| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© The Author 2006. Published by Oxford University Press.
BRIEF COMMUNICATION |
IGF1 Gene Polymorphism and Risk for Hereditary Nonpolyposis Colorectal Cancer
Affiliations of authors: Department of Epidemiology (MZ, CIA, XG, IMC, JSJ, MLF), Department of GI Medicine and Nutrition (PML), Department of Surgical Oncology (MAR-B), University of Texas MD Anderson Cancer Center, Houston, TX
Correspondence to: Marsha L. Frazier, PhD, Department of Epidemiology, University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd., Box 189, Houston, TX 77030 (e-mail: mlfrazier{at}mdanderson.org).
 |
ABSTRACT |
|---|
 TopNotes Abstract References |
|---|
Hereditary nonpolyposis colorectal cancer (HNPCC) is an autosomal dominant disorder caused by germline mutations in DNA mismatch repair (MMR) genes. Insulin-like growth factor-I (IGF-I) is involved in colorectal carcinogenesis, and elevated plasma IGF-I levels are associated with sporadic colorectal cancer (CRC) risk. We investigated the relationship between IGF1 promoter cytosine-adenine (CA) dinucleotide–repeat polymorphism length and CRC risk in 121 MMR gene mutation carriers using Cox regression and Kaplan-Meier analysis. All statistical tests were two-sided. Time to onset for CRC increased for each decrease in CA-repeat number (median = 19 repeats, range = 12–22 repeats; hazard ratio [HR] = 1.17, 95% confidence interval [CI] = 1.05 to 1.31; P = .006). Patients carrying a CA
17 repeat allele had a statistically significantly higher CRC risk (HR = 2.36; 95% CI = 1.28 to 4.36; P = .006) than all others and were younger at onset (44 years versus 56.5 years; P = .023). These findings indicate a statistically significant association between shorter IGF1 CA-repeat lengths and increased risk for CRC in HNPCC. This is the first report, to our knowledge, to show that IGF1 variant genotypes modify risk of a hereditary form of cancer.
Hereditary nonpolyposis colorectal cancer (HNPCC) is an autosomal dominant disorder caused by DNA mismatch repair (MMR) gene mutations with hMLH1 and hMSH2 being the most frequently mutated (1). The syndrome is characterized by an increased risk of a variety of cancers, with colorectal cancer (CRC) and endometrial cancer being the most common (2). The incidence of HNPCC is one in 1000 in the general population and one in 100 in individuals with CRC (3).
The age of CRC onset varies considerably among MMR gene mutation carriers, which may be partially explained by a combination of genetic and environmental factors. For example, a previous study from our group (4) showed that adverse polymorphic genotypes of the cyclin D1 gene influence age-associated risk for CRC in HNPCC. The cyclin D1 gene guanine (G) to adenine (A) polymorphism at codon 242 (nucleotide 870) in exon 4 enhances alternate splicing, increases the levels, and prolongs the half-life of the alternative cyclin D1 transcript (4). MMR mutation carriers with one or two copies of the cyclin D1 A allele develop CRC 11 years earlier, on average, than MMR gene carriers with the GG phenotype (4).
Insulin-like growth factor (IGF)-I is a polypeptide that has previously been associated with sporadic CRC. Numerous in vitro and animal studies of CRC (5–8) have implicated IGF-I in cell transformation, tumor growth, metastasis, and poor prognosis. In addition, epidemiologic studies (5,6) have indicated that high plasma IGF-I plays a role in energy balance (energy intake versus expenditure), which has also been shown to influence risk for CRC. IGF-I also activates and induces nuclear accumulation of cyclin D1, a required step for initiation of the cell cycle (9,10).
Although IGF-I plasma levels fluctuate due to age and physiologic and behavior factors, such as nutrition, exercise, and smoking, some studies (11,12) also implicate a genetic component in the interindividual variability in IGF1 gene expression. A gene polymorphism that consists of numerous cytosine-adenine (CA) dinucleotide repeats is located in the 5' untranslated promoter region of the IGF1 gene (1 kilobase from the transcriptional initiation site). This polymorphism is thought to alter promoter activity and, thus, influence the transcription rate of IGF1 (12). CA repeats in a promoter have also been reported to influence transcription in the growth-related genes acetyl-coenzyme carboxylase and egfr, in which the length of the repeat is inversely associated with gene transactivation and, therefore, protein levels (13,14). To date, no in vitro studies have evaluated whether modifying the length of the CA repeats alters IGF1 gene expression.
