Keywords

1 Endoscopic Diagnosis of Pancreatic Cancer

1.1 Background

Pancreatic cancer has a poor prognosis, with a 5-year survival rate of 8–9%, which is the lowest of all cancers. The reason for this is that early detection is challenging, and at the time of diagnosis, it is already an unresectable advanced cancer. If pancreatic cancer can be detected at an early stage (i.e., less than 2.0 cm), the prognosis is relatively good. Therefore, accurate early diagnosis is crucial for improving the prognosis of pancreatic cancer. With recent advances in endoscopic equipment and technology, endoscopy has become indispensable for the diagnosis and treatment of pancreatobiliary diseases. Among them, endoscopic ultrasonography (EUS) and endoscopic retrograde cholangiopancreatography (ERCP) play a significant role in the early detection of pancreatic cancer.

EUS is an ultrasound (US) technique in which the tip of the endoscope is equipped with a high-frequency transducer. To observe pancreatic lesions in real-time through the gastrointestinal tract, it is possible to obtain high spatial resolution and high image resolution without being affected by gastrointestinal gas artifacts. Therefore, EUS is extremely useful for detecting small pancreatic tumors. Furthermore, EUS-related procedures such as contrast-enhanced EUS (CE-EUS), EUS elastography, and EUS-guided fine needle aspiration (EUS-FNA) enable a more accurate diagnosis.

The established usefulness of EUS-FNA for pancreatic masses has reduced the chances of performing diagnostic ERCP with a risk for post-ERCP pancreatitis (PEP). However, recently, there have been increasing reports of pancreatic carcinoma in situ with no mass and only pancreatic duct stenosis from Japan. Serial pancreatic juice aspiration cytological examination (SPACE) using ERCP is useful for diagnosis, and the need for ERCP has been reconfirmed.

Thus, it is important to use EUS and ERCP for the diagnosis of pancreatic cancer. In the first section, we introduce the role of the two modalities in the clinical practice of pancreatic cancer.

1.2 Endoscopic Ultrasound (EUS)

1.2.1 Type of EUS

There are two types of EUS, radial, and convex types, depending on the distally attached ultrasonic probe. The radial EUS image was visualized 360° perpendicular to the scope axis. The pancreatic duct can be visualized on the long axis. It has the advantage that changes in the pancreatic duct diameter are easy to detect. However, it is sometimes difficult to visualize the transitional part from the pancreatic head to the pancreatic body (pancreatic neck) and visualize the end of the pancreatic tail. On the other hand, the convex EUS image is visualized parallel to the scope axis. It has the advantage that it is easy to understand the relationship between the lesion and the blood vessel because the blood vessel easily aligns with the axis of the scope. In addition, the pancreatic neck can be visualized. The major difference between the two types is that radials, in contrast to the convex type, does not allow fine needle aspiration (FNA). It is essential to understand the characteristics, advantages, and disadvantages of the radial and convex types and use them appropriately according to the situation.

1.2.2 Detection of Pancreatic Cancer

EUS is considered the most sensitive diagnostic imaging modality for detecting solid pancreatic lesions. Summarizing 12 previously published reports, the sensitivity of EUS for the detection of pancreatic cancer was 98%, which was superior to that of US (72.5%), computed tomography (CT; 74%), and magnetic resonance imaging (MRI; 83%) [1,2,3,4,5,6,7,8,9,10,11,12,13]. Thus, EUS is particularly useful in detecting small pancreatic lesions due to its high resolution. In recent years, the usefulness of multi-detector-row CT (MDCT) in diagnosing pancreatic cancer has been reported [14]. However, when detecting pancreatic cancer of ≤20 mm, the detection rate of MDCT decreased to 50%, while EUS had a high detection rate of more than 90% [15, 16]. Several reports have shown that EUS can detect pancreatic cancer that was not identified on other modalities. Krishna et al. reported that the sensitivity of EUS for detecting pancreatic malignancy when MDCT findings were indeterminate was 85%, with a specificity of 58% [17]. Therefore, EUS should be performed in patients suspected of having pancreatic cancer even if MDCT does not detect a mass.

Recently, the results of Japanese pancreatic cancer registries showed that even for pancreatic cancer ≤20 mm, there was a difference in the 5-year survival rate between < 10 mm and 10–20 mm (80.4% versus 50%, respectively) [18]. Kamada et al. reported that the detection rates of US, contrast-enhanced CT, and EUS for pancreatic cancer of <10 mm were 30%, 30%, and 100%, respectively [1]. Based on the above, EUS plays a crucial role in detecting small pancreatic cancer.

