western countries.

Ultrasound examination is very accurate in detecting intrahepatic stones, since they appear as heterogeneous and echoic foci centred on the intrahepatic ducts, or as a “comet-tail” artefact due to ultrasound reverberation (Fig 1) [19, 20]. The artefact is not mobile, as opposed to pneumobilia. The “comet-tail” artefact may be due to intrahepatic lithiasis or to the associated cholangiopathy.

If endoscopic retrograde cholangiopancreatography (ERCP) has been considered as the “gold standard” for diagnosing bile duct stone, magnetic resonance cholangiopancreatography (MRCP) is a non-invasive alternative technique that has been shown to be equivalent to ERCP in choledocholithiasis diagnosis and superior to ERCP in intrahepatic lithiasis diagnosis [21‐23]. The diagnosis of biliary stone is based on the presence of round or oval shape signal voids in the lumen of the bile ducts on heavily T2-weighted sequences (Figs. 2, 3, 4, 5, 6 and 7). However, it is occasionally difficult to diagnose stones when the surrounding liquid is not present. On T1-weighted sequences, stones present with a spontaneous hyperintensity and, recently, studies have shown the superiority of these sequences over the T2-weighted sequences for the detection of biliary stones [24]. Discrepancies between MR and ultrasound have been reported as MR is not able to detect very small stones (Fig 1), while ultrasound may be suboptimal in case of massive intraluminal stones (Fig 7). A complementary exploration by ultrasound and MR should be recommended.

Computed tomography (CT) performance for the diagnosis of intrahepatic lithiasis depends on the calcium content of bile stones. Cholesterol stones present with a typical hypodense aspect and are poorly visible. MR and ultrasound are superior to CT is such cases.

When bile duct dilatations are present, they appear as mild to moderate. Rarely, patients may present with more severe dilatations (<10 %) [14]. Such severe bile duct dilatation may only involve one or two liver segments or may be diffuse (Figs. 3, 4, 5, 6 and 7). In such cases, and as opposed to ductal plate malformations (Caroli disease, congenital hepatic fibrosis, congenital cystic anomalies of the common bile duct, etc.), which are directly related to abnormal embryological development of the bile ducts, abnormalities are related to a biliogenesis disorder developed on initially normal ducts. Anomalies are a consequence of the chronic alteration of the bile composition, which results in damage to the biliary epithelium. Bile duct abnormalities are demonstrated as unifocal or multifocal non-cystic large spindle-shaped bile duct dilatations [14] (Figs. 2, 3, 4, 5, 6 and 7). However, ductal plate malformations, particularly the Caroli disease, have to be ruled out by the absence of specific features, such as the “central dot sign”, dilated bile ducts associated with focal area of cystic ectasia, or the fact that no biliary dilatation can be found in LPAC syndrome without underlying biliary stones [14, 25]. Other possible differential diagnosis is bile duct dilatations related to focal obstacle on the biliary ducts. Traditionally, downstream-acquired stenosis (iatrogenic, biliodigestive anastomosis, sclerosing cholangitis, cholangiocarcinoma) may lead to obstructive dilatations that appear more central without intrahepatic lithiasis [14].

In mice, the multidrug resistance (MDR) glycoproteins that mediate the translocation of phosphatidylcholine across the canalicular membrane of the hepatocyte are called MDR2. Whereas the main feature in MDR2 knock-out mice, which corresponds to the equivalent animal model of human MDR3 deficiency [26, 27], is sclerosing cholangitis, controversies exist whether a genetically determined dysfunction of MDR3 plays a pathogenic role in primary biliary cirrhosis and primary sclerosing cholangitis (PSC) in humans. Pauli-Magnus et al. [28] found no genetic argument supporting the role of MDR3 in PSC. Since then, concepts in PSC understanding have evolved and many authors consider that PSC may represent a mixed bag of diseases of different aetiologies in which several genes such as ABCB4/MDR3 may play a disease modifier role [29]. To support this conceptual view, our group recently reported for the first time, in a series of 13 patients with MDR3 deficiency, imaging presentations mimicking sclerosing cholangitis in two patients at MR imaging [30] (Figs. 8 and 9). They corresponded to small duct fibro-obliterative lesions at pathology, and may be due to the direct toxic effect of biliary acids on epithelium. To our knowledge, this is the only report of such association of MDR3 deficiency and secondary sclerosing cholangitis but the two patients presented with recurrent cholangitis and not LPAC syndrome per se.

All complications associated with chronic cholangitis and/or cholelithiasis have been described in patients with LPAC syndrome: intrahepatic cholangiocarcinoma (IHCC) (Fig. 10), portal hypertension, cholangitis and abscess formation (Fig 11), hepatic fibrosis or cirrhosis. IHCC is a rare primary liver tumour (10-20 %) [31, 32]. Several risk factors have been identified and differ in western and Asian populations: primary sclerosing cholangitis, congenital biliary abnormalities and hepatolithiasis [32]. In most cases, no underlying risk factor is found. Recently, Tougeron et al. [9] reported two cases of IHCC in different and unrelated families with MDR3 deficiency. In both cases, no argument supporting the direct relation between ABCB4 mutations and tumorigenesis was found and IHCC may be considered as a consequence of the chronic biliary abnormalities. Genetic polymorphisms in biliary transporters genes have been studied but, to date, no relation has been established between IHCC and ABCB4 mutations [4].