Endocr Pathol (2014) 25:6579

DOI 10.1007/s12022-013-9295-2

Neuroendocrine Tumors of the Pancreas: Current Concepts

and Controversies
Michelle D. Reid & Serdar Balci & Burcu Saka &
N. Volkan Adsay

Published online: 16 January 2014

# Springer Science+Business Media New York 2014

Abstract In the past decade, the clinico-pathologic characteristics of neuroendocrine tumors (NETs) in the pancreas
have been further elucidated. Previously termed islet cell
tumors/carcinomas or endocrine neoplasms, they are now
called pancreatic NETs (PanNETs). They occur in relatively
younger patients and may arise anywhere in the pancreas.
Some are associated with von HippelLindau, MEN1, and
other syndromes. It is now widely recognized that, with the
exception of tumorlets (minute incipient neoplasms) that occur in some syndromes like MEN1, all PanNETs are malignant, albeit low-grade, and although they have a protracted
clinical course and overall 10-year survival of 6070 %, even
low-stage and low-grade examples may recur and/or metastasize on long-term follow-up. Per recent consensus guidelines
adopted by both European and North American NET Societies (ENETS and NANETs) and WHO-2010, PanNETs are
now graded and staged separately, unlike previous classification schemes that used a combination of grade, stage, and
adjunct prognosticators in an attempt to define benign behavior or malignant categories. For staging, the ENETs
proposal may be more applicable than CAP/AJCC, which is
based on the staging of exocrine tumors. Current grading of
PanNETs is based on mitotic activity and ki-67 index. Other
promising prognosticators such as necrosis, CK19, c-kit, and
others are still under investigation. It has also been recognized
that PanNETs have a rather wide morphologic repertoire
including oncocytic, pleomorphic, ductulo-insular, sclerosing,
and lipid-rich variants. Most PanNETs are diagnosed by fine
M. D. Reid : S. Balci : B. Saka : N. V. Adsay
Department of Pathology, Emory University School of Medicine,
Atlanta, GA, USA
N. V. Adsay (*)
Department of Pathology, Emory University Hospital, 1364 Clifton
Rd NE, Room H178, Atlanta, GA 30322, USA
e-mail: nadsay@emory.edu

needle aspiration biopsy, in which single, monotonous

plasmacytoid cells with fair amounts of cytoplasm and distinctive neuroendocrine chromatin are diagnostic. Molecular
alterations of PanNETs are also very different than that of
ductal or acinar tumors. Loss of expression of DAXX and
ATRX proteins has been recently identified in 45 %. Along
with these improvements, several controversies remain, including grading, value of current cutoff ranges, and the best
methods for counting ki-67 index (manual count by computercaptured image may be the most practical for the time being).
More important is the controversial use of the term carcinoma, which was previously employed in WHO-2004 only for
invasive and metastatic cases but has now been made synonymous with grade 3 group of tumors. It is becoming clear that
grade 3 group comprises two distinct categories: (1) differentiated but proliferatively more active tumors which typically
have ki-67 indices in the 2050 % range and (2) true poorly
differentiated NE carcinomas as defined in the lung, with ki67 typically >50 %. Further studies are needed to address
these controversial aspects of PanNETs.
Keywords Pancreas . Neuroendocrine tumor . WDNET .
PDNEC

Introduction
Many new developments have taken place in our understanding of the biology, morphologic, and molecular characteristics
of neuroendocrine tumors of the pancreas. In 2010, the World
Health Organization (WHO) proposed a new terminology for
these tumors [1], which has had far-reaching effects on their
current diagnosis and management. As a very important and
positive development, the grading and staging parameters for
these tumors have now been separated from their name and
diagnostic sub-categories. In the ensuing article, we will

66

review these new developments, their limitations and controversies, as well as recent advances in the diagnosis, classification, and management of these tumors.

Definition and Terminology

Neuroendocrine tumors (NETs) of the pancreas are derived
from, and composed of, epithelial cells with phenotypic and
ultrastructural neuroendocrine differentiation. A spectrum of
neoplasms with neuroendocrine differentiation occurs in the
pancreas, ranging from well-differentiated to undifferentiated.
Differentiated examples are essentially counterparts of carcinoid tumors elsewhere, now designated as well-differentiated
neuroendocrine tumors (WDNETs). The term carcinoid however has not been employed in the pancreas, except in the case
of metastases from the gastrointestinal tract or rare serotoninproducing examples. Instead, in the past, pancreatic neuroendocrine tumors were termed islet cell tumors/carcinomas due
to their resemblance to islets of Langerhans, the endocrine
component of this organ. However, it is now presumed that
these tumors do not necessarily arise from the islets, and in
fact, many may actually originate from the neuroendocrinetype cells residing in the pancreatic ductal epithelium. They are
also regarded as a part of the age-old APUDoma concept
since they are fundamentally characterized by amine precursor
uptake and decarboxylation activity. Such APUD cells have
endocrine, autocrine, paracrine, and neuromodulatory functions.
In the WHO-2004 classification system, these tumors were
referred to as pancreatic endocrine neoplasms. However, in
the WHO-2010 system, endocrine tumor was modified to
neuroendocrine tumor based on the belief that the latter term
was a more accurate indicator of the cells neural as well as
endocrine phenotype and, more importantly, to distinguish
them from the conventional endocrine cells of the adrenal,
thyroid, and pituitary gland.
Although it is widely agreed that the term neuroendocrine
neoplasm would be a more accurate categorical designation for
these neoplasms, the term NET was adopted by the WHO in
2010 primarily because the acronym had already been incorporated into the very names of the societies dedicated to studying
them (i.e., the European Neuroendocrine Tumor Society
(ENETS) and the North American Neuroendocrine Tumor
Society (NANETs)). Thus, these tumors are currently referred
to as pancreatic neuroendocrine tumors (PanNETs). The acronym P-NET is used by some authors; however, because the
latter can be confused with the primitive neuroectodermal tumor
acronym, PNET, the former designation is preferred.
Poorly differentiated neuroendocrine carcinomas are currently also included in the generic category of PanNETs and
are recognized as the high-grade (grade 3 [G3]) version of
these tumors. However many authors advocate making a
distinction between ordinary (well-differentiated) PanNETs

Endocr Pathol (2014) 25:6579

and these poorly differentiated neuroendocrine carcinomas,

regarding them as a separate category in the pancreas (see
the following discussion).
It should be noted that, unlike the WHO-2004 system, in
the WHO-2010 (Table 1) as well as in current practice, the
terminology of NETs is no longer modified by the tumors
stage and grade. Previously, the term endocrine carcinoma
was employed for cases showing extra-pancreatic spread or
metastasis. However, this was problematic because a percentage of cases that were classified as benign demonstrated
malignant clinical behavior. This necessitated a shift in the
designation of cases that were originally classified as benign
but in whom metastases were discovered shortly after. While
this issue was resolved with the WHO-2010, another potentially problematic terminology shift was inadvertently created
because in the latter classification carcinoma was made
synonymous only with G3 (poorly differentiated) NETs. Unfortunately some practitioners still use the term carcinoma
for metastatic NETs even when they are well differentiated
(G1 or 2 tumors).

Clinical Aspects
PanNETs are relatively rare tumors, accounting for 2 % of all
pancreatic neoplasms and affecting one to two individuals per
1,000,000/year. However, with improving technology, more
and more cases come to attention as incidentalomas [2].
Tumors are most frequently seen in adults but may rarely
occur in children. Patients are typically 3060 years (with a
mean of 50 years). Men and women are equally affected but
high-grade (G3) carcinomas often occur in older males.
PanNETs more commonly arise in the pancreatic head but
can involve any part of the pancreas.
WDNETs may also arise in a background of familial syndromes including von HippelLindau (VHL) syndrome, tuberous sclerosis complex (TSC), neurofibromatosis type 1
(NF1), and multiple endocrine neoplasia type 1 (MEN1). In
Table 1 Classification system for pancreatic neuroendocrine tumors
Modified from WHO 2010
Differentiation

Grade

Mitosesa/10
|HPF

Ki-67b
proliferation
|index (%)

Well-differentiated PanNET

Grade 1
Grade 2
Grade 3

<2
2-20
>20

<3
3-20
>20

Poorly-differentiated PanNEC

WHO World Health Organization, HPF high power field, PanNET pancreatic neuroendocrine tumor, PanNEC pancreatic neuroendocrine carcinoma (large and small cell type)
a

In at least 50 HPFs

MIB1 antibody

Endocr Pathol (2014) 25:6579

MEN1 and VHL, tumors are frequently multifocal and occur

in a younger age group [3, 4]. Up to 80 % of MEN1 patients
develop multiple (<0.5 cm) WDNETs (microadenomatosis) as
well as proliferating or dysplastic islets, but these are usually
associated with at least one or more macrotumors (> 0.5 cm)
(Fig. 1) [3]. Up to 17 % of patients with VHL syndrome
develop PanNETs. Most of these are non-functioning and
60 % have clear cell morphology due to intracytoplasmic lipid
[4, 5]. Mutation of the VHL gene has been identified in these
patients and plays a role in disease development. Although
lipid-rich cells were once thought to be pathognomonic of
VHL-related WDNETs, similar cells have now been identified
in a subset of WDNETs that are either MEN1-related or nonsyndrome-associated [6]. Up to 10 % of patients with NF1
develop WDNETs, which have an increased affinity for the
ampullary region.

