Despite the potential importance of the cell cycle and apoptosis pathways in brain tumor etiology, little has been published regarding brain tumor risk associated with common gene variants in these pathways. Using data from a hospital-based case-control study conducted by the National Cancer Institute between 1994 and 1998, we evaluated risk of glioma (n = 388), meningioma (n = 162), and acoustic neuroma (n = 73) with respect to 12 single nucleotide polymorphisms from 10 genes involved in apoptosis and cell cycle control: CASP8, CCND1, CCNH, CDKN1A, CDKN2A, CHEK1, CHEK2, MDM2, PTEN, and TP53. We observed significantly decreased risk of meningioma with the CASP8 Ex14-271A>T variant [odds ratio (OR)AT, 0.8; 95% confidence interval (95% CI), 0.5-1.2; ORAA, 0.5; 95% CI, 0.3-0.9; Ptrend = 0.03] and increased risk of meningioma with the CASP8 Ex13+51G>C variant (ORGC, 1.4; 95% CI, 0.9-2.1; ORCC, 3.6; 95% CI, 1.0-13.1; Ptrend = 0.04). The CT haplotype of the two CASP8 polymorphisms was associated with significantly increased risk of meningioma (OR, 1.7; 95% CI, 1.1-2.6), but was not associated with risk of glioma or acoustic neuroma. The CCND1 Ex4-1G>A variant was associated with increased risk for glioma, and the Ex8+49T>C variant of CCNH was associated with increased risk of glioma and acoustic neuroma. The MDM2 Ex12+162A>G variant was associated with significantly reduced risk of glioma. Our results suggest that common variants in the CASP8, CCND1, CCNH, and MDM2 genes may influence brain tumor risk. Future research in this area should include more detailed coverage of genes in the apoptosis/cell cycle control pathways. (Cancer Epidemiol Biomarkers Prev 2007;16(8):1655–61)

The processes of tumor initiation, growth, and proliferation are thought to be largely influenced by genes that regulate vital cellular functions (1). In normal cells, DNA damage is recognized by cellular mechanisms, and responses to prevent the propagation of errors include nucleotide damage repair, activation of checkpoints to arrest the cell cycle, and/or cell apoptosis (2). Hereditary defects in genes that regulate the crucial pathways of cell cycle control and apoptosis have been associated with a number of human malignancies, including tumors of the brain and nervous system (1). The importance of apoptosis and cell cycle control in brain tumor etiology has also been suggested by the frequent presence of mutations or loss of expression of apoptosis/cell cycle control genes such as PTEN, CDKN2A, MDM2, TP53, NF1, and RB in brain tumors (3).

Despite the potential importance of the cell cycle and apoptosis pathways in brain tumor etiology, very little has been published regarding brain tumor risk associated with more common gene variants in these pathways, with the exception of the TP53 gene, for which some evidence of differences in glioma and meningioma risk with common polymorphisms and haplotypes has been noted (4-7), and the CCND gene, for which no association with meningioma was observed (8). Using data from a hospital-based case-control study conducted by the National Cancer Institute (NCI) between 1994 and 1998, we evaluated risk of three brain tumor types (glioma, meningioma, and acoustic neuroma) with respect to 12 single nucleotide polymorphisms (SNP) from 10 genes involved in apoptosis and cell cycle control: CASP8 (rs13113, rs1045485), CCND1 (rs678653, rs603965), CCNH (rs2266690), CDKN1A (rs1801270), CDKN2A (rs3731249), CHEK1 (rs506504), CHEK2 (rs2267130), MDM2 (rs769412), PTEN (rs701848), and TP53(rs1042522). These polymorphisms were selected for their common occurrence in the population, and for potential functional relevance signaled by nonsynonymous amino acid changes or occurrence in exonic or promoter regions of the gene (Table 1).

Table 1.

Apoptosis and cell cycle genes and SNPs evaluated in the NCI adult brain tumor study

