Lessons From McCune-Albright Syndrome–Associated Intraductal Papillary Mucinous Neoplasms: : GNAS-Activating Mutations in Pancreatic Carcinogenesis

GNAS-activating mutations are reported in intraductal papillary mucinous neoplasms (IPMNs) and in McCune-Albright syndrome, characterized by fibrous dysplasia, precocious puberty, and café au lait spots. Recently, IPMNs have been described as a McCune-Albright syndrome–associated tumor, present in about 15% of patients. The aim of the present work was to assess the prevalence of polyostotic fibrous dysplasia and McCune-Albright syndrome among patients operated on for presumptive sporadic IPMNs. All patients operated on for IPMNs between January 1, 2007, and December 31, 2012, with available imaging were retrospectively screened for polyostotic fibrous dysplasia based on their preoperative abdominal or thoracoabdominal spiral computed tomography images. Systematic screening of 272 patients operated on for IPMNs revealed 1 patient with axial and peripheral polyostotic fibrous dysplasia and café au lait spots on clinical examination suggestive of McCune-Albright syndrome. This patient had been operated on for an unusually large invasive colloid adenocarcinoma (pT3N0M0 R0) derived from an intestinal subtype GNAS-mutated IPMN. The patient underwent adjuvant chemotherapy with gemcitabine for 6 months and was alive without recurrence 6 years later. Besides providing additional evidence of a syndromic IPMN as a feature of McCune-Albright syndrome, this observation is further evidence of the functional oncogenic consequences of GNAS mutations in the pancreas.

Somatic activating mutations of the stimulatory G protein α subunit encoded by the GNAS gene (Online Mendelian Inheritance in Man [OMIM] 139320) have recently been reported in up to 70% of pancreatic intraductal papillary mucinous neoplasms (IPMNs).^1-4 Along with KRAS (OMIM 190070), it is one of the 2 most prevalent mutations among these tumors.

Based on screening of patients with McCune-Albright syndrome (McAS), IPMNs and other hepatobiliary neoplasms have recently been described as McAS–associated tumors, present in about 15% of patients.^5,6 McCune-Albright syndrome is a rare disorder characterized by polycystic fibrous dysplasia, precocious puberty, and café au lait spots caused by the same GNAS dominant–activating mutations,⁷ which are somatic and postzygotic, with mosaic distribution.

The aim of this work was to assess the prevalence of polyostotic fibrous dysplasia and McAS among patients operated on for presumptive sporadic IPMNs. The functional and clinical consequences associated with this specific phenotype were also evaluated.

Permission from the Comité d’Evaluation de l’Ethique des projets de Recherche Biomédicale (CEERB) Paris Nord Institutional Review Board and written informed consent from the patients were obtained before data review and analysis. A database at the Department of Pathology, Hôpital Beaujon, Clichy, France, was screened for surgical specimens of IPMNs obtained between January 1, 2007, and December 31, 2012. Demographic variables, the radiological workup, and pathology data were obtained from a prospective database, with additional retrospective medical record review.

For each patient, the availability of preoperative abdominal or thoracoabdominal spiral computed tomography images was assessed. Patients with available imaging were screened for polyostotic fibrous dysplasia using bone window (160-500 Hounsfield units [Hu]) and scout view acquisition of their preoperative abdominal or thoracoabdominal spiral computed tomography images. When imaging features of fibrous dysplasia were observed, patients were asked to attend a medical consultation to review their complete medical history, undergo a comprehensive clinical examination, and be screened for McAS. Screening included complete skin examination, thyroid ultrasonography, full-body digital radiography (EOS; EOS Imaging), pituitary magnetic resonance imaging, and pituitary function tests.

