| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Departments of Medicine*
and
Pathology,
Veterans Affairs Medical Center
and Baylor College of Medicine, Houston, Texas; and the Guro
Hospital,
Korea University College of
Medicine, Seoul, Korea
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Microsatellite instability (MSI), a form of genomic instability, was first described in hereditary nonpolyposis colorectal cancer.9 Since then, MSI has been reported in a variety of familial and sporadic human cancers.10 Microsatellites are ubiquitous, short, repetitive DNA sequences widely and randomly distributed throughout the human genome, with unknown function.11 Microsatellite instability in tumors is identified when alleles of novel sizes are detected in microsatellite sequences derived from cancer DNA that are not present in normal tissues of the same individual. Gastric cancers, both familial and sporadic, have been shown to exhibit various levels of MSI (17%–59%).7,8,10
Intestinal metaplasia (IM) is generally believed to be a preneoplastic lesion of the stomach, which is epidemiologically linked to chronic H. pylori infection.12,13 Identification of genetic changes common to both gastric IM and carcinomas should broaden our understanding of the molecular pathways of gastric carcinogenesis. In contrast to colorectal cancer, current knowledge of these intermediate genetic changes is still superficial and largely limited to reports on mutations of p53 and APC genes in IM and dysplasia.14,15 Recently, several studies16-18 reported MSI in areas of IM from gastric cancer patients, suggesting that MSI may be an early event in the multistep progression of gastric carcinogenesis. However, the issue of whether foci of IM in individuals without cancer also harbor MSI was not addressed. Because most gastric cancer patients have evidence of IM in the stomach and not all patients with IM will develop cancer, this question has important biological implications and perhaps clinical applications. We examined IM tissues obtained from patients with gastric cancer and from individuals without gastric malignancy at the time of examination, to elucidate the significance and chronology of MSI in the multistep gastric carcinogenesis pathway.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Gastric tissue samples were selected from archival pathological specimens of patients undergoing upper gastrointestinal endoscopy in Guro Hospital, Seoul, Korea. Endoscopic biopsies were obtained by a standard mapping protocol that included six biopsies from corpus and six biopsies from gastric antrum. Three different groups of patients, including 30 patients with sporadic gastric cancer, 26 patients with peptic ulcer diseases (17 with gastric ulcers and 9 with duodenal ulcers), and 19 patients with chronic gastritis, were studied. All patients showed extensive metaplasia in the stomach. Ulcer patients also had evidence of chronic gastritis on histological examination.
Histological Examination
Gastric carcinomas were classified into diffuse and intestinal types as defined by Lauren.19 For nontumor tissues (IM and control), the Genta stain20 was used for grading of H. pylori infection, chronic and active inflammation, and IM according to the updated Sydney classification for gastritis.21 Samples showing IM were further stained with high iron diamine/Alcian blue, pH 2.5, for subtyping as described by Filipe et al.22 Type III or incomplete IM was diagnosed only if sulfomucin was detected in the cytoplasm of cells with absorptive cell morphology; otherwise IM was scored as a complete type. Active H. pylori infection was identified by the presence of typical organisms in any of the available gastric biopsies of the same patient by histology.
Preparation of Genomic DNA
Serial 5-µm-thick sections of selected tissue blocks were
obtained on glass slides, and the areas of interest were microdissected
after matching with an adjacent section stained with hematoxylin and
eosin or Genta stain. Microdissection was performed to enrich the DNA
content from the tumor or metaplastic tissues23
and to
minimize DNA contamination by other inflammatory or stromal cell
nuclei. For tumor samples, tissues were selected to contain more than
90% of the representative tumor in the section. One or two IM samples
were selected from biopsy sites that contained sufficient IM tissue for
microdissection (more than 70% of the gastric epithelium replaced by
IM on the tissue section), and typically they were associated with only
mild chronic inflammation and did not display atypical or dysplastic
features (Figure 1, A and B)
. IM tissues
from cancer patients were chosen from an area distant from dysplasia or
the tumor (embedded in different tissue blocks). Corresponding samples
of gastric mucosa that were free of tumor, metaplasia, and dysplasia
were obtained from each case as control tissue. Therefore,
cancer patients had three sets of samples (IM, tumor, and control)
whereas noncancer patients had two sets (IM and control). After tissue
deparaffinization, DNA was extracted with the QIAamp Tissue Kit
(Qiagen, Chatsworth, CA) following the manufacturer’s instructions.
