Body mass index: Relationship to clinical, pathologic and features of microsatellite instability in endometrial cancer

Gynecologic Oncology 104 (2007) 535 – 539 www.elsevier.com/locate/ygyno

Body mass index: Relationship to clinical, pathologic and features of microsatellite instability in endometrial cancer Carolyn K. McCourt a , David G. Mutch a , Randall K. Gibb a , Janet S. Rader a , Paul J. Goodfellow b , Kathryn Trinkaus c , Matthew A. Powell a,⁎ a

Department of Obstetrics and Gynecology, Washington University School of Medicine, Saint Louis, MO, USA b Department of Surgery, Washington University School of Medicine, Saint Louis, MO, USA c Division of Biostatistics, Washington University School of Medicine, Saint Louis, MO, USA Received 31 May 2006 Available online 14 November 2006

Abstract Objectives. There is a well known association between obesity and endometrial cancer. We sought to examine the relationships between body mass index (BMI), as a measure of obesity, and known demographic, clinical, and molecular characteristics of microsatellite instability and MLH1 promoter methylation in a cohort of patients with endometrial cancer. Methods. Corpus cancer specimens were prospectively obtained from 473 consecutively enrolled patients between 1992 and 2004. Clinical and pathologic data were extracted from review of the medical record. Microsatellite instability (MSI) was evaluated in all tumors, and methylation of the MLH1 promoter was determined for MSI positive tumors. Results. The median (SD) age and BMI were 64.8 years (11.9) and 33.5 (9.4), respectively. Histology included 376 endometrioid (79%), 69 serous/clear cell or mixed (15%), and 28 sarcomas (6%). Median BMI was 32.4 for endometrioid, 31.0 for serous/clear cell or mixed, and 27.8 for sarcomas (p = 0.14). BMI was negatively associated with age at surgery (p < 0.01). The remainder of analyses excluded sarcoma histology. BMI was associated with stage of disease; patients with stage I/II disease had significantly higher BMI than those with stage III/IV disease (32.6 vs. 30.6; p = 0.02). In relation to molecular features of endometrial cancer, BMI was significantly different between MSI positive tumors compared to MSI negative tumors (30.3 vs. 32.7; p = 0.02). MSI was also significantly different between tumor histology, occurring with a higher frequency in Type I than Type II tumors (p < 0.01). Conclusions. The majority of endometrial cancer patients are obese. Those with higher BMI are more likely to be younger, present with early stage disease, and have MSI negative tumors. © 2006 Elsevier Inc. All rights reserved. Keywords: Uterine cancer; Microsatellite instability; Body mass index

Introduction Endometrial cancer is the most common gynecologic malignancy in the United States. Obesity is one of several well-known risk factors for endometrial cancer, and in the U.S., obesity is a major public health problem of epidemic proportion [1,2]. Recent statistics reveal that 33% of women over age 20 in the U.S. are obese (BMI > 30) [3]. The association between ⁎ Corresponding author. Department of Obstetrics and Gynecology, Division of Gynecologic Oncology, 4911 Barnes Jewish Hospital Plaza, Saint Louis, MO 63110, USA. Fax: +1 314 362 2893. E-mail address: [email protected] (M.A. Powell). 0090-8258/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2006.09.019

increased weight and endometrial cancer is hypothesized to be the result of a mitogenic milieu of relative unopposed or elevated peripheral estrogen. This environment develops due to both aromatization of androgens to estrogens in adipose tissue, as well as decreased sex hormone binding globulin [4]. Molecular studies of endometrial cancer have identified several genetic alterations which are thought to play a critical role in the development of malignancy [5]. Microsatellite instability (MSI) is a molecular phenotype originally elucidated in hereditary nonpolyposis colon cancer (HNPCC), and was subsequently found in association with 20–30% of sporadic endometrial cancers [6]. Microsatellites are tandem repeats of simple nucleotide sequences that occur throughout the human

