| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Laboratory of Pathology,*
National Cancer
Institute; the Surgery Branch,
National Cancer
Institute; the Department of Radiology,
Warren G. Magnuson Clinical Center; and the Urologic Oncology
Branch,§
National Cancer Institute, National
Institutes of Health, Bethesda, Maryland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The gene for VHL disease has been localized to chromosome 3p25.5 and was identified in 1993.5 The two-hit theory of Knudson6 predicts that in a familial cancer syndrome such as VHL disease, the genotype of each neoplasm should consist of an allele with an inherited germ-line mutation and loss of the second wild-type allele through allelic deletion. In fact, genetic studies of different VHL disease-associated tumors, including hemangioblastomas,7 renal cell carcinomas,8-10 pheochromocytomas,11 pancreatic microcystic adenomas,12 and endolymphatic sac tumors,13 demonstrated loss of heterozygosity (LOH) at chromosome 3p at the VHL gene region. VHL gene deletion, however, has never been studied in VHL-associated pancreatic NETs, and it is therefore unknown whether these tumors show genetic evidence of VHL disease-related pathogenesis.
In this study, we performed macroscopic, histopathological, and immunohistochemical evaluation of 30 pancreatic NETs in 14 VHL patients. Furthermore, to investigate VHL gene alterations in VHL NETs, DNA from tumor and normal pancreatic tissue in 6 patients was studied for allelic deletions of the VHL gene by fluorescence in situ hybridization (FISH) and/or polymerase chain reaction (PCR)-single-strand conformational polymorphism analysis (SSCP).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Fourteen VHL patients (eight females and six males; mean age, 35
years; range, 18 to 48 years) with solid pancreatic lesions were
selected from the group of familial VHL patients followed at the
National Cancer Institute, National Institutes of Health14
(Table 1)
. Each patient had a documented
germ-line mutation in the VHL gene.
|
. Four
patients were treated by enucleation, five patients were treated by
distal/partial pancreatectomy, two patients underwent total
pancreatectomy, two patients were treated by pancreaticoduodenectomy
(Whipple's procedure), and one patient had an exploratory laporotomy
and biopsy of pancreatic and liver lesions. From one patient (patient
2), additional pancreatic NETs were obtained at autopsy (Table 1)
. Morphological Examination of Tumors
Thirty formalin-fixed, paraffin-embedded pancreatic tumors were
obtained from the files of the Laboratory of Pathology, National Cancer
Institute, National Institutes of Health. All tumors were grossly
examined and measured. Tumors were evaluated on hematoxylin and eosin
(H&E) stain, periodic acid-Schiff (PAS), and PAS-diastase (PAS-D)
stains. Immunohistochemistry stains for cytokeratin AE1/AE3
(1:300/1:100) and chromogranin A (1:1600) (both from Boehringer
Mannheim, Indianapolis, IN); synaptophysin (1:1000) (Zymed, San Diego,
CA); neuron-specific enolase (NSE) (1:200), glucagon (1:900),
pancreatic polypeptide (1:8000), somatostatin (1:3000), gastrin
(1:2000) (DAKO, Carpinteria, CA) S100 (1:8000), and insulin (1:50)
(BioGenex, San Ramon, CA) were performed in all tumors in which
sufficient tissue was available (Table 2)
. Tumor cells were scored as positive
by immunohistochemistry stain when they showed moderate-to-marked
intensity of staining in the appropriate distribution for the marker,
with adequate tissue controls. Electron microscopy was performed in
five cases (patients 1, 3, 4, 7, and 8). For electron microscopic
examination, 2.5% glutaraldehyde-fixed, osmium-postfixed tumor tissue
was embedded in Maraglas 655 (Ladd Research Industries, Burlington,
VT). Thin sections were stained with uranyl acetate and lead citrate
and were reviewed in a Philips CM10 transmission electron microscope.