|
All samples were analyzed by a set of eight microsatellite
markers, which included the five reference panel markers recommended by
the National Cancer Institute workshop on MSI for cancer detection and
familial predisposition.23
There were three mononucleotide
(A)n repeat markers (BAT25, BAT26, and BAT40) and
five dinucleotide (CA)n repeat markers (D2S123,
D5S346, D13S170, D17S250, and TP53).24
Oligonucleotides
were synthesized by Life Technologies Inc. (Gaitherburg, MD). One of
the oligonucleotide primers was end labeled with
-32P-labeled ATP, using T4 polynucleotide
kinase (Amersham, Arlington Heights, IL). Polymerase chain reaction
(PCR) amplifications were then carried out in 50-µl reaction volumes,
containing 1x PCR reaction buffer (GeneAmp; Perkin-Elmer, Branchburg,
NJ), 20 pmol each of deoxynucleotide triphosphate (dATP, dCTP, dTTP,
and dGTP; Promega, Madison, WI), 50 pmol of both labeled and unlabeled
primer, and 0.25 U of Taq polymerase (AmpliTaq Gold, Perkin
Elmer). PCR products were diluted in equal volumes of formamide gel
loading buffer (80% formamide, 0.1% xylene cyanol, 0.1% bromophenol
blue, 2 mmol/L ethylenediaminetetraacetic acid), denatured at 95°C,
and separated through 5.6 mol/L urea, 32% formamide, 7%
polyacrylamide gels.24
MSI-positive samples were confirmed
by 7% polyacrylamide and 5.6 mol/L urea sequencing gels. MSI was
defined as a band shift in either of the two alleles or the appearance
of a band with different size in the tumor or IM sample (Figure 2)
. All autoradiograms were independently
interpreted by two investigators (W. K. L. and A.
R. S.). In discordant cases, the opinion of a third observer was
obtained. PCR was repeated in samples displaying MSI, for confirmation
of the results. A specimen was defined as high-level MSI (MSI-H) if
more than 30% of the markers examined showed instability, low-level
MSI (MSI-L) if less than 30% of the markers displayed instability, and
microsatellite stable (MSS) if none of the markers exhibited
MSI.23
|
The 
2
or Fisher-exact test and unpaired
Student’s t-test were used for analysis of most of the
categorical or numerical data, respectively. In addition, the
Mann-Whitney U test was used in the comparison of
nonparametric differences between two groups, whereas the
Kruskal-Wallis test was applied in the analysis of differences between
three groups. A P value of less than 0.05 was considered
statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Thirty gastric cancer patients (21 males and 9 females; mean age 60.1 years, range 38–79 years) were studied. None of them had a family history of gastric cancer or hereditary nonpolyposis colorectal cancer syndrome.25 Of all carcinomas, 21 (70%) were of the intestinal type, and 9 (30%) were of the diffuse type. Tumors were located in the distal stomach or antrum in 16 cases (53%), and active H. pylori infection was detected in 27 cases (90%).
Among the 30 tumors examined, MSI-H was detected in 8 (26.7%) and
MSI-L in 15 tumor samples (50%). All MSI-H tumors were of intestinal
type and had evidence of an active H. pylori infection in
the stomach. Of eight tumors displaying high-level MSI, seven were
located in the distal stomach (Table 1)
.
H. pylori infection was more frequent in patients with MSI-L
than in those with MSS tumors (57% versus 100%;
P = 0.02). The mean ages of patients with MSI-H, MSI-L,
or MSS tumors were similar (Table 1)
.
|
A total of 75 IM samples (30 from patients with gastric cancer and
45 from noncancer patients) were studied. Type III IM was found in 40%
of the patients with gastric carcinoma and in 24% of the cancer-free
patients, a difference that did not reach statistical significance
(Table 3)
. Among noncancer patients, type
III IM was detected in 35% of patients with gastric ulcer, 22% with
duodenal ulcer, and 16% with simple chronic gastritis.
|
. Notably,
compared with tumor samples, IM specimens were more often MSS (52%
versus 23.3%; P = 0.009) and less
frequently showed high-level instability (26.7% versus
9.3%; P = 0.03; Table 2
). Moreover, IM samples from
cancer patients had a higher frequency of MSI-H when compared with IM
tissues from cancer-free patients, but the difference was not
statistically significant (13.3% versus 6.7%;
P = 0.4; Table 2
). Among the 30 cancer patients,
microsatellites were more frequently altered in IM coexisting with
tumors that showed instability (Figure 3)
. IM with instability at one or more
loci was found in 75% of the cases with MSI-H tumor and 60% of the
cases with MSI-L, but none was found in the patients with MSS tumors.
