Familial risk of endometrial cancer after exclusion of families that fulfilled Amsterdam, Japanese or Bethesda criteria for HNPCC

J. Lorenzo Bermejo1,*, F. L. Büchner1 and K. Hemminki1,2

1 Division of Molecular Genetic Epidemiology, German Cancer Research Center (DKFZ), Heidelberg, Germany; 2 Department of Biosciences at Novum, Karolinska Institute, Huddinge, Sweden

Received 29 September 2003; revised 2 December 2003; accepted 22 December 2003


    ABSTRACT
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
Background:

Endometrial cancer is the second most common lesion within hereditary non-polyposis colorectal cancer (HNPCC) syndrome. The importance of the non-HNPCC genetic predisposition to endometrial cancer is unclear, and the familial aggregation of endometrial cancer after exclusion of HNPCC families may offer valuable clues about the involvement of non-HNPCC-related genes.

Patients and methods:

The families of the nationwide Swedish Family-Cancer Database were classified as HNPCC families according to the Amsterdam I or II, the modified Amsterdam, the Japanese and the Bethesda criteria. Standardized incidence ratios (SIRs) for endometrial cancer when parents or siblings were diagnosed with cancer at the most common sites were calculated before and after exclusion of HNPCC families.

Results:

The proportion of individuals in the families with endometrial cancer was highest when the criteria required three cancers within a family or multiple HNPCC-related cancers in the same individual. Consideration of the Amsterdam or the Japanese criteria hardly reduced the familial aggregation of endometrial cancer. After exclusion of families that fulfilled the Bethesda criteria, SIRs were significant when the parents were diagnosed with endometrial or thyroid gland cancers; 75.7% (95% confidence interval 60% to 99.1%) of the familial cases of endometrial cancer were not related to HNPCC according to the Bethesda criteria. The reduction of SIRs for cancers at the colon, pancreas, prostate and ovary was limited when the Bethesda criteria were applied. However, the Bethesda criteria identified most of the familial aggregation when endometrial cancers were diagnosed before the age of 55 years.

Conclusions:

The data suggest that additional effects, not related to HNPCC, contribute to the familial aggregation of endometrial cancer.

Key words: Amsterdam criteria, Bethesda guidelines, endometrial cancer, familial risk, HNPCC, Japanese criteria


    Introduction
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
Most endometrial cancers are sporadic [1] and they are often associated with the estrogen–progestin imbalance: chronic exposure of women to estrogen during their lives, with the exception of pregnancy periods, is a well-established risk factor [2]. The best documented form of familial endometrial cancer is a part of the hereditary non-polyposis colorectal cancer (HNPCC) syndrome, characterized by an early age of onset [16]. After colorectal cancer, endometrial cancer is the second most common tumor in HNPCC and, in women, it can even exceed the incidence of colorectal cancer [7]. The importance of genetic predisposition to endometrial cancer that is not related to HNPCC remains unclear.

Different clinical criteria have been established to classify HNPCC families. The Amsterdam criteria I [8], the modified Amsterdam criteria [9, 10] and the Japanese criteria [11] only consider colorectal cancer (Table 1). Endometrial, small bowel, ureter and renal pelvis cancers are included in the Amsterdam II criteria [5]. The Bethesda guidelines, developed to guide the analysis of microsatellite instability, additionally consider stomach, ovarian and hepatobiliary tract cancers, as well as colon adenomas [7]. Even sebaceous gland adenomas, also features of Muir–Torre syndrome, cancers of the nervous system and keratoacanthomas have been associated with the HNPCC syndrome [57, 12].


View this table:
[in this window]
[in a new window]
 
Table 1. Clinical criteria for diagnosis of hereditary non-polyposis colorectal cancer (HNPCC)
 
The subject of this study is the familial aggregation of endometrial cancer that is not related to HNPCC. The association of endometrial cancer with any cancer in families was assessed before and after exclusion of HNPCC families. The specific questions were the association of endometrial cancer with cancer at other sites and the dependence of these associations on HNPCC. We used the 2002 update of the nationwide Swedish Family-Cancer Database, which offers unique possibilities for reliable estimation of familial risk because the data on family relationships and cancers are obtained from registered sources of practically complete coverage and free from bias [13].


