Home Duke University Press
QUICK SEARCH:   [advanced]


     
  Home | Help | Feedback | Subscriptions | Archive | Search | Table of Contents

First published on June 18, 2008
This version was published on August 1, 2008
Neuro Oncol 2008 10(4):553-559; DOI:10.1215/15228517-2008-026
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
10/4/553    most recent
15228517-2008-026v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lönn, S.
Right arrow Articles by Inskip, P. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Duke University Press

Basic and Translational Investigations

Genetic variation in insulin-like growth factors and brain tumor risk

Stefan Lönn, Nathaniel Rothman, William R. Shapiro, Howard A. Fine, Robert G. Selker, Peter M. Black, Jay S. Loeffler, Amy A. Hutchinson and Peter D. Inskip

Division of Cancer Epidemiology and Genetics (S.L., N.R., P.D.I.) and Neuro-oncology Branch (H.A.F.), National Cancer Institute, Bethesda, MD; Barrow Neurological Institute, St. Joseph's Hospital and Medical Center, Phoenix, AZ (W.R.S.); Western Pennsylvania Hospital, Pittsburgh, PA (R.G.S.); Brigham and Women's Hospital, Boston, MA (P.M.B.); Department of Radiation Oncology, Massachusetts General Hospital, Boston, MA (J.S.L.); Division of Cancer Epidemiology and Genetics, Core Genotyping Facility, Advanced Technology Center, SAIC-Frederick, Inc., National Cancer Institute-Frederick, Frederick, MD (A.A.H.); USA

Address all correspondence to Stefan Lönn, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Room 7053, 6120 Executive Blvd., Bethesda, MD 20892-7238, USA (Stefan.Lonn{at}ki.se).


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Many studies support a role for insulin-like growth factors (IGFs) in the regulation of tumor cell biology. We hypothesized that single-nucleotide polymorphisms (SNPs) in IGF genes are risk factors for glioma and meningioma. To test the hypothesis, we examined associations of brain tumor risk with nine variants in five IGF genes in a hospital-based case-control study. The study was conducted at hospitals in Boston, Phoenix, and Pittsburgh between 1994 and 1998. Eligible cases were individuals (18 years or older) newly diagnosed with glioma or meningioma. Controls were selected among patients who were admitted to the same hospitals for a variety of nonmalignant conditions and frequency matched to cases by hospital, age, sex, race, and distance from residence. The present analysis was restricted to non-Hispanic whites. DNA was extracted from blood samples collected from 354 glioma cases, 133 meningioma cases, and 495 control individuals. We evaluated nine SNPs in five IGF genes (IGF1, IGF1R, IGF2, IGF2R, and IGFBP3). The majority of the analyzed IGF SNPs did not display statistically significant associations with glioma or meningioma. For glioma, one IGF1R SNP (rs2272037) indicated a possible association. No indications of association were seen for glioblastoma, but for low-grade gliomas, the odds ratio under a dominant model was 0.56 (95% confidence interval [CI], 0.35-0.90) for IGF1 rs6220, 2.98 (95% CI, 1.65-5.38) for IGF1R rs2272037, and 1.60 (95% CI, 0.90-2.83) for IGF1R rs2016347. Overall, our results do not provide strong evidence of associations of brain tumor risk with IGF polymorphic variants but identify several associations for glioma that warrant further examination in other, larger studies.

Key Words: central nervous system • glioma • insulin-like growth factor • meningioma • single nucleotide polymorphism


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
The insulin-like growth factor (IGF) system comprises two ligands (IGF-1 and IGF-2), the IGF-1 and IGF-2 receptors, six binding proteins (IGFBP-1 to -6), and various IGFBP-related peptides.1 IGF-1 is the major physiological mediator of the effect of growth hormone and therefore has a strong influence on cell proliferation and differentiation. It also inhibits apoptosis by blocking initiation of the apoptotic pathway.1 The IGF-1 receptor (IGF-1R) mediates the action of IGF-1 and is involved in oncogenic transformation processes.1 IGFBPs modulate the interaction between IGF-1 and IGF-1R but also have independent effects on cell growth.1-3 IGFBP-3 has an inhibitory effect on IGF-1 activity and also acts as an apoptotic agent.1 IGFBP-3 also has been recognized to exhibit a number of growth-promoting effects.

