Indian Journal for the Practising Doctor

Special Article

Shabnam Bashir and Neel Shah

ISSN No: 0973-516X

---

Dr Shabnam Bashir, MBBS, is senior resident, Dr Neel Shah, MS, FRCS, Senior Consultant and Laproscopic GE Surgeon, Indraprastha Apollo Hospital, New Delhi.
Contact: Dr Shabnam Bashir, Email: shabu_8171@yahoo.com

---

Cancer of the pancreas is not a common form of malignancy in ours country. However, it is a relatively silent form of cancer till late stages. In the vast majority of cases, the condition is diagnosed when already in advanced stage and unresectable. Early diagnosis depends primarily on a high degree of suspicion. Patients generally complain of intermittent back pain which may be considered innocuous by the patientor or the doctor, thus no back pain should be taken lightly especially in the past-40 people.

Cancer of the pancreas is a relatively silent but fatal form of cancer. It is usually diagnosed very late when not resectable, is generally resistant to chemotherapy, and has a very poor prognosis. At the same time, there are no widely available screening tests. Strong suspicion is a key to early diagnosis at a surgically-resectable stage.

Frequency

Globally pancreatic cancer ranks 13th in incidence and 8th as a cause of cancer death.

Table: Comparative Global Incidence

Region	Incidence/100,000 population
American blacks	13
Koreans	11
American whites	8
India	2

Most countries have an incidence of 8-12 cases per 100,000 per year. The highest incidence in the world is found in black males in the USA followed by native Hawaiian males and men of Korean, Czech, Latvain and New Zealand Maori ancestry. In India, pancreatic cancer is quite infrequent.

Table: Estimated no. of new cases and deaths for pancreatic cancer, 2002) (USA)

New Cases		New Deaths
Total	30,300	Total	29,700
Male	14,700	Male	14,500
Female	15,600	Female	15,200

Cancer of the pancreas is the fourth leading cause of cancer death in the United States, killing about 30,000 Americans each year.

In the USA, pancreatic cancer accounts for 2% of new cancers in both sexes, with 5% new cancer deaths among males and 6% among females. According to the National Cancer Institute figures, the average ageadjusted incidence rates in 2001 were 12.5 (men) and 9.8 (women) per 100,000 population, respectively. The respective average annual age-adjusted death rates for pancreatic cancer were 12.2 and 9.3 per 100,000 population, respectively for males and females. Because of the lethality of pancreatic cancer annual deaths equal the annual incidence.

More than 10,000 people die each year in Britain from pancreatic cancer. The disease is not only common it is also extremely difficult to treat. Even in such developed countries, the disease remains largely untreatable. Aptly, thus, cancer of the pancreas has been called “the challenge of the twenty-first century.”

Genetic factors Cancer of the pancreas is a genetic disease caused by mutations in DNA. These changes can be inherited or they can be acquired. The former explain why cancer of the pancreas runs in some families, and the latter may arise spontaneously during cell replication or by exposure to carcinogens as found in cigarette smoke.

A special technique, Spectral Karyotyping, colours mutated DNAs differently. The arrows in the picture (below) mark multi-colored chromosomes which have exchanged parts of their genetic material (DNA). These changes, and other genetic abnormalities observed in pancreatic cancer are currently being investigated vigorously in various hospitals, particularly the Johns Hopkins, Baltimore. Pancreatic cancers cluster in some families; estimates are that 10% of pancreatic cancers are hereditary. Many of these occur as part of rare medical syndromes; however, isolated hereditary predisposition is also seen (Familial pancreatic cancer). Syndromes associated with hereditary form of pancreatic cancers are given below:

1) Familial pancreatic cancer: It has now become clear that pancreatic cancer by itself runs in some families. In one study, 7.8% of patients reported a positive family history, as against 0.6% among controls (Odds ratio: 1). Thus, relatives of patients themselves have an increased risk for developing pancreatic cancer; the highest risk is however seen in closer relatives (Odds ratio: 5.25 in comparison with 1 for over-all relations). The inheritance pattern is autosomal dominant. Scientists are working diligently to discover the reasons why pancreatic cancer runs in some families. Recently, a susceptibility locus has been found in relation to chromosome 4q32-34.

(Photo Courtsey Johns Hopkins, Baltimore)

2) Familial breast cancer gene (BRCA2): BRCA2 is the second familial breast cancer gene identified. It has recently been shown that as many as 10% of pancreas cancers are caused by inherited defects in this gene. One particular defect in BRCA2 (a mutation called 6174 del T) is found in about 1% of individuals of Ashkenazi Jewish descent which may explain the higher rate of pancreatic cancer observed in Jews as compared to Catholics and Protestants. Some 6% of Ashkenazi Jews have also 11307k mutation/polymorphism, increasing their chances of getting cancers of the gastrointestinal system including pancreas. BRCA2 – a complex gene encoding 3350 aminoacids – was discovered on chromosome 13q12. It is an important component of the pathway that protects cells from the effects of DNA damage. The majority of mutations lead to protein truncation, and cancer develops when 2nd copy is lost. Additionally, BRCA2 has been found to be a tumour-depressive gene. BRCA2 mutation is the most common genetic identification marker in patients with familial pancreatic cancer. Apart from its relation with pancreatic cancer, BRCA2 is associated with an increased incidence of male breast cancer, prostatic cancer, cancers of gall bladder, bile duct and stomach, and multiple myeloma.

3) Peutz-Jeghers syndrome: This is a very rare hereditary syndrome in which affected family members develop polyps in their small intestines and pigmented spots on their lips. These polyps are masses of tissue that protrude from the normal surface of the intestine. The syndrome has an autosomal dominant inheritance mapped to a locus on chromosome 19p13.3. Mutations have been reported in LKB1/STK11 gene 354. Patients with Peutz-Jeghers syndrome have an increased risk of developing different malignancies (relative risk: 18), particularly pancreatic cancer, than the general population.

