Congenital biliary dilatation may consist of 2 disease entities

Journal of Pediatric Surgery (2011) 46, 1503–1509

www.elsevier.com/locate/jpedsurg

Congenital biliary dilatation may consist of 2 disease entities Mei Diao a , Long Li a,⁎, Wei Cheng b,⁎⁎ a

Department of Pediatric Surgery, Capital Institute of Pediatrics, Beijing 100020, P. R. China Department of Paediatric Surgery, Monash Children's, Southern Health, Victoria, Department of Paediatrics and Department of Surgery, Faculty of Medicine, Nursing and Health Sciences, Monash University, Clayton, Victoria 3168, Australia b

Received 17 September 2010; revised 5 December 2010; accepted 20 December 2010

Key words: Congenital biliary dilatation; Pancreaticobiliary malunion; Pancreatitis; Distal common bile duct

Abstract Background/Purpose: This study aims to establish the possible mechanisms of pathogenesis of congenital biliary dilatation and to classify the disease accordingly. Methods: Radiologic features of congenital biliary dilatation and pancreaticobiliary malunion in 107 affected children were examined and correlated with laboratory results. Relative lengths/diameters were calculated to provide comparison between children of different ages. Intraluminal pressures of common bile duct (CBD) were measured intraoperatively. Results: The minimal relative diameters of distal CBD negatively correlated with the maximal relative diameters/lengths of dilated CBD, the maximal relative diameters of common hepatic duct, and left/right hepatic ducts. The intraluminal pressure in patients with a stenotic distal CBD (stenotic group) was significantly higher than that in patients with a nonstenotic distal CBD (nonstenotic group). The narrower the distal CBD, the more deranged the liver function. Conversely, serum/bile amylase levels were more elevated in the nonstenotic group. Common channel protein plugs were only found in the nonstenotic group, whereas common hepatic duct strictures, intrahepatic duct dilatations, and calculi were detected more frequently in the stenotic group. Conclusion: We propose to categorize congenital biliary dilatation into 2 subgroups: (1) cystic type with stenotic distal CBD associated with deranged liver function and common hepatic duct stricture and (2) fusiform type with nonstenotic distal CBD associated with pancreatitis and common channel protein plugs. Different underlying pathologies of each group require different operative strategies. © 2011 Elsevier Inc. All rights reserved.

⁎ Corresponding author. L. Li, Department of Pediatric Surgery, Capital Institute of Pediatrics, Beijing, P. R. China. Tel.: +86 10 85695669. ⁎⁎ Corresponding author. W. Cheng, Department of Paediatric Surgery, Monash Children's, Southern Health, Department of Paediatrics and Department of Surgery, Faculty of Medicine, Nursing and Health Sciences, Monash University, Clayton, Victoria, Australia. Tel.: +61 3 9594 5674; fax: +61 3 9594 6495. E-mail addresses: [email protected] (L. Li), [email protected] (W. Cheng). 0022-3468/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2010.12.022

Currently, pancreaticobiliary malunion (PBMU) and reflux of pancreatic juice into the common bile duct (CBD), weakening of bile duct wall, and distal CBD obstruction are thought to contribute to congenital biliary dilatation (commonly known as choledochal cyst) [1,2]. Previous study in adults with PBMU demonstrates a negative correlation between the minimal diameter of the distal CBD and the maximal diameter of proximal CBD [3]. The current study aims to assess the CBD morphologies in children with congenital biliary dilatation and to establish a

1504 possible relationship between the underlying pathology and morphology.

1. Methods One hundred seven consecutive children with congenital biliary dilatation and PBMU who underwent cyst excision and Roux-Y hepatojejunostomy between October 2000 and October 2009 were studied. Ethical approval from the Ethics Committee of Capital Institute of Pediatrics and written parental consents were obtained. Intraoperative biliary manometry was performed before cholangiography and dissection. Biliary pressure in millimeters of mercury was recorded on a manometer (Biomedical manometer BL-420F; Tai Meng Technique Co Ltd, Cheng Du, China) connected to a transducer (Biomedical transducer JH-2; Beijing Institute of Space MedicalEngineering Co Ltd, Beijing, China). A 14G needle (diameter, 2.1 mm) attached to a saline-filled tube was inserted directly into the dilated CBD. Fig. 1 illustrates intraoperative biliary manometry. The resting intraluminal CBD pressure was measured (Fig. 1A). After bile collection for amylase assessment, normal saline was infused into dilated CBD by an infusion pump (Medfusion SY-1200; Smiths Medical Co Ltd, Zhe Jiang, China) at a rate of 0.5 mL/min, and the perfusion pressure was then recorded (Fig. 1B). The intraluminal pressure increased as normal

