Introduction
Introduction
The widespread availability of endoscopic ultrasound (EUS) has facilitated evaluation
of the pancreas and extra-hepatic biliary system. EUS to date has been shown to be
highly sensitive in the detection of both choledocholithiasis (especially in patients
with small stones and non-dilated bile ducts) and gallbladder microlithiasis, however
the use of this technique in relation to endoscopic retrograde cholangiography (ERCP)
and laparoscopic surgery in gallstone disease remains somewhat confusing. EUS has
the added advantage of accurately discriminating CBD obstruction due to choledocholithiasis
and other causes such as small ampullary tumours, cholangiocarcinoma, benign extra-heaptic
cholangitis, congenital bile duct or peri-ampullary abnormalities (which together
account for 10 - 20 % of cases of CBD obstruction in patients with suspicion of CBD
stones). This review highlights technical aspects of examining the extra-hepatic biliary
duct system and the clinical performance and results of EUS in this context and proposes
strategies in relation to its use in association with ERCP and surgery in common clinical
practice. The use of EUS in patients with a asymptomatic dilated common bile duct
(CBD) is also discussed.
Technical aspects
Technical aspects
Visualisation of the extra-hepatic bile ducts using EUS demands a systematic approach
and is best performed under mild intravenous sedation with mizadolam and/or propofol.
With the patient in the left lateral decubitus position the transducer is placed in
the distal portion of the second segment of the duodenum and withdrawn to the ampullary
region. Above the distal common bile duct (CBD) is seen along with the terminal portion
of the main pancreatic duct (Fig. [1]). Slight proximal movements allow visualisation of the intra-pancreatic CBD which
is often located close to water balloon; the latter may sometimes need to be deflated
so as to not compress the CBD (Fig. [1]a). The CBD and common hepatic duct (CHD) are examined in longitudinal sections (Fig.
[2]) as far proximally as the hepatic duct with the transducer in the genu superius
and/or the duodenal bulb. The proximal duodenum also affords examination of the porta hepaticus with the cystic duct and the fundus and neck of the gallbladder. Positioning in the
duodenal bulb yields access to the gallbladder and CBD from its origin in the hilum
to the convergence of cystic and hepatic ducts and the proximal portion of the CBD
(Fig. [5]a). Using this method, the hepatic duct and CBD can be visualized in 95 - 100 % of
patients [1]
[2]
[3]
[4]. It is also mandatory to examine the gall bladder fundus via the transgastric route
with the transducer in the gastric antrum.
Standard imaging frequencies of 7.5 and 12 MHz are employed allowing resolution to
1 mm thus allowing for the detection of very small stones - microlithiasis. However,
it is important to start with a lower frequency to ensure that biliary disease at
a distance from the transducer is not missed. Slight repositioning of the patient
in the semi-prone position may be required to bring the CBD in line with the ultrasound
field of view. While it is sometimes difficult to examine the CHD at the level of
the hepatic hilum, especially using radial instruments, use of linear scopes facilitates
longitudinal views at this level. Another not infrequent, difficulty occurs following
endoscopic sphincterotomy or choledochoduodenostomy, where the presence of air within
the CBD makes it difficult to examine the peri-ampullary region and CHD, respectively.
Fig. 1
aLongitudinal section of the common bile duct inside the pancreas with the cystic duct
junction proximally and the MPD distally. b Junction of the common bile duct (CBD) and the main pancreatic duct (MPD) at the
level of the ampulla of Vater.
Fig. 2 The transducer is in the duodenal bulb and the common hepatic duct (HD) is visualised
below the inflated balloon and the cystic duct junction below the hepatic duct. IVC:
inferior vena cava.
Performance of EUS in the detection of choledocholithiasis
In a prospective and competitive blinded study in a group of patients with CBD obstruction
the accuracy of EUS in the diagnosis choledocholithiasis was 100 % (Figs. [3], [4]) [1]. Moreover, endosonography was significantly more sensitive than transabdominal ultrasonography
(US) or CT [1]. The diagnostic performance of EUS for choledocholithiasis is summarized in Table
[1] [5]
[6]
[7]
[8]
[9]
[10]
[11]. Data from these largely prospective studies include a large number of patients
(n = 1470). EUS demonstrated a high sensitivity (a mean of 93 %) with an excellent
specificity level (> 95 %) in these patients (Table [1]) who globally, with the exception of one series [8], presented an intermediate risk of CBD stones (20 - 50 %). The performance of EUS
in recognising choledocholithiasis is not related to stone size or the diameter of
the CBD (Figs. [3], [4]) [5]
[9] and high degrees of accuracy are also achieved with linear EUS [12]. In addition, the learning curve required to obtain satisfactory results in diagnosing
choledocholithiasis is relatively short as physicians with less than 1 years experience
achieve a high degree of skill in this indication [13].
