DOI: https://doi.org/10.21203/rs.3.rs-2472585/v1
Background: Dilation of common bile duct (CBD) is mostly pathological and mainly occurs secondary to mechanical causes. We aimed to explore the prevalence of CBD dilation in Intraductal Papillary Mucinous Neoplasms of the pancreas (IPMN-P) patients referred to EUS.
Methods: a retrospective study of all patients who had an EUS diagnosis of IPMN from 2011 to 2019 at Galilee Medical Center were extracted. Control group including patients with other types of pancreatic cysts.
Results: Overall, 158 patients were included in the study, 117 patients (74%) diagnosed with IPMN (group A) and 41 patients (26%) diagnosed with other pancreatic cysts (group B). CBD dilation was significantly correlated with IPMN patients as compared to group B (OR 3.8, P=0.01). Classifying IPMN to subtypes, only main duct-IPMN significantly correlated with CBD dilation (OR 40, P < 0.0001), and a trend for significance with mixed-IPMN (OR 3.6, P=0.06). Adjusting for confounders (resected gallbladder and age), main duct-IPMN remained significantly correlated with CBD dilation (P<0.001).
Conclusion: main duct-IPMN was significantly correlated with dilated CBD. Assessment of the pancreas is warranted in encountered cases of dilated CBD without obvious mechanical cause.
With the increased availability and use of cross-sectional imaging, diverse incidental findings are now diagnosed frequently including pancreatic cysts which are encountered in over 2% of computed tomography and magnetic resonance imaging examinations; a frequency that rise with age [1] [2]. About 38% of them are Intraductal Papillary Mucinous Neoplasms (IPMN-P) [3]. Typically, they are formed by intraductal neoplastic proliferations of columnar, mucin-containing cells with a variable degree of papillary formation and cyst formation [4] [5]. IPMN-P grows from cells of the main duct or the branch ducts of the pancreas, and accordingly are classified into three types based on involvement of pancreatic ducts: main duct IPMN (MD-IPMN), branch duct IPMN (BD-IPMN) and mixed type IPMN (M-IPMN) [6] [7]. MD-IPMN affects any segment or diffusely the main pancreatic duct and is more aggressive with higher malignant potential than BD-IPMN [8]. On the other hand, BD-IPMN usually arises in the uncinate process, but may involve the tail of the pancreas, but it has a lower malignant potential than MD-IPMN (9,10). Mixed duct type behaves like MD-IPMN [7] [9]. The adult common bile duct accepted upper normal diameter measurement is 6 to 7 mm. Although its width is supposed to be stable in the absence of resistance to bile flow, however it increases with age but remains within normal limits and increases post cholecystectomy [10]. A big study by Perret RS et al, reached the conclusion that in the overwhelming majority of patients 60 years of age or older the mean diameter of the CBD remained less than 6–7 mm [11]. A systematic review evaluating incidental common bile duct dilation identified a cause in about 33% where the most common causes were CBD stone, chronic pancreatitis, and periampullary diverticulum. Coexisting dilation of both bile and pancreatic ducts usually designated as double duct sign, suggests obstructing pancreatic disease, especially malignancy which may also be accompanied by obstructive jaundice [12]. Notably, common bile duct dilatation was reported in recent years to be an independent predictor of malignancy in IPMN-P patients [13]. Also a study seeking computed tomography features suggestive of malignant or invasive IPMNs identified dilatation of CBD to be more common in malignant IPMN-P [14]. However, the prevalence of CBD dilation in the different types of IPMN-P was not reported yet. A finding that may guide further investigation for CBD dilation without evident mechanical cause. The aim of our study was to explore the prevalence of CBD dilation in IPMN-P patients referred to endoscopic ultrasound (EUS) in our institution
We performed a case control retrospective study, using the databases of Galilee Medical Center. All patients above 18 years old who were referred to EUS and who had EUS morphological diagnosis of pancreatic cysts from 1.4.2011 till 1.10.2019 were included in the study. Patients were excluded if they had hepato-biliary malignancy, obstructive jaundice, chronic pancreatitis, and abnormal liver function tests. Extracted data included demographic variables (age, gender), smoking, personal or family history of malignancy, EUS findings, presence of gallbladder, CBD stones, CBD width by EUS and the morphological diagnosis. All procedures were carried out via linear echoendoscope (Pentax-Japan), model 3870 and performed by a single gastroenterologist with more than 10 years’ experience in the field of Endoscopic Ultrasound. Patients were placed in the left lateral decubitus position and were sedated with intravenous midazolam and propofol according to the decision of the endoscopist. All methods were performed in accordance with the relevant guidelines and regulations
The diagnosis of pancreatic cystic lesions in our study was based on two criteria, the morphological EUS criteria and the biochemical cyst analysis (amylase and Carcinoembryonic antigen (CEA)), as follows:
Four types of pancreatic cysts were reported in our study. The morphologic sonographic diagnosis of the pancreatic cysts in our study was made according to the following characteristics: 1) IPMNs: MD-IPMNs defined by pancreatic duct dilatation in EUS without obstructing lesion or stone while branch duct type is characterized by cystic dilation of branch ducts connected to the main pancreatic duct, whereas, mixed type represents a combination of both types [15]. 2) Mucinous cystic neoplasm (MCN) defined by multiple locules, or unilocular, most commonly located in the body or tail of pancreas and [15] [16]. 3) Serous cyst adenoma (SCA) defined by focal lesion that may be located anywhere in the pancreas resembling a honeycomb with pathognomonic central calcification in up to 20% of cases [17], and 4) Pseudocyst (PC) defined by extra pancreatic cyst with thin muddy-brown debris after an episode of moderate to severe pancreatitis [18].