Experimental studies have shown that alterations in IGF function can influence cellular transformation and tumor cell proliferation. IGF1, IGF2, and IGF1R genes have all been reported to be overexpressed in glioma and meningioma as well as in a wide range of other human cancers, including breast, leukemia, lung, thyroid, and prostate.4 IGFs, together with their receptors and binding proteins, have been reported to be associated with cancer risk.5,6 Epidemiological studies have suggested that genetic variation in IGF1, IGF1R, and IGFBP3 may be related to breast, prostate, and colorectal cancer risk.7-10 In vitro studies have demonstrated that IGF receptors and binding proteins promote mitogenesis and differentiation in glial cells, oligodendrocytes, neuronal cells, adult stem cells, and brain explants and regulate axon myelination.4 Furthermore, observations in the literature suggest that IGF gene pathways show similar expression and functional features during fetal development and tumorigenesis.11 There is, however, little epidemiologic data concerning the possible involvement of IGF signaling in the development of brain tumors in humans. A recent small prospective study indicated an inverse association between glioma and IGF-1 serum levels.12 We hypothesized that polymorphisms in IGF genes are risk factors for glioma and meningioma. To test the hypothesis, we examined associations with several IGF gene variants in the context of a case-control study.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
The methods for this case-control study have been described in detail previously.13 In brief, the study was conducted at hospitals in Boston, Phoenix, and Pittsburgh between 1994 and 1998. Each of the three hospitals is a referral center for the diagnosis and treatment of brain tumors. The study was restricted to adults (18 years or older) who received care at one of the participating hospitals, resided within 50 miles of the hospital (or within Arizona, in the case of the Phoenix center), and could understand English or Spanish. Institutional review boards at the National Cancer Institute and all participating hospitals approved the protocol, and written informed consent was obtained from each participant.

Cases and Controls
Eligible cases were defined as patients with a primary intracranial glioma or meningioma during the study period. All cases had to be diagnosed with a microscopically confirmed tumor within the 8 weeks preceding hospitalization at a participating hospital; most (80%) were enrolled within 3 weeks of initial diagnosis. In total, 88% of the glioma cases (n = 489) and 98% of the meningioma cases (n = 197) participated in the study. DNA extracted from blood samples was available for 388 glioma cases and 162 meningioma cases, of which 354 and 133, respectively, were non-Hispanic whites. We restricted the present analysis to non-Hispanic whites.

The controls were patients who were admitted to the same hospitals as the cases for a variety of nonmalignant conditions. The most common reasons for hospitalization among the controls were injuries (25%) and disorders of the circulatory (22%), musculoskeletal (22%), digestive (12%), and nervous (7%) systems. They were frequency matched to the total group of patients with tumors (including acoustic neuroma) according to hospital, age (in 10-year strata), sex, race or ethnic group, and proximity of their residence to the hospital. Of the eligible controls, 86% (n = 799) participated. DNA extracted from blood samples was available for 553 controls, of which 495 were non-Hispanic whites.

All participating cases and controls were interviewed by trained nurses. The structured, computerized, in-person interview included detailed questions related to medical and reproductive history, including exposure to diagnostic and therapeutic radiation, and various environmental risk factors, including occupational exposures, cellular telephone use, and sociodemographic characteristics.

Selection of Polymorphisms and Laboratory Analyses
Single-nucleotide polymorphisms (SNPs) in IGF genes were selected initially based on the allele frequency, potential functional importance as indicated by a non-synonymous amino acid change, occurrence in an exon or promoter region, or associations with other cancers in the literature;6-8 however, several intronic SNPs also were included as potential markers. Nine SNPs in five IGF genes were evaluated (Table 1).


View this table:
[in this window]
[in a new window]

  		Table 1. Basic information for single-nucleotide polymorphisms analyzed in the study

 

DNA was extracted from peripheral white blood cells from blood samples by GenoType, Ltd. (United Kingdom) using a phenol-chloroform method as described by Daly et al.14 Genotyping was conducted by the Core Genotyping Facility at the National Cancer Institute (Gaithersburg, MD, USA), using TaqMan (Applied Biosystems, Foster City, CA, USA) methods. Descriptions for assay-specific methods can be found at the National Cancer Institute SNP500Cancer Web site (http://snp500cancer.nci.nih.gov).

Quality-control measures included 75 study duplicates (two samples for each individual, all of whom were study subjects) interspersed throughout the batches for all assays and in 68 samples from three individuals who were not study subjects (processed in identical fashion as samples from study subjects). In addition, laboratory assay-specific positive controls for the three possible genotypes and one DNA-negative control were included on each assay plate.

