| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Biology and Genetics |
Departments of Surgery [G. La., B. S., P. K., S. G.] and Laboratory Medicine and Pathobiology [H. O., M. R.], the Centre for Research in Women’s Health [G. Li., S. A. N.], University of Toronto, and Centre for Cancer Genetics, Samuel Lunenfeld Research Institute [G. La., H. O., M. R., S. G.] and Department of Pathology and Laboratory Medicine [H. O., M. R.], Mount Sinai Hospital, Toronto, Ontario, M5G 1X5 Canada
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
20% of pancreatic carcinomas are
resectable at presentation (4)
. Moreover, the majority of
patients with resected pancreatic cancers develop local recurrences and
metastases and die of their disease within a short time after surgery
(5)
. Pancreatic cancer is associated with various
environmental and lifestyle risk factors (6)
, occupational
exposures (7)
, and medical conditions
(8, 9, 10)
. However, the etiology of this nearly uniformly
fatal disease still remains poorly understood. Two lines of evidence have recently emerged to suggest that susceptibility to pancreatic cancer development may be inherited. The first line of evidence comes from case reports describing multiple individuals with pancreatic cancer in the same family (11) . Several epidemiological studies have also demonstrated that a family history of pancreatic cancer is an important risk factor for the disease. In a case control study, Falk et al. (12) showed an increased risk of pancreatic cancer among persons reporting the occurrence of any cancer in a close relative (odds ratio = 1.86; 95% CI,3 1.42–2.44). This risk was highest when the reported cancer was a pancreatic tumor (odds ratio = 5.25; 95% CI, 2.08–13.21). In a different case-control study in the French-Canadian population, 7.8% of cases but only 0.6% of controls reported a family history of pancreatic cancer (13) . Furthermore, in an Italian study, a family history of pancreatic cancer was associated with a relative risk of 2.8 (95% CI, 1.3–6.3) for the disease, even after adjustment for smoking, alcohol, dietary factors, and medical history (14) .
The second line of evidence comes from the observation that pancreatic cancer occurs in excess of expected frequencies in several familial cancer syndromes, which are associated with germ-line mutations in various cancer predisposing genes. The familial atypical multiple mole melanoma (3) syndrome is characterized by the development of multiple nevi and malignant melanomas (15) . Germ-line p16 mutations have been shown to cosegregate with melanoma in a subset of familial atypical multiple mole melanoma kindreds showing linkage to 9p21 (16 , 17) . Pancreatic adenocarcinoma is the second most common cancer in these families (18) , and kindreds with p16 mutations that impair protein function have a 13-fold increased risk of pancreatic cancer (19) . An excess of pancreatic cancer also occurs in breast-ovarian cancer families with BRCA1 and BRCA2 mutations (20, 21, 22, 23, 24) and in kindreds with HNPCC (25, 26, 27, 28, 29) , which is caused by germ-line mutations of hMSH2, hMLH1, and other mismatch repair genes (30, 31, 32, 33, 34, 35, 36, 37) .
It is estimated that 5–10% of pancreatic cancers are
associated with a strong familial predisposition (13
, 38)
,
but the proportion of pancreatic cancers attributable to inherited
mutations of the above-mentioned genes remains largely unknown.
Therefore, the present study was undertaken to determine the
contribution of familial cancer predisposing genes to pancreatic cancer
susceptibility. The specific objectives were twofold: (a) to
estimate the proportion of unselected pancreatic cancer cases that
belong to families with hereditary cancer syndromes; and (b)
to estimate the frequency of germ-line mutations of the p16,
BRCA1, BRCA2, hMSH2, and
hMLH1 genes in pancreatic cancer patients with a personal or
family history of cancer.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
All patients were asked to provide a blood sample (20 ml) for mutation analysis and complete a self-administered family history questionnaire to document the current ages, ages at death, causes of death, and ages at diagnosis of any cancer and the types of cancer found in any member of their first-, second-, or third-degree relatives. Patients were also asked to indicate their ethnic origin. Questionnaires were most often completed by patients at home and mailed back to the investigators. This was necessary given the poor health of most patients and the need to contact other relatives to obtain information. A personal or telephone interview with the patient (or spouse, if the patient died during the course of the study) was occasionally conducted to complete the questionnaires. Information provided in the questionnaires was verified, and missing information was completed in a second telephone interview with all patients. Every attempt was made to verify cancer diagnoses in relatives through examination of hospital records and/or pathology reports. The relative or a proxy was approached for consent to obtain this information from medical records departments. Data thus obtained was used to construct a three-generation pedigree.