EUS images of pancreatic cancer are visualized as a well-circumscribed hypoechoic mass with irregular margins. It may also be accompanied by dilation of the pancreatic duct distal to the mass, dilation of the surrounding branch duct, and retention of cysts. It is sometimes difficult to distinguish between benign mass-forming pancreatitis and pancreatic cancer. Contrast-enhanced EUS, which will be described later, is useful for the differential diagnosis; pancreatic cancer is visualized as a hypovascular pattern, and mass-forming pancreatitis is visualized as a homogenous isovascular pattern [19].

1.2.3 Differential Diagnosis of Solid Pancreatic Mass

1.2.3.1 Contrast-Enhanced EUS (CE-EUS)

CE-EUS can image the blood vessel flowing into the lesion in real-time and obtain a clear contrast with the surrounding pancreatic parenchyma. CE-EUS has been reported to be useful for the differential diagnosis of pancreatic tumors because it enables qualitative diagnosis of tumors and detailed blood flow evaluation. In Japan, Sonazoid®, a second-generation low-sound pressure system intravenous contrast agent, is widely used. This contrast agent consists of gas-filled microbubbles of approximately 2–5 μL in diameter, encapsulated by a phospholipid or lipid shell. In principle, EUS can detect the secondary harmonic component generated from the contrast agent, so that an ultrasonic contrast agent can be directly captured as a signal, and a contrast image can be obtained.

The mass is evaluated as hyperenhancement, isoenhancement, hypoenhancement, or nonenhancement, depending on how much the mass contrasts with the surrounding pancreatic parenchyma. Pancreatic cancer mainly shows a hypoenhancement pattern, mass-forming pancreatitis shows an isoenhancement pattern, and endocrine tumors show a hyperenhancement pattern. Kitano et al. [16] reported that the sensitivity and specificity of pancreatic cancer showing hypoenhancement patterns were 95% and 89%, respectively, in 277 patients with pancreatic tumors. On the other hand, in the same study, the sensitivity and specificity of endocrine tumors showing hypervascular pattern were 79% and 99%, respectively. The sensitivity and specificity of tumorigenic pancreatitis showing an isovascular pattern were 78% and 95%, respectively. Furthermore, in the meta-analysis, the sensitivity and specificity of CE-EUS for pancreatic cancer were 93% and 80%, respectively, showing promising results [20]. Interestingly, regarding small lesions of < 2 cm, MDCT had a sensitivity of 70.6% and a specificity of 91.9%, whereas CE-EUS had a sensitivity of 91.2% and a specificity of 94.4%. CE-EUS was significantly superior to MDCT [16].

T staging and N staging are important factors in deciding on the course of treatment for pancreatic cancer. In T staging, it is essential to diagnose the presence or absence of infiltration of pancreatic cancer into the surrounding blood vessels. Imazu et al. [21] reported that the diagnostic rate of T staging was significantly improved by adding CE-EUS (92%) compared to EUS alone (69%). In addition, they described that contrast enhancement was particularly useful in determining portal vein infiltration because the portal vein wall could be visualized more clearly. CE-EUS has also been reported to be useful in diagnosing N staging [22].

1.2.3.2 Elastography (EG)

US elastography is a technique for imaging or quantifying tissue elasticity. An image of strain, which has a negative correlation with tissue elasticity, is called strain elastography (strain-EG). Strain-EG is originally a qualitative examination that is evaluated by color pattern, but quantitative tissue elasticity diagnosis such as strain ratio and histogram analysis is possible by processing the image. Giovannini et al. [23] classified the EUS-EG color patterns into five categories and evaluated the ability to distinguish between benign and malignant pancreatic masses. As a result, the sensitivity and specificity of EUS elastography to differentiate benign from malignant pancreatic masses were 92.3% and 80.0%, respectively, compared to 92.3% and 68.9%, respectively, for the conventional B-mode images. They concluded that EUS elastography is superior to conventional B-mode imaging and appears to distinguish benign from malignant pancreatic masses with high sensitivity, specificity, and accuracy. Although this color pattern evaluation is visual and easy to understand, the interpretation tends to be subjective. Therefore, it should be recognized that the results obtained will be less objective.

There is a strain ratio (SR) that semi-quantitatively analyzes the image information of the EUS-EG. SR can be evaluated objectively, and its usefulness has been reported. Iglesias-Garcia et al. [24] evaluated the ability to distinguish between benign and malignant pancreatic masses using SR and reported that the diagnostic ability was high, with a sensitivity of 91.2%, a specificity of 91.0%, and an accuracy rate of 91.1%.