Peutz-Jeghers Syndrome includes:

• Perioral pigmentation
• Pigmentation of fingers
• Upper and lower gastrointestinal hamartomatous lesions,
• Small bowel and pancreatic cancers
• Colorectal cancers
• SCTAT

4) Familial melanoma: The Familial Atypical Multiple Mole Melanoma (FAMMM) syndrome is a rare hereditary syndrome in which affected family members develop skin moles and melanomas (an aggressive form of skin cancer). These patients also have an increased risk of developing pancreatic cancer.

5) Hereditary colon cancer: The Hereditary Non-polyposis Colorectal Cancer (HNPCC) syndrome strikes as many as 1 in 200 individuals and is characterized by the inherited predisposition to develop colon cancer, endometrial cancer, stomach cancer, ovarian cancer, and other genitourinal cancers. Patients with HNPCC may also have an increased risk of developing pancreatic cancer. Indeed, the DNA finding typical of HNPCC, called microsatellite instability, has recently been reported in a small (about 4%) fraction of pancreas cancers. Inheritance is autosomal dominant, and genes involved are MLH1, MSH2 and MSH6. Hereditary colon cancer runs in families, is early in onset and not associated with polyposis.

6) Hereditary pancreatitis: This rare disease is characterized by the development of recurrent episodes of severe chronic pancreatitis starting at an early age. The gene responsible for hereditary pancreatitis, called the trypsinogen gene, has been mapped to chromosome 7q35. This gene codes a cationic trypsinogen, that when mutated, fails to inactivate trypsin resulting in autodigestion of the pancreas. Chronic epithelial injury and regeneration increase the risk for pancreatic cancers. The patients are at increased risk for pancreatic pseudocysts, pancreatic exocrine failure, diabetes mellitus, and pancreatic cancers. Indian scientists at the Centre for Cellular and Molecular Biology (CCMB) and the Asian Institute of Gastroenterology have done commendable research on various genes involved in familial chronic pancreatitis. Earlier they had identified the involvement of the mutation of SPINK1 (which normaly acts as an inhibitor in pancreas, but when mutated reduces the inhibitory unction of trypsin leading to autodigestion of pancreas). Subsequently, they discovered a second candidate gene – cathepsin B – for chronic pancreatitis.

MALIGNANT MELANOMA

Inherited risk factors are thought to account for approximately 5% to 10% of melanoma syndromes such as dysplastic nevus syndrome, familial atypical multiple molemalignant melanoma syndrome (FAMMM), and melanoma-astrocytoma syndrome. Three putative melanoma-susceptibility genes have been suspected:

CMM1 mapped to chromosome 1p36 without an identifiable gene and 2nd locus mapped to 9p21 (culprit geneCDKN2). Finally, germ line mutations in the CDK4 gene on chromosome 12q14 have been documented in rare melanoma-prone families.

Hereditary predisposition to melanoma should be suspected in a patient with findings of invasive melanoma in at least two first-degree relatives and/or multiple atypical nevi in the patient. Early age at melanoma diagnosis and multiple primary melanomas are typical of melanoma-prone families.

The FAMMM syndrome is associated with pancreatic cancer. Studies have found that the risk for pancreatic cancer in these families is elevated by a factor of 13.4. Astrocytomas and sarcomas occur to excess in some melanoma families. In a study of 159 familial cases of pancreatic cancer 12% were found to show FAMMM cutaneous phenotypes, and in half of them, the CDKN2A germline mutation was identified. . Findings from this study support the existence of the FAMMM-PC syndrome caused by the CDKN2A germ line mutation. Other recent study has linked melanoma/ pancreatic cancer predisposition to a missense mutation (Gly93Trp).

Risk factors for pancreatic cancer Certain factors are known or suspected to increase the likelihood of getting the genetic mutations that may potentially result in pancreatic cancer (Box).

a) Known Risk Factors:

i) Cigarette smoking: Cigarette smoke contains a large number of carcinogens. Therefore, smoking is one of the biggest risk factors for developing pancreatic cancer. People who smoke have at least a 2-fold increased risk of developing pancreatic cancer. Early start of smoking increases the chance further: smoking during 20’s has been associated with a 3 fold increased risk of pancreatic cancer.

ii) Age: Age is the most significant risk factor for pancreatic cancer. As with prostatic cancer, the risk of developing pancreatic cancer increases linearly with age. Over 80% of the cases develop between the ages of 60 and 80. However, in the presence of genetic predisposition cases may occur before 50 years of age. Thus it is always prudent to make a thorough search in persons with back pain, abdominal pain and jaundice and those with familial history of any cancer.

iii) Race: Studies in the United States have shown that pancreatic cancer is more common in the African-Americans (blacks) than it is in the white population. Some of this increased risk may be due to different dietary habits and to higher prevalence of cigarette smoking.

iv) Gender: Men are more vulnerable to pancreatic cancer compared to women of the same age. Thus male-female ratio is about 1.5:1. This may be, in part, because men are more likely to smoke than women. This gender differential is maintained in all regions and races. The incidence among the American males has been slowly dropping over the past 2 decades, while the incidence among females has increased slightly. These trends represent the effect of changing smoking rates for men and women.

v) Religion: Pancreatic cancer is proportionally more common in Jews than the rest of the population. This may be because of a particular inherited mutation in the breast cancer gene (BRCA2) which runs in some Jewish families.

vi) Chronic pancreatitis: Long-term inflammation of the pancreas (pancreatitis) has been linked to cancer of the pancreas. A multicentre study of 2000 patients with chronic pancreatitis showed a 26-fold incease in the risk of developing pancreatic cancer. This risk increased linearly with passage of time, with 4% of patients who had chronic pancreatitis for 20 years developing pancreatic cancer. The reason for this association is not clear, but it is greatest in patients with inherited chronic pancreatitis. The relative risk of pancreatic cancer in the latter is increased more than 50-fold, and the cumulative risk rate for pancreatic cancer by age 70 is 40%. The mean age of diagnosis is 57 years as against 69 years in those without hereditary pancreatitis.

v) Diabetes: Diabetes increase chances of getting pancreatic cancer. Meta-analysis of 30 studies has concluded that patients with diabetes of at least 5 years’ duration have a 2-fold increased risk of developing pancreatic carcinoma. The reasons for this are not clear.

vi) Gastrectomized patients: Patients with partial gastrectomy appear to have an increased risk for developing pancreatic cancer.

vii) Diet: Diets high in cholesterol and nitrosamines, and fried foods may increase the risk of pancreatic cancer. High meat diets have been found to be related to a higher risk. The type of meat also may have aetiological relation – Swedish research spanning 17 years, at the Karolinska Institute in Stockholm, suggests that red meat increases the risk while the diet of poultry meat might cut down on the risk. Processed meats, fish or eggs showed no significant association.