Statistical Analysis
Allele frequencies in SNPs among controls were assessed for deviation from Hardy-Weinberg equilibrium (HWE). Associations between SNPs and risk of brain tumors were assessed using unconditional logistic regression to estimate odds ratios (ORs) and calculate associated 95% likelihood-based confidence intervals (CIs). All SNPs were analyzed under a dominant model, but a codominant relationship was assumed when numbers permitted (homozygous variant frequency >1% among the controls). The analyses were restricted to non-Hispanic whites and adjusted for the matching variables (hospital, age, sex, and proximity of their residence to the hospital).

Stratified analyses were performed by sex and age (two groups: <50 years and ≥50 years). Glioblastomas and low-grade gliomas were analyzed separately. The tumor grade of gliomas was classified according to the guidelines of Kleihues et al.15 There were 171 glioblastoma cases (48% of all gliomas) and 98 low-grade gliomas (28% of all gliomas). The low-grade glioma group included 34 oligodendrogliomas, 29 astrocytomas, 14 neuronal-glial tumors, 12 mixed gliomas, and 9 other low-grade gliomas.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Genotyping was successful on an average of 89% of all variants of the total collected blood samples analyzed. Missing values, primarily the result of insufficient quantity or concentration of high-quality DNA, or poor amplification for a specific locus, were equally likely to be from case or control samples. We achieved 98%-100% agreement among replicates and duplicate samples for all assays (Table 1). Two of the analyzed SNPs showed significant deviations from HWE (Table 1). The IGF1 rs6220 SNP had more than expected heterozygotes, and the IGF1R rs2272037 SNP had fewer than expected heterozygotes. The concordance in the quality-control measures for the two SNPs with significant departure from HWE was high (Table 1).

Frequencies of characteristics of brain tumor cases and controls are presented in Table 2. Meningioma cases were more often female compared with controls or glioma cases. The glioma and meningioma cases tended to be older than the controls.


View this table:
[in this window]
[in a new window]

  		Table 2. Distribution of cases and controls with respect to selected characteristics

 

The risks of glioma and meningioma associated with IGF polymorphic variants are presented in Table 3. The majority of the analyzed IGF genes did not display statistically significant associations with glioma or meningioma. For glioma, only one SNP (IGF1R gene rs2272037) indicated an association for both heterozygous and homozygous carriers (p for trend = 0.04); however, the OR was greater for heterozygotes than for homozygous variants. The OR under the dominant model was 1.58 (95% CI, 1.15-2.15). Under the dominant model, the IGF1 (rs6220) variant was significantly associated with glioma risk (OR 0.74; 95% CI, 0.56-0.98). Meningioma was not strongly associated with any of the genotypes examined.


View this table:
[in this window]
[in a new window]

  		Table 3. Risk of brain tumors in relation to insulin-like growth factor (IGF) polymorphic variants

 

Table 4 displays the results for gliomas separately for glioblastoma and low-grade glioma. No statistically significant associations between glioblastoma and the analyzed SNPs were detected, but indications of associations were seen between low-grade glioma and one IGF1 SNP (rs6220) and two IGF1R SNPs (rs2272037 and rs2016347). The OR under a dominant model was 0.56 (95% CI, 0.35-0.90) for rs6220, 2.98 (95% CI, 1.65-5.38) for rs2272037, and 1.60 (95% CI, 0.90-2.83) for rs2016347. The rs2272037 was the only SNP that displayed a statistically significant trend (p = 0.03); however, the OR was greater for the heterozygous carriers than for the homozygous carriers.


View this table:
[in this window]
[in a new window]

  		Table 4. Insulin-like growth factor (IGF) polymorphic variants and risk of glioblastoma (GBM) and low-grade glioma (LGG)

 

For several SNPs, the gender-specific analysis yielded a stronger association among men compared to women. Although data were sparse in the gender-specific analysis, the positive association observed for IGF1R (rs2272037) with low-grade glioma was stronger among men with OR 5.35 (95% CI, 1.97-14.57) for heterozygous carriers and OR 3.09 (95% CI, 0.95-10.00) for homozygous carriers compared to women with OR 2.51 (95% CI, 1.09-5.80) and OR 1.21 (95% CI, 0.39-3.70), respectively. Indication of a possible stronger association among men compared to the combined analysis was also present for IGF1 (rs2162679) and another IGF1R gene (rs2137680). Stratifying the analysis by age (<50 years, ≥50 years) did not indicate heterogeneity of risk for glioblastoma, low-grade glioma, or meningioma (results not shown).

