Fighting colorectal cancer: molecular epidemiology differences among Ashkenazi and Sephardic Jews and Palestinians

H. Darwish1, I. E. Trejo3, I. Shapira4, S. Oweineh1, M. Sughayer1,2, L. Baron4, E. Aljadeff9, M. Silbermann10, W. Sweidan2,6, D. Zilberg7, Z. Halpern7,8, H. Hibshoosh5 and N. Arber4,7,8,+

1 Department of Biochemistry, Al Quds University; 2 Abu-Dies, Muqased Hospital, East Jerusalem; 6 Augusta Victoria Hospital, Palestinian Authority; 3 Department of Pathology, 4 GI Oncology Unit, 7 Department of Gastroenterology, Tel Aviv Sourasky Medical Center; 8 Tel Aviv University, Tel Aviv; 9 Department of Mathematics, Technion-Israel Institute of Technology, Haifa, Israel; 10 Middle East Cancer Consortium, Israel; 5 Department of Pathology, Columbia University, New York, NY, USA

Received 29 October 2001; revised 14 December 2001; accepted 9 January 2002


    Abstract
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
Background:

To evaluate and compare differences in the molecular genetics among high-risk (Ashkenazi Jews), intermediate-risk (Sephardic Jews) and low-risk (Palestinians) groups for colorectal cancer who live in the same geographical region.

Patients and methods:

The 1995–1996 records from the Tel Aviv Medical Center and Muqased hospital (East Jerusalem) randomly identified patients with colorectal cancer. There were 25 patients from each ethnic group. Epidemiological data were obtained from interviews with the patients and from their hospital charts. The levels of cyclin D1, ß-catenine, p27, p53, Ki-67 and Her-2/neu proteins were determined by immunohistochemistry. The main outcome measures were the association between gene expression and colorectal incidence in the different ethnic groups.

Results:

Ashkenazi Jews have the highest rate of colorectal cancer, and are diagnosed at an early stage compared with Palestinians (72% and 33% of the cases are in Dukes’ A and B, respectively), and, hence, this may explain the better 5-year survival rate among this group. Sephardic Jews are diagnosed at a more advanced stage, the tumors are poorly differentiated and they lack p27. Palestinians have significantly higher cyclin D1 levels. There was a statistically significant inverse correlation between the expression of ß-catenine and cyclin D1, as well as p53 and p27 (P <0.05).

Conclusions:

Increased expression of cyclin D1, p53, Ki-67, ß-catenine and Her-2/neu, and decreased expression of p27 may be important events in the three ethnic groups with colorectal cancer. The lower mortality rate among Ashkenazi Jews may be partially explained by their better molecular biology profile.

Key words: Ashkenazi Jews, colorectal cancer, molecular epidemiology, Palestinians, Sephardic Jews


    Introduction
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
The onset of colorectal cancer (CRC) is less of an ‘event’ than the result of successive ‘stages’ in the protracted process of colorectal carcinogenesis [1]. Evidence is accumulating that shows, in addition to these changes, many types of tumors undergo abnormal events that directly affect the cell cycle (proliferation and apoptosis). Recent studies conducted by the group from Tel Aviv have demonstrated the importance of cell-cycle proteins in the multi-step process of gastrointestinal carcinogenesis [26].

Among Jews in Israel, CRC is the leading cause of death from cancer [7] while the Palestinian population appears to be protected from it [7]. It has long been noted that Israelis of European origin are more susceptible to CRC then AsianAfrican born Jews (42 per 100 000 versus 25 per 100 000, mean age standardized rates for 1989–1993) [7]. Studies from the United States and Australia have also corroborated the findings that Jewish populations, particularly those of European origin, are at high risk for CRC [8]. The incidence of the disease among Palestinians is estimated to be 7 per 100 000, approximately one-fifth that of Jews in Israel [7].