Diets high in fruits and vegetables may reduce the risk of pancreatic cancer. This is attributable to the antioxidants in these food items. Vegetables have been found to have a stronger protective effecton than fruits. The incidenceof pancreatic cancer is higher in people with increased energy consumption. Obesity is considered a possible risk factor for pancreatic cancer and may contribute to the higher incidence of disease among the American blacks. Alcohol is not an independent risk factor unless associated with chronic pancreatitis. Coffee consumption also is not an independent risk factor.

Table 1: Causes of Pancreatic Cancer

Underlying cause	% of pancreatic cancer
Smoking	30%
Dietary factors	20%
Hereditary	5-10%
Chronic pancreatitis	<5%
Unknown	40%

Pathophysiology

Of all the pancreatic tumours, 95% develop from the exocrine portion (mainly the ductal epithelium, but also acinar cells, connective tissue, lymphatic tissue), and the rest in the endocrine part. Approximately 75% of all pancreatic carcinomas occur within the head of pancreas; 15-20% in the body and 5-10% in the tail.

Some 80-95% of pancreatic tumours have mutations in the KRAS2 gene, and 85- 98% have mutations, deletions, or hypermethylation in the CDKN2 gene, 50% mutations in TP53 and about 55% homozygous deletions or mutations of Smad4.

Some of the mutations may be found in the pancreatic cancer precursors (30% of patients with chronic pancreatitis have detectable mutations in TP16 and 10% LRAS mutations.

Spread and metastasis: A major concern when diagnosing a pancreatic cancer is whether or not the cancer has already metastasized outside the pancreas. The location of the metastases will determine whether the patient has loco-regional or metastatic disease. The location of the metastases will also determine whether the cancer is surgically resectable or unresectable. There are certain sites that pancreatic cancer may spread to that usually, but not always, eliminate surgery as a treatment option. They are:

1) Lymph nodes: Metastases to lymph nodes does not automatically eliminate surgery as a treatment option. The location of the affected lymph nodes makes a big difference. For example, the lymph nodes in the groove between the duodenum and the pancreas are a very common site of metastases. These are considered locoregional; they are routinely removed during the Whipple surgical procedure. However, the spread of cancer to more distant lymph nodes, such as lymph nodes closer to the liver, may mean the tumor is unresectable.

Typically, pancreatic cancer first metastasizes to regional lymph nodes, then to liver, and less commonly, to the lungs. It may invade surrounding visceral organs (duodenum, stomach and colon).

2) Liver: Metastases to the liver are a common finding especially with tumors in the tail and the body of the pancreas. Usually, if there is evidence of hepatic metastasis, surgery will not be an option.

3) Coeliac plexus: Coeliac plexus is grouped around the aorta and passes through opening in the diaphragm. [It is these nerves that cause back pain when pressed upon by a growing tumour. Therefore, this is the site where an alcohol is injected during an alcohol nerve block to stop the sensation of pain].

Table 2: Patterns of Neoplastic Spread for Pancreatic Cancer

Haematogenous	50%
Lymphatic	50%
Local Infiltration	30%

Fig 2: Diagramatic presentation of metastatic sites in pancreatic cancer.

4) Superior mesenteric vessels: Superior mesenteric artery and vein carry blood to and from the bowels and are closely associated with the pancreas (These blood vessels run between the uncinate process and head of the pancreas). Therefore, they may become involved by the spreading of the tumour.

5) Ligament of Treitz: This thin muscle wraps around the small intestine where the duodenum and jejunum meet. It passes behind the pancreas and is attached above to the spine and the diaphragm. Thus it is apt to be invaded by the adjoining pancreatic cancer.

6) Portal vein: This is another important blood vessel that runs right next to the pancreas. If the cancer has spread to involve the portal vein, the cancer may be considered unresectable. However, the surgeon may decide that surgery can proceed. If so, he can remove the affected portion of the portal vein and replace it with an artificial graft, where possible.

Clinical features in pancreatic cancer:

It should be noted that pancreatic cancer is a silent killer; it does not cause any symptom in early stage. Symptoms are manifested only when the tumour has grown to a sufficiently large size.

The common presenting fearures may be:

• Unexplained loss of weight and appetite
• Generalized fatigue
• Sensation of pain in the upper abdomen or back
• Jaundice with dark urine and pale stools

Patients typically report the gradual onset of nonspecific symptoms such as anorexia, malaise, nausea, fatigue and pain.

Pain is the most common presenting symptom – typically it is mid-epigastric in location, with radiation of the pain sometimes occurring to the mid- or lower back region. Back radiation indicates retroperitoneal invasion of the splanchnic nerve plexus by the tumour.

The onset of diabetes mellitus within the previous year is sometimes associated with pancreatic carcinoma. However, only about 1% of cases of newonset diabetes mellitus in adults are actually related to occult pancreatic cancer.

The most characteristic sign of cancer head of pancreas is painless obstructive jaundice, which may precede abdominal pain. The patient is the early to notice change in the colour of urine and feces, but he or his family usually can not notice jaundice till it reaches 6-8 mg%. On the contrary, the clinician can recognize jaundice when the total bilirubin is just 2.5-3 mg%. Pruritus is also an early symptom which may precede yellowishness of mucosae or skin.

Depression is usually found in patients with cancer (particularly abdominal) but is more common with pancreatic cancers. In fact, depression may be the presentation or the most prominent presenting symptom. This is partly attributable to the delayed diagnosis. Migratory thromboplebitis (Trousseau sign) and venous thrombosis occur with high frequency in pancreatic cancer.

Signficant weight loss is a characteristic feature. Weight loss is related to anorexia or subclinical malabsorption from pancreatic exocrine insufficiency caused by pancreatic duct obstruction. Patients with jaundice may have a palpable gallbladder (Courvoisers sign) and the evidence of skin excoriations from pruritus.