M. Diao et al. saline was infused into the CBD until a plateau was reached. It then decreased as normal saline flowed into the common channel and duodenum. The perfusion pressure was defined as the highest pressure reached after infusing normal saline. Intraoperative cholangiogram, endoscopy (cholangioscopy and/or cholecystoscopy) was performed to establish the anatomy of intrahepatic bile ducts, junctions of CBDs, pancreatic ducts, common channels, and the presence of intraluminal protein plugs/calculi. The biliary strictures were corrected and protein plugs/calculi were removed by saline irrigation after cholangiogram and/or endoscopic investigations. Intraoperative cholangiograms were reviewed with our pediatric radiologists. Relative diameter and relative length, defined as the ratio of the bile duct diameter and length to the height of the second lumbar vertebra, were introduced to compare the measurements between children of different ages [4]. The laboratory results were correlated with the radiologic features.

1.1. Statistical analysis Data were analyzed with SPSS 13.0 software package (SPSS, Chicago, IL). Pearson correlation was used to correlate the minimal relative diameters of the distal CBD with the maximal relative diameter and maximal relative length of the dilated CBDs, the maximal relative diameter of the common hepatic duct, and left and right hepatic ducts. The Student t test was performed to compare the age, minimal relative diameter of the distal CBD, maximal relative length of the common channel, intraluminal CBD pressure, liver function parameters, and bile and serum amylase levels between groups with different relative diameters of the distal CBD and between different Todani's subtypes, respectively. χ2 test was applied to compare the morbidity of common hepatic duct stricture, intrahepatic duct dilatation and calculi, common channel protein plugs, and clinical symptoms between groups with different relative diameters of the distal CBD and between different Todani's subtypes, respectively. A P b .05 was considered statistically significant.

2. Results

Fig. 1 Illustration of intraoperative biliary manometry. A, Resting and perfusion pressures in a patient with stenotic distal CBD. B, Resting and perfusion pressures in a patient with nonstenotic distal CBD.

One hundred seven patients (male/female, 29:78; mean age, 3.96 years; range, 7 days to 18 years) with congenital biliary dilatation and PBMU were recruited into this study. According to the Todani's classification, there were 92 type I and 15 type IV cysts. No type II or III was found in our series. Patients with Caroli disease were excluded from the current study. The relationship between the CBD outlet size and the cyst morphology was firstly examined. We found that the minimal relative diameters of distal CBD (CBD outlet) negatively correlated with the maximal relative diameter and length of the dilated CBD, the maximal relative

Congenital biliary dilatation

1505

Fig. 2 Correlations between the minimal relative diameters of distal CBDs and the maximal relative diameters of dilated CBDs (A), the maximal relative lengths of dilated CBDs (B), the maximal relative diameters of common hepatic ducts (C), the maximal relative diameters of left hepatic ducts (D), and the maximal relative diameters of right hepatic ducts (E).

diameter of common hepatic duct, and left and right hepatic ducts (Fig. 2A-D, E; R = −0.98, −0.94, −0.93, −0.97, −0.96, P b .001, respectively). In summary, the narrower the CBD outlet, the more severe the CBD dilatation.

We adopted the definition of incomplete obstructive jaundice in the current national Clinical Textbook of Medical Diagnosis, that is, a total bilirubin of 171 μmol/L or greater [5], and arbitrarily divided our patients into 2 groups (≥171

1506

M. Diao et al.

Fig. 3 Morphologic features of congenital biliary dilatation with stenotic vs nonstenotic distal CBDs. A, Congenital biliary dilatation with stenotic distal CBD. B, Congenital biliary dilatation with nonstenotic distal CBD.