CE-EUS and EUS elastography may provide additional information on the diagnosis of pancreatic cancer, in addition to the yield from EUS-FNA. CE-EUS helps to identify EUS-FNA targets and reduces the need to repeat FNA [16]. The specificity of EUS-FNA may be improved when used in combination with EUS elastography [25]. If the pancreatic mass with negative EUS-FNA findings is a hypovascular mass with CE-EUS or a hard mass with EUS elastography, repeating EUS-FNA is recommended.

1.2.4 Definitive Diagnosis of Pancreatic Cancer

1.2.4.1 EUS-FNA

Pathological diagnosis is essential for diagnosing pancreatic cancer and in deciding on courses of treatment for pancreatic cancer. EUS-FNA is a technique for collecting tissue under the guidance of EUS, and Villmann et al. [26] reported the clinical application of EUS-FNA for pancreatic lesions in 1992. EUS-FNA is widely performed for the pathological diagnosis and staging of solid pancreatic masses owing to its high diagnostic ability and safety. A meta-analysis of EUS-FNA in pancreatic tumors reported that the sensitivity of EUS-FNA was 85–89%, and the specificity was 96–99% [27]. In addition, the diagnostic ability of EUS-FNA for lymph node swelling is also high, with a sensitivity of 86.8% and a specificity of 95.8% [28].

Factors that influence the diagnosis of EUS-FNA are classified into three factors: patient factors, procedural factors, and other factors.

Patient factors include size, characterization, location, and background pancreatic parenchyma. Regarding the size of the pancreatic masses, it was reported that the diagnostic ability of EUS-FNA was lower for smaller lesions. The accuracy rate was 93.4% for lesions > 20 mm, 83.5% for 10 mm to 20 mm, and 82.5% for <10 mm [29].

Procedural factors include the needle’s diameter and shape, puncture, and suction method. Generally, the thicker the needle, the larger the sample collected; however, the maneuverability deteriorates. On the other hand, the thinner the needle, the better the operability, but the smaller the amount of sample collected. Therefore, a thick needle is selected when many tissues, such as for immunostaining and genetic testing, are required. A thinner needle is selected for puncture from the duodenum, where the scope’s curvature becomes strong. In recent years, needles for the histological examination have been developed for an endoscopic ultrasound-guided fine needle biopsy (EUS-FNB), and their usefulness has been reported. The shape of these needles’ tips has been devised, such as the franceen shape and other shapes with side holes and core traps. Recent reports indicated that EUS-FNB has a significantly lower number of punctures and a higher diagnosis rate than EUS-FNA [30].

Recently, cancer genomic medicine using next generation sequencing (NGS) has become widespread. In a report comparing the success of NGS between FNA needles and FNB needles, FNB needles could collect samples that were significantly more suitable for NGS than FNA needles (90.9% versus 66.9%, p = 0.02) [31].

Regarding the puncture, various measures have been taken to improve the accuracy rate of EUS-FNA. For example, the Door knocking method, in which a tissue sample is taken into the needle by hitting the needle strongly against the mass, is useful for a hard tumor with significant fibrosis. The fanning technique has also been reported, in which tissue is collected from different parts of the tumor by moving the needle in a fan shape and can be diagnosed with a small number of punctures [32].

As for the suction pressure, applying negative pressure using a 20 ml syringe is common. In addition, the high negative pressure method using a 50 ml pressure-resistant syringe [33], slow-pull method using the capillary phenomenon by slowly pulling the stylet [34], the wet-suction method in which the inside of the needle is filled with saline [35], and the non-suction method are available. However, there is no consensus on the best suction method among them, and it is important to change the suction method according to the mass characteristics.

Other factors include processing of the sample collected and additional genetic testing. It has been reported that rapid on-site examination (ROSE) in the presence of a pathologist or cytotechnologist not only improves the accuracy rate of EUS-FNA, it also reduces complications with the decrease in the number of punctures [36]. Nevertheless, Iwashita et al. recommend macroscopically identified white tissue of more than 4 mm in the sample collected by EUS-FNA to improve the accuracy rate if ROSE is not available (macroscopic on-site quality evaluation: MOSE) [37]. By combining EUS-FNA with K-ras mutation, the diagnostic ability of EUS-FNA can be increased. In a meta-analysis by Fuccio et al., adding K-ras mutations to inconclusive EUS-FNA cases reduced false-negative rates by 55% and increased diagnostic sensitivity by 8.1% [38].

