This study demonstrated the variety of findings seen on PTCC in a relatively large patient cohort with suspected BA. 55 studies were normal, and thus avoided further invasive investigations requiring general anesthesia and laparotomy in these young patients. The most common diagnosis in those with an abnormal PTCC was BA; the two most common patterns were i) failure to opacify the CHD associated with a narrowed CBD, and ii) an isolated small GB. The second most common diagnosis was Alagille syndrome, all of whom demonstrated patent but narrowed CBD and lack of CHD opacification.
Percutaneous transhepatic cholecysto-cholangiography in this center is considered part of the diagnostic work-up of neonates and infants presenting with conjugated hyperbilirubinemia, especially those suspected of BA. It is a minimally invasive alternative to the intra-operative cholangiography and histologic examination of the duct remnants of the porta hepatis, which are considered the standard criterion for the diagnosis of BA [3, 6,7,8,9]. It is crucial to practice careful technique, using high quality imaging and proper intraluminal GB puncture, to prevent contrast extravasation, which obscures visualization and can compromise procedural success. There is no minimal patient weight or size for this procedure, the only requirement is to have a GB or cystic structure that can be punctured. Successful training can be achieved in centers with high volume of these procedures, usually by dedicated pediatric interventional radiologists. Identification of a normal biliary tree appearance and drainage during diagnostic work-up is of major importance as it obviates the need for surgery [10,11,12,13, 17].
The normal length of the GB in the first year of life has been described to be approximately 1.5 to 3.5 cm [18,19,20,21], however it can be difficult to differentiate between a small and a contracted GB. The spectrum of US findings in BA includes a small, absent or an atretic GB without a lumen [21]. A collapsed or non-visualized GB on US is found in 50–60% of BA patients [16, 22]. In the group with normal PTCC, 73% of the patients had a normal GB on their pre-procedure diagnostic US, the remainder had a small or contracted GB, therefore a small or contracted GB is not a contraindication for a PTCC and can be seen in normal studies. The majority of patients (94%) with abnormal PTCC findings had small or contracted GB on pre-procedure diagnostic US, one patient had an absent GB (cyst noted at the procedure day) and one had a small cyst. Only one patient with BA had a normal GB reported in pre-procedure diagnostic US. Based on this, independent of the size, presence, or absence of a GB in the pre-procedure diagnostic US, the GB is re-assessed at the time of the procedure. If nothing is visible to target, a liver biopsy is obtained, or the procedure is aborted. Attempts to access intrahepatic bile ducts are not routinely performed.
When the clinical, laboratory and sonographic findings remain indeterminate, a PTCC is recommended [17]. A PTCC can delineate the anatomy of the biliary system to either exclude BA or help direct further diagnostic steps. Lack of CHD opacification with CBD narrowing is not pathognomonic for any condition, as it may be associated with BA or hypoplasia [Figs. 2, 6,7]. 16 cases had this pattern, 11 of whom (61%) were diagnosed with BA. However, the pattern was also found in biliary hypoplasia/Alagille syndrome and Alpha-1 antitrypsin deficiency. Similar findings with Alagille syndrome have been found by others [16].
A morphologically abnormal GB associated with a non-communicating cyst in either an intra-hepatic or extra-hepatic location is most consistent with cystic BA. Cystic BA was seen in 12% of BA cases, which is consistent with previous reports [23]. The main differential diagnosis of this condition is a choledochal cyst, which can be associated with dilated intra-hepatic bile ducts, a larger cyst, and normal appearing GB [23].
The presence of normal opacification and morphology on PTCC makes the diagnosis of BA very unlikely, and helps avoid further invasive investigations, such as laparotomy with intraoperative cholangiography, in most patients. These results are consistent with previous studies which have identified up to 72.2% reduction of other invasive diagnostic procedures using PTCC [10, 16]. On later follow-up, none were subsequently diagnosed as BA. In our practice, a normal PTCC does not change the decision to perform a planned liver biopsy during the same visit to IR as histologic assessment will aid in the differential diagnosis of the ongoing cholestasis.
The GGT levels were significantly higher in patients with an abnormal PTCC compared to a normal PTCC. In jaundiced neonates and infants presenting with an elevated GGT, an abnormal cholangiography may be likely. High GGT cholestasis has been described in Alagille syndrome, PFIC type 3 and frequently in BA [3]. In the diagnosis of BA, elevated GGT has diagnostic accuracies at varying cut-off levels, and is particularly useful in patients < 120 days old; it is incorporated into diagnostic work-up for BA [17, 24, 25]. In this study the levels of GGT were significantly higher in patients diagnosed with BA compared to other diagnoses. This study’s results further support the use of GGT in the diagnostic work-up of suspected BA.
There are several limitations to this study. This was a retrospective study, and occasional suboptimal patient records or imaging reports were found. The different PTCC findings were not quantified or calculated, which makes it difficult to define normal measurements. With new equipment and IR technology, it is feasible to perform measurements on visualized structures and potentially establish normal parameters – this could be the focus of future studies. It is possible that some BA cases with abnormal or absent GB on diagnostic US, were not included as they may have proceeded directly to surgery, therefore, patients with narrowed CBD, as opposed to other patterns may be over-represented in this study’s cohort. The differentiation between a contracted and small GB in the pre-procedure diagnostic US can be confusing. This terminology was quoted verbatim from radiological reports by a large cohort of diagnostic radiologists and is based on their interpretation of the images and not necessarily on any absolute difference in measurements, however most of the time both terms described a GB less than 1.5 cm in length. Radiation dose metrics were not available for many of the PTTC studies. The total number of patients presenting with conjugated hyperbilirubinemia or the total number of cases with BA during the study period were not determined; an earlier study over 10 years of observation found 212 and 35 patients respectively [5]. Lastly, there may be potential inaccuracies when collating patients’ diagnoses and treatment outcomes from clinical, pathological, and radiological data.
In conclusion, PTCC is a minimally invasive imaging procedure to assess the biliary system in neonates with hyperbilirubinemia and suspected BA. With increasing frequency, it is added to the diagnostic work-up of this condition [3, 16, 17, 26, 27]. It is especially valuable when normal, as it can avoid unnecessary surgery. This paper describes in a large patient cohort abnormal patterns that warrant further investigations, with a low procedure related complication rate. It is important for the pediatric (interventional) radiologist performing this procedure, the pediatric surgeon and pediatric hepatologist, to be familiar with the different patterns encountered, and to collaborate meaningfully in the management of these patients.
