Original Scientific Reports

Clinical and Molecular Biologic Characteristics of Early-onset Versus Late-onset

Colorectal Carcinoma in Filipinos
Gemma B. Uy, M.D.,
1
Leoncio L. Kaw, M.D.,
1
Corazon K. Punzalan, M.D.,
2
R. Ireneo Luis C. Querol, M.D.,
1
Elena V. Koustova, Ph.D.,
1
Mark W. Bowyer, M.D.,
1
Christine M. Hobbs, M.D.,
3
Leslie H. Sobin, M.D.,
3
David C. Wherry, M.D.
1
1
Department of Surgery, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, 29814-4799 Bethesda, Maryland, USA
2
Division of Colorectal Surgery, University of the PhilippinesPhilippine General Hospital, Taft Avenue, 1100 Manila, Philippines
3
Gastrointestinal Section, Armed Forces Institute of Pathology, 20307 Bethesda, Maryland, USA
Published Online: January 8, 2004
Abstract. A case-control study of Filipino patients who underwent surgical
resection for colorectal cancer (CRC) during a 1-year period was under-
taken. Thirty-five patients under age 40 years were identified. Paraffin
blocks of these and 35 randomly selected patients over age 40 underwent
histologic and immunohistochemical evaluation. Markers chosen for
evaluation included the apoptosis-associated gene products (p53 and bcl-
2), a tumor proliferation activity-related factor (Ki-67), and the markers
(MLH1 and MSH2) of DNA microsatellite instability (MSI). Results were
correlated with age and the stage and location of the tumor. The average
age of the early-onset group was 30.7 years compared to the late-onset
group at 67.0 years; and the male/female ratio was equivalent. The younger
patients had a significantly higher Dukes stage, the tumors were more
poorly differentiated, and they were more frequently of the mucinous and
signet ring cell histopathologic type. Expression of p53 was higher in the
younger patients (p < 0.001) and was independent of the degree of differ-
entiation or the stage of the tumor. No differences of expression were noted
for the other markers measured. The increased frequency of CRC in Fili-
pino patients less than 40 years of age offers a unique opportunity to gain a
better understanding of carcinogenesis, which might be exploited during
diagnosis and management. The differences noted between the early- and
late-onset CRC are provocative and provide an impetus for increased
screening in Filipinos.
Colorectal cancer (CRC) remains a major concern throughout the
world. Despite advances in medicine, there has been no significant
decline in the incidence and mortality rates over the last few de-
cades. In the United States, the colon and rectumare the third most
common sites of newcancers, and CRCis the third leading cause of
cancer deaths for both men and women, with an estimated 148,300
new cases and 56,600 deaths from this disease between 1979 and
1998 [1]. Although less commonly encountered in Asia than in
Western countries, CRC remains a challenge in the Philippines.
Statistics from 1998 showed that it ranked fourth in cancer inci-
dence and fifth as a leading cause of cancer death [2].
Colorectal cancer is primarily a disease of older age groups.
Younger patients are considered an interesting subset, as their can-
cers generally behave differently from those in older patients and
are commonly linked to a familial predisposition. In Western popu-
lations, CRC in the young (< 40 years old) comprise around 2% to
8%of all the diagnosed cases [3]. In young Filipinos, several reports
have shown a significantly higher frequency, in the range of 17%to
36% [46]. A recent 7-year review by Kaw et al. [5] of 1277 CRC
patients showed that 17% were under the age of 40. In addition, it
was found that these patients present at more advanced stages of
the disease.
The onset of CRC involves an interplay of genetic and environ-
mental factors, which explains the wide variability of incidence
among races. Known predisposing factors include adenomatous
polyps, inflammatory bowel disease (ulcerative colitis and Crohns
disease), and inherited disorders such as familial adenomatous
polyposis (FAP) and hereditary nonpolyposis colorectal cancer
(HNPCC). HNPCC has been well studied and involves a distinct
subset of patients. They tend to be younger, have more right-sided
tumors, and display a higher incidence of synchronous and meta-
chronous CRC[7]. HNPCCis thought to involve an alternate path-
way of tumorigenesis that is associated with microsatellite instabil-
ity and is found in 4% to 6% of the population [8]. Interestingly,
documented HNPCC cases are rare in the Philippines. So far, only
one case has been reported [9]. Perhaps this low number is due to
the lack of family histories in Filipino medical records or because
cases do not meet the Amsterdam HNPCC criteria. The rare oc-
currence of polyps and inflammatory bowel disease (IBD) in Fili-
pinos and other unique characteristics of Filipino CRC patients
make this population ideal for exploring possible variations in tu-
mor biology.
This study employed immunohistochemistry (IHC) to examine
the differences between early-onset and late-onset CRC in Filipi-
nos. Markers involved at various stages of tumor development were
studied. Markers of apoptosis included p53 and bcl-2. Mutation of
p53 is considered a relatively late event along the pathway and is a
determinant of progression from adenoma to carcinoma, whereas
bcl-2 is involved early on. The bcl-2 gene is a known inhibitor of Correspondence to: David C. Wherry, M.D.