The main complications include bleeding, pancreatitis, and perforation. According to a systematic review by Wang et al. [39], the frequency of EUS-FNA complications in the pancreas was 1.03%, of which pancreatitis was 0.44%, bleeding was 0.10%, and perforation was 0.01%. In solid pancreatic tumors, the overall complication rate is reported to be 0.82%, which is lower for pancreatitis (0.35%), bleeding (0.07%), and perforation (0.01%). Thus, EUS-FNA is regarded as a very safe procedure. However, it should be noted that EUS-FNA related complications for pancreatic cystic lesions are slightly higher at 2.8%.

Recently, there have been some case reports of needle tract seeding (NTS) by EUS-FNA. The frequency is unknown, but in a retrospective multicenter study by Yane et al. [40], NTS was found in 3.4% (6/176) of patients who underwent EUS-FNA before pancreatic tail resection; there was no significant difference in prognosis between the EUS-FNA group and the non-EUS-FNA group (48 months versus 43.9 months, P = 0.392), but it was reported that the frequency was not negligible. Gao et al. [41] summarized 33 cases extracted from previous reports; among them, there were 28 cases of NTS in EUS-FNA/FNB for pancreatic cancer, all of which were tumors of the pancreas’ body and tail. The NTS site was found on the stomach wall (particularly the posterior wall) in all cases, and the median size was 25 mm (range: 4–50 mm). Most were localized in the submucosa and exhibited the morphology of submucosal tumors. Although the detection strategy and treatment for NTS have not been clarified, surgical resection is possible by early detection, and some cases have a good prognosis. It is important to follow up with NTS in mind after surgery carefully. The factors of NTS are considered to be associated with the number of punctures, size and shape of the needle, and tumor characteristics. However, it is unclear which factor is involved due to few reports. Therefore, it is important to avoid unnecessary EUS-FNA or reduce the number of punctures of the pancreatic body and tail tumors that do not include the puncture route in the surgical resection area [42]. FNB needles have been reported to have better diagnostic rates with a smaller number of punctures than FNA needles, and choosing FNB needles may be one way to avoid NTS.

1.3 Case Presentation (Fig. 18.1)

A 70-year-old male visited our hospital with an elevated tumor marker (CA19-9 level: 109 U/mL). Abdominal contrast-enhanced CT revealed no dilation of the pancreatic duct and no tumorous lesions (Fig. 18.1a). EUS showed a hypoechoic lesion with a diameter of 10 mm at the pancreatic body (Fig. 18.1b). CE-EUS showed a hypovascular pattern (Fig. 18.1c), and EUS-EG showed a dominant blue pattern with heterogeneity (Fig. 18.1d). We suspected pancreatic cancer based on these findings and subsequently performed EUS-FNA (Fig. 18.1e). Pathological findings revealed pancreatic adenocarcinoma (Fig. 18.1f). The patient underwent distal pancreatectomy and was diagnosed with pancreatic cancer (pT1b, pN0, pM0, pStage IA; UICC 8th edition).

Fig. 18.1
figure 1

(a) CE-CT revealed no dilation of the pancreatic duct and no tumorous lesions. (b) EUS showed a hypoechoic lesion with a diameter of 10 mm at the pancreatic body (yellow arrowhead). (c) CE-EUS showed a hypovascular pattern (yellow arrowhead). (d) EUS-EG showed a dominant blue pattern with heterogeneous (yellow arrowhead). (e) EUS-FNA was performed for a hypoechoic lesion in the pancreatic body. (f) Pathological findings revealed adenocarcinoma

Full size image

1.4 Endoscopic Retrograde Cholangiopancreatography (ERCP)

In the clinical practice of pancreatic cancer, ERCP is considered as a therapeutic modality (biliary drainage) mainly for obstructive jaundice and cholangitis due to the spread of EUS-FNA. In some small pancreatic cancers with a tumor diameter <10 mm, it may be challenging to identify the mass even with EUS, and EUS-FNA is not indicated in such cases. Furthermore, in cases such as carcinoma in situ of the pancreas that presents only with pancreatic duct stenosis, a definitive pathological diagnosis can be obtained only by pancreatic juice cytology using ERCP. The usefulness of preoperative ERCP cytology for early diagnosed pancreatic cancer cases, such as Stage 0–1, has been reported in Japan. In addition, continuous cytology with endoscopic nasopancreatic drainage (ENPD) tube placement (SPACE) is useful for diagnosing patients with smaller pancreatic cancer [43, 44]. The advantages of SPACE are that a sufficient amount of pancreatic juice can be obtained and that multiple pancreatic juice cytology can be performed, which can be expected to maximize the diagnostic yield. The sensitivity of SPACE at Stage 0 was 70–100%, and the smaller the tumor diameter, the better [45]. In addition, in the case of localized pancreatic duct stenosis without mass lesions, the sensitivity was 82%, the specificity was 100%, and the accuracy rate was 95%. If early pancreatic cancer is suspected in patients with pancreatic duct stenosis whose mass cannot be visualized by EUS, ERCP cytology and SPACE should be actively performed.

