Introduction

The pathogenesis of choledochal cyst (CDC) is still in controversy. Spitz [1] successfully established an experimental model of cystic dilatation by ligation of distal end of CBD in newborn lambs, indicating obstruction of distal CBD is one of the etiologies [2, 3]. PBMU is a common pathological change. It allows free regurgitation of pancreatic juice to CBD, which may cause inflammation, destroy biliary epithelium and underlying muscle, and result in weakness of cyst wall, and eventually CBD dilatation and/or distal stenosis [4]. Therefore, PBMU was considered as one of the etiologies. However, the hypothesis is challenged. First, fusiform or cylindrical dilatation was only constructed in animal models of PBMU [5, 6]. Second, more severe reflux merely induced milder dilatation, while bile amylase level was lower with more dilated cysts [7]. Additionally, CBD dilatation does not occur in some patients with PBMU [8]. And for neonates and infants, secretory response of pancreas to stimulation is extremely poor, and the concentrations of amylase and lipase in duodenal fluid are negligible [9, 10].

Papilla of Vater represents embryonic hepatic diverticulum, and occurs prior to the development of pancreaticobiliary ductal union [11]. Li and Yamataka [12] first reported in 2001 that ectopic distal location of papilla of Vater (EDLPV) was closely associated with PBMU and proposed that ectopic distal budding of hepatic diverticulum in early embryos was one of the pathogeneses of CDC [12,13,14,15]. In addition, there were some studies on correlation between PBMU and EDLPV [16, 17]. Kim [18] also reported a 30-month-old CDC patient with EDLPV in 2011. The purpose of this study was to investigate the correlation between EDLPV and clinical characteristics of CDCs.

Patients and methods

Patients

In this retrospective study, CDC patients underwent definitive operation at Children’s Hospital Capital Institute of Pediatrics between January 2020 and March 2022 were reviewed. The exclusion criteria were as follows: 1. Location of papilla of Vater was indistinguishable by intraoperative cholangiography (IOC), e.g., papilla location was obscured by a large cyst; and 2. External drainage or endoscopic retrograde cholangiopancreatography (ERCP) was performed before surgery. The study was approved by Ethical Committee of Capital Institute of Pediatrics.

Methods

Preoperative abdominal ultrasonography, computed tomography (CT), laboratory tests, and IOC were conducted routinely. Follow information was extracted from electronic medical record system: demographics, clinical manifestations, imaging examination results, and laboratory test results.

Papilla location was determined by the surgeon according to IOC during operation and reviewed by another researcher before analysis. For prenatally diagnosed patients, age at onset, expressed as a negative number, was defined as the period from the time of prenatal diagnosis to the date of birth, and duration of disease was defined as the period from the time of prenatal diagnosis to the date of surgery. Advanced fibrosis was defined as Ludwig stage of 3 or 4 [19]. The sizes of cyst, common hepatic duct (CHD), common channel (CC), and intrahepatic bile duct (IHD) were measured first according to IOC, secondly according to ultrasonography and CT. To eliminate the effect of age and magnification on parameters, a relative index was calculated as previously described [12]. Sludge and protein plugs were comprehensively detected by ultrasonography, CT, IOC, and intraoperative irrigation.

Patients were categorized into 3 groups based on location of major papilla (Fig. 1): Group 1 (G1): papilla in the middle third of second part of duodenum; Group 2 (G2): papilla in the distal third of second part, the junction between second and third part, and the beginning of third part of duodenum; and Group 3 (G3): papilla from the middle of third part to fourth part of duodenum.

Fig. 1
figure 1

Location of papilla of Vater by cholangiography: A papilla in the middle of second part of duodenum (Group 1); B papilla in the junction between second and third part of duodenum (Group 2); C papilla in the beginning of third part of duodenum (Group 2); D papilla in the middle of third part of duodenum (Group 3); E papilla in the junction between third and fourth part of duodenum (Group 3); F papilla in the fourth part of duodenum (Group 3). Arrows indicated the location of papilla

Full size image

Statistical analysis

Statistical analysis was carried out using SPSS version 24.0 (IBM, Armonk, NY, USA). Normally distributed continuous variables expressed as mean ± SD were compared by ANOVA, while non-normally distributed variables expressed as interquartile range were compared by nonparametric KruskalWallis test. Categorical variables described as proportions were compared by χ2 test or Fisher’s exact test. p < 0.05 was considered statistically significant.