All blood samples and questionnaire data were obtained after informed consent was obtained in accordance with protocols approved by the University of Toronto Human Ethics committee.
Risk Classification
Pedigrees/cases were classified based on the criteria outlined in
Table 1
. Pedigrees were classified as high risk/familial or intermediate
risk/familial if they satisfied criteria for various inherited cancer
syndromes, namely, familial melanoma, familial breast-ovarian cancer
syndrome, HNPCC, or familial pancreatic cancer. Patients were
classified as intermediate risk/nonfamilial if they developed
pancreatic cancer before 50 years of age or if they had multiple
primary cancers without an associated family history of cancer. If the
pedigrees did not meet any of these criteria, they were considered low
risk and excluded from further genetic testing. Our Mount Sinai
criteria for HNPCC are modified from the Amsterdam criteria
(26)
and include: (a) three individuals in at
least two successive generations with at least one colorectal cancer
and two others with either gastrointestinal, genitourinary, or
gynecological cancers with no age limit for the cancer diagnosis; or
(b) any colorectal cancer patient diagnosed at <35 years of
age with or without a family history of cancer; or (c) any
individual with multiple primary cancers of the sites associated with
HNPCC with or without a family history of cancer; and (d) a
clinical diagnosis of familial adenomatous polyposis is excluded
(39)
.
|
. All high-risk and intermediate-risk patients underwent testing for
p16 and BRCA2. The former is a small gene (3
exons) and is technically simple to screen (40)
. Previous
studies suggest a role for BRCA2 in pancreatic cancer even
in the absence of a family history of breast and/or ovarian cancer
(41
, 42)
. We have previously screened 39 Jewish pancreatic
cancer patients for the BRCA1 185delAG founder mutation and
found that, unlike BRCA2 6174delT, BRCA1
germ-line mutations probably do not contribute to disease in patients
with sporadic pancreatic
cancer.4
Hence, BRCA1 testing was confined to patients with family
histories suggestive of breast-ovarian cancer syndrome. Likewise,
mismatch repair gene expression was analyzed only in patients with a
personal or family history suggestive of HNPCC.
|
Germ-Line Mutation Analysis
p16.
PCR was performed in a 20-µl reaction mixture containing 200 ng of
DNA, 2 µl of 10x PCR buffer [1x PCR buffer = 10
mM Tris-HCl (pH = 8.3), 50 mM
KCl, and 0.01% gelatin], 1.25–3.0 mM
MgCl2, 1 µl of 10 mM
deoxynucleotide triphosphates, 0.5 µl each of forward and reverse
primers (20 µM), 8.8 µl of 12% DMSO, and 1 unit of
AmpliTaq polymerase (Perkin-Elmer, Branchburg, NJ). Exons 1
and 2 of
p16 were amplified using previously described primers
(X1.31F and X1.26R for exon 1, and X2.62F and X2.42R for exon 2; Ref.
17
). In some cases, it was necessary to amplify exon 1 in
two smaller fragments using a new set of internal primers: 1AR,
5'-TCGGTGCGTTGGGCAGCG-3'; and 1BF, 5'-CGGGGTCGGGTAGAGGAG-3'. PCR was
performed in a GeneAmp 9600 thermal cycler (Perkin-Elmer, Foster City,
CA) as follows: denaturation at 94°C for 4 min; 30 cycles each of
denaturation at 94°C for 15 s, annealing at 55–67°C for
20 s, and extension at 72°C for 20 s; and a final extension
at 72°C for 10 min. PCR products were electrophoresed on 2% agarose
gels and purified using the Qiaex II gel extraction kit (Qiagen,
Mississauga, Ontario, Canada). Purified PCR products were cycle
sequenced to directly screen for mutations using the Thermosequenase
radiolabeled dideoxy chain termination kit (Amersham Life Science,
Cleveland, OH). 33P-labeled sequencing products
were separated by electrophoresis on 6% denaturing polyacrylamide gels
(80 W, 1.5–3 h), dried (80°C for 1 h), and exposed to
autoradiographic film (BioMax; Eastman Kodak, Rochester, NY) overnight
at room temperature.
BRCA1 and BRCA2
PTT.