WORLD
Journal of
SURGERY
2004 by the Societe
Internationale de Chirurgie
World J. Surg. 28, 117123, 2004
DOI: 10.1007/s00268-003-7281-4
apoptosis and therefore allows accumulation of genetic alterations
needed for neoplastic development [10].
The percentage of proliferating tumor cells was assessed using
Ki-67 antibody. This antibody is directed against a nuclear antigen
that is expressed only by proliferating cells [11]. Finally, alterations
in mismatch repair (MMR) proteins were examined using antibod-
ies to MLH1 and MHS2. As shown in several studies, IHCprovides
a rapid, accurate means of identifying tumors with MSI [1214].
Materials and Methods
Materials
This study was approved by the Institutional Review Board of the
Uniformed Services University of the Health Sciences (USUHS) in
Bethesda, Maryland, USA. Surgical pathology reports from Janu-
ary 1, 1999 to December 31, 1999 of the Department of Pathology
at the Philippine General Hospital (PGH) were reviewed. A total
of 35 paraffin blocks of resected adenocarcinoma of the colon or
rectum from patients aged 40 years were identified. These pa-
tients comprised the early-onset CRC group. Another 35 patients
> 40 years of age were randomly selected and comprised the late-
onset CRC group. Blocks from the carcinomas of each of these
patients were also obtained. The data collected included patient
age and sex, location of the primary tumor, depth of bowel wall
invasion, and involvement of lymph nodes. Family history, pres-
ence of distant metastasis, and outcome were not available fromthe
records. The tumors were then staged using Dukes classification
[15]. The location of the primary tumor was divided into three sites:
right colon (from the cecum to the transverse colon), left colon
(from the splenic flexure to the sigmoid colon), and rectum (recto-
sigmoid junction, rectum, and anorectum excluding the anus).
The blocks of formalin-fixed tissue fromthe subjects were coded
and shipped to the USUHS for immunohistologic analysis by inves-
tigators blinded to the patients data. These paraffin blocks were
reembedded, sectioned, and mounted on slides (American His-
tolabs, Gaithersburg, MD, USA) for immunohistochemistry and
hematoxylin and eosin (H&E) staining. The H&E-stained slides
were sent to the Division of Gastrointestinal Pathology at the
Armed Forces Institute of Pathology (AFIP), Washington, DC for
blinded histologic evaluation of the degree of differentiation, his-
topathologic type, and depth of bowel wall invasion.
Immunohistochemical Analysis
The IHC analysis of p53, Ki-67, bcl-2, MLH1, and MSH2 was per-
formed. Optimal concentrations of the various antibodies were de-
termined in pilot experiments. The slides were deparaffinized in
xylene and rehydrated in alcohol of graded concentrations and dis-
tilled water. Antigen demasking was performed by heating the
slides with Target Retrieval Solution (Dako, Carpinteria, CA,
USA). Incubation with Peroxidase Suppressor (Pierce, Rockford,
IL, USA) was done to quench any endogenous peroxidase activity.
Primary antibodies were applied and incubated overnight at 4C
at the following concentrations: p53, 1:500 (rabbit anti-p53 onco-
protein polyclonal antibody; Chemicon, Temecula, CA, USA); Ki-
67, 1:500 (mouse anti-Ki-67 monoclonal antibody, Chemicon); bcl-
2, 1:500 (mouse anti-bcl-2a oncoprotein monoclonal antibody,
Chemicon); MLH1, 1:500 [rabbit polyclonal immunoglobulin G
(IgG), Oncogene Research Products, Boston, MA, USA); MSH2,
1:500 (rabbit polyclonal IgG, Oncogene Research Products).
Horseradish peroxidase (HRP)-labeled secondary antibodies
(Pierce) were then applied at a 1:2500 dilution for Ki-67, bcl-2, and
MSH2 and at 1:5000 dilution for p53 and MSH1; they were incu-
bated for 30 minutes at room temperature. TSA Biotin System
(PerkinElmer Life Sciences, Boston, MA, USA) was used for signal
amplification. Cells that reacted with the antibodies were visual-
ized using the HRP chromogenic substrate Vector VIP (Vector
Laboratories, Burlingame, CA, USA). After each incubation, the
slides were extensively washed with phosphate-buffered saline
(PBS). Negative controls consisted of slides for which the primary
antibody was omitted.
Immunoreactivity was evaluated by counting the cells positive
for the various antibodies using Image Pro Plus 4.1 (Media Cyber-
netics, Silver Spring, MD, USA) at 100magnification for p53 and
Ki-67 and at 400for bcl-2. For MLH1 and MSH2 staining, normal
colorectal tissue adjacent to the tumor served as positive controls.
Loss of expression of these proteins was recorded when positive
nuclear immunostaining was observed in normal tissue or lympho-
cytes but not in malignant cells [14, 16].