Patients with end-stage disease may have ascites, a palpable abdominal mass, hepatomegaly from liver metastases or splenomegaly from portal vein obstruction. Diagnosis: Pancreatic cancer often presents clinically with non-specific signs and symptoms such as pain, jaundice and weight loss. In these situations the diagnosis of pancreatic cancer may not strike the doctor’s mind and, even when suspected, pancreatic cancer can be difficult to detect and diagnose.

Differential Diagnosis

• Abdominal Aortic Aneurysm
• Ampullary Carcinoma
• Bile Duct Strictures
• Bile Duct Tumors
• Cholangitis
• Cholecystitis
• Choledochal Cysts
• Choledocholithiasis
• Cholelithiasis
• Duodenal Ulcers
• Gastric Cancer
• Gastric Ulcers
• Neoplasms of the Endocrine Pancreas
• Pancreatitis, Acute
• Pancreatitis, Chronic

Other Problems to be Considered:

• Intestinal ischemia
• Gastric lymphoma
• Pancreatic lymphoma
• Hepatocellular carcinoma (hepatoma)

The first step in the diagnosis is obtaining a complete family and medical history. One should particularly enquire about the presence of cancer (pancreatic or other) in first degree relatives (father, mother, brother, and sister).

History is followed by detailed physical examination in cluding presence of jaundice, abnormal growth or ascites. Blood, urine and stools tests are the preliminary lab tests.

Medical imaging studies are focused on lesion detection, diagnosis, and staging of disease. A variety of techniques can be used to establish a diagnosis. These techniques include transcutaneous ultrasound (TUS), computed tomography (CAT) scan, endoscopic ultrasound (EUS), magnetic reasonance imaging (MRI), endoscopic retrograde cholangiopancreatography (ERCP), and positron emission tomography (PET).

Diagnostic techniques used for pancreatic cancer:

• Angiography
• Ca 19-9
• CT Scan (Endoscopic Ultrasound (EUS)
• Fine Needle Biopsy (FNA)
• Percutaneous Transhepatic
• Cholangiography
• Endoscopic Retrograde cholangiopancreatography, (ERCP)
• Positron emission tomography, (PET)

The choice of imaging modality in a particular situation depends largely upon the local availability and expertise in the procedure. Additional considerations include accuracy of the imaging procedure for providing staging information, its ablity to simultaneously obtain tissue for a biopsy, and its capacity to facilitate palliative therapeutic procedures such as biliary stent placement or celiac neuroloysis.

Computed Tomography: Because of its ubiquitous availability and ability to image the whole abdomen and pelvis, abdominal CT scanning is usually the mainstay of initial diagnostic modalities used for assessing patients suspected to have pancreatic carcinoma.

The quality of CT scanners has been rapidly evolving. Older generation scanners can detect 70-80% of pancreatic carcinomas. Unfortunately, 40-50% of tumors smaller than 3 cm are missed, and these are the tumors most likely to be resectable. Newer models using spiral (ie, helical) CT scanning with multiple detectors and dual-phase contrast enhancement have significantly improved the sensitivity and specificity of abdominal CT findings in patients with pancreatic carcinoma. Dual-phase spiral CT findings are approximately 80% accurate for helping determine the resectability potential of pancreatic carcinoma. However, small tumors can still be missed even with the most advanced CT scanning currently available.

Detection of small pancreatic tumors with CT is primarily based on the difference in enhancement between tumor and normal parenchyma. Most ductal adenocarcinomas enhance less than the surrounding normal parenchyma. Additional findings show a focal change in texture or an abrupt change in the calibre of the pancreatic and/or bile duct. By using a dedicated technique, tumors as small as 5 mm can be detected. Accurate definition of the relationship of the tumor with the superior mesenteric artery, celiac axis, superior mesenteric portal venous confluence, and the detection of extrapancreatic disease will determine whether the tumor is potentially resectable.

The accuracy for CT to predict unresectability is well established. CT scanning can also be used to direct fineneedle aspiration of pancreatic masses.

USG: The pancreas may be involved with primary neoplasia of the exocrine and endocrine components of the gland, which creates a variety of sonographic appearances representative of the spectrum of neoplasia. In expert hands, sonography is an excellent modality for evaluation of all aspects of imaging of these tumors (ie lesion detection, diagnosis, and staging of diseas) and when used in conjunction with colour and spectral Doppler, ultrasonography reliably predicts unresectability of pancreatic cancer on the basis of vascular involvement, lymphadenopathy, and liver metastases. As with other imaging modalities, prediction of resectability is less reliable predominantly regarding microscopic tumor deposits in normal-sized lymph nodes.

Even though transcutaneous ultrasonography (TUS) is less expensive and generally more readily available than CT scanning, it has less utility in pancreatic carcinoma than CT scanning because the pancreas is often obscured by overlying gas. Additionally, the depth of the pancreas from the abdominal wall limits transcutaneous ultrasonic imaging to lower frequency (2-5 MHz), and thus, a lower-resolution ultrasonographic image is obtained. Therefore, TUS can help detect only 60- 70% of pancreatic carcinomas, and similar to CT scanning, more than 40% of the lesions smaller than 3 cm are missed.

TUS, however, is very useful as an initial screening test in evaluating patients who present with possible obstructive jaundice. By helping to detect intrahepatic or extrahepatic bile duct dilation, abdominal ultrasonography can help rapidly and accurately assess whether or not a patient has a biliary obstruction. However, other studies, such as abdominal CT scanning, EUS, ERCP, or magnetic resonance cholangiopancreatography (MRCP), usually should then be performed to definitively diagnose the source of biliary obstruction.

Fig 1: Computerized tomographic scan showing a pancreatic adenocarcinoma of the pancreatic head. The gallbladder (gb) is distended because of biliary obstruction. The superior mesenteric artery (sma) is surrounded by tumor, making this an unresectable T4 lesion.

Fig 2: Abdominal CT scan of a small, vaguely seen, 2-cm pancreatic adenocarcinoma (mass) causing obstruction of both the common bile duct cbd) and pancreatic duct (pd).