In addition, IM samples from patients with MSI-H tumors had
significantly more altered loci than IM samples from patients with
MSI-L or MSS tumors (P = 0.003).
|
|
. Other gastric histological
features, such as degree of acute inflammation, chronic inflammation,
severity, and subtypes of IM, were not significantly different in
patients with MSI-H, MSI-L, or MSS IM. The mean ages of the three
groups of patients were similar. It is important that, when
microsatellite loci were altered in both tumor and IM, the patterns of
affected markers were usually similar; however, discrepancies were
observed in three cases (Table 4
and
Figure 2
).
|
For each marker analyzed, MSI was more frequently detected in
tumor tissues than in IM (Figure 4)
. In
general, dinucleotide repeat markers were more frequently altered than
mononucleotide repeat markers, and alterations in the BAT26 and TP53
markers were significantly more frequent in tumor than in IM
(P = 0.02). The results of our expanded panel of
eight microsatellite markers and the NCI-recommended panel are listed
in Table 5
. The numbers of patients
classified into MSI-H were similar in both panels, but the number of
cases with low-level instability increased when the expanded panel was
used. Interestingly, when instability in the BAT markers was compared
with all other markers, slightly over one third of the MSI-H tumors and
IM tissues had BAT instability, and a minority of the MSI-L tumors and
IM tissues displayed BAT alterations (Table 6)
.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Together, these findings suggest an
early and important role of MSI in the gastric carcinogenesis pathway,
and they may define a subset of individuals particularly susceptible to
gastric cancer.
Although previous reports16-18
had addressed the issue of
MSI in premalignant gastric lesions, the conclusions were based on
relatively small numbers of cancer patients, and analyses of IM in
patients without gastric cancer were not performed. Two studies from
Japan16,17
examined IM tissues from patients with
intestinal-type gastric cancer and detected MSI in 33% (3 of 9) and
27% (4 of 15) IM tissues, respectively. All cases showed a single
altered locus that was identical in IM and tumor, and MSI was limited
to incomplete-type IM. In the current study, microsatellites were
similarly affected in both complete and incomplete IM subtypes (Table 3)
. Hamamoto et al17
examined multiple sections
from gastrectomy specimens and demonstrated the topographical
distribution of microsatellite alterations at a single locus (D1S191)
in different IM samples surrounding gastric cancer. In contrast, we
identified MSI in IM tissues from gastric endoscopic biopsies that were
remote from the primary tumor or in patients without cancer, indicating
a potentially significant role of MSI in premalignant metaplastic areas
that are not in close proximity to the tumor.
In contrast to gastric tumors without MSI, MSI-H gastric tumors tended to exhibit characteristic features such as predominant distal location and were of intestinal type, similar to a previous report.8 The pathogenetic mechanisms of MSI-H tumors are better defined than those of MSI-L and MSS tumors. MSI-H tumors may harbor frameshift mutations in coding regions of some cancer-related genes, such as BAX, IGFRII, TGFßRII, hMSH3, and hMSH6,18,26,27 thereby favoring cancer development. The clinical significance of MSI-L in tumors is less well understood. In the current study, low levels of instability were frequently detected in both tumors and IM samples, which may be attributed to the use of the expanded microsatellite markers panel. BAT or mononucleotide marker alterations are good indicators of MSI-H colorectal cancer with underlying DNA mismatch repair deficiencies, and discordances between mononucleotide simple repeats and dinucleotide repeats appear to be unusual.23,28,29 Slightly over one third of the MSI-H gastric tumors and IM tissues had alterations in the BAT markers. Therefore, the majority of MSI-H cases were scored on their dinucleotide repeat instability. The reason for the discordance in instability between the BAT loci and the dinucleotide loci is not clear.
Interestingly, the expression of DNA mismatch repair proteins in mismatch-competent cells might be transiently suppressed in the presence of oxidative stress.30 In the gastric mucosa, reactive oxygen species are commonly released in inflamed gastric mucosa as a result of chronic H. pylori infection, and they could be responsible for the DNA mismatch repair deficiency underlying some MSI-H gastric cancers.31 All tumor and IM samples with MSI-H in this study had evidence of active H. pylori infection in the stomach. Whether inflammation resulting from chronic H. pylori infection can account for the frequent low-level instability and predominant dinucleotide repeat alterations might deserve further evaluation. Nonetheless, it is tempting to speculate that chronic H. pylori infection, by its effects on cell proliferation and apoptosis,32,33 may accelerate the accumulation of genetic alterations in the gastric mucosa that would then be carried on to IM and cancer. This was supported by the observation of Wu et al, who reported a significant association between H. pylori infection and gastric cancer with MSI-H.34 It is interesting that H. pylori was present in the great majority of stomachs with IM (MSI-positive and MSS) and in 100% of patients with MSI-positive tumors, but was absent in about 50% of the stomachs with MSS tumors. This difference can be explained by several possibilities: 1) MSS tumors might develop through molecular pathways that characterize a specific background gastric mucosa that constitutes an unfavorable environment for H. pylori; 2) patients with MSI-positive tumors might be infected with more virulent H. pylori strains; 3) host factor(s) determining susceptibility to MSI development might affect the ability of H. pylori to persist in the gastric environment of patients that ultimately develop MSI-positive tumors. A possibly important role of host susceptibility factors in the development of MSI-positive gastric cancer was suggested by previous studies recognizing a significantly higher frequency of MSI in gastric cancer samples from Korean individuals when compared with tumors from American or Colombian patients.35
Altered microsatellites have been identified in Barrett’s esophagus36 and non-neoplastic colonic mucosa of ulcerative colitis patients.37 Together with the findings of MSI in IM, altered microsatellites have been detected in the most common premalignant gastrointestinal lesions and are likely to play an early role in the carcinogenesis pathway of gastrointestinal malignancies. Although gastric cancer is a common disease, molecular markers for early diagnosis of the disease are lacking. Given the early involvement of MSI in the multistep gastric carcinogenesis model, detection of MSI may serve as a surrogate marker for the risk of gastric cancer development. It might be helpful to identify high-risk patients, by determining MSI of preneoplastic lesions, such that close monitoring or potential intervention can be instituted. Because the majority of patients with IM will not progress to cancer and only a proportion of these patients harbored MSI, it is conceivable that patients with IM displaying MSI are at greater risk of developing gastric cancer than those without instability.
