The American Journal of Surgical Pathology 25(8): 1047–1053, 2001
, Philadelphia
Familial Fibrocystic Pancreatic Atrophy With
Endocrine Cell Hyperplasia and
Pancreatic Carcinoma
Kenneth A. Meckler, M.D., Teresa A. Brentnall, M.D.,
Rodger C. Haggitt, M.D., David Crispin, B.A., David R. Byrd, M.D.,
Michael B. Kimmey, M.D., and Mary P. Bronner, M.D.
Understanding the pathology of familial pancreatic carcinoma
may provide important insights into pancreatic tumorigenesis.
We now describe in detail the pancreatic pathology of an au-
tosomal dominant pancreatic carcinoma kindred with distinct
clinical, genetic, and pathologic manifestations differing from
all other reported forms of sporadic or familial pancreatic neo-
plasia. Affected individuals develop a prodrome of diabetes
mellitus, pancreatic exocrine insufficiency, and characteristic
pancreatic imaging abnormalities. Eleven family members have
undergone total pancreatectomy, revealing a unique and char-
acteristic fibrocystic, lobulocentric pancreatic atrophy. This
was patchy to diffuse in distribution and was invariably asso-
ciated with a nesidioblastosis-like endocrine cell hyperplasia.
All but one resected pancreas demonstrated glandular epithelial
dysplasia: 10 had low-grade dysplasia (pancreatic intraductal
neoplasia grade II of III or PanIN II) and seven also had high-
grade dysplasia (pancreatic intraductal neoplasia grade III of III
or PanIN III). Dysplasia was multifocal in small- to medium-
sized duct-like structures within areas of acinar atrophy, mi-
crocystic change, and mucinous hyperplasia. Two pancreata
had carcinomas of multiple and unusual histologic subtypes,
including small cell undifferentiated carcinoma and giant cell
anaplastic carcinoma. The findings in this kindred yield impor-
tant information on a distinctive and previously unrecognized
pancreatic cancer precursor. Recognition of this entity may
help identify additional kindreds and perhaps the underlying
genetic defect. As is the case for other familial cancers, the as
yet unknown specific genetic defect may have wider implica-
tions for pancreatic cancer in general.
Key Words: Pancreatic carcinoma—Familial—Fibrocystic
pancreatic atrophy—Pancreatic dysplasia—PanIN—Endocrine
cell hyperplasia—Nesidioblastosis.
Am J Surg Pathol 25(8): 1047–1053, 2001.
From the Departments of Pathology (K.A.M., R.C.H., D.C., M.P.B.),
Gastroenterology (T.A.B., R.C.H., M.B.K., M.P.B.), and Surgery
(D.R.B.), University of Washington Medical Center, Seattle,
Washington, U.S.A.
This work is dedicated to the memory of our esteemed colleague, Dr.
Rodger C. Haggitt, without whose contributions, this study would not
have been possible.
Address correspondence and reprint requests to Mary P. Bronner,
MD, University of Washington Medical Center, Department of Pathol-
ogy, Room BB-220, 1959 NE Pacific Street, Box 356100, Seattle, WA
98195-6100, U.S.A.; e-mail: bronner@u.washington.edu
1047
© 2001 Lippincott Williams & Wilkins, Inc.
Pancreatic cancer ranks fourth as a cause of cancer
death in the United States, and approximately 28,200
Americans will have died from it in the year 2000.11,13
The majority of pancreatic carcinomas appear to be spo-
radic, and cigarette smoking and chronic pancreatitis are
important risk factors.13,18 Despite its prevalence, in-
creasing incidence, and recognition of risk factors, we
know little about the early events in pancreatic tumori-
genesis. This remains a major obstacle in the fight
against this disease. Its high mortality principally results
from a lack of effective screening modalities to detect
early, curable disease. Most pancreatic carcinomas are
discovered only after they are pathologically advanced
and much of the precursor lesion(s) has been destroyed.
These combined factors have made early diagnosis, in-
tervention, and study of early pancreatic tumorigenesis
difficult.
The histology of familial pancreatic carcinoma, not to
mention its precursor lesions, has also not been well
characterized.19 To begin to address this deficiency, the
detailed pathology is now presented from one of the
largest pancreatic carcinoma kindreds described to date.
The clinical genetic aspects6 and the clinical manage-
ment of the family3 have been previously reported and
are summarized below, but the kindred’s distinctive pan-
creatic pathology has not been heretofore fully described.