or b171 μmol/L). We found the minimal distal CBD relative diameter of the patients with a total bilirubin of 171 μmol/L or greater (0.13 ± 0.05) was significantly narrower than that of the patients with total bilirubin less than 171 μmol/L (0.27 ± 0.06, P b .001). We hence labeled these groups stenotic and nonstenotic (Fig. 3A, B). We found that the patients from the stenotic group presented at a much younger age than those in the nonstenotic group (2.44 ± 2.06 years vs 7.53 ± 3.92 years, P b .001). All 11 neonatal patients belonged to the stenotic group. Seven of them were diagnosed antenatally. The maximal relative length of the common channel in the stenotic group was 1.45 ± 0.43, which was significantly longer than 0.69 ± 0.25 in the nonstenotic group (P b .001). The CBD intraluminal resting and perfusion pressures of the stenotic group (30 ± 8 and 65 ± 16 mm Hg) were significantly higher than those of the nonstenotic group (10 ± 8 and 52 ± 13 mm Hg; Table 1, P b .001). The relative diameter of the distal CBD also influenced the liver function tests; that is, the narrower the distal CBD, the more deranged the liver function (Table 2, P b .001). On the other hand, the nonstenotic group was associated with significantly elevated serum and bile amylase levels (Table 1, P b .001) indicative of pancreatitis. In addition to the derangements of liver functions, CBD outlet size also appeared to determine the ductal pathologies. Protein plug formation was detected in the common channels in 63% (20/32) of the patients in the nonstenotic group but were absent in the 75 patients examined from the stenotic group (Table 3, P b .001). On the other hand, combined intrahepatic and extrahepatic bile duct dilatation (Todani's type IV) was noted in 11 (15%) of 75 patients in the stenotic group. Nine (82%) of the 11 patients had additional common hepatic duct strictures, and 5 of them (56%) had intrahepatic calculi. Todani's type IV pathology was found in 4 (13%) of 32 patients in the nonstenotic group, but no intrahepatic calculi was detected in this group (Table 3, 7% vs 0%, P = .14, not statistically significant). Common hepatic duct stricture occurred more frequently in the stenotic group than in the nonstenotic group (Table 3, 12% vs 0%, P = .04). Although the frequency of abdominal pain in each group was similar (stenotic, 69% [52/75]; nonstenotic, 72% [23/32]; P = .79, Table 3), more patients from the stenotic group presented with jaundice (87%) and a palpable mass

Table 1 The mean intraluminal resting and perfusion pressures, bile, and serum amylase levels in children with congenital biliary dilatations New classification

Resting pressure (mm Hg) Perfusion pressure (mm Hg) Bile amylase (U/L) SAMY (U/L), rr: 25-125 U/L

Todani's classification

Stenotic group (n = 75)

Nonstenotic group (n = 32)

P

Type I (n = 92)

Type IV (n = 15)

P

30 ± 8 65 ± 16 2398 ± 1173 104 ± 77

10 ± 8 52 ± 13 77,354 ± 43,759 660 ± 212

b.001 b.001 b.001 b.001

21 57 21,774 261

40 ± 6 90 ± 16 43,463 ± 73,624 328 ± 430

b.001 b.001 .28 .57

SAMY indicates serum amylase; rr, reference range.

± 10 ± 11 ± 33,729 ± 259

Congenital biliary dilatation Table 2

1507

The mean liver function parameters in children with congenital biliary dilatations New classification

TB (μmol/L), rr: 3.4-20 μmol/L ALT (U/L), rr: b40 U/L AST (U/L), rr: b40 U/L ALP (U/L), rr: b400 U/L GGT (U/L), rr: 7-50 U/L

Todani's classification

Stenotic group (n = 75)

Nonstenotic group (n = 32)

P

Type I (n = 92)

Type IV (n = 15)

P

225 141 144 619 389

73 73 80 509 188

b.001 b.001 b.001 b.001 b.001

177 ± 83 119 ± 50 124 ± 48 584 ± 79 325 ± 141

199 ± 71 130 ± 46 129 ± 42 597 ± 70 353 ± 126

.34 .47 .70 .56 .46

± 35 ± 45 ± 44 ± 70 ± 124

± 56 ± 10 ± 13 ± 18 ± 16

TB indicates total bilirubin; ALT, alanine transaminase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; GGT, γ-glutamyl transpeptidase; rr, reference range.

(23%) than those from the nonstenotic group (31% and 6%, P b .001 and P = .04, respectively). We compared our new classification with that of Todani's. Although both classifications predict the intraluminal pressures of the bile ducts, the new classification, in addition, was able to differentiate the relative length of the common channel, changes in liver function, pancreatitis, presenting symptoms, and age (Tables 1-3) between the groups. Todani's classification, however, better predicted intrahepatic dilatation, presence of stones, and common hepatic duct stricture (Table 3). The median follow-up period of stenotic and nonstenotic groups was 47 months (range, 14-101 months) and 42 months (range, 15-99 months), respectively. As of this date, no instances of postoperative cholestasis, intrahepatic stone formation, cholangitis, protein plugs/calculi formation in common channels, or pancreatitis were observed in either group.