Studies on the association between IGF1 CA repeat length and serum IGF-I levels have been inconsistent (12,15–29). The inconsistency may be due to the strategy used for these genotype–phenotype studies; most are based on a comparison of serum IGF-I levels between homozygous carriers of the CA 19 polymorphism versus other genotypes. The fact that this analytic strategy is common may also explain why the influence of the length of the IGF1 CA-repeat polymorphism on cancer risk has been reported inconsistently, perhaps reflecting differences in allele frequencies greater than or less than 19 repeats among the different populations studied.
We investigated the relationship between IGF1 CA-repeat polymorphism length and age-associated risk for HNPCC, because it is not yet known whether IGF-I has a role in this hereditary form of CRC. The study included 121 individuals (from 59 families) with confirmed germline hMLH1 or hMSH2 gene mutations from the University of Texas M.D. Anderson Cancer Center HNPCC registry (30). Written informed consent for this study was obtained from all participants (4). The study was approved by the M.D. Anderson Cancer Center Internal Review Board. More than half of the participants (Table 1) were females (53.7%), MSH2 mutation carriers (53.7%), and Caucasian (85%). The mean age of CRC onset was 43.9 years (range = 23–84 years). Individuals with CRC (45%) were older than non-CRC individuals (mean age = 57.6 years versus 46.9 years), more likely to be deceased (34.5% versus 3%), and more often the proband in a family (71% versus 12%).
|
The IGF1 dinucleotide genotyping was performed using the forward 5'-GCTAGCCAGCTGGTGTTATT-3' primer and the reverse 5'-ACCACTCTGGGAGAAGGGTA-3' primer, flanking the polymorphic (CA)n repeat upstream of the transcription start site (11). Polymerase chain reaction (PCR) and gel electrophoresis were performed as previously described (31). The cyclin D1 gene G to A polymorphism at codon 242 (nucleotide 870) in exon 4 was assayed, and the IGF-I CA-repeat number was determined by nucleotide sequencing of selected PCR products, as described previously (4). The frequencies of the CA-repeat allele numbers (with percentages of the total of 242 alleles) were as follows: 12 (0.8%), 14 (0.8%), 15 (0.5%), 17 (3%), 18 (12%), 19 (58%), 20 (20%), 21 (4.1%), and 22 (0.8%).
For statistical analysis, the age of onset of colorectal cancer was defined as the patient's age at diagnosis or, for the unaffected MMR mutation carriers, the age at last follow-up. The CRC age of onset was used as the outcome, whereas the CA-repeat number in the IGF-1 gene, cyclin D1 genotype (homozygous wild-type GG versus heterozygous AG and homozygous mutant AA), sex, and ethnicity were independent variables. The individual's IGF1 (CA)n repeat was modeled at the genotype level (N = 121). Kaplan-Meier survival analysis was used to assess differences in time to onset of CRC among carriers with varying CA-repeat numbers, the log-rank test (which gives equal weight to all failures) was used to evaluate the homogeneity of the survival curves, and Cox proportional hazard regression model analysis was used to estimate the association between CRC risk and CA-repeat number. To assess proportional hazards assumptions, we checked whether or not the Schoenfeld residuals were significantly correlated with age. No departures from proportional hazards assumptions were observed for any statistically significantly associated markers. To allow for possible correlation of time to cancer onset for individuals of the same family (due to unmeasured variables, such as shared household, environment, dietary exposures, and same gene mutations) we used a robust variance correction available in Stata (the cluster function). This action adjusted the variance of the Cox proportional hazards homogeneity tests by accounting for correlation in time to onset for CRC within families. Analyses were performed using Intercool Stata 8.0 (Stata Corp., College Station, TX).
Because the most commonly reported CA-repeat length is 19, we initially created the following genotype categories for comparison: CA19/CA19 versus CAno19/CAno19, CA19/CAany versus CAno19/CAno19, CA19/CA19 versus CA19/CAno19, CA19/CAno19 versus CAno19/CAno19, CA19/CA19 versus CA19/<19, CA19/CA<19 versus CA<19/CA<19, CA19/CA19 versus CA<19/CA<19, CA19/CA19 versus CA19/CA>19, CA19/CA>19 versus CA>19/CA>19, and CA19/CA19 versus CA>19/CA>19. Statistically nonsignificant results were obtained in all the above analyses that stratified by CA19 alleles (Supplementary Table 1; available at http://jncicancerspectrum.oxfordjournals.org/jnci/content/vol98/issue3).