|
Six NETs from VHL patients 1, 4, and 11–14 and a pancreatic adenocarcinoma from patient 11 were analyzed for VHL gene deletion by PCR-SSCP analysis using a single nucleotide polymorphic marker (104/105) at the VHL gene locus.10,15 Three patients (patients 1, 4, and 11) were informative for the marker; ie, normal DNA showed two different alleles (heterozygosity). Briefly, tumor and normal pancreatic cells were procured by a modified tissue microdissection procedure using 5-µm histological sections from frozen or formalin-fixed, paraffin-embedded tissue.16 Cells were procured with a 30-gauge needle and immediately resuspended in 10-µL solution containing 10 mmol/L Tris-HCl, pH 8.0, 1 mmol/L ethylenediamine tetraacetic acid, pH 8.0, 1% Tween 20, 0.1 mg/ml proteinase K, and incubated overnight at 37°C. The mixture was boiled for 10 minutes to inactivate the proteinase K, and 10% of this solution was used for PCR analysis. The PCR amplification reaction was carried out for 30 cycles at 95°C for 30 seconds, 70°C for 30 seconds, using primers to the single-nucleotide polymorphism upstream of the coding region of the VHL gene: upstream, 5'-AGT GGA AAT ACA GTA ACG AGT TGG CCT-3'; downstream, 5'-GTC CCA GTT CTC CGC CCT CCG GGG CAT-3'.15 Labeled amplified DNA was mixed with an equal volume of formamide loading dye (95% formamide, 20 mmol/L ethylenediamine tetraacetic acid, 0.05% bromphenol blue, and 0.05% xylene cyanol) and analyzed on SSCP gel.10,15 The samples were denatured for 5 minutes at 95°C and loaded onto a gel consisting of 6% acrylamide (49:1 acrylamide:bis), 5% glycerol, and 0.6x Tris-borate ethylenediamine tetraacetic acid. Samples were electrophoresed at 8 W at room temperature overnight. Gels were transferred to 3-mm Whatman paper and dried, and autoradiography was performed with Kodak X-OMAT film (Eastman Kodak, Rochester, NY). The complete or near complete (90% decreased intensity) absence of one allele on acrylamide gel was interpreted as LOH. Each result was reproduced two to three times.
Four tumors from VHL patients 1, 2, 7, and 10 were analyzed for
VHL gene deletion by FISH. Briefly, a touch preparation was
performed from fresh or frozen tumors. FISH was performed using a P1
plasmid clone containing the entire VHL gene as a
probe.17
DNA was labeled with digoxigenin-11-dUTP by nick
translation (Boehringer Mannheim). Biotin-labeled 
-satellite
centromeric probe specific for chromosome 3 (Oncor, Gaithersburg, MD)
was used as a control. Slides were denatured in 70% formamide/2x
standard saline citrate at 72°C for 2 minutes and dehydrated in
ethanol series of 70, 80, 90, and 100%. The probes were denatured at
70°C for 10 minutes and then incubated at 37°C for 30 minutes for
preannealing. DNA (250 µg) was applied to the slide and allowed to
hybridize overnight in a humidified chamber at 37°C. Detection was
performed using anti-digoxigenin rhodamine and avidin-fluorescein
isothiocyanate. Hybridization signals were scored using a Zeiss
Axiophot fluorescence microscope, and three-color images were captured
on a Photometrics cooled-charge-coupled device camera (Photometrics,
Tucson, AZ) using IP Lab image software (Signal Analytics Corporation,
Vienna, VA). At least 100 interphases with strong hybridization signals
were scored for each tumor. The presence of more than 20% cells with
only one VHL signal was interpreted as an allelic deletion
(loss of one copy of the VHL gene). The tumor cells retained
both centromeric probes. Normal frozen pancreatic tissue control showed
both centromeric probes and less than 2% of cells with one
VHL signal.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Thirty tumors were identified in 14 patients. Seven patients had
multiple pancreatic tumors, and seven patients had a single tumor. The
tumors were located in the head, body, and/or tail of the pancreas
(Table 1)
. All tumors were well circumscribed and varied in size from
0.4 to 8 cm (median, 2 cm). The color was tan, red/brown, gray, or
yellow (Figure 1)
. Invasion of adjacent
organs was not observed.
|
Histologically, 28 tumors were confined to the pancreas. One tumor
in patient 1 showed a direct extension to a peripancreatic lymph node.
