Article Text

other Versions


Genetic unrelatedness of co-occurring pancreatic adenocarcinomas and IPMNs challenges current views of clinical management
  1. Aldo Scarpa1,
  2. Francisco X Real2,3,
  3. Claudio Luchini1
  1. 1 Department of Diagnostics and Public Health-Section of Pathology, ARC-Net Research Center, University and Hospital Trust of Verona, Verona, Italy
  2. 2 Epithelial Carcinogenesis Group, Spanish National Cancer Research Centre-CNIO, Madrid, Spain
  3. 3 Departament de Ciències Experimentals i de la Salut, Universitat Pompeu Fabra, Barcelona, Spain
  1. Correspondence to Professor Aldo Scarpa, ARC-Net Research Centre and Department of Diagnostics and Public Health-Section of Pathology, University of Verona, Verona I-37134, italy; aldo.scarpa{at}

Statistics from

Intraductal papillary mucinous neoplasms (IPMNs) represent the most common cystic preinvasive lesion in the pancreas, which can progress to invasive pancreatic ductal adenocarcinoma (PDAC). Patients with radiological findings consistent with IPMN are strictly monitored with imaging techniques, and surgical resection is recommended in case of clinical and/or radiological ‘high-risk stigmata’ or ‘worrisome features’.1 2 Recent research has suggested the potential role of molecular analysis of cyst fluid to discriminate IPMNs from other types of pancreatic cysts and from invasive lesions.3–6

Notably, recent evidences pointed out the presence of a so-called ‘field effect’ of pancreatic carcinogenesis. This concept indicates the occurrence of seemingly unrelated patches of non-neoplastic cells that may be partly shared with a tumour lesion.7 Therefore, within the pancreas not only a preinvasive lesion is at risk of progression to malignancy but independent preinvasive and invasive lesions may develop as well. In other words, the ‘field effect’ might separately predispose to both preinvasive and invasive lesions. Along this line, Pea and colleagues, sequenced DNA from a cohort of IPMNs and PDAC from the same patients, showing that a significant proportion of the recurrent cancers was likely independent from the primary resected IPMNs.8

The existence of a field-effect carcinogenesis in the pancreas has also been suggested by the work of Shindo and colleagues, who reported that germ line mutations in PDAC susceptibility genes are commonly identified in patients with pancreatic cancer, even when a family history of cancer is lacking.9 Some of these mutations are therapeutically targetable, and they may be missed if current family history guidelines are the criterion used to determine the appropriateness of gene testing.9 Similarly, a field effect may also result from lifestyle exposures relevant to PDAC (ie, tobacco and alcohol), as has been demonstrated in oral cancers.10–12

The manuscript of Felsenstein and colleagues, published in this issue of Gut, further corroborates this concept. With next-generation sequencing, they analysed a large cohort of IPMNs and co-occurring PDACs, using a specific panel of cancer driver genes.13 The main finding was that 18% of co-occurring IPMNs and PDACs were likely independent (see figure 1A); this result supports the concept of a ‘field effect’ of pancreatic carcinogenesis. Two additional findings presented in this manuscript merit consideration: (1) there is a striking genetic heterogeneity in IPMNs, even with respect to well-characterised driver genes and (2) all colloid carcinomas (CCs) were likely related to their associated IPMNs (see figure 1A).

Figure 1

(A) Co-occurring intraductal papillary mucinous neoplasm (IPMN) and invasive pancreatic ductal adenocarcinoma (PDAC). In 18% of cases, co-occurring IPMN and PDAC are genetically independent despite being very close (upper part of the figure). If the invasive adenocarcinoma has features of colloid carcinoma (CC), it is more likely to be genetically related to the co-occurring IPMN (lower part of the figure). (B) Flow chart of clinical management of patients with IPMN (upper part of the figure) and implications derived from recent evidences of possible genetic unrelatedness of co-occurring PDAC and IPMN (lower part of the figure). High-risk stigmata: (1) obstructive jaundice in a patient with cystic lesion of the head of the pancreas, (2) enhancing mural nodule >5 mm and (3) main pancreatic duct >10 mm; worrisome features: (1) Clinical (pancreatitis); (2) imaging (cyst >3 cm, enhancing mural nodule <5 mm, thickened/enhancing cyst walls, main duct size 5–9 mm, abrupt change in calibre of pancreatic duct with distal pancreatic atrophy, lymphadenopathy, increased serum level of CA19-9, and cyst growth rate >5 mm/2 years).1 EUS, endoscopic ultrasound.