Cell Type and Functionality Status

Based on clinical presentation, PanNETs were historically
separated into two clinical categories, functional versus
non-functional, based on the symptoms elicited by the predominant hormone the tumors secrete into the bloodstream.
Clinical presentation and prognosis were once felt to be
strongly linked to tumor functionality. However, it has now
become clear that many tumors secrete multiple hormones
albeit with variable detectability and functional impact. Additionally, tumors may change their hormone productivity over
time. Moreover, with recent improvements in imaging modalities, more non-functional PanNETs are being discovered, and
so the value of a functionality-based classification system has
lost support.

67

Functional tumors are still named based on primary clinical

presentation and serologic activity (e.g., gastrinoma and
insulinoma) and NOT on immunohistochemical hormone expression. Insulinomas often present with the classical Whipple triad due to excessive insulin production [7]. Gastrin
overproduction is associated with ZollingerEllison syndrome, while overproduction of vasoactive intestinal peptide
(VIP) can cause severe watery diarrhea, hypokalemia, and
achlorhydria, leading to death.
Hormone status has also been shown to be associated with
biologic behavior, which is indirectly linked to the stage at
which a hormone-producing tumor causes clinical symptoms.
For example, most insulinomas have benign behavior, which
is largely attributable to their early detection. Over 70 % of
insulinomas are <2 cm; however, they cause striking insulinrelated alterations in glucose metabolism which typically
leads to early diagnosis. Patients have the classical Whipple
triad, features of which include (1) low blood glucose and (2)
symptoms and signs of hypoglycemia, which (3) resolve with
blood glucose elevation [7]. Conversely, most glucagonomas
are aggressive and have already metastasized at diagnosis.
Although these tumors are typically >2 cm, they are often
not detected early because glucagon levels are not clinically
elevated. Having said this, it is possible that cell type may
have additional yet to be defined biologic characteristics
that lead to divergent prognoses in these tumors.
Recently, small PanNETs occurring on the wall of pancreatic ducts, usually of serotonin-producing type, have come to
clinical attention due to duct obstruction sometimes mimicking intraductal papillary mucinous neoplasms [8, 9].

Classification and Sub-grouping

In the previous WHO-2004 classification system, PanNETs
had been sub-categorized as having benign behavior, as
having uncertain behavior, or as frank carcinoma (with
extrapancreatic spread or metastasis) based on an amalgamation of grade, stage, spread status, and adjunct prognosticators (vascular and perineural invasion). This was a fairly
unique approach for any type of tumor in the body. However,
in the WHO-2010 system, grade and stage were separated out,
as is the case for tumors of other organs. The latter is discussed
in detail in the following sections.

Grading

Fig. 1 In the background of atrophy and hyperplastic islets, there is an

evolving neuroendocrine tumor that is 3.0 mm in size (microadenoma)
(hematoxylin & eosin stain, 200 magnification). This patient had a
history of multiple endocrine neoplasia type 1 syndrome

In the current 2010-WHO classification system, pancreatic neuroendocrine tumors are divided into well-differentiated [grade 1
(G1) and 2 (G2)] NETs (WDNETs) and poorly differentiated
(G3) neuroendocrine carcinoma (PDNECs). This system gives
the distinct impression that PDNECs are in continuum with

68

ordinary (well-differentiated) PanNETs, which is not necessarily the case. Evolving evidence strongly suggests that G3
tumors should be regarded separately and as such will be
discussed separately later.
The grading of NETs is based on two calculations: (1)
mitotic count and (2) ki-67 labeling index (using the MIB1
antibody). G1 NETs are defined as having a ki-67 index of
<3 % and <2 mitoses/10 high power fields (HPF) or 2 mm2.
G2 NETs have a ki-67 index of 320 % or 220 mitoses/10
HPFs (Table 1) [1]. The WHO recommends that for mitotic
counts, at least 50 HPFs should be counted, and for ki-67
index, a minimum of 500 tumor cells should be counted in
tumor hot spots. For grade-discordant cases (based on differences in mitotic count and ki-67 index), the higher grade
should be used.
There are several areas of ambiguity in this grading system,
which are now being clarified. For example, in the original
guidelines, it was unclear which category (G1 versus G2) to
place ki-67 indices of 23 %, a range in which a substantial
number of cases fell. The cutoff value has now been clarified
by NANETs as <3 %, indicating that cases within this range
are actually G1.
It should be noted that in the current grading scheme,
morphology is, for the most part, unfortunately disregarded.
It has now become clear that a group of morphologically welldifferentiated NETs that has comparatively high ki-67 indices
of >20 % exists. These well-differentiated but proliferatively
high-grade tumors typically have a ki-67 index of <50 %.
Unfortunately, per 2010-WHO guidelines, these cases are
classified as high-grade (G3) neuroendocrine carcinomas,
along with full-blown PDNECs (which have a far worse
prognosis), but they behave in a clinically worse fashion than
typical G2 tumors (see Biologic Behavior) [10]. Additionally, a study by Sorbye et al. recently indicated that true
PDNECs of the gastrointestinal tract, with their aggressive
behavior and platinum sensitivity, may be better defined by ki67 indices of >50 % [11]. Therefore, it is becoming clear that
the current G3 group defined as >20 % will have to be split
into two separate categories in the future in order to distinguish the well-differentiated G3 tumors (ki-67, 2050 %)
from the poorly differentiated carcinomas (G4?; ki-67 >50 %).
At the other end of the spectrum, for the distinction of G1
from G2 tumors, some groups advocate using 5 % as the cutoff,
which may be a more accurate and reproducible cutoff. While
further studies are needed to clarify this issue, the current 2 %
cutoff remains widely used. Additionally, the current cutoff
which is widely used and endorsed by the WHO-2010 is often
referred to as the 2 % cutoff in publications; however, as
recently clarified by NANETS, it is actually applied as <3 %
and 3 % in practice and as such it is the 3 % cutoff.
It should be noted that other proposed grading systems are
also in place. Hochwald and Klimstra had proposed a two-tier
grading system for PanNETs based on necrosis and mitotic

Endocr Pathol (2014) 25:6579

count and separated these neoplasms into low-grade (showing

no necrosis and mitoses <2/50 HPFs) and intermediate-grade
tumors (with necrosis and/or mitoses >2/50 HPFs) [12]. While
many studies have found necrosis to be a strong prognosticator, others have failed to confirm this observation. In addition,
immunoexpression of cytokeratin 19 [1315] and c-kit [16]
have also shown strong prognostic value in some studies.
Vascular and perineural invasion are also believed to be
features of aggressive behavior. They are not a part of the
official grading system but are instead reported separately.
Some studies failed to find a correlation between vascular
invasion and survival, which is partly attributable to the fact
that PanNETs are highly vascularized tumors, and (similar to
the challenges encountered in other endocrine neoplasia) it is
often difficult to distinguish true vascular invasion from
tumor-related vascularity. We advocate a purist attitude when
reporting vascular invasion.

Issues with Grading

If the ki-67 index is calculated properly, it is seldom, if ever,
lower than the mitotic count (i.e., ki-67 almost always trumps
mitotic count). This is not surprising, considering that ki-67
covers the whole cycle of the mitotic process. Additionally,
counting mitoses appears to be more challenging than calculating ki-67 index. Mitotic activity is highly variable depending on the type of microscope used. Additionally, strict criteria
should be used to count mitotic figures, and one should avoid
overcalling pyknotic nuclei and other mimics mitoses. Stromal
density and section thickness also significantly affect mitotic
count by artificially altering the cellular density fitting a given
field area, thereby changing the denominator, whereas for
ki-67, these factors are non-issues since the denominator is the
number of cells counted and not the field area.
However, the calculation of ki-67 index is not free from
challenges. A variety of methods are used for counting with
several well-known and some not so obvious advantages
and disadvantages. The pitfalls of the eyeballing method
have been aptly illustrated by Tang et al. who showed poor
interobserver agreement with this methodology [17].
Eyeballing is now discouraged unless a case has an unequivocally low or high index. Automated systems pose their own
unique limitations, including operator (in)experience and the
softwares ability to detect and account for overlapping cells,
staining of non-tumor cells (including lymphocytes), and
background hemosiderin or other brown pigment. Additionally, with automated systems, hot spot selection is highly
operator dependent and requires proper training to ensure
accurate results. The cost of automated instruments may also
be prohibitive, limiting widespread accessibility. The need for
batching may also be a shortcoming. If these challenges are
overcome or carefully addressed, automated counting may

Endocr Pathol (2014) 25:6579

prove to be a fairly reliable method. Of note, currently, there is

a separate billing code for automated counting. The direct
real-time eye-count approach, which is widely used in hematology, has also proven difficult in grading PanNETs, unless
a grid is used. However, even with a grid, cell distribution and
overlap make the test difficult to perform. We have found that,
in our current practice, the most practical, cost-effective, and
reproducible method for counting ki-67-positive cells is a
manual count done either on a computer screen or as a captured static photomicrograph of tumor hot spots, where negative and positive tumor cells are visualized and immediately
crossed off once counted. The latter can be done with a digital
camera attached to the microscope, a setup that is far less costly
than automated methods and takes an average of 6 min [18].
Naturally, tumor heterogeneity and subjectivity in hot spot
determination are additional factors that may lead to variations
in results. Whether to count the less intensely labeled cells as
positive is also a question. For this, normalization with the
background tissue is helpful. There should be no background
staining in the stroma or cytoplasm in a properly performed ki67 assay and light brown nuclei should generally be
disregarded. In tumors with abundant lymphocytes or other
confounding cell types, ki-67 indices may be falsely elevated.
In such cases, dual staining with neuroendocrine markers may
be necessary, especially if the labeling index appears to be
close to a cutoff, in particular 20 %.
Despite these challenges, ki-67 has nonetheless been
shown in numerous studies to correlate highly with clinical
outcome. Perhaps counter-intuitively, it has proven to be more
practical and reproducible than mitotic count, and is as, if not
more applicable, than other quantitative immunohistochemical assays that are the norm in other organs such as the breast
(ER/PR/her2neu) and stomach (EGFR).