PTT was used to screen for mutations in these large genes. About 60% of BRCA1 and BRCA2 are encoded by a large exon 11 (44 , 45) , and this was screened first by amplifying overlapping fragments from genomic DNA. If no truncated protein was observed in exon 11, the remaining coding region was examined by reverse transcription-PCR and PTT, using cDNA as template. PCR was performed using previously described overlapping primers and conditions in a 50-µl reaction volume containing 100 ng of DNA as template (44) . For reverse transcription-PCR, 150 ng of cDNA were substituted for DNA, and the reaction volume was reduced to 30 µl.
Protein products were subsequently synthesized in a coupled in vitro transcription/translation reaction (TnT/T7 coupled rabbit reticulocyte lysate system; Promega) in a reaction mixture containing 2.4 µl of PCR product, 0.24 µl of TNT buffer, 0.12 µl of amino acid minus methionine, 4.8 units of RNasin, 0.12 µl of T7 polymerase, 0.40 µl of [35S]methionine (Amersham Life Sciences), and 2.6 µl of rabbit reticulocyte lysate. The mixture was incubated at 30°C for 2 h. Proteins were separated by SDS-PAGE (13% SDS-PAGE, 280 V, 3 h). Gels were fixed (70% water, 20% methanol, 10% acetic acid) for 30 min, dried at 60°C for 2 h, and exposed to autoradiographic film as described above. All putative mutations were confirmed by cycle sequencing of an independent PCR product, as described above.
Heteroduplex Analysis.
In addition to PTT, heteroduplex analysis was used to rapidly screen
appropriate cases (patients of Ashkenazi Jewish descent) for known
founder mutations, namely, BRCA2 6174 delT, BRCA1
185delAG, and BRCA1 5382 insC. PCR was performed using
previously described primers and conditions (42
, 46
, 47)
in a 50-µl volume containing 100 ng of DNA as template.
PCR products were mixed with 6 µl of sucrose dye (0.25% bromphenol blue, 0.25% xylene cyanol, 40% sucrose in water), and heteroduplex formation was carried out by denaturing the PCR products at 95°C and slow cooling to room temperature. Products were separated by electrophoresis on 12% polyacrylamide or mutation detection enhancement gels (FMC Bioproducts, Rockland, ME; 300 V, 4 h), stained with ethidium bromide, and visualized with UV light.
hMSH2 and hMLH1
All available paraffin-embedded tumor tissues from pancreatic
cancer patients with a family history suggestive of HNPCC were cut into
6-µm sections, and immunohistochemical analysis was carried out as
described previously (48)
. Cancers were considered to have
loss of hMSH2 or hMLH1 expression if there was no detectable nuclear
staining of neoplastic cells. Intact nuclear staining of surrounding
nonneoplastic epithelium, stromal cells, or lymphocytes was a
prerequisite for evaluation of sections and served as an internal
positive control.
All germ-line mutations were confirmed by repeat testing of an independent PCR product.
Statistical Analyses
SPSS software was used for statistical analyses. Mean ages at
diagnosis in mutation-positive and -negative families are indicated
with corresponding SDs and were compared using the Student’s
t test. Ethnic differences and mutation frequencies in the
various risk categories were compared with the Fisher’s exact test.
All Ps are two-sided, unless otherwise stated.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Risk Status
Twenty-seven of 102 patients (26%) satisfied criteria for various
familial cancer syndromes. Of these, 8 patients were classified as high
risk, and 19 were classified as intermediate risk. The distribution of
study patients in various risk groups is shown in Fig. 1
. The largest group was composed of 16 patients who satisfied our
criteria for familial pancreatic cancer. Of these patients, 12 patients
had one other affected relative with pancreatic cancer, 2 patients had
two other affected relatives with pancreatic cancer, and 2 patients had
three other affected relatives with pancreatic cancer. The next largest
group was made up of nine patients fulfilling our Mount Sinai criteria
for HNPCC. None of these patients satisfied the more stringent
Amsterdam criteria (26)
. Seven patients fulfilled criteria
for the breast-ovarian cancer syndrome. Four of these patients had two
relatives with breast and/or ovarian cancers, and three patients had
more than two relatives with breast and/or ovarian cancers. The
smallest familial cancer syndrome group was comprised of three patients
with a family history of melanoma. Only one patient had more than one
relative affected with melanoma.
|
In contrast, the intermediate-risk/nonfamilial group was made up of 10
patients who were diagnosed at age 
50 years and 1
patient with two primary cancers (pancreatic cancer-melanoma), without
an associated family history of cancer.