Several studies have compared EUS and ERCP in a blinded fashion (ERCP performed subsequent
to EUS and by two different operators [14]
[6]
[10]). The sensitivity of ERPC was found to be 79 - 90 % compared to 88 - 100 % for EUS
and false-negative results were observed more frequently with ERCP. It is important
to note that the false-negative results with ERCP were due to small stones located
within dilated bile ducts, whereas false-negatives for EUS consisted of stones located
in the proximal portion of the CHD or the intra-hepatic ducts during the examination.
The excellent specificity for EUS in choledocholithiasis has also been shown by Napoleon
et al. in patients with suspected CBD stones [15] where 238 patients who were stone-free at EUS were prospectively followed at 1 year.
Follow-up revealed that 97 % of patients remained free of CBD stones thus demonstrating
the high negative predictive value of EUS [15].
In patients with unexplained acute pancreatitis the use of EUS with high resolution
affords an increased chance of detecting micolithiasis - a frequent cause of pancreatitis
[16]
[17]
[18]
[19]. Here, the role of EUS in relation to ERCP, which is both a diagnostic and therapeutic
procedure, is under discussion. Three studies emphasized that EUS should replace ERCP
as the first procedure in patients with mild to moderate acute pancreatitis to avoid
unnecessary ERCP and thus diminish ERCP-induced morbidity [17]
[18]
[19]
[20]. One interesting study devoted to this situation suggested that the rate of morbidity
and mortality could be reduced by systematically using EUS in case of acute pancreatitis
followed by ERCP with sphincterotomy when EUS has demonstrated CBD stones [21]. The efficacy and safety of EUS in this indication was confirmed [22] in a prospective randomized study in 140 patients with acute pancreatitis of a suspected
biliary origin. Patients were randomised to EUS or ERCP within 24 hours from admission.
When EUS detected choledocholithiasis (EUS group), therapeutic ERCP was performed
during the same endoscopy session. EUS was successful in all patients in the EUS group,
whereas cannulation of the CBD during ERCP was unsuccessful in 10 patients (14 %)
in the ERCP group (P = 0.001). Combined percutaneous transabdominal US and ERCP failed
to detect CBD stones in 6 patients in the ERCP group and overall morbidity rate was
7 % and 14 % (NS) in the EUS and ERCP groups, respectively [22]. Given the efficacy and relative simplicity of EUS in patients with pancreatitis
of suspected biliary origin, it should be proposed as a screening method especially
patients with low to moderate risks of a biliary origin. The combination of EUS and
eventually ERCP, in the event of finding a CBD stone, during the same session is highly
recommended to avoid a subsequent endoscopic session. Whether MRC will become sufficiently
accurate for the detection of small stones in this setting remains to be shown.
Table 1 Performance of EUS in the detection of choledocholithiasis
| Author |
Study design |
No. patients |
Frequency of
choledocholithiasis |
Sensitivity
% |
Specificity
% |
Diagnostic
accuracy % |
| Amouyal, 1994 [5]
|
Prospective/Blind |
62 |
32 (52 %) |
97 |
100 |
98 |
| Shim, 1995 [6]
|
Prospective/Blind |
132 |
28 (21 %) |
89 |
100 |
97 |
| Palazzo, 1995 [7]
|
Retrospective/Blind |
422 |
152 (36 %) |
95 |
98 |
96 |
| Prat, 1996 [8]
|
Prospective/Blind |
119 |
78 (66 %) |
93 |
97 |
95 |
| Sugiyama, 1997 [9]
|
Prospective/Blind |
142 |
51 (36 %) |
96 |
100 |
99 |
| Norton, 1997 [10]
|
Prospective/Blind |
50 |
24 (48 %) |
88 |
96 |
92 |
| Canto, 1998 [11]
|
Prospective/Blind |
64 |
19 (30 %) |
84 |
95 |
92 |
Fig. 3
aSmall stone (3 mm) with typical acoustic shadowing (black arrow) in the distal portion
of a non-dilated CBD. PV = portal vein. b Small stone (5 mm) within a dilated CBD (18 mm). c Three very small calculi (1 mm in size) located within the dependant portion of the
lower part of the CBD which is not dilated (5 mm).