Metastases to peripancreatic lymph nodes were not observed. Patient 3
had a biopsy-documented liver metastasis (tumor 8). Scattered islets,
nerves, and exocrine ductules were seen within the tumor in 50% of the
cases. The tumors showed solid, trabecular, and/or glandular
architecture (Table 2
; Figure 2, A and B
). Twenty-two tumors showed prominent stromal collagen bands (Figure 2C)
, and one tumor demonstrated calcification (Table 2)
. Congo red
stain for amyloid was negative in the stroma. Angioinvasion was not
present. The cytoplasm was eosinophilic in 8 tumors (Figure 2A)
,
amphiphilic in 4 tumors, clear in 7 tumors (Figure 2, B and C)
,
eosinophilic and clear in 5 tumors, and amphiphilic and clear in 6
tumors. Thus, 60% (18 of 30) of tumors demonstrated cells with clear
cytoplasm (Figure 2, B and C)
. PAS and PAS-D stains demonstrated
glycogen in a minority of cases. The nuclear features characteristic
for NETs were seen in all of the cases. Seventeen of 28 (61%) tumors
showed prominent focal nuclear atypia (Figure 2C)
. Mitoses were
infrequent; the mitotic rate did not exceed 2 mitoses per 10 high-power
fields. Evaluation of the pancreas for other lesions was possible in 11
cases. Additional lesions observed included benign serous cysts (5
cases), microcystic adenomas (3 cases), and a pancreatic adenocarcinoma
(1 case) (Table 1)
. Nesidioblastosis and hyperplasia of the islets of
Langerhans were not observed in adjacent pancreas.
|
All 18 tumors that were evaluated by synaptophysin stain showed
moderate-to-marked cytoplasmic positivity (Table 2
; Figure 3A
). Chromogranin A stain was positive in
14 of 18 tumors evaluated. S100 (Figure 3B)
and NSE stains were
positive in all tumors evaluated (15 of 15 and 11 of 11 tumors,
respectively). Cytokeratin AE1/AE3 demonstrated positivity in 16 of 16
tumors. Fifteen of 23 tumors evaluated with hormonal markers (glucagon,
pancreatic polypeptide, somatostatin, insulin, and gastrin) showed
negative results. Four tumors were positive with multiple hormonal
markers. Six tumors were positive for pancreatic polypeptide, 4 for
somatostatin, 2 for insulin, and 1 for glucagon. The majority of these
stains showed only focal positivity in tumor cells (Figure 3, C and D)
.
All 23 tumors were negative for gastrin. Prominent small vessels in
NETs were demonstrated by a vascular marker, CD34 (QBEND 10, Immunotech
Inc., Westrock, MA).
|
Electron microscopy was performed in six tumors from five
patients. Electron-dense neurosecretory granules (Figure 2D)
were
demonstrated in all tumors. Neoplastic clear cells showed abundant
intracytoplasmic lipid droplets. In addition, irregular, concentric,
electron-dense phospholipid or lipoprotein membranes consistent with
"myelin figures" were observed in clear cells (Figure 2D)
. Both
primary pancreatic and metastatic tumor in the liver in patient 3
showed tumor cells that were surrounded by basal laminae and contained
large amounts of lipid and numerous neuroendocrine granules.
Cytoplasmic glycogen was detected in one tumor.
VHL Gene Allelic Deletion Results
In each of the three informative patients (patients 1, 4, and 11)
studied by PCR-SSCP analysis, LOH of one VHL gene allele was
found in DNA from pancreatic NET and not in the DNA extracted from
adjacent normal pancreatic parenchyma. Figure 4A
illustrates representative results of
SSCP analysis of DNA amplified across a single nucleotide polymorphic
site at the VHL gene from two NETs in patients 1 and 4. For
each patient, both alleles are present in lanes N1 and
N2 (Figure 4)
, respectively, containing DNA procured from
the adjacent normal pancreas. In contrast, loss of the lower allele is
detected in lane T1 in patient 1, and loss of the upper
allele is detected in lane T2 in patient 4, containing DNA
from pure populations of microdissected NET cells (Figure 4)
. In
patient 11, both pancreatic adenocarcinoma and NET were analyzed. LOH
at the VHL gene was not detected in pancreatic
adenocarcinoma in contrast to an NET in the same patient and to all
NETs in the study. In four VHL patients (patients 1, 2, 7, and 10),
allelic deletion of one copy of the VHL gene was detected by
FISH in interphase touch preparations of pancreatic NETs. Figure 4
,
B–D, illustrates representative results of FISH analysis in
tumors from patients 1, 7 and 10.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Twelve percent of all VHL patients (n = 256)
evaluated at the National Cancer Institute had solid pancreatic lesions
by imaging studies.14
VHL patients with pancreatic NETs in
the current study were significantly younger (mean age, 35 years) than
sporadic patients with NETs (mean age, 58 years).24
All
tumors were hormonally nonfunctional. The pancreatic NETs in VHL
patients were often multiple, both microscopic and macroscopic in
size, and were located throughout the pancreas (Table 1)
.