Statistical Analysis
The clinicopathologic characteristics (sex, tumor location, degree
of differentiation, Dukes stage) were compared between the early-
and late-onset groups using the z test (Microsoft Excel 2000). Ex-
pression of p53, bcl-2, and Ki-67 was stratified into five groups: 1%
to 20%, 21%to 40%, 41%to 60%, 61%to 80%, and 81%to 100%.
The distribution of cases between the two onset groups was ana-
lyzed using Chi-Square Trend with Graphpad Prism Version 2.0
(Graphpad Software, San Diego, CA, USA). The
2
test or Fishers
exact test, whichever was appropriate, was used to correlate antigen
expression with the degree of differentiation and tumor stage
(SPSS Program for Windows Version 11.0, Statistical Package for
Social Sciences; SPSS, Chicago, IL, USA). Significance was defined
as p < 0.05.
Results
One patient from the late-onset group was excluded because the
paraffin block showed no tumor cells upon reviewat the AFIP. The
mean age for the early-onset group was 30.7 years (range 1940
years), whereas that for the late-onset group was 67 years (range
5182 years). Of the 35 patients in the younger group 23 were men
(65%), whereas only 19 of the 34 (55.9%) in the older group were of
the male sex. In both groups, most of the cancers were in the rectum
(57% and 56%); and although a higher percentage of CRCs was
right-sided in the younger group (29% vs. 15%), the difference did
not reach statistical significance.
A statistically significant difference between the two groups was
noted, with regard to the degree of differentiation. Younger pa-
tients had a higher percentage of poorly differentiated CRCs
(34.2%vs. 11.8%), whereas the older patients had a higher percent-
age of moderately differentiated CRCs (64.7%vs. 28.6%). The mu-
cinous type of adenocarcinoma was more common in the early-
onset group, as was the presence of signet ring cells (p < 0.05).
Coexistent granulomatous infections, such as schistosomiasis and
tuberculous lymphadenitis, were noted in 2.9% and 7.2% of cases,
respectively (Table 1).
The depth of tumor invasion and degree of differentiation re-
118 World J. Surg. Vol. 28, No. 2, February 2004
ported in this study were based on the findings of the AFIP on their
review of the tumor slides. The presence or absence of distant me-
tastasis at the time of surgery was not part of the surgical pathology
reports, so staging was limited to using Dukes systeminstead of the
TNM classification. As shown in Table 1, the younger patients had
pathologically more aggressive disease and a significantly higher
percentage of stage C tumors than did the older group (p = 0.037).
p53
Expression of p53 was graded based on the percentage of tumor
cells that were immunoreactive with p53 antibody (Fig. 1) and cat-
egorized into various groups using 20% range intervals. The
younger group showed a higher percentage of p53-positive cells in
their tumors than did the older group (p =0.0004). Interestingly, in
most of the younger patients 61% to 80% of tumor cells expressed
p53, whereas most of the older patients had a relatively equal dis-
tribution of p53 expression over the 1% to 20%, 21% to 40%, and
41% to 60% ranges (Fig. 2). The degree of differentiation did not
correlate with Dukes stage.
bcl-2
The percentage of tumor cells that were bcl-2-positive was noted to
be similar for the two groups (Fig. 3). Altogether, 63% of the pa-
tients under age 40 and 65%of those over age 40 were in the 1%to
20%range, and 23%of the former and 32%of the latter were in the
21% to 40% range. No patient had a tumor with more than 60%
bcl-2-positive cells (Fig. 4). As with p53, there was no correlation
with the other clinicopathologic variables.
Ki-67
The Ki-67 score was the percent of tumor cells counted that were
Ki-67-positive (Fig. 5). There was substantial heterogeneity among
the scores in both groups, with a mean Ki score of 36.06 23.28 for
the early-onset group and 43.95 23.55 for the late-onset group
(Fig. 6). The mean Ki score was not statistically different between
the two groups. The Ki score was found to be independent of dif-
ferentiation, location, and stage of the tumor.
MSH1 and MSH2
We used IHCto detect MMRdefects, and we considered failure of
tumor cells to react with either MLH1 or MSH2 antibody as indica-
tive of defective MMR (Table 2). In the early-onset group 8 of 35
tumors (22.9%) failed to react with one of the two antibodies, as did
7 of 34 tumors (20.6%) in the late-onset group. One tumor from a
patient in the early-onset group did not react with either antibody.
No statistical correlation was drawn with regard to age, degree of
tumor differentiation, stage, or mucinous type between cases with
defective or intact MMR.
Discussion
Colorectal cancer is a major health problemand remains one of the
leading causes of cancer deaths worldwide. It is predominantly a
Table 1. Demographic and clinicopathologic data of early- and
late-onset colorectal cancer.
