Mucinous Carcinomas of the Colon and Rectum Show
Higher Rates of Microsatellite Instability and Lower
Rates of Chromosomal Instability
A Study Matched for T Classification and Tumor Location
Yoshihiro Kazama, M.D.
Toshiaki Watanabe, Ph.D.
Takamitsu Kanazawa, M.D.
Tomohiro Tada, M.D.
Junichiro Tanaka, M.D.
Hirokazu Nagawa, Ph.D.
Department of Surgical Oncology, University of
Tokyo, Tokyo, Japan.
Address for reprints: Yoshihiro Kazama, M.D., De-
partment of Surgical Oncology, University of To-
kyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, Japan. Fax:
(011) 81-3-3811-6822. E-mail: ykazama-tky@
Received September 3, 2004; revision received
December 26, 2004;accepted January 10, 2005.
BACKGROUND. The clinicopathologic significance of mucinous carcinomas (Muc)
of the colon and rectum has been widely discussed, but there have been few
studies on Muc regarding genetic and epigenetic alterations. The current study
analyzed genetic and epigenetic alterations of Muc to clarify their differences from
well differentiated adenocarcinomas (WD).
METHODS. Thirty-nine cases of Muc and 39 cases of WD were investigated. Cases
of WD were matched with cases of Muc for T classification and tumor location.
Microsatellite instability (MSI) status and loss of heterozygosity (LOH) of four loci
(2p, 5q, 17p, 18q) were evaluated. The methylation status of the hMLH1 promoter
region in Muc was also examined.
RESULTS. “MSI tumors” were defined as those that showed MSI-high, and “chro-
mosomal instability (CIN) tumors” were defined as those that showed LOH but not
MSI-high. MSI tumors were significantly more frequent in Muc (30.8%) than in WD
(5.1%). CIN tumors were significantly less frequent in Muc (53.8%) than in WD
(87.2%). In Muc, MSI tumors were significantly more frequent in the proximal
colon (55.6%) than in the distal colon (9.5%). Also, methylation of the hMLH1
promoter region in Muc was significantly more frequent in MSI tumors (83.3%)
than in CIN tumors (27.8%) (P ? 0.0077).
CONCLUSIONS. When matched for T classification and tumor location, Muc shows
higher rates of MSI and lower rates of CIN than WD.. Muc shows different
characteristics according to tumor location, and methylation of the hMLH1 pro-
moter region strongly correlates with Muc tumors showing MSI. Cancer 2005;103:
2023–9. © 2005 American Cancer Society.
KEYWORDS: mucinous carcinoma, colorectal cancer, microsatellite instability, chro-
mosomal instability, hMLH1, methylation, T classification, tumor location.
between 5% and 15% of all colorectal cancers.1,2The clinicopatho-
logic significance of Muc has been widely discussed,2–5but there have
been few studies on Muc regarding genetic alterations, partly because
the number of cases is too small.
Colorectal cancers can progress through two pathways of
genomic instability: microsatellite instability (MSI) and chromosomal
instability (CIN).6Colorectal cancers with MSI often present a Muc
phenotype,7but, to the best of our knowledge, there have been no
studies that clearly indicate that Muc shows MSI more frequently
than well differentiated adenocarcinomas (WD), which account for
ucinous carcinomas of the colon and rectum (Muc) are charac-
terized by an extracellular mucin production and account for
© 2005 American Cancer Society
Published online 5 April 2005 in Wiley InterScience (www.interscience.wiley.com).
the majority of colorectal cancers. Further, there have
been no studies that compared the frequency of CIN
in Muc with that in WD. Such studies are required to
investigate the characteristics of genomic instability in
The methylation of the promoter region of some
specific genes such as hMLH1, p16, MGMT and APC
may be involved in colorectal carcinogenesis. Recent
studies indicate that methylation of the promoter re-
gion of hMLH1 is responsible for most, if not all of the
sporadic colorectal cancers with MSI.8–11However, in
the case of Muc, few reports on this methylation are
The aim of the current study is to analyze genetic
and epigenetic alterations of Muc and clarify their
differences from WD matched for T classification and
tumor location. Muc and WD were matched for T
classification and tumor location as these clinicopath-
ologic features influence MSI and CIN status.11,12To
the best of our knowledge, this is the first report on
Muc genetic alterations that used a control group
matched for clinicopathologic features. Our group
evaluated MSI status and the LOH of four loci (2p, 5q,
17p, 18q) that have been reported in most cases of
colorectal carcinomas, and we investigated the char-
acteristics of genomic instability. In addition, as an
epigenetic alteration of Muc, the methylation status of
the hMLH1 promoter region was evaluated.