536

C.K. McCourt et al. / Gynecologic Oncology 104 (2007) 535–539

genome. Due to the nature of these repetitive sequences, microsatellites are prone to replication errors secondary to slippage of the DNA polymerase complex. Normally, replication errors are corrected by a DNA mismatch repair system, however, malfunction of this system results in persistence of replication errors within microsatellites, resulting in microsatellite instability. MLH1, PMS2, MSH2, MSH6, and MSH3 are key proteins involved in the highly conserved mechanism of DNA mismatch repair [7]. While MLH1 expression is often lost in the setting of MSI, mutations in this gene are not commonly seen in sporadic endometrial cancer. Rather, methylation of its promoter region results in loss of expression of this protein, and this phenomenon occurs in approximately 75% of sporadic endometrial cancers with MSI [8,9]. Previous studies have examined some clinical and pathologic features associated with microsatellite instability, but results have been conflicting and limited by small sample sizes [10– 13]. In addition, none have examined the role of MLH1 expression with these clinical and pathologic variables. We sought to examine the relationships between body mass index (BMI), as a measure of obesity, and the clinical, pathologic, and molecular features of DNA mismatch repair, specifically MLH1 promoter methylation, in endometrial cancer patients. Materials and methods

Table 1 Demographics for 473 uterine cancer cases Variable Age (missing = 1) Overall <60 60–80 >80 Race Caucasian African American Other a Stage I/II III/IV Unknown/Unstaged Histology Endometrioid Serous/Clear cell/Mixed Sarcoma BMI (missing = 73) Overall <25 25–40 >40 Microsatellite Stable Unstable–MLH1 methylated Unstable–MLH1 unmethylated a

This study was approved by the Human Studies Committee at Washington University School of Medicine. Patients were approached about participation in collection and banking of tumor tissue prior to surgery. Tumor tissue was obtained from women with primary uterine corpus cancers at the time of hysterectomy. The study population is representative of our clinic population. There was no selection for specific histologic types, tumor grade, or stage, nor was there selection of sporadic or inherited forms of uterine cancer. Information on patient height and weight, from which BMI was calculated, was obtained from either the anesthesia record, electronic database with admission vitals from the day of surgery, or office vitals from the visit immediately prior to surgery. Demographic and pathologic data were obtained by review of medical records and pathology reports. Polymerase chain reaction (PCR)-based analysis of microsatellite stability was performed as previously described [14,15], utilizing the NCI designated markers for MSI detection in colorectal cancer (BAT25, BAT26, D2S123, D5S346, D17S250) [16]. Tumors were classified as having MSI if two or more markers revealed novel sized alleles. Normal DNA was prepared from peripheral blood leukocytes, or for a small number of cases from normal myometrium. Methylation status of the MLH1 promoter was assessed in all MSI tumors by combined bisulfite restriction analysis (COBRA) of cytosines

N (%)

Median (SD)

472 (99.8) 163 (34.5) 259 (54.9) 50 (10.6)

64.8 (11.9) – – –

391 (82.7) 80 (16.9) 2 (0.4)

– – –

339 (72) 111 (23) 23 (5) 376 (79.5) 69 (14.6) 28 (5.9)

– – –

400 (84.6) 72 (18) 233 (58.3) 95 (23.8)

33.5 (9.4) – – –

339 (71.7) 95 (20.1) 39 (8.2)

– – –

Native American, Asian.

−229, −231 (BstUI digestion for − 250 and − 252 relative to the ATG initiation codon), and − 241 (Sau3AI digestion for the cytosine at − 262 relative to the ATG) [17]. Fisher's Exact test and Wilcoxon–Mann–Whitney test, as well as Kruskal–Wallis test were utilized for statistical analysis of categorical and continuous variables, respectively. Multivariate logistic regression was utilized to examine the associations between different variables on presenting with advanced stage disease, while controlling for potential confounders based on univariate analysis. A p value of <0.05 was considered statistically significant.

Results Four hundred seventy-three corpus cancer specimens were collected between 1992 and 2004. Microsatellite instability was assessed in all tumors, and methylation of the MLH1 promoter region was determined in all MSI positive tumors (Fig. 1).

Fig. 1. Microsatellite schema.