    Patients and methods
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
The Swedish Family-Cancer Database was created in the middle of the 1990s by linking census information, death notifications and the administrative family register at Statistics Sweden to the Swedish Cancer Registry. The database includes all persons who were born in Sweden after 1931 to their biological parents. The database was updated at the end of 2002 to include more than 10.2 million individuals. The Swedish Cancer Registry relies on separate compulsory notifications of cases from clinicians who diagnosed a neoplasm, and pathologists/cytologists. A four-digit diagnostic code according to the 7th revision of the International Classification of Diseases (ICD-7) was used since 1992 and ICD-O-2. In this study, small intestine, colon, colon in situ and rectal cancers included only adenocarcinomas; upper aerodigestive tract cancer included lip, tongue, salivary gland, mouth and pharynx; eye cancer included melanoma only; and leukemia codes included ICD-7 204–209. The percentage of histologically or cytologically verified cases of cancer was close to 100% [13].

Only families from the database with at least four generations were included in the study. If one founder parent of the family was missing or married several times, the family was not considered. Families were classified as HNPCC families according to the Amsterdam criteria I or II, the modified Amsterdam criteria, the Japanese criteria and the Bethesda criteria, as described in Table 1.

Follow-up started at birth, immigration date or first year of the study (1961), whichever came latest. Follow-up terminated at diagnosis of cancer, death, emigration date, 31 December 2000 or the age of diagnosis as specified in the study. The standardized incidence ratio (SIR) was used to estimate familial relative risk. SIR was based on only parents diagnosed (SIRparent) or only siblings diagnosed (SIRsibling) as probands. The SIR was calculated as the ratio of observed to expected number of cases. The expected numbers were computed from age (5 year bands), parity (six groups from ‘any parturition’ to ‘more than five parturitions’), socioeconomic status (six groups), age at first birth (five groups, 5 year bands between ‘before age of 20’ and ‘after age of 35’) and residential area (four groups) standardized incidence rates. The confidence intervals (CIs) were calculated assuming a Poisson distribution [14]; 99% CIs are presented to address the issue of multiple comparisons.


    Results
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
The database contained 509 985 families with at least four generations; the mean size of the families was 14.1 individuals. The age of the individuals in the first generation was not limited, but the maximum age of individuals in the second generation was 68 years. Among the over 6.1 million individuals who constituted the families, 42 070 persons were diagnosed with colorectal cancer and 26 689 persons were diagnosed with endometrial cancer.

The number and size of families that fulfilled the different clinical criteria are presented in Table 2. Amsterdam I criteria required three cases of colorectal cancer within a family, and classified 18 families. Amsterdam II criteria required three cases of colorectal, endometrial, small bowel, ureter or renal pelvis cancer within a family, and classified 37 families. Among the 5593 persons in the second generation diagnosed with colorectal cancer, 35 (0.6%) belonged to families that fulfilled the Amsterdam I criteria and 48 (0.9%) belonged to families that fulfilled the Amsterdam II criteria. Set 1 of the modified Amsterdam criteria and set B of the Japanese criteria required two diagnosed individuals in a family. The modified Amsterdam criteria classified 382 families and the Japanese criteria classified 279 families. Condition 2 of the Bethesda criteria required multiple HNPCC-related cancers in one individual, and classified 2826 families. Condition 3 of the Bethesda criteria rested on two cases within a family, and classified 368 families. Condition 4 of the Bethesda criteria was based on one colorectal or endometrial cancer before the age of 45 years, and classified 4403 families. Condition 7 of the Bethesda criteria considered early-onset colorectal adenomas, and classified 259 families.


View this table:
[in this window]
[in a new window]
 
Table 2. Number, size, number of cases of endometrial cancer and percentage of individuals with endometrial cancer in families classified according to different clinical criteria for hereditary non-polyposis colorectal cancer (HNPCC)
 
The number and percentage of individuals with endometrial cancer in the classified families is also presented in Table 2. The percentage of individuals with endometrial cancer was highest in Amsterdam II families (4.06%), followed by families that fulfilled set 2 of the modified Amsterdam criteria (1.65%) and families that fulfilled condition 2 of the Bethesda criteria (1.33%).

Table 3 shows the number of cases of endometrial cancer, SIRparent and 95% CIs for women when their parents were diagnosed with cancer at any site. SIRparent in the general population was significant (95% CI did not include 1) for endometrial, thyroid gland, colon, prostate and ‘all’ cancer sites, with a range from 2 to 1.09. The 99% CI of SIRparent did not include 1 for the endometrium, colon, prostate and ‘all’ sites. The lowest confidence limit of the SIRparent was >0.9 for pancreas, urinary bladder and colon in situ. After exclusion of Amsterdam I or Amsterdam II families, the SIRs were almost identical to those for the general population, and are not presented. SIRs were only slightly modified when HNPCC families were excluded according to the modified Amsterdam criteria or the Japanese criteria. If Bethesda families were excluded, SIRparent decreased, and was only significant for the thyroid gland (1.95), the endometrium (1.92) and ‘all’ sites (1.07); the 99% CI of SIRparent for the endometrium did not include 1 when the Bethesda criteria were applied. The Bethesda criteria identified 17 pairs where both a mother and a daughter had endometrial cancer. Among the 17 daughters, eight were diagnosed with endometrial cancer before the age of 45 years, but none was affected by multiple endometrial cancer.