MATERIALS AND METHODS
For the current study, our group selected and investi-
gated 78 pairs of matched normal colonic mucosa and
tumor specimens from colorectal carcinoma surgically
resected at the Department of Surgical Oncology, To-
kyo University Hospital between 1990 and 2003. This
study was approved by the Ethics Committee of Tokyo
University Hospital. Included patients were consid-
ered to have sporadic tumors because their clinical
presentation and family history did not suggest a di-
agnosis of either familial adenomatous polyposis13
or hereditary nonpolyposis colorectal carcinoma
(HNPCC).14Thirty-nine of the tumors were diagnosed
as Muc, whereas the other tumors were WD. Samples
were classified according to the 1989 World Health
Organization (WHO) classification. The cases of WD
were matched with the cases of Muc for T classifica-
tion and tumor location (Table1). T classification was
according to the TNM classification of malignant tu-
mors (5th edition, 1997).15Six pairs of Muc and 15
pairs of WD were flash frozen in liquid nitrogen and
stored at ?80 °C until analysis. The others were for-
malin-fixed, paraffin-embedded samples. In the fixed
samples, tissues were microdissected by the following
method. Several glands in the neoplastic lesions and
adjacent normal mucosa were carefully dissected mi-
croscopically from a 20 ?m-thick section of paraffin-
embedded block, with reference to the adjacent he-
matoxylin and eosin (H & E)-stained section. DNA was
extracted from each specimen with DNAeasy kit (Qia-
gen, Tokyo, Japan) after deparaffinization by sodium
dodecyl sulfate(SDS) and proteinase-K.
Allelotyping PCR and Detecting Allelic Loss of Markers
DNA from microdissected tumor specimens and flash-
frozen samples were used for allelotyping PCR using
fluorescent-labeled primers for four loci (5q:D5S346,
D5S107, D5S2055; 2p:D2S123; 17p:p53penta, D17S796;
18q: D18S474). In addition to these primer sets,
D17S250, BAT25 and BAT26 were used to determine
MSI status.16Primer sequences for D5S346, D2S123,
and D17S250 have been described previously.17
Primer sequences for other loci were obtained
from the Genome Database and the database of the
Whitehead Institute. All primer pairs were end labeled
with fluorochromes 6-carboxy-fluorescein (FAM),
or NED (Applied Biosystems, Tokyo, Japan). The PCR
was performed in 10 ?L reaction volumes containing
10 ? PCR Gold Buffer (Applied Biosystems, Tokyo,
Japan), 2.5 mM MgCl2,200 ?M deoxynucleotide
triphosphates mixture, 0.5 ?M of each primer, and
20–40 ng of extracted DNA and 0.4 units of AmpliTaq
Gold? DNA polymerase (Applied Biosystems, Foster
City, CA). The DNA was amplified in a thermal cycler
(GeneAmp? PCR system 9700, Applied Biosystems)
and PCR was performed according to the protocol as
follows: 10 minutes at 95 °C for polymerase activation;
40 cycles at 94 °C for 30 seconds, 56 °C for 1 minute
Clinicopathologic Features of Colorectal Carcinomas
Muc (n ? 39)WD (n ? 39)
62.9 ? 13.3 64.0 ? 9.3
Muc: Mucinous carcinomas; WD: Well differentiated adenocarcinomas; Proximal colon: cecum, as-
cending colon, and transverse colon; Distal colon: descending colon, sigmoid colon, and rectum.
2024 CANCER May 15, 2005 / Volume 103 / Number 10
and 72 °C for 1minute; then an additional 30 minutes
at 70 °C. After denaturation by heating at 95 °C for 5
minutes, PCR products were electrophoresed and an-
alyzed on an automated sequencer [ABI PRISM? 3100
genetic analyzer (Applied Biosystems)], and the fluo-
rescent signals from alleles of different size were re-
corded and analyzed using GeneScan? version 3.1 and
Genotyper? version 2.1 software (both software pack-
ages, Applied Biosystems).