Mutation Analysis
The results of the mutation analysis are summarized in Table 2
.
p16.
Thirty-eight high-risk and intermediate-risk cases were analyzed for
mutations. One p16 mutation was identified, a missense
mutation that leads to an amino acid substitution, isoleucine to serine
at codon 49 (I49S). This mutation was identified in a high-risk patient
with a strong family history of melanoma. We have characterized the
functional significance of this mutation in a yeast-two hybrid assay
(49)
and found that this variant has compromised binding
to cdk4 when compared with wild-type p16 protein (50)
.
Another p16 sequence variant, which leads to an alanine to threonine substitution at codon 148 (A148T), was identified in a patient with a family history of breast-ovarian cancer. This previously reported variant neither impairs p16 protein function nor segregates with disease in melanoma kindreds, and it is found in normal controls. It has therefore been designated a polymorphism (17 , 51) . No p16 mutations were identified in the 11 patients who belonged to the intermediate-risk/nonfamilial group.
BRCA1 and BRCA2.
The majority (80%) of mutations in these genes are nonsense or
frameshift mutations that lead to truncated proteins (52)
.
Hence, PTT was used to screen for mutations. Like p16,
BRCA2 was also tested in all high-risk and intermediate-risk
patients. Only those patients with a personal or family history of
breast and/or ovarian cancer who did not also have BRCA2
mutations were screened for BRCA1 alterations. Four
mutations were identified [one in BRCA1 (5382 insC) and
three in BRCA2 (all 6174delT)], all in patients of
Ashkenazi Jewish descent. All these mutations are frameshifts and lead
to truncated, nonfunctional proteins (52)
. The patient
with the BRCA1 mutation satisfied intermediate-risk criteria
for breast-ovarian cancer. Two of the three patients harboring
BRCA2 mutations were classified as high risk for
breast-ovarian cancer syndrome, and one was classified as intermediate
risk/nonfamilial. This latter patient had early-onset pancreatic cancer
(age at diagnosis, 35 years) but had been adopted, and family
history was not available.
In addition, another BRCA2 sequence variant, which leads to an alanine to threonine substitution at codon 2951 in exon 22 (A2951T), was identified in one patient with a family history of pancreatic cancer. This variant is a previously described polymorphism (53) .
hMSH2 and hMLH1.
Although there are six mismatch repair genes, only five
(hMSH2, hMLH1, hPMS1,
hPMS2, and hMSH6/GTBP) are mutated in the
germ-line of HNPCC family members. Furthermore, the majority (>90%)
of mutations in these kindreds involve either hMSH2 or
hMLH1 (30, 31, 32, 33, 34, 35, 36, 37)
. Because the inactivation of
mismatch repair genes is associated with complete loss of corresponding
mismatch repair protein expression, immunohistochemistry using
monoclonal antibodies against mismatch repair gene products can be used
to identify tumors with mismatch repair deficiency (54)
.
We have recently demonstrated a high correlation between mismatch
repair protein expression and germ-line mutation status in tumors from
HNPCC patients (48)
. Hence, this technique was used to
analyze patients satisfying HNPCC criteria.
Adequate tumor and normal tissue for immunohistochemistry was available for only three patients who met the criteria for HNPCC, and these were tested for hMSH2/hMLH1 expression. All three tumors demonstrated intact expression of these mismatch repair proteins. Therefore, these cases were not considered for direct germ-line mutation analysis of hMSH2 and hMLH1.
To summarize, 5 of 38 (13%) tested pancreatic cancer patients had
germ-line mutations in three different cancer predisposing genes.
Pedigrees of mutation carriers are shown in Fig. 2
. Of the five mutations, three were identified in cases with high-risk
family histories (3 of 8 tested, 38%). Of the remaining two mutations,
one was identified in a patient with an intermediate-risk family
history (1 of 19 tested, 5%), and the other was identified in a young
Jewish patient with an unknown family history (1 of 11 tested, 9%).
The mutation frequency was significantly different between the
high-risk and intermediate-risk groups (3 of 8 versus 2 of
30; Fisher’s exact P = 0.05). Although
mutation carriers tended to be younger, there was no statistically
significant difference in age at diagnosis between mutation carriers
(48.2 ± 10.9 years) and noncarriers (58.7 ±12.4
years).