The combined clinical, genetic, and pathologic features
of this kindred are distinctive and unique in comparison
with other reported forms of familial and sporadic pan-
creatic neoplasia.
Pancreatic disease in this kindred is inherited as an
autosomal dominant trait with high penetrance (Fig. 1),
but the causative gene for this kindred’s disorder has not
yet been identified. The literature describes a few pan-
creatic carcinoma kindred. Most are associated with a
recognized genetic syndrome, such as the following: 1)
hereditary pancreatitis related to mutations in the cat-
ionic trypsinogen gene; 2) hereditary nonpolyposis colo-
1048 K. A. MECKLER ET AL.
FIG. 1. Pedigree of family X with 62 individuals spanning five generations. This demonstrates the autosomal-dominant
inheritance of familial fibrocystic pancreatic atrophy with endocrine cell hyperplasia and ductal dysplasia and carcinoma.
rectal cancer (HNPCC) related to mutations in one of the
mismatch repair genes, most commonly the MLH-1 or
MSH-2 genes; or 3) familial atypical mole-malignant
melanoma often related to p16 mutations, a cell cycle
control gene.14,19,27 The current kindred has none of the
phenotypes associated with these syndromes. Further-
more, no cytogenetic alterations have been observed, and
molecular genetic abnormalities known to be important
in either sporadic or familial pancreatic cancer have not
been detected in the current family’s germline DNA,3,6
including alterations in K-ras,5 the cationic trypsinogen
gene,27 the mismatch repair genes, or the p16, p53,
DPC4, or APC genes.14,19 Finally, genotyping and link-
age analysis (report in preparation) indicate that the lo-
cation of the gene is not consistent with involvement by
BRCA-1 or -2.
Affected family members characteristically develop a
clinical prodrome of diabetes mellitus and exocrine pan-
creatic insufficiency, without overt clinical pancreati-
tis.3,6 This clinical prodrome has developed over a broad
time span in the affected family members, ranging from
weeks to decades, before the diagnosis of pancreatic car-
cinoma or dysplasia. Because the onset of clinical symp-
toms does not necessarily predict whether the patient has
developed pancreatic neoplasia, we have developed a
Am J Surg Pathol, Vol. 25, No. 8, 2001
surveillance strategy, which has been previously de-
scribed. The surveillance program uses characteristic en-
doscopic retrograde cholangiopancreatography (ERCP)
(Fig. 2A) or endoscopic ultrasound (EUS) abnormalities,
or both, to help identify affected family members who
have developed pancreatic neoplasia and for whom total
pancreatectomy is indicated.3 The distinctive pathology
of the pancreatic cancer precursor lesions and the carci-
nomas in this kindred forms the basis of this report.
MATERIALS AND METHODS
The pedigree of the 62-member, five-generation fam-
ily is shown in Figure 1. University of Washington
Human Subject’s Review Board approval and informed
patient consent were obtained for genetic testing per-
formed in this study. Based on their clinical prodrome,
11 members were determined to be carriers of the as yet
unknown mutant allele and underwent total pancreatec-
tomy.3,6 All pancreata were submitted in their entirety
for examination. The tissues were fixed in 10% neutral
buffered formalin, routinely processed into paraffin
blocks, sectioned, and stained with hematoxylin and eo-
sin. All sections were systematically evaluated for pan-
creatitis, patterns of fibrosis, and epithelial and endocrine
 FAMILIAL FIBROCYSTIC PANCREATIC ATROPHY
FIG. 2. (A) Pancreatogram from patient no. II.9 reveals
an irregular main pancreatic duct (arrow) and cystically
dilated side branches (arrowheads). (B) Gross total pan-
createctomy specimen from patient no. III.19 showing ir-
regular nodularity of the pancreatic parenchyma and the
attached loop of duodenum arching around the pancreatic
head. (C) Low-power view of the distinctive lobulocentric
fibrocystic change with minimal inflammation in patient no.