Fig. 4 Small stone (6 mm in size) impacted in the distal CBD beneath a stenosis of the duct
with thickened wall due to cholangitis.
EUS or magnetic resonance cholangiography?
EUS or magnetic resonance cholangiography?
The use of non-invasive techniques to detect CBD stones is an appealing option and
magnetic resonance cholangiography (MRC) is gaining experience in this indication.
MRC is especially interesting in patients with low to intermediate risk of choledocholithiasis
(and in those who pose contraindications for sedation endoscopy) where interventional
strategies are less likely to be required. However, patients with such risk for CBD
stones are more likely to have small stones which may be difficult to detect without
EUS (the risk groups for CBD stones are given in Table [2]). Good comparative data on accuracy of MRC in the detection of choledocholithiasis
are to date lacking. A high sensitivity of 100 % with a poorer specificity of 73 %
were initially reported using MRC [23] and this was reported to be independent of stone size. However, in a recent study
by Kondo et al the sensitivities of EUS, MRC, and CT scan were 100 %, 88 %, and 88
%, respectively [24]. The differences between EUS and MRC were largely explained by stone size as the
sensitivity of MRC in detecting stones above and below 5 mm was 100 % and 67 %, respectively
[24].
Table 2 Risk of presence of common bile duct (CBD) stones in patients with suspected choledocholithiasis
according to clinical, biological and morphological criteria
|
Chance of
CBD stones |
Clinical parameters |
Biological parameters |
CBD diameter |
|
Low risk
|
2 - 3 % |
no associated clinical history |
normal |
≤ 7 mm |
|
Intermediate risk
|
20 - 50 % |
acute ascending cholangitis
pancreatitis |
↑ ALP ≤ twice UPL*
↑ GGT
↑ ALT/or ↑ AST |
8 - 10 mm |
|
High risk
|
50 - 80 % |
acute ascending cholangitis
jaundice |
↑ ALP > twice UNL* |
> 10 mm |
| *UPL: the upper normal value |
| ALP: alkaline phosphatase; AST: aspartate transaminase; ALT: alanine transaminans |
Performance of EUS in the detection of gallbladder
microlithiasis
Performance of EUS in the detection of gallbladder
microlithiasis
EUS is also excellent in the diagnosis of unrecognised gallbladder microlithiasis
with a sensitivity of almost 100 %. In patients with a high suspicion of cholelithiasis
following two negative trans-abdominal ultrasounds EUS is the diagnostic tool of choice.
The value of EUS in such circumstances was underlined in patients with acute pancreatitis,
where the sensitivity of EUS was 96 % for the detection of gallbladder microlithiasis
compared to 67 % obtained with microscopic examination of aspirated duodenal bile
following stimulation by cholecystokinin (Fig. [5]) [25]. These results have been subsequently confirmed [26]. Indeed in a recent multicentre prospective study involving 213 patients with acute
pancreatitis endosonography was the sole method establishing the diagnosis of biliary
pancreatitis in 15 % of patients [27].
Fig. 5 a Microlithiasis (1 mm) in the neck of the gallbladder. b Micro-lithiasis of the gallbladder with typical ”stared sky” aspect.
EUS prior to laparoscopic cholecystectomy
EUS prior to laparoscopic cholecystectomy
The prediction of CBD stones in patients with symptomatic choledocholithiasis may
be estimated with excellent precision by combining a number of clinical, biological
and morphological criteria [28]
[29]. Three different risk groups may be determined (Table [2]) with risk of finding choledocholithiasis of between 2 to 80 %. The probability
of finding stones in the CBD of these patients decreases as the period between the
onset of symptoms and exploration increases because of spontaneous evacuation of CBD
stones. Spontaneous evacuation of CBD stones was estimated to be approximately 21
% as evaluated by EUS and ERCP performed between 6 hours and 27 days, respectively
following the onset of symptoms [30]. The spontaneous passage of stones occurred in two-thirds of patients between a
period of 6 hours and 3 days but stone migration continued up to 27 days following
initial EUS [30]. Thus, even in patients with predictive factors suggesting a high risk of CBD stones,
not all have choledocholithiasis at examination cholecystectomy [31]. The delay between symptom onset and exploration should therefore be considered
in analysis of patients at risk of CBD stones as the further the delay, the lower
the chance of positive results.