The cyst fluid was analyzed for amylase and CEA level. Unfortunately, reported sensitivities and specificities of chemical analyses have broad ranges, making interpretation difficult [19] [20]. The American society of gastrointestinal endoscopy (ASGE) guidelines recommend a cut-off level of CEA of > 192 ng/mL for pancreatic mucinous cysts (MCA and IPMN) and CEA < 5 ng\mL for serous cysts. [18] However, it doesn’t recommend a cut-off level for cyst amylase. Similarly, the World Gastroenterology Guidelines, report a CEA level > 192 ng\mL for mucinous cysts: IPMN and MCN (sensitivity 73%, specificity 84%) and < 5 ng\mL for serous cysts (sensitivity 100% and specificity 86%) (WGO Global Guideline Pancreatic Cystic Lesions 2019). For PC, The European Study Group on Cystic Tumors of the Pancreas guidelines reported that low level amylase may exclude pancreatic pseudocysts (amylase < 250 U/L; sensitivity 44%, specificity 98%) [21]. Therefore, in our study, the agreement between the morphologic sonographic diagnosis and the FNA cyst fluid analysis results were set as CEA > 192 ng\Ml for mucinous cysts while CEA < 5 ng\mL for diagnosis of SCA and amylase > 250 unit\lit for PC.
Univariate descriptive statistic was used to compare patients with IPMN diagnosis and other pancreatic cystic lesions. Data was reported as mean ± standard deviation for quantitative continuous variables, and frequencies (percentages) for categorical variables. Univariate regression was used to estimate odds ratio (OR) of baseline factors and multiple linear regression analysis was used to eliminate the effect of confounders. A threshold for statistical significance was set at a P value < 0.05. All analyses were performed by an experienced statistician using the statistical analysis software (SAS Vs 9.4 Copyright (c) 2016 by SAS Institute Inc., Cary, NC, USA).
Overall, 158 patients were diagnosed with pancreatic cystic lesions according to the EUS diagnostic criteria mentioned above. We identified 117 patients (74%) diagnosed with IPMN (group A) and 41 patients (26%) diagnosed with other pancreatic cysts (group B). Among group B, 9 patients were diagnosed with MCN, 26 patients with SCA and 6 patients with PC (Fig. 1). The mean cyst levels of amylase and CEA among each cyst type in our cohort were compatible to those reported by the professional guidelines which further supporting the morphological diagnosis; for IPMN (amylase of 72204 U/L, CEA of 1325 ng\mL), for MCA (CEA of 1369 ng\mL), for SCA (amylase of 5 U/L) and for PC (amylase of 110045 U/L). The average age was 74.6 ± 10.4 and 61.1 ± 14 for groups A and B, respectively. Sixty-eight patients (58.1%) were males in group A as compared to 23 patients (56.1%) in group B. Notably, only one-fifth of patients have gallbladder stone by EUS and 80% of patients have their gallbladder in situ, while no patients had cholestatic liver diseases or CBD stones per EUS. Table 1 demonstrated the baseline characteristics the endoscopic findings.
