Hereditary pancreatic cancer: related syndromes and clinical perspective

From a histological point of view [1], almost 95% of pancreatic cancers are pancreatic ductal adenocarcinomas (PDAC). PDAC has the worst prognosis among the major cancers and it constitutes the fourth leading cause of cancer death in the developed world. Such is its relevance that PDAC will be the third cause of cancer mortality in European Union in 2017, with 43,800 expected deaths among women and 43.600 among men [2]. In the United States the prospects also are not looking very hopeful, with 53,670 new cases and 43,090 deaths due to pancreatic cancer predicted this year [3]. The average lifetime risk of pancreatic cancer for both men and women is about 1 in 65 (1.5%) and the five year survival rate is 7% [4], which is related to advanced stage at diagnosis in the majority of the cases.

Pancreatic neuroendocrine tumors (PNETs) are infrequent neoplasms that represent approximately 1-2% of all pancreatic cancers [5]. Estimated incidence for PNETs is less than 1/100,000 per year although their relative indolent nature could underestimate these numbers [5, 6]. Five year survival rate in PNETs is about 42%, which is in concordance with predominant diagnosis of low-intermediate grade tumors [7].

Older age constitutes one of the main risk factors in the development of PDAC. Median age of onset of PDAC is 71 years [8]. Tobacco is the most important recognized toxic risk factor, doubling the risk [9]. Alcohol abuse [10], chronic pancreatitis [11], dietary factors, obesity, exposure to different agents and Diabetes Mellitus type 2 (DM2) also increase risk for PDAC [12]. Apart from these factors, family history also can influence in the probability of PDAC development [13]. Different reports that have evaluated incidence of pancreatic cancer in relatives have found that first degree relatives have almost a two-fold increased risk of developing PDAC and also that this risk seems to be proportional to the number of first-degree relatives with PDAC [14, 15]. It is estimated that a hereditary component may be implicated in nearly 10% of all PDAC cases [16, 17], but currently in less than 20% of them a defined hereditary cancer predisposition syndrome with increased risk of PDAC development can be identified.

Conversely, in PNETs, there are no well recognized modifiable risk factors [18]. The majority of them are considered sporadic tumors and in about 20% of the cases a cancer hereditary syndrome can be recognized [19].

Main hereditary cancer syndromes related to an increased risk in PNETs development are multiple endocrine neoplasia type 1 (MEN1), Von-Hippel-Lindau disease (VHL) and neurofibromatosis type 1 (NF-1). Although less frequently, a relation between PNETs and Tuberous sclerosis complex (TSC) has been suggested. Predominantly PNETs in hereditary syndromes are grade 1 (PNET G1, Ki-67 index <3%) and grade 2 (PNET G2, Ki-67 index 3-20%) tumors [20].

MEN1, which is also referred as Wermer syndrome [21], is clinically characterized by the classical triad of tumors of the parathyroid glands, the pancreatic islet cells, and the anterior pituitary and it is inherited in an autosomal dominant manner with high penetrance. MEN1 gene is on chromosome 11q13 and encodes menin protein [22]. Parathyroid tumors, resulting in primary hyperparathyroidism are the most common feature of MEN1 and occur in approximately 95% of patients [23]. PNETs occur in 20-75% and anterior pituitary tumors occur in 30% of patients [24]. Multifocality is one of the main features of PNETs in this syndrome [25]. Most of PNETs are non-functioning tumors [26], and insulinomas are the second group in frequency order. More rarely, glucagonomas and vipomas have been described among MEN1 related PNETs [27].

VHL disease is an autosomal dominant condition secondary to mutations in VHL gene, and it is typically associated with pheochromocytomas, renal cell carcinomas, central nervous system hemangioblastomas, endolymphatyc sac tumors and cystic pancreatic lesions [28]. Most of these pancreatic cysts are considered benign in a sense that they do not alter pancreatic function or they do not have an increased risk of malignancy towards PDAC. Presence of PNETs has been described in around 5-17% of patients with VHL disease, and they are almost always single non-functioning tumors [29].

NF type 1 is a disorder inherited in an autosomal dominant manner and it is caused by mutations in NF1 gene [30]. Neurofibromas, café-au-lait spots, Lisch nodules and freckles in the underarms are typical features of this syndrome, and they are included in well defined clinical diagnosis criteria [31]. PNETs have been described in less than 1% of patients with NF type 1, with somatostatin and insulin secreting tumors being the most commonly associated lesions [32, 33]. Data about increased risk of PDAC and NF type 1 are scarce and inconclusive, with isolated cases reported in the literature [34].

TSC is a rare entity inherited in an autosomal dominant manner, characterized mainly by multiple hamartomatous lesions, epilepsy and intellectual disability, and it is produced by mutations in TSC1 and TSC2 genes [35]. Although there is scarcity of data about TSC and increased risk of PNETs, insulinomas and non-functioning tumors have been reported in patients with TSC [36].