It is accepted that patients categorized to the different risk-groups (Table [2]) merit different therapeutic strategies (Table [2] and treatment algorithm). In fact, treatment options also depend on local experience
of both surgeons and endoscopists and the wishes of the patients; once the advantages,
risks and probabilities of success of the different methods have been clearly explained
to them [32]:
-
Patients in the low-risk group may be treated using laparoscopic cholecystectomy without any preoperative
examination of the CBD. It seems unreasonable to propose 100 explorations (of whatever
kind) of the CBD in order to identify stones in 2 or 3 patients. Some surgeons carry
out systematic intra-operative opacification in order to examine the anatomy of the
biliary tree and/or to verify a patent CBD.
-
Those with a clearly identified high-risk may be selected at the outset for cholecystectomy using classical laparotomy. This
is rarely indicated when the diagnosis of choledocholithiasis is certain following
trans-abdominal ultrasound, and when the surgeon is not skilled in laparoscopic extraction
of CBD stones and the patient is young and/or does not wish to have an endoscopic
sphincterotomy. Endoscopic retrograde opacification may be performed for diagnosis
and treatment in the presence of acute ascending cholangitis or in the case of severe
acute pancreatitis observed within the first 72 hours; the same applies to poor candidates
for surgery (see algorithm). In other cases, EUS may be proposed particularly if several
days have elapsed between the onset of symptoms and treatment in order to verify CBD
patency. In patients with persistent CBD stones, EUS can determine the size, location
and number of stones, whether they are impacted or mobile, the diameter of the CBD
and cystic ducts, type of the cystic duct implantation (normal or deep) and finally
the state of the gallbladder (acute endosonographic severe cholecystitis with cholecysto-bulbar
fistula is a contraindication to laparoscopic cholecystectomy). The above elements
are helpful to the surgeon in selecting the best extraction strategy. If a cause other
than stones is found at EUS such as cholangiocarcinoma, ampullary or pancreatic tumours,
endosonographic loco-regional extension can also be performed [14].
-
For patients at intermediate-risk, the most appropriate current strategy comprises preoperative EUS possibly followed
by either endoscopic sphincterotomy during the same session in the event of CDB stones
or laparoscopic extraction if the surgeon is skilled in this manner and in the absence
of factors predictive of conversion to laparotomy (see algorithm).
EUS after Cholecystectomy
EUS after Cholecystectomy
Residual or recurrence lithiasis of the CBD following cholecystectomy occurs in 1
- 5 % of cases [33]. In routine practice, such a situation is suspected when there is either biliary
colic associated with transient abnormalities in the hepatic or pancreatic blood tests
or persistent abnormalities in blood tests with or without a dilated CBD at trans-abdominal
ultrasound. This situation is quite frequent in daily practice and is complicated
by the fact that the main differential diagnosis, i. e. sphincter of Oddi dysfunction
(SOD), is impossible to differentiate on clinical, biochemical or evolutionary criteria.
Residual or recurrent lithiasis is potentially serious if not recognised and untreated
due to the risk of ascending cholangitis or severe pancreatitis. It is therefore important
to diagnose it when it is suspected. SOD on the other hand is different in that the
majority of complications stem from diagnostic efforts using ERCP, especially if endoscopic
sphincterotomy is performed [34]
[35]
[36]
[37] (the latter of which is necessary in the event of the discovery of a dilated CBD
in order avoid misdiagnosing small stones not seen at simple injection of contrast
medium [38]) or using sphincter of Oddi manometry. In fact, in case of SOD, ERCP and manometry
are accompanied by a serious risk of acute pancreatitis [34]
[35]
[36]
[37]. The paradox in this situation is that endoscopic sphincterotomy is the treatment
of choice in terms of efficacy, morbidity and death rate with respect to residual
or recurrent lithiasis [39], while this is an effective but also very dangerous treatment (about 20 % of acute
pancreatitis) in cases of SOD [34]
[35]
[36]
[37]. Thus, caution should be applied prior to endoscopic sphincterotomy where the diagnosis
of residual lithiasis and SOD is not apparent. EUS is clearly the examination of choice
in this delicate situation and its use appears even clearer in patients at high risk
of SOD, i. e. in young subjects with narrow CBD and no evidence of ascending cholangitis.
As regards MRCP, its role in this sub-group of patients with thin bile ducts is worthy
of evaluation and prospective comparison with EUS (see above).