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The peak incidence of pancreatic cancer is in the seventh and eighth
decades of life, and males are at a slightly higher risk than females.
The mean age of our patients was 60.6 years (range, 30–87 years).
Although pancreatic cancer is rare before age 50 years
(2)
, patients diagnosed at 50 years of age and males are
overrepresented in our study population and constitute 17% and 64% of
the total series, respectively. Furthermore, our series has a greater
proportion of Jewish patients when compared to the general population
of Ontario (14% of total versus <2%; Ref.
58
). Although certain populations such as New Zealand
Maoris and Hawaiians are characterized by the highest incidence rates
for pancreatic cancer (59)
, the incidence of this disease
in Israeli Jews is comparable to that in several developed countries
(60)
. Therefore, our study population probably reflects
the referral patterns inherent in the institutions from which our
patients were recruited because all eligible patients were approached
for the study, regardless of age, sex, ethnicity, or family history. As
a result, the results of the present study should be cautiously
extrapolated to other populations.
A large proportion of patients (26%) was classified as being at
increased risk for familial cancer syndromes. This is not surprising
given our broad criteria, some of which are less stringent than most
published definitions (17
, 23
, 26)
. Germ-line mutations
were identified in 5 of 38 (13%) tested cases. Four of the five
mutations were identified in patients with strong family histories
consistent with the corresponding cancer syndrome, with many affected
relatives, and satisfying multiple high- and intermediate-risk criteria
(Table 2
; Fig. 2
). The exception was a young Jewish patient who had
been adopted. The presence of a BRCA2 6174delT mutation in
this patient highlights the value of studying such individuals even in
the absence of a known family history.
Our findings are in keeping with previous studies in which germ-line p16 mutations have only been reported in pancreatic cancer patients with an associated family history of melanoma (15 , 19) . Of 38 high- and intermediate-risk patients tested, only 1 had a strong family history of melanoma, and this individual carried a p16 mutation. Two other patients with a single relative affected with melanoma, 24 patients with family histories of other cancer syndromes, and 11 patients with early-onset disease and double primary cancers did not harbor p16 mutations.
A family history of pancreatic cancer has been noted to predict the presence of a BRCA2 mutation (23 , 61) , and pancreatic cancers are more frequent in breast cancer families with BRCA2 mutations than in families without mutations (23) . Although other studies have suggested a role for germ-line BRCA2 mutations in the development of sporadic pancreatic cancer (41 , 42) , all BRCA2 mutations and the single BRCA1 mutation in our series were identified in patients with strong family histories for breast-ovarian cancer (with the exception of the adopted patient discussed earlier). Moreover, all these individuals were of Ashkenazi Jewish descent, although the BRCA2 6174delT mutation has independent origins in both Ashkenazi Jewish and non-Jewish populations (62) . No BRCA2 mutations were identified in patients satisfying criteria for other inherited cancer syndromes, including familial pancreatic cancer, or in the remaining patients in the intermediate-risk/nonfamilial group. No BRCA1 mutations were identified in four patients with histories suggestive of breast-ovarian cancer syndrome and lacking BRCA2 mutations, in whom the entire BRCA1 coding sequence was screened.
The BRCA2 6174delT founder mutation has been previously identified in 10% of Ashkenazi Jewish patients with apparently sporadic pancreatic cancers seen at our institution (42) . Of the 14 Jewish (including 1 Sephardic Jewish) patients in the present series, 3 satisfied criteria for breast-ovarian cancer syndrome (2 were classified as high risk, and 1 was classified as intermediate risk). In contrast, only 4 of 88 non-Jewish patients had a family history of breast-ovarian cancer (3 of 14 versus 4 of 88; Fisher’s exact P = 0.05). Of the remaining Jewish patients, one was characterized as high risk for familial pancreatic cancer, one was classified as intermediate risk for other familial cancer syndromes, two were classified as intermediate risk/nonfamilial, and seven were classified as low risk. Four of 13 (31%) Ashkenazi Jewish patients were mutation carriers (three BRCA2 6174delT mutations and one BRCA1 5382 insC mutation), and all those in whom family history was available met familial breast-ovarian cancer syndrome criteria. No BRCA2 mutations were identified in the remaining high-, intermediate- and low-risk Jewish patients in whom the entire gene was screened. These patients also did not harbor BRCA1 185delAG or 5382 insC mutations or p16 mutations.
