In recent years, the incidence of pancreatic cancer has been increasing in industrialized nations, and it now occurs with a frequency of 8–10 per 100,000 population and is more common in males. Pancreatic cancer accounts for about 5 % of all cancer deaths.
Up to 6 % of individuals with pancreatic carcinoma may have a positive family history of the condition, and there is up to fivefold increased risk of this cancer in the first-degree relatives of index cases (Fernandez et al. 1994; Ghadirian et al. 2002). However, familial clustering is relatively rare, but the Tumours may occur as part of the spectrum of malignancies in Lynch syndrome, in hereditary cutaneous malignant melanoma families, in families segregating BRCA1, and especially BRCA2 mutations. It also occurs as part of Li–Fraumeni syndrome, and occurs with increased frequency in Peutz–Jeghers syndrome and in ataxia telangiectasia (Flanders and Foulkes 1996; Lim et al. 2004; Beggs et al. 2010; Mehenni et al. 2006). The occurrence of pancreatic cancer in kindred with several cases of breast cancer is a significant predictor of a germline mutation in BRCA2 (Ozcelik et al. 1997). In addition, families have been reported in which several cases of cancer of the pancreas have occurred in a sibship, albeit with late age at onset. Although autosomal recessive inheritance has been suggested (Friedman and Fialklow 1976), there is evidence of autosomal dominant inheritance in most affected families (Ehrenthal et al. 1987; Bartsch et al. 2012; Lynch et al. 1990; Solomon et al. 2012; Evans et al. 1995), and dominant inheritance with incomplete penetrance would be a unifying explanation. The age at onset, histology, sex distribution, and survival of familial pancreatic cancer appear to be similar to those of nonfamilial cases. Mutations in CDKN2A (encoding p16) on chromosome 9p21 are present in some families with multiple cases of cutaneous malignant melanoma, and in some families, the risk of pancreatic cancer is very high, notably in individuals with the Dutch founder mutation (Goldstein et al. 1995a, b, 2004; Ghiorzo et al. 1999; Lynch et al. 2002; de Vos tot Nederveen Cappel et al. 2003; Bartsch et al. 2002; Vasen et al. 2000). By contrast, familial pancreatic cancer without other familial cases of melanoma is not associated with CDKN2A mutations (Lal et al. 2000; Bartsch et al. 2002).
Mutations in the breast cancer susceptibility gene, BRCA2, are found in some familial pancreatic cancer kindreds (Murphy et al. 2002; Hahn et al. 2003), although several of the mutations were BRCA2:6174delT, and studies that did not include a high proportion of Ashkenazi Jews tended to have lower BRCA2 mutation frequencies.
Nevertheless, there is an excess of pancreatic cancers in breast cancer families carrying BRCA2 mutations, SIR
5.79 in individuals carrying the mutation, and there is also an increase in risk in carriers of mutations in BRCA1, SIR 4.11 (Mocci et al. 2013; Thompson and Easton 2002), although some recent data found a fourfold increased risk of pancreatic cancer in BRCA2 mutation carriers but not in BRCA1 mutation carriers (Moran et al. 2012). There also seems to be a small increased risk of pancreatic cancer in mutation-negative families with clustering of breast cancer (BRCAX), SIR 1.31 (Mocci et al. 2013).
Like BRCA2, PALB2 is a pancreatic cancer susceptibility gene; although the lifetime risk is not known, it is likely to be at least 5 % in carriers of deleterious PALB2 mutations (Axilbund and Wiley 2012). Notably, many of the families contain women affected by both breast and pancreatic cancer, and studies of both familial and sporadic pancreatic cancer have shown that these mutations are uncommon (reviewed in Tischkowitz and Xia 2010).
Thus, sequencing isolated pancreatic cancer cases to look for PALB2 mutations is likely to have a very low yield. There is some evidence, however, that such mutations could influence response to certain chemotherapeutic agents (Villarroel et al. 2011).
The gene responsible for ataxia telangiectasia, ATM, has also been identified as a pancreatic cancer susceptibility gene, although as for PALB2, it is responsible for less than 5 % of very strongly familial pancreatic cancer families and only a very small proportion of all cases (Roberts et al. 2012).
The Fanconi anemia genes have been implicated in pancreatic cancer, but the true contribution of these genes is uncertain; FANCA mutations do not seem to be important overall (Rogers et al. 2004), although two truncating mutations in FANCC, both associated with loss of heterozygosity in the Tumours, were seen in a series of 421 pancreatic cancer cases seen at the Mayo Clinic (Couch et al. 2005).