III.19. Note the patchy acinar atrophy and intralobular mi-
crocyst formation with enveloping fibrosis: hematoxylin
and eosin. (D) Distinctive lobulocentric fibrocystic change
with minimal inflammation in patient no. III.19. Note the
dense collagenization of the lobules, acinar atrophy, and
microcystic glandular change within the lobules: hema-
toxylin and eosin. (E) Pronounced area of intralobular fi-
brocystic change without dysplasia in patient no. IV.31:
hematoxylin and eosin. (F) Mucinous metaplasia occurred
frequently (black arrows) and was always present when
dysplasia was identified. Note the focal LGD developing in
an area of mucinous metaplasia/hyperplasia (long white
arrow). Note also the endocrine cell hyperplasia infiltrating
within and between the glands (short white arrow) in pa-
tient no. III.14: hematoxylin and eosin. (G) Squamous
metaplasia of the intralobular microcysts was occasionally
identified in patient no. III.12: hematoxylin and eosin. (H)
Prominent endocrine cell hyperplasia between and within
the glands was a consistent feature in patient no. III.14:
hematoxylin and eosin.
cell alterations including cystic change, atrophy, meta-
plasia, hyperplasia, dysplasia, and carcinoma.
To better evaluate the endocrine cell proliferation seen
in this material, deparaffinized sections were immuno-
1049
stained with a panel of antibodies to endocrine markers,
including monoclonal antibody to insulin (Dako,
Carpinteria, CA, USA) at a dilution of 1:2000, poly-
clonal antibody to glucagon (Instar, Stillwater, MN,
USA) at a dilution of 1:1000, monoclonal antibody to
somatostatin (Dako, Carpinteria, CA, USA) at a dilution
of 1:1000, polyclonal antibody to pancreatic polypeptide
(Chemicon, Temecula, CA, USA) at a dilution of 1:200,
and polyclonal antibody to gastrin (Dako, Carpinteria,
CA, USA) at a dilution of 1:2000. Negative controls
consisted of substitutions of mouse ascites fluid for the
primary antibody. Sections were subjected to heat-
induced epitope retrieval using a microwave oven, as
previously described.10 Antigens were localized using a
FIG. 3. (A) LGD (also known as PanIN II or atypical duc-
tal hyperplasia) arising in a duct with mucinous hyperpla-
sia. Within the focus of LGD note the nuclear enlarge-
ment, stratification, hyperchromatism, and preserved
nuclear polarity, whereby the long axis of the nuclei re-
mains perpendicular to the basement membrane in pa-
tient no. III.14: hematoxylin and eosin. (B) HGD (also
known as PanIN III or ductal carcinoma in situ). Note the
architectural complexity of the intraluminal dysplastic epi-
thelium. The higher magnification inset in the upper left-
hand corner better shows the marked nuclear pleomor-
phism and the loss of nuclear polarity, whereby the long
axis of the nuclei is no longer perpendicular to the base-
ment membrane nor is there any regular orientation of the
nuclei to one another in patient no. III.17: hematoxylin and
eosin. (C) Invasive pancreatic ductal adenocarcinoma,
not otherwise specified with irregular, angulated, and in-
filtrative glands in a desmoplastic stroma in patient no.
II.9: hematoxylin and eosin. (D) Giant cell anaplastic car-
cinoma with sarcoma-like spindle cell and multinucleated
giant cell change in patient no. II.9: hematoxylin and eo-
sin. (E) Small cell undifferentiated carcinoma with a
stippled chromatin pattern, nuclear molding, and incon-
spicuous nucleoli in patient no. II.9: hematoxylin and eosin.
Am J Surg Pathol, Vol. 25, No. 8, 2001
1050 K. A. MECKLER ET AL.
standard avidin-biotin method with 3,3⬘-diamino-
benzidine as the chromogen and nickel chloride enhance-
ment, as previously described.10 The degree of staining
was assessed by two pathologists (M.P.B. and K.A.M.).
Immunoreactivity was assessed as absent, weak or focal,
moderate, or strong and diffuse.