Staging
The staging of PanNETs is now done separately from the
grade, unlike the WHO-2004 system where the two were
combined to determine the category.
There are currently several different staging protocols for
PanNETs. The one from The American Joint Commission on
Cancer (AJCC)/Union of International Cancer Control
(UICC)/College of American Pathologists (CAP) [1921]
differs from that of the European Neuroendocrine Tumor
Society (ENETS) [22] system (Table 2). The AJCC has essentially adopted the staging used for exocrine tumors, which
is problematic for pT3 tumors, which are defined in part by
peripancreatic soft tissue involvement. Because the pancreas
has very irregular lobules, no capsule, and fat placement
throughout, it is often difficult, if not impossible, to assess
peripancreatic soft tissue invasion in these tumors [2325].
Furthermore, most PanNETs protrude from the pancreas even

69
Table 2 AJCC/UICC/CAP and ENETS TNM classification systems for
pancreatic neuroendocrine tumors
T stage

AJCC/UICC
[19, 20]/CAP TNM

ENETS [22] TNM

T1
T2
T3

Confined to pancreas, <2 cm

Confined to pancreas, >2 cm
Peripancreatic spread,
without major vascular
invasion
Invasion of adjacent organs
or major vessels

Confined to pancreas, <2 cm

Confined to pancreas, 24 cm
Confined to pancreas, >4 cm,
or invades the duodenum
or bile duct
Invasion of adjacent organs or
major vessels

T4

when they are small, especially those located in the tail. That is
probably why many studies have failed to identify a correlation between survival and T stage. In contrast, the ENETS
staging system relies predominantly on more reproducible
criteria such as tumor size rather than poorly defined and
irreproducible parameters like peripancreatic extension.
A recent study of 1,072 post-surgical PanNET patients
showed that the ENETS system was superior to the
AJCC/UICC system for stratifying risk of death and creating
risk-based treatment guidelines [26, 27]. Despite the proven
discrepancies between the two staging systems, there appears
to be no adverse effect on diagnosis and management [28]. In
our current practice, we typically provide the stage by both
schemes, separately.

Early PanNETsIncipient Neoplasia/Dysplasia

Extensive study of patients with MEN1 has led to the recognition of so-called precursor lesions of PanNETs. MEN1 patients
often develop proliferative or enlarged islets, some of which
acquire hormonal and morphologic clonality and achieve a size
beyond that of normal islets. The latter are regarded as dysplastic and are akin to pulmonary tumorlets. Naturally, it is
difficult to define the dysplasia and determine the point at
which such proliferations become autonomous and neoplastic.
These lesions often acquire an enveloping dense fibrous band,
a feature of neoplastic transformation commonly encountered
in endocrine organs. They also often show uniform staining
with pancreatic hormones. Lesions that appear fully established
but are <0.5 cm are designated as microadenomas. Loss of
heterozygosity for the MEN1 gene occurs in the established
microadenomas, while the precursor lesions appear to be heterozygous [29]. Microadenomatosis and proliferating islets may
also occur in the absence of hereditary syndromes [29] and
produce increased glucagon and insulin. Occasionally, solitary
minute NETs occur sporadically and are incidental findings in
resections performed for other pathologies. In the absence of
proliferative islets or hereditary syndromes, we report them as
minute, incidental NET, x.x cm (microadenoma).

70

It was in MEN1 syndrome that the idea of preneoplastic

precursors of pancreatic NETs was first fashioned as it gave
researchers the unique opportunity to study these tumors in
their earliest form. The term nesidioblastosis was coined by
Laidlaw [30] who believed that pancreatic adenomas (islet cell
tumors) were a secondary reactive process in ductal epithelium that occurred after an awakening of their islet and duct
building capability. He named these pluripotent cells
nesidioblasts (Greek for islet builder) and used the term
nesidioblastosis to describe islet cell proliferation and
nesidioblastoma for islet cell adenomas.
Others later proposed that nesidioblastosis might explain
the patterns of disease seen in familial multiple endocrine
adenomatosis (MEA) syndrome [31, 32]. For almost a century, nesidioblastosis has remained a part of the dysplasia/early
neoplastic transformation concept of PanNETs. Adult forms
of diffuse nesidioblastosis (abnormality of cells) have also
been implicated in persistent hyperglycemia [33].

Characteristics of Ordinary PanNETs (WDNET)

Fast-forward to the present day, where our knowledge of the
clinical aspects, biology, and associated alterations in these
tumors has expanded significantly: so how do we now view
and evaluate PanNETs?
Gross Features
PanNETs can involve any part of the pancreas. They are
typically well circumscribed and in sporadic cases are often
solitary. Their appearance may vary due to the contents of the
tumor. The vast majority are tan-brown and fleshy due to the
dense cellularity and relative stromal paucity (Fig. 2). Few
examples contain a more abundant stromal component and
may appear more yellow-white and firm/sclerotic. Some cases
have a complete capsule, but in most cases the capsule is
incomplete or non-existent. The vast majority of PanNETs
are solid tumors, but about 5 % present with cystic

Fig. 2 Well-differentiated pancreatic neuroendocrine tumor is seen arising in the pancreatic body as a solitary, well-circumscribed tan yellow
mass. Note the absence of hemorrhage and necrosis

Endocr Pathol (2014) 25:6579

degeneration, typically with clear serosanguinous contents,

and must be differentiated radiologically from cystic pancreatic tumors. Some examples may have extensive blood (socalled peliotic variant) and appear grossly hemorrhagic.
PanNETs, especially those in the tail, often protrude from
the pancreatic surface and have a nodular bulge or polypoid
appearance on gross examination.
Histopathology
Unlike the more common pancreatic ductal adenocarcinoma,
PanNETs are mostly histologically stroma poor with morphologic features that are dependent on their level of differentiation. They are typically well circumscribed with variable
amounts of capsule-like tissue. In WDNETs, cells are arranged as sheets of monotonous epithelial cells, with anastomosing cords, nests, and trabeculae, coarse salt and pepper
chromatin, and a fair amount of cytoplasm that may be eccentric and plasmacytoid (Fig. 3). Glandular structures, tubuloacinar units, and pseudorosettes may also be seen, and the
latter may be so striking as to lead to misinterpretation as
solid-pseudopapillary neoplasms (SPNs). Mitotic activity is
usually scant. If mitoses are easily found, the possibility of an
acinar cell neoplasm or other high-grade tumor should be
excluded.
The correlation of morphologic findings with hormone
production (functionality status) is unreliable. Perhaps one
exception is the close association of psammoma bodies and
glandular pattern with immunohistochemical positivity for
somatostatin.
There are morphologic variants of PanNETs that are peculiar and diagnostically challenging. The lipid-rich variant of
NETs is characterized by abundant foamy/microvesicular cytoplasm, which creates a picture virtually indistinguishable
from adrenocortical cells (Fig. 4). Some but not all of these
are associated with VHL syndrome. Interestingly, the cytoplasmic lipid frequently pushes and compresses the nucleus
peripherally, thus obscuring the classical nuclear features of a
NET, and may lead to their misdiagnosis as metastatic renal
cell carcinoma or adrenal tumor on morphology. Some
PanNETs have more uniform cytoplasm and may also show
prominent rhabdoid [34] or signet ring features. WDNETs are
also notorious for having oncocytic features characterized by
abundant granular, eosinophilic cytoplasm, and a single prominent nucleolus, reminiscent of hepatocellular carcinoma
(Fig. 5). When the oncocytic variant metastasizes to the liver,
it can be easily misdiagnosed as primary hepatocellular carcinoma especially if appropriate immunohistochemical stains
are not performed. Unlike other morphologic variants, the
oncocytic PanNET has been found to be more aggressive in
some studies. Some tumors show a distinct small round cell
appearance and have a very high nucleus-to-cytoplasm ratio.
However, the nuclear molding, extensive necrosis, and high

Endocr Pathol (2014) 25:6579

71

Fig. 3 a Well-differentiated
pancreatic neuroendocrine tumor,
grade 1. The tumor is stroma poor
with an organoid arrangement of
bland cells and intervening thinwalled blood vessels
(hematoxylin & eosin stain, 400
magnification). b Tumor cells are
strongly positive for
chromogranin A (400
magnification)

mitotic activity seen in the small-cell variant of PDNEC are

not typically seen. A peliotic variant of PanNETs is also
described, in which pools of blood and dilated blood vessels
are conspicuous both grossly and microscopically. Marked
cystic degeneration is seen in 5 % of tumors, and in such
cases, there is typically a cuff of ordinary PanNET on the cyst
wall, making the diagnosis rather straightforward.
PanNETs may show marked nuclear pleomorphism, socalled endocrine atypia, similar to that seen in other endocrine
organs (Fig. 6). This can be so extensive that it can lead to
misdiagnosis as adenocarcinoma [35]. Tumors may rarely show
morphologic features mimicking those of paragangliomas [36].
Typically, these paraganglioma-like tumors show more
marked endocrine atypia (Fig. 7). The tumors presence in
the pancreas, its coexpression of keratin and neuroendocrine
markers, as well as the absence of basophilia and granularity
typically seen in paraganglioma help to clinch the diagnosis. In
some PanNETs with endocrine atypia, the degenerative
symplastic cells (akin to those seen in symplastic leiomyomas

or ancient schwannomas) can be so abundant and atypical that

the case can be misdiagnosed as a high-grade malignancy.
Some PanNETs contain abundant ductal elements, deemed
by some as the ductulo-insular variant of PanNET. The
ducts in such cases show obvious benign cytology. It is
debatable whether these are entrapped native ductules proliferating secondary to the local effects of the tumor or represent
an unusual transdifferentiation of tumor cells. The former
appears to be the more plausible explanation. These tumors
do not seem to behave differently than conventional PanNETs.
However, if the ductular elements have severe atypia, the
remote possibility of a mixed ductalneuroendocrine carcinoma ought to be considered, but these are exceedingly uncommon tumors.
Although tumors are usually stroma poor, they may exhibit
marked sclerosis, especially those that secrete serotonin [37].
Necrosis is not a typical feature of PanNETs and, when
present, is usually focal, involves single cells, or may be
comedo-like. If there is extensive tumor necrosis, the