3. Discussion Our study corroborated with a previous study that showed the relationship between distal CBD diameter and dilated

CBD in adults [3]. Moreover, our results demonstrated that CBD outlet size determines the associated pathology. The stenotic type is associated with more severe CBD dilatation (often cystic in morphology), higher CBD pressure, more deranged liver functions, intrahepatic ductal dilatation, and calculi (Fig. 4). On the other hand, the nonstenotic type (often fusiform in morphology) is associated with possible bidirectional bile/pancreatic juice reflux, pancreatitis, protein plugs, and calculi formation in the common channels (Fig. 4). A previous study comparing 13 children with cystic CBD dilatation and 12 children with fusiform CBD dilatation showed that the cystic group was younger, and the serum bilirubin, aspartate aminotransferase, γ-glutamyl transpeptidase, alkaline phosphatase, and intraluminal pressure were higher than those with fusiform dilatation. The differences were not statistically significant, probably because of the small sample size [6]. In our preliminary study with a small group of patients, a significant higher intraluminal pressure, narrower distal CBD, and lower bile amylase level were found in children with cystic CBD dilatation compared with those with fusiform CBD dilatation [7]. Our study, with a larger sample size, confirmed the findings observed in the previous studies.

Table 3 Patients' age at presentation, biliary morphologies, complications, and clinical symptoms in children with congenital biliary dilatations New classification

Age (y) Relative diameter of distal CBD Relative length of common channel Intrahepatic duct dilatations Common hepatic duct strictures Intrahepatic duct stone formation Protein plugs in common channel Clinical Abdominal pain symptoms Jaundice Palpable mass

Todani's classification

Stenotic group (n = 75)

Nonstenotic group (n = 32)

P

Type I (n = 92)

Type IV (n = 15)

2.44 ± 2.06 0.13 ± 0.05 1.45 ± 0.43

7.53 ± 3.92 0.27 ± 0.06 0.69 ± 0.25

b.001 b.001 b.001

4.00 ± 3.38 0.17 ± 0.08 1.21 ± 0.53

3.74 ± 4.88 0.16 ± 0.09 1.29 ± 0.41

.80 .58 .55

11 (15%) 9 (12%) 5 (7%) 0 52 (69%) 65 (87%) 17 (23%)

4 (13%) 0 0 20 (63%) 23 (72%) 10 (31%) 2 (6%)

.77 .04 .14 b.001 .79 b.001 .04

0 0 0 16 65 60 13

15 (100%) 9 (60.0%) 5 (33.3%) 4 (27%) 10 (67%) 15 (100%) 6 (40%)

b.001 b.001 b.001 .39 .76 b.01 .02

(17%) (71%) (65%) (14%)

P

1508

M. Diao et al.

Fig. 4 Working model of the possible pathogenesis of congenital biliary dilatation and proposed surgical strategies based on the new classification. CHD indicates common hepatic duct; IHD, intrahepatic duct; BPJ, biliary-pancreatic junction.

The different underlying pathologies of each type require different operative strategies (Fig. 4). In the stenotic type, the cyst excision should be extended to the level of the stenotic segment. The distal CBD stump is stenotic, and ligation is optional. In our series where no distal CBD ligation was performed in the stenotic type, no postoperative pancreatic juice leak was encountered. However, proximal bile duct fashioning and irrigation are required to minimize postoperative cholestasis, stone formation, and cholangitis [8]. On the other hand, in the nonstenotic type, excision should be performed just proximal to the biliary-pancreatic junction to avoid injury to the pancreatic duct. The nonstenotic distal CBD stump must be ligated to prevent pancreatic juice leak. Careful exploration and irrigation of the common channel are necessary to minimize recurrent postoperative pancreatitis [8]. Our medium-term follow-up results suggest that postoperative complications were minimized if the morphology-based surgical strategy was adhered to. The mean age of the stenotic group was significantly younger than that of the nonstenotic group, and all neonatal CBD dilatations were associated with stenotic distal CBD, suggesting that there may be 2 distinct but overlapping entities of congenital biliary dilatations. This finding is in agreement with the previous report of a high incidence of the cystic form of CBD dilatation in neonates (cystic/fusiform ratio, 20:1 in neonates vs 1.67:1 in patients of all ages) [9]. Contrary to the current etiological theory, PBMU and reflux do not apply to all congenital biliary dilatations, especially the cystic CBD dilatation. Animal experiments also suggest a heterogenous pathogenesis of congenital biliary dilatation. Ligation of the distal CBD in lambs led to cystic CBD dilatation, reminiscent of the cystic biliary dilatations with stenotic outlet noted in our series [10]. The increased intraluminal pressure in patients with a stenotic distal CBD prevented reflux of pancreatic juice into the biliary system, further contributing to the choledochal cystic dilatation,