RESULTS
Clinical
The clinical details of the kindred are presented in the
pedigree (Fig. 1). Indicators of disease include onset of
diabetes mellitus, exocrine pancreatic insufficiency with
steatorrhea, absence of overt clinical pancreatitis, and
characteristic ERCP or EUS findings, or both. These
included parenchymal cystic change and irregularities of
the pancreatic ductal system (Fig. 2A). Eleven affected
family members were identified by these clinical char-
acteristics who underwent total pancreatectomy (Figs. 1
and 2B). The resections were performed before the de-
velopment of carcinoma but after the development of
precursor dysplasia, also termed pancreatic intraductal
neoplasia (PanIN) grades II and III (of III) in eight of the
patients.3 Three patients were referred before the devel-
opment of our surveillance protocol: two patients already
had cancer at the time of referral, and the third patient
underwent a prophylactic pancreatectomy because she
developed prodromal insulin-dependent diabetes. At re-
section this latter patient had fibrocystic pancreatic atro-
phy without dysplasia or carcinoma.3,6
Of the remaining family, only three additional mem-
bers have undergone pathologic tissue examination. Au-
topsy material was available for review on one patient
(patient no. II.1, Fig. 1). The pancreatic sections from
this autopsy showed invasive pancreatic ductal adeno-
carcinoma, in addition to the family’s characteristic pre-
cursor pathology, namely, the fibrocystic pancreatic at-
rophy and endocrine cell hyperplasia. This patient’s
cancer extensively invaded into the spleen and was meta-
static to multiple lymph nodes and had invaded into the
patient’s diaphragm at the time of autopsy. The second
family member (patient no. II.11, Fig. 1) also died of
pancreatic adenocarcinoma. His pancreatectomy material
performed more than 30 years earlier is not available for
review; however, by report it showed usual-type ductal
adenocarcinoma, without mention of any additional non-
tumorous pancreatic change. The third family member
(patient no. III.18, Fig. 1) had a wedge hepatic biopsy
that on review showed metastatic pancreatic ductal ad-
enocarcinoma. The adjacent liver in this patient as well
as in the autopsy hepatic sections in patient no. II.1 and
a hepatic biopsy on patient no. III.17 were all normal. No
changes of fibropolycystic hepatic disease were ob-
served. The gallbladder and cystic duct also appeared
Am J Surg Pathol, Vol. 25, No. 8, 2001
normal on the autopsy sections of patient no. II.1, as did
the resected gallbladder from patient nos. III.17 and
III.11 and the bile duct sections on all of the remaining
11 total pancreatectomy resections comprising this re-
port. To our knowledge, no additional preoperative fine-
needle aspirations, biopsies, or other surgical procedures
or autopsies were performed on members of this family.
Gross Pathology
Grossly, all pancreata demonstrated patchy to diffuse
parenchymal fibrosis, prominent grossly visible intra-
ductal calcifications, and parenchymal microcyst forma-
tion, ranging in size from a few millimeters to 1–2 cm
(Fig. 2B). The two pancreata with carcinomas also con-
tained grossly visible masses, measuring approximately
4 cm in one patient and approximately 3.5 cm in the
other. The 3.5-cm lesion was grossly cystic.
Microscopic Pathology
Microscopically, all pancreata demonstrated a charac-
teristic fibrocystic pancreatic atrophy that was distrib-
uted in a patchy to diffuse pattern. Microcystic change
appeared to develop from ectatic intralobular pancreatic
ducts or acini, or both, and was surrounded by concentric
bands of dense fibrous tissue. Inflammatory pancreatitis
was mild to absent (Fig. 2C–E) and, where present, con-
sisted of a mixture of lymphocytes, plasma cells, neutro-
phils, eosinophils, and mast cells, with rare lymphoid
follicles. Occasional intralobular microcysts contained
inspissated and focally calcified mucus. In pancreata
with extensive and diffuse fibrocystic change, the intra-
lobular ducts were frequently markedly ectatic (Fig. 2E).
Mucinous metaplasia of small intralobular ducts (Fig.
2F, black arrows) was invariably present. Squamous
metaplasia (Fig. 2G) was less frequently seen. Acinar
atrophy, which was usually marked, was present in all
cases.
All pancreata demonstrated a nesidioblastosis-like en-
docrine cell hyperplasia with ductocentric endocrine cell
proliferation. In addition to endocrine cell proliferation
within the interstitium, numerous single endocrine cells
and small clusters insinuated themselves between the
ductal epithelial cells and their subjacent basement mem-
branes (Fig. 2F, short white arrow, and Fig. 2H). Ran-
dom endocrine cells demonstrated marked nuclear atyp-
ia. Immunohistochemistry demonstrated an unremark-
able immunophenotype, akin to that observed in normal
pancreatic islets, including a predominance of insulin-
positive cells, along with scattered glucagon, somatostat-
in, and rare pancreatic polypeptide and gastrin-
immunoreactive cells.
Low-grade dysplasia (LGD), which has also been
termed atypical ductal hyperplasia (ADH) or pancreatic
 FAMILIAL FIBROCYSTIC PANCREATIC ATROPHY
intraductal neoplasia grade II (PanIN II),2,15–17 was iden-
tified in 10 of 11 pancreata (Fig. 3A). This lesion was
defined by characteristic cytoarchitectural changes.2,15–17
The architectural changes include mild irregularity of the
intraluminal ductal epithelial profiles with either flat epi-
thelium or short epithelial tufts or papillae extending into
the glandular lumina. Cytologic changes included
nuclear enlargement, hyperchromatism, crowding, and
stratification, without loss of nuclear polarity, such that
the nuclei retained their perpendicular orientation to the
basement membrane.