Because the shorter CA-repeat length is thought to lead to higher levels of IGF1 expression (12), we considered the allele with the smallest IGF1 CA-repeat number for each individual as the dominant allele. The genotype frequencies are provided (Supplementary Table 2; available at http://jncicancerspectrum.oxfordjournals.org/jnci/content/vol98/issue3). When the smallest allele from each individual was modeled as a continuous variable using the Cox model, a statistically significant association between CA-repeat length and time to CRC onset was observed (for every decrease by one CA-repeat length, hazard ratio [HR] = 1.17, 95% CI = 1.05 to 1.31; P = .006). Adjusting for cyclin D1 and ethnicity gave similar results (HR = 1.18, 95% CI = 1.03 to 1.33; P = .02). For example, those with a 14-CA repeat have a 1.186 or 2.70 higher risk of CRC than those with a 20-CA repeat.
We also examined the distribution of the IGF1 genotypes. None of the patients had more than one CA17 allele and those with one CA17 allele had greater risk than those with genotypes not having a CA 17 allele (CA
18), making genotypes with a CA17 versus CA18 a logical cut point. Using the CA 18 genotype as the referent group, those with a CA17 allele had a statistically significantly higher CRC risk before (HR = 2.36, 95% CI = 1.28 to 4.36; P = .006) and after adjustment for cyclin D1 (HR = 3.15, 95% CI = 1.68 to 5.91; P<.001). Kaplan-Meier survival analysis showed a 12.5-year difference in the median age of onset between the CA17 and CA18 groups, with the median ages being 44 and 56.5 years, respectively. The log-rank test (P = .023) demonstrated a statistically significant difference between the curves (Fig. 1).
|
An intriguing result was obtained when we evaluated the associations between the IGF1 CA polymorphism length and cyclin D1 adverse genotypes after stratifying by MMR gene mutation (Table 2). For the subset of hMLH1 carriers, no association between CRC risk and IGF1 CA-repeat length was observed; however, the association between CRC risk and cyclin D1 gene mutation was statistically significant (HR = 2.68, 95% CI = 1.06 to 6.77; P = .037). Among hMSH2 carriers, a statistically significant association between CRC risk and IGF1 CA-repeat length was observed (HR = 4.22, 95% CI =1.48 to 12.03; P = .007). However, there was no significant association between CRC risk and cyclin D1 (HR =1.67, 95% CI = 0.67 to 4.14; P = .27). These results indicate a combined association of cyclin D1 and IGF1 CA
17 genotypes on risk for CRC that varies by MMR mutation type. The risk associated with cyclin D1 was most pronounced among hMLH1 carriers, whereas the association with IGF1 was stronger in hMSH2 carriers. A test for interaction between cyclin D1 and IGF1 was not statistically significant, indicating that larger studies are needed to evaluate the combined effects of cyclin D1 and IGF1. Consistent with our findings, a previous in vitro study demonstrated that following DNA damage, the resulting cell cycle arrest occurs if the MMR genes can induce the degradation of cyclin D1 (32).
|
Our findings indicate a statistically significant association between shorter IGF1 CA-repeat lengths and increased risk for CRC in MMR gene mutation carriers. The findings are consistent with the trends observed in the reports of sporadic CRC and the IGF1 gene polymorphism (26,29) but not with those examining the association of the IGF1 polymorphism with breast and prostate cancers (15,16,22,24,25,27,28). This discrepancy is most likely caused by different carcinogenic mechanisms; breast and prostate cancers are hormonally regulated malignancies and IGF1 gene expression in those instances may also be regulated by estrogen and/or androgen. In addition, an obvious difference between our study population and others is that all of our subjects had a major underlying genetic defect in a MMR gene, which may have made them more susceptible to the modifying effects of the IGF1 polymorphism. Another possibility is that the discrepancies among studies may also lie in the study design and analysis performed. In survival analysis, each subject (affected or unaffected) can provide more information about the impact of the genetic factors on time-to-onset than would be afforded by the binary outcome of the case–control design (33). Furthermore, most of the other studies classified the CA-repeat length polymorphism by the presence (19/19) or absence (non-19/non-19) of the most common allele, perhaps reflecting differences in allele frequencies greater than or less than 19 repeats among populations (13–15,20–27).