Asymptomatic dilatation of the common bile duct
Asymptomatic dilatation of the common bile duct
Although in a routine clinical setting the significance of a dilated CBD in the absence
of symptoms pertaining to biliary or pancreatic disease is extremely frequent, little
research data is available on this subject. Several mechanisms may result in dilatation
of the extra-hepatic bile ducts including strictly benign disorders - post cholecystectomy
(Fig. [6]), dilatation secondary to certain drugs (opiates…) or incidental anomalous junction
of the pancreaticobiliary duct (AUPBD) - where follow-up explorations are indeed not
required. Indeed, although many experts felt that a large CBD diameter was physiological
in the elderly and while the diameter increases significantly with age, 98 % of all
ducts remained below 6 to 7 mm, the commonly accepted upper range of normal, in a
large prospective series of asymptomatic patients (n = 1018) aged 60 to 94 years (analysis
performed using transabdominal US) [40]. However, a number of situations exist where ”significant” pathologies results in
an asymptomatic dilated CBD including asymptomatic choledocholithiasis, malignant
and benign (e. g., sclerosing cholangitis) strictures and choledochal cysts. Indeed,
asymptomatic choledocholithiasis has been estimated to be as high as 19 % in some
surgical series [41] - thus underlying the need for thorough CBD exploration in patients undergoing laparoscopic
cholecystectomy (as discussed above). Recently, Kim et al. [42] studied 77 patients with an asymptomatic internal CBD diameter > 7 mm at US and
without a definite causative lesion on the latter. Of these, 49 underwent ERCP and
28 underwent follow-up US or CT instead of ERCP. Patients with increased bilirubin
or with a previous history of upper abdominal surgery including cholecystectomy were
excluded. Follow-up examinations revealed: a) ERCP: no lesion in 20 patients (41 %),
juxtapapillary duodenal diverticulum (JDD) in 11 (23 %), benign stricture in ten (20
%), distal CBD mass in two (4 %), choledochal cyst in two (4 %), anomalous junction
of the pancreaticobiliary duct (AUPBD) in two (4 %), and choledochal cyst with AUPBD
in two (4 %). There were no differences in age or in alkaline phosphatase or gamma-glutamyl
transpeptidase levels between the patients who had causative lesions revealed at ERCP
and those who did not; b) among the 28 patients who did not undergo ERCP, the CBD
returned to normal in12 and eight had no change in CBD diameter on follow-up US. Among
eight patients who underwent CT, there were four with normal findings, one with JDD,
and three with suspected choledochal cysts [42]. Thus a significant cause of biliary tract lesions in asymptomatic adults with dilatation of the CBD
can be found and certain such as choledochal cysts require careful surveillance and
preventive surgery [43]. Most often, a dilated CBD in such circumstances results from an incidental discovery
at US or CT - performed for another unrelated reason. While these examinations in
themselves can help in ruling out obvious causes (e. g., early malignant obstruction
or a large asymptomatic stone) endosonography appears to be the method of choice as
it capable of providing a diagnosis in almost all cases; In fact, EUS was found to
provide an accurate explanation for CBD dilatation from an unexplained cause in 70
of the 76 patients (92 %) [44].
Fig. 6 Dilatation of the intra-pancreatic portion of the CBD (which measures 7 mm). Note
that the common hepatic duct is dilated (CHD) as well as the remnant of the cystic
duct.
Algorithm Diagnostic and therapeutic algorithm for patients with symptomatic cholecystolithiasis
and suspicion of common bile duct (CBD) stones. ES: endocopic sphincterotoma.
Conclusions
Conclusions
Despite the development of MRCP and the laparoscopic extraction of CBD stones, the
place of EUS in the diagnosis of CBD stones remains important. The examination combines
the best performance and almost zero morbidity and unlike EUS for other indications,
the results in relation to choledocholithiasis depend little on the experience of
the operator. Its role in relation to other diagnostic techniques depends principally
on the environment (experience of the surgeon and interventional endoscopist) and
the circumstances of the diagnosis. The main advantages of EUS are its sensitivity
and specificity even in patients with small calculi in a non-dilated CBD. In the future
however, other imaging modalities may challenge EUS. The role of EUS versus that of
ERCP has been correctly defined; a comparison with MRC needs to be evaluated prospectively
in a controlled group of patients who pose serious frequent diagnostic dilemmas, i.
e. those with a non-dilated CBD and small stones, because these are patients who represent
difficult therapeutic problems in terms of feasibility for laparoscopic extraction
and in terms of the potential morbidity of laparoscopic extraction and endoscopic
sphincterotomy. Finally, when EUS is chosen as the imaging modality to identify CBD
stones, it seems logical to perform ERCP during the same session to enable therapeutic
extraction. Thus, training in both biliary EUS and therapeutic ERCP seems to be a
promising goal for young endoscopists who want to specialize in interventional endoscopy.