Although we were unable to test all cases satisfying Mount Sinai criteria for HNPCC, the finding of intact hMSH2/hMLH1 expression in the three screened tumors is not entirely unexpected because other studies have previously demonstrated that germ-line mismatch repair mutations are uncommon in families not satisfying the Amsterdam criteria for HNPCC (63 , 64) . Although there was no significant difference in age of onset between p16, BRCA1, and BRCA2 mutation carriers and noncarriers, the former tended to be much younger, in keeping with other studies that noted that BRCA2 mutations associated pancreatic cancers (the majority of mutations detected in our series) occur at a younger age than compared to the general population (23 , 42) .
Inherited predisposition to cancer is often manifested by early age at onset or multiple primary tumors in the same individual (65) . We therefore analyzed the p16 and BRCA2 genes in an additional 11 patients with early-onset disease or multiple primary cancers without associated family histories of cancer (intermediate risk/nonfamilial). Only one patient (unknown family history) had a mutation, suggesting that these variables, in isolation, are probably not associated with an increased inherited risk of pancreatic cancer related to germ-line mutations in these two genes. A similar pattern has been noted for early-onset colorectal cancers (66) , but not for early-onset breast cancers (67 , 68) .
There are several possible explanations for the absence of detectable germ-line mutations in most high- and intermediate-risk cases. First, the methods used for mutation screening may not be adequately sensitive. Direct sequencing has been shown to accurately detect essentially all types of genetic variations (69) and was used to screen for p16 mutations. Exon 3 was not analyzed because no disease-associated germ-line mutations have been reported in this small fragment to date (15 , 70) . PTT (used to screen BRCA1 and BRCA2) may miss some missense mutations, gross chromosomal aberrations, and regulatory region silencing mutations where only the wild-type allele would be amplified or the mutant RNA is unstable. Although not as sensitive as sequencing the complete coding sequence, PTT has nevertheless been shown to be a rapid and efficient method of screening for mutations in the large BRCA1 and BRCA2 genes (44) .
More than 80% of tumors from patients with HNPCC are characterized by the presence of deficient mismatch repair as seen by the presence of alterations in simple repetitive DNA sequences called MSI. Because only 10–15% of sporadic colorectal cancers demonstrate MSI, examination of tumors for MSI is a useful complementary screening test to identify patients with possible germ-line mismatch repair gene mutations (31 , 71, 72, 73) However, tumor samples from many pancreatic cancer patients consist only of fine-needle aspiration biopsies (with no normal tissue for comparison) or have poor neoplastic cellularity (less than 50%) and are hence unreliable for MSI testing (74) . Therefore, immunohistochemistry for hMSH2 and hMLH1 expression was used to screen tumor samples from patients that met our HNPCC criteria for mismatch repair deficiency. We have recently shown immunohistochemistry to have 97% sensitivity and 100% specificity for identifying cancers with hMSH2 and hMLH1 inactivation (48) .
Taken together, although some patients may harbor mutations in exon 3 of p16 or in other mismatch repair genes, the low frequency of germ-line mutations in intermediate-risk patients in our series is unlikely to be due solely to the methods used for mutation screening.
Second, our family history classification may suffer from recall bias because patients with cancer diagnoses tend to recall familial cancers at a higher frequency than noncancer patients. They also selectively assign the primary site of their cancer to tumors in relatives, leading to the suggestion of an inherited cancer syndrome, when, in fact, none exists (14) . In our study, efforts were made to minimize this source of bias by confirming reported cancers in relatives. Third, the apparent familial aggregation of cancers in some families may represent chance association of sporadic cases. Fourth, young age of onset and multiple primary cancers by themselves are not likely to be important risk factors. Fifth, these families may harbor mutations in other genes predisposing to pancreatic cancer such as STK11 (75) and the cationic trypsinogen gene (76) , which were not tested in the present study. Sixth, and most likely, there are probably as yet unidentified genes predisposing to pancreatic cancer, particularly in patients meeting only familial pancreatic cancer criteria.
In summary, we report that 13% of our pancreatic cancer cases with personal or family histories suggestive of inherited predisposition occur in association with germ-line mutations of the p16, BRCA1, and BRCA2 genes. Importantly, 80% (four of five) of mutations were identified in Jewish patients, three of whom had family histories of breast-ovarian cancer, whereas only one p16 mutation was identified in a non-Jewish patient with a family history of melanoma. We conclude that known causes of genetic susceptibility are an important risk factor in a small proportion of pancreatic cancer patients, particularly when associated with a strong family history of familial cancer syndromes. The absence of identifiable mutations in a large proportion of seemingly high- and intermediate-risk cases suggests that there are as yet unidentified genes predisposing to this disease.