In conclusion, we have demonstrated a high frequency of MSI in gastric IM from patients with and without gastric cancer. Taking into consideration the progressive increase in MSI frequency from premalignant to malignant lesions, our results suggest the early involvement and continuous accumulation of MSI in gastric cells that have entered the multistep gastric carcinogenesis pathway. The role of detection of MSI in premalignant gastric lesions as a surrogate marker of risk of gastric cancer development warrants further investigation.
|
Footnotes |
|---|
Supported by the Department of Veterans Affairs and by generous support from Hilda Schwartz. W. K. L. is partly supported by the Li Po Chun Charitable Trust Fund Overseas Postgraduate Scholarships of Hong Kong SAR.
Accepted for publication November 6, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  G. Capella, G. Pera, N. Sala, A. Agudo, F. Rico, G. Del Giudicce, M. Plebani, D. Palli, H. Boeing, H B. Bueno-de-Mesquita, et al. DNA repair polymorphisms and the risk of stomach adenocarcinoma and severe chronic gastritis in the EPIC-EURGAST study Int. J. Epidemiol., December 1, 2008; 37(6): 1316 - 1325. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. K. Leung, E. K.W. Ng, F. K.L. Chan, W. Y. Chan, K.-f. Chan, A. C.M. Auyeung, C. C.H. Lam, J. Y.W. Lau, and J. J.Y. Sung Effects of Long-term Rofecoxib on Gastric Intestinal Metaplasia: Results of a Randomized Controlled Trial Clin. Cancer Res., August 1, 2006; 12(15): 4766 - 4772. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Rugge, G Bersani, R Bertorelle, G Pennelli, V M Russo, F Farinati, D Bartolini, M Cassaro, and V Alvisi Microsatellite instability and gastric non-invasive neoplasia in a high risk population in Cesena, Italy J. Clin. Pathol., August 1, 2005; 58(8): 805 - 810. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. An, I.-S. Choi, J. C. Yao, S. Worah, K. Xie, P. F. Mansfield, J. A. Ajani, A. Rashid, S. R. Hamilton, and T.-T. Wu Prognostic Significance of CpG Island Methylator Phenotype and Microsatellite Instability in Gastric Carcinoma Clin. Cancer Res., January 15, 2005; 11(2): 656 - 663. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Akiyama, C. Maesawa, S. Ogasawara, M. Terashima, and T. Masuda Cell-Type-Specific Repression of the Maspin Gene Is Disrupted Frequently by Demethylation at the Promoter Region in Gastric Intestinal Metaplasia and Cancer Cells Am. J. Pathol., November 1, 2003; 163(5): 1911 - 1919. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M M Walker Is intestinal metaplasia of the stomach reversible? Gut, January 1, 2003; 52(1): 1 - 4. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Hartmann, L. Zanardo, T. Bocker-Edmonston, H. Blaszyk, W. Dietmaier, R. Stoehr, J. C. Cheville, K. Junker, W. Wieland, R. Knuechel, et al. Frequent Microsatellite Instability in Sporadic Tumors of the Upper Urinary Tract Cancer Res., December 1, 2002; 62(23): 6796 - 6802. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-H. Lee, S. C. Abraham, H.-S. Kim, J.-H. Nam, C. Choi, M.-C. Lee, C.-S. Park, S.-W. Juhng, A. Rashid, S. R. Hamilton, et al. Inverse Relationship between APC Gene Mutation in Gastric Adenomas and Development of Adenocarcinoma Am. J. Pathol., August 1, 2002; 161(2): 611 - 618. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. K. Sood, R. Holmes, M. J. C. Hendrix, and R. E. Buller Application of the National Cancer Institute International Criteria for Determination of Microsatellite Instability in Ovarian Cancer Cancer Res., June 1, 2001; 61(11): 4371 - 4374. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