Endoscopic ultrasonography (EUS) obviates the physical limitations of TUS by placing a high-frequency ultrasonographic transducer on an endoscope, which is then positioned in the stomach or duodenum endoscopically to help visualize the head, body, and tail of the pancreas.

Fig 3: Endoscopic ultrasound of a 2.2-cm pancreatic adenocarcinoma of the head of the pancreas obstructing the common bile duct (CBD) but not invading the portal vein (PV) or superior mesenteric vein (SMV). Findings from endoscopic ultrasound–guided fine-needle aspiration revealed a moderately to poorly differentiated adenocarcinoma. Abdominal CT findings did not show this mass, and an attempt at endoscopic retrograde cholangiopancreatography at another institution was unsuccessful.

Additionally, because of the proximity of the pancreas to the EUS transducer, high-frequency ultrasonography (7.5-12 MHz) can be used to produce very high-resolution (submillimeter) images. Where expert EUS is available, it has proven to be the most sensitive and specific diagnostic test for pancreatic cancer.

In numerous series, EUS has detection rates of 99-100% for all pancreatic carcinomas, including those smaller than 3 cm. EUS is as accurate as ERCP or MRCP for assessing the etiology of obstructive jaundice.

An additional significant diagnostic advantage is EUS-guided fine-needle aspiration, which allows for the simultaneous cytologic confirmation of pancreatic carcinoma at the time of EUS diagnosis.

EUS appears to be equivalent to dual-phase spiral CT scanning for assessing tumor resectability potential.

Endoscopic ultrasonography (EUS) is used to explore a distal duct obstruction. It is used to explore the ampullary region of the pancreas. EUS-guided biopsy provides a cytologic diagnosis when a mass is identified.

Pancreatography: If a diagnosis is not achieved with CT, MRCP (mageticresonance cholangiopancreatography) or cholangiography, such as ERCP (endoscopic retrograde cholangiopancreatography) or percutaneous transhepatic cholangiography, is used to define the obstruction. Cholangiography also allows brushing for cytologic evaluation.

ERCP is highly sensitive for helping detect pancreatic carcinoma. Of patients with pancreatic adenocarcinoma, 90-95% have abnormalities on ERCP findings. However, the changes observed on ERCP are not always highly specific for pancreatic carcinoma and can be difficult to differentiate from changes observed in patients with chronic pancreatitis.

ERCP is more invasive than the other diagnostic imaging modalities available for pancreatic carcinoma. ERCP also carries a 5-10% risk of significant complications with the procedure.

Brush cytology and forceps biopsy at the time of ERCP have been used to diagnose pancreatic carcinoma histologically; however, in most series, the yield of a cytologic diagnosis with these procedures has been less than 50%.

ERCP findings provide only limited staging information, but ERCP does have the advantage of allowing for therapeutic palliation of obstructive jaundice with either a plastic or metal biliary stent.

Fig 4: T staging for pancreatic carcinoma. T1 and T2 stages are confined to the pancreatic parenchyma. T3 lesions invade local structures such as the duodenum, bile duct, and/or major peripancreatic veins, and T4 lesions invade surrounding organs (eg, stomach, colon, liver) or invade major arteries such as the superior mesenteric or celiac arteries.

Fig 5: Hematoxylin and eosin stain of a pancreatic carcinoma. Note the intense desmoplastic response around the neoplastic cells. The large amount of fibrotic reaction in these tumors can make obtaining adequate tissue by fine-needle aspiration difficult.

Fig 6: Cytologic samples from fine-needle aspirations (rapid Papanicolaou stain) of pancreatic adenocarcinomas. (A) Well differentiated, (B) moderately differentiated, (C) moderate to poorly differentiated, (D) poorly differentiated tumor.

The role of MRI in pancreatic cancer has been less well studied than the role of CT scanning. It does not appear to be superior to spiral CT scanning. However, in patients with jaundice, MRCP can be used as a noninvasive method for imaging the biliary tree and pancreatic duct.

Whether MRCP is as sensitive and specific for pancreaticobiliary pathology as other procedures is still being investigated. Because of the difficulty of working within intense magnetic fields, MRI is limited by the inability to perform MRI-directed needle aspirations; however, this technology is undergoing rapid change.

Fast and ultrafast dynamic MRI using the paramagnetic contrast agents and fat-suppression technique have improved the ability of MRI to image the pancreas. While some studies suggest that helical CT is superior to MRI in staging and detection of pancreatic cancer, another study has found that MRI can detect more small tumors than CT.

Positron emission tomography (PET) scanning uses 18Ffluorodeoxyglucose (FDG) to image both the primary tumor and metastatic disease. PET scanning can be especially useful in looking for occult metastatic disease. Its role in pancreatic cancer evaluation and management is still under investigation. False-positive PET scans have been reported in pancreatitis.

Biopsy: Although all the imaging techniques may reveal a suspicious mass in the pancreas, the “gold standard” for diagnosing pancreatic cancer remains histopathology. Tissue for microscopic examination can be obtained by fine needle biopsy, by tissue needle cone biopsy or by excisional biopsy at the time of laparotomy. Fine-needle aspiration (FNA) is a cytologic technique in which cells are aspirated from a tumour using a needle and syringe with the application of negative pressure. The technique can also be performed using image-directed guidance (EUS-guided, CTguided) and is particularly helpful in the diagnosis of relatively inaccessible lesions. The aspirated tissue consists of disaggregated cells rather than intact tissue. Diagnosis of malignancy, therefore, usually depends on detection of abnormal intracellular features, such as nuclear pleomorphism, and so the margin of error in diagnosis of many tumour types is higher than with other biopsy techniques.

The necessity of obtaining a cytologic or tissue diagnosis of pancreatic cancer prior to operation remains controversial.

Some centres advocate the practice of aggressively operating on all patients thought to have pancreatic cancer and argue against a preoperative tissue diagnosis. The contention in these centres is that negative findings after preoperative fine-needle aspiration may just be a sampling error and, thus, should not stop a pancreaticoduodenectomy if a potentially resectable pancreatic neoplasm is strongly suggested based on preoperative testing. Additionally, preoperative attempts at fineneedle aspiration or biopsy of the pancreas might contaminate the peritoneum with tumor cells.