IPMN patients |
Other pancreatic cysts type |
P value |
|
---|---|---|---|
Number of patients |
117 |
41 |
|
Age (years), mean ± SD |
74.6 ± 10.4 |
61.1 ± 14 |
< 0.001 |
Gender, N (%) • Male • Female |
49 (41.9) 68 (58.1) |
18 (43.9) 23 (56.1) |
0.8 |
Family history of pancreatic cancer, N (%) |
3 (2.6) |
1 (2.4) |
0.1 |
Smoking, N (%) |
6 (5.1) |
2 (4.9) |
0.1 |
Background liver diseases, N (%) • Cholestatic liver disease • Other liver disease (HCV) |
0 6 (5.1) |
0 0 |
0.3 |
Personal history of cancer, N (%) |
22 (18.8) |
5 (12.2) |
0.4 |
Alcohol consumption, N (%) |
1 (0.9) |
2 (4.9) |
0.1 |
Gallbladder stone by EUS, N (%) |
23 (19.7) |
3 (7.3) |
0.08 |
Signs of chronic pancreatitis, N (%) |
2 (1.7) |
2 (4.9) |
0.2 |
Gallbladder in situ, N (%) |
93 (80) |
37 (90.2) |
0.1 |
CBD dilation by EUS, N (%) |
34 (29.1) |
4 (9.7) |
0.01 |
Thirty-four patients (29.1%) in group A had dilated CBD as compared to only 4 patients (9.8%) in group B. CBD dilation was significantly correlated with IPMN patients (group A) as compared to the other pancreatic cyst types (Group B) (OR 3.8, 95% CI 1.3–11.5, P = 0.01). Classifying the IPMN group to subtypes, 16 patients (13.7%) had MD-IPMN, 76 patients (65%) had BD-IPMN and 25 patients (21.3%) had M-IPMN. When comparing each sub-type of IPMN to the other pancreatic cyst types (group B), we found that only MD-IPMN have statistically significant correlation with dilated CBD dilation (OR 40, 95% CI 7.9–200, P < 0.0001), and a trend for significance with M-IPMN (OR 3.6, P = 0.06), while no correlation was noticed with BD-IPMN (P = 0.2). Further analysis comparing the prevalence between the three IPMN groups, CBD dilation was significantly higher in the MD-IPMN patients compared to the BD-IPMN and M-IPMN (OR 19.2, P < 0.0001 for MD-IPMN vs. BD-IPMN, OR 11.1, P = 0.002 for MD-IPMN vs. M-IPMN), while no significant difference was found between BD-IPMN and M-IPMN (OR 0.6, P = 0.3) (Table 2).
Parameter |
Odds Ratio |
95% Confidence Limit |
P value |
---|---|---|---|
MD-IPMN vs. group B |
40 |
7.9–200 |
< 0.0001 |
BD-IPMN vs. group B |
2.1 |
0.64–6.8 |
0.2 |
M-IPMN vs. group B |
3.6 |
0.93–13.9 |
0.06 |
MD-IPMN vs. BD-IPMN |
19.2 |
4.8–77 |
< 0.0001 |
MD-IPMN vs. M-IPMN |
11.1 |
2.4–51.4 |
0.002 |
BD-IPMN vs. M-IPMN |
0.6 |
0.2–1.6 |
0.3 |
Three parameters showed significant correlation with CBD dilation including IPMN diagnosis (OR 3.8, P = 0.01), resected gallbladder (OR 7.7, P < 0.001) and age (OR 1, P = 0.006) (Table 3). The mean CBD width in patients with CBD dilation was significantly higher compared to patients without CBD dilation (9.64 ± 2.3 mm vs. 4.5 ± 0.7 mm). Adjusting for other explanatory confounder parameters that might be associated with CBD dilation (resected gallbladder and age) using multiple linear regression model, MD-IPMN remained significantly correlated with CBD dilation, while BD-IPMN and M-IPMN lost their association with CBD dilation (Table 4).