Assessment of MSI and LOH
We classified tumors as MSI positive when the PCR
product of tumor DNA revealed at least one peak that
was not visible in the PCR product of the correspond-
ing normal tissue DNA (Fig. 1). We used the criteria of
the National Cancer Institute) workshop16to classify
MSI and microsatellite stability (MSS) using the five
primers that are commonly accepted in estimating
MSI status. MSI-high is determined if two or more of
the five markers exhibit instability. MSI-low is deter-
mined if only one of the five markers exhibits insta-
For each marker, the normal DNA sample was
used to determine the allele size for the corresponding
subject. For heterozygous samples, alleles were de-
fined as the two peaks of the greatest height. Allele
ratios were calculated as described previously18, 19(Fig.
1). Because incomplete losses are commonly observed
and may reflect either normal cell contamination or
tumor heterogeneity, a ratio of ? 0.60 was considered
to be indicative of LOH. Homozygous cases were con-
sidered not informative for LOH.
Methylation-Specific PCR (MSP)
MSP exploits the effect of sodium bisulfite on DNA,
which is the efficient conversion of unmethylated cy-
tosine to uracil without a change in methylated cyto-
sine. Consequently, after treatment, the methylated
and unmethylated alleles have different sequences
that can be used to design allele-specific primers.
Genomic DNA was modified with sodium bisulfite as
The amount of DNA was small, and its condition
was probably poor for long-term formalin fixation, so
our group first amplified modified DNAs with flanking
PCR primers that amplify bisulfite-modified DNA but
do not preferentially amplify methylated or unmeth-
ylated DNA.8PCR was performed in 10 ?L reaction
volumes containing 10 ? PCR Gold Buffer(Applied
Biosystems), 2.5 mM MgCl2,200 ?M deoxynucleotide
triphosphates mixture, 0.5 ?M of each primer, 2 ?L of
modified DNA and 0.4 units of Amplitaq Gold DNA
polymerase(Applied Biosystems). Primer sequences of
the hMLH1 promoter region were 5?-GAGTAGTTTT-
TTTTTTAGGAGTGAAG-3? (sense) and 5?-AAAAACTA-
TAAAACCCTATACCTAATCTA-3? (antisense). The first
PCR was performed according to the protocol as fol-
lows: 10minutes at 95 °C for polymerase activation; 40
cycles at 94 °C for 30 seconds, 54 °C for 1 minute and
72 °C for 1minute; then an additional 30 minutes at 70
°C. Then 1 ?L from the first PCR product was ampli-
fied with primers as described below. Primer se-
quences for unmethylated reaction were 5?-TTTTGAT-
GTAGATGTTTTATTAGGGTTGT-3? (sense) and 5?-
ACCACCTCATCATAACTACCCACA-3? (antisense), and
for methylated reaction, they were 5?-ACGTAGAC-
GTTTTATTAGGGTCGC-3? (sense) and 5?-CCTCATCG-
TAACTACCCGCG-3? (antisense). The second PCR was
performed according to the protocol as follows:
10minutes at 95 °C for polymerase activation; 40 cycles
at 94 °C for 30 seconds, 55 °C for 1 minute and 72 °C
for 1minute; then an additional 30 minutes at 70 °C.
Primer pairs were end labeled with fluorochromes
FAM or NED, and PCR products were evaluated with
an ABI Genetic Analyzer. We determined the methyl-
ation status when a distinct peak of methylated or
unmethylated product was displayed. The locus of
hMLH1 was classified as unmethylated if the height of
the methylated band was ? 10% of the unmethylated
band, or methylated if the height of the methylated
band was ? 10% of the unmethylated band. All PCRs
were performed with positive controls for both un-
methylated and methylated alleles. The positive con-
trol for methylated alleles consisted of placental DNA
treated in vitro with excess Sss I methyltransferase [M.
Sss I (New England Biolabs, Beverly, MA)], generating
DNA completely methylated at all CpG sites, and this
was then treated with sodium bisulfite. The positive
control for unmethylated alleles consisted of sodium
bisulfite-treated placental DNA. The assays were per-
formed at least twice to ensure reproducibility of re-
The Fisher exact test was used to determine statistical
significance of observed differences between groups.