High-grade dysplasia (HGD), also known as carci-
noma in situ or pancreatic intraductal neoplasia III
(PanIN III),2,15–17 was identified in 7 of 11 pancreata
(Fig. 3B). This lesion was also defined by characteristic
cytoarchitectural changes. 2,15–17 The architectural
changes included more irregular and complex intralumi-
nal ductal epithelial proliferations than described above
for LGD, including long anastomosing papillae, luminal
bridging, and necrotic debris within the lumen. Cytologic
changes included more marked nuclear atypism with
higher nuclear:cytoplasmic ratios than LGD, vesicular or
hyperchromatic chromatin, frequently enlarged and ir-
regular nucleoli, and loss of nuclear polarity, whereby
the nuclei lost their orderly orientation to each other and
no longer remained perpendicular to the basement mem-
brane. Finally, HGD frequently demonstrated a higher
mitotic and apoptotic rate than did LGD.
In all cases ductal dysplasia occurred multifocally
within many separate pancreatic lobules and was seen
only in lobules affected by the above-described fibrocys-
tic pancreatic atrophy. Dysplasia appeared to develop
within the microcystically dilated intralobular glands and
ducts because that is where the most focal and earliest
LGD changes were identified (Fig. 3A). Dysplasia ap-
peared to progress from flat, nonstratified cuboidal epi-
thelium lining the cysts (Fig. 2E), to a flat lining of
hyperplastic columnar cells with basal, single nuclei and
tall mucin caps (Fig. 2F, black arrows, 3A), to low-grade
dysplastic cells (Fig. 2F, long white arrow, 3A). In all
cases with HGD, LGD was also present, frequently in
direct continuity with the HGD. In both cases with in-
vasive carcinoma, HGD was also present in direct con-
tinuity with the invasive tumor.
Four subtypes of invasive carcinoma were present in
two patients, including usual-type ductal adenocarci-
noma not otherwise specified, along with small cell un-
differentiated carcinoma, giant cell anaplastic carcinoma,
and mucinous cystadenocarcinoma (without mesenchy-
mal stroma). One of the two pancreata with grossly vis-
ible masses demonstrated mucinous cystadenocarci-
noma, and the other, a combination of the remaining
three mentioned subtypes (Fig. 3C–E). These latter three
tumors appeared to be separate primaries, as they were
not grossly contiguous within the resected pancreas.
1051
The relationship of patient age to severity and extent
of neoplasia in this family’s pancreatectomy specimens
were assessed. There was no correlation over the decades
(ages 31–57) during which the family members were
diagnosed with varying degrees on dysplasia and/or can-
cer. This may be related to confounding risk factors for
pancreatic disease within the family members, such as
concurrent alcoholism or cigarette smoking, or to the fact
that the inheritance pattern demonstrates the medical ge-
netic phenomenon of anticipation, wherein the age of
disease onset becomes progressively younger with each
passing generation.
DISCUSSION
Limited observations exist regarding the precursor le-
sions of pancreatic carcinoma. This is true regardless of
whether familial or sporadic types are considered. Fur-
thermore, virtually no data exist on the natural history of
these precursor lesions in the pancreas. Although pan-
creatic ductal proliferative lesions have been described in
resected pancreata with sporadic-type pancreatic neopla-
sia,2,14,16,17,20 very few descriptions of familial pancre-
atic carcinoma have appeared in the literature, and es-
sentially nothing is known about their precursor pathol-
ogy.19 To begin to fill this void, we have presented the
detailed and distinctive pathologic findings, inclusive of
the precursor changes, from a large familial pancreatic
neoplasia kindred with more than 60 members and span-
ning five generations. Recognition of this distinctive pa-
thology may permit pathologists to identify additional
families at risk and may ultimately lead to discovery of
the genetic defect and a better understanding of pancre-
atic tumorigenesis in general.
