**3. Investigations of pancreato-biliary cancers**

The imaging modalities involve in the detection, staging and management of pancreatic cancer are computer tomography (CT), magnetic resonance imaging (MRI), endoscopic retrograde cholangiopancreatography (ERCP), endoscopic ultrasound (EUS) and positron emission tomography (PET) (Clarke et al., 2003; Chang et al., 2009; Peddu et al., 2009). The diagnosis can potentially be made by any, or a combination of the above modalities. The roles and relative importance of these imaging modalities have changed over the last few decades and continue to change with rapid technological advancement in medical imaging.

Base on current best available evidence, CT should be used as first line for diagnosis, staging and the assessment of resectability in pancreato-biliary cancer. MRI should be reserved for patients with iodine contrast allergy or who cannot be exposed to radiation or to be used as an adjunct to CT in patients with suspicious liver lesions that need to be to better characterized (Clarke et al., 2003; Peddu et al., 2009). MR cholangio-pancreatography (MRPC) is an essential part of the evaluation for cholangiocarcinoma as it can identify the luminal involvement of the cancer as well as the road map of the biliary tree (Patel, 2006). Such information is not only important in local staging but also critical in determining respectability, type of surgery and/or identifying the dominant obstructive ductal system for biliary drainage. In selected cases, cholangioscopy is helpful by providing direct endoscopic visualization of the intra-ductal lesion that responsible for the biliary stricture (Patel, 2006). The recent development of SpyGlass cholangioscopy system has also allowed tissue sample under direct vision.

EUS should be used for local staging and assessment of resectability if it remains inconclusive on non-invasive imaging modalities (Chang et al., 2009; Iglesias Garcia et al., 2009). It should also be used in patients with a high clinical suspicion of a lesion that has not been clearly demonstrated using other modalities. EUS-FNA should also be the biopsy route of choice in patients where a tissue diagnosis or tissue from regional lymph nodes may alter the course of treatment, or if neo-adjuvant treatment is contemplated. If there is disagreement between CT and EUS images, then laparotomy and trial of dissection should be considered (Chang et al., 2009). PET/CT should be used selectively such as when metastatic disease is suspected but has not been demonstrated with other imaging modalities (Nguyen and Bartholomeusz, 2011; Serrano et al. 2010). The availability and local expertise of each imaging modality will also influence their use. A suggested management algorithm for patients with suspected pancreato-biliary cancer is shown in Figure 1.

## **3.1 Tumour markers**

The role of tumour marker in the diagnosis and management of pancreato-biliary cancers remains controversial (Balzano and Di Carlo, 2008). Carbohydrate antigen 19-9 (CA19-9), which is caused by the up-regulation of glycosyl transferase genes, is the most commonly used marker and can provide useful diagnostic and prognostic information (Duffy et al.

Pancreato-Biliary Cancers – Diagnosis and Management 465

Fig. 2. Examples of EUS-guide FNA of a pancreatic mass in the body (panel A), a liver lesion

contrast, for lesions in the pancreatic body and tail, where the needle track is not resected, the risks and benefits of pre-operative biopsy should be carefully assessed on an individual basis. Due to its anatomical position, tissue acquisition from biliary lesion via EUS guided FNA is more difficult and in general, the diagnostic yield is lower than that for pancreatic cancer and is dependent on the location of the lesion. As it is easier to visualize and access the distal biliary lesions, the diagnostic yield is significantly higher in distal compared with

In cases where the diagnosis of the biliary stricture remains unclear after conventional MDCT, MRI and EUS evaluation, directly visualization of the appearance of the ductal strictures and biopsy can be helpful in differentiating benign from malignant disorders (Figure 3). Although video "mother-baby" cholangioscope provides high quality images, it is fragile and often lack of accessory channel for tissue sampling (Nguyen, 2009; Nguyen et al., 2009). The diagnostic yield of malignancy based on cholangioscopic appearance of the intra-ductal lesion varied from 70% to 88% (Nguyen, 2009; Nguyen et al., 2009). Currently, tissue sampling is only possible with the single-operator disposable SpyGlass system, which has a 1.2mm accessory channel. Although SpyGlass guided tissue sample is successful in up to 96% of cases, its overall accuracy in confirming a malignant stricture is only modest (49% of cases) (Nguyen, 2009; Nguyen et al., 2009). This is mainly due to the poor sensitivity of SpyGlass guided biopsy in the diagnosis of malignancy from extrinsic cancers (8%) as

compared to that of intrinsic cancers (66%) (Nguyen, 2009; Nguyen et al., 2009).

