Differences in phosphatidylcholine and bile acids in bile from Egyptian and UK patients with and without cholangiocarcinoma

Abstract

Background

Cholangiocarcinoma (CC) is a fatal malignancy, the incidence of which is increasing worldwide, with substantial regional variation. Current diagnostic techniques to distinguish benign from malignant biliary disease are unsatisfactory. Metabolic profiling of bile may help to differentiate benign from malignant disease. No previous studies have compared the metabolic profiles of bile from two geographically and racially distinct groups of CC patients.

Objectives

This study aimed to compare metabolic profiles of bile, using in vitro proton magnetic resonance spectroscopy, from CC patients from Egypt and the UK, and from patients with CC and patients with non-malignant biliary disease.

Methods

A total of 29 bile samples, collected at cholangiography, were analysed using an 11.7-T system. Samples were from eight CC patients in either Egypt (n = 4) or the UK (n = 4) and 21 patients with benign biliary disease (choledocholithiasis [n = 8], sphincter of Oddi dysfunction [n = 8], primary sclerosing cholangitis [n = 5]).

Results

Bile phosphatidylcholine (PtC) was significantly reduced in CC patients. Egyptian CC patients had significantly lower biliary PtC levels compared with UK patients. Taurine- and glycine-conjugated bile acids (H-26 and H-25 protons, respectively) were significantly elevated in bile from patients with CC compared with bile from patients with benign diseases (P = 0.013 and P < 0.01, respectively).

Conclusions

Biliary PtC levels potentially differentiate CC from benign biliary disease. Reduced biliary PtC in Egyptian compared with UK patients may reflect underlying carcinogenic mechanisms.

Introduction

Cholangiocarcinoma (CC) is the second most common primary hepatic tumour worldwide and its incidence rates have risen steeply around the world over the past 30 years.^1–7 Cholangiocarcinoma is usually fatal as a result of its late presentation, the difficulty of diagnosis and lack of a curative therapy other than surgical resection. Its diagnosis is challenging as there are no sensitive and specific biomarkers and imaging cannot accurately differentiate benign from malignant biliary disease. Most cases present too late for resection and, consequently, CC has a poor prognosis.^1,2 Most known risk factors for CC are associated with chronic inflammation of the biliary tree.^2,7,8 Primary sclerosing cholangitis (PSC) is the most common known predisposing condition for CC in Western nations.⁹ In parts of the world where CC is most common, such as in Thailand, the malignancy has been linked to chronic infestation of the biliary tree with liver flukes. However, most patients do not exhibit any known risk factors and there are no known markers of risk for the development of biliary malignancy. Rates of CC vary greatly among different areas of the world, possibly reflecting the distribution of risk factors and their interaction with host genetic differences that may increase risk for carcinogenesis.^2,7,8 We postulate that an imbalance of lipids and toxic bile acids in bile may have a pathogenic role in cholangiocarcinogenesis.

Magnetic resonance spectroscopy (MRS) allows the study of metabolite changes in biofluids and tissues. We, and others, have previously reported that in vitro MRS of bile may be able to differentiate between benign and malignant biliary tract disease.^9–11 We have also shown that bile acid levels are significantly higher, and phosphatidylcholine (PtC) levels significantly reduced, in bile from UK CC patients compared with bile from UK patients with benign biliary disease.¹⁰ There are no published studies to date investigating the metabolic profiles of bile from Egyptian CC patients, or comparing the bile profiles of two racially distinct patient groups. Using in vitro MRS, the objectives of this study were to:

1. compare the metabolic profiles of bile from CC patients from Egypt and the UK;

2. compare the metabolic profiles of bile from CC patients and bile from patients with non-malignant biliary disease, and

3. identify potential biomarkers of CC.