1.5 Case Presentation (Fig. 18.2)

A 60-year female visited our hospital with mild upper abdominal pain. CE-CT revealed dilatation of the pancreatic duct in the body and tail of the pancreas, but no tumorous lesions were detected (Fig. 18.2a). EUS showed dilatation of the main pancreatic duct in the pancreatic body and tail. Stenosis was suspected on the proximal side of the dilated part of the main pancreatic duct. At the site where the caliber of the pancreatic duct changed, there were no mass regions (Fig. 18.2b). ERP revealed focal stenosis with distal dilation of the pancreatic duct in the tail of the pancreas (Fig. 18.2c). SPACE was performed (Fig. 18.2d), and malignant findings were obtained (Fig. 18.2e). Thus, laparoscopic distal pancreatectomy was performed. High-grade pancreatic intraepithelial neoplasm (PanIN) was observed in the resected specimen, consistent with stenosis of the main pancreatic duct. Finally, the patient was diagnosed with pancreatic carcinoma in situ (pTis, pN0, pStage 0; UICC 8th edition).

Fig. 18.2
figure 2

(a) CE-CT revealed dilatation of the pancreatic duct in the body and the tail of the pancreas, but no tumorous lesions were detected. (b) EUS showed dilatation of the main pancreatic duct in the pancreatic body and tail (red arrowhead). Stenosis (yellow arrow) was suspected on the proximal side of the dilated part of the main pancreatic duct. There were no mass regions consistent with stenosis of the main pancreatic duct. (yellow arrowhead: normal diameter of the main pancreatic duct on the head side of the stenosis). (c) ERP revealed focal stenosis (yellow arrow) with distal dilation of the pancreatic duct in the tail of the pancreas. (d) SPACE was performed using an endoscopic nasopancreatic drainage tube. (e) Malignant findings were obtained by SPACE

Full size image

2 Endoscopic Diagnosis of Cholangiocarcinoma

2.1 Background

Cholangiocarcinoma (CC) is an uncommon gastrointestinal malignancy originating from the epithelial lining of the biliary tract. CC accounts for approximately 3% of all gastrointestinal malignancies [46]. Nevertheless, CC is the most common malignant tumor of the biliary tract and is the second most common primary hepatic cancer [47]. Incidence varies worldwide, with the highest known rates in Southeast Asia and much lower rates in the Western world [48] with relatively higher rates among the elderly with a male predominance [49]. The clinical course, localization, and histological analysis of CC usually represent challenging issues for diagnosis and management.

Although surgery and liver transplantation are the main curative options for CC, at the time of diagnosis, most patients have advanced stages at diagnosis with unresectable disease, resulting in a poor prognosis with low 5-year overall survival [50]. Approximately 30% of patients considered resectable on the initial imaging are shown to be unresectable on surgical exploration [51].

Anatomically, CC is usually categorized into intrahepatic (iCC), distal (dCC), or perihilar (pCC) subtypes. The latter can be further classified based on the involvement pattern of biliary ducts, according to the Bismuth-Corlette classification, into four different types. pCC represents about 50–60%, dCC 40%, and iCC < 10% of CC cases [52]. iCC incidence appears to be increasing in Western countries [53]. This increase may be explained to some extent by the progress in diagnostic procedures; however, the rise of viral hepatitis and fatty liver disease may have largely impacted this rising incidence [54].

CCs are grouped morphologically into either mass-forming, periductal-infiltrating, or intraductal-growing subtypes [55]. Histologically, most CCs are adenocarcinomas (90%). Other variants include papillary adenocarcinoma, squamous cell carcinoma, signet-ring type, intestinal-type adenocarcinoma, and undifferentiated carcinoma [56].

Although the definite etiology is not clearly understood, several risk factors for CC are well-described, including cholangitis, particularly primary sclerosing cholangitis (PSC), inflammatory bowel disease (both ulcerative colitis and Crohn’s disease), parasitic infections, choledochal cysts, hepatolithiasis, choledocholithiasis, hepatitis C and B viral infections, liver cirrhosis regardless of the cause, toxic agents such as thorotrast, diabetes, obesity, heavy alcohol use, and smoking [57].