Other surgeons are hesitant to perform an operation with as much potential for morbidity as a pancreaticoduodenectomy on patients without a positive tissue or cytologic diagnosis of cancer. Additionally, tissue diagnosis is almost always required prior to initiation of chemotherapy, radiation therapy (whether palliative or neoadjuvant), or nonoperative palliation of obstructive jaundice using permanent metallic stents.

Studies of the risk of peritoneal contamination with CT-guided biopsy have suggested that this risk is actually very low. EUS-guided fine-needle aspiration provides the additional advantage of aspiration through tissue that would ultimately be included in the operative field should the patient undergo resection.

Additionally, the histology of a pancreatic tumour can change the surgical approach to the tumor. For example, a pancreatic lymphoma should be treated medically rather than with operative resection.

Finally, patients and their families usually want a definitive diagnosis prior to making major therapeutic decisions.

EUS-guided fine-needle aspiration has proven to be the most effective means for making a definitive cytologic diagnosis of pancreatic carcinoma. Using EUS-guided fineneedle aspirations, a cytologic diagnosis can be made in 85-95% of patients. A recent study has also suggested that transcutaneous aspiration may be associated with a higher risk of peritoneal tumour spread than aspiration with EUS. Thus, for potentially resectable tumours, EUS-guided fine-needle aspiration is the preferred biopsy technique, if it is available and if a biopsy needs to be obtained.

The yield of CT-guided fine-needle aspiration or biopsy findings is approximately 50-85% in the lesions that are visible on CT scanning.

Histologic Findings: Of all pancreatic cancers, 80% are adenocarcinomas of the ductal epithelium. Only 2% of tumors of the exocrine pancreas are benign. Less common histologic appearances of exocrine pancreatic cancers include giant cell carcinoma, adenosquamous carcinoma, microglandular adenocarcinoma, mucinous carcinoma, cystadenocarcinoma, papillary cystic carcinoma, acinar cystadenocarcinoma, and acinar cell cystadenocarcinoma. Very rarely, primary connective tissue cancers of the pancreas can occur. The most common of these is primary pancreatic lymphoma.

Cystic neoplasms of the pancreas account for fewer than 5% of all pancreatic tumors. These consist of benign serous cystadenomas, premalignant mucinous cystadenomas, and cystadenocarcinomas.

Patients can also develop tumours of the islet cells of the pancreas. These can be functionally inactive islet cell carcinomas or benign or malignant functioning tumors such as insulinomas, glucagonomas, and gastrinomas.

Other tests: Angiography is useful to determine if the vessels around the pancreas are involved by the tumour and a estimation of a tumour marker in blood (as CA19-9) can be useful in following the effectiveness of treatment.

Diagnostic algorithm

Evaluation algorithm followed in the most specialized centres of the USA and UK is given below:

Most patients suspected of having pancreatic carcinoma are initially studied with transcutaneous abdominal ultrasonography and/or spiral CT scanning (usually not with dual-phase contrast thincut pancreatic protocols). Patient management thereafter can vary from institution to institution depending on local expertise and interest.

If patients have obvious hepatic metastatic disease based on initial TUS or CT findings, they undergo a CT- or TUSguided biopsy of one of the liver metastases and then proceed to palliative therapy. Patients with a suggested or definite pancreatic mass observed on abdominal CT scanning or TUS or those who are still considered to have pancreatic cancer but do not have an obvious pancreatic mass need to have more definitive imaging studies. This can be done using high-quality thin-cut CT scanning with dual phase contrast and/or by other procedures such as endoscopic ultrasonography.

Transcutaneous ultrasonography is available at the secondary care level (at subdistrict hospitals) in Jammu & Kashmir, but CTs are found at tertiary level only. Highquality endoscopic ultrasonography and EUS-guided fine-needle aspiration is not readily available, In centres where it is available, when a pancreatic mass is observed, EUS-guided fine-needle aspiration is performed to confirm the disease cytologically. At the same time, the patient is staged using EUS to determine resectability potential.

Patients thought to have resectable tumors based on EUS findings proceed directly to operative intervention. If tumors are deemed unresectable based on EUS findings and patients have obstructive jaundice, they proceed directly to therapeutic stent placement with ERCP while under the same endoscopic sedation. Most patients undergo dedicated pancreas protocol CT scanning to complete preoperative staging if the initial CT scan was not of adequate quality.

Patients with unresectable disease are offered chemotherapy for their disease. In institutions without EUS and EUS-guided fine-needle aspiration capabilities, spiral CT scanning with CT-guided pancreatic fineneedle aspiration or biopsy plays the central role in evaluation.

The most difficult clinical situation in which to diagnose pancreatic carcinoma is in the patient with underlying chronic pancreatitis. In this situation, all of the above imaging studies may show abnormalities that may not help differentiate between pancreatic carcinoma and chronic pancreatitis. Even tumour markers can be elevated in patients with chronic pancreatitis. In these patients, one must often combine multiple imaging modalities, close clinical follow-up, serial imaging studies, and occasionally empiric resection to diagnose an underlying pancreatic carcinoma.

Staging: The diagnosis of pancreatic cancer is not complete unless staging is done. Subsequent decisions about treatment will be based upon the stage assigned.

The results of various diagnostic tests will indicate how far the cancer has progressed and determine the stage. However some centres advocate performing a staging laparoscopy or laparotomy before proceeding to attempted resection. A few centres also advocate performing intraoperative laparoscopic ultrasonography to help further assess the tumor stage and to look for occult metastases. Using this approach, a significant number of patients are found to have previously unsuspected peritoneal or liver metastases. The operations avoided by staging laparoscopy largely depend on how aggressively and accurately the patient was staged preoperatively.

Generally speaking, different stages carry different prognoses (see table below). The charts reproduced below are commonly used to stage pancreatic tumors.This staging process is based on the modifications of the American joint Committee on Cancer (AJCC 2002).

Table: 5 yr relative survival by stage of diagnosis; USA, 2001.