The more remarkable hereditary cancer predisposition syndromes with increased risk of PDAC are: hereditary breast and ovarian cancer syndrome (HBOC), familial melanoma (FM), Lynch syndrome (LS), familial adenomatous polyposis (FAP), Peutz-Jeghers syndrome (PJS) and Li-Fraumeni syndrome (LFS) [37]. Also ATM gene has been defined as a potential predisposing factor of PDAC in heterozygous carriers [38]. They account for almost 10-15% of PDAC familial cases, defined by a minimum of two PDAC diagnoses among first degree relatives [39]. The remaining 85-90% of familial cases with pancreatic cancer aggregation lacks these hereditary cancer predisposition cancer syndromes [40] and they are defined as familial pancreatic cancer (FPC). Considering that in the majority of families with PDAC susceptibility a responsible gene mutation will not be identified, several multigene panel and/or whole genome sequencing studies have been designed [41]. Multiple gene panels for PDAC include among others, BRCA1, BRCA2, PALB2, CDKN2A, MLH1, MSH2, MSH6, PMS2, EPCAM, ATM, APC, STK11, PRSS1 and TP53 genes [42]. A PACGENE Consortium study [43] included samples from 727 unrelated probands with PDAC family history (521 met FPC criteria) who were tested for mutations in BRCA1, BRCA2, PALB2 and CDKN2A. They found that the prevalence of pathogenic variants among included probands was: BRCA1, 1.2%; BRCA2, 3.7%; PALB2, 0.6% and CDKN2A 2.5%. As a consequence of this approach, many families that meet classical definition of FPC are now being reclassified into specific hereditary cancer syndromes. Therefore, we could affirm that FPC is currently a diagnosis of exclusion, strictly reserved for those families with 2 or more first-degree relatives with PDAC in the absence of a recognizable syndrome or genetic disorder. Mean age of onset of pancreatic neoplasia in FPC seems to be slightly lower (64-65 years) than sporadic pancreatic cancer cases [44].

Metastatic PDAC patients are usually treated with chemotherapy [102]. Current options in patients with good performance status are FOLFIRINOX (a platinum containing regimen) or gemcitabine/nab-paclitaxel in first line setting [103, 104] and treatment decision depends on patients’ comorbidities or expected toxicities profiles. Tumors harboring somatic or germline pathogenic variants in genes related to DNA double strand damage repair, such as BRCA1, BRCA2, PALB2 or ATM, have been associated to better responses to platinum-based chemotherapy schedules [105]. Platinum compounds generate double DNA strand breaks that cannot be repaired when homologous recombination related genes are affected [106]. This benefit in platinum based schedules has also been reported in patients with PDACs related to HBOC syndrome and constitutes the basis for tailored therapy clinical trials [107‐109].

Poly-ADP ribose polymerases (PARP) are involved in single DNA strand break repair. Those tumors harboring homologous recombination genes mutations can draw upon this salvage pathway in order to repair DNA damage [110]. Therefore, inhibition of PARP mediated pathway could lead to tumor cell destruction (synthetic lethality concept) in the presence of a pathogenic BRCA variant. Olaparib was the first PARP inhibitor approved as maintenance therapy in advanced high grade serous ovarian carcinomas that have platinum sensitive recurrences [111]. Olaparib and its copartners (veliparib, rucaparib) alone or in combination with platinum based chemotherapy have shown high activity in BRCA/PALB2 mutated pancreatic cancers and they are object of study in current open randomized clinical trials [108, 112‐114].

It is hypothesized that tumors with high genomic instability may benefit more from immunotherapy checkpoint inhibitors, especially from the program death 1/program death-ligand 1 (PD1/PD-L1) axis agents (nivolumab, pembrolizumab) [115]. Tumors harboring high genomic instability are related to a higher mutational load which potentially can increase the number of neoepitopes that are exposed and virtually generate a specific anticancer immune response [116]. Genomic instability is one of the main characteristic of Lynch syndrome related tumors and it has also been correlated with tumors with somatic alterations in BRCA and PALB2 genes [117]. It has to be emphasized that a phase II study of pembrolizumab in patients with colorectal cancer showed no tumor responses in the group with mismatch repair proficient tumors, and an impressive 50% of objective responses in patients with mismatch repair deficient tumors [118]. Also, PD-1 inhibition treatment approximation has shown promising results in a phase II study including patients with different gastrointestinal cancers: those with mismatch repair deficient non colorectal tumors had an immune-related objective response rate of 71%. This fact should be taken into consideration when designing clinical trials of immunotherapy in pancreatic cancer and also in treatment decision in patients with metastatic PDACs related to hereditary cancer syndromes [119]. Figure 2 summarizes these features of PDAC in the context of hereditary cancer syndromes and their potential implication in targeted therapies development.

The majority of neuroendocrine tumors have somatic mutations in MEN1, ATRX, DAXX and/or in genes involved in phosphoinositide 3-kinase, AKT, and mammalian target of rapamycin (PI3K/AKT/mTOR) pathway [120, 121]. The presence of specific somatic or germline mutations in PNETs and their correlation with better responses to different multitargeted inhibitors is object of research. This way is being explored in a prospective phase II trial [122], which is recruiting patients diagnosed of low-intermediate grade neuroendocrine tumors; patients with germline/somatic MEN1 mutations are assigned to sunitinib (multityrosine kinase inhibitor) treatment and those with germline/somatic NF1/VHL/TSC mutations are treated with everolimus (mTOR pathway inhibitor).

PDAC is the more frequent histological subtype of pancreatic cancer. Even though the majority of pancreatic cancer cases are considered as sporadic, it is estimated that about 10% of them have a familial component. FPC is defined as a family with 2 or more first-degree relatives with pancreatic cancer. The majority of families with multiple cases of pancreatic cancer do not have an identifiable causative gene or syndrome and strictly they meet FPC definition. A small subset of hereditary pancreatic cancer (20%) is attributable to known inherited cancer predisposition syndromes. PNETs are the second type of pancreatic cancer in terms of incidence and about 20% of them have an involved inherited condition. Recognition of germline mutations in patients with pancreatic tumors does not only suppose an impact in genetic counseling process, since it also may affect treatment decisions and predict response to specific therapies.