Gene symbolGene nameLocationSNP 500 AA or nt variant IDRS numberBasis of selection
Epidemiologic associationReferences
Prevalence*/functional considerations
CASP8 Caspase-8, apoptosis-related cysteine peptidase 2q33-q34 Ex13+51G>C (D285H) rs1045485 >0.1 prevalence Breast cancer (20) 
     Amino acid change   
CASP8   Ex14-271A>T rs13113 >0.1 prevalence NHL (19) 
     Exonic region   
     Selected for haplotype   
CCND1 Cyclin D1 11q13 Ex4-1G>A rs603965 >0.1 prevalence Bladder cancer, colorectal cancer, head and neck cancer, lung cancer, leukemia, NHL (24, 25) 
     Affects protein splicing   
CCND1   Ex5+852C>G rs678653 >0.1 prevalence   
     Exonic region   
     Selected for haplotype   
     >0.1 prevalence   
CCNH Cyclin H 5q13.3-q14 Ex8+49T>C rs2266690 Amino acid change   
   V270A  Gene involved in cell-cycle control   
CDKN1A Cyclin-dependent kinase inhibitor 1A 6p21.2 Ex2+98C>A rs1801270 >0.1 prevalence  (37) 
   S31R  Amino acid change   
     Gene alterations in brain tumors   
CDKN2A Cyclin-dependent kinase inhibitor 2A 9p21 Ex3-16G>A rs3731249 Amino acid change   
   A148T  Gene alternations in brain tumors   
CHEK1 CHK1 checkpoint homologue 11q24-q24 Ex13+76A>G rs506504 Exonic region   
   I471V  Gene involved in cell-cycle control   
CHEK2 CHK2 checkpoint homologue 22q12.1 IVS9-200G>A rs2267130 >0.1 prevalence   
     Gene involved in cell-cycle control   
MDM2 Mdm2, transformed 3T3 cell double minute 2, p53 binding protein 12q14.3-q15 Ex12+162A>G rs769412 Exonic region  (37) 
     Negative regulator of TP53   
     Gene alterations in brain tumors   
PTEN Phosphatase and tensin homologue 10q23.3 1515 bp 3′ of STP C>T rs701848 >0.1 prevalence  (37) 
     3′ region often contains regulatory elements important for transcription and translation   
     Gene alterations in brain tumors   
        
TP53 Tumor protein p53 17p13.1 Ex4+119C>G rs1042522 >0.1 prevalence  (37) 
   P72R  Amino acid change   
     Gene alterations in brain tumors   
Gene symbolGene nameLocationSNP 500 AA or nt variant IDRS numberBasis of selection
Epidemiologic associationReferences
Prevalence*/functional considerations
CASP8 Caspase-8, apoptosis-related cysteine peptidase 2q33-q34 Ex13+51G>C (D285H) rs1045485 >0.1 prevalence Breast cancer (20) 
     Amino acid change   
CASP8   Ex14-271A>T rs13113 >0.1 prevalence NHL (19) 
     Exonic region   
     Selected for haplotype   
CCND1 Cyclin D1 11q13 Ex4-1G>A rs603965 >0.1 prevalence Bladder cancer, colorectal cancer, head and neck cancer, lung cancer, leukemia, NHL (24, 25) 
     Affects protein splicing   
CCND1   Ex5+852C>G rs678653 >0.1 prevalence   
     Exonic region   
     Selected for haplotype   
     >0.1 prevalence   
CCNH Cyclin H 5q13.3-q14 Ex8+49T>C rs2266690 Amino acid change   
   V270A  Gene involved in cell-cycle control   
CDKN1A Cyclin-dependent kinase inhibitor 1A 6p21.2 Ex2+98C>A rs1801270 >0.1 prevalence  (37) 
   S31R  Amino acid change   
     Gene alterations in brain tumors   
CDKN2A Cyclin-dependent kinase inhibitor 2A 9p21 Ex3-16G>A rs3731249 Amino acid change   
   A148T  Gene alternations in brain tumors   
CHEK1 CHK1 checkpoint homologue 11q24-q24 Ex13+76A>G rs506504 Exonic region   
   I471V  Gene involved in cell-cycle control   
CHEK2 CHK2 checkpoint homologue 22q12.1 IVS9-200G>A rs2267130 >0.1 prevalence   
     Gene involved in cell-cycle control   
MDM2 Mdm2, transformed 3T3 cell double minute 2, p53 binding protein 12q14.3-q15 Ex12+162A>G rs769412 Exonic region  (37) 
     Negative regulator of TP53   
     Gene alterations in brain tumors   
PTEN Phosphatase and tensin homologue 10q23.3 1515 bp 3′ of STP C>T rs701848 >0.1 prevalence  (37) 
     3′ region often contains regulatory elements important for transcription and translation   
     Gene alterations in brain tumors   
        
TP53 Tumor protein p53 17p13.1 Ex4+119C>G rs1042522 >0.1 prevalence  (37) 
   P72R  Amino acid change   
     Gene alterations in brain tumors   

Abbreviation: NHL, non–Hodgkin's lymphoma.

*

Based on SNP500.

Study Setting and Population

A detailed description of study methods can be found elsewhere (9). Briefly, subjects for a case-control study of brain tumors were enrolled between 1994 and 1998 from three hospitals located in Phoenix, Arizona; Boston, Massachusetts; and Pittsburgh, Pennsylvania. Each hospital specializes in the treatment of brain tumors and serves as a regional referral center for the diagnosis. The study protocol was approved by the institutional review board of each participating institution, and written informed consent was obtained from each patient or proxy.

Eligible patients were 18 years or older with a first intracranial glioma, meningioma (ICD-O-2 codes 9530-9538), or acoustic neuroma (ICD-O-2 codes 9560) diagnosed at the hospital or during the 8 weeks preceding hospitalization. Ninety-two percent of eligible brain tumor patients agreed to participate, and 489 patients with glioma, 197 with meningioma, and 96 patients with acoustic neuroma were enrolled. All diagnoses of glioma and meningioma and 96% of acoustic neuromas were confirmed by microscopy at the hospital of diagnosis.