The value of P ? 0.05 was considered statistically
Characteristics Of Genomic Instability
Our group examined the characteristics of the
genomic instability of Muc and WD. Previous research
had indicated that MSS and MSI-L tumors have a
common molecular background,21so we defined “MSI
tumors” as those that showed MSI-H.
The consequence of CIN is an imbalance in the
number of chromosomes per cell (aneuploidy) and an
Matched Study for Mucinous Carcinomas/Kazama et al. 2025
FIGURE 1. Case examples show MSI,
LOH, and methylated band. (A) shows
extra bands in tumor, and (B) shows loss
of one allele in tumor. In these cases,
the tumor DNA sample is shown in the
upper panel, and the corresponding nor-
mal DNA is shown in the lower panel. (C)
shows unmethylated band in one tumor
and methylated band in another tumor.
2026CANCER May 15, 2005 / Volume 103 / Number 10
enhanced rate of LOH.22To investigate allelic losses by
LOH, we used seven polymorphic markers mapped
closely to key tumor suppressor genes that are be-
lieved to be lost during colon carcinogenesis.23We
therefore defined “CIN tumors” as those that showed
LOH at one or more of these sites but did not show
In Muc, 12 cases (30.8%) were MSI tumors, 21
cases (53.8%) were CIN tumors, and 6 cases (15.4%)
were classified “other tumors” that showed neither
LOH nor MSI-H (Fig. 2). MSI tumors were significantly
more frequent in Muc than in WD (5.1%) (P ? 0.0032),
whereas CIN tumors were significantly less frequent in
Muc than in WD (87.2%) (P ? 0.0024).
It is noteworthy that Muc showed different char-
acteristics according to tumor location (Fig. 3). In
Muc, MSI tumors were significantly more frequent in
the proximal colon (55.6%) than in the distal colon
(9.5%) (P ? 0.042). CIN tumors were less frequent in
the proximal colon (38.9%) than in the distal colon
(66.7%), although no significant difference was found
(P ? 0.11).
In WD, all MSI tumors were located in the proxi-
mal colon. However, the frequency of CIN tumors was
high both in the proximal colon (83.3%) and in the
distal colon (85.7%).
Three cases of Muc could not be evaluated according
to the methylation status of the hMLH1 promoter
FIGURE 2. Genomic instability patterns of Muc and WD are summarized. MSI tumors were those that showed MSI-H, CIN tumors were those that showed LOH
but not MSI-H, whereas Other tumors were those that showed neither LOH nor MSI-H.
FIGURE 3. Genomic instability patterns of Muc in the proximal colon and the distal colon are summarized.
Matched Study for Mucinous Carcinomas/Kazama et al.2027
region. As shown in Table 2, 15 of 36 cases (41.7%) of
Muc exhibited methylation of the hMLH1 promoter
region. This methylation was significantly more fre-
quent in MSI tumors (83.3%) than in CIN tumors
(27.8%) (P ? 0.0077) and other tumors (0%) (P
? 0.0042). It was also significantly more frequent in
the proximal colon (61.1%) than in the distal colon
(22.2%) (P ? 0.040) (Table 3).
The current study is the first to clearly indicate that the
Muc phenotype shows MSI more frequently than the
WD phenotype. Only one previous study reported that
Muc showed MSI more frequently than nonmucinous
carcinomas of the colon and rectum (non-Muc),24but
in that study, the “non-Muc” group was not matched
for tumor location with the “Muc” group. This may
have some influence on the comparison of MSI status,
because Muc is located in the proximal colon more
frequently than non-Muc,2,4,5and colorectal cancers
located in the proximal colon show a high frequency
of MSI.11In our study, we took this tendency into
consideration. The cases of WD (control group) were
matched for tumor location with the cases of Muc. In
spite of this matching, Muc showed MSI significantly
more frequently than WD, so our study indicated that
the Muc tumorigenesis strongly correlates with MSI.