It is important to point out that although the shorter IGF1 CA repeat is associated with increased risk for CRC, another possibility is that the increased risk is actually due to a polymorphism in another gene or a locus within the IGF1 gene that is in linkage disequilibrium with the CA repeat. We examined the region surrounding the IGF1 gene and found that the closest upstream gene was phenylalanine hydroxylase (357 779 nucleotides away) and downstream was pro-melanin–concentrating hormone (198 075 nucleotides away). Running Ensembl with the HapMap database shows no linkage disequilibrium between single nucleotide polymorphisms (SNPs) in the IGF1 gene and neighboring genes. However, linkage disequilibrium does exist within the IGF1 gene and therefore our findings do not rule out the possibility that the association of IGF1 promoter allelic variants is due to a SNP or other locus within the IGF1 gene.
Our study had several potential limitations. It relied on a small sample size, with Caucasians as the predominant ethnic group. However, the association between repeat length and CRC risk was statistically significant. Referral bias could have influenced our results: more severe and younger case patients are often referred to our institution because M.D. Anderson Cancer Center is a tertiary care center. Because individuals are not referred on the basis of their genotypes, referral bias would not falsely indicate an association but could change the age at onset distribution, because CRC patients at our site are often younger than those in the general population.
Future in vitro and epidemiologic studies would provide insights about the genotype–phenotype correlation between the IGF1 CA-repeat numbers and the corresponding serum IGF-I levels in MMR mutation carriers. These future studies, although important, have their own sets of problems: IGF-I serum measurements usually rely on one-time point measurement done mostly on individuals who already have cancer. In these patients, changes in IGF-I levels may be a result of the disease rather than a precursor to it. This area is understudied, and further investigations are needed to understand the relationship between IGF1 genotype and IGF-I expression.
In summary, this report provides the first evidence, to our knowledge, for a role of the IGF1 gene in modifying the risk of a hereditary form of cancer. The association between IGF1 and CRC risk may be more pronounced in this study population because the individuals have a high predisposition to develop CRC as a result of their underlying MMR gene mutations. The IGF1 gene may thus be an important target for chemopreventive strategies for HNPCC.
|
NOTES |
|---|
|
TopNotesAbstractReferences |
|---|
M. Zecevic's current address is The Lancet, 360 Park Ave. S., New York, NY.
Supported by National Cancer Institute grant CA70759 (to M.L. Frazier), National Institutes of Health Cancer Center support grant CA 16672 (to Dr. John Mendelsohn, PI), and National Cancer Institute Cancer Education and Career Development Program R25 CA 57730 (to Dr. Robert M. Chamberlain, PI).
We thank Robert M. Chamberlain, PhD, and Sara Strom, PhD, for their insightful discussions and editorial assistance.
|
REFERENCES |
|---|
|
TopNotesAbstractReferences |
|---|
(1) De la Chapelle A. Genetic predisposition to colorectal cancer. Nat Rev Cancer 2003;4:769–80.
(2) DeFrancisco J, Grady WM. Diagnosis and management of hereditary non-polyposis colon cancer. Gastrointest Endosc 2003;58:390–408.[CrossRef][Medline]
(3) Umar A, Risinger JI, Hawk ET, Barrett JC. Testing guidelines for hereditary non-polyposis colorectal cancer. Nat Rev Cancer 2004;4:153–8.[ISI][Medline]
(4) Kong S, Amos CI, Luthra R, Lynch PM, Levin B, Frazier ML. Effects of cyclin D1 polymorphism on age of onset of hereditary nonpolyposis colorectal cancer. Cancer Res 2000;60:249–52.
(5) Giovannucci E. Insulin, insulin-like growth factors and colon cancer: a review of the evidence. J Nutr 2001;131 11 Suppl:3109S–20S.