Despite the advances in cross-sectional imaging techniques, the diagnosis and differentiation of malignant bile duct strictures remain challenging. Endoscopic approaches are often required for definitive histological diagnosis in addition to precise local staging and resectability assessment in early-stage disease when radiological features are uncertain.

Endoscopic evaluation includes various procedures such as endoscopic retrograde cholangiography (ERC), endoscopic ultrasound (EUS), intraductal ultrasound (IDUS), direct cholangioscopy, and probe-based confocal laser endomicroscopy (pCLE).

2.2 Endoscopic Retrograde Cholangiography and Associated Procedures

Traditionally, ERC has been the first-line procedure for suspected CC, allowing complete evaluation of the extrahepatic biliary tree, better understanding of site and length of biliary strictures, and providing cytological and/or histological diagnosis upon which management strategies can be planned.

Bile aspiration is an affordable, easy, old-fashion ERC-assisted technique for obtaining cytological analysis using a catheter at the level of the bile duct stricture to aspirate 10–15 mL of bile [58]. However, it has low sensitivity for the detection of malignancy, ranging from 6 to 32% [59]. On the other hand, another novel approach for early detection of CC depends on bile analysis for tumor proteins “proteomics” with the concept that carcinoma takes place at the biliary epithelium and tumor-related proteins can be detectable in bile. In a comparative study, bile proteomic analysis discriminated benign conditions (choledocholithiasis and PSC) from CC with high accuracy [60].

ERC-assisted biliary brushing for cytology remains the most commonly used method for histological diagnosis of CC at the time of ERC. Despite being an easy, straightforward, and highly specific technique depending on a wire-guided cytology brush directed into the biliary stricture, its sensitivity for the diagnosis of potentially malignant strictures has been unsatisfactory, ranging from 30 to 57% [59]. New generations of cytobrushes with increased length, size, and bristle stiffness were investigated with the aim of increasing ERC-based tissue yield, but the results were disappointing [61]. Similarly, many modifications of the cytobrushing technique have been tried to increase its sensitivity. For instance, dilatation of the stricture before and after brushing was studied, but with poor results [62]. On the other hand, obtaining successive brush specimens has been shown to increase tissue yield [62, 63]. In a recent prospective study, the sensitivity of biliary cytobrushing was reported to have increased to 84.3% when combined with biliary aspiration before and after the brushing technique [64]. Furthermore, advanced cytological techniques such as fluorescence in situ hybridization (FISH) [65] and flow cytometry have improved sensitivity when combined with conventional biliary cytology [66]. FISH depends on the use of fluorescence-labeled probes to detect chromosomal abnormalities in the form of aneuploidy or polyploidy in cells obtained via routine biliary brushings. Benign strictures have been predictively differentiated from malignant ones in PSC patients by optimizing the performance of FISH testing of multiple specimens of the biliary tract [67]. However, these advanced techniques are not usually available or widely approved.

ERC-assisted endobiliary forceps are assumed to provide deeper tissue samples with increased diagnostic sensitivity. Its sensitivity to detect malignant bile duct strictures was shown to be higher, ranging from 43 to 81% [68]. However, a recent meta-analysis showed that this increase was only demonstrated when biopsy was combined with brushing, with almost the same pooled sensitivity for both of them (45% for brushing versus 48% for endobiliary forceps) [69]. Nevertheless, this technique is more challenging with a risk of major bleeding [70] and perforation [71] and thus requires a high degree of experience to perform safely.

2.3 Endoscopic Ultrasound

Although magnetic resonance cholangiopancreatography (MRCP) is the primary non-invasive tool for pancreaticobiliary systems, EUS has a comparable impact on the management of CC. EUS provides a detailed examination of the extrahepatic biliary tree and surrounding structures, making it a valuable tool for the diagnosis and accurate staging of extrahepatic CC with a lower complication rate when compared to ERCP [72]. Moreover, EUS evaluates associated portal lymphadenopathy [56] with a high degree of accuracy using non-invasive real-time EUS elastography. In addition, EUS-FNA has been more accurate than CT and positron emission tomography (PET)-CT for the evaluation of regional lymph node metastasis [73] which in turn has largely influenced the selection of different management lines [74].