Fig. 4 Well-differentiated pancreatic neuroendocrine tumor, lipid-rich

variant. Tumor cells are large with abundant, clear, vacuolated cytoplasm
and oval nuclei lacking the salt and pepper chromatin typical of these
tumors (hematoxylin & eosin stain, 400 magnification)

Fig. 5 Well-differentiated pancreatic neuroendocrine tumor, oncocytic

variant. This example shows tumor cells with abundant granular eosinophilic cytoplasm and nuclei with prominent central nucleoli (hematoxylin
& eosin stain, 200 magnification)

72

Fig. 6 Well-differentiated pancreatic neuroendocrine tumor, pleomorphic

variant. Tumor cells show marked nuclear enlargement, pleomorphism, and
bizarre nuclei (hematoxylin & eosin stain, 400 magnification)

possibility of an acinar cell neoplasm or high-grade neuroendocrine carcinoma ought to be considered.

Poorly Differentiated Neuroendocrine Carcinomas

(PDNEC)
In the WHO-2010, PDNECs are classified under the rubric of
grade 3 PanNETs along with proliferatively active ordinary
PanNETs. However, it is probably more accurate to regard
poorly differentiated neuroendocrine carcinomas (PDNEC) as
a separate entity or, at least, as the undifferentiated version of
differentiated PanNETs. These are essentially defined as their
counterparts in the lung. They are extremely aggressive tumors
that have often disseminated by the time of diagnosis. Median

Fig. 7 Well-differentiated pancreatic neuroendocrine tumor with

paraganglioma-like features. Tumor cells show focal pleomorphism
and a prominent nested pattern reminiscent of a paraganglioma (hematoxylin & eosin stain, 100 magnification)

Endocr Pathol (2014) 25:6579

survival is typically less than 2 years. Platinum-based therapeutic agents have shown some promise in controlling their
growth; however, their overall prognosis remains grim.
These poorly differentiated, high-grade (G3) carcinomas are
subdivided based on cell size into small- and large-cell variants.
The large-cell variant is more common, often large, and characterized by large cells with prominent nucleoli and variable
cytoplasm. Mitotic activity is brisk (often >40/10 HPFs) and
necrosis is often extensive. The small-cell variant shows small to
intermediate cells with coarse salt and pepper chromatin, high
nucleus-to-cytoplasm ratio, inconspicuous nucleoli, prominent
nuclear molding, and crush artifact (Fig. 8). Mitotic figures are
easily identifiable and there is extensive tumor necrosis.
Better delineation of PDNECs from ordinary G3 PanNETs
has been elucidated in some recent studies. Well-differentiated
but highly proliferative PanNETs typically have a ki-67 index
below 4050 % [10], whereas PDNECs typically show ki-67
>55 %, are characterized by highly aggressive behavior, and
require platinum-based therapy [11].
It should be noted here that PDNECs are very uncommon
tumors, and most cases that are diagnosed as such prove to be
acinar cell carcinomas once studied more carefully [38].

Ancillary Studies
Immunohistochemistry
PanNETs are essentially defined by the production of cytoplasmic neuroendocrine granules, which are commonly
highlighted by immunohistochemical markers chromogranin,
synaptophysin, and CD56 (Fig. 3). Ordinary well-differentiated PanNETs tend to show stronger more diffuse staining
with neuroendocrine markers than PDNECs. Synaptophysin
is highly sensitive but less specific. Chromogranin is the most
specific neuroendocrine marker but is the least sensitive of the
three. CD56 antibody (123c3) which is directed against neural

Fig. 8 Small-cell neuroendocrine carcinoma of the pancreas. Tumor

cells are arranged as tight groups and sheets of small cells with high
nuclear to cytoplasmic ratio (hematoxylin & eosin stain, 200
magnification)

Endocr Pathol (2014) 25:6579

cell adhesion molecules is the least specific neuroendocrine

marker but is nonetheless useful in the differential diagnosis of
these tumors. Neuroendocrine markers are also positive in
large-cell carcinomas, but staining intensity and distribution
are not as robust as in WDNETs.
Additionally, NETs strongly express pancytokeratin as a
marker of epithelial differentiation. Cytokeratin 19 (CK19),
which is regarded as a marker of ductal lineage in the pancreas, is sometimes expressed by PanNETs and has been found in
some studies to be a marker of more aggressive behavior
[1315]. Because of the latter, some authors advocate doing
a CK19 stain on all PanNETs. CD117 (c-kit) has also been
proposed as an independent adverse prognostic marker in
PanNETs; however, its use is not routine [16].
Focal acinar differentiation (with trypsin or chymotrypsin
positivity) has also been reported in PanNETs, but it is usually
present as isolated cells. Approximately 45 % of sporadic
PanNETs show loss of expression of DAXX and ATRX
immunostains, which correlates with mutations in the DAXX
and ATRX genes [39]. Loss of Dpc4, a common phenomenon in
ductal adenocarcinoma of pancreas, is only rarely (if ever) seen
in neuroendocrine tumors, which suggests that inactivation of
the SMAD4/DPC4 gene is a rare event in these tumors.
CD99, a transmembrane glycoprotein encoded by the MIC-2
gene, and progesterone receptor (PR) are also positive in a
subset of PanNETs, in a manner similar to non-neoplastic islet
cells. The nuclear transcription factor islet-1 (isl1) is frequently
expressed in PanNETs [40, 41]. Pancreatic duodenal homeobox
1 (PDX-1), a homeodomain transcription factor, plays a regulatory role in early pancreatic development and is a marker of
pancreatic PanNET (in both primary and metastatic tumors).
Some studies have found isl1 and PDX-1 to be helpful in the
differential diagnosis of pancreatic NETs from non-pancreatic
WDNETs at metastatic sites such as liver, but their sensitivity
and specificity have not been rigorously tested [42].
Immunohistochemical analysis of PanNETs often lacks correlation with serologic status and, in some cases, may yield
opposite results, presumably due to the rapid release and dispersal
of the hormone product without cellular accumulation, but with
retention of other hormones produced. Therefore, correlation
between functional status and immunohistochemical hormone
expression remains an imperfect science, and the use of immunohistochemistry to assess hormone production is unreliable.
In a recent study, small- and large-cell PDNECs were shown
to be genetically related but distinct from ordinary PanNETs
[43]. P53 is positive in both small-cell (100 %) and large-cell
NECs (90 %) but is negative in well-differentiated PanNETs.
Rb protein is lost in 6090 % of PDNECs, and for tumors that
retain Rb, there is usually concurrent loss of p16 staining,
indicating that the two are mutually exclusive in NEC. Smalland large-cell NECs show abnormal immunolabelling with p53
and Rb in 95 and 74 % of cases, respectively, which correspond
to intragenic mutated TP53 and retinoblastoma RB-1 genes

73

[43]. Conversely, WDNETs retain Rb and p16. PDNECs retain

DAXX and ATRX immunostains, unlike ordinary PanNETs in
which these are frequently lost. PDX-1 is positive in up to 40 %
of NECs, with large-cell tumors showing more positivity.
PAX8 is positive in 50 % of NETs (including PDNECs and
WDNETs).
BCL2 protein, which is overexpressed in small-cell carcinoma of the lung, is also overexpressed in PDNECs (100 % of
small-cell and 50 % of large-cell tumors) but is negative in G1
tumors and variably expressed in G2 WDNETs [43]. The
platinum-based chemotherapeutic agents used to treat smallcell lung carcinoma exert their effects by inducing apoptosis
via a BCL2 regulated pathway. This suggests that BCL2
expression by PDNECs could potentially be exploited with
BCL2 antagonists similar to those used in the lung.
Electron Microscopy
On ultrastructural analysis, NETs contain membrane-bound,
dense core neurosecretory granules, which historically have
been highly useful in their diagnosis. However, ultrastructural
analysis is no longer accessible in most institutions and has
been superseded by immunohistochemistry. Electron microscopy may be useful in very poorly differentiated tumors or
cases with an amphicrine phenotype.
Molecular Studies
PanNETs associated with MEN1 and VHL syndromes show
mutations in the MEN1 and VHL genes. In MEN1, the MEN1
(menin) gene located on chromosome 11q13 is mutated.
Among sporadic PanNETs, somatic MEN1 mutation or loss
of heterozygosity at the MEN1 locus is seen in a significant
number of cases [39, 4447]. MEN1 mutation is associated
with a better prognosis. PDNECs however are infrequently
associated with MEN1[1]. Deletion of the VHL gene occurs in
up to 25 % of sporadic WDNETs.
Among sporadic PanNETs, inactivating somatic mutations
of the DAXX (death-domain associated protein) and ATRX
(alpha thalassemia/mental retardation syndrome X-linked)
genes have been identified in 45 % of cases. DAXX and
ATRX proteins are crucial for telomere maintenance [48],
and their associated genes are implicated in chromatin remodeling [39]. When mutated, DAXX and ATRX result in immunohistochemical loss of their corresponding proteins.
Approximately 15 % of PanNETs show genetic mutations
in the mammalian target of rapamycin (mTOR) cell signaling pathway proteins [39]. The mTOR pathway alteration
represents a unique therapeutic target for drugs like everolimus which are directed specifically at this pathway [49].
Chromosomal gains and losses have also been identified in
PanNETs, and multiple chromosomal abnormalities are associated with a worse prognosis [50].