Patients with suspected or confirmed diagnosis of pancreato-biliary malignancy should be assessed by a multidisciplinary team and stratified as resectable, borderline resectable,

**4. Therapeutic approaches for pancreato-biliary cancers** 

(panel B) and pathological celiac node (panel C).

**3.3 Cholangioscopy** 

proximal lesions (81% vs 59%) (Mohamadnejad et al. 2011).

Fig. 1. Suggested algorithm for the evaluation and management of patients with suspected pancreatic.

2010). Its sensitivity (70%-90%) and specificity (43%-91%) for diagnosing pancreatic cancer are only modest and can be falsely increased by high serum bilirubin (Duffy et al. 2010). However, for those with confirmed pancreatic cancer, high serum CA19-9 is associated with a worse survival (Park et al., 2008). Similarly, in patients who undergo curative resection for pancreatic cancer, a normalizing post-operative CA19-9 level is associated with a longer median and disease-free survival compared to those with persistently high level (Duffy et al., 2008; Balzano and Di Carlo, 2008).

#### **3.2 Tissue sampling**

A distinct advantage of EUS is its ability to obtain tissue via fine needle aspiration (FNA). This approach is superior to percutaneous biopsy (via US or CT guided) in the investigation of pancreato-biliary malignancies with higher diagnostic yield (84% vs. 62%) and significantly lower risk of tumour seeding from the needle tract (<2% vs. 16%) (Paquin et al., 2005). Apart from biopsy of the primary tumour, it also has the ability to biopsy lymph nodes, liver lesions and ascitic fluid, which is critical in accurate staging and avoiding unnecessary resection (Figure 2). For pancreatic head lesions, the possibility of seeding is eliminated, as the needle track is included in the resection specimen (Yamao et al., 2005). In

Fig. 1. Suggested algorithm for the evaluation and management of patients with suspected

2010). Its sensitivity (70%-90%) and specificity (43%-91%) for diagnosing pancreatic cancer are only modest and can be falsely increased by high serum bilirubin (Duffy et al. 2010). However, for those with confirmed pancreatic cancer, high serum CA19-9 is associated with a worse survival (Park et al., 2008). Similarly, in patients who undergo curative resection for pancreatic cancer, a normalizing post-operative CA19-9 level is associated with a longer median and disease-free survival compared to those with persistently high level (Duffy et

A distinct advantage of EUS is its ability to obtain tissue via fine needle aspiration (FNA). This approach is superior to percutaneous biopsy (via US or CT guided) in the investigation of pancreato-biliary malignancies with higher diagnostic yield (84% vs. 62%) and significantly lower risk of tumour seeding from the needle tract (<2% vs. 16%) (Paquin et al., 2005). Apart from biopsy of the primary tumour, it also has the ability to biopsy lymph nodes, liver lesions and ascitic fluid, which is critical in accurate staging and avoiding unnecessary resection (Figure 2). For pancreatic head lesions, the possibility of seeding is eliminated, as the needle track is included in the resection specimen (Yamao et al., 2005). In

pancreatic.

al., 2008; Balzano and Di Carlo, 2008).

**3.2 Tissue sampling** 

Fig. 2. Examples of EUS-guide FNA of a pancreatic mass in the body (panel A), a liver lesion (panel B) and pathological celiac node (panel C).

contrast, for lesions in the pancreatic body and tail, where the needle track is not resected, the risks and benefits of pre-operative biopsy should be carefully assessed on an individual basis. Due to its anatomical position, tissue acquisition from biliary lesion via EUS guided FNA is more difficult and in general, the diagnostic yield is lower than that for pancreatic cancer and is dependent on the location of the lesion. As it is easier to visualize and access the distal biliary lesions, the diagnostic yield is significantly higher in distal compared with proximal lesions (81% vs 59%) (Mohamadnejad et al. 2011).