Parameter Early onset Late onset p
No. of patients 35 34
Age (years), mean 30.7 6.3 67 9.0
Male 23 (65.7%) 19 (55.9%) NS
Female 12 (34.3%) 15 (44.1%) NS
Location of tumor
Right side 10 (29%) 5 (15%) NS
Left side 5 (14%) 10 (29%) NS
Rectum 20 (57%) 19 (56%) NS
Differentiation
Well 13 (37.1%) 8 (23.5%) NS
Moderate 10 (28.6%) 22 (64.7%) 0.003
Poor 12 (34.2%) 4 (11.8%) 0.027
Dukes stage
A 1 (2.9%) 3 (8.8%) NS
B 8 (22.9%) 14 (41.2%) NS
C 26 (74.3%) 17 (50%) 0.037
Other tumor characteristics
Mucinous adenocarcinoma 10 (27.8%) 1 (2.9%) 0.004
Signet ring type 1 (2.9%) 0
With signet rings 8 (22.9%) 0 0.003
With Schistosoma eggs 2 (5.7%) 0 NS
With tuberculous lymphadenitis 4 (11.4%) 1 (2.9%) NS
Fig. 1. p53-positive cells (dark staining) in a patient with adenocarcinoma
of the sigmoid colon.
Fig. 2. Comparison of p53 staining for early- and late-onset colorectal car-
cinoma groups. Squares: 40 years old; circles: > 40 years old.
119 Uy et al.: Colorectal Carcinoma in Filipinos
disease of older age groups, with 90%of cases occurring in persons
over 50 years of age and a peak incidence during the seventh decade
of life. However, it can also occur at younger ages, especially in
patients with a family history of the disease. Traditionally, CRC in
the young has been considered to have a worse prognosis, partly
because of its more aggressive character and partly due to delay in
diagnosis. However, recent studies have shown a similar survival
rate between the young and old groups when compared stage by
stage [3]. Mitry et al. [17] showed better survival for patients less
than age 45 using stage-stratified survival rates. This was explained
by a significantly lower postoperative mortality rate and more fre-
quent use of adjuvant therapy or palliative chemotherapy.
Reviews by Talens et al. [6] and Kaw et al. [5] of CRC patients
treated at the Philippine General Hospital showed a considerably
higher frequency of CRC in the young (8% and 17%, respectively)
than has been found in Western populations. In the Western litera-
ture, patients younger than 40 years of age comprise only 2%to 8%
of the total population [3]. The higher frequency of early-onset
CRC in the Philippines provides an opportunity to compare and
contrast age-specific genetic and molecular biologic characteris-
tics. To this end, established markers of various tumorigenic prop-
erties (i.e., p53, bcl-2, MLH1, MSH2, Ki-67) were studied in speci-
mens obtained from patients under and over the age of 40. To our
knowledge, this represents the first such attempt to quantify differ-
ences between these two age groups.
In our study, the younger patients had a higher frequency of
Fig. 3. bcl-2-positive cells (dark staining) in a patient with adenocarcinoma
of the colon.
Fig. 4. Comparison of bcl-2 staining between early- and late-onset colorec-
tal carcinoma groups. Squares: 40 years old; circles: > 40 years old.
Fig. 5. Ki-67-positive cells (dark staining) in a patient with adenocarci-
noma of the rectum.
Fig. 6. Comparison of Ki-67 staining between early- and late-onset colo-
rectal carcinoma groups. Squares: 40 years old; circles: > 40 years old.
Table 2. MLH1 and MSH2 staining characteristics of early- versus
late-onset colorectal cancer.
MLH1/MSH2 staining Early onset Late onset
+/+ 27 27
+/- 6 5
-/+ 1 2
-/- 1 0
Total 35 34
Defective MMR
a
22.9% 20.6%
Intact MMR
b
77.1% 79.4%
MMR: mismatch repair.
a
Failed to stain positively for either MLH1 or MSH2 antibody.
b
Stained positively for both MLH1 and MSH2 antibodies.
120 World J. Surg. Vol. 28, No. 2, February 2004
poorly differentiated tumors, cancers with a mucinous component,
cancers with signet ring cells, and cancers diagnosed at advanced
stages. Poorly differentiated adenocarcinomas comprised 34.2%of
the CRCs in the younger age group compared to only 11.8% in the
older group. It has been shown that poorly differentiated adeno-
carcinomas are associated with a poorer survival rate than are well
differentiated tumors [18]. In our patient population, 74.3% of tu-
mors in the younger age group were Dukes stage Cversus 50.0%in
the older age group. This is important, as the prognosis is directly
related to the tumor stage at operation. According to U.S. data, the
5-year survival rate for a localized tumor is an excellent 89.7%. This
figure drops dramatically to 64.4% with spread to regional lymph
nodes and to a dismal 8.3% with distant metastasis [1]. Because of
the difficulties related to patient follow-up, there are no available
local data that pertain to the survival rates of Filipino cancer pa-
tients at different stages of the disease.