74

Differential Diagnosis
Differential diagnoses of PanNETs include all primary pancreatic neoplasms with diffuse, cellular sheet-like proliferation
as well as metastatic tumors.
Acinar cell carcinoma is one such primary pancreatic tumor
that is derived from exocrine acinar cells and shows acinar
differentiation as evidenced by positivity for trypsin, chymotrypsin, lipase, other pancreatic enzymes, and BCL10. Tumor
cells have abundant, granular, zymogen-rich cytoplasmic granules, coarse chromatin, and prominent, sometimes cherry red,
central nucleoli. A subset of acinar cell carcinomas is morphologically almost indistinguishable from PanNETs. In addition,
most acinar cell carcinomas focally express neuroendocrine
markers either as scattered individual cells or large zones or
even as a very-well-formed distinct component (mixed carcinomas). The presence of exuberant mitotic activity in the setting
of seemingly bland monotonous epithelial cells is more in
keeping with acinar cell carcinoma than a differentiated
PanNET, in which mitotic figures are often few and far between. Additionally, acinar cell carcinomas show prominent
central nucleoli, which are typically less prominent in
PanNETs. Most helpful though is the neuroendocrine chromatin pattern. Morphologic distinction is facilitated by immunohistochemistry. Acinar cell carcinoma is negative or only focally positive for neuroendocrine markers and shows positivity for
pancreatic enzymes, while NETs show opposite results.
Pancreatoblastomas are morphologically similar to acinar cell
carcinoma and may also be confused with PanNETs. However,
immunohistochemistry as well as the finding of classical
squamoid morules should facilitate distinction.
Morphologic distinction of mixed acinarneuroendocrine
carcinoma, an acinar cell neoplasm that shows over 30 %
neuroendocrine differentiation, can be especially challenging
both on morphology as well as immunohistochemistry as
tumor cells stain with both acinar and neuroendocrine
markers. The key to distinguishing this variant from PanNET
is the mixed neuroendocrine and acinar immunophenotype as
well as the distribution of the neuroendocrine component,
which should not be significant in these mixed tumors.
Another key differencel is SPN. Both PanNET and SPN can
show solid sheet-like pattern and nested growth (Fig. 9). Typically, the nesting is more vague in SPNs, and both patterns
blend, rather imperceptibly, with each other, creating a more
mesenchymal-like appearance. Striking overlapping of tumor
nuclei is more prominent in SPN and the presence of large
cytoplasmic vacuoles also favors SPN. On close examination,
one should note the absence of the salt and pepper chromatin
of neuroendocrine tumors and the presence of fine, open,
powdery chromatin along with longitudinal nuclear grooves,
which are distinctive findings in SPN. In addition, the presence
of PAS-positive, diastase-resistant hyaline globules is supportive of SPN, although similar globules can occasionally be seen

Endocr Pathol (2014) 25:6579

in PanNETs [51]. Immunohistochemistry is extremely helpful

in making the diagnosis as SPNs, in contrast to PanNETs, show
only focal or weak positivity for keratin and chromogranin and
strongly express nuclear -catenin. It should be kept in mind
that SPNs may express neuroendocrine markers CD56 and
synaptophysin, sometimes diffusely. This is a potential source
of misinterpretation of immunohistochemistry unless one pays
close attention to the distinguishing morphologic features. It
should also be noted that some PanNETs exhibit striking
pseudopapillary pattern mimicking SPNs.
The clear cell or lipid-rich variant of WDNET can potentially be misdiagnosed as metastatic clear-cell renal cell carcinoma (RCC) involving the pancreas. Metastasis of clear-cell
RCC to the pancreas is a known phenomenon and is one of the
most frequent tumors to metastasize to the pancreas [52].
Additionally, the lipid-rich variant of WDNET often fails to
show the classical nuclear features of neuroendocrine neoplasms, making diagnosis more challenging (Fig. 4). As mentioned earlier, the oncocytic variant of WDNET may resemble
hepatocellular carcinoma (HCC) and, if it first presents as liver
metastasis, can be misdiagnosed as HCC.
Plasmacytomas are known to involve the pancreas, and for
WDNETs showing marked plasmacytoid features, distinction
from these hematopoietic neoplasms can be especially challenging. It is important to note that in plasmacytoma, the chromatin
distribution has a clock-face pattern with a distinctive perinuclear
Hof. In WDNET, the chromatin distribution is different and the
perinuclear Hof is absent. Plasmacytomas stain positively for
CD138 and show clonal immunoglobulin, light chain restriction,
with negativity for keratin and neuroendocrine markers.
The large-cell variant of poorly differentiated (G3) neuroendocrine carcinoma resembles poorly differentiated adenocarcinoma and may be misdiagnosed as such unless suspected
on morphologic assessment. Immunohistochemical expression of neuroendocrine markers in addition to keratin supports
the former. The small-cell variant of PDNEC may resemble
metastatic small-cell carcinoma of lung. TTF1 stain is not
helpful in distinction as small-cell carcinoma is TTF1positive in a variety of extra-pulmonary sites. Clinical information and a history of previous carcinoma are especially
important in the accurate diagnosis of such cases.
The crushed and molded tumor cells of small-cell PDNEC
may resemble a high-grade lymphoma. Morphologic distinction is often impossible and immunohistochemistry is required
for diagnosis.
Small, round, blue cell tumors may involve the pancreas
primarily or secondarily. These include desmoplastic small,
round-cell tumor (DRCT) [53] as well as primitive
neuroectodermal tumors (PNETs) [54]. The latter also expresses keratin in addition to CD99, both of which are also
expressed by WDNETs. However, the salt and pepper chromatin distribution is absent in DRCT and PNETs, which also
show characteristic molecular alterations (EWS-WT1 gene

Endocr Pathol (2014) 25:6579

75

Fig. 9 Solid pseudo-papillary neoplasm of the pancreas mimicking a

pancreatic neuroendocrine tumor. a There is an organoid arrangement of
(stroma-poor) tumor cells with morphology similar to a well-differentiated pancreatic neuroendocrine tumor (hematoxylin & eosin stain, 100

magnification). b Tumor cells show nuclear and cytoplasmic positivity for

-catenin immunostain (shown on the right) in contrast to the acinar cells
on the left, which do not express nuclear -catenin (200 magnification)

fusion and t(11;22), respectively) that help in distinction.

Merkel cell carcinoma may rarely metastasize to the pancreas
and in one report was present with a synchronous PanNET
(insulinoma) [55]. Awareness of a previous history of this
disease should prompt appropriate immunohistochemical
studies to exclude this differential.
Because of their brisk mitotic activity and morphologic
resemblance to neuroendocrine cells, acinar cell carcinoma
may be mistaken for large-cell neuroendocrine carcinoma. A
panel of immunohistochemical stains including both neuroendocrine and acinar markers (trypsin, chymotrypsin) can
help to distinguish the two.

ratio, crushed artifact, and nuclear molding. There is extensive

single-cell and background necrosis in addition to brisk mitotic
activity seen both on smears and cellblock. Tumor cells express
pancytokeratin as well as neuroendocrine markers, which help
to support the diagnosis. For large-cell neuroendocrine carcinoma, tumor cells are usually large with variable cytoplasm as
well as prominent central nucleoli. On cytology, these tumors
may resemble a poorly differentiated adenocarcinoma and may
not show the classical salt-and-pepper chromatin of WDNETs
or the nuclear molding of their small-cell counterpart. If a
neuroendocrine primary is not suspected at the time of evaluation, this diagnosis may be easily missed.
The grading of PanNETs has been performed on FNA by
immunolabelling cell blocks with ki-67 [56, 57]. One might
argue that cytologic samples are notoriously hypocellular and
therefore lack the requisite minimum of 500 tumor cells

Cytology
Fine needle aspiration (FNA) has been used in the cytologic
diagnosis of primary and metastatic PanNETs for many years. It
is typically performed by endoscopic ultrasound-guided FNA,
which is successful at identifying NETs in up to 90 % of cases.
The cytologic features of these tumors are distinctive. For
ordinary, well-differentiated (G1 and 2) NETs, tumor cells are
classically singly dispersed or clustered as bland-appearing
monotonous cells with round to oval nuclei and variable cytoplasm, which may or may not have distinctive plasmacytoid
features (Fig. 10). In addition, multiple cytoplasmic vacuoles
may be seen. On Papanicolaou stain, tumor cells show the
classical salt-and-pepper chromatin distribution, which supports the diagnosis (Fig. 10). Pseudorosettes are visible not only
on cellblock but also on smears. Immunohistochemical analysis
of cellblock material often demonstrates tumor cell positivity
for keratin as well as neuroendocrine markers.
In poorly differentiated neuroendocrine carcinoma of smallcell type, the cytologic features are similar to the lung counterpart. Tumor cells are often crushed and small to intermediate in
size with irregular nuclear borders, high nucleus-to-cytoplasm