C.K. McCourt et al. / Gynecologic Oncology 104 (2007) 535–539

537

Table 2 Pathologic and molecular characteristics of uterine cancers by BMI Variable Race b (N = 471) Caucasian African American Histology (N = 473) Endometrioid Serous/Clear cell/ Mixed Sarcoma Grade-Endometrioid (N = 376) 1 2 3 Stage (N = 446) I/II III/IV Unknown/ Unstaged Microsatelliteb (N = 446) Stable Unstable MLH1 promoter b (N = 128) Methylated Unmethylated a b c d

BMI <25

BMI 25–40

BMI >40

Missing Median (SD)

67 4

198 35

76 18

50 23

31.5 (9.2) 34.2 (10.4)

61 7

185 43

86 9

44 11

32.4 (9.7) 31.1 (8.0)

4

5

0

18

27.8 (7.2)

p value a < 0.01

0.14 0.45 c

0.16 Fig. 2. Age at surgery and BMI. 34 17 10

88 64 33

52 25 9

21 16 7

33.1 (10.8) 31.8 (8.2) 30.7 (8.2) 0.02 d

48 19 1

167 58 3

74 13 8

40 10 5

32.6 (9.5) 30.6 (7.9) 44.91 (12.5) 0.02

45 23

157 71

77 18

39 16

32.7 (9.9) 30.3 (8.2) 0.53

15 8

54 17

11 7

11 5

31.3 (7.6) 28.4 (9.6)

p values represent comparisons between medians. Excludes one Asian and one Native American. Excludes carcinosarcoma, leiomyosarcoma, stromal sarcoma. Excludes unknown stage/unstaged.

Demographic data are displayed in Table 1. Total hysterectomy, bilateral salpingoophorectomy, pelvic washing, and at least lymph node sampling were performed in patients with evidence of, or suspicion of, myometrial invasion. Staging information was evaluable in 451/473 (95%) patients. Information was available to calculate BMI in 400/473 (85%) patients. In those 73 patients for whom BMI could not be calculated, height was the missing data point. Overall median age (standard deviation) was 64.8 years (± 11.9), and overall median BMI was 33.5 (± 9.4). African Americans comprised approximately 17% of the cohort, while 82.7% were Caucasian. Tumor histology included 376 (79.5%) endometrioid adenocarcinomas, 69 (14.6%) serous carcinomas, clear cell carcinomas, or mixed histology, and 28 (5.9%) sarcomas. Due to small sample size and inherent disparate tumor characteristics, patients with sarcomatous histology were excluded from the remainder of the analyses. Body mass index was utilized as a measure of obesity. Table 2 illustrates that BMI was not significantly different when comparing histologic type (p = 0.14). In contrast, African American women had a greater median BMI compared to Caucasian women (34.2 vs. 31.5, p < 0.01). Women with disease confined to the uterus were also noted to have a significantly higher BMI than those with advanced disease (32.6 vs. 30.6, p = 0.02). Microsatellite instability was related to body mass index. Women with MSI positive tumors had a lower BMI than

those women whose tumors retained microsatellite stability (30.3 vs. 32.7, p = 0.02). However, BMI did not differ significantly based on MLH1 promoter methylation (31.3 vs. 28.4, p = 0.53). There was a moderately strong negative correlation between BMI and age at the time of surgery, with a Spearman rank correlation of − 0.28. Fig. 2 illustrates declining BMI with advancing age. When examining the cohort in regard to microsatellite instability (Table 3), MSI occurred more frequently in Caucasian women compared to African American women (31% vs. 18%, p = 0.05). There was significantly more MSI detected in tumors with endometrioid histology than those with serous, clear cell, or mixed histology (32% vs. 10%, p < 0.01). Additionally, within endometrioid tumors, MSI was significantly different between tumor grade, with the majority of MSI accounted for in grade 1 and 2 tumors (p < 0.01). Microsatellite instability was not related to FIGO stage or age at surgery. However, there was a difference in median age between women with MSI tumors with MLH1 promoter methylation. Specifically, women with MLH1 methylation were older than women without MLH1 methylation (median age 67 years vs. 57.3 years, p < 0.01).

Table 3 Relationship between tumor microsatellite instability and clinicopathologic variables a Variable

MSI negative

MSI positive

Age median (SD) (N = 472) Race b (N = 443) Caucasian African American Histology (N = 445) Endometrioid Serous/Clear cell/Mixed Grade-endometrioid (N = 376) 1 2 3 Stage (N = 446) I/II III/IV Unknown/Unstaged

64.9 (12.4)

63.4 (10.8)

257 58

115 13

255 62

121 7

144 66 45

51 56 14

231 75 12

98 25 5

a b

p value 0.63 0.05

< 0.01

< 0.01

0.52

Excludes carcinosarcoma, leiomyosarcoma, stromal sarcoma. Excludes one Asian and one Native American.