The genetic makeup of Jews and Arabs is almost identical: the two groups share common prehistoric ancestors and have a common geographical and ecological milieu. Israel is therefore an ideal laboratory to test these concepts because of a well defined population of subgroups that have retained their ethnicity. The intra-population variations in incidence, morbidity and mortality suggest that at least some genetic differences partially account for these obvious differences in the degree of susceptibility to CRC. Furthermore, there is a genetic susceptibility for germline predominant mutations among the Jewish population for developing breast (BRCA 1, 2 mutations) and colon (APC) cancers, which are known in the literature as the Ashkenazi Jewish mutations. It is also possible that lifestyle differences, and in particular, differences in diet, may be a protecting factor among Palestinians (as in the economically underdeveloped world), and a risk factor among the Ashkenazi Jews (as in the economically developed world).

These patterns of biological markers associated with CRC have not yet been evaluated prospectively in a well-designed study based on different ethnic populations. This is the first study to do so, setting out to define the molecular differences in CRC patients from highly susceptible Ashkenazi Jews, intermediate-risk Sephardic Jews and low-risk Palestinians.


    Material and methods
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
The 1995–1996 records of the Tel Aviv Medical Center were searched. Four hundred and twelve Ashkenazi and 198 Sephardic patients with tumors were identified. The records of Muqased Hospital (East Jerusalem), which is a referral center for Moslems who reside in the West Bank, identified 82 patients with CRC. Tissues from familial adeomatous polyposis or hereditary nonpolyposis colorectal cancer patients were excluded. Twenty-five patients with CRC from each ethnic group were chosen at random. Only patients with full clinical, laboratory and epidemiology data were included in the study. Epidemiological data were obtained from interviews with the patients and from their hospital charts. Follow-up information was obtained from reviewing the medical records from the hospital tumor registries, and from yearly contacts with the treating physicians or patients. The 5-year survival was obtained from the hospital records or by telephone contact with the patient’s physician or family. All the selected patients agreed to participate in the study.

Immunohistochemistry was used to determine levels of cyclin D1, ß-catenine, p27, p53, Ki-67 and Her-2/neu proteins.

Immunohistochemistry
All immunohistochemical analyses utilized the avidin–biotin complex immunoperoxidase technique (Figure C1). Tissue sections (5 µm) were mounted on poly-L-lysine-coated slides. After deparaffinization in Americlear (Baxter, McGaw Park, IL, USA) and absolute ethanol, sections were hydrated through a series of graded alcohol, distilled water, and phosphate-buffered saline at pH 7.4. The slides were immersed in 10 mM citrate buffer (pH 6) and heated by microwave to enhance antigen exposure for a total of 10 min. Potential background signals were blocked using goat or horse serum for 20 min, and primary anti-human antibodies were added and incubated in a high-humidity chamber. These included mouse monoclonal antibodies to cyclin D1 (2 h at 25°C), p53 (overnight at 4°C), and Ki-67 (1 h at 25°C) (Immunotech, Inc., Westbrook, ME, USA), ß-catenine (2 h at 25°C) and p27Kip1 (1 h at 25°C) (Transduction Laboratories, Lexington, KY, USA), and Her-2/neu (1 h at 25°C) (Oncogene Research Products, Cambridge, MA, USA). Although all the concentrations of primary antibodies gave good nuclear staining, the optimal concentrations that produced a minimal background were: a 1:200 dilution for cyclin D1 and p27; a 1:5 dilution for p53; a 1:75 dilution for ß-catenine; a 1:400 dilution for Ki-67; and a 1:50 dilution for Her-2/neu. Positive controls included breast adenocarcinoma for cyclin D1, Her-2/neu, Ki-67 and p53, gastric adenocarcinoma for p27, and colon adenocarcinoma for ß-catenine.