The estimated ORs were similar in the crude and adjusted analysis, indicating that the matching variables had limited influence on the results. The results did not change materially when the analysis included all racial or ethnic groups. Sequentially excluding subgroups of controls based on reasons for hospitalization did not change any overall results.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Several environmental factors have been suggested to increase the risk of brain tumors,16,17 but few have been studied with strong or consistent evidence of causality. Variation in IGF function should be considered as a possible candidate in brain tumor etiology. To our knowledge, no previous epidemiologic study has investigated genetic variation in the IGF pathway in relation to brain tumor risk, and our results therefore cannot be compared directly with other studies. A recent small prospective epidemiologic study indicated an inverse association between glioma and serum levels of IGF-1,12 and experimental data support the possibility that IGFs are related to glioma development and progression.4 Our investigation did not indicate an association between meningioma and IGF polymorphic variants. For glioma, no association was seen between glioblastoma and IGF polymorphic variants, but a possible association was detected for low-grade glioma. The associations were mainly seen for the IGF1R gene. IGF-1R binds IGFs with a high affinity and plays a critical role in transformation events.18 It is highly overexpressed in most malignancies, where it functions as an antiapoptotic agent by enhancing cell survival.

This is the first study exploring the hypothesis that alterations in IGF pathways are risk factors for brain tumors, and the study has several notable strengths. The results are based on one of the largest brain tumor case-control studies with DNA. Cases were identified continuously during the study period through collaboration with the treating clinics, and a rapid recruitment of cases was therefore possible. The rapid ascertainment is essential in a study of brain cancers because of the severity of the disease and the relatively short survival time. The participation rate was high, and the collection of blood samples very soon after brain tumor diagnosis minimizes the influence of a survival bias associated with IGF genotypes.

The study has several limitations as well, and there is reason for caution in interpreting the results. Two of the analyzed SNPs showed significant departure from HWE, and these included SNPs with non-null associations. It has been reported that HWE-violating SNPs more often show significant associations than SNPs without HWE violation.19 There are several reasons why HWE may be violated, including genotyping error, chance, and population structure. In the present study, the quality-control data indicate high reproducibility of results for the two SNPs with HWE-violation. It is not likely that the HWE violation is a chance finding, but we cannot exclude the possibility. If we assume HWE for controls in the two SNPs according to the strategy presented by Chen et al.,20 the OR shifts toward unity but still indicates an association between IGF and low-grade glioma. In addition, discrepant HWE results do not mean that postulated associations should be dismissed, but they should hint at the need for caution in interpretation and more evidence and replication. We evaluated nine SNPs in four tumor groups or subgroups, and the only significant associations were only marginally significant, so they may well be due to chance; replication is clearly needed. The selected SNPs in our study did not fully cover the IGF pathway and additional SNPs should be analyzed, including more IGF genes, for example, IGFBP2 and IGFBP5. Selection bias could be a source of spurious associations in a hospital-based case-control study if one or more of the gene variants evaluated is associated with one or more of the diseases constituting the control series; however, sequential removal of each major control group based on reason for hospitalization did not materially change the results.

In conclusion, we report a possible association between IGF polymorphic variants and the risk of low-grade glioma. Our results are not robust, and the association between IGF polymorphisms and brain tumors needs to be considered further in large, well-designed studies with more comprehensive coverage of the IGF genes.


   Acknowledgments
 
This project has been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health, under contract N01-CO-12400. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. government.

Received for publication November 2, 2007. Accepted for publication March 3, 2008.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  • Jones JI, Clemmons DR. Insulin-like growth factors and their binding proteins: biological actions. Endocr Rev. 1995;16: 3-34.[Abstract/Free Full Text]

  • Khandwala HM, McCutcheon IE, Flyvbjerg A, Friend KE. The effects of insulin-like growth factors on tumorigenesis and neoplastic growth. Endocr Rev. 2000;21: 215-244.[Abstract/Free Full Text]

  • Sepp-Lorenzino L. Structure and function of the insulin-like growth factor I receptor. Breast Cancer Res Treat. 1998;47: 235-253.[CrossRef][Web of Science][Medline]

  • Russo VC, Gluckman PD, Feldman EL, Werther GA. The insulin-like growth factor system and its pleiotropic functions in brain. Endocr Rev. 2005;26: 916-943.[Abstract/Free Full Text]

  • Renehan AG, Zwahlen M, Minder C, O'Dwyer St, Shalet SM, Egger M. Insulin-like growth factor (IGF)-I, IGF binding protein-3, and cancer risk: systematic review and meta-regression analysis. Lancet. 2004;363: 1346-1353.[CrossRef][Web of Science][Medline]

