BACKGROUND & AIMS Screening of high-risk groups for pancreatic cancer has not been adopted because of concerns regarding specificity and sensitivity. Suitability of a combination of 3 novel molecular screening techniques was investigated. METHODS Pancreatic juice was extracted from 146 patients with pancreatic ductal adenocarcinoma, chronic pancreatitis, or biliary tract stones. p53 mutations were analyzed by using a modified yeast functional assay, K-ras status was analyzed using mutation-specific real-time PCR and the proportion of p16(INK4a) promoter methylation was estimated using comparative methylation-specific real-time PCR. RESULTS p53 mutations were detected in 20 of 48 (42%) cancer cases, none of 49 controls, and 2 of 49 (4%) patients with pancreatitis. K-ras mutations were detected in 31 of 57 (54%) cancer patients, 13 of 61 (21%) controls, and 23 of 67 (34%) patients with pancreatitis. Twenty-six of 42 (62%) cancer patients had promoter methylation levels > 12%, compared with 3 of 24 (13%) controls, and 2 of 26 (8%) with pancreatitis. Mutations in p53 or high-level p16(INK4a) promoter methylation occurred in 29 of 36 (80%) patients with cancer, 3 of 24 (13%) controls, and 3 of 22 (13%) with pancreatitis. Three patients (8%) of 36 with cancer; 14 of 24 (58%) controls, and 13 of 22 (59%) patients with pancreatitis had no marker. The gallstone disease patients had a high rate of positive K-ras mutations, possibly reflecting the fact that they were not disease free. CONCLUSIONS Combination molecular analysis increased the discrimination between patients with malignant and benign disease. This level of discrimination would allow patients in high-risk groups to be stratified from negligible risk to over 50% probability of an early cancer.

Background: Mutations in the PRSS1 gene explain most occurrences of hereditary pancreatitis (HP) but many HP families have no PRSS1 mutation. Recently, an association between the mutation N34S in the pancreatic secretory trypsin inhibitor (SPINK1 or PSTI) gene and idiopathic chronic pancreatitis (ICP) was reported. It is unclear whether the N34S mutation is a cause of pancreatitis per se, whether it modifies the disease, or whether it is a marker of the disease. Patients and methods: A total of 327 individuals from 217 families affected by pancreatitis were tested: 152 from families with HP, 108 from families with ICP, and 67 with alcohol related CP (ACP). Seven patients with ICP had a family history of pancreatitis but no evidence of autosomal dominant disease (f-ICP) compared with 87 patients with true ICP (t-ICP). Two hundred controls were also tested for the N34S mutation. The findings were related to clinical outcome. Results: The N34S mutation was carried by five controls (2.5%; allele frequency 1.25%), 11/87 (13%) t-ICP patients (p=0.0013 v controls), and 6/7 (86%) affected (p<0.0001 v controls) and 1/9 (11%) unaffected f-ICP cases. N34S was found in 4/108 affected HP patients (p=0.724 v controls), in 3/27 (11%) with wild-type and in 1/81 (1%) with mutant PRSS1, and 4/67 ACP patients (all p>0.05 v controls). The presence of the N34S mutation was not associated with early disease onset or disease severity. Conclusions: The prevalence of the N34S mutation was increased in patients with ICP and was greatest in f-ICP cases. Segregation of the N34S mutation in families with pancreatitis is unexplained and points to a complex association between N34S and another putative pancreatitis related gene.

The prevalence of pancreatic cancer in the general population is too low--even in high-prevalence areas such as Northern Europe and North America (8-12 per 10(5) population)--relative to the diagnostic accuracy of present detection methods to permit primary screening in the asymptomatic adult population. The recognition that the lifetime risk of developing pancreatic cancer for patients with hereditary pancreatitis (HP) is extremely high (20% by the age of 60 years and 40% by the age of 70 years) poses considerable challenges and opportunities for secondary screening in those patients without any clinical features of pancreatic cancer. Even for secondary screening, the detection of cancer at a biological stage that would be amenable to cure by surgery (total pancreatectomy) still requires diagnostic modalities with a very high sensitivity and specificity. Conventional radiological imaging methods such as endoluminal ultrasound and endoscopic retrograde pancreatography, which have proved to be valuable in the early detection of early neoplastic lesions in patients with familial pancreatic cancer, may well be applicable to patients with HP but only in those without gross morphological features of chronic pancreatitis (other than parenchymal atrophy). Unfortunately, most cases of HP also have associated gross features of chronic pancreatitis that are likely to seriously undermine the diagnostic value of these conventional imaging modalities. Pre-malignant molecular changes can be detected in the pancreatic juice of patients. Thus, the application of molecular screening in patients with HP is potentially the most powerful method of detection of early pancreatic cancer. Although mutant (mt) K-ras can be detected in the pancreatic juice of most patients with pancreatic cancer, it is also present in patients with non-inherited chronic pancreatitis who do not progress to pancreatic cancer (at least in the short to medium term), as well as increasingly in the older population without pancreatic disease. Nevertheless, the presence of mt-K-ras may identify a genuinely higher-risk group, enabling additional diagnostic imaging and molecular resources to be focussed on such a group. What is clear is that prospective multi-centre studies, such as that being pursued by the European Registry of Hereditary Pancreatitis and Familial Pancreatic Cancer (EUROPAC), are essential for the development of an effective secondary screening programme for these patients.