Subsequently, the Vectastain Elite ABC kit was used (Vector Laboratories, Burlingame, CA, USA) according to the manufacturer’s instructions. Color development was accomplished with a 0.375 mg/dl solution of a 3,3'-diamino-benzidine tetrahydrochloride (Sigma Chemical Co., St Louis, MO, USA) containing 0.003% hydrogen peroxide. The specificity of the antibodies was demonstrated by inhibition of immunohistochemical staining in positive controls, by pre-incubating the antibody with 1 mg of the immunizing peptides for 1 h at 40°C, representing about a 100-fold excess of peptide over antibody.

Interpretation of immunohistochemical staining
An experienced surgical pathologist (L.E.T.) interpreted the staining. Nuclear staining was considered positive if the chromogen was clearly detected in at least 10% of the nuclei within a microscopic field. Immunoreactivity was considered positive if the chromogen was detected in at least 5% of the nuclei within a microscopic field. Staining intensity included four scales: no staining (scale level 0); weakly positive and comparable with adjacent non-neoplastic epithelium (scale 1); moderately positive (scale 2); and strongly positive (scale 3). Scales 0 and 1 were regarded as negative and scales 2 and 3 as positive. Positive and negative controls were included within each batch of slides. To confirm reproducibility, 25% of the slides were chosen at random and another level of the same tissue was stained once again in the same batch. All batches were coded and blindly scored twice. Duplicate slides gave similar results.

Statistical analysis
The proportions of samples expressing the different genes from different histological categories were computed, and then compared across categories for selected factors (including gender, age, cigarette smoking, alcohol consumption, differentiation, Dukes’ stage, and mortality). When comparing proportions positive for staining across histological categories, Fisher’s exact test and chi-squared tests were employed [9]. McNemar’s test [10] was used for testing asymmetry in the association between the different genetic markers.


    Results
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
The epidemiological data in the three subgroups are summarized in Table 1. There were 39 males (52%) and 36 females (48%), with a median age of 66 years. The majority of tumors in the Ashkenazi Jews (72%) and 50% in the Sephardic Jews were in Dukes’ stages A and B, while those of the Palestinians group (67%) were in stages C and D. Tumor grade was similar in Ashkenazi Jews and Palestinians, while almost all tumors (92%) in the Sephardic patients were well differentiated. Smoking was most common among Palestinians, and alcohol consumption was very low in all three groups. Evidently, the 5-year survival is significantly higher among the Ashkenazi Jews (70%) compared with the Sephardic Jews (47%) and Palestinians (30%).


View this table:
[in this window]
[in a new window]
 
Table 1. Epidemiological data in the different subgroups
 
Molecular genetics abnormalities are summarized in Table 2. There were no statistical differences among the three groups in the expression of Ki-67 (~90%), p53 (~70%) and Her-2/neu (24–33%). The Sephardic Jews had a significantly lower expression of p27 (P <0.05), and the Palestinians a significantly higher cyclin D1 compared with both the other two groups. There was a significant inverse correlation between ß-catenine and cyclin D1, and p53 and p27 expression (P <0.05).


View this table:
[in this window]
[in a new window]
 
Table 2. Cell-cycle abnormalities among the different groups. Values are proportions with 95% confidence intervals
 
Increased level of cyclin D1 and decreased level of p27 were associated with a poorer prognosis. The levels of Ki-67, ß-catenine, and Her-2/neu did not predict outcome. Age, gender and tumor stage were not significantly associated with any of the above abnormalities.


    Discussion
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
In the present preliminary study, the clinical importance of several cell-cycle control proteins in Palestinians, Sephardic and Ashkenazi Jews was evaluated prospectively.

As in other published series [1, 1118], increased expression of cyclin D1, p53, Ki-67, ß-catenine and Her-2/neu, as well as decreased expression of p27 are all important events that may play a role in patients’ prognosis. Thus, cyclin D1 amplification [17], increased expression of cyclin D1 in tumors from the entire gastrointestinal tract [1820], and decreased expression of p27 in gastric [20] and colonic adenocarcinomas [21] are associated with a poorer prognosis.