Fig. 10 Fine needle aspiration of a well-differentiated pancreatic neuroendocrine tumor showing singly dispersed, bland tumor cells with
plasmacytoid morphology, oval nuclei, and salt and pepper chromatin
(Papanicolaou stain, 400 magnification)

76

Endocr Pathol (2014) 25:6579

required for calculation. Additionally, since NETs are fraught

with proliferative heterogeneity, is the index/grade generated by
an FNA (or even a core biopsy) reliable? Few have adequately
addressed this question, but there are studies that support grading of these tumors on biopsy-generated samples despite
intratumoral heterogeneity [58]. In a recent cytologic study,
grading results obtained by manual counting of ki-67-positive
cells correlated positively with indices obtained by automated
image cytometer, as well as corresponding resections [56].
These findings suggest that the grading of PanNETs on cytologic material is applicable, and in situations in which such
samples are the only ones available for evaluation (particularly
in patients with metastatic disease), they may present the only
opportunity for classification and treatment stratification.

months, 5-year survival 22 %) than G2 NETs (median survival

63 months, 5-year survival 61 %) but less aggressive than
PDNEC (median survival 15 months, 5-year survival 17 %)
[10]. Similar differences in survival have been shown by
others [10, 11, 59, 60]. This was also confirmed by a study
from Europe which showed that for all gastrointestinal G3
NECs survival times were longer when the ki-67 index was
2055 versus over 55 %, and the latter group was also more
responsive to platinum-based therapy [11]. Accordingly, we
currently regard G3 NETs as two distinct groups: (1) well
moderately differentiated PanNET with proliferative index of
G3, which typically has ki-67 >20 and <50 %, and (2) poorly
differentiated neuroendocrine carcinoma, which typically has
ki-67 >50 %.

Biologic Behavior

Treatment Options

Ordinary PanNETs (G1 and G2) are considered low-grade

malignancies (with the exception of tumorlets described earlier), with an overall 10-year survival of 6070 %. In the
previous (WHO-2004) classification system as well as earlier
studies, an attempt was made to recognize a benign behavior category. However, after more recent studies with longer
clinical follow-up, it became clear that even the incidentally
detected and resectable PanNETs that are <2 cm develop
recurrences and metastasis in 715 % of cases on long-term
follow-up. The cases that are low stage (T1N0M0) and low
grade (G1) have a very good prognosis with an estimated 10year survival of >95 %.
The prognosis of the G3 category is more complicated,
largely due to the fact that this category encompasses at least
two distinct entities: ordinary PanNETs with high proliferative
rate and true PDNECs (Fig. 11). Preliminary results from
Basturk et al. indicate that G3 cases with ordinary PanNET
morphology are clearly more aggressive (median survival 32

It is generally agreed among experts that, if possible, all

functioning PanNETs should be resected [61, 62], with the
exception of patients with MEN1 and ZollingerEllison syndrome or those with small (<2 cm) non-functioning PanNETs.
The role of surgery in MEN1 remains controversial because
patients often have multifocal microscopic disease requiring
extensive resections.
For incidental PanNETs that are <1 cm, even watchful
waiting has been proposed, but this is controversial, considering that even these can recur or metastasize, albeit the incidence is low (<5 %). However, these data are based on
resected cases, and it is not known how high the incidence
of progression would be for unresected cases. For PanNETs
<2 cm, surgical enucleation, if feasible, often appears curative,
but since this procedure is relatively recent, more extended
follow-up is required to confirm this [61, 62]. For larger
tumors, resection is indicated, if clinically feasible. The treatment options for those with liver metastasis include hepatic

Fig. 11 Well-differentiated pancreatic neuroendocrine tumor (grade 3 by ki67 index) in contrast to the poorly differentiated neuroendocrine carcinoma
in Fig. 8. a By mitotic count (2/10 high power fields), this tumor was

classified as grade 1 per 2010 WHO grading scheme (hematoxylin & eosin
stain, 400 magnification). b The corresponding ki-67 stain yielded a proliferation index of 25 %, consistent with a grade 3 tumor (400 magnification)

Endocr Pathol (2014) 25:6579

resection (cryosurgery) and transplantation, and hepatic artery

embolization or radiofrequency ablation for non-surgical
candidates.
The role of chemotherapy (or targeted therapy) has been
debatable. Since PanNETs are slow-growing tumors, they
often do not respond to conventional cytotoxic chemotherapeutic agents. Recently, targeted therapies have become a
serious alternative consideration. For tumors with alterations
of the mTOR pathway, the drug everolimus has shown promise as a therapeutic agent [63] and is currently in phase IV
clinical trials (http://www.nanets.net/research/current-clinicaltrials). Cabozantinib, a tyrosine kinase inhibitor, and sunitinib,
a vascular endothelial growth factor inhibitor, have also
shown similar promise (http://www.nanets.net/research/
current-clinical-trials).
For patients with PDNECs, cisplatin and etoposide-based
regimens are recommended and patients often show 4070 %
(albeit brief) response rate [61]. The recent demonstration of
BCL2 overexpression by PDNECs [43] suggests that BCL2
antagonists may prove useful in their treatment in a manner
similar to their current use in small cell carcinoma of lung,
which also expresses BCL2.

Summary of Key Elements to Include in a Pathology

Report
Resection of a Primary Pancreatic Neuroendocrine Tumor
Diagnosis and differentiation: well-differentiated PanNET
versus poorly differentiated PanNEC
Tumor grade: based on mitoses/10 HPF and ki-67 proliferation index (%)
Tumor size
Lymphvascular invasion
Perineural invasion
Large vessel invasion
Margin status (for resections)
TNM stage (if a resection); specify method used (CAP/
AJCC vs ENETS)
Comment: Tumor grade, mitotic count/10 HPF, ki-67 index
(%), necrosis, neuroendocrine marker expression, cytokeratin
19 expression (optional)
Resection or Biopsy of Metastatic Tumor (Presumed)
from Pancreas
Diagnosis and differentiation: well-differentiated PanNET
versus poorly differentiated PanNEC
Margin status: for resections
Comment: tumor grade [based on mitoses/10 HPF, ki67 proliferation index (%)], additional (optional) stains
(PDX-1, isl1)

77

References
1. Klimstra D, Arnold R, Capella C, Hruban R, Kloppel G, Komminoth
P, Solcia E, Rindi G (2010) Neuroendocrine neoplasms of the pancreas. In: Bosman F, Carneiro F, Hruban R, Theise N (eds) WHO
classification of tumours of the digestive system, 3rd edn. IARC,
Lyon, pp 322-326
2. Vagefi PA, Razo O, Deshpande V, McGrath DJ, Lauwers GY, Thayer
SP, Warshaw AL, Fernandez-Del Castillo C (2007) Evolving patterns
in the detection and outcomes of pancreatic neuroendocrine neoplasms:
the Massachusetts General Hospital experience from 1977 to 2005.
Arch Surg 142 (4):347-354. doi:10.1001/archsurg.142.4.347
3. Anlauf M, Schlenger R, Perren A, Bauersfeld J, Koch CA, Dralle H,
Raffel A, Knoefel WT, Weihe E, Ruszniewski P, Couvelard A,
Komminoth P, Heitz PU, Kloppel G (2006) Microadenomatosis of
the endocrine pancreas in patients with and without the multiple
endocrine neoplasia type 1 syndrome. Am J Surg Pathol 30 (5):
560-574. doi:10.1097/01.pas.0000194044.01104.25
4. Lubensky IA, Pack S, Ault D, Vortmeyer AO, Libutti SK, Choyke
PL, Walther MM, Linehan WM, Zhuang Z (1998) Multiple neuroendocrine tumors of the pancreas in von Hippel-Lindau disease
patients: histopathological and molecular genetic analysis. Am J
Pathol 153 (1):223-231. doi:10.1016/S0002-9440(10)65563-0
5. Hoang MP, Hruban RH, Albores-Saavedra J (2001) Clear cell endocrine pancreatic tumor mimicking renal cell carcinoma: a distinctive
neoplasm of von HippelLindau disease. Am J Surg Pathol 25 (5):
602-609
6. Fryer E, Serra S, Chetty R (2012) Lipid-rich (clear cell) neuroendocrine tumors of the pancreas in MEN I patients. Endocr Pathol 23
(4):243-246. doi:10.1007/s12022-012-9221-z
7. Whipple AO, Frantz VK (1935) Adenoma of islet cells with hyperinsulinism: a review. Ann Surg 101 (6):1299-1335
8. Shi C, Siegelman SS, Kawamoto S, Wolfgang CL, Schulick RD,
Maitra A, Hruban RH (2010) Pancreatic duct stenosis secondary to
small endocrine neoplasms: a manifestation of serotonin production?
Radiology 257 (1):107-114. doi:10.1148/radiol.10100046
9. Kenney B, Singh G, Salem RR, Paterno F, Robert ME, Jain D (2011)
Pseudointraductal papillary mucinous neoplasia caused by microscopic
periductal endocrine tumors of the pancreas: a report of 3 cases. Hum
Pathol 42 (7):1034-1041. doi:10.1016/j.humpath.2010.09.018
10. Basturk O, Yang Z, Tang LH, Hruban R, McCall C, Adsay V,
Krasinskas A, Jang KT, Bellizzi A, Shi C, Klimstra DS (2013)
Increased (> 20 %) Ki67 proliferation index in morphologically well
differentiated pancreatic neuroendocrine tumors (PanNETs) correlates with decreased overall survival (abstract). Modern Pathology
26:423A-423A
11. Sorbye H, Welin S, Langer SW, Vestermark LW, Holt N, Osterlund P,
Dueland S, Hofsli E, Guren MG, Ohrling K, Birkemeyer E, ThiisEvensen E, Biagini M, Gronbaek H, Soveri LM, Olsen IH, Federspiel
B, Assmus J, Janson ET, Knigge U (2013) Predictive and prognostic
factors for treatment and survival in 305 patients with advanced
gastrointestinal neuroendocrine carcinoma (WHO G3): the
NORDIC NEC study. Annals of Oncology: Official Journal of the
European Society for Medical Oncology/ESMO 24 (1):152-160.
doi:10.1093/annonc/mds276
12. Hochwald SN, Zee S, Conlon KC, Colleoni R, Louie O, Brennan
MF, Klimstra DS (2002) Prognostic factors in pancreatic endocrine
neoplasms: an analysis of 136 cases with a proposal for low-grade
and intermediate-grade groups. J Clin Oncol 20 (11):2633-2642
13. Deshpande V, Fernandez-del Castillo C, Muzikansky A, Deshpande
A, Zukerberg L, Warshaw AL, Lauwers GY (2004) Cytokeratin 19 is
a powerful predictor of survival in pancreatic endocrine tumors. Am J
Surg Pathol 28 (9):1145-1153
14. Jain R, Fischer S, Serra S, Chetty R (2010) The use of Cytokeratin 19
(CK19) immunohistochemistry in lesions of the pancreas,

78

15.