Parameter |
Odds Ratio |
95% Confidence Limit |
P value |
---|---|---|---|
Age |
1 |
1.01–1.08 |
0.006 |
IPMN diagnosis |
3.8 |
1.3–11.5 |
0.01 |
Gender Male vs Female |
0.4 |
0.18–0.9 |
0.2 |
Family history of pancreatic cancer |
1.3 |
0.15–11.8 |
0.8 |
Background liver disease |
1.8 |
0.3–9.8 |
0.5 |
Personal history of cancer |
2.2 |
0.9–5.3 |
0.08 |
Smoking |
0.6 |
0.09-4 |
0.6 |
Alcohol |
0.4 |
0.01–13.6 |
0.6 |
Gallbladder stone by EUS |
1.5 |
0.6–3.9 |
0.3 |
Resected gallbladder |
7.7 |
3.2–18.8 |
< 0.001 |
Parameter |
Odds Ratio |
95% Confidence Limit |
P value |
---|---|---|---|
MD-IPMN vs. group B |
25.6 |
4.6-142.9 |
0.0002 |
MD-IPMN vs. BD-IPMN |
9.6 |
4.6–83.3 |
< 0.0001 |
MD-IPMN vs. M-IPMN |
16.3 |
3-88.1 |
0.001 |
BD-IPMN vs. group B |
1.3 |
0.3–4.9 |
0.6 |
M-IPMN vs. group B |
1.6 |
0.3–7.9 |
0.6 |
BD-IPMN vs. M-IPMN |
0.8 |
0.3–2.7 |
0.7 |
Our results clearly show a strong association between IPMN-P and CBD dilatation compared to the other pancreatic cyst types with OR 3.8 and P = 0.01. After classifying IPMN into the three different subtypes, only MD-IPMN remained with statistically significant correlation with dilated CBD dilation with OR 40 and P < 0.0001, while no correlation was seen with BD-IPMN (P = 0.2). M-IPMN showed a trend for significance correlation with CBD dilation with P = 0.06). Importantly, after adjusting for age and cholecystectomy that may by their self-lead to CBD dilation, only MD-IPMN remained significantly correlated with CBD dilation. It is not uncommon for radiologists or other practitioners dealing with bile ducts to encounter incidental common bile duct dilatation. Actually, in the absence of a relevant clinical scenario where dilation is anticipated, this may pose a challenge which usually leads to a further investigation in an attempt to elucidate the cause behind this dilation. Endoscopic Ultrasound study showed that although CBD dilates after the age of 70 years, the normal CBD does not exceed 7.6 mm even in the most elderly patient with gallbladder in situ, and the single additional factor contributing to CBD dilation was cholecystectomy [22]. Probably this association originates from coexistence of IPMN-P with Intraductal Papillary Mucinous neoplasm of Bile duct (IPMB) or it reflects tow manifestations of Intraductal Papillary Mucinous Pancreato-biliary neoplasm involving the pancreatic and biliary ducts. On the other hand, it may be that the high viscosity of the pancreatic fluid and a change of its dynamics constitutes some kind of resistance to the bile duct flow leading to this dilatation [13] without actual involvement of the bile ducts in the same disease. Biliary tract IPMN (IPMB) is well known as a papillary and/or cystic lesion of the intra and extra hepatic biliary tract [23] [24]. Thus presumably, in the absence of bile duct obstruction that may lead to dilatation, we could assume that these biliary papillary tumors may be biliary counterparts of IPMN-P. Several reports have already showed that IPMN of pancreas and IPMB share some pathological, clinical and survival characteristics including the same gastrointestinal phenotypes, intraductal papillary neoplastic proliferation of mucin containing cells with occasional mucin hypersecretion and frequent progression to tubular adenocarcinoma and mucinous carcinoma [24] [25] [26] [27]. Furthermore some authors believe that IPMB and IPMN of pancreas represent different organs involvement of the same disease, as both biliary and pancreatic ducts develop from the same primordium, and this could be explained by the effect of field cancerization [28]. Strauss A et al have found that the most important implication of finding dilated CBD in IPMN patients is that it represents the best independent predictor of malignancy in these patients with a PPV of 96.4%. The addition of CBD dilation to the international guidelines improved malignancy detection significantly by over 60% without impairing specificity [13]. This finding needs thorough investigation as it may probably identify patients at high risk for malignant transformation. The other side of the same coin is that the differential diagnosis of incidental finding of dilated CBD, should include coexistent IPMN and should urge us to proceed with additional investigation by magnetic resonance imaging and/or EUS to uncover this potentially premalignant disease. The limitations were that our study was conducted in a single center by one experienced endosonographist who described only a sonographic finding of dilated CBD without reporting the relation of this finding to malignancy prediction and that we had no bile fluid analysis, neither biopsy of the bile ducts was taken, thus leaving the mechanism behind this dilation unknown. In conclusion: we found a significant novel correlation of CBD dilation with MD-IPMN. further prospective multicenter study should be performed to conform our findings, in addition, another prospective study designed to obtain biopsies of the bile duct wall to uncover the pathological basis of the duct dilation, probably by identifying mucin containing cells should be performed.
Acknowledgment: none
Author Contributions: Tawfik Khoury and Wisam Sbeit contributed to the design of the manuscript. Tawfik Khoury and Wisam Sbeit contributed to data analysis and interpretation. All authors contributed to data collection and analysis. Tawfik Khoury and Wisam Sbeit contributed to critical revision of the manuscript and all authors approved the final version to be published.
Availability of data and material: The data are found at the gastroenterology department at Galilee Medical Center with the corresponding author (Tawfik Khoury), Nahariya, Israel, and will be available upon reasonable request.
Conflict of interest: The authors declare no conflict of interest regarding this manuscript
Funding sources: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors
Ethics approval: The study was approved by the local institutional ethic committee on 13/06/2019 (0097-19-Nahariya Hospital Research (NHR)). Written informed consent was waived due to the retrospective non-interventional study design.