Ashkenazi Jews have the highest rate of CRC, yet the 5-year survival rate was significantly better among this group (70%) compared with Sephardic Jews (47%) and Palestinians (30%). The better 5-year survival rate is probably due to the early diagnosis of the disease among Ashkenazi Jews (72% in Dukes’ A and B) as compared with only 33% among Palestinians. The comparisons of the immunohistochemical data among the three groups do not compare ‘like with like’, since the CRC comparison among the different groups was not of the same stage or grade. It is possible that Palestinians and Sephardic Jews show a poorer 5-year survival because 67% and 50% of their tumors, respectively, were at a more advanced stage compared with 28% of tumors in the Ashkenazi group. This difference in survival rate can therefore be attributed to earlier stage at the time of diagnosis. Indeed, screening programs for early detection of CRC are more popular among Ashkenazi Jews than among Sephardic Jews, whereas screening in fact does not exist among Palestinians. At the same time a significant decrease in the level of p27 among Sephardic Jews and a significant increase in cyclin D1 in Palestinians may partly contribute to the poorer prognosis among these groups.

It was recently demonstrated that ß-catenine regulates the expression of cyclin D1 [22]; however, in our series there was an inverse relationship in their expression. There might be other factors, still undetermined, that regulate the interaction between these two genes.

An interesting finding is the very high rate (92%) of well differentiated tumors among Sephardic Jews, as compared with Ashkenazi Jews (32%) and Palestinians (38%) (P <0.05). These findings have not been described previously. This difference might be an artifact of multiple comparisons in a relatively small study population. Since this difference was statistically significant it might be that another, yet unidentified, molecular marker exists in the former and not in the latter groups.

It is not clear from the present data what the bases are for the difference in CRC incidence among the three groups. However, what is indisputable is that noticeable histological and molecular differences were evident in tumors of patients from the three groups. Other epidemiological (and in particular nutrition) and molecular factors are involved, which might further explain the observed differences. The number of patients in each group, however, is too small and the results should be confirmed in larger scale studies.


    Acknowledgements
 
The authors are indebted to Dr Wendy Atkin and Dr Bernard Levin for their invaluable professional and scientific contribution, and to Mrs Heather Rockman for editorial assistance. Funding from the Middle East Cancer Consortium (to H.D. and N.A.), and the Israel Cancer Association (to N.A.) supported this study.



View larger version (92K):
[in this window]
[in a new window]
 
Figure 1. Example of p27, cyclin D1, Her-2/neu, p53, ß-catenine and Ki-67-positive immunoreactivity staining in the nuclei of colorectal tumor cells.

 

    Footnotes
 
+ Correspondence to: Prof. N. Arber, GI Oncology Unit, Department of Gastroenterology, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel 64239. Tel: +972-3-6974968/280; Fax: +972-3-6974622; E-mail: arber{at}post.tau.ac.il, nadir{at}tasmc.health.co.il Back


    References
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
1. Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990; 61: 759–767.[ISI][Medline]

2. Arber N, Hibshoosh H, Moss SF et al. Increased expression of cyclin D1 is an early event in multistage colorectal carcinogenesis. Gastroenterology 1996; 110: 669–674.[ISI][Medline]

3. Arber N, Sutter T, Miyake M et al. Increased expression of cyclin D1 and the Rb tumor suppresser gene in c-K-ras transformed rat enterocytes. Oncogene 1996; 12: 1903–1908.[ISI][Medline]

4. Arber N, Doki Y, Han EKH et al. Antisense to cyclin D1 inhibits the growth and tumorigenesis of human colon cancer cells. Cancer Res 1997; 7: 1569–1574.