The current study clearly indicated that the Muc
phenotype shows CIN less frequently than the WD
phenotype. Previously published studies had indi-
cated that notable correlations are found between the
extension of carcinomas into the colonic wall and the
presence of LOH,12so this may have some influence
on the comparison of LOH status between Muc and
WD, because Muc invades more extensively at the
time of discovery than non-Muc.2,3Therefore, in our
study, the cases of WD (control group) were matched
for T classification of tumors with cases of Muc. This
matching made it clear that the Muc phenotype shows
CIN less frequently than the WD phenotype. This cor-
responds to a previously reported study in which LOH
on chromosome 17p and 18q was less frequent in Muc
than in non-Muc.25
Our study demonstrates that Muc shows different
characteristics according to tumor location. In our
study, Muc in the proximal colon showed MSI signif-
icantly more frequently than in the distal colon. One
previous study reported that MSI-positive Muc did not
show a proclivity for the proximal colon.24This dis-
crepancy may be caused by selecting different micro-
satellite markers to determine MSI status. In our
study, we used the criteria of the National Cancer
Institute16workshop using the five primers that are
commonly accepted in estimating MSI status, but in
that previous study these five primers were not used.
In our study, Muc in the proximal colon showed
CIN less frequently than in the distal colon but with no
significant difference. This corresponds to a previous
study that reported that LOH on chromosome 17p or
18q was generally less frequent in the proximal colon
than in the distal colon.25A further study using a large
number of tumors may be required to show a signifi-
In our study, methylation of the hMLH1 promoter
region was found in 15 of 36 cases (41.7%) of Muc, and
this strongly correlated with MSI tumors. Also, meth-
ylation of this region was found significantly more
frequently in the proximal colon than in the distal
colon. These data underline the finding that Muc in
the proximal colon shows MSI-H more frequently
than in the distal colon, because methylation of the
hMLH1 promoter region is responsible for most of the
sporadic colorectal cancers with MSI-H.8–11
A previous study reported that MSI-H was found
in 14 of 14 (100%) colorectal carcinomas with full
methylation of the hMLH1 promoter region, whereas
it was found in only 2 of 26 (7.7%) colorectal carcino-
mas with partial methylation of this region, and that
full methylation of this region may play a crucial role
in the carcinogenesis of MSI-H colorectal carcino-
mas.11In our study, 11 (73.3%) of 15 cases of Muc with
methylation of this region showed MSI-H. This sug-
gests that most of the methylation of the hMLH1 pro-
moter region in Muc may be full methylation, and this
may cause a high frequency of MSI-H in Muc.
In summary, we showed that MSI was signifi-
cantly more frequent in Muc than in WD matched for
T classification and tumor location. We also showed
Correlation between Methylation of hMLH1 Promoter Region and
Genomic Instability in the Cases of Muc
P ? 0.007
P ? 0.0042
Correlation between Methylation of hMLH1 Promoter Region and
Tumor Location in the Cases of Muc
14P ? 0.040
2028 CANCER May 15, 2005 / Volume 103 / Number 10
that CIN was significantly less frequent in Muc than in
matched WD. In addition, methylation of the hMLH1
promoter region strongly correlated with Muc tumors
showing MSI. Further, Muc showed different charac-
teristics according to tumor location. In the proximal
colon, ? 50% of the cases of Muc showed MSI,
whereas in the distal colon, 67% of such cases showed
CIN and only 10% showed MSI.
1. Symonds DA, Vickery AL. Mucinous carcinoma of the colon
and rectum. Cancer. 1976;37:1891–1900.
2.Nozoe T, Anai H, Nasu S, Sugimachi K. Clinicopathological
characteristics of mucinous carcinoma of the colon and
rectum. J Surg Oncol. 2000;75:103–107.
3.Green JB, Timmcke AE, Mitchell WT, Hicks TC, Gathright JB,
Jr., Ray JE. Mucinous carcinoma—just another colon can-
cer? Dis Colon Rectum. 1993;36:49–54.
4.Hanski C. Is mucinous carcinoma of the colorectum a dis-
tinct genetic entity? Br J Cancer. 1995;72:1350–1356.
5.Umpleby HC, Ranson DL, Williamson RC. Peculiarities of
mucinous colorectal carcinoma. Br J Surg. 1985;72:715–718.
6. Lengauer C, Kinzler KW, Vogelstein B. Genetic instability in
colorectal cancers. Nature. 1997;386:623–627.
7.Kim H, Jen J, Vogelstein B, Hamilton SR. Clinical and patho-
logical characteristics of sporadic colorectal carcinomas
with DNA replication errors in microsatellite sequences.
Am J Pathol. 1994;145:148–156.
8. Herman JG, Umar A, Polyak K, Graff JR, Ahuja N, Issa JP, et
al. Incidence and functional consequences of hMLH1 pro-
moter hypermethylation in colorectal carcinoma. Proc Natl
Acad Sci U S A 1998;95:6870–6875.