Our findings have several possible clinical implications. First, they stress the importance of a prospectively obtained family history in characterizing inherited cancer risk. Second, the ability to identify a population of individuals at increased risk for pancreatic cancer based on inherited mutations affords the opportunity to develop and validate novel screening protocols that would allow early detection and intervention at a potentially curable stage of disease, thereby improving survival. Based on our findings, we would suggest genetic screening only in those patients with significant family histories of familial cancer syndromes known to be associated with pancreatic adenocarcinoma. Genetic testing can also be offered to relatives because determining the carrier status of at-risk family members may lead to a significant reduction in anxiety for noncarriers. Furthermore, mutation carriers can undergo screening for melanoma and breast cancer according to current guidelines (77 , 78) . However, it is important to emphasize that there are currently no established clinical protocols for pancreatic cancer screening; hence, there is at least a potential psychological risk associated with the identification of carriers predisposed to a nearly fatal disease.
As a result of the above and the fact that the lifetime risk of pancreatic cancer in mutation carriers has not been defined, we strongly believe that genetic and clinical screening of pancreatic cancer patients and family members should only be conducted in research settings with appropriate pre- and post-test counseling. Lastly, our study indicates that further work is necessary to characterize other unknown inherited susceptibility genes in pancreatic cancer.
ACKNOWLEDGMENTS
We thank the patients and families who agreed to participate in
this study; the surgeons and medical and radiation oncologists for
assistance in the initial identification of study patients; Cecilia
Cabral-Bolarinho, John White, Xiu-Hong Shi, Susan Lau, and Hyeja Kim
for technical advice and assistance; and Lisa Madlensky for helpful
discussion.
|
FOOTNOTES |
|---|
1 Supported in part by the National Cancer
Institute of Canada [S. A. N., H. O., M. R., S. G.] with funds
provided by the Canadian Cancer Society; and by the Edward Christie
Stevens Fellowship and the Joseph M. West Family Memorial Fund
(Department of Postgraduate Medicine awards to [G. L.]. 
2 To whom requests for reprints should be
addressed, at Suite 1225, Mount Sinai Hospital, 600 University Avenue,
Toronto, Ontario, M5G 1X5 Canada. Phone: (416) 586-8550; Fax: (416)
586-8932; E-mail: sgallinger{at}mtsinai.on.ca
3 The abbreviations used are: CI, confidence
interval; HNPCC, hereditary nonpolyposis colorectal cancer; PTT,
protein truncation test; MSI, microsatellite instability.
4 H. Ozcelik, unpublished data.
Received 7/ 7/99. Accepted 11/12/99.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. W. Dewald, T. C. Smyrk, E. C. Thorland, R. R. McWilliams, D. L. Van Dyke, J. G. Keefe, K. J. Belongie, S. A. Smoley, D. L. Knutson, S. R. Fink, et al. Fluorescence In Situ Hybridization to Visualize Genetic Abnormalities in Interphase Cells of Acinar Cell Carcinoma, Ductal Adenocarcinoma, and Islet Cell Carcinoma of the Pancreas Mayo Clin. Proc., September 1, 2009; 84(9): 801 - 810. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. R. Ferrone, D. A. Levine, L. H. Tang, P. J. Allen, W. Jarnagin, M. F. Brennan, K. Offit, and M. E. Robson BRCA Germline Mutations in Jewish Patients With Pancreatic Adenocarcinoma J. Clin. Oncol., January 20, 2009; 27(3): 433 - 438. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. Wescott and A. K. Rustgi Pancreatic Cancer: Translating Lessons from Mouse Models and Hereditary Syndromes Cancer Prevention Research, December 1, 2008; 1(7): 503 - 506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P Ghaneh, E Costello, and J P Neoptolemos Biology and management of pancreatic cancer Postgrad. Med. J., September 1, 2008; 84(995): 478 - 497. [Full Text] [PDF]
|
|
|
|
 |
|