Although usually well circumscribed and confined to the pancreas, VHL-associated NETs can metastasize, as evidenced by one case in this study. Recently, four other VHL patients referred to the National Cancer Institute had documented liver metastases from a pancreatic NET. These four cases are reported elsewhere,14 but they were not included in the present study, because the primary pancreatic tumor was diagnosed by radiology only and was not available for evaluation. The median primary tumor diameter in patients with metastases was 5 cm, compared with a median primary tumor diameter of 2 cm for patients without evidence of metastatic disease. Therefore, the size of the primary VHL NET appears to be related to the risk of metastatic disease.14 Similarly, three of the six NETs in VHL patients reported by the Mayo Clinic were malignant.4
Histologically, all VHL-associated pancreatic lesions demonstrated an architecture characteristic for NETs. Neuroendocrine differentiation was confirmed by synaptophysin, chromogranin A, NSE, and S100 staining, and/or small, dense core granules shown by electron microscopy. Negative immunostaining for pancreatic and gastrointestinal hormones was observed in 15 of 23 of tumors (65%), and 8 of 23 (35%) NETs demonstrated focal positivity for pancreatic polypeptide, somatostatin, insulin, and/or glucagon. Small vessels, stromal collagen bands, and focal nuclear atypia were prominent in the majority of our VHL cases. The distinguishing feature of VHL NETs was clear-cell morphology that was present in 60% of the tumors, regardless of size. The cytoplasmic clearing was attributed to prominent lipid globules and myelin figures demonstrated by electron microscopy. A minority of VHL-associated NETs showed cytoplasmic glycogen on PAS and PAS-D stains or on electron microscopy.
NETs in VHL patients may be a diagnostic challenge because other VHL-associated tumors, such as renal cell carcinoma, hemangioblastoma, pancreatic microcystic (serous) adenoma, and epididymal cystadenoma, which contain numerous small vessels and clear cells, are histologically similar. The most common lesions to be considered in the differential diagnosis of VHL NETs are microcystic adenoma of the pancreas and metastatic renal cell carcinoma. Microcystic adenomas and NETs may occur in the same pancreas, and solid areas of microcystic adenoma can be hard to distinguish from clear-cell NETs in VHL patients.14,25 However, pancreatic microcystic adenomas are usually glycogen rich and lack dense core granules.26 Clear-cell renal cell carcinomas tend to form sheets of clear cells and microcysts separated by thin fibrovascular septae rather than the broad collagen bands common in NETs; renal cell carcinomas also have cytoplasmic glycogen or lipid and are negative for neuroendocrine immunohistochemistry markers. Two additional, uncommon pancreatic tumors may also be considered. The first, clear-cell "sugar" tumor of pancreas, similar to "sugar" tumor of lung, demonstrates abundant cytoplasmic glycogen and is HMB-45-positive and cytokeratin, NSE, and chromogranin A-negative.27 The second, clear-cell carcinoma of the pancreas, is exceedingly rare and demonstrates significant pleomorphism and invasiveness.26
The presence of pancreatic NETs, as well as pheochromocytomas in patients with VHL, suggests that VHL disease may represent a continuum of multiple endocrine neoplasia.4,18,20 Multiple pancreatic NETs commonly occur in multiple endocrine neoplasia type 1 (MEN1).24,28,29 However, along with similarities, several distinguishing features may be seen in pancreatic pathology of patients with VHL disease and MEN1. First, NETs are found in 82 to 100% of MEN1 patients,24 as compared with an incidence of 12 to 17% in VHL disease.4,14 Second, hormonally functional NETs are rare in VHL disease4,21 but are relatively common in MEN1;28 all tumors were nonfunctional in this study. Finally, pancreatic resection specimens in MEN1 patients demonstrate multiple microadenomas and nesidioblastosis in 30% of MEN1 cases.29 VHL patients in this study had a mean of two pancreatic NETs. Although scattered islets and ductules were frequent within VHL-associated NETs, nesidioblastosis and/or NETs less than 0.4 cm in size were not observed in adjacent pancreas.