538

C.K. McCourt et al. / Gynecologic Oncology 104 (2007) 535–539

We then constructed multivariate logistic regression models including key variables, such as BMI, histology, grade, and race. As expected, tumor histology and grade were independent predictors, with non-endometrioid histology (OR 3.54, 95%C.I. 1.95–6.44), and increasing grade (OR 2.01–6.44, 95%C.I. 1.06–11.95) associated with stage III or IV disease. Additionally, BMI was an independent predictor of stage of disease (OR 0.97, 95%C.I. 0.94–0.99). Specifically, decreasing BMI was associated with presenting with FIGO stage III or IV disease at the time of surgery. Discussion Endometrial cancer was one of the first cancers recognized as being related to obesity [18]. Overall the cohort we studied was obese, with a median BMI of 33.5. As reported previously, obesity was associated with earlier stage disease [19,20]. BMI, however, was not significantly different with respect to either histologic type or endometrioid FIGO grade in our study. Additionally, in this cohort, BMI was significantly different between Caucasians and African Americans, with African Americans being more obese. Body mass index was inversely correlated to age at the time of surgery. This relationship was noted in two recent studies examining obesity and other clinical/ pathologic features of endometrial cancer [20,21]. BMI is undoubtedly the most accessible and easily comparable measurement of adiposity across studies and populations. Indeed it is the best suited for this retrospective analysis. However, studies have shown that BMI may not be an optimal indicator of body fat composition in the elderly, and may underestimate body fat percentage [22]. The association between increasing age and decreasing BMI may be due to several factors. One reason may involve a so-called cohort effect, which represents a generational difference between the prevalence of obesity. The observed difference may also be due to premature death of obese women, and thus selective survival of those less obese. Additionally, declining BMI may simply result from weight loss that occurs as women age [23–25]. The relationship between microsatellite instability and obesity has previously been explored in colon cancer. Slattery et al. found that a subset of women with a high BMI were more likely to have tumors which were MSI negative. In their study, they also examined other surrogate markers of estrogen exposure, namely pregnancy, hormone use, and total ovulatory months [26]. All showed significant differences between MSI positive and negative colon cancers, with estrogen exposure associated with MSI negative colon tumors. The relationship between obesity and MSI, however, was not observed in men. The authors concluded that estrogens prevented MSI positive tumors, whether the source of estrogen was endogenous or exogenous. We also found an association between obesity and microsatellite instability, in which higher BMI correlated with MSI negative endometrial tumors. Earlier work suggested that estrogen has a protective effect in the proximal colon, possibly by inhibiting the MSI phenotype, and thus colon tumorigenesis [27–29]. Our data indicating an apparent inverse correlation between estrogen levels and MSI lend support to the theory that

estrogen might protect against MSI-positive cancers. Alternatively, it may be that estrogen driven tumorigenesis is distinctly different from tumorigenesis driven by loss of DNA mismatch repair. As a mitogen, estrogen's effect in the uterus on cellular proliferation and cancer promotion may result in the accumulation of genetic abnormalities, while loss of mismatch repair may confer growth advantages as well as increased mutation rates. As seen in previously published data from the same cohort, MSI tumors were found more frequently in endometrioid tumors [30]. Approximately 75% of MSI positive endometrioid tumors displayed hypermethylation of the MLH1 promoter, indicating that this is a more common mechanism of tumorigenesis than in Type II endometrial cancers. Additionally, MSI was significantly different between FIGO grading of endometrioid tumors, with 46% of Grade 2 tumors showing MSI, compared to 26% and 24% of Grade 1 and Grade 3 tumors, respectively. Age did not differ significantly based on microsatellite instability (64.9 vs. 63.4, p = 0.63). Maxwell et al. also found no relationship between age and MSI in a group of 131 patients with endometrioid endometrial cancer, 29 of which displayed MSI [13]. Conversely, in a study by Yehuda et al., participants were examined for the development of MSI by comparing microsatellite markers in peripheral blood lymphocytes at two different time points spanning 10 years. MSI was not detected at any of the eight markers in the 23 young (aged 25–35) individuals, while 17 of 38 older (aged 65–85) individuals developed MSI at 3 or more loci (p = 0.001) [31]. The results of this study, however, have not been replicated. Some investigators have additionally explored the potential role of the CpG island methylator phenotype as a causative factor in the development of MSI [32]. In sporadic endometrial cancer, as well as sporadic colorectal cancer, epigenetic inactivation of the MLH1 gene via promoter hypermethylation appears to be the most frequent mismatch repair aberration in MSI tumors [8,9]. In the subset of women with MSI positive tumors in our cohort, age was significantly different based on MLH1 promoter methylation, with methylation occurring more frequently in older women (57.3 vs. 67 years). Yiu et al. also determined that MLH1 promoter methylation increased with age among MSI positive tumors in a cohort of 230 consecutive patients treated for primary colorectal cancer [33]. The literature related to age and MSI, therefore, appears contradictory. Perhaps some of the discrepancy is due to a lack of standardization of the loci tested when defining microsatellite instability. All of the studies mentioned thus far have used a variety of microsatellite markers, with very little overlap. In the present study, we used markers designated by the NCI Workshop on Microsatellite Instability for cancer detection in colorectal cancer [16], as they have been studied previously, and help correlate with other data. The literature regarding this issue in colorectal cancer does, however, show an apparent association between increasing age and MSI positive tumors [26,29,33]. In addition to microsatellite instability, we also examined MLH1 promoter methylation, the mechanism most commonly associated with the MSI phenotype in sporadic endometrial