For selected IPMN specimens, the type of duct involvement was determined by macroscopic and microscopic examinations and was classified as main duct, branch duct, or mixed-type IPMNs. Intraductal papillary mucinous neoplasms were graded as mild, moderate, or high-grade dysplasia (carcinoma in situ) or as invasive carcinoma according to World Health Organization criteria.⁸ The intraductal components were classified as 4 distinct epithelial subtypes of gastric, intestinal, pancreatobiliary, or oncocytic based on their epithelial morphology on routine hematoxylin-eosin-safran staining and their mucin profile on immunochemistry (MUC1, MUC2, and MUC5AC [1/400]; BD Biosciences). In addition, the human equilibrative nucleoside transporter 1 (hENT1) immunohistochemical status was assessed (clone SP120 [1/400]; Spring Biosciences).

DNA was extracted from formalin-fixed, paraffin-embedded tumor specimens using a tissue kit (QIAamp; Qiagen) according to the manufacturer’s instructions. GNAS and KRAS status (codons 201 and 12, 13, and 61, respectively) were assessed by electropherogram analysis (SNaPshot Multiplex System; Applied Biosystems). Primers used are listed in the Table. Detailed experimental settings are available on request from the author.

Among 287 patients operated on for IPMNs between January 1, 2007, and December 31, 2012, in the Department of Hepato-Pancreato-Biliary Surgery, Hôpital Beaujon, Clichy, France, 272 patients with available imaging were retrospectively screened for polyostotic fibrous dysplasia based on their preoperative abdominal or thoracoabdominal spiral computed tomography images using bone window (160-500 Hu) and scout view acquisition. Among 272 patients operated on for IPMNs with spiral computed tomography images available, 1 patient (0.4%) demonstrated imaging features of polyostotic fibrous dysplasia, including axial (ie, pelvis, spine, sternum, rib, and skull) and peripheral (ie, femur and humerus) involvement (Figure 1A). This 62-year-old man had initially been seen with abdominal pain related to a bulky right flank mass. He had lost 14 kg during the past 18 months. Computed tomography (Figure 1B), magnetic resonance imaging, and duodenoscopy revealed a heterogeneous 11-cm duodenal infiltrating pancreatic mass, with biliary and pancreatic duct dilatation and without distant metastasis or vascular involvement. After pancreatoduodenectomy, pathological examination (Figure 2A) revealed an invasive pancreatic colloid adenocarcinoma (pT3N0M0 R0) arising from an intestinal subtype IPMN (ie, MUC1 negative, MUC2 positive, and MUC5AC positive by immunohistochemistry) (Figure 2B) that was hENT1 negative. Genetic analysis of the IPMN revealed a GNAS-activating mutation (c.602G>A [p.R201H]) absent in the adjacent normal pancreas (Figure 3) but present in peripheral blood leucocytes. No KRAS mutation was present in the IPMN specimen. The patient underwent adjuvant chemotherapy with gemcitabine for 6 months and was alive without recurrence 6 years later. The patient’s medical history revealed peripheral and central polyostotic fibrous dysplasia, responsible for multiple fractures and deformation. Clinical examination showed a large café au lait spot (Figure 4) on the right side of the back at the level of the abdominal scar, intramuscular myxomas, and a multinodular thyroid goiter confirmed by ultrasonography. No pituitary adenoma was visible on magnetic resonance imaging. Taken together, these results suggested that the patient at the initial examination had manifested a syndromic IPMN associated with McAS.