In a large study by Mohamadnejad et al., EUS has succeeded in detecting malignancy in 100% of dCC cases in addition to the higher sensitivity of EUS-FNA in dCC than in pCC (81% versus 59%, respectively) [75]. However, another study has investigated the potential role of EUS-FNA as a first-line treatment for patients with suspected pCC and revealed a higher sensitivity (79%) with 82% accuracy [76]. Moreover, the sensitivity for CC detection in patients with negative brush cytology was improved using EUS-FNA, as reported in a study by DeWitt et al. [77]. This emphasizes the important role of EUS in the early diagnosis of both dCC and pCC, particularly if other procedures were inconclusive. There is a large concern about the risk of tumor seeding through EUS-FNA, which in one study has reached 83% versus 8% in patients without prior EUS-FNA sampling [78]. Some centers may discourage transplantation for those patients who are at risk of peritoneal metastasis [65]. Hence, the benefits of EUS-FNA of primary tumors must be weighed against the risk of tumor dissemination.

2.4 Intraductal Ultrasound

ERC-assisted wire-guided small-diameter high-frequency (20 MHz) probes are introduced intraductal for better evaluation of biliary stricture and for obtaining fine details [79]. Malignancy criteria by IDUS include a hypoechoic mass with irregular margins invading surrounding tissues, bile duct wall interruption, and asymmetrical wall thickening. In addition, loss of the hyperechoic line between the tumor and nearby vessel is considered as vascular invasion [80]. IDUS was reported to be more accurate than endobiliary forceps biopsy and cytology in detecting bile duct malignancy [81]. IDUS was shown to be more accurate than standard EUS for T staging of malignant biliary strictures (IDUS 77.7%; EUS, 54.1%) but not for N staging (IDUS 62%; EUS, 62.5%) [82]. When IDUS was combined with cholangioscopy, the accuracy for evaluation of Bismuth-type hilar lesions was 95–100% [83]. However, in patients with PSC, IDUS has failed to differentiate between inflammatory and malignant strictures [84], which in turn requires an aggressive workup including cholangioscopy and tissue acquiring techniques. Currently, the substantial progress in the cholangioscopy system has limited the role of IDUS in the evaluation of ductal CC, which is now less frequently performed in many centers.

2.5 Cholangioscopy

The lumen of the bile duct can be directly visualized using cholangioscopy accompanied by targeted biopsies with relative endoscopic efficiency for the discrimination of suspicious malignant lesions. Cholangioscopic criteria highly suggestive of malignancy include dilated and tortuous vessels, intraductal nodules and masses, and infiltrative or ulcerated strictures [85]. Cholangioscopy-guided tissue acquisition appears to be superior to ERCP modalities for tissue sampling, with an overall success rate of up to 90% [86,87,88].

With the rapidly growing progress in cholangioscopy systems, the current generation of digital single-operator cholangioscopy (SOC) (SpyGlass DS; Boston Scientific, Marlborough, MA) has been the most frequently used diagnostic tool for indeterminate biliary strictures, and the easier to maneuver properties have widened their use even beyond tertiary centers [65]. In a single-center study, SOC showed a high sensitivity and specificity (88% and 94%, respectively) for definitive diagnosis in patients with indeterminate biliary lesions [89]. Nonetheless, in a recent meta-analysis, the pooled sensitivity and specificity for detection of CC using SOC were 66.2% (95% confidence interval (CI), 59.7–72.3%) and 97% (95.0% CI, 94.0–99.0%), respectively, and that for biliary strictures with negative prior brushings and biopsies were 74.7% (95% CI, 63.3–84.0%) and 93.3% (95% CI, 85.1–97.8%), respectively [90].

In the case of PSC, studies are conflicting about the role of cholangioscopy in differentiating ductal strictures, either benign or malignant. While a study by Awadallah et al. has shown unsatisfactory results for the discrimination of bile duct malignancy using cholangioscopy [91], Tischendorf et al. revealed that transpapillary cholangioscopy has higher sensitivity and specificity (92% and 93%, respectively) for malignant stricture discrimination in patients with PSC [92]. However, the technique is usually challenging with an increased risk of cholangitis, especially in patients with PSC [93].

2.6 Chromocholangioscopy and Narrow Band Imaging

Chromocholangioscopy depends on the same principle as standard chromoendoscopy for discriminating dysplastic lesions along the gastrointestinal tract. It enhances visualization through selective dye uptake and highlights alterations in the mucosal surface pattern of the bile ducts [94].