Stage at diagnosis	5 yr Survival rate
Localized	16%
Regional	7%
Distant	2%
All stages	4%

Definition of Tumor (T)
TX	Primary tumor cannot be assessed
T0	No evidence of primary tumor
Tis	In situ carcinoma
T1	Tumor limited to the pancreas, 2 cm or less in greatest dimension
T2	Tumor limited to the pancreas, more than 2 cm in greatest dimension
T3	Tumor extends directly into any of the following: duodenum, bile duct, peripancreatic tissues
T4	Tumor extends directly into any of the following: stomach, spleen, colon, adjacent large vessels

In the TNM classification, the stage of a pancreatic cancer (exocrine portion) is determined by 3 factors: T: (location and size of tumour), N: (evidence of metastases in lymph nodes close to the cancer) , and M: (evidence of distant metastasis). A value is assigned for each of these factors based on the diagnostic test results.

At initial presentation, only 20% of patients present with stage I disease, 40% present with locally advanced disease, and 40% present with disease metastatic to nodes or distant sites.

To date, studies show that EUS is approximately 70-80% accurate for correctly staging pancreatic carcinoma. EUS appears to better assess involvement of the portal vein/superior mesenteric vein. Spiral CT scanning with dual-phase contrast probably has similar or better overall accuracy and is especially good at assessing major arterial involvement. EUS is better than CT scanning to help detect abnormal lymph nodes around the pancreas and celiac axis. Furthermore, with the addition of EUSguided fine-needle aspiration, EUS can help cytologically document metastatic disease in suggestive lymph nodes.

Definition of Regional Lymph Nodes (N)
NX	Regional lymph nodes cannot be assessed
N0	No regional lymph node metastasis
N1	Regional lymph node metastasis pN1a= metastasis in a single regional lymph node pN1b= metastasis in multiple regional lymph nodes
Definition of Distant Metastasis (M)
MX	Distant metastasis cannot be assessed
M0	No distant metastasis
M1	Distant metastasis

Stage Grouping for Primary Tumors of the Exocrine Pancreas
Stage	Tumor	Regional Lymph Nodes	Distant Metastases
Stage 0	Tis	N0	M0
Stage IA	T1	N0	M0
Stage IB	T2	N0	M0
Stage IIA	T3	N0	M0
Stage IIB	T1 T2 T3	N1 N1 N1	M0 M0 M0
Stage III	T4	any N	M0
Stage IV	any T	any N	M1

Screening tests

Although there are no screening tests currently available to screen the general population, investigators in the USA are working on developing some tests. It is hoped that these new tests may prove effective, particularly when applied to select groups of patients known to have an increased risk of developing pancreatic cancer such as individuals with a strong family history.

Many tumour markers have been identified, some of them are available for use while others are still being studied.

1. Carbohydrate antigen 19-9:

CA19-9 is the most useful available tumour marker for following the progression of the disease but is only 80% accurate in identifying patients with pancreatic cancer. The value of CA19-9 in therapy monitoring is established; in most followup studies the sensitivity reached 100% in detection of recurrent disease. Due to the inability of CA19-9 to identify early potentially curable disease, several other markers have been studied including SPAN-1, CA 242, CA-50, CEA, DUPAN-2, elastase-1, tissue polypeptide antigen and tissue polypeptidespecific antigen. However, these markers have not performed nearly as well as the CA19-9.

The major useful tumour marker for pancreatic carcinoma is the carbohydrate antigen 19-9 (CA 19-9). It is a murine monoclonal antibody originally made against colorectal cancer cells. The CA 19-9 antigen is a sialylated oligosaccharide that is most commonly found on circulating mucins in cancer patients. It is also normally present within the cells of the biliary tract and can be elevated in acute or chronic biliary disease. In healthy people 5-10% lack the enzyme to produce CA 19-9.

The reference range of CA 19-9 is less than 33-37 U/mL. Of patients with pancreatic carcinoma, 75-85% have elevated CA 19-9 levels. In the absence of biliary obstruction or benign pancreatic disease, a CA 19-9 value greater than 100 U/mL is highly specific for malignancy, usually pancreatic.

Evaluation of CA 19-9 levels has been used as an adjunct to imaging studies for helping determine the resectability potential of pancreatic carcinoma. Fewer than 4% of patients with a CA 19-9 level of more than 300 U/mL have been found to have resectable tumors.

Unfortunately, CA 19-9 is least sensitive for small early-stage pancreatic carcinomas and thus has not proven to be effective for the early detection of pancreatic cancer or as a screening tool.

An elevated CA 19-9 level is found in 0.2% of an asymptomatic population older than 40 years. Of these elevations, 80% are false-positive results. If only symptomatic patients are studied, 4.3% have elevated CA 19-9 levels. Two thirds of these results are false positive. To date, no standardized role has been found for CA 19- 9 measurements in pancreatic carcinoma, and the usefulness of this practice must still be classified as only a supplement to other diagnostic modalities.

An ideal pancreatic cancer screening test should be a safe, inexpensive, highly accurate test that reliably diagnoses pancreas cancer at a stage when it is not causing symptoms in the patient. This would provide an opportunity to take appropriate and effective action to treat and potentially cure the disease. Currently, we cannot offer a screening test for pancreas cancer that could meet these demands. However, the potential for screening this disease is considerably improved compared to just a few years ago.

First of all, we know more about which individuals have a genetic predisposition for developing this disease. Prior to the 1990’s, it was not widely appreciated how often pancreatic cancer is an inherited disease. Now we know that individuals with abnormalities in certain genes, such as BRCA2, p16, HNPCC, and individuals with histories of familial pancreatitis and Peutz-Jeghers syndrome are all predisposed to pancreatic cancer. These are important discoveries. However, individuals who inherit damage to one of these genes still represent only a modest proportion of all those individuals at risk of developing pancreatic cancer.

Genetic testing for mutations in CDKN2 is now available clinically, however, technically not used for screening. Four healthy carriers of CDKN2 mutations, all children of parents who died of pancreatic carcinoma, are currently being followed in the USA. Regular endoscopic ultrasound of the pancreas is ongoing and pancreatic juice is being collected for analysis of telomerase activity, Kras mutations, CA19-9, and other candidate biomarkers. K-ras mutations, detected by PCR, was found to contribute to the clinical decision process when the cancer is clinically suspected.