(6) Kaaks R. Nutrition, energy balance and colon cancer risk: the role of insulin and insulin-like growth factor-I. IARC Sci Publ 2002;156:289–93.[Medline]
(7) Pollak MN, Schernhammer ES, Hankinson SE. Insulin-like growth factors and neoplasia. Nat Rev Cancer 2004;4:505–518.[CrossRef][ISI][Medline]
(8) Sandhu MS, Dunger DB, Giovannucci EL. Insulin, insulin-like growth factor-I (IGF-I), IGF binding proteins, their biologic interactions, and colorectal cancer. J Natl Cancer Inst 2002;94:972–80.
(9) Hamelers IH, van Schaik RF, Sipkema J, Sussenbach JS, Steenbergh PH. Insulin-like growth factor I triggers nuclear accumulation of cyclin D1 in MCF-7S breast cancer cells. J Biol Chem 2002;277:47645–52.
(10) Frederick TJ, Wood TL. IGF-I and FGF-2 coordinately enhance cyclin D1 and cyclin E-cdk2 association and activity to promote G1 progression in oligodendrocyte progenitor cells. Mol Cell Neurosci 2004;25:480–92.[CrossRef][ISI][Medline]
(11) Rotwein P, Pollock KM, Didier DK, Krivi GG. Organization and sequence of the human insulin-like growth factor I gene. J Biol Chem 1986;261:4828–32.
(12) Rosen CJ, Kurland ES, Vereault D, Adler RA, Rackoff PJ, Craig WY, Witte S, Rogers J, Bilezikian JP. Association between serum insulin growth factor-I (IGF-I) and a simple sequence repeat in IGF-I gene: implications for genetic studies of bone mineral density. J Clin Endocrinol Metab 1998;83:2286–90.
(13) Tae HJ, Luo X, Kim KH. Roles of CCAAT/enhancer-binding protein and its binding site on repression and derepression of acetyl-CoA carboxylase gene. J Biol Chem 1994;269:10475–84.
(14) Gebhardt F, Zanker KS, Brandt B. Modulation of epidermal growth factor receptor gene transcription by a polymorphic dinucleotide repeat in intron 1. J Biol Chem 1999;274:13176–80.
(15) Yu H, Li BD, Smith M, Shi R, Berkel HJ, Kato I. Polymorphic CA repeats in the IGF-I gene and breast cancer. Breast Cancer Res Treat 2001;70:117–22.[CrossRef][ISI][Medline]
(16) Missmer SA, Haiman CA, Hunter DJ, Willett WC, Colditz GA, Speizer FE, et al. A sequence repeat in the insulin-like growth factor-1 gene and risk of breast cancer. Int J Cancer 2002;100:332–6.[CrossRef][ISI][Medline]
(17) Jernstrom H, Chu W, Vesprini D, Tao Y, Majeed N, Deal C, et al. Genetic factors related to racial variation in plasma levels of insulin-like growth factor-1: implications for premenopausal breast cancer risk. Mol Genet Metab 2001;72:144–54.[CrossRef][ISI][Medline]
(18) Jernstrom H, Deal C, Wilkin F, Chu W, Tao Y, Majeed N, et al. Genetic and nongenetic factors associated with variation of plasma levels of insulin-like growth factor-I and insulin-like growth factor-binding protein-3 in healthy premenopausal women. Cancer Epidemiol Biomarkers Prev 2001;10:377–84.
(19) Allen NE, Davey GK, Key TJ, Zhang, S, Narod SA. Serum insulin-like growth factor I (IGF-I) concentration in men is not associated with the cytosine-adenosine repeat polymorphism of the IGF-I gene. Cancer Epidemiol Biomarkers Prev 2002;11:319–20.
(20) Kato I, Eastham J, Li B, Smith M, Yu H. Genotype-phenotype analysis for the polymorphic CA repeat in the insulin-like growth factor-I (IGF-I) gene. Eur J Epidemiol 2003;18:203–9.[Medline]
(21) Rietveld I, Janssen JA, Hofman A, Pols HA, van Duijn CM, Lamberts SW. A polymorphism in the IGF-I gene influences the age-related decline in circulating total IGF-I levels. Eur J Endocrinol 2003;148:171–5.[Abstract]
(22) Figer A, Karasik YP, Baruch RG, Chetrit A, Papa MZ, Sade RB, Rizel S, Friedman E, Riezel S. Insulin-like growth factor I polymorphism and breast cancer risk in Jewish women. Isr Med Assoc J 2002;4:759–62.[ISI][Medline]
(23) Giovannucci E, Haiman CA, Platz EA, Hankinson SE, Pollak MN, Hunter DJ. Dinucleotide repeat in the insulin-like growth factor-I gene is not related to risk of colorectal adenoma. Cancer Epidemiol Biomarkers Prev 2002;11:1509–10.
