Introduction

Biliary strictures (BS) are relatively common after pediatric liver transplantation with a reported incidence of 10–35 % [1, 2]. BSs are frequently managed percutaneously, thus eliminating the need for more invasive surgical interventions in most cases [3–8].

The small liver volume in partial liver transplant recipients, the absence of intrahepatic biliary dilatation, and the presence of multiple biliary anastomosis render such interventional procedures challenging given the need for long fluoroscopy time and large radiation dose. Another valid consideration is the need for multiple procedures and reintervention in the majority of patients.

The radiation exposure from interventional radiology procedures is particularly relevant when treating children because of their greater radio sensitivity compared with adults; children have also a relatively long life span during which this potentially augmented oncological risk could be expressed [9, 10]. The radiation risk associated with these procedures may be deterministic or stochastic. Deterministic effects are those that occur once a specific dose threshold is exceeded and including radiation skin injury. In contrast, the increased oncological risk is a stochastic effect and is therefore related to the cumulative radiation dose with no defined threshold.

The step lightly campaign [11, 12] recently focused on improving radiation safety during pediatric interventional radiology procedures. The main consideration is the need for future research as well as development a larger baseline patient database regarding pediatric interventional radiology procedures to meet the main goal of evaluating doses delivered to pediatric patients who undergo interventional procedures. However, knowledge of the magnitude of radiation exposure in pediatric interventional radiology is limited.

This article reports the radiation exposure for interventional procedures performed in liver-transplanted children with BS who were treated percutaneously in a single centre. We also compare the differences in radiation exposure between procedures performed using an image intensifier-based system (IIDS) and procedures performed with flat panel detector-based system (FPDS).

Materials and Methods

Institutional review board approval for this study was not required due to its retrospective nature and solely record-based approach. No financial support was provided for this study. No specific informed consent was requested for this retrospective survey; however, informed consent to the single procedure was obtained in all cases from a parent or from the child’s legal representative.

Patient Population

From January 2008 to March 2013, a total of 1,195 biliary interventional procedures in 244 patients were performed in the Interventional Radiology Department of a single transplant centre. One hundred seventy consecutive procedures were performed in 34 pediatric liver transplant recipients and constitute our study group. No pediatric biliary procedures were excluded from our survey. A retrospective review of the dose-area product (DAP) and fluoroscopy time values for each of those procedures was performed. Patient characteristics are listed in Tables 1 and 2.

Table 1 Characteristics of 34 liver-transplanted children with BS treated with percutaneous management
Full size table
Table 2 Characteristics of children with BS treated with percutaneous management compared with both type of imaging system
Full size table

Angiographic Equipment

Procedures were performed in an angiographic suite with an IIDS (Advantax; General Electric Medical Systems, Milwaukee, Wis. USA) from January 2008 to July 2010. The procedures performed after July 2010 were performed using an FPDS (Innova 4100; General Electric Medical Systems). In both angiographic systems, DAP and fluoroscopy time were measured using a dual-channel DIAMENTOR M4-KDK DAP/dose meter transmission ion chamber (PTW; Freiburg, Germany), which was fixed to the collimator that had valid calibration and quality-control certificate checked every 6 months. As a routine, DAP and fluoroscopy time values were archived into our Radiology Information System and Picture Archiving and Communication System (Centricity RIS 4.2i; General Electric Medical Systems, Milwaukee, Wis. USA) at the end of every interventional procedure.

Interventional Procedure

Standard dose-reduction measures were routinely employed. This included adequate collimation limited only to the area of interest and low object-to-detector and source-to-image distances. Decrease of the object-to-detector distance was limited by the fact that most children had a left lateral split-liver transplant. Given that the only possible approach here is the anterior subxiphoid approach, an air-gap of at least 20 cm from the abdominal wall to the detector is unavoidable during the puncture and catheterization phase. However, this distance was decreased during bilioplasty or cholangiography. In all procedures using either of the angiographic systems, a low fluoroscopy level without magnification was used as routine. High-detail fluoroscopy and/or magnification was used only in selected technically challenging cases and then only during the most critical steps. Pulsed fluoroscopy at 7.5 images/s was used with the FPDS, whereas it was not possible to vary images rate with the IIDS. The “last image hold feature,” which displays the last active fluoroscopic image, was always used because this enables image review without additional fluoroscopic exposure. The majority of the procedures were performed using only the anteroposterior projection; however, in same cases oblique projections were used during some intraprocedural steps.