Few studies have investigated the role of chromocholangioscopy. A study by Maetani et al. differentiated benign from malignant biliary epithelium depending on the degree of methylene blue uptake; malignant tissue showed null uptake, while normal and dysplastic epithelium showed higher degrees (90% and 69%, respectively) [95]. In another study by Hoffman et al., after 55 patients underwent chromocholoscopy, normal epithelium was homogenously stained, while inflammatory and dysplastic lesions showed heterogeneous dark staining and benign strictures, such as post-liver transplant and PSC, were weakly stained [96]. On the other hand, the accuracy of chromocholangioscopy was affected by staining of the mucin and exudates overlying the biliary lesion with the biliary epithelium hiding beneath them [97].

Narrow band imaging (NBI) depends on filtering the white light into blue and green colors with different wavelengths, resulting in enhancement of vascular and surface patterns of the biliary mucosa. In a prospective multicentric study, NBI improved the cholangioscopy ability to distinguish malignant from benign lesions in 34 of 38 patients with indeterminate biliary lesions [98]. In another study, compared to white light cholangioscopy, NBI was significantly better for vascular and surface patterns of biliary lesions [99].

2.7 Confocal Laser Endomicroscopy

Probe-based confocal laser endomicroscopy (pCLE) is a novel imaging technique that provides a microscopic view of the surface epithelium and up to 250 μm of the lamina propria in real-time [100]. For biliary imaging, a confocal miniprobe is either passed within a carrying catheter through the channel of the ERCP or through the instrument channel of a cholangioscope. Intravenous fluorescein contrast is used to highlight the vasculature and extracellular matrix of examined tissues sparing the nuclei that appear dark. Low-power lasers illuminate tissues and detect the reflected fluorescent light, providing real-time images for evaluation and incorporating dynamic information such as blood flow, contrast uptake, and leakage [94, 101]. This technology seems to be useful in differentiating neoplastic from benign biliary strictures. Using a combination of specific pCLE criteria highlighting malignancy (including thick white bands (>20 mm), thick dark bands (>40 mm), dark clumps or epithelial structures) have succeeded in discriminating malignant strictures with high sensitivity (97%) but with low specificity (33%) owing to false-positive cases with inflammation related mainly to prior stent placement [102]. Hence, pCLE can considerably increase the sensitivity of detection of malignant biliary lesions and appears to be a promising diagnostic method.

2.8 Case Presentation (Figs. 18.3 and 18.4)

A 72-year-male visited our hospital with obstructive jaundice. MRCP revealed a common bile duct stricture (Fig. 18.3a). ERC revealed stenosis with upstream dilatation of the common bile duct, and IDUS showed bile duct wall interruption and asymmetrical wall thickening consistent with stenosis (Fig. 18.3b, c). Thus, cholangiocarcinoma was suspected; subsequently, brush cytology (Fig. 18.3d) and biopsy were performed under fluoroscopy (Fig. 18.3e). Finally, a plastic stent was placed into the common bile duct (Fig. 18.3f). However, no malignant findings were observed. We decided to perform cholangioscopy using a digital single-operator cholangioscopy (SpyGlass DS). The cholangioscopic images showed that the most suspicious malignancy was irregularly dilated and tortuous vessels, irregular mucosa, and lumen narrowing, and easy bleeding (Fig. 18.4b, c, d). Targeted biopsies were performed for stenosis under direct vision. A diagnosis of cholangiocarcinoma could be made with a biopsy sample.

Fig. 18.3
figure 3

(a) MRCP revealed common bile duct stricture. (b) ERC revealed stenosis with upstream dilatation of the common bile duct. (c) IDUS image showed bile duct wall interruption and asymmetrical wall thickening consistent with stenosis. (d) Brush cytology was performed for stenosis of the common bile duct. (e) Biopsy was performed under fluoroscopy. (f) A plastic stent was placed into the common bile duct

Full size image
Fig. 18.4
figure 4

(a) Cholangioscopic image showed the normal mucosa at the hilar portion. (b–d) Cholangioscopic images showed irregularly dilated and tortuous vessels, irregular mucosa, and lumen narrowing, and easy bleeding

Full size image

2.9 Conclusion

EUS- and ERCP-related procedures have become essential modalities in detecting, providing a definitive diagnosis of, and deciding upon courses of treatment for pancreaticobiliary carcinoma. Although EUS plays a central role in detecting and diagnosing pancreatic cancer, ERCP should be actively performed for pancreatic duct stenosis in which a mass cannot be visualized by EUS. In the diagnosis of cholangiocarcinoma, cytology, and biopsy under ERCP have limitations, but the advent of SOC will facilitate cholangioscopy and improve the diagnostic ability. To reduce deaths from pancreatobiliary carcinoma, new promising diagnostic methods are expected in the future.