Some clinical risk factors associated with pancreatic
neoplasia are known. Clinically overt chronic pancreati-
tis is well recognized to be associated with the develop-
ment of sporadic-type pancreatic carcinoma.8,18 The cur-
rent family’s clinical prodrome of familial diabetes mel-
litus and pancreatic exocrine insufficiency leading to
pancreatic carcinoma is a distinguishing feature that was
previously unknown.3,6
A few pancreatic carcinoma kindreds have been re-
ported, but the pancreatic pathology of these syndromes
has not been well characterized.19 Most of the reported
kindred to date demonstrate an association with a genetic
syndrome, such as hereditary pancreatitis, ataxia-
telangiectasia, HNPCC, and familial atypical mole-
malignant melanoma (FAMM).14,19,27 The kindred re-
ported herein had none of the syndromic phenotypes of
these conditions, including absence of overt clinical pan-
creatitis with abdominal pain developing at a young age,
as seen in hereditary pancreatitis, absence of the vascular
or neurologic alterations that characterize ataxia-
telangiectasia, absence of colonic or endometrial adeno-
Am J Surg Pathol, Vol. 25, No. 8, 2001
1052 K. A. MECKLER ET AL.
carcinomas or other syndromic tumors seen in HNPCC,
and lack of the melanocytic neoplasms of FAMM. Fur-
ther, no cytogenetic alterations have been observed in the
current kindred’s chromosomes, and molecular genetic
alterations previously reported in familial and sporadic
pancreatic cancer5,14,19,27 have also been excluded from
the family’s germline DNA, including abnormalities in
K-ras, the cationic trypsinogen gene, the mismatch repair
genes, and the p16, p53, APC, or DPC4 genes.3,6 Geno-
typing and linkage analysis work in progress indicate
that the location of the gene is not consistent with in-
volvement of BRCA-1 or -2.
The distinctive clinical prodrome in this family per-
mitted identification of those who had inherited the as
yet unknown mutant allele, and 11 members have under-
gone total pancreatectomy. Early in our care of this fam-
ily and before the surveillance protocol had been estab-
lished, two patients presented with already advanced and
incurable pancreatic carcinoma. However, nine subse-
quent members of the kindred underwent prophylactic
pancreatectomy before the development of malignancy.
This series of 11 patients provided a rare opportunity
to study the early events of this family’s pancreatic
tumorigenesis.
All pancreata studied demonstrated a characteristic
and readily recognizable lobulocentric fibrocystic pan-
creatic atrophy that was usually most prominent in the
pancreatic tails. The cysts appeared to develop from pro-
gressively dilated intralobular pancreatic ducts and/or
acini that communicate with the main pancreatic duct, as
demonstrated on pancreatograms (Fig. 2A). The micro-
cysts were encased in dense fibrous tissue that main-
tained the underlying pancreatic lobular architecture. The
cysts often contained inspissated and focally calcified
mucus, suggesting that they may be the result of chronic
ductal obstruction. Acinar atrophy invariably accompa-
nied the above changes. Interestingly, these fibrocystic
background changes occurred in association with only
mild or no inflammatory pancreatitis.
The peculiar and multiple synchronous varieties of
pancreatic carcinoma identified in two family members
described herein, including cystadenocarcinoma, small
cell undifferentiated carcinoma, anaplastic giant cell car-
cinoma, and usual-type ductal carcinoma, is remarkable.
This pronounced tumor heterogeneity suggests that the
mutator phenotype or mutation rate may be accelerated
in this familial pancreatic cancer syndrome.
Another consistent finding was prominent endocrine
cell hyperplasia, which occurred in a ductocentric pattern
reminiscent of the endocrine cell change in children with
organic hyperinsulinism producing hypoglycemia, a con-
dition termed nesidioblastosis.28 Because our kindred is
composed of adults with diabetes mellitus and hypergly-
cemia, the physiologic opposite of nesidioblastosis, we
have used the term “nesidioblastosis-like” to describe the
Am J Surg Pathol, Vol. 25, No. 8, 2001
ductocentric endocrine cell hyperplasia. The morpho-
logic features of nesidioblastosis have also been ob-
served in patients with Zollinger-Ellison syndrome, cys-
tic fibrosis, chronic pancreatitis, endocrine cell tumors,
and normal infants.9 Clinical nesidioblastosis has been
described in adults, although this appears to be exceed-
ingly rare.7 To our knowledge, nesidioblastosis-like en-
docrine cell proliferation has not been reported in pa-
tients with sporadic pancreatic carcinoma, although
diabetes mellitus and varying degrees of endocrine cell
hyperplasia have been recognized but remain poorly