Previous sequencing papers indicated that KRAS and GNAS are the most commonly mutated genes in IPMNs.14–17 However, in this preinvasive tumour category, there are also important differences in the genetic background and progression. Indeed, recent data support the existence of two distinct molecular progression pathways in IPMN.18 Pancreatobiliary-type IPMNs are highly associated with PDAC, sharing its typical mutation profile,19 while intestinal-type IPMNs and associated CCs typically harbour GNAS mutations, as intestinal villous adenomas.20 The study of the Johns Hopkins group adds a significant part to this knowledge. On one hand, it confirms the intimate correlations between IPMN and CC, on the other hand shows that IPMN and PDAC, even if concurrent, may be ‘neighbours but not always relatives’, as also reported in the title with a fashion manner.13

Recent work using genetically engineered mice has highlighted the possible relevance of the cell-of-origin. While IPMNs are recapitulated through the introduction of genetic alterations in pancreatic ductal cells in two different models,20–22 pancreatic intraepithelial neoplasms appear to arise predominantly through the transdifferentiation of acinar cells.23–25 These findings may help to understand the ‘neighbours but not always relative’ conundrum.

One of the main implications of the work of Felsenstein and colleagues, which could impact on the management of patients with IPMN, is that the molecular analysis of cyst fluid may not always be representative of the real biological risk of malignant transformation in a pancreas with IPMN. Indeed, there is a significant proportion of patients in whom this analysis may lead to underestimation of the risk of PDAC development. In fact, the close to 20% proportion of PDAC developed independently from a coexisting IPMN represents a challenging scenario due to the potential clinical impact of missing an early PDAC. On independent replication, these findings should be considered in the follow-up of patients with IPMN, where molecular analysis of cyst fluid might be considered only as part of the clinical workup (see figure 1B). Cytological or microbiopsy approaches from different pancreatic areas, the integration of molecular analysis of cyst fluid with circulating tumour cells, the encouraging application of liquid biopsy26 and a strict radiological follow-up represent the most promising tools for a comprehensive risk stratification of patients with IPMNs in the near future. Preventive and early intervention strategies should also consider the risk of co-occurrence of both neoplasms, IPMN and PDAC, especially considering that current evidence points to the existence of shared epidemiological risk factors.27

Another important aspect of IPMNs and PDACs regards tumour heterogeneity. We already know that PDAC harbour extensive intertumour and intratumour heterogeneity. But it is also of interest the reported intratumour heterogeneity of the IPMNs analysed by Felsenstein and colleagues. Comparing the alterations in two distinct regions of the same IPMN, one adjacent to—and the other distant from—the co-occurring PDAC, they show that PDACs were more likely related to the neighbouring IPMN area and more likely ‘independent’ from the distant areas. These findings highlight the existence of significant intra-IPMN heterogeneity and raise the possibility of polyclonality. This again raises concerns about the representativeness of tissue sampling strategies.

In conclusion, the finding that IPMN and PDAC, even if co-occurring in the same pancreas, are not necessarily genetically associated or represent the evolution of a same oncogenetic process has important biological and clinical implications. A high probability of phylogenetic relationship is present only in the case of co-occurring IPMN and colloid carcinoma. The general assumption of the relatedness of IPMN and PDAC must be abandoned, in order to avoid a potential underestimation of the risk of malignant transformation in patients with IPMN.


View Abstract


  • Contributors All authors contributed to the writing of the article.

  • Funding This study was funded by Associazione Italiana per la Ricerca sul Cancro, grant 5x1000 n.12182.

  • Competing interests None declared.

  • Patient consent Not required.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Pancreas
    Matthäus Felsenstein Michaël Noë David L Masica Waki Hosoda Peter Chianchiano Catherine G Fischer Gemma Lionheart Lodewijk A A Brosens Antonio Pea Jun Yu Georgios Gemenetzis Vincent P Groot Martin A Makary Jin He Matthew J Weiss John L Cameron Christopher L Wolfgang Ralph H Hruban Nicholas J Roberts Rachel Karchin Michael G Goggins Laura D Wood