View this table:
[in this window]
[in a new window]
 
Table 3. Risk for endometrial cancer in offspring by parental cancer before and after exclusion of hereditary non-polyposis colorectal cancer (HNPCC) families according to the modified Amsterdam, the Japanese and the Bethesda criteria
 
Data for cancer among siblings are presented in Table 4. SIRsibling was significant for the endometrium (2.31), the ovary (1.85) and the colon (1.73). The lowest confidence limit of the SIRsibling was 0.97 for breast cancer. SIRs were minimally affected when families that fulfilled the modified Amsterdam criteria were excluded. Families that fulfilled the Japanese criteria included one sibling pair of concordant endometrial cancer, and the SIRsibling for endometrium was no longer significant in the remaining families. If Bethesda families were excluded, the SIRsibling was not significant for any cancer site.


View this table:
[in this window]
[in a new window]
 
Table 4. Risk for endometrial cancer among siblings before and after exclusion of hereditary non-polyposis colorectal cancer (HNPCC) families according to the modified Amsterdam, the Japanese and the Bethesda criteria
 
The SIRparent for concordant endometrial cancer before specific ages is shown in Figure 1. Exclusion of the families by the Japanese or the modified Amsterdam criteria resulted in SIRs similar to those of the general population. The Bethesda criteria removed early-onset endometrial cancers, but the SIRs were significant for cancers diagnosed beyond the age of 55 years.



View larger version (18K):
[in this window]
[in a new window]
 
Figure 1. Standardized incidence ratio (SIR) for endometrial cancer when mothers were diagnosed with endometrial cancer at restricted ages before and after exclusion of hereditary non-polyposis colorectal cancer (HNPCC) families according to the modified Amsterdam, the Japanese and the Bethesda criteria (discontinuous lines represent non-significant SIRs).

 

    Discussion
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
The familial aggregation of endometrial cancer in the Swedish Family-Cancer Database has been analyzed before and significant risks were found for daughters of mothers diagnosed with endometrial and colon cancer [15]. The present update of the database includes 937 new cases of endometrial cancer in daughters, as well as 1753 new cases of endometrial cancer in mothers, which permits a more accurate estimation of risks at early ages, and thus a closer analysis of the relationship between endometrial cancer and the HNPCC syndrome. Other features of this study were the standardization for parity and age at first childbirth and the analysis of the risk of endometrial cancer when the proband was affected with endometrial cancer before a specified age. To deal with the issue of multiple comparisons, 99% CIs were computed. However, 99% CIs imposed conservative criteria, since some well-established aggregations, such as endometrial cancer between sisters (Table 4), were not significant at the 99% confidence level. Significant findings at the 95% confidence level are examined for biological plausibility and consistency with previous literature in this Discussion.

Only families with at least four generations were included in the study, but the maximum age of individuals in the second generation (68 years) is a limitation of the present analysis. The frequency of patients with colorectal cancer in the second generation fulfilling the Amsterdam criteria I was 0.6% (95% CI 0.4% to 0.9%) and 0.9% (95% CI 0.6% to 1.1%) of the patients that fulfilled the Amsterdam II criteria. These figures are in agreement with the frequency of 1.2% (95% CI 0.1% to 2.2%) reported for HNPCC in a recent Swedish study [16]. However, this study relied entirely on registered data of complete coverage. The issue of ascertainment bias, when the clinical criteria require three cases within a family, has been discussed previously [17]. Another limitation of this study was that no signet-ring histology could be considered (condition 6 of the Bethesda criteria); however, this omission was unlikely to have any large effect on the results.