Results

Demographic data

Between January 2020 and March 2022, a total of 380 CDCs underwent definitive surgery in our center, and 53 patients were dropped out in accordance with exclusion criteria (8 for nonvisible papilla location, 31 for external drainage, and 14 for ERCP), leaving 327 for the study.

Of 327 patients, 38 (11.62%) were allocated to G1, 168 (51.38%) to G2, and 121 (37.00%) to G3 (Table 1). EDLPV was observed in 289 (G2 and G3, 88.38%).

Table 1 Comparison among three groups in clinical data
Full size table

Age

Median age at onset was − 3.50 (1.63, 30.80) months. At onset, G3 patients were much younger than both G1 (p < 0.001) and G2 (p < 0.001) patients (G1 vs. G2 vs. G3, 20.52 vs. 19.47 vs. − 3.40 months, p < 0.001) (Table 1). Median age at surgery was 20.23 (1.87, 44.93) months. G3 patients were markedly younger at surgery than both G1 (p < 0.001) and G2 (p < 0.001) patients (G1 vs. G2 vs. G3, 31.18 vs. 27.75 vs. 3.43 months, p < 0.001) (Table 1). There were no significant differences between G1 and G2 patients (age at onset, p = 1.000; and age at surgery, p = 0.967).

Prenatal diagnosis

One hundred and forty-seven patients (44.95%) were diagnosed prenatally (G1 vs. G2 vs. G3, 26.32% vs. 36.31% vs. 62.81%, p < 0.001) (Table 1). A significantly greater proportion of prenatal diagnosis was demonstrated in G3 compared with both G1 (p < 0.001) and G2 (p < 0.001). However, a significant difference was not displayed between G1 and G2 (p = 0.242).

Cyst size

Mean relative cyst length was 3.35 ± 1.61 (G1 vs. G2 vs. G3, 2.38 vs. 2.93 vs. 4.25, p < 0.001) (Table 1). G3 patients had approximately 1.79-fold longer cysts than G1 patients (p < 0.001) and 1.45-fold longer than G2 patients (p < 0.001), and G2 patients had approximately 1.23-fold longer cysts than G1 patients (p = 0.035).

Mean relative cyst diameter was 1.93 ± 1.25 (G1 vs. G2 vs. G3, 1.18 vs. 1.60 vs. 2.62, p < 0.001) (Table 1), which was largest in G3 patients (p < 0.001 vs. G2), second in G2 patients (p = 0.041 vs. G1), and smallest in G1 patients (p < 0.001 vs. G3).

Two hundred and sixty-seven patients (81.65%) had cystic dilatation (G1 vs. G2 vs. G3, 65.79% vs. 78.57% vs. 90.91%, p = 0.001) (Table 1). G3 patients more frequently had cystic dilatation compared with both G1 (p < 0.001) and G2 (p = 0.005), while there was no statistical discrepancy between G1 and G2 (p = 0.095).

Size of common channel

Mean relative CC length was 0.65 ± 0.34 (G1 vs. G2 vs. G3, 0.52 vs. 0.69 vs. 0.64, p = 0.022) (Table 1). Compared with G1 patients, G2 patients had a longer CC (p = 0.007) and G3 patients tended to have a longer CC (p = 0.099), while G3 and G2 patients had comparable lengths (p = 0.225). No significant differences were noted in relative CC diameter among three groups (p = 0.191).

Diameter of common hepatic duct

Mean relative CHD diameter was 0.45 ± 0.31 (G1 vs. G2 vs. G3, 0.36 vs. 0.42 vs. 0.52, p = 0.012) (Table 1). There was marked dilatation in G3 patients compared to G1 patients (p = 0.014) and to G2 patients (p = 0.015), while there was no significant difference between G1 and G2 patients (p = 0.281).

Diameter of intrahepatic bile duct

Of 327 patients, whether IHD diameter is > 3 mm can be determined in 258 (> 3 mm, 85.66%, 221/258; G1 vs. G2 vs. G3, 91.96% vs. 84.62% vs. 84.78%, p = 0.541), and whether ≥ 8 and 10 mm in 268 (≥ 8 mm, 30.97%, 83/268, G1 vs. G2 vs. G3, 31.43% vs. 38.13% vs. 20.21%, p = 0.015; ≥ 10 mm, 23.88%, 64/268, G1 vs. G2 vs. G3, 22.86% vs. 30.94% vs. 13.83%, p = 0.011).

