Pathology, genetics and precursors of human and experimental pancreatic neoplasms: An update

Abstract

Over the past decade, there have been substantial improvements in our knowledge of pancreatic neoplasms and their precursor lesions. Extensive genetic analyses, recently using high-throughput molecular techniques and next-generation sequencing methodologies, and the development of sophisticated genetically engineered mouse models closely recapitulating human disease, have improved our understanding of the genetic basis of pancreatic neoplasms. These advances are paving the way for refined, molecular-based classifications of pancreatic neoplasms with the potential to better predict prognosis and, possibly, response to therapy. Another major development resides in the identification of subsets of pancreatic exocrine and endocrine neoplasms which occur in the context of hereditary syndromes and whose genetic basis and tumor development have been at least partially defined. However, despite all molecular progress, correct and careful morphological characterization of tissue specimens both in the context of experimental and routine diagnostic pathology represents the basis for any further genetic investigation or clinical decision. This review focuses on the current and new concepts of classification and on the current models of tumor development, both in the field of exocrine and endocrine neoplasms, and underscores the importance of applying standardized terminology to allow adequate data interpretation and promote scientific exchange in the field of pancreas research.

Introduction

The pancreas is an organ with dual function exerted by an exocrine and endocrine cell compartment. Thus, the neoplasms arising therefrom display different phenotypes in terms of morphology and biology, and are driven by distinct genetic alterations. In order to improve the prognosis and treatment options, especially for those tumors with unfavorable outcome, it is essential to identify precursor lesions and their driving molecular alterations. With this knowledge we will improve our understanding of the tumors' natural history, we may then better select the patients who profit from existing therapeutic approaches and might finally define new therapeutic targets.

The development of sophisticated genetically engineered mouse models that closely recapitulate human disease, together with the extensive use of high-throughput molecular techniques and next-generation sequencing methodologies have improved our understanding of the genetic basis of pancreatic neoplasms and may prepare our way for clinic-translational approaches. The bottom line for these approaches is the consistent use of established, internationally accepted classification systems and a unanimous nomenclature both in research and routine pathology.

The 2010 World Health Organization classification [1] of pancreatic tumors is widely accepted. It separates the tumors in epithelial (including exocrine and neuroendocrine tumors) and non-epithelial neoplasms and subtypes them according to their biological behavior in benign, premalignant and malignant (Table 1).

Human neoplasms can be further classified in sporadic, familial and hereditary. A neoplasm occurring in individuals who do not have a germline mutation that confers increased susceptibility to cancer or without a family history of cancer is defined as sporadic. Cancer that occurs in families more often than would be expected by chance and usually at an earlier age compared to the general population is defined as familial. Shared environmental or lifestyle factors play a role in familial cancer. If a germline mutation predisposes to cancer development, the term hereditary neoplasm is used (for definitions see also www.cancer.gov/dictionary).

Among the pancreatic neoplasms the sporadic epithelial non-neuroendocrine tumors are most frequent and account for about 90%. They display either an acinar or ductal cell phenotype. In addition, there are some rare tumors whose phenotypic appearance is not attributable to one of the known cell types. Genetically, the neoplasms with a ductal phenotype usually harbor KRAS mutations (“KRAS-positive”), while those with other phenotypes such as the acinar cell carcinomas, neuroendocrine neoplasms, serous cystic neoplasms and solid pseudopapillary neoplasms have a different genetic profile. In this review, the pancreatic neoplasms and their precursors are classified according to their KRAS-status and their cellular phenotype (see Table 2). In addition, their counterparts in genetically engineered mice are briefly described for comparison. Finally, we present a conceptual approach to the development of pancreatic neoplasms.