Materials and methods

Patients

Sample collection was organized by authors MSHA and MMEC, following relevant local ethics committee approval. Bile samples from 29 patients were acquired at endoscopic retrograde cholangiopancreatography (ERCP). Eight patients had CC. Four of these were Egyptian and underwent ERCP at the National Liver Institute, Egypt, carried out by GAB. The other four patients were from the UK; their bile samples were collected at Hammersmith Hospital, Imperial College London by AVT and DSB. The diagnosis of CC was confirmed by a combination of cholangiography, cross-sectional imaging, elevated tumour markers (CA19-9, CA-125) and positive histology and/or cytology. Of the Egyptian CC patients, three had stage II disease (local invasion) and one had stage III disease (involving regional lymph nodes). Their body mass indices (BMIs) ranged from 23 to 26. All four of the UK patients with CC had stage II disease and BMIs of 24–27. A total of 21 patients had non-malignant biliary disease, including choledocholithiasis (n = 8), sphincter of Oddi dysfunction (SOD, n = 8), and PSC (n = 5). They were diagnosed by AVT and AWS. Three patients in the control group were Egyptian. They were diagnosed with choledocholithiasis (n = 2) and PSC (n = 1). Table 1 describes the patient demographics. Patients in the non-malignant disease group had BMIs in the range of 22–28. Patients in the choledocholithiasis subgroup had BMIs in the range of 26–28. None of the patients had cholangitis at the time of bile collection, as evidenced by clinical assessment and negative cultures. None of the patients had diabetes.

Table 1.

Demographic and clinicopathological characteristics of patients enrolled in the study

Disease	Cholangiocarcinoma	Choleodocholithiasis	SOD	PSC
Patients, n	8	8	8	5
Age, years, mean (range)	68 (56–82)	59 (33–79)	50 (27–62)	39 (28–66)
Sex
Male, n	5	0	0	5
Female, n	3	8	8	0

SOD, sphincter of Oddi dysfunction; PSC, primary sclerosing cholangitis

Bile collection and storage

Approximately 4 ml of bile were obtained at ERCP, after cannulation of the common bile duct and before contrast injection. Bile samples were stored in the dark at −20 °C prior to MRS analysis.

Sample preparation for bile MRS study

A sample of 600 µl of bile was transferred to a 17.5-cm, 5-mm diameter nuclear magnetic resonance (NMR) tube containing a 4-mm diameter ‘insert’ tube containing 50 µl of an external reference standard (15 µl of phosphate buffer 0.2 M/35 µl sodium trimethylsilyl-propionate [TSP] in deuterium oxide [D₂O] solution).⁹

Magnetic resonance spectroscopy

Magnetic resonance spectroscopy experiments were performed using a JEOL 500-MHz NMR spectrometer (JEOL Ltd, Tokyo, Japan) and a 11.7 Tesla superconducting magnet (Oxford Instruments, Plc, Oxford, UK). The ¹H NMR spectra were obtained using a pulse–collect sequence (pulse angle, 90 °; repetition time, 22 s; relaxation delay, 20 s; number of scans, 32) in combination with a pre-saturation technique and Hahn spin-echo sequencing.

Spectral interpretation

Bile spectra were analysed using the KnowItAll Informatics System, Version 7.9 (Bio-Rad Laboratories, Philadelphia, PA, USA). Peaks were referenced to the chemical shift of TSP at 0.0 ppm. Spectra were autophased and subsequently manually phased after the removal of the water peak (4.5–5.2 ppm), using the software flat-line function. A baseline correction was applied using the spline algorithm. Peaks were assigned according to established databases and previously published literature.^12,13 For relative quantification, peak areas were manually integrated and expressed as a percentage ratio to the total spectral signal (range: 10–0.2 ppm).

Statistical analysis

Statistical analysis was performed using spss for Windows Version 14.0 (SPSS, Inc., Chicago, IL, USA). Because of the small sample sizes of the groups, no assumptions of normality were made and non-parametric tests (Kruskal–Wallis and Mann–Whitney U-tests) were used to compare disease groups.

Results

Phosphatidylcholine concentration was reduced in bile from CC patients compared with bile from patients with benign biliary diseases. The median value of the PtC proton resonance at 1.2 ppm (tail group) was significantly lower in CC bile compared with SOD bile (P < 0.01). In addition, PtC resonance at 4.3 ppm (α-CH2) was significantly lower in bile from Egyptian CC patients than in bile from UK CC patients (P = 0.02). Taurine-conjugated (H-26 proton) and glycine-conjugated (H-25 proton) bile acids were significantly elevated in bile from patients with CC compared with bile from patients with benign biliary diseases (P = 0.013 and P < 0.01, respectively). This difference was significantly higher in CC bile compared with bile from SOD (P = 0.012 and P < 0.01, respectively) and PSC (P < 0.01 and P < 0.01, respectively) patients. Tables 2 and 3 show the median values, interquartile ranges and P-values for comparisons of the major biliary metabolites. Figures 1 and 2 compare bile acids and PtC between the groups.