All procedures were performed by two interventional radiologists with 11 and 8 years of experience in abdominal interventional radiology, respectively. Percutaneous transhepatic cholangiogram was usually performed with a subxiphoid approach using a 20G needle, which was advanced under ultrasound and fluoroscopic guidance into a peripheral bile duct. After cholangiography, the biliary tree was catheterized using an Accustick Introducer System (Boston Scientific, Natick, MA) over a Cope wire (Cook, Bjaeverskov, Denmark). The stricture was crossed with .035- or .038-in. hydrophilic guidewires, and a transanastomotic biliary catheter ranging from 5 to 8.5F with side holes above and below the stricture was then placed. An external biliary drainage catheter was placed in cases where it was not possible to negotiate the anastomosis so as to allow biliary decompression and decrease any inflammatory components of the stricture. This also provided access for the second attempt to cross the stricture, with the latter being performed a few days later. In 33 of 34 patients, a transanastomotic biliary catheter was placed in the first procedure. An external drainage catheter was placed in 1 patient with a severe anastomotic stricture. The anastomosis was successfully crossed 5 days later, at which point a transanastomotic biliary catheter was placed. The latter procedure was included in the group with cholangiogram and biliary catheter exchange or removal. The first session of balloon dilatation was usually performed 1 week later after a cholangiogram through a 5 or 6F sheath. In every bilioplasty session, three transanastomotic dilatations lasting 10 min each were performed. Transanastomotic biliary catheters (6–10F) were customized by adding additional sideholes and placed through the anastomosis after every session of dilatation. As part of our departmental protocol, a minimum of three dilatation sessions were performed at intervals of 4 weeks. At each bilioplasty session, ballooning was performed using a balloon catheter, which was upsized by 1 mm each session until a maximum diameter of 10 mm was reached. Catheters were not replaced whenever resolution of the stricture was apparent during cholangiography through a sheath or whenever a minimal residual stenosis <20 % of the expected lumen caliber was present. In all cases, catheters were removed only when complete bile duct decompression was evident within 3 min of cholangiography.

Statistical Analysis

For the purpose of this study, the procedures were classified into three categories: percutaneous transhepatic cholangiogram and biliary catheter placement (PTBD); cholangiogram and balloon dilatation; and cholangiogram and biliary catheter change or removal.

The DAP and fluoroscopy time for each biliary procedure performed in pediatric liver transplant recipients was reviewed retrospectively. The fluoroscopy time of each interventional procedure was recorded in seconds. DAP [or Kerma area product (Gy cm2)], was also recorded and was considered as a surrogate measurement of the entire amount of energy delivered to the patient by the radiation beam during the procedure.

Primary data were recorded using Excel 2007 (Microsoft, Richmond, WA). For each of the procedure categories, distribution of DAP and fluoroscopy time was analyzed through descriptive statistical analysis and expressed as mean ± SD. DAP and fluoroscopy time differences for each procedure category between the two angiographic systems were compared using Wilcoxon rank-sum test (Mann–Whitney). To assess differences of DAP between the two angiographic systems and adjusted for the effect of the fluoroscopy time, a multiple linear regression model was performed. DAP was considered a dependent variable. The independent variables were either categorical, as in the case of the angiographic system used, or continuous, as in the case of fluoroscopy time. The continuous variable time was centered at the mean, and the baseline category of the angiographic system was the FPDS. The estimate of the regression coefficients was reported, and the confidence interval (CI) was 95 %. All statistical results were considered significant when p < 0.05 was reached.

Results

One hundred seventy consecutive interventional procedures in 34 patients were included in the study period: PTBD (n = 40), cholangiogram and balloon dilatation (n = 55), and cholangiogram and biliary catheter change or removal (n = 75). Overall mean DAP values for the three procedures were as follows: PTBD = 21.63 ± 23.70, cholangiogram and balloon dilatation = 10.13 ± 11.70, and cholangiogram and catheter removal = 5.89 ± 11.12 Gy cm2. Mean fluoroscopy times were as follows: PTBD = 857 ± 511, cholangiogram and balloon dilatation = 419 ± 285, and cholangiogram and catheter removal = 254 ± 315 s. As expected, DAP and fluoroscopy times varied depending on the complexity of the procedures, and the latter was deemed a cause of intracategorical and intercategorical variation.

Ninety-two procedures were performed using the IIDS: PTBD (n = 22), cholangiogram and balloon dilatation (n = 32), and cholangiogram and biliary catheter change or removal (n = 38). Mean DAP values for the three procedures were as follows: PTBD = 30.89 ± 21.96, cholangiogram and balloon dilatation = 14.06 ± 13.40, and cholangiogram and catheter removal = 10.38 ± 14.20 Gy cm2. Mean fluoroscopy times were PTBD = 807 ± 436, cholangiogram and balloon dilatation = 470 ± 317, and cholangiogram and catheter removal = 344 ± 401 s.

Seventy-eight procedures were performed using the FPDS: PTBD (n = 18), cholangiogram and balloon dilatation (n = 23), and cholangiogram and biliary catheter change or removal (n = 37). Mean DAP values for the three procedures were as follows: PTBD = 10.31 ± 21.12, cholangiogram and balloon dilatation = 4.65 ± 5.44, and cholangiogram and catheter removal = 1.27 ± 1.91 Gy cm2. Mean fluoroscopy times were PTBD = 918 ± 598, cholangiogram and balloon dilatation = 347 ± 219, and cholangiogram and catheter removal = 163 ± 146 s (Table 3).