Recently, somatic activating GNAS mutations have been described in up to 70% of IPMNs,^1-4,9 representing one of the most frequent genetic events together with KRAS mutations. To date, the clinical and functional consequences of these mutations have remained largely unknown. The mutations have been described in other hepatobiliary neoplasms such as intraductal papillary neoplasms of the bile duct,¹⁰ cholangiocarcinoma,¹¹ and hepatocellular adenoma and carcinoma,⁵ although much less frequently. Hepatopancreatobiliary neoplasms, including hepatocellular adenoma, choledochal cyst, and IPMNs, have recently been reported in about 30% of patients with McAS.⁶ None of these lesions were previously described as invasive or responsible for cancer-related mortality. Herein, we describe the first reported invasive IPMN to date with a proven GNAS mutation arising in a patient with McAS, and this observation stresses 2 important points. First, the unusually large volume of the tumor, although weakly aggressive, suggests a slow-growing neoplasm that would be accessible to screening and prophylactic surgery. This observation may indicate the need for systematic magnetic resonance imaging screening of hepatopancreatobiliary neoplasms in patients with McAS to propose, when possible, surgery before the occurrence of invasive pancreatic cancer, although this remains to be formally demonstrated. Second, this GNAS-mutated IPMN displayed an intestinal phenotype with a colloid invasive component. Convincing observations^1,9 associate GNAS mutations with an intestinal phenotype, although this association remains unexplained. A colloid invasive pattern is present in about 25% of IPMN-derived cancers and arises from intestinal-type IPMNs.¹² In addition, colloid IPMN-derived adenocarcinomas are associated with improved long-term survival compared with tubular IPMN-derived adenocarcinomas or pancreatic ductal adenocarcinomas.^13,14 Consequently, whether the presence of GNAS mutations or the absence of KRAS mutations could be responsible for this colloid pattern of invasive cancer remains to be studied in more depth. Indeed, little is known about the functional consequences of GNAS-activating mutations, especially in the pancreas; however, in a colon cancer cell line model, stable transfection of GNAS-mutated plasmid was associated with a phenotype differentiation similar to that of intestinal-type IPMNs (ie, mucin production¹⁵), without promoting cell growth in vitro or in vivo. Further studies should focus on the prognostic consequences of the GNAS intestinal-associated phenotype, comparing the outcome of these patients with that of patients having tubular GNAS-mutated tumors and GNAS wild-type tumors. Additional evidence will come from in vitro or in vivo experiments studying the functional consequences of GNAS overexpression in pancreas cell lines or animal models.

Overall, this work indicates that IPMNs are an McAS-associated tumor and suggests that GNAS is a potential pancreatic oncogene, although this needs to be formally demonstrated on larger scale and in vitro or in vivo studies. Benign-looking bone or skin lesions in patients seen with IPMNs should not be overlooked and must be further investigated because they may reveal McAS.

Accepted for Publication: January 31, 2014.

Corresponding Author: Sebastien Gaujoux, MD, PhD, Assistance Publique–Hôpitaux de Paris, Department of Hepato-Pancreato-Biliary Surgery, Hôpital Beaujon, 100 Blvd du Général Leclerc, 92118 Clichy CEDEX, France (sebastien.gaujoux@gmail.com).

Published Online: June 4, 2014. doi:10.1001/jamasurg.2014.535.

Author Contributions: Dr Gaujoux had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Parvanescu, Cros, Ronot, Ruszniewski, Gaujoux.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Parvanescu, Cros, Ronot, Grybek, Gaujoux.

Critical revision of the manuscript for important intellectual content: Cros, Ronot, Hentic, Couvelard, Levy, Chanson, Ruszniewski, Sauvanet, Gaujoux.

Statistical analysis: Parvanescu, Ronot, Gaujoux.

Obtained funding: Gaujoux.

Administrative, technical, or material support: Couvelard, Sauvanet, Gaujoux.

Study supervision: Cros, Hentic, Levy, Chanson, Ruszniewski, Sauvanet, Gaujoux.

Conflict of Interest Disclosures: None reported.

Funding/Support: Ms Parvanescu reported receiving a grant from the Société Française de Chirurgie Digestive. This work was supported in part by the Fonds d’Aide à la Recherche et à l’Evaluation from the Société Nationale Française de Gastroentérologie.

Role of the Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: Agnès Linglart, MD, PhD, provided helpful discussion and technical assistance. We thank the staff of the Department of Pathology at Hôpital Beaujon and the Department of Oncogenetics at Hôpital Bichat.