Table 2.

Summary of the results of spectral integral values of major metabolites in bile

	All CC patients (n = 8)	Egyptian CC patients (n = 4)	UK CC patients (n = 4)	SOD patients (n = 8)	Choledocholithiasis patients (n = 8)	PSC patients (n = 5)
H-25 taurine-conjugated bile acids	1.2	0.8	1.6	0.6	0.6	0.4
	(0.3–2.1)	(0.6–1.5)	(0.3–2.0)	(1.2–1.5)	(0.1–1.3)	(0.1–0.9)
H-26 taurine-conjugated bile acids	0.9	1.6	0.9	0.6	0.5	0.5
	(0.6–2.5)	(0.6–2.5)	(0.8–1.3)	(0.5–0.8)	(0.1–2.8)	(0.3–0.7)
H-25 glycine-conjugated bile acids	1.3	1.5	1.3	0.7	0.8	0.6
	(0.8–3.5)	(0.8–3.5)	(1–1.6)	(0.5–0.9)	(0.2–1.5)	(0.4–0.7)
PtC tail group	24.0	19.4	24.9	26.5	28.2	29.5
(CH2)2	(9.5–26.3)	(9.5–24.9)	(22.7–26.3)	(24.8–27.1)	(11.4–34.6)	(13.5–30.0)
PtC head group	4.0	4.2	3.9	3.8	4.5	4.8
N-(CH3)3	(3.1–11.0)	(3.2–11.0)	(3.1–4.3)	(3.4–4.4)	(2.3–6.0)	(2.4–5.2)
PtC	2.7	2.8	2.7	3.2	2.6	3.4
β-CH2	(1.6–4.9)	(1.6–4.9)	(2.1–3.5)	(2.0–3.7)	(2.1–3.5)	(1.8–3.7)
PtC	1.0	0.8	1.4	0.6	0.7	0.94
α-CH2	(0.5–1.3)	(0.2–1.1)	(1.2–1.5)	(0.4–1.1)	(0.2–1.2)	(0.6–1.5)

Values are presented as medians and interquartile ranges

CC, cholangiocarcinoma; SOD, sphincter of Oddi dysfunction; PSC, primary sclerosing cholangitis; PtC, phosphatidylcholine

Table 3.

Comparison between the different groups for spectral integral values of bile acids metabolites, phosphatidylcholine, choline, lactate and cholesterol metabolites in bile

	CC vs. all groups	Egyptian vs. UK CC patients	CC vs. SOD	CC vs. choledcholithiasis	CC vs. PSC
H-18 bile acids	<0.01	0.15	<0.01	0.14	0.057
H-21 bile acids	0.78	0.77	0.46	0.46	0.30
H-25 taurine-conjugated bile acids	0.15	0.25	0.11	0.09	0.057
H-26 taurine-conjugated bile acids	0.013	1.0	0.012	0.059	<0.01
H-25 glycine-conjugated bile acids	<0.01	0.056	<0.01	0.059	<0.01
Total conjugated bile acids	0.96	0.30	0.79	0.53	0.76
PtC tail group (CH2)2	0.06	0.15	<0.01	0.74	0.10
PtC head groupN-(CH3)3	0.09	0.38	0.53	0.09	0.38
PtC, β-CH2	0.39	1.0	0.40	0.83	0.46
PtC, α-CH2	0.25	0.02	0.09	0.21	0.60
Choline, α-CH2	0.76	0.15	0.40	0.46	0.46

Values depicted in bold are significant at P < 0.05

CC, cholangiocarcinoma; SOD, sphincter of Oddi dysfunction; PSC, primary sclerosing cholangitis; PtC, phosphatidylcholine

Figure 1.

Distribution of (A) H-25 glycine-conjugated bile acid and (B) H-26 taurine-conjugated bile acid amongst the study groups. *Cholangiocarcinoma vs. benign biliary disease, P < 0.05. SOD, sphincter of Oddi dysfunction; PSC, primary sclerosing cholangitis

Figure 2.