The two-hit tumor suppressor gene theory of Knudson6 predicts that in a familial cancer syndrome such as VHL disease, the genotype of each neoplasm should consist of one allele with an inherited germ-line mutation and loss of the second wild-type allele, which occurs through chromosomal deletion. Therefore, LOH or mutation at the VHL gene should be detectable in the lesion if the lesion represents a VHL-associated neoplasm. LOH at chromosome 3p has been detected in VHL-associated renal cell carcinomas,8-10 hemangioblastomas,7 pheochromocytomas,11 endolymphatic sac tumors,13 and pancreatic microcystic adenomas.12 In this study, allelic deletion of the second copy of the VHL gene was detected in NETs from all six VHL patients with known germ-line mutations. Furthermore, in one case (patient 11), VHL gene heterozygosity was retained in a pancreatic adenocarcinoma that was associated with a NET with a VHL gene deletion. The presence of allelic deletions of the VHL gene in pancreatic NETs provides direct molecular evidence for a role of the VHL gene in their tumorigenesis and establishes NET as an independent tumor type of VHL disease. In contrast, retention of heterozygosity in adenocarcinoma, coupled with the absence of other VHL patients with pancreatic carcinoma in our series, suggests that VHL gene alterations are unlikely to be associated with the development of pancreatic adenocarcinoma in VHL patients.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Accepted for publication April 10, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. M. Hassan, A. Phan, D. Li, C. G. Dagohoy, C. Leary, and J. C. Yao Family History of Cancer and Associated Risk of Developing Neuroendocrine Tumors: A Case-Control Study Cancer Epidemiol. Biomarkers Prev., April 1, 2008; 17(4): 959 - 965. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Garbrecht, M. Anlauf, A. Schmitt, T. Henopp, B. Sipos, A. Raffel, C. F Eisenberger, W. T Knoefel, M. Pavel, C. Fottner, et al. Somatostatin-producing neuroendocrine tumors of the duodenum and pancreas: incidence, types, biological behavior, association with inherited syndromes, and functional activity Endocr. Relat. Cancer, March 1, 2008; 15(1): 229 - 241. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Blandamura, A. Parenti, B. Famengo, A. Canesso, P. Moschino, C. Pasquali, S. Pizzi, V. Guzzardo, and V. Ninfo Three cases of pancreatic serous cystadenoma and endocrine tumour J. Clin. Pathol., March 1, 2007; 60(3): 278 - 282. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. M. Linehan, P. A. Pinto, R. Srinivasan, M. Merino, P. Choyke, L. Choyke, J. Coleman, J. Toro, G. Glenn, C. Vocke, et al. Identification of the Genes for Kidney Cancer: Opportunity for Disease-Specific Targeted Therapeutics Clin. Cancer Res., January 15, 2007; 13(2): 671s - 679s. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. R Woodward and E. R Maher Von Hippel-Lindau disease and endocrine tumour susceptibility. Endocr. Relat. Cancer, June 1, 2006; 13(2): 415 - 425. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R Chetty and S L Asa Pancreatic endocrine tumour with cytoplasmic keratin whorls. Is the term "rhabdoid" appropriate? J. Clin. Pathol., October 1, 2004; 57(10): 1106 - 1110. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. B. Marcos, S. K. Libutti, H. R. Alexander, I. A. Lubensky, D. L. Bartlett, M. M. Walther, W. M. Linehan, G. M. Glenn, and P. L. Choyke Neuroendocrine Tumors of the Pancreas in von Hippel-Lindau Disease: Spectrum of Appearances at CT and MR Imaging with Histopathologic Comparison Radiology, December 1, 2002; 225(3): 751 - 758. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. H. Mohr, A. O. Vortmeyer, Z. Zhuang, S. K. Libutti, M. M. Walther, P. L. Choyke, B. Zbar, W. M. Linehan, and I. A. Lubensky Histopathology and Molecular Genetics of Multiple Cysts and Microcystic (Serous) Adenomas of the Pancreas in von Hippel-Lindau Patients Am. J. Pathol., November 1, 2000; 157(5): 1615 - 1621. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Shen, Z. Zhuang, D. J. Gersell, and F. A. Tavassoli Allelic Deletion of VHL Gene Detected in Papillary Tumors of the Broad Ligament, Epididymis, and Retroperitoneum in von Hippel-Lindau Disease Patients International Journal of Surgical Pathology, July 1, 2000; 8(3): 207 - 212. [Abstract] [PDF]
|
|
|
|
 |
|
 S. J. Marx CLINICAL REVIEW 109: Contrasting Paradigms for Hereditary Hyperfunction of Endocrine Cells J. Clin. Endocrinol. Metab., September 1, 1999; 84(9): 3001 - 3009. [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