Compared with G3, G2 patients were inclined to have dilatation of IHD (≥ 8 mm, p = 0.004; and ≥ 10 mm, p = 0.003, respectively) (Table 1). The rates of diameter ≥ 8 mm and ≥ 10 mm in G1 had tendency to be higher compared with G3 (≥ 8 mm, p = 0.180; and ≥ 10 mm, p = 0.217, respectively) and be lower compared with G2 (≥ 8 mm, p = 0.462; and ≥ 10 mm, p = 0.348, respectively), but the differences were not significant. However, a comparison of relative IHD diameter among three groups did not reveal significant differences (G1 vs. G2 vs. G3, 0.39 vs. 0.44 vs. 0.49, p = 0.194).

Sludge and protein plugs

Sludge was detected in 255 patients in CBD (77.98%). The incidence tended to be higher in G3, although no significant differences were noted among three groups (G1 vs. G2 vs. G3, 71.05% vs. 75.00% vs. 84.30%, p = 0.093) (Table 1). G3 patients had a higher incidence than G2 and G1 (G3 vs. G1 + G2, 84.30% vs. 74.27%, p = 0.035).

Intrahepatic sludge was found in 38 patients (11.62%) (G1 vs. G2 vs. G3, 7.89% vs. 7.74% vs. 18.18%, p = 0.018) (Table 1). G3 patients had a higher incidence compared with G2 (p = 0.007), and no significant differences between G1 and either G2 (p = 1.000) or G3 (p = 0.129) were observed.

Protein plugs in CC were observed in 102 patients (31.19%). The occurrence in G3 was distinctly lower compared with both G2 (16.53% vs. 38.69%, p < 0.001) and G1 (16.53% vs. 44.74%, p < 0.001) (Table 1), with no significant difference between G1 and G2 (p = 0.492).

Perforation

Prevalence of perforation was 11.01% (36/327) (G1 vs. G2 vs. G3, 2.63% vs. 9.52% vs. 15.70%, p = 0.055). Compared with G2 and G1, G3 patients had a higher risk (G3 vs. G1 + G2, 15.70% vs. 8.25%, p = 0.035). G3 patients tended to have a higher risk than both G2 and G1, although there were no significant differences among three groups (Table 1).

Clinical manifestations

Fewer G3 patients complained of pancreatitis, vomiting, and abdominal pain compared with both G1(9.92% vs. 28.95%, p = 0.004; 38.84% vs. 63.16%, p = 0.009; and 29.75% vs. 68.42%, p < 0.001, respectively) and G2 (9.92% vs. 27.98%, p < 0.001; 38.84% vs. 65.48%, p < 0.001; and 29.75% vs. 58.33%, p < 0.001, respectively) (Table 1), and differences between G1 and G2 did not reach significance (p = 0.904, p = 0.787, and p = 0.251, respectively).

Liver function

Detailed data were demonstrated in Table 2. The levels of total bilirubin and directed bilirubin increased progressively from G1 to G2 patients (G1 vs. G2, p = 0.013, p = 0.007, respectively), reaching the highest values in G3 patients (G2 vs. G3, p < 0.001, p < 0.001, respectively). The levels of γ-glutamyl transferase (GGT) and total bile acid (TBA) were significantly elevated in G3 patients compared with both G1 (p = 0.001, p = 0.025, respectively) and G2 (p < 0.001, p = 0.040, respectively) patients; and the levels of albumin, amylase, and prealbumin were markedly lower in G3 patients compared with both G1 (p = 0.002, p < 0.001, and p < 0.001, respectively) and G2 (p < 0.001, p < 0.001, and p = 0.003, respectively) patients; however, there were no significant differences when G1 patients were compared with G2 patients (GGT, p = 1.000; TBA, p = 0.830; albumin, p = 1.000; amylase, p = 1.000; and prealbumin, p = 0.386, respectively). Cholinesterase level was significantly lower in G3 patients compared with G2 patients (p = 0.046), while no significant differences were noted between G1 and either G2 (p = 1.000) or G3 (p = 0.214) patients. There were no significant differences in the levels of alanine aminotransferase, aspartate aminotransferase, and alkaline phosphatase among three groups (p = 0.708, p = 0.386, and p = 0.449, respectively).

Table 2 Comparison among three groups in biological and pathological data
Full size table

Prevalence of advanced fibrosis was 10.19% (33/327). Advanced fibrosis had a tendency to be associated with a more distal location of papilla. Especially for prenatally detected patients (n = 147), G3 patients had a higher risk compared with G2 (13.16% vs. 1.67%, p = 0.015) (Table 2). The same trend was observed for postnatally detected patients, but the values did not reach significance (p = 0.267).