understood.1,4,21,24
The endocrine cell hyperplasia also occurred in the
context of acinar atrophy in all pancreata studied from
this kindred. It has been demonstrated previously that
small and/or large pancreatic duct obstruction may result
in both acinar atrophy and endocrine cell hyperplasia.25
Therefore, it is possible that duct obstruction may be an
early event in our kindred that subsequently produces
acinar atrophy, lobular microcystic change, and endo-
crine cell hyperplasia. The ductocentric pattern of endo-
crine cell hyperplasia in this family may not be surpris-
ing, as it has been suggested that pancreatic endocrine
cell proliferation occurs via differentiation of putative
pancreatic ductal stem cells into endocrine cells.26
Of particular interest is the compelling experimental
animal data that endocrine cell proliferation enhances
ductal-type carcinogenesis in N-nitrobis(2-oxopropyl)amine
(BOP)-treated hamsters.22,23 In this model the acinar and
ductal tissues are ablated chemically, leaving only endo-
crine cells. When these ablated animals are then given
the carcinogen BOP, the same ductal-type pancreatic ad-
enocarcinoma develops in both the ablated animals with
only endocrine cells as in control animals with intact
pancreata.22 In other studies pure endocrine cell clones
from male hamsters have been transplanted into the sali-
vary glands of female hamsters. The transplanted endo-
crine cells, traceable to the originating male clones by
virtue of their Y chromosomes, also develop into pan-
creatic ductal-type adenocarcinoma within the salivary
gland of the female animals after BOP induction.23 Thus,
endocrine cell hyperplasia may be an important event in
pancreatic carcinogenesis, not only in this kindred but in
pancreatic carcinogenesis in general.
The lobulocentric fibrocystic pancreatic atrophy and
endocrine cell hyperplasia in this kindred are similar to
the pancreatic pathology in infants with trisomy 13, in
which there is intralobular pancreatic duct ectasia, ductu-
loinsular complexes, and metaplasia of intralobular
ducts.12 This similarity raises the obvious question of
whether the inherited mutant allele in this kindred is on
chromosome 13. Future genetic studies will address this
issue; however, at this point no karyotypic cytogenetic
abnormalities have been identified on chromosome 13 or
elsewhere.
 FAMILIAL FIBROCYSTIC PANCREATIC ATROPHY
The extent of neoplastic change (none to LGD to HGD
to carcinoma) did not demonstrate a significant correla-
tion with increasing patient age. This may be related to
several factors, including the following: 1) the relatively
short time span over which this kindred developed pan-
creatic neoplasia, ranging from ages 31–57; 2) the trend
of decreasing age of disease phenotype with successive
generations, a clinical genetic phenomenon known as
anticipation6; 3) other confounding risk factors, such as
alcohol- or smoking-induced pancreatic alterations in
some of the family members; and 4) chance statistical
variance because of the relatively limited number of pa-
tients with neoplastic change.
In conclusion, we have presented the pathologic fea-
tures of one of the largest pancreatic carcinoma kindreds
described to date. Affected members of this family de-
velop a characteristic clinical prodrome and demonstrate
unique pancreatic pathology with lobulocentric fibrocys-
tic change, acinar atrophy, and endocrine cell hyperpla-
sia in association with ductal proliferative dysplastic pre-
malignant change or invasive carcinoma. We further be-
lieve that the findings in this kindred strengthen the
argument that ductal dysplasia, particularly HGD, is a
precursor of pancreatic carcinoma. The affected indi-
viduals in families such as this, who have abnormalities
detected at screening and surveillance, may require a
prophylactic pancreatectomy to prevent death from pan-
creatic cancer.3 We hope that the unique clinical pro-
drome and pancreatic pathology of this kindred will help
clinicians and pathologists identify other families at
risk. 䊐 crine cancer of the pancreas: their patterns and suggested biologic
REFERENCES
1. Balkau B, Barrett-Connor E, Eschwege E, et al. Diabetes and
pancreatic carcinoma. Diabetes Metab 1993;19:458–62.
2. Brat DJ, Lillemoe KD, Yeo CJ, et al. Progression of pancreatic
intraductal neoplasia to infiltrating adenocarcinoma of the pan-
creas. Am J Surg Pathol 1998;22:163–6.
3. Brentnall TA, Bronner MP, Byrd DR, et al. Effective early diag-
nosis and treatment of familial pancreatic cancer. Ann Intern Med
1999;131:247–55.
4. Calle EE, Murphy TK, Rodriguez C, et al. Diabetes mellitus and
pancreatic cancer mortality in a prospective cohort of United States
adults. Cancer Causes Control 1998;9:403–10.