(24) DeLellis K, Ingles S, Kolonel L, McKean-Cowdin R, Henderson B, Stanczyk F, Probst-Hensch NM. IGF1 genotype, mean plasma level and breast cancer risk in the Hawaii/Los Angeles multiethnic cohort. Br J Cancer 2003;88:277–82.[CrossRef][ISI][Medline]
(25) Li L, Cicek MS, Casey G, Witte JS. No association between genetic polymorphisms in IGF-I and IGFBP-3 and prostate cancer. Cancer Epidemiol Biomarkers Prev 2004;13:497–8.
(26) Slattery ML, Samowitz W, Curtin K, Ma KN, Hoffman M, Caan B, Neuhausen S. Associations among IRS1, IRS2, IGF1, and IGFBP3 genetic polymorphisms and colorectal cancer. Cancer Epidemiol Biomarkers Prev 2004;13:1206–14.
(27) Wen W, Gao YT, Shu XO, Yu H, Cai Q, Smith JR, Zheng W. Insulin-like growth factor-I gene polymorphism and breast cancer risk in Chinese women. Int J Cancer 2005;113:307–311.[CrossRef][ISI][Medline]
(28) Tsuchiya N, Wang L, Horikawa Y, Inoue T, Kakinuma H, Matsuura S, et al. CA-repeat polymorphism in the insulin-like growth factor-I gene is associated with increased risk of prostate cancer and benign prostatic hyperplasia. Int J Oncol 2005;26:225–31.[ISI][Medline]
(29) Wong HL, Delellis K, Probst-Hensch N, Koh WP, Van Den Berg D, Lee HP, et al. A new single nucleotide polymorphism in the insulin-like growth factor I regulatory region associates with colorectal cancer risk in Singapore Chinese. Cancer Epidemiol Biomarkers Prev 2005;14:144–51.
(30) Frazier ML, O'Donnell FT, Kong S, Gu X, Campos I, Luthra R, et al. Age-associated risk of cancer among individuals with N-acetyltransferase 2 (NAT2) mutations and mutations in DNA mismatch repair genes. Cancer Res 2001;61:1269–71.
(31) Frazier ML, Sinicrope FA, Amos CI, Cleary KR, Lynch PM, Levin B, et al. Loci for efficient detection of microsatellite instability in hereditary nonpolyposis colorectal cancer. Oncol Rep 1999;6:497–505.[ISI][Medline]
(32) Lan Z, Sever-Chroneos Z, Strobeck MW, Park CH, Baskaran R, Edelmann W, et al. DNA damage invokes mismatch repair-dependent cyclin D1 attenuation and retinoblastoma signaling pathways to inhibit CDK2. J Biol Chem 2002;277:8372–81.
(33) Wijsman EM, Amos CI. Genetic analysis of simulated oligogenic traits in nuclear and extended pedigrees: Summary of GAW10 contributions. Genet Epidemiol 1997;14:719–35.[CrossRef][ISI][Medline]
Manuscript received June 3, 2005; revised October 17, 2005; accepted October 21, 2005.
Correspondence about this Article
CiteULike 
Connotea 
Del.icio.us What's this?
J Natl Cancer Inst 2006 98: 1664-1665.
This article has been cited by other articles:
|
|
 |
|
  P. D. Ryan and P. E. Goss The Emerging Role of the Insulin-Like Growth Factor Pathway as a Therapeutic Target in Cancer Oncologist, January 1, 2008; 13(1): 16 - 24. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. T. Campbell, L. Edwards, J. R. McLaughlin, J. Green, H. B. Younghusband, and M. O. Woods Cytochrome P450 17A1 and Catechol O-Methyltransferase Polymorphisms and Age at Lynch Syndrome Colon Cancer Onset in Newfoundland Clin. Cancer Res., July 1, 2007; 13(13): 3783 - 3788. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Reeves, C. Meldrum, and R. J. Scott Re: IGF-1 Gene Polymorphism and Risk for Hereditary Nonpolyposis Colorectal Cancer. J Natl Cancer Inst, November 15, 2006; 98(22): 1664 - 1665. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||