The risk profile of Filipino CRCpatients is different fromthat of
Western patients. Polyps are found in only 4.7% cases of resected
CRC specimens from Filipino patients [5]. In addition, ulcerative
colitis and Crohns disease are uncommon in Filipinos, as are re-
ported cases of HNPCC and FAP [9]. Our study did reveal a high
frequency of coexisting granulomatous infections, such as tubercu-
lous lymphadenitis and schistosomiasis, especially in the younger
age group. As suggested by Chen et al. [19], these infections could
be involved in the multistep process of colorectal tumorigenesis by
causing genetic alterations similar to those seen in ulcerative colitis.
In colon cancers, as in others, there is progressive DNA damage
involving a randomand multistep process that includes inactivation
or deletion of tumor suppressor genes and activation of oncogenes
[20]. CRCmay develop along one of several genetic pathways, with
areas of overlap [21, 22]. This entire process may take several years,
and during this time frame preventive interventions at various
stages may significantly alter the progression to invasive cancer.
New insights regarding the genetic basis of this disease will poten-
tially have application in the clinical management of CRCpatients.
It will be extremely valuable if molecular biologic factors that favor
development along certain pathways can be identified for direct
screening. Several prognostic markers may help us understand the
subcellular mechanisms in colorectal tumorigenesis and identify
those persons at increased risk of developing this type of cancer.
The protein p53 was tested in this study, as it is the most commonly
observed marker of a poor prognosis for patients with CRCs, oc-
curring in approximately 75% [23]. This cell cycle regulatory pro-
tein induces apoptosis in cases of DNA damage. It also has tumor
suppressor activity. Mutations of p53 result in prolonged half-life
and accumulation of the protein to levels detectable immunohisto-
chemically in cancer cell nuclei. Overexpression of p53 reflects loss
of normal growth regulation. Comparison of the percentages of
p53-positive tumor cells between our two groups showed higher
expression (p < 0.01) in the younger group than in the older one.
This may provide an explanation for the more aggressive nature of
CRCin young Filipinos. Indeed, as Schwandner et al. [24] found in
their study, p53 was an independent predictor of recurrence and
survival. In addition, Gallego et al. [25] noted a significantly shorter
disease-free survival for their p53-positive group. In our study, p53
expression did not correlate with stage. This may be attributed to
the central role of p53, as mutation occurs just before the tumor
gains its invasive properties, so its presence is not stage-dependent.
One important clinical implication of tumors with deficient p53 is
that they show enhanced sensitivity to doxorubicin and resistance
to 5-fluorouracil (5-FU) [26]. This may explain treatment failures
when 5-FU is used as the first line of adjuvant chemotherapy.
Another marker used in this study was bcl-2, an antiapoptotic
protein that is thought to be involved early in the genetic pathway
by acting synergistically with other genes to promote neoplastic
changes. In a study by Sinirope et al. [10], bcl-2 positivity was local-
ized to the basal epithelium in normal mucosa and hyperplastic
crypts but was found superficially in dysplastic and malignant cells.
An inverse correlation was found between bcl-2 and p53 positivity,
but only in adenomas and not in carcinomas. This suggests down-
regulation of bcl-2 by mutant p53 in premalignant polyps. In this as
in other studies, no correlation was shown between p53 expression
and other clinicopathologic variables [10, 27].
The proliferative activity of the tumor cells was evaluated using
antibody to Ki-67 protein. Our results concur with other studies
that have shown the percentage of tumor cells expressing Ki-67 to
be independent of age, stage, and differentiation in CRC [11, 28,
29].
The discovery of MSI and its association with HNPCC has given
new insights into the tumor biology of CRC. It offers an alternate
pathway to the established adenomacarcinoma sequence [30].
MSI is due to failure of the DNA mismatch repair system to repair
errors during DNA replication, thereby leading to the accumula-
tion of mutations [31, 32]. Knowledge of MSI has also broadened
our approach to the clinical management of patients with heredi-
tary susceptibility to development of CRC. However, it has been
shown that MSI is not exclusive to either CRC or HNPCC [33].
Although MSI is found in more than 90% of HNPCC patients, it is
also present in 10% to 15% of sporadic CRCs [34]. The rarity of
HNPCCin the population (46%) explains why it accounts for only
16% of all cases of MSI. The other 84% of cases are sporadic. The
critical implication of MSI status is that it is common in a unique
subset of CRC patients whose tumors tend to be more proximal in
location, poorly differentiated, mucinous, and associated with peri-
tumoral lymphocytic infiltration and DNA diploidy. In addition,
these patients have been shown to have a better prognosis than
patients with non-MSI-related adenocarcinomas and are sensitive
to 5-FU-based chemotherapy [3538]. Still, these patients are in
need of long-term surveillance, as they are at increased risk for
metachronous CRC as well as extracolonic tumors.
