An Algorithmic Approach to the Diagnosis of Pancreatic Neoplasms by ProQuest


More Info
									              An Algorithmic Approach to the Diagnosis of
                         Pancreatic Neoplasms
                                David S. Klimstra, MD; Martha B. Pitman, MD; Ralph H. Hruban, MD

● Context.—The pancreas gives rise to an array of distinct                 Conclusions.—By using a series of dichotomous deci-
neoplasms that can be solid, cystic, or intraductal and can             sions, the differential diagnosis of a pancreatic neoplasm
recapitulate the various lines of differentiation present in            can be narrowed, and in cases of the more common neo-
the normal gland.                                                       plasms, accurate classification can be achieved. Uncom-
  Objective.—To develop an algorithmic approach to the                  mon neoplasms not accounted for by this approach are
diagnosis of pancreatic neoplasms that simplifies their                  also discussed, and the additional diagnostic information
pathologic evaluation.                                                  needed for complete pathologic reporting is presented.
  Data Sources.—We reviewed literature related to the                      (Arch Pathol Lab Med. 2009;133:454–464)
classification of pancreatic neoplasms on the basis of their
gross, histologic, and immunohistochemical features.

T   o help classify pancreatic neoplasms, we have devel-
     oped a simple diagnostic algorithm that integrates the
gross and radiographic features with the histologic and
                                                                        pancreatic neoplasms are infiltrating ductal adenocarci-
                                                                        nomas or closely related variants thereof (see below).2 The
                                                                        other more commonly encountered entities each make up
immunohistochemical findings. By following the steps de-                 no more than 1% to 5% of pancreatic neoplasms, but since
tailed in this review, the pathologist can accurately classify          the latter entities include the most treatable (or even cur-
most pancreatic neoplasms, although some exceptions do                  able) neoplasms to arise in the pancreas, their proper rec-
occur, either because the rarity of a neoplasm would ren-               ognition and separation from ductal adenocarcinoma is
der the algorithm overly complex if it were included, or                very important.
because some common neoplasms rarely demonstrate un-                       The classification of pancreatic neoplasms is based in
usual features. Nonetheless, this conceptual approach can               large part on their gross configuration (solid, cystic, intra-
help delineate the interrelationships among entities and it             ductal) and the lines of cellular differentiation they exhibit
can advance our understanding of the fundamental fea-                   (ductal, acinar, endocrine).3 Based upon the histologic fea-
tures that define each entity.                                           tures (especially in the case of ductal differentiation), the
                                                                        differentiation present in a pancreatic neoplasm may be
                THE BASIS FOR CLASSIFYING                               evident, but in some circumstances additional studies are
                 PANCREATIC NEOPLASMS                                   needed (Table 3).
  Pancreatic epithelial neoplasms exhibit a wide range of                                Direction of Differentiation
pathologic features. The current classification of pancreatic
                                                                           The most common direction of differentiation observed
epithelial neoplasms is listed in Table 1.1 Of course, some
                                                                        in pancreatic neoplasms is toward a ductal phenotype.
of the entities included are very uncommon or are regard-
                                                                        Ductal differentiation is defined as the recapitulation of the
ed as variants of other entities from which they do not
                                                                        characteristics of normal ducts, that is, gland or tubule
differ significantly in terms of treatment or prognosis.
                                                                        formation and mucin production.3 Mucin can be demon-
Thus, this extensive classification schema can be con-
To top