Discussion

Current study discovered that the more distal location of papilla, the larger cysts, echoing our previous report [12]. In addition, EDLPV was associated with the morphology of cyst, corresponding to our previous report [15]. The more distal location of papilla, the more elongated CBD, as result of which CBD wall might get weaker and distal end narrower. Longer cysts possibly mechanically increased resistance to bile flow. Distal obstruction might eventually become heavier anatomically and mechanically, giving rise to more dilatation of CBD.

A longer CC was also correlated with EDLPV. In our study, relative CC length was longer in G3 and G2 compared with G1 (G1 vs. G2 + G3, 0.52 ± 0.24 vs. 0.67 ± 0.3512, p = 0.008), consistent with our previous investigation [12]. PBMU issues from a long CC, indicating PBMU correlated with EDLPV. Based on our current data, with the exception of no significant difference between patients with papilla in the junction between third and fourth part of duodenum and G1 patients (0.32 ± 0.17 vs. 0.52 ± 0.24, p = 0.111), others with EDLPV had longer CC (0.69 ± 0.35 vs. 0.52 ± 0.24, p = 0.006). For some patients with a large cyst, CC might be obscured, especially for those with papilla in the junction between third and fourth part and the fourth part of duodenum. Missing data may lead to bias.

Critically, given that patients with more distal location of papilla were most frequently detected prenatally, CBD dilatation occurred during fetal life, and cysts were large enough that they were most diagnosed prenatally. While patients with more proximal location were most frequently diagnosed postnatally due to relatively smaller cysts. Furthermore, patients with more distal location were much younger, for location of papilla determined at 22 days of gestation [11]. Both abnormal embryonic development and relatively larger cysts contribute to younger age at onset and surgery.

Despite the fact that both G3 and G2 patients had EDLPV, there were plenty of intriguing disparities between them. The heavier distal obstruction might give rise to less reflux of bile into pancreatic duct, thereby decreasing the occurrences of pancreatitis, abdominal pain, and vomiting. Meanwhile, the composition ratio of bile to pancreatic juice mixed in CC might be relatively lower, probably contributing to a lower incidence of protein plugs in CC. More severe distal obstruction and younger age, when patients tend to be more pliable, might be responsible for higher rates of cystic dilatation, intrahepatic sludge, and advanced fibrosis, more elevated biliary obstructive indicators, and lower levels of cholinesterase [20], albumin, and prealbumin.

Surprisingly, with a larger CBD and CHD, G3 patients had a lower occurrence of IHD dilatation compared with G2, but IHD diameter was not smaller. More severe distal obstruction might result in a more dilated cyst and CHD, which, in the meantime, might serve as a cushion and relieve pressure on IHD. Weakness of CBD wall might be associated with younger age [1]. In the experimental study by Miyano [2], fusiform dilatation with IHD dilatation was induced by ligation of distal CBD in mature animals (10 dogs, 5 rabbits, and 10 rats), while cystic dilatation without IHD dilatation in 10 immature rats. In our study, IHD dilatation was less common in G3 patients, but the diameter of dilated IHD was larger when the pressure in cyst was sufficiently high. Relatively slighter obstruction and buffering, and older age in G1 patients might be the reason why there was no significant difference compared to G3.

Hepatic diverticulum, arising from ventral primitive foregut, develops into bile duct, ventral pancreas, and liver, while dorsal pancreas originates from the opposite during embryonic development [11]. Location of papilla of Vater depicts the budding of hepatic diverticulum in early embryonic life. As duodenum rotates and becomes C-shaped, ventral pancreas fuses with dorsal pancreas [21]. Meanwhile, superior branch of ventral pancreas joining dorsal pancreatic duct forms main pancreatic duct [22]. Major papilla generally enters the left-lateral-posterior middle wall of descending portion of duodenum.

Ectopic distal budding of hepatic diverticulum, resulting in a longer distance from ventral pancreas to hepatic plate and to dorsal pancreas, might be responsible for the stretch and weakness of CBD and CC, giving rise to distal stenosis and a longer CC, ultimately, dilatation and PBMU [12, 13]. Therefore, we postulated that EDLPV might be one of the pathogeneses of CDC. Further research is required to validate and explore the underlying molecular mechanism.

In conclusion, EDLPV was closely correlated with clinical characteristics of CDCs. Patients with a more distal location of papilla had longer cysts and common channels, more severe clinical features, considerably younger age, and a higher proportion of prenatal diagnosis.