5. Arber N, Han EKH, Sgambato A et al. Transformatiom of rat enterocytes by a c-K-ras oncogene increases resistance to growth inhibition and apoptosis induced by sulindac sulfide. Gastroenterology 1997; 113: 1892–1900.[ISI][Medline]

6. Kornmann M, Arber N, Korc M. Inhibition of basal and mitogen-stimulated pancreatic cancer cell growth by cyclin D1 antisense is associated with loss of tumorigenicity and potentiation of cytotoxicity to cisplatinum. J Clin Invest 1998; 101: 344–352.[Abstract/Free Full Text]

7. Rozen P, Lynch HT, Figer A et al. Israel Cancer Registry, Cancer in Israel, Ministry of Health, Jerusalem, 1997. Familial colon cancer in the Tel-Aviv area and the influence of ethnic origin. Cancer 1987; 60: 2355–2362.[ISI][Medline]

8. Kune S, Kune GA, Watson LF. The Melbourne CRC study: incidence findings by age, sex, site migrants and religion. Int J Epidemiol 1986; 15: 483–493.[Abstract]

9. Brier SS. Analysis of contingency tables under cluster sampling. Biometrika 1980; 67: 591–596.[ISI]

10. McNemar Q. Note on the sampling error of the difference between correlated proportions or percentages. Psyohometrika 1994; 12: 153–157.

11. Vogelstein B, Fearon ER, Hamilton SR et al. Genetic alterations during colorectal tumor development. N Engl J Med 1998; 319: 525–532.[Abstract]

12. Hamilton SR. The molecular genetics of colorectal neoplasia. Gastroenterology 1993; 105: 3–7.[ISI][Medline]

13. Ilyas M, Thomlinson IPM. Genetic pathways in colorectal cancer. Histopathology 1996; 28: 389–399.[ISI][Medline]

14. Reale MA, Fearon ER. Gene defects in colorectal tumorigenesis. In Young GP, Rozen P, Levin B (eds): Prevention and Early Detection of Colorectal Cancer. London: WB Saunders Company Ltd 1996; 63–86.

15. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer. Cell 1996; 87: 159–170.[ISI][Medline]

16. Pines J. Cyclins: wheels within wheels. Cell Growth Differ 1991; 2: 305–310.[ISI][Medline]

17. Jiang W, Kahn SM, Tomita N et al. Amplification and expression of the human cyclin D1 gene in esophageal cancer. Cancer Res 1992; 52: 2980–2983.[Abstract]

18. Arber N, Gammon MD, Hibshoosh H et al. Overexpression of cyclin D1 occurs in both squamous carcinomas and adenocarcinomas of the esophagus and in adenocarcinomas of the stomach. Hum Pathol 1993; 30: 1087–1092.

19. Banno S, Yoshikawa K, Nakamura S et al. Monoclonal antibody against PRAD1/cyclin D1 stains nuclei of tumor cells with translocation or amplification at the BCL-1 locus. Jpn J Cancer Res 1994; 85: 918–926.[ISI][Medline]

20. Yasui Y, Kudo S, Semba H et al. Reduced expression of cyclin-dependent kinase inhibitor p27kip1 is associated with advanced stage and invasiveness of gastric carcinomas. Jpn J Cancer Res 1997; 88: 625–629.[ISI][Medline]

21. Loda M, Cukor B, Tam SW et al. Increased proteasome-dependent degradation of the cyclin-dependent kinase inhibitor p27 in aggressive colorectal carcinomas. Nat Med 1997; 3: 231–234.[ISI][Medline]

22. Shtutman M, Zhurinsky J, Simcha I et al. The cyclin D1 gene is a target of the beta-catenin/LEF-1 pathway. Proc Natl Acad Sci USA 1999; 96: 5522–5527.[Abstract/Free Full Text]





This Article
Abstract
Full Text (PDF)
E-letters: Submit a response
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (5)
Disclaimer
Request Permissions
Google Scholar
Articles by Darwish, H.
Articles by Arber, N.
PubMed
PubMed Citation
Articles by Darwish, H.
Articles by Arber, N.