16.

17.

18.

19.
20.
21.

22.

23.

24.

25.

26.

27.

28.

Endocr Pathol (2014) 25:6579

gastrointestinal tract, and liver. Appl Immunohistochem Mol
Morphol 18 (1):9-15. doi:10.1097/PAI.0b013e3181ad36ea
Schmitt AM, Anlauf M, Rousson V, Schmid S, Kofler A, Riniker F,
Bauersfeld J, Barghorn A, Probst-Hensch NM, Moch H, Heitz PU,
Kloeppel G, Komminoth P, Perren A (2007) WHO 2004 criteria and
CK19 are reliable prognostic markers in pancreatic endocrine tumors.
Am J Surg Pathol 31 (11):1677-1682. doi:10.1097/PAS.
0b013e31805f675d
Zhang L, Smyrk TC, Oliveira AM, Lohse CM, Zhang S, Johnson
MR, Lloyd RV (2009) KIT is an independent prognostic marker for
pancreatic endocrine tumors: a finding derived from analysis of islet
cell differentiation markers. Am J Surg Pathol 33 (10):1562-1569.
doi:10.1097/PAS.0b013e3181ac675b
Tang LH, Gonen M, Hedvat C, Modlin IM, Klimstra DS (2012)
Objective quantification of the Ki67 proliferative index in neuroendocrine tumors of the gastroenteropancreatic system: a comparison of
digital image analysis with manual methods. Am J Surg Pathol 36
(12):1761-1770. doi:10.1097/PAS.0b013e318263207c
Bagci P, Ohike N, Dursun N, Jang K, Tajin T, Basturk O, Konh SY,
Ducato L, Reid M, Adsay V (2012) Comparative analysis of different
methodologies for ki67 in pancreatic neuroendocrine tumors (abstract). Modern Pathol 25;441A-441A
AJCC (2010) Cancer staging manual, 7th edn edn. Springer, New York
Sobin L, Gospodarowicz M, Wittekind C (2009) UICC: TNM classification of malignant tumors, 7th edn. Wiley-Blackwell, Oxford
CAP (2013) College of american pathologists cancer protocols and
checklists. College of American Pathologists. http://www.cap.org/.
Accessed 04/10/2013
Rindi G, Kloppel G, Alhman H, Caplin M, Couvelard A, de Herder
WW, Erikssson B, Falchetti A, Falconi M, Komminoth P, Korner M,
Lopes JM, McNicol AM, Nilsson O, Perren A, Scarpa A, Scoazec JY,
Wiedenmann B (2006) TNM staging of foregut (neuro)endocrine
tumors: a consensus proposal including a grading system. Virchows
Arch 449 (4):395-401. doi:10.1007/s00428-006-0250-1
Adsay NV, Bagci P, Tajiri T, Oliva I, Ohike N, Balci S, Gonzalez RS,
Basturk O, Jang KT, Roa JC (2012) Pathologic staging of pancreatic,
ampullary, biliary, and gallbladder cancers: pitfalls and practical
limitations of the current AJCC/UICC TNM staging system and
opportunities for improvement. Seminars in Diagnostic Pathology
29 (3):127-141. doi:10.1053/j.semdp.2012.08.010
Saka B, Oliva I, Bandyopadhyay S, Basturk O, Balci S, Maithel S,
Kooby D, Sarmiento J, Staley C, El-Rayes B, Choi H, Knight J,
Krasinskas A, Adsay V (2014) Will the pT1 and pT2 pancreas cancer
please stand up? Peripancreatic soft tissue is involved in most
pancreatic ductal adenocarcinomas (PDAC), negating its value as a
staging parameter and necessitating a new staging scheme (abstract).
Modern Pathology 27(Supplement), in press
Saka B, Balci S, Bagci P, Maithel S, Kooby D, Sarmiento J, Staley C,
El-Rayes B, Choi H, Knight J, Goodman M, Krasinskas A, Adsay V
(2014) Proposal for a new and prognostically valuable tumor size
based T-stage for pancreatic adenocarcinoma (abstract). Modern
Pathology 27(Supplement), in press
Rindi G, Falconi M, Klersy C, Albarello L, Boninsegna L, Buchler
MW, Capella C, Caplin M, Couvelard A, Doglioni C, Delle Fave G,
Fischer L, Fusai G, de Herder WW, Jann H, Komminoth P, de Krijger
RR, La Rosa S, Luong TV, Pape U, Perren A, Ruszniewski P, Scarpa A,
Schmitt A, Solcia E, Wiedenmann B (2012) TNM staging of neoplasms
of the endocrine pancreas: results from a large international cohort study.
J Natl Cancer Inst 104 (10):764-777. doi:10.1093/jnci/djs208
Rindi G, de Herder WW, O'Toole D, Wiedenmann B (2006)
Consensus guidelines for the management of patients with digestive
neuroendocrine tumors: why such guidelines and how we went about
it. Neuroendocrinology 84 (3):155-157. doi:10.1159/000098006
Liszka L, Pajak J, Mrowiec S, Zielinska-Pajak E, Golka D, Lampe P
(2011) Discrepancies between two alternative staging systems
(European Neuroendocrine Tumor Society 2006 and American Joint

29.

30.
31.

32.
33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

Committee on Cancer/Union for International Cancer Control 2010) of

neuroendocrine neoplasms of the pancreas. A study of 50 cases. Pathol
Res Pract 207 (4):220-224. doi:10.1016/j.prp.2011.01.008
Anlauf M, Perren A, Kloppel G (2007) Endocrine precursor lesions
and microadenomas of the duodenum and pancreas with and without
MEN1: criteria, molecular concepts and clinical significance.
Pathobiology 74 (5):279-284. doi:10.1159/000105810
Laidlaw GF (1938) Nesidioblastoma, the islet tumor of the pancreas.
Am J Pathol 14 (2):125-134 125
Vance JE, Stoll RW, Kitabchi AE, Williams RH, Wood FC, Jr. (1969)
Nesidioblastosis in familial endocrine adenomatosis. JAMA 207 (9):
1679-1682
Heitz PU, Kasper M, Polak JM, Kloppel G (1979) Pathology of the
endocrine pancreas. J Histochem Cytochem 27 (10):1401-1402
Kloppel G, Anlauf M, Raffel A, Perren A, Knoefel WT (2008) Adult
diffuse nesidioblastosis: genetically or environmentally induced?
Hum Pathol 39 (1):3-8. doi:10.1016/j.humpath.2007.09.010
Perez-Montiel MD, Frankel WL, Suster S (2003) Neuroendocrine
carcinomas of the pancreas with 'Rhabdoid' features. Am J Surg
Pathol 27 (5):642-649
Zee SY, Hochwald SN, Conlon KC, Brennan MF, Klimstra DS
(2005) Pleomorphic pancreatic endocrine neoplasms: a variant commonly confused with adenocarcinoma. Am J Surg Pathol 29 (9):
1194-1200
Garbrecht N, Anlauf M, Schmitt A, Henopp T, Sipos B, Raffel A,
Eisenberger CF, Knoefel WT, Pavel M, Fottner C, Musholt TJ, Rinke
A, Arnold R, Berndt U, Plockinger U, Wiedenmann B, Moch H,
Heitz PU, Komminoth P, Perren A, Kloppel G (2008) Somatostatinproducing neuroendocrine tumors of the duodenum and pancreas:
incidence, types, biological behavior, association with inherited syndromes, and functional activity. Endocr Relat Cancer 15 (1):229-241.
doi:10.1677/ERC-07-0157
McCall CM, Shi C, Klein AP, Konukiewitz B, Edil BH, Ellison TA,
Wolfgang CL, Schulick RD, Kloppel G, Hruban RH (2012)
Serotonin expression in pancreatic neuroendocrine tumors correlates
with a trabecular histologic pattern and large duct involvement. Hum
Pathol 43 (8):1169-1176. doi:10.1016/j.humpath.2011.09.014
Basturk O, Hruban R, Adsay V, Krasinskas A, Jang KT, Liu X,
Vakiani E, Zhang L, Frankel W, Giordano T, Bellizzi A, Chen JH,
Deshpande V, Yang Z, Klimstra DS (2013) High-grade neuroendocrine carcinomas of the pancreas: a clinicopathologic analysis of 60
cases (abstract). Modern Pathology 26:422A-423A
Jiao Y, Shi C, Edil BH, de Wilde RF, Klimstra DS, Maitra A,
Schulick RD, Tang LH, Wolfgang CL, Choti MA, Velculescu VE,
Diaz LA, Jr., Vogelstein B, Kinzler KW, Hruban RH, Papadopoulos
N (2011) DAXX/ATRX, MEN1, and mTOR pathway genes are
frequently altered in pancreatic neuroendocrine tumors. Science
331 (6021):1199-1203. doi:10.1126/science.1200609
Agaimy A, Erlenbach-Wunsch K, Konukiewitz B, Schmitt AM,
Rieker RJ, Vieth M, Kiesewetter F, Hartmann A, Zamboni G,
Perren A, Kloppel G (2013) ISL1 expression is not restricted to
pancreatic well-differentiated neuroendocrine neoplasms, but is also
commonly found in well and poorly differentiated neuroendocrine
neoplasms of extrapancreatic origin. Mod Pathol 26 (7):995-1003.
doi:10.1038/modpathol.2013.40
Graham RP, Shrestha B, Caron BL, Smyrk TC, Grogg KL, Lloyd RV,
Zhang L (2013) Islet-1 is a sensitive but not entirely specific marker
for pancreatic neuroendocrine neoplasms and their metastases. Am J
Surg Pathol 37 (3):399-405. doi:10.1097/PAS.0b013e31826f042c
Chan ES, Alexander J, Swanson PE, Jain D, Yeh MM (2012) PDX-1,
CDX-2, TTF-1, and CK7: a reliable immunohistochemical panel for
pancreatic neuroendocrine neoplasms. Am J Surg Pathol 36 (5):737743. doi:10.1097/PAS.0b013e31824aba59
Yachida S, Vakiani E, White CM, Zhong Y, Saunders T, Morgan R,
de Wilde RF, Maitra A, Hicks J, Demarzo AM, Shi C, Sharma R,
Laheru D, Edil BH, Wolfgang CL, Schulick RD, Hruban RH, Tang