C.K. McCourt et al. / Gynecologic Oncology 104 (2007) 535–539

cancers. There was no correlation seen between BMI and the presence of MLH1 promoter methylation in our endometrial cancer population. In summary, these data support the hypothesis that estrogens likely prevent the MSI phenotype, and continued research into the possible relationship between sex steroid hormonal factors and the mismatch repair system is warranted.

[16]

[17]

References [18] [1] International Agency for Research on Cancer. IARC handbooks of cancer prevention. Weight control and physical activity. Lyon: International Agency for Research on Cancer; 2002. [2] Kaaks R, Lukanova A, Kurzer MA. Obesity, endogenous hormones, and endometrial cancer risk: a synthetic review. Cancer Epidemiol, Biomarkers Prev 2002;11:1531–43. [3] Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity among US adults, 1999–2000. J Am Med Assoc 2002;288: 1723–7. [4] Parazzini F, LaVecchia C, Bocciolone L, Franceschi S. The epidemiology of endometrial cancer. Gynecol Oncol 1991;41:1–16. [5] Lax SF. Molecular genetic pathways in various types of endometrial carcinoma: from a phenotypical to a molecular-based classification. Virchows Arch 2004;444:213–23. [6] Goodfellow P, Buttin B, Herzog T, Rader J, Gibb R, Swisher E, et al. Prevalence of defective DNA mismatch repair and MSH6 mutation in an unselected series of endometrial cancers. Proc Natl Acad Sci 2003;100 (10):5908–13. [7] Harfe BD, Jinks-Robertson S. DNA mismatch repair and genetic instability. Annu Rev Genet 2000;34:359–99. [8] Simpkins S, Bocker T, Swisher E, Mutch D, Gersell D, Kovatich A, et al. MLH1 promoter methylation and gene silencing is the primary cause of microsatellite instability in sporadic endometrial cancers. Hum Mol Genet 1999;8(4):661–6. [9] Esteller M, Levine R, Baylin SB, Ellenson LH, Herman JG. MLH1 promoter hypermethylation is associated with the microsatellite instability phenotype in sporadic endometrial carcinomas. Oncogene 1998; 17:2413–7. [10] Basil J, Goodfellow P, Rader J, Mutch D, Herzog T. Clinical significance of microsatellite instability in endometrial carcinoma. Cancer 2000;89 (8):1758–64. [11] Hirasawa A, Aoki D, Inoue J, Imoto I, Susumu N, et al. Unfavorable prognostic factors associated with high frequency of microsatellite instability and comparative genomic hybridization analysis in endometrial cancer. Clin Cancer Res 2003;9:5675–82. [12] Lax S, Tashiro H, Slebos R, Ellenson L. The frequency of p53, K-ras mutations, and microsatellite instability differs in uterine endometrioid and serous carcinoma. Cancer 2000;88(4):814–24. [13] Maxwell LG, Risinger JI, Alvarez AA, Barrett CJ, Berchuck A. Favorable survival associated with microsatellite instability in endometrioid endometrial cancers. Obstet Gynecol 2001;97:417–22. [14] Peiffer SL, Herzog TJ, Tribune DJ, Mutch DG, Gersell DJ, Goodfellow PJ. Allelic loss of sequences from the long arm of chromosome 10 and replication errors in endometrial cancers. Cancer Res 1995;55(9):1922–6. [15] Kowalski LD, Mutch DG, Herzog TJ, Rader JS, Goodfellow PJ. Mutational analysis of MLH1 and MSH2 in 25 prospectively acquired