Several studies have shown the usefulness of IHC using MLH1
and MSH2 antibodies to detect MMR defects. IHC is a rapid, sen-
sitive, specific method for screening large numbers of patients. This
contrasts with genetic analysis of MSI status, which is expensive,
time-consuming, and requires specialized equipment [12]. IHChas
shown a predictive value of 100%for detecting high-frequency MSI
and 96.7% for detecting microsatellite stability and low-frequency
MSI, as reported by Lindor et al. [13]. Stone et al. [14] showed 96%
sensitivity using IHC for detecting MSI compared to molecular
analysis with the polymerase chain reaction method using the five
microsatellite loci defined by the National Cancer Institute guide-
lines. Thus, IHC can be utilized to screen for patients needing fur-
ther testing to confirm germ line mutations [39, 40]. Our study
found a somewhat higher proportion of MSI cases in the younger
group (22.9%) than in the older group (20.6%). In addition, loss of
expression of MSH2 was more commonly seen than loss of expres-
sion of MLH1 (75%vs. 25%). Lanza et al. [37] divided patients into
two groups: those with or suspected of having HNPCC and those
without a family history suggestive of HNPCC. This study showed a
nearly equal frequency of loss of MLH1 and MSH2 in the former
121 Uy et al.: Colorectal Carcinoma in Filipinos
group, whereas patients with sporadic CRCs more frequently lost
MLH1 expression (87.5%).
Hypermethylation of the MLH1 promoter region has been sug-
gested as the primary mechanism in the development of sporadic
CRC[26]. Our findings stimulate interest in further study of MMR
defects in Filipino patients. For instance, one could investigate
whether Filipinos have a unique type of mutation or some kind of
familial clustering of CRCcases that does not meet established cri-
teria of known syndromes. An association between MSI-positive
tumors and certain unique clinicopathologic features was not
found in this study. Loss of expression of one of the markers of
defective MMR (MLH1 or MSH2) was the same for both the
younger and older patients. Location of the tumor and the degree
of differentiation did not differ significantly between intact and de-
fective MMRcases. Although MLH1 and MSH2 are the most com-
mon genes involved in this pathway, other genetic mutations may
occur. Testing for all such mutations in future studies may aid in the
further elucidation of CRC tumorigenesis.
There is still much to learn about the molecular behavior of co-
lorectal cancer. Other markers must be tested to lay out a clearer
viewof the genetic alterations involved in the development of these
cancers. The concept of de novo carcinoma is emerging and pro-
vides an alternate pathway to the established adenomacarcinoma
sequence [41]. Future studies comparing Filipinos with age- and
stage-matched non-Filipinos may reveal divergent pathways of co-
lorectal tumorigenesis. A survey of survival differences among
Asian emigrants to the United States has already shown that Fili-
pino immigrants were more likely to be diagnosed with advanced
CRC and had worse survival rates than Chinese, Japanese, and
non-Hispanic white immigrants [42]. Because access to structured
medical care has been eliminated as a factor for this delay in diag-
nosis, a genetic predisposition may explain the particular aggres-
siveness of this disease in Filipinos.
Our study was limited by its retrospective nature, lack of clinico-
pathologic data, and poor patient follow-up. These issues are being
addressed in an ongoing prospective study of the molecular biology
of CRC in Filipinos. With most patients presenting at an advanced
stage, the need for earlier diagnosis cannot be overstated. It is
hoped that the knowledge that we gain fromthis and future studies
of the molecular biology of CRC allows an earlier diagnosis and
more targeted management of our patients.
Re sume . Cette e tude cas-te moin concerne des patients philippins ayant eu
une re section chirurgicale pour cancer colorectal pendant une pe riode
dobservation dun an. Trente cinq patients de moins de 40 ans ont e te
identifie s. Lexamen histologique et immunohistochimique apre `s fixation a`
la paraffine a inte resse ces pie `ces de re sections et compare a` ceux de 35
autres patients se lectionne s au hasard. Les marqueurs choisis pour
e valuation comprenaient les produits ge ne tiques associe s a` lapoptose (p53
et bcl-2), un facteur dactivite de prolife ration tumorale (Ki-67), et les
marqueurs (MLH1 et MSH2) de linstabilite microsatellite dADN (MSI).
Les re sultats ont e te corre le s avec lage, le stade et la localisation de la
tumeur. Lage moyen du groupe porteur de cancer dapparition pre coce a
e te de 30.7 ans compare a` celui du groupe dapparition tardive qui a e te de
67 ans; le sexe ratio a e te de 1. Les patients les plus jeunes avaient un stade
de Dukes significativement plus avance , des cancers plus indiffe rencie s, et
e taient plus souvent mucineux ou contenant des cellules a` bague de chaton.
Lexpression P53 e tait plus forte chez les patients plus jeunes (p < 0.001),
inde pendamment du degre de de diffe renciation ou du stade de la tumeur.
On na retrouve aucune diffe rence dans lexpression des autres marqueurs
tumoraux mesure s. Lincidence plus e leve e de cancer colorectal chez les
patients philippins de moins de 40 ans est une situation unique pour
pourvoir mieux comprendre la carcinogene `se et qui pourrait e tre exploite e
pour ame liorer le diagnostic et la prise en charge. Les diffe rences note es
entre les cancers dapparition pre coce et tardive sont provocatrices et nous
incitent a` proposer un programme acce le re de de pistage.