9. Cunningham JM, Christensen ER, Tester DJ, Kim CY, Roche
PC, Burgart LJ, et al. Hypermethylation of the hMLH1 pro-
moter in colon cancer with microsatellite instability. Cancer
10. Kuismanen SA, Holmberg MT, Salovaara R, Schweizer P,
Aaltonen LA, de La Chapelle A, et al. Epigenetic phenotypes
distinguish microsatellite-stable and -unstable colorectal
cancers. Proc Natl Acad Sci U S A. 1999;96:12661–12666.
11. Miyakura Y, Sugano K, Konishi F, Ichikawa A, Maekawa M,
Shitoh K, et al. Extensive methylation of hMLH1 promoter
region predominates in proximal colon cancer with micro-
satellite instability. Gastroenterology. 2001;121:1300–1309.
12. Iino H, Fukayama M, Maeda Y, Koike M, Mori T, Takahashi
T, et al. Molecular genetics for clinical management of colo-
rectal carcinoma. 17p, 18q, and 22q loss of heterozygosity
and decreased DCC expression are correlated with the met-
astatic potential. Cancer 1994;73:1324–1331.
13. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal
cancer. Cell. 1996;87:159–170.
14. Vasen HF, Watson P, Mecklin JP, Lynch HT. New clinical
criteria for hereditary nonpolyposis colorectal cancer
(HNPCC, Lynch syndrome) proposed by the International
Collaborative group on HNPCC. Gastroenterology. 1999;116:
15. Sobin LH, Fleming ID. TNM Classification of Malignant
Tumors, 5th edition. Union Internationale Contre le Cancer
and the American Joint Committee on Cancer. Cancer 1997;
16. Boland CR, Thibodeau SN, Hamilton SR, Sidransky D, Eshle-
man JR, Burt RW, et al. A National Cancer Institute Work-
shop on Microsatellite Instability for cancer detection and
familial predisposition: development of international crite-
ria for the determination of microsatellite instability in colo-
rectal cancer. Cancer Res 1998;58:5248–5257.
17. Umetani N, Sasaki S, Watanabe T, Ishigami H, Ueda E,
Nagawa H. Diagnostic primer sets for microsatellite insta-
bility optimized for a minimal amount of damaged DNA
from colorectal tissue samples. Ann Surg Oncol. 2000;7:276–
18. Cawkwell L, Bell SM, Lewis FA, Dixon MF, Taylor GR, Quirke
P. Rapid detection of allele loss in colorectal tumours using
microsatellites and fluorescent DNA technology. Br J Can-
19. Medintz IL, Lee CC, Wong WW, Pirkola K, Sidransky D,
Mathies RA. Loss of heterozygosity assay for molecular de-
tection of cancer using energy-transfer primers and capil-
lary array electrophoresis. Genome Res. 2000;10:1211–1218.
20. Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB.
Methylation-specific PCR: a novel PCR assay for methyl-
ation status of CpG islands. Proc Natl Acad Sci U S A.
21. Laiho P, Launonen V, Lahermo P, Esteller M, Guo M, Her-
man JG, et al. Low-level microsatellite instability in most
colorectal carcinomas. Cancer Res. 2002;62:1166–1170.
22. Nowak MA, Komarova NL, Sengupta A, Jallepalli PV, Shih Ie
M, Vogelstein B, et al. The role of chromosomal instability in
tumor initiation. Proc Natl Acad Sci U S A 2002;99:16226–
23. Boland CR, Sato J, Appelman HD, Bresalier RS, Feinberg AP.
Microallelotyping defines the sequence and tempo of allelic
losses at tumour suppressor gene loci during colorectal can-
cer progression. Nat Med. 1995;1:902–909.
24. Messerini L, Vitelli F, De Vitis LR, Mori S, Calzolari A, Palmi-
rotta R, et al. Microsatellite instability in sporadic mucinous
colorectal carcinomas: relationship to clinico-pathological
variables. J Pathol. 1997;182:380–384.
25. Kern SE, Fearon ER, Tersmette KW, Enterline JP, Leppert M,
Nakamura Y, et al. Clinical and pathological associations
with allelic loss in colorectal carcinoma [corrected]. JAMA.
Matched Study for Mucinous Carcinomas/Kazama et al.2029