[19]

[20]

[21]

[22]

[23]

[24]

[25] [26]

[27] [28]

[29]

[30]

[31]

[32] [33]

539

RER+ endometrial cancers. Genes Chromosomes Cancer 1997;18: 219–27. Boland CR, Thibodeau SN, Hamilton SR, Sidransky D, Eshleman JR, Burt RW, et al. A National Cancer Institute Workshop on microsatellite instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 1998;58(22):5248–57. Whelan A, Babb S, Mutch D, Rader J, Herzog T, Todd C, 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. Calle EE, Kaaks R. Overweight, obesity and cancer: epidemiological evidence and proposed mechanisms. Nat Rev, Cancer 2004;4:579–91. Anderson B, Conner JP, Andrews JI, Davis CS, Buller RE, Sorosky JI, et al. Obesity and prognosis in endometrial cancer. Am J Obstet Gynecol 1996;174:1171–8. Everett E, Tamimi H, Greer B, Swisher E, Paley P, Mandel L, et al. The effect of body mass index on clinical/pathologic features, surgical morbidity, and outcome in patients with endometrial cancer. Gynecol Oncol 2003;90:150–7. Pavelka JC, Ben-Shachar I, Fowler JM, Ramirez NC, Copeland LJ, Eaton LA, et al. Morbid obesity and endometrial cancer: surgical, clinical, and pathologic outcomes in surgically managed patients. Gynecol Oncol 2004;95:588–92. Visser M, van den Heuvel E, Deurenberg P. Prediction equations for the estimation of body composition in the elderly using anthropometric data. Br J Nutr 1994;71:823–33. Seidell JC, Visscher TLS. Body weight and weight change and their health implications for the elderly. Eur J Clin Nutr 2000;54(Suppl 3): S33–9. Lehmann AB, Bassey EJ. Longitudinal weight changes over four years and associated factors in 629 men and women aged over 65. Eur J Clin Nutr 1996;50:6–11. Elia M. Obesity in the elderly. Obes Res 2001;9(Suppl 4):244S–8S. Slattery ML, Potter JD, Curtin K, Edwards S, Ma KN, Anderson K, et al. Estrogens reduce and withdrawal of estrogens increase risk of microsatellite instability-positive colon cancer. Cancer Res 2001;61:126–30. Potter JD. Hormones and colon cancer. J Natl Cancer Inst 1995; 87:1039–40. Broeders MJ, Lambe M, Baron JA, Leon DA. History of childbearing and colorectal cancer risk in women aged less than 60: an analysis of Swedish routine registry data 1960–1984. Int J Cancer 1996;66:170–5. Breivik J, Lothe RA, Meling GI, Rognum TO, Borresen-Dale AL, Gaudernack G. Different genetic pathways to proximal and distal colorectal cancer influenced by sex-related factors. Int J Cancer 1997;74:664–9. Goodfellow PJ, Buttin BM, Herzog TJ, Rader JS, Gibb RK, Swisher E, et al. Prevalence of defective DNA mismatch repair and MSH6 mutation in an unselected series of endometrial cancers. Proc Natl Acad Sci U S A 2003;100:5908–13. Yehuda AB, Globerson A, Krichevsky S, Bar On H, Kidron M, Friedlander Y, et al. Ageing and the mismatch repair system. Mech Ageing Dev 2000;121:173–9. Toyota M, Issa JP. CpG island methylator phenotypes in aging and cancer. Semin Cancer Biol 1999;9:349–57. Yiu R, Qiu H, Lee SH, Garcia-Aguilar J. Mechanisms of microsatellite instability in colorectal cancer patients in different age groups. Dis Colon Rectum 2005;48:2061–9.