Endocr Pathol (2014) 25:6579

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

LH, Klimstra DS, Iacobuzio-Donahue CA (2012) Small cell and

large cell neuroendocrine carcinomas of the pancreas are genetically
similar and distinct from well-differentiated pancreatic neuroendocrine tumors. Am J Surg Pathol 36 (2):173-184. doi:10.1097/PAS.
0b013e3182417d36
Gortz B, Roth J, Krahenmann A, de Krijger RR, Muletta-Feurer
S, Rutimann K, Saremaslani P, Speel EJ, Heitz PU, Komminoth
P (1999) Mutations and allelic deletions of the MEN1 gene are
associated with a subset of sporadic endocrine pancreatic and
neuroendocrine tumors and not restricted to foregut neoplasms.
Am J Pathol 154 (2):429-436. doi:10.1016/S0002-9440(10)
65289-3
Hessman O, Lindberg D, Skogseid B, Carling T, Hellman P, Rastad J,
Akerstrom G, Westin G (1998) Mutation of the multiple endocrine
neoplasia type 1 gene in nonfamilial, malignant tumors of the endocrine pancreas. Cancer Res 58 (3):377-379
Toliat MR, Berger W, Ropers HH, Neuhaus P, Wiedenmann B (1997)
Mutations in the MEN I gene in sporadic neuroendocrine tumours of
gastroenteropancreatic system. Lancet 350 (9086):1223. doi:10.
1016/S0140-6736(05)63453-8
Wang EH, Ebrahimi SA, Wu AY, Kashefi C, Passaro E, Jr., Sawicki
MP (1998) Mutation of the MENIN gene in sporadic pancreatic
endocrine tumors. Cancer Res 58 (19):4417-4420
Heaphy CM, de Wilde RF, Jiao Y, Klein AP, Edil BH, Shi C,
Bettegowda C, Rodriguez FJ, Eberhart CG, Hebbar S, Offerhaus
GJ, McLendon R, Rasheed BA, He Y, Yan H, Bigner DD, ObaShinjo SM, Marie SK, Riggins GJ, Kinzler KW, Vogelstein B,
Hruban RH, Maitra A, Papadopoulos N, Meeker AK (2011)
Altered telomeres in tumors with ATRX and DAXX mutations.
Science 333 (6041):425. doi:10.1126/science.1207313
Yao JC, Shah MH, Ito T, Bohas CL, Wolin EM, Van Cutsem E,
Hobday TJ, Okusaka T, Capdevila J, de Vries EG, Tomassetti P,
Pavel ME, Hoosen S, Haas T, Lincy J, Lebwohl D, Oberg K (2011)
Everolimus for advanced pancreatic neuroendocrine tumors. N Engl
J Med 364 (6):514-523. doi:10.1056/NEJMoa1009290
Wood L, Adsay N, Hruban R (2013) Molecular pathology of pancreatic cancer. In: Cheng L, Eble J (eds) Molecular surgical pathology. Springer, New York, pp 17-42
Meriden Z, Shi C, Edil BH, Ellison T, Wolfgang CL, Cornish TC,
Schulick RD, Hruban RH (2011) Hyaline globules in neuroendocrine
and solid-pseudopapillary neoplasms of the pancreas: a clue to the
diagnosis. Am J Surg Pathol 35 (7):981-988. doi:10.1097/PAS.
0b013e31821a9a14
Adsay NV, Andea A, Basturk O, Kilinc N, Nassat H, Cheng JD
(2004) Secondary tumors of the pancreas: ana analysis of a surgical
and autopsy database and review of the literature. Virchows Archiv
444(6):527-535
Bismar TA, Basturk O, Gerald WL, Schwarz K, Adsay NV (2004)
Desmoplastic small cell tumor in the pancreas. Am J Surg Pathol 28
(6):808-812

79
54. Movahedi-Lankarani S, Hruban RH, Westra WH, Klimstra DS
(2002) Primitive neuroectodermal tumors of the pancreas: a report
of seven cases of a rare neoplasm. Am J Surg Pathol 26 (8):10401047
55. Bachmann J, Kleeff J, Bergmann F, Shrikhande SV, Hartschuh W,
Buchler MW, Friess H (2005) Pancreatic metastasis of Merkel cell
carcinoma and concomitant insulinoma: case report and literature
review. World J Surg Oncol 3:58. doi:10.1186/1477-7819-3-58
56. Fung AD, Cohen C, Kavuri S, Lawson D, Gao X, Reid MD (2013)
Phosphohistone h3 and ki-67 labeling indices in cytologic specimens
from well-differentiated neuroendocrine tumors of the gastrointestinal tract and pancreas: a comparative analysis using automated image
cytometry. Acta Cytol 57 (5):501-508. doi:10.1159/000351475
57. Larghi A, Capurso G, Carnuccio A, Ricci R, Alfieri S, Galasso D,
Lugli F, Bianchi A, Panzuto F, De Marinis L, Falconi M, Delle Fave
G, Doglietto GB, Costamagna G, Rindi G (2012) Ki-67 grading of
nonfunctioning pancreatic neuroendocrine tumors on histologic samples obtained by EUS-guided fine-needle tissue acquisition: a prospective study. Gastrointest Endosc 76 (3):570-577. doi:10.1016/j.
gie.2012.04.477
58. Yang Z, Tang LH, Klimstra DS (2011) Effect of tumor heterogeneity
on the assessment of Ki67 labeling index in well-differentiated neuroendocrine tumors metastatic to the liver: implications for prognostic stratification. Am J Surg Pathol 35 (6):853-860. doi:10.1097/PAS.
0b013e31821a0696
59. Velayoudom-Cephise FL, Duvillard P, Foucan L, Hadoux, J,
Chougnet CN, Leboulleux, S, Malka D, Guigay J, Goere D,
Debaere T, Caramella C, Schlumberger M, Planchard D, Elias D,
Ducreux M, Scoazec JY, Baudin E (2013) Are G3 ENETS neuroendocrine neoplasms heterogeneous? Endocr Relat Cancer 20 (5):649657. doi:10.1530/ERC-13-0027
60. McCall CM, Shi C, Cornish TC, Klimstra DS, Tang LH, Basturk O,
Mun LJ, Ellison TA, Wolfgang CL, Choti MA, Schulick RD, Edil
BH, Hruban RH (2013) Grading of well-differentiated pancreatic
neuroendocrine tumors is improved by the inclusion of both Ki67
proliferative index and mitotic rate. Am J Surg Pathol 37 (11):16711677. doi:10.1097/PAS.0000000000000089
61. Kulke MH, Anthony LB, Bushnell DL, de Herder WW, Goldsmith
SJ, Klimstra DS, Marx SJ, Pasieka JL, Pommier RF, Yao JC, Jensen
RT (2010) NANETS treatment guidelines: well-differentiated neuroendocrine tumors of the stomach and pancreas. Pancreas 39 (6):735752. doi:10.1097/MPA.0b013e3181ebb168
62. Metz DC, Jensen RT (2008) Gastrointestinal neuroendocrine tumors:
pancreatic endocrine tumors. Gastroenterology 135 (5):1469-1492.
doi:10.1053/j.gastro.2008.05.047
63. Yao JC, Phan AT, Chang DZ, Wolff RA, Hess K, Gupta S, Jacobs C,
Mares JE, Landgraf AN, Rashid A, Meric-Bernstam F (2008) Efficacy
of RAD001 (everolimus) and octreotide LAR in advanced low- to
intermediate-grade neuroendocrine tumors: results of a phase II study. J
Clin Oncol 26 (26):4311-4318. doi:10.1200/JCO.2008.16.7858