Study controls were patients admitted to the same hospitals as cases for a variety of nonneoplastic conditions, including injuries (25%), circulatory system disorders (22%), musculoskeletal disorders (22%), and digestive disorders (12%). Controls were frequency matched in a 1:1 ratio to all brain tumor patients based on age in years (18-29, 30-39, 40-49, 50-59, 60-69, 70-79, and 80-99 years); race/ethnicity (non-Hispanic White, Hispanic, African-American, other); sex (male, female); hospital (Phoenix, Boston, Pittsburgh); and residential proximity to the hospital in miles (0-4, 5-14, 15-29, 30-49, 50 or more). 799 control patients, who represented 86% of all contacted controls, were enrolled.

Blood samples were collected and sent for genotyping for 388 patients with glioma (79%), 162 patients with meningioma (82%), 73 patients with acoustic neuroma (76%), and 553 controls (69%). The main obstacle to obtaining blood samples was subject refusal, with non-participation in the blood draw being higher for controls (24%) than for cases (14%).

Processing of Blood Samples and Genotyping

DNA was extracted from blood samples using a phenol-chloroform method (10), and genotyping was conducted by the NCI Core Genotyping Facility using a medium-throughput TaqMan assay. Sequence data and assay conditions are provided online9

(11).

Quality control (QC) specimens for the study included 15 to 34 samples from each of three individuals (A, B, C) who were not study participants (QC-A, n = 34; QC-B, n = 20; QC-C, n = 15) and duplicate samples from 89 study subjects. These specimens were collected and processed in a manner similar to study samples and were interspersed among all genotyping assays in a masked fashion. Percentage agreement between the three nonstudy replicates ranged from 97.8% to 100% for all SNPs. Concordance for duplicates was lowest for CHEK2 IVS9-200G>A (91.8%), but ranged from 97.2% to 100% for the remaining SNPs, with 100% concordance for nine of the 12 SNPs. In addition to QC specimens, each plate of 368 specimens included homozygous wild-type, heterozygous, and homozygous variant positive controls and one DNA negative control.

Statistical Analyses

Statistically significant departure from Hardy-Weinberg equilibrium for controls was assessed using the χ2 test. For each polymorphism, unconditional logistic regression was used to calculate odds ratios (OR) and 95% confidence intervals (95% CI) for each tumor type, adjusted for the study matching factors of age, sex, race/ethnicity, hospital, and residential proximity to the hospital. Because controls were frequency matched to all tumor types, all controls were used in the models for each tumor type. Models were run under the assumption of codominant inheritance (AA versus Aa versus aa) and dominant inheritance (AA versus Aa or aa). A score test of linear trend was conducted for each SNP using a three-level ordinal variable. To account for the fact some of our results could be false-positive findings due to chance, we calculated the probability that our findings were false positives using the false discovery rate, which reflects the expected ratio of false-positive findings to the number of significant findings (12). We applied the false discovery rate method to the P values for trend as this allows for the fewest number of comparisons and thus degrees of freedom to assess the additive model of risk. For evaluating the false discovery rate, all sets of Ptrend (glioma, meningioma, and acoustic neuroma) were evaluated simultaneously.

All analyses were run separately by gender and were repeated in non-Hispanic whites (89% of the study population). Given that the protein MDM2 is a strong down-regulator of TP53, we also assessed potential interaction between polymorphisms in these two genes.

Haplotypes formed by the two CASP8 polymorphisms (D′ = 1.0, r2 = 0.12) and the two CCND1 polymorphisms (D′ = 0.89, r2 = 0.34) were analyzed within non-Hispanic whites. The HaploView program, version 3.32, was used to estimate haplotype block structure in controls. Haplotype frequencies were estimated using the estimation-maximization algorithm (13), and overall differences between cases and controls (adjusted for age and sex) were assessed using the global score test in HaploStat (R version 2.2.0). The effects of individual haplotypes were estimated using unconditional logistic regression, assuming an additive model and using posterior probabilities of the haplotypes as weights to update the regression coefficients in an iterative manner (14).

Blood samples were available for 1,176 of the 1,581 participants in the brain case-control study. The genotyping success rate for the 12 SNPs ranged from 93.9% to 99.1%. Analysis of all SNPs in the control population did not detect deviation from Hardy-Weinberg equilibrium. Individuals with less education, and those in the oldest age group (70-90 years), were less likely to have samples sent for genotyping (Table 2). Otherwise, the distribution of demographic characteristics for genotyped subjects was similar to the distribution for all study subjects. Relative to controls, subjects with glioma were proportionately more often male, whereas subjects with meningioma and acoustic neuroma were more often female. A lower proportion of subjects with meningioma and acoustic neuroma were in the youngest age bracket (<18 years) compared with controls.

Table 2.

Demographic characteristics in participants with and without genotyping: NCI adult brain tumor study, 1994 to 1998

CharacteristicGlioma* (n = 388)Meningioma* (n = 162)Acoustic