Resumen. Se realizo un estudio de control de casos de pacientes filipinos
sometidos a reseccion por cancer colo-rectal (CCR) en un periodo de un
ano. Treinta y cinco pacientes menores de 40 anos fueron identificados. Se
hizo el estudio histologico e inmunohistoqu mico de e stos y de otros 40
pacientes mayores de 40 anos seleccionados al azar. Los marcadores
escogidos para la evaluacion incluyeron los productos del gen asociado con
la apoptosis (p53 y bel-2); un factor relacionado con la actividad tumoral
(Ki-67); y los marcadores (MlH1) y MSH2) de inestabilidad microsatelital
de ADN (MSI). Los resultados fueron correlacionados con la edad, el
estado y la ubicacion del tumor. La edad promedio de los pacientes con
presentacion temprana de la enfermedad fue 30.7 anos, en tanto que la
edad promedio del grupo con presentacion tard a fue 67 anos; la relacion
masculino a femenino fue equivalente. Los pacientes mas jovenes mostraron
un estado de Duke mas alto, sus tumores eran mas pobremente
diferenciados y mas comunmente del tipo histopatolo gico mucinoso y en
anillo de sello. La expresio n de P53 aparecio mas alta en el grupo de
pacientes mas jovenes (p <0.001) e independiente del grado de diferenciacion
del tumor. No se encontraron diferencias en la expresio n de los otros
marcadores. La mayor frecuencia de CCR en los pacientes filipinos
menores de 40 anos ofrece una oportunidad unica para lograr una mejor
comprension del proceso de carcinogenesis que podr a ser explotada tanto
en el diagnostico como en el manejo de la neoplasia. Las diferencias
observadas entre los grupos de menor y mayor edad en cuanto a la
presentacion del CCRson provocativas y dan fundamento a incrementar el
tamizaje en los filipinos.
References
1. Colorectal Cancer and Early Detection. Cancer Facts and Figures 2002,
Surveillance Research, Atlanta, American Cancer Society, 2002:2027
2. Laudico AV, Esteban DB, Reyes LM, et al. Philippine Cancer Facts
and Estimates, Manila. Philippine Cancer Society 1998;1998:944
3. Parramore JB, Wei JP, Yeh KA. Colorectal cancer in patients under
forty: presentation and outcome. Am. Surg. 1998;64:563567
4. Inoturan E, Yap N, de Guzman R, et al. Adenocarcinoma of the colon
and rectum: a reviewof 53 cases. Philippine J. Surg. Spec. 2000;55:710
5. Kaw L, Punzalan C, Crisostomo A, et al. Surgical pathology of colorec-
tal cancer in Filipinos: implications for surgical practice. J. Am. Coll.
Surg. 2002;195:188195
6. Talens E, Recio S, Magsanoc C, et al. Cancer of the colon, rectum and
anal canal: an eight year review of cases. Philippine J. Surg. Spec. 1991;
46:5359
7. Baba S. Recent advances in molecular genetics of colorectal cancer.
World J. Surg. 1997;21:678687
8. Bresalier R. Malignant and premalignant lesions of the colon. In Gren-
dell J, McQuaid K, Friedman S, editors, Current Diagnosis and Treat-
ment in Gastroenterology, 1st edition, Stamford, CT, Appleton &
Lange, 1996;367391
9. Paz CR. Lynch Syndrome II. Acta Med Philippina 1991;5862.
10. Sinirope F, Ruan SB, Cleary K, et al. Bcl-2 and p53 oncoprotein expres-
sion during colorectal tumorigenesis. Cancer Res. 1995;55:237241
11. Porschen R, Kriegel A, Langen C, et al. Assessment of proliferative
activity in carcinomas of the human alimentary tract by Ki-67 immuno-
staining. Int. J. Cancer 1991;47:686691
12. Ward R, Meagher A, Tomlinson I, et al. Microsatellite instability and
the clinicopathological features of sporadic colorectal cancer. Gut
2001;48:821829
13. Lindor N, Burgart L, Leontovich O, et al. Immunohistochemistry ver-
sus microsatellite instability testing in phenotyping colorectal tumors. J.