The International Collaborative Group on HNPCC established the Amsterdam I criteria to provide a basis for uniformity in collaboration studies [8]. The criteria were restrictive, since extracolonic malignancies, clearly an important component of HNPCC, were not considered, and small families were unlikely to fulfil the criteria [3]. Only 18 families from the database were classified as Amsterdam I families. On the other hand, when Amsterdam I criteria were met, the chance of HNPCC was high, i.e. Amsterdam I criteria aimed at specificity more than sensitivity. Although Amsterdam I criteria are widely accepted and used in clinical counselling [5], some investigators consider the criteria to be too strict [7, 18], and alternatives have been developed. The modified Amsterdam criteria [810] require two diagnosed relatives, and consider early-onset endometrial cancer. The Japanese criteria [11] also classify families with two cases, and additionally take into consideration the clinical features of HNPCC: early-onset, multiple synchronous or metachronous colorectal cancers, and right colon involvement. The Bethesda guidelines proposed by Rodriguez-Bigas et al. [7] target the evaluation of colorectal tumors for microsatellite instability or mismatch repair gene testing. The Bethesda guidelines are less restrictive and more sensitive, but also less specific, than the Amsterdam criteria. The most recent Amsterdam II criteria [5], developed by the International Collaborative Group on HNPCC, include colorectal, endometrial, small bowel, ureter and renal pelvis cancers, but keep the requirement to three cases in the family. In this study, 37 families were classified according to Amsterdam II and 7296 families were classified according to the Bethesda criteria as HNPCC families. The proportion of individuals in the Amsterdam II families with endometrial cancer was 4.06%, compared with 0.89% in the Bethesda families. Syngal et al. [19] compared different clinical criteria for HNPCC by using the sensitivity and specificity for mutations in MSH2 and MLH1 as a distinctive feature. A comparison of the clinical criteria based on the risks for specific sites, such as the endometrium, would be also of great interest.

We found significant risks for endometrial cancer when the relatives were diagnosed with colon, endometrial, thyroid gland, ovarian and prostate cancers. The lowest limit of the 95% CI was >0.9, even when parents were diagnosed with pancreatic, urinary bladder and colon in situ cancers (Table 3), and when sisters were diagnosed with breast cancer (Table 4). The link between pancreatic, urinary bladder and colon in situ cancers with HNPCC is well established, while the inclusion of the breast among the HNPCC cancer sites remains controversial. The likelihood of HNPCC contribution in the association of endometrial cancer with other cancers depended on the age of onset and the specific cancer site. SIRs for the colon were not significant after exclusion of the Bethesda families, but the lowest confidence limit remained high, and some association between colon and endometrial cancer, independent of HNPCC, is probable, although it is also plausible that the clinical criteria did not encompass all HNPCC cases. A similar interpretation held when parents were diagnosed with colon in situ, pancreas, prostate and urinary bladder cancers. SIRs when sisters were affected with breast or ovarian cancer remained relatively high, even in the Bethesda-negative cases. These data suggest that HNPCC plays a minor role, if any, in the association of breast/ovary and endometrial cancer. In general, the reduction of SIRs when the Bethesda families were excluded depended largely upon the age at diagnosis. If cancers were diagnosed at any age in parents (Table 3), SIRs remained significant for endometrial and thyroid gland cancers after exclusion of the Bethesda families. In all, 75.7% (95% CI 60% to 99.1%) of the familial cases of endometrial cancer were probably not related to HNPCC. If only conditions 1–3 of the Bethesda criteria were considered, as has recently been recommended [19], the proportion of endometrial cancers not related to mismatch repair gene mutations increased to as much as 87.1%. A higher proportion of the familial aggregation was recognized by the Bethesda criteria if cancers were diagnosed before the age of 68 years in siblings (Table 4). Results presented in Figure 1 show the performance of the Bethesda criteria for endometrial cancer diagnosed at different ages. The SIR for familial endometrial cancer, not related to HNPCC, increased with increasing ages of diagnosis. Infrequent mismatch repair gene alterations (i.e. MSH6 mutations) could explain in part the aggregation of late-onset endometrial cancer [20].

In conclusion, the data suggest that germline mismatch repair mutations explain a limited amount of the familial aggregation of endometrial cancer. The remaining genetic and environmental causes of the aggregation are a challenge for future studies.


    Acknowledgements
 
The Family-Cancer Database was created by linking registers maintained at Statistics Sweden and the Swedish Cancer Registry.