 R. E Brand, M. M Lerch, W. S Rubinstein, J. P Neoptolemos, D. C Whitcomb, R. H Hruban, T. A Brentnall, H. T Lynch, M. I Canto, and Participants of the Fourth International Symposium Advances in counselling and surveillance of patients at risk for pancreatic cancer Gut, October 1, 2007; 56(10): 1460 - 1469. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Ghaneh, E. Costello, and J. P Neoptolemos Biology and management of pancreatic cancer Gut, August 1, 2007; 56(8): 1134 - 1152. [Full Text] [PDF]
|
|
|
|
 |
|
 G. M. Petersen, M. de Andrade, M. Goggins, R. H. Hruban, M. Bondy, J. F. Korczak, S. Gallinger, H. T. Lynch, S. Syngal, K. G. Rabe, et al. Pancreatic cancer genetic epidemiology consortium. Cancer Epidemiol. Biomarkers Prev., April 1, 2006; 15(4): 704 - 710. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C J van Asperen, R M Brohet, E J Meijers-Heijboer, N Hoogerbrugge, S Verhoef, H F A Vasen, M G E M Ausems, F H Menko, E B Gomez Garcia, J G M Klijn, et al. Cancer risks in BRCA2 families: estimates for sites other than breast and ovary J. Med. Genet., September 1, 2005; 42(9): 711 - 719. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Garber and K. Offit Hereditary Cancer Predisposition Syndromes J. Clin. Oncol., January 10, 2005; 23(2): 276 - 292. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. B. Lowenfels and P. Maisonneuve Epidemiology and Prevention of Pancreatic Cancer Jpn. J. Clin. Oncol., May 1, 2004; 34(5): 238 - 244. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Gangi, J. G. Fletcher, M. A. Nathan, J. A. Christensen, W. S. Harmsen, B. S. Crownhart, and S. T. Chari Time Interval Between Abnormalities Seen on CT and the Clinical Diagnosis of Pancreatic Cancer: Retrospective Review of CT Scans Obtained Before Diagnosis Am. J. Roentgenol., April 1, 2004; 182(4): 897 - 903. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Liede, B. Y. Karlan, and S. A. Narod Cancer Risks for Male Carriers of Germline Mutations in BRCA1 or BRCA2: A Review of the Literature J. Clin. Oncol., February 15, 2004; 22(4): 735 - 742. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. F. Parker, S. R. Florell, A. Alexander, J. A. DiSario, P. J. Shami, and S. A. Leachman Pancreatic Carcinoma Surveillance in Patients With Familial Melanoma Arch Dermatol, August 1, 2003; 139(8): 1019 - 1025. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. M. Petersen and R. H. Hruban Familial Pancreatic Cancer: Where Are We in 2003? J Natl Cancer Inst, February 5, 2003; 95(3): 180 - 181. [Full Text] [PDF]
|
|
|
|
|
|
S. A. Hahn, B. Greenhalf, I. Ellis, M. Sina-Frey, H. Rieder, B. Korte, B. Gerdes, R. Kress, A. Ziegler, J. A. Raeburn, et al. BRCA2 Germline Mutations in Familial Pancreatic Carcinoma J Natl Cancer Inst, February 5, 2003; 95(3): 214 - 221. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Thompson, D. F. Easton, and the Breast Cancer Linkage Consortium Cancer Incidence in BRCA1 Mutation Carriers J Natl Cancer Inst, September 18, 2002; 94(18): 1358 - 1365. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. M. Murphy, K. A. Brune, C. Griffin, J. E. Sollenberger, G. M. Petersen, R. Bansal, R. H. Hruban, and S. E. Kern Evaluation of Candidate Genes MAP2K4, MADH4, ACVR1B, and BRCA2 in Familial Pancreatic Cancer: Deleterious BRCA2 Mutations in 17% Cancer Res., July 1, 2002; 62(13): 3789 - 3793. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F X Real, N Malats, G Lesca, M Porta, S Chopin, G M Lenoir, and O Sinilnikova Family history of cancer and germline BRCA2 mutations in sporadic exocrine pancreatic cancer Gut, May 1, 2002; 50(5): 653 - 657. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. M. Anderson and J. E. Harris Selected Features of Nonendocrine Pancreatic Cancer Experimental Biology and Medicine, June 1, 2001; 226(6): 521 - 537. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E EFTHIMIOU, T CRNOGORAC-JURCEVIC, N R LEMOINE, and T A BRENTNALL Inherited predisposition to pancreatic cancer Gut, February 1, 2001; 48(2): 143 - 147. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