Clin. Oncol. 2002;20:10431048
14. Stone JG, Robertson D, Houlston RS. Immunohistochemistry for
MSH2 and MLH1: a method for identifying mismatch repair deficient
colorectal cancer. J. Clin. Pathol. 2001;54:484487
15. Goldberg R. Gastrointestinal tract cancers. In Casiato D, Lowitz B,
editors, Manual of Clinical Oncology, 4th edition, Philadelphia, Lippin-
cott Williams & Wilkins, 2000;182196
16. Paraf F, Gilquin M, Longy M, et al. MLH1 and MSH2 protein immu-
nohistochemistry is useful for detection of hereditary non-polyposis co-
lorectal cancer in young patients. Histopathology 2001;39:250258
122 World J. Surg. Vol. 28, No. 2, February 2004
17. Mitry E, Benhamiche AM, Jouve JL, et al. Colorectal adenocarcinoma
in patients under 45 years of age: comparison with older patients in a
well-defined French population. Dis. Colon Rectum 2001;22:380387
18. Bresalier R, Kim Y. Malignant neoplasms of the large intestine. In
Feldman M, Scharschmidt B, Sleisenger M, editors, Sleisenger and
Fordtrans Gastrointestinal and Liver Disease: Pathophysiology/
Diagnosis/Management, 6th edition, Philadelphia, Saunders, 1998;
19231924
19. Chen MC, Chang PY, Chuang CY, et al. Colorectal cancer and schis-
tosomiasis. Lancet 1981;1:971973
20. Alberts D. Reducing the risk of colorectal cancer by intervening in the
process of carcinogenesis: a status report. Cancer J. 2002;8:208221
21. Ilyas M, Tomlinson IPM. Genetic pathways in colorectal cancer. His-
topathology 1996;28:389399
22. Fearon E, Vogelstein B. A genetic model for colorectal tumorigenesis.
Cell 1990;61:759767
23. Fahy B, Bold R. Epidemiology and molecular genetics of colorectal
cancer. Surg. Oncol. 1998;7:115123
24. Schwandner O, Schiedeck THK, Bruch HP, et al. Apoptosis in rectal
cancer: prognostic significance in comparison with clinical, histopatho-
logic, and immunohistochemical variables. Dis. Colon Rectum2000;43:
12271236
25. Gallego MG, Acenero MJ, Ortega S, et al. Prognostic influence of p53
nuclear overexpression in colorectal carcinoma. Dis. Colon Rectum
2000;43:971975
26. Chung D. The genetic basis of colorectal cancer: insights into critical
pathways of tumorigenesis. Gastroenterology 2000;119:854865
27. Pereira H, Silva S, Juliao R, et al. Prognostic markers for colorectal
cancer: expression of p53 and bcl2. World J. Surg. 1997;21:210213
28. Kyzer S, Gordon P. Determination of proliferative activity in colorectal
carcinoma using monoclonal antibody Ki67. Dis. Colon Rectum 1997;
40:322325
29. Shepherd N, Richman P, England J. Ki-67 derived proliferative activity
in colorectal adenocarcinoma with prognostic correlations. J. Pathol.
1988;155:213219
30. Goldstein NS. Recent pathology related advances in colorectal adeno-
carcinomas. Eur. J. Surg. Oncol. 2001;27:446450
31. Sedwick WD, Markowitz S, Veigl M. Mismatch repair and drug re-
sponses in cancer. Drug Resist. Update 1999;2:295306
32. Grady W, Markowitz S. Genetic and epigenetic alterations in colon
cancer. Annu. Rev. Genomics Hum. Genet. 2002;3:101128
33. Boland CR, Thibodeau S, Hamilton S, et al. A National Cancer Insti-
tute workshop on microsatellite instability for cancer detection and fa-
milial predisposition: development of international criteria for the de-
termination of microsatellite instability in colorectal cancer. Cancer
Res. 1998;58:52485257
34. Peltomaki P, Vasen H, Jass JR. Hereditary nonpolyposis colorectal
cancer. In Hamilton SR, Aaltonen LA, editors, WHO Classification of
Tumours. Pathology and Genetics of Tumours of the Digestive System
Lyon, IARC Press, 2000;126129
35. Hemminki A, Mecklin JP, Jarvinen H, et al. Microsatellite instability is
a favorable prognostic indicator in patients with colorectal cancer re-
ceiving chemotherapy. Gastroenterology 2000;119:921928
36. Gryfe R, Kim H, Hsieh E, et al. Tumor microsatellite instability and
clinical outcome in young patients with colorectal cancer. N. Engl. J.
Med. 2002;342:6977
37. Lanza G, Gafa R, Maestri I, et al. Immunohistochemical pattern of
MLH/MSH2 expression is related to clinical and pathological features
in colorectal adenocarcinomas with microsatellite instability. Mod.
Pathol. 2002;15:741749
38. Lynch HT, Smirk TC, Watson PP, et al. Genetics, natural history, tumor
spectrum, and pathology of hereditary nonpolyposis colorectal cancer:
an updated review. Gastroenterology 1993;104:15351549
39. Dieumegard B, Grandjouan S, Sabourin JC, et al. Extensive molecular
screening for hereditary non-polyposis colorectal cancer. Br. J. Cancer
2000;82:871880
40. Ponz de Leon M, Pedroni M, Benati P, et al. Hereditary colorectal can-
cer in the general population: from cancer registration to molecular
diagnosis. Gut 1999;45:3238
41. Muto T, Nagawa H, Watanabe T, et al. Colorectal carcinogenesis: his-
torical review. Dis. Colon Rectum 1997;40:8085
42. Lin S, Clarke C, Prehn A, et al. Survival differences among Asian sub-
populations in the United States after prostate, colorectal, breast and
cervical carcinomas. Cancer 2002;94:11751182
123 Uy et al.: Colorectal Carcinoma in Filipinos