Individuals in melanoma-prone kindreds should perform monthly skin selfexamination, possibly with the assistance of a spouse or other relative. Comprehensive dermatologic evaluation by a knowledgeable dermatologist should be done semiannually. Any suspicious lesions should be excised. Sunburn should be avoided, and use of ultraviolet A/B-blocking sunscreens is encouraged.

Secondly, we now know much more about the changes in the DNA and other molecules in the pancreas cells that give rise to pancreatic cancer. Most of these changes are not inherited and occur as the result of factors such as smoking, diet, and age. Importantly, pancreatic cancer researchers are characterizing these specific changes in DNA and other molecules. It is hoped that over the next few years, specific and extremely sensitive screening tests will be developed. Such a screening test will be able to detect pancreatic cancer at an early stage when it still cannot be visualized using stateof- the-art diagnostic imaging techniques. Once such a screening panel is in place, it can be offered to individuals who, from their family history, know that they are at particular risk of developing this disease.

Currently many new screening tools are being investigated. Carcinoembryonic antigen (CEA) is a high molecular weight glycoprotein found normally in fetal tissues. It has commonly been used as a tumour marker in other gastrointestinal malignancies. However, it has minimal utility in pancreatic carcinoma. The reference range is less than or equal to 2.5 mg/mL. Only 40-45% of patients with pancreatic carcinoma have elevations in CEA levels. Multiple other benign and malignant conditions can lead to elevated CEA levels; thus, CEA is not a sensitive or specific marker for pancreatic cancer.

Pure pancreatic juice (PPJ) cytology and tumour marker (CEA) determination has been evaluated for the diagnosis of early pancreatic cancer. The authors conclude that combined cytology and CEA determination in PPJ, with a diagnostic accuracy of 93%, could be useful for the diagnosis of pancreatic cancer even in early stages.

Other tumour markers which are currently being studied for early diagnosis of pancreatic cancer include CA 242 and CA 50. Studies comparing CA-242 levels in benign and malignant pancreatic tumours found no difference and concluded that this marker can not compare with CA 19-9. In other studies the two tumour markers had almost a similar prognostic value.

Tissue polypeptide specific antigens (TPS) has been studied as a marker for differentiation between chronic pancreatitis and pancreatic cancer. Almost complete discrimination of the two was attained; the researchers concluded that TPS is more useful than CA 19-9 in the differential diagnosis of the two conditions.

References

1. Am Cancer Society, Surveillance Research, 1998, NCI – Surveillancve, Epidemiology, End Result Programme, 2001.
2. Am Cancer Society, Cancer Facts & Figures, 2002.
3. Banfi G. Bravi S. Ardemagni A. Zerbi A. CA 19.9, CA 242 and CEA in the diagnosis and follow-up of pancreatic cancer. International Journal of Biological Markers. 11(2):77-81, 1996 Apr-Jun.
4. Bassin A, P. O’Leary. The first report of a successful ancreaticoduodenectomy. The American Surgeon, Atlanta, 1995, 61: 845- 846.
5. Birkmeyer et al. Analysis of pancreatic cancer mortality. New England J Medicine, 2002 Apr 11;346 (15):1128-37)
6. Cancer Research Vol 57, 1537-1546, 1997
7. Falk RT, Pickle LW, Fonthan ET. Lifesyle risk factors for pancreatic cancer in Louisiana – a Case Control Study. Am J Epid 1988;128:324-36
8. Ghadrian P, Boyle P, Simand A. Repeated familial aggregation of pancreatic cancers in a population-basedf case-control sudy in Montreal, Canada. Int J Pharmaceutics 1991;10:183-96
9. Gattani AM. Mandeli J. Bruckner HW. Tumor markers in patients with pancreatic carcinoma. Cancer. 78(1):57-62, 1996 Jul 1.
10. Giardello FM, Welsh SB, Hamiltone E. Mutations of the LKB1 (STK11) gene in PJS patients. J Med Genetics 1998;36:365-8
11. Gordon TA, Bowman HM, Tielsch JM, Bass EB, Burleyson GP, Cameron JL. State-wide regionalization of pancreaticoduodenectomy and its effect on in-hospital mortality. Ann Surg. 1998 Jul;228(1):71-8.
12. Hasholzner U. Stieber P. Reiter W. Zimmermann A. Hofmann K. Schalhorn A. CA 242 in comparison with established tumour markers in colorectal, pancreatic and lung cancer. Anticancer Research. 19(4A):2477-80, 1999 Jul-Aug.
13. Ichihara T. Nomoto S. Takeda S. Nagura H. Sakamoto J. Kondo K. Horisawa M. Nakao A. Clinical usefulness of the immunostaining of the tumor markers in pancreatic cancer. Hepato- Gastroenterology. 48(40):939-43, 2001 Jul-Aug.
14. Luken SI, Peterson GM, Gruler SB. Familial colorectal cancer linked to a hypermutable gene in APC. Nat Gene 1997; 17: 97-83.
15. Matsumoto S. Harada H. Tanaka J. Ochi K. Seno T. Tsurumi T. Kunichika K. Evaluation of cytology and tumor markers of pure pancreatic juice for the diagnosis of pancreatic cancer at early stages. Pancreas. 9(6):741-7, 1994 Nov.
16. Micke O, Bruns F, Schafer U, Kurowski R, Horst E, Willich N. CA 19-9 in the therapy monitoring and follow-up of locally advanced cancer of the exocrine pancreas treated with radiochemotherapy. : Anticancer Res. 2003 Mar-Apr;23(2A):835-40.
17. Nazli O. Bozdag AD. Tansug T. Kir R. Kaymak E. The diagnostic importance of CEA and CA 19-9 for the early diagnosis of pancreatic carcinoma. Hepato- Gastroenterology. 47(36):1750-2, 2000 Nov- Dec.
18. NCI Surveillance, Epidemiology, End Result Programme, 1992-97
19. NCI SEER Programme, 1979-1998. Am Cancer Society, Surveillance Research, 2002.
20. Peters J H