Distribution of phosphatidylcholine (PtC) tail group amongst the various biliary diseases. Cholangiocarcinoma vs. all groups, P = 0.06; cholangiocarcinoma vs. SOD, P < 0.05. SOD, sphincter of Oddi dysfunction; PSC, primary sclerosing cholangitis

Discussion

Cholangiocarcinoma incidence rates are increasing worldwide, with substantial regional variation. Current diagnostic techniques to distinguish benign from malignant biliary tract disease are unsatisfactory, but the metabolic profiling of bile may differentiate benign from malignant biliary disease. This is the first study to compare the metabolic profiles of bile from CC patients from two geographically and racially distinct groups. We found that bile PtC is significantly reduced and taurine-conjugated (H-26) and glycine-conjugated (H-25) bile acids are significantly elevated in bile from patients with CC, compared with bile from patients with benign diseases.

Theoretically, several factors may have impacted our results. The bile lipid profile can be affected by the individual's metabolic condition and weight. None of the patients had diabetes and, although patient BMIs were marginally higher in the benign disease group, particularly in those with choledocholithiasis, the overall difference was small. Recent dietary habits can also affect the bile metabonome, albeit in an unpredictable manner. All patients were fasted for ≥8 h prior to the collection of bile. Recent weight loss may also potentially affect the biliary lipid profile. Most patients in both groups reported having lost weight in the 3 months prior to bile collection at ERCP, but the amount of weight lost was not verifiable in most cases and thus any potential impact on our data is unclear.

We have confirmed our previous findings of significantly higher bile acid levels and significantly lower PtC levels in bile from CC patients compared with bile from patients with benign biliary disease.^9,10 Further, we have established that these findings are not confined to White UK patients, but also occur in another distinct racial group comprising Egyptian patients. Phosphatidylcholine is the major phospholipid in mammalian cells and is essential for membrane structure, signal transduction and lipoprotein metabolism.¹⁴ It is synthesized in the hepatocyte endoplasmic reticulum from choline and subsequently transported into the biliary canaliculus by the flippase multidrug-resistant protein 3 (MDR3).¹⁵ Its main function in bile is to form mixed micelles with bile acids and cholesterol, essential for the emulsification of fats. Phosphatidylcholine has also been shown to have a cytoprotective role in the biliary epithelium and may reduce the cellular toxicity of bile acids.^16,17 Peak integrals for tail group PtC assignments were lower in bile from CC patients than in bile from patients with other biliary diseases. However, this difference did not reach statistical significance except with SOD.

We also found that PtC peak assignment, at 4.3 (α-CH₂),was significantly lower in Egyptian than UK patients with CC. The reduction in biliary PtC observed from our proton MR results may be potentially explained by impaired transport of PtC. Genetic variations in biliary transporter proteins, such as MDR3, would lead to reduced phospholipid export and a subsequent reduction in biliary PtC. This would disrupt micellar formation, leading to the formation of bile acid monomers which are more toxic to the cholangiocyte cell membrane, cause inflammation and, in theory, can induce carcinogenesis.^18,19 Studies have shown that mdr2 knockout mice (the mice analogue of MDR3) produce bile with high bile acids and low PtC content. These mice develop both hepatocellular carcinoma and CC.²⁰ The finding of reduced PtC in Egyptian patients with CC is consistent with the hypothesis that underlying biliary transporter differences exist in these patients.

Another significant finding of this study is that the peak integral values of both glycine- and taurine-conjugated bile acids were significantly higher in CC bile compared with bile from disease controls. The greatest difference was seen when comparing CC bile with bile from the SOD and PSC patient groups. We propose that bile acids may play a role in the pathogenesis of CC. Bile acids cause oxidative DNA damage, induce frequent apoptosis and, after repeated exposure, select for cells resistant to apoptosis, which leads to DNA mutation and carcinogenesis.^19–21 Bile acids may also promote direct cholangiocyte growth by activating the epidermal growth factor receptor and, via the mitogen-activated protein kinase signalling pathway, leading to disordered cell cycling and cholangiocyte proliferation.^22,23 Thus, elevated conjugated bile acids in CC bile may play an aetiopathogenic role and the relative ratio of conjugated bile acids and phospholipids may also serve as a potential biomarker of the disease. In conclusion, this early study addresses the feasibility of using MRS to assess metabolite changes in bile for diagnostic purposes and to provide insights into the pathogenesis of CC. Magnetic resonance spectroscopy of bile may have a future role in delineating biomarker panels for the diagnosis of CC and may potentially inform the direction of future genetic studies on disease susceptibility in different populations.

Conflicts of interest

None declared.