    FOOTNOTES
 
* Correspondence to: Dr J. Lorenzo Bermejo, Division of Molecular Genetic Epidemiology, German Cancer Research Center (DKFZ), Im Neuenheimer Feld 580, D-69120 Heidelberg, Germany. Tel: +49-6221-421805; Fax: +49-6221-421810; E-mail: J.Lorenzo{at}dkfz.de Back


    REFERENCES
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
1. Whelan AJ, Babb S, Mutch DG et al. MSI in endometrial carcinoma: absence of MLH1 promoter methylation is associated with increased familial risk for cancers. Int J Cancer 2002; 99: 697–704.[CrossRef][ISI][Medline]

2. Emons G, Fleckenstein G, Hinney B et al. Hormonal interactions in endometrial cancer. Endocr Relat Cancer 2000; 7: 227–242.[Abstract/Free Full Text]

3. Lynch HT, Smyrk T, Lynch J. An update of HNPCC (Lynch syndrome). Cancer Genet Cytogenet 1997; 93: 84–99.[CrossRef][ISI][Medline]

4. Ichikawa Y, Lemon SJ, Wang S et al. Microsatellite instability and expression of MLH1 and MSH2 in normal and malignant endometrial and ovarian epithelium in hereditary nonpolyposis colorectal cancer family members. Cancer Genet Cytogenet 1999; 112: 2–8.[CrossRef][ISI][Medline]

5. 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: 1453–1456.[Medline]

6. Lynch HT, de la Chapelle A. Genetic susceptibility to non-polyposis colorectal cancer. J Med Genet 1999; 36: 801–818.[Abstract/Free Full Text]

7. Rodriguez-Bigas MA, Boland CR, Hamilton SR et al. A National Cancer Institute Workshop on Hereditary Nonpolyposis Colorectal Cancer Syndrome: meeting highlights and Bethesda guidelines. J Natl Cancer Inst 1997; 89: 1758–1762.[Free Full Text]

8. Vasen HF, Mecklin JP, Khan PM, Lynch HT. The International Collaborative Group on Hereditary Non-Polyposis Colorectal Cancer (ICG-HNPCC). Dis Colon Rectum 1991; 34: 424–425.[ISI][Medline]

9. Bellacosa A, Genuardi M, Anti M et al. Hereditary nonpolyposis colorectal cancer: review of clinical, molecular genetics, and counseling aspects. Am J Med Genet 1996; 62: 353–364.[CrossRef][ISI][Medline]

10. Benatti P, Sassatelli R, Roncucci L et al. Tumour spectrum in hereditary non-polyposis colorectal cancer (HNPCC) and in families with ‘suspected HNPCC’. A population-based study in northern Italy. Colorectal Cancer Study Group. Int J Cancer 1993; 54: 371–377.[ISI][Medline]

11. Fujita S, Moriya Y, Sugihara K et al. Prognosis of hereditary nonpolyposis colorectal cancer (HNPCC) and the role of Japanese criteria for HNPCC. Jpn J Clin Oncol 1996; 26: 351–355.[Abstract]

12. Aarnio M, Sankila R, Pukkala E et al. Cancer risk in mutation carriers of DNA-mismatch-repair genes. Int J Cancer 1999; 81: 214–218.[CrossRef][ISI][Medline]

13. Cancer Incidence in Sweden, 2000. Stockholm: National Board of Health and Welfare 2002; 2002: 118.

14. Esteve J, Benhamou E, Raymond L. Statistical methods in cancer research. Volume IV. Descriptive epidemiology. IARC Sci Publ 1994; 128: 1–302.[Medline]

15. Hemminki K, Vaittinen P, Dong C. Endometrial cancer in the family-cancer database. Cancer Epidemiol Biomarkers Prev 1999; 8: 1005–1010.[Abstract/Free Full Text]

16. Olsson L, Lindblom A. Family history of colorectal cancer in a Sweden county. Fam Cancer 2003; 2: 87–93.[CrossRef][Medline]

17. Hemminki K. Recharacterization of hereditary nonpolyposis colorectal cancer families from a population-based series of cases. J Natl Cancer Inst 2001; 93: 651–652.[Free Full Text]

18. Potter JD. Colorectal cancer: molecules and populations. J Natl Cancer Inst 1999; 91: 916–932.[Abstract/Free Full Text]

19. Syngal S, Fox EA, Eng C et al. Sensitivity and specificity of clinical criteria for hereditary non-polyposis colorectal cancer associated mutations in MSH2 and MLH1. J Med Genet 2000; 37: 641–645.[Abstract/Free Full Text]

20. Wijnen J, de Leeuw W, Vasen H et al. Familial endometrial cancer in female carriers of MSH6 germline mutations. Nat Genet 1999; 23: 142–144.[CrossRef][ISI][Medline]





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
Disclaimer
Request Permissions
Google Scholar
Articles by Bermejo, J. L.
Articles by Hemminki, K.
PubMed
PubMed Citation
Articles by Bermejo, J. L.
Articles by Hemminki, K.