intrahepatic ductal dilatation, and deranged liver function tests. In our series, significantly longer common channels were found in the stenotic group. We speculate that the stenotic (cystic) subtype may be formed at an early fetal stage. The hepatic diverticulum (the future ampulla of Vater) is ectopically located in a more distal position in the descending duodenum (D2) [4]. Fetal growth may elongate the common channel. The weak bile duct wall and immature hepatic parenchyma in the early embryonic phase may contribute to extrahepatic and intrahepatic bile duct dilatation under high intraluminal pressure (Fig. 4). On the other hand, choledochopancreaticostomy in puppies was found to be associated with only mild chemical reaction and resulted in fusiform rather than cystic CBD dilatation [11]. This model is reminiscent of our patients with fusiform biliary dilatation and a nonstenotic distal CBD. Evidence for reflux of pancreatic juice into the CBD was the elevated amylase level in the bile of patients with a nonstenotic distal CBD. The fusiform subtype may be formed at a later developmental stage, accelerated by the pancreatic reflux through the common channel and the nonstenotic distal CBD, at a time when the mature bile duct wall and hepatic parenchyma are less pliable. The elevated serum amylase level and pancreatitis are probably caused by reflux of bile into the pancreatic duct. Clinically, the bidirectional reflux most likely contributes to cholangitis and fusiform dilatation of CBD on one hand and pancreatitis and protein plugs formation in the common channel on the other (Fig. 4). Conventional Todani's classification categorizes congenital biliary dilatations according to their morphologies. However, Todani's type II and type III pathology are rarely encountered in clinical practice. Caroli disease is a different disease entity that requires different management. In addition, Todani's classification is descriptive and does not address the underlying mechanisms of congenital biliary dilatation. Hence, it fails to provide a surgical guideline specific to the subtypes. The proposed classification better

Congenital biliary dilatation predicts liver function, pancreatitis, common channel length, patients' age, and the intrahepatic biliary pathology. It simplifies the traditional classification. Moreover, it provides an evidence-based management strategy for congenital biliary dilatations.

References [1] Babbitt DP, Starshak RJ, Clemett AR. Choledochal cyst: a concept of etiology. Am J Roentgenol Radium Ther Nucl Med 1973;119(1): 57-62. [2] Glenn F, McSherry CK. Congenital segmental cystic dilatation of the biliary ductal system. Ann Surg 1973;177(6):705-13. [3] Nomura T, Shirai Y, Wakai T, et al. Narrow portion of the terminal choledochus is a cause of upstream biliary dilatation in patients with anomalous union of the pancreatic and biliary ducts. World J Gastroenterol 2005;11(41):6503-7.

1509 [4] Li L, Yamataka A, Wang YX, et al. Ectopic distal location of the papilla of Vater in congenital biliary dilatation: implications for pathogenesis. J Pediatr Surg 2001;36(11):1617-22. [5] Chen WB, Wang YC, Wang HD, et al. Liver function tests. In: Chen WB, Wang YC, editors. Clinical textbook of medical diagnosis. 5th ed. Beijing: People's Health Publishing House; 2001. p. 400. [6] Davenport M, Basu R. Under pressure: choledochal malformation manometry. J Pediatr Surg 2005;40(2):331-5. [7] Li L, Wang DY, Chen YC, et al. The changes of morphology and intraluminal pressure in choledochal cyst. Chin J Pediatr Surg 2000;21(4):214-6. [8] Miyano T, Yamataka A, Kato Y, et al. Choledochal cysts: special emphasis on the usefulness of intraoperative endoscopy. J Pediatr Surg 1995;30(3):482-4. [9] Lane GJ, Yamataka A, Kohno S. Choledochal cyst in the newborn. Asian J Surg 1999;22:310-2. [10] Spitz L. Experimental production of cystic dilatation of the common bile duct in neonatal lambs. J Pediatr Surg 1977;12(1):39-42. [11] Miyano T, Suruga K, Suda K. “The choledocho-pancreatic long common channel disorders” in relation to the etiology of congenital biliary dilatation and other biliary tract disease. Ann Acad Med Singapore 1981;10(4):419-26.