  • Furstenberger G, Senn HJ. Insulin-like growth factors and cancer. Lancet Oncol. 2002;3: 298-302.[CrossRef][Web of Science][Medline]

  • Al-Zahrani A, Sandhu MS, Luben RN, et al. IGF1 and IGFBP3 tagging polymorphisms are associated with circulating levels of IGF1, IGFBP3 and risk of breast cancer. Hum Mol Genet. 2006;15: 1-10.[Abstract/Free Full Text]

  • Cheng I, Stram DO, Penney KL, et al. Common genetic variation in IGF1 and prostate cancer risk in the Multiethnic Cohort. J Natl Cancer Inst. 2006;98: 123-134.[Abstract/Free Full Text]

  • Wong HL, Delellis K, Probst-Hensch N, 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: 144-151.[Abstract/Free Full Text]

  • Chen C, Freeman R, Voigt LF, Fitzpatrick A, Plymate SR, Weiss NS. Prostate cancer risk in relation to selected genetic polymorphisms in insulin-like growth factor-I, insulin-like growth factor binding protein-3, and insulin-like growth factor-I receptor. Cancer Epidemiol Biomarkers Prev. 2006;15: 2461-24616.[Abstract/Free Full Text]

  • Wang H, Fuller GN, Zhang W. Insulin-like growth factors and insulin-like growth factors binding proteins in CNS tumors. In: Zhang W, Fuller GN, eds. Genomic and Molecular Neuro-oncology. Sudbury, MA: Jones and Bartlett Publishers; 2004: 119-130.

  • Lönn S, Inskip PD, Pollak MN, Weinstein SJ, Virtamo J, Albanes D. Glioma risk in relation to serum level of insulin-like growth factors. Cancer Epidemiol Biomarkers Prev. 2007;16: 844-846.[Abstract/Free Full Text]

  • Inskip PD, Hatch EE, Stewart PA, et al. Study design for a case-control investigation of cellular telephones and other risk factors for brain tumors in adults. Radiat Prot Dosim. 1999;86: 45-52.[Abstract]

  • Daly AK, Steen VM, Fairbrother KS, Idle JR. CYP2D6 multiallelism. Methods Enzymol. 1996;272: 199-210.[Web of Science][Medline]

  • Kleihues P, Cavenee WK, ed. Pathology and Genetics of Tumours of the Nervous System. Lyon, France: International Agency for Research on Cancer; 2000.

  • Inskip PD, Linet MS, Heineman EF. Etiology of brain tumors in adults. Epidemiol Rev. 1995;17: 382-414.[Free Full Text]

  • Ohgaki H, Kleihues P. Epidemiology and etiology of gliomas. Acta Neuropathol. 2005;109: 93-108.[CrossRef][Medline]

  • Singleton JR, Randolph AE, Feldman EL. Insulin-like growth factor I receptor prevents apoptosis and enhances neuroblastoma tumorigenesis. Cancer Res. 1996;56: 4522-4529.[Abstract/Free Full Text]

  • Trikalinos tA, Salanti G, Khoury MJ, Ioannidis JP. Impact of violations and deviations in Hardy-Weinberg equilibrium on postulated gene-disease associations. Am J Epidemiol. 2006;163: 300-309.[Abstract/Free Full Text]

  • Chen J, Chatterjee N. Exploiting Hardy-Weinberg equilibrium for efficient screening of single SNP associations from case-control studies. Hum Hered. 2007;63: 196-204.[CrossRef][Web of Science][Medline]


    Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



    This Article
    Right arrow Abstract Freely available
    Right arrow Full Text (PDF)
    Right arrow All Versions of this Article:
    10/4/553    most recent
    15228517-2008-026v1
    Right arrow Alert me when this article is cited
    Right arrow Alert me if a correction is posted
    Right arrow Citation Map
    Services
    Right arrow Similar articles in this journal
    Right arrow Similar articles in PubMed
    Right arrow Alert me to new issues of the journal
    Right arrow Download to citation manager
    Right arrow reprints & permissions
    Citing Articles
    Right arrow Citing Articles via Web of Science (1)
    Right arrow Citing Articles via Google Scholar
    Google Scholar
    Right arrow Articles by Lönn, S.
    Right arrow Articles by Inskip, P. D.
    Right arrow Search for Related Content
    PubMed
    Right arrow PubMed Citation
    Social Bookmarking
    Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
    What's this?

      Home | Help | Feedback | Subscriptions | Archive | Search | Table of Contents