One family has been described with a specific form of hereditary pancreatic carcinoma, where there is clear architectural distortion within the pancreas, and a high incidence of dysplasia. In some cases, preventive pancreatectomy has been carried out, and incipient or early cancers were identified. The gene was linked to the telomeric regions of chromosome 4q (Eberle et al. 2002), and subsequently a missense mutation leading to a conserved proline to serine change at amino acid position 239 in the gene encoding palladin was identified in this single family (Pogue-Geile et al. 2006). This finding has not been universally accepted by other researchers in the field, however (Salaria et al. 2007; Slater et al. 2007; Zogopoulos et al. 2007; Klein et al. 2009).
To date, this has been the only family described with this gene mutation, and an interesting feature is precancerous dysplasia and prominent stromal fibrosis and so, early presentations include diabetes mellitus. Pancreatic cancer is reported to occur in familial adenomatous polyposis, although the Tumours usually actually arise from the ampulla of Vater.
There is an increased risk of adenocarcinoma of the pancreas in hereditary pancreatitis, a well-described but uncommon autosomal dominant condition associated with attacks from childhood of recurrent pancreatitis in affected family members (Kattwinkel et al. 1973). The condition has been mapped to chromosome 7q, and mutations in the cationic trypsinogen gene PRSS1 have been detected in affected individuals (Whitcomb et al. 1996).
One survey showed that among 112 families in 14 European countries (418 affected individuals), 58 (52 %) families carried a p. R122H mutation, 24 (21 %) had the p. N29I mutation, and 5 (4 %) had the p. A16V mutation. Other mutations were very rare, but 19 % of all families had no identified mutations in PRSS1 (Howes et al. 2004). Other genetic causes of chronic pancreatitis, such as CFTR (Sharer et al. 1998; Pezzilli et al. 2003) or SPINK1 (Witt et al. 2001) mutations, do not appear to frequent in individuals with pancreatic cancer (Malats et al. 2001; Pezzilli et al. 2003; Teich et al. 2003), but there could be instances where it would be worthwhile to look for specific mutations, such as p. EF508 in CFTR and p. N34S in SPINK1 in individuals with pancreatic cancer in the context of chronic pancreatitis (Flanders and Foulkes 1996).
Pancreatic cancer has been described in a case of Williams syndrome, possibly secondary to hypercalcemia (Jensen et al. 1976). Pancreatoblastoma has been described in the BWS (Koh et al. 1986) and rarely, in familial adenomatous polyposis (Abraham et al. 2001). Overexpression of IGF2 may underlie these cancers in both conditions (Kerr et al. 2002).
Screening for pancreatic cancer is extremely difficult because minor abnormalities detected by screening may be difficult to interpret as there is no clear premalignant lesion, and there are no well-recognized protocols. However, abdominal and endoscopic ultrasonography and endoscopic retrograde cholangiopancreatography (ERCP) have been suggested as possible screening methods, although the latter is probably too invasive except for the most high-risk families, where structural abnormalities may be identified (Brentnall et al. 1999); serum CA 19.9 does not appear to be suitably sensitive, but islet amyloid polypeptide may detect early pancreatic cancer. MRI cholangiopancreatography and endoscopic ultrasound are likely to be the most acceptable choices, and a recent meeting of experts suggests that either or both are indicated for high-risk individuals (Canto et al. 2013).
Interestingly, this expert group limited those suitable for screening to first- degree relatives (FDRs) of patients with pancreatic cancer from a familial pancreatic cancer kindred with at least two affected FDRs; patients with Peutz–Jeghers syndrome; and CDKN2A, BRCA2, and Lynch syndrome mutation carriers with ≥1 affected FDR (Verna et al. 2010; Giardiello and Trimbath 2006; de Vos tot Nederveen Cappel 2011; Brand et al. 2007; Canto et al. 2011; Harinck et al. 2010).
Somatic KRAS mutations are common in pancreatic carcinomas, and their detection in pancreatic juice may also be a helpful diagnostic marker; a combination of such indicators may be developed for use in screening for pancreatic cancer in time (Lynch et al. 1996; Urrita and DiMagno 1996), but as yet no clinically useful tests of this type have emerged. Next-generation sequencing has identified recurrent mutations in GNAS in pancreatic cysts that arise in intraductal papillary mucinous neoplasm (IPMN). These mutations may provide a helpful clue when deciding which pancreatic cysts to excise, as the same mutations were found in the associated invasive lesions (Wu et al. 2011).