5. DiGiuseppe JA, Hruban RH, Offerhaus GJA, et al. Detection of
K-ras mutations in mucinous pancreatic duct hyperplasia from a
patient with a family history of pancreatic carcinoma. Am J Pathol
1994;144:889–95.
6. Evans JP, Burke W, Chen R, et al. Familial pancreatic adenocar-
cinoma: association with diabetes and early molecular diagnosis. J
Med Genet 1995;32:330–5.
1053
7. Farley DR, van Heerden JA, Myers JL. Adult pancreatic nesidio-
blastosis: unusual presentations of a rare entity. Arch Surg 1994;
129:329–31.
8. Finch MD, Howes N, Ellis I, et al. Hereditary pancreatitis and
familial pancreatic cancer. Digestion 1997;58:564–9.
9. Goudswaard WB, Houthoff HJ, Koudstaal J, et al. Nesidioblastosis
and endocrine hyperplasia of the pancreas: a secondary phenom-
enon. Hum Pathol 1986;17:46–54.
10. Gown A, de Wever N, Battifora H. Microwave based antigenic
unmasking: a revolutionary new technique for routine immunohis-
tochemistry. Appl Immunohistochem 1993;22:934–44.
11. Greenlee RT, Murray T, Bolden S, et al. Cancer statistics, 2000.
CA Cancer J Clin 2000;50:7–33.
12. Hashida Y, Jaffe R, Yunis EJ. Pancreatic pathology in trisomy 13:
specificity of the morphologic lesion. Pediatr Pathol 1983;1:
169–78.
13. Howard TJ. Pancreatic adenocarcinoma. Curr Probl Cancer 1996;
20:281–2.
14. Hruban RH, Petersen GM, Ha PK, et al. Genetics of pancreatic
cancer: from genes to families. Surg Oncol Clin North Am 1998;
7:1–23.
15. Hruban RH, Wilentz RE, Goggins M, et al. Pathology of incipient
pancreatic cancer. Annu Oncol 1999;10(suppl 4):9–11.
16. Hruban RH, Adsay NV, Albores-Saavedra J, et al. Pancreatic in-
traepithelial neoplasia (PanIN): a new nomenclature and classifi-
cation system for pancreatic duct lesions. Am J Surg Pathol
2001;25:579–86.
17. Longnecker DS. The quest for preneoplastic lesions in the pan-
creas. Arch Pathol Lab Med 1994;118:226.
18. Lowenfels AB, Maisonneuve P, Lankisch PG. Chronic pancreatitis
and other risk factors for pancreatic cancer. Gastroenterol Clin
North Am 1999;28:673–85.
19. Lynch HT, Smyrk TS, Kern SE, et al. Familial pancreatic cancer:
a review. Semin Oncol 1996;23:251–75.
20. Obara T, Saitoh Y, Maguchi H, et al. Multicentric development of
pancreatic intraductal carcinoma through atypical papillary hyper-
plasia. Hum Pathol 1992;23:82–5.
21. Pour PM, Permert J, Mogaki M, et al. Endocrine aspects of exo-
significance. Am J Clin Pathol 1993;100:223–30.
22. Pour PM, Kazakoff K. Stimulation of islet cell proliferation en-
hances pancreatic ductal carcinogenesis in the hamster model. Am
J Pathol 1996;149:1017–25.
23. Pour PM, Weide L, Liu G, et al. Experimental evidence for the
origin of ductal-type adenocarcinoma from the islets of Langer-
hans. Am J Pathol 1997;150:2167–80.
24. Silverman DT, Schiffman M, Everhart J, et al. Diabetes mellitus,
other medical conditions and familial history of cancer as risk
factors for pancreatic cancer. Br J Cancer 1999;80:1830–7.
25. Traverso WL, Bockman DE, Pleis SK, et al. Nesidioblastosis as a
mechanism to prevent fibrosis-induced diabetes after pancreatic
duct obstruction. Pancreas 1993;8:325–9.
26. Wang RN, Klöppel G, Bouwnes L. Duct to islet cell differentiation
and islet cell growth in the pancreas of the duct-ligated adult rats.
Diabetologia 1995;38:1405–11.
27. Whitcomb DC. The spectrum of complications of hereditary pan-
creatitis: is this a model for future gene therapy? Gastroenterol
Clin North Am 1999;28:525–41.
28. Wigglesworth JS, Singer DB. Textbook of Fetal and Perinatal
Pathology, ed 1. Malden, MA: Blackwell Scientific Publications,
1991:1036–46.
Am J Surg Pathol, Vol. 25, No. 8, 2001