Table 3 DAP (mean ± SD) and fluoroscopy time (mean ± SD) for 170 biliary interventional procedures performed in 34 liver transplanted children with BS treated with percutaneous management
Full size table

The difference in mean of DAP values on comparison of the IIDS group with the FPDS in the three categories achieved statistically significance. The p values achieved for the PTBD group, the cholangiography and balloon dilatation group, and the cholangiography and biliary catheter change or removal group corresponded to 0.001, 0.0002, and 0.00001, respectively.

The multiple linear regression models showed a significant difference in DAP between the two angiographic systems for all three different interventional radiological procedures considered (Table 4; Figs. 1, 2, 3). Another interesting observation seen in all three interventional procedures was that the increase in magnitude of exposure dose over time was always significantly greater in the IIDS compared with the FPDS. No deterministic or stochastic complications were observed in our study group with a mean follow-up of 42 months (range 4–72).

Table 4 Multiple linear regression analysis performed to assess the differences of DAP between the two angiographic systems adjusted for the effect of the fluoroscopy time
Full size table
Fig. 1
figure 1

Multiple linear regression model performed to assess the differences of DAP between the two angiographic systems, adjusted for the effect of fluoroscopy time, in patients who underwent PTBD

Full size image
Fig. 2
figure 2

Multiple linear regression model performed to assess the differences of DAP between the two angiographic systems, adjusted for the effect of fluoroscopy time, in patients who underwent cholangiography and balloon dilatation

Full size image
Fig. 3
figure 3

Multiple linear regression model performed to assess the differences of DAP between the two angiographic systems, adjusted for the effect of fluoroscopy time, in patients who underwent cholangiography and biliary catheter change or removal

Full size image

Discussion

The results of our survey show that radiation doses during biliary interventional procedures in liver transplanted children vary according to the complexity of the procedure and include a wide variation of DAP and fluoroscopy time. Complicated procedures require more irradiation time and deliver greater radiation exposure to patients compared with noncomplicated procedures.

Biliary interventional procedures in this specific pediatric cohort of liver transplant patients can be challenging even for experienced radiologists. In our series, a partial liver transplant was present in 30 of 34 patients; this inevitably poses a greater technical challenge given the small liver volume. In our cohort, sonographically evident dilatation of bile ducts was present in only 60 % of cases, thus making PTBD access more difficult to achieve. Multiple punctures and/or attempted puncture of the main duct in the perihilar region are often necessary. In rare cases, the use of a microguidewire and microcatheter, usually reserved for selective vascular catheterization, may be necessary due to the small caliber of the bile ducts. A practical issue of note is the presence of two separate bilioenteric anastomoses draining segments II and III in ≤41 % of patients with left lateral split-liver transplantation [13]. Such patients might develop BS in one or in both anastomoses. Knowledge of the number of hepaticojejunostomies performed is therefore mandatory in planning PTBD.

No deterministic or stochastic complications were observed in our experience or have been reported in the currently published literature. However, this might be due to limited long-term follow-up. The Society of Interventional Radiology–Cardiovascular Interventional Radiological Society of Europe international guideline on patient radiation management states that fluoroscopy time should not be used to monitor patient irradiation during interventional procedures; however, fluoroscopy time might provide an indication of procedure complexity, but it does not always correlate with other dose metrics [14, 15]. It is difficult to compare the results of our survey with those of other studies, mainly because the data available are limited and inhomogeneous. Data on radiation exposure is not available in the largest series currently available on biliary interventional procedures in pediatric liver recipients. Righi et al. [16] report lower DAP in three biliary procedures performed in three pediatric liver transplant recipients. However, our levels of radiation exposure is less than the established dose limits for adults proposed by Miller et al. [17], Hart et al. [18], and Kloeckner et al. [19].

Accurate potential risk estimation for stochastic effects, such as cancer and leukemia, should be quantified using the patient-specific Monte Carlo simulation for effective dose. One study in pediatric patients undergoing cardiac catheterization [20] showed that DAP is suitable for estimation of the effective dose with good accuracy. This was not feasible in our study because it was retrospective and also because inaccurate values would be derived due to the inhomogeneous X-ray beam geometry (beam direction and field size variations during the procedure). Prospective studies are therefore needed to explore this argument with more accuracy.

Some studies have reported that the X-ray dose is lower using an FPDS, whereas others have indicated that it is greater or similar to that in with conventional IIDS [21–25]. In our survey, mean DAP in the three categories of biliary procedures was statistically lower in the FPDS group compared with the IIDS group. Larger studies, with more patients and more procedures, are necessary to confirm these findings. If our findings are confirmed, using an FPDS would be in accordance with as low as reasonably achievable principles [26].

In conclusion, biliary procedures in pediatric liver transplant recipients can be challenging and might require long fluoroscopy time and large radiation exposure. In the years to come, longer follow-up will yield more information on the occurrence of deterministic or stochastic complications. The use of an FPDS, as opposed to an IIDS, should be considered for pediatric interventional radiology procedures after verification of our findings by other investigators.

Systematic recording of fluoroscopy time and radiation exposure at the end of the interventional procedure is mandatory for future studies on this issue and is also an essential step in determining local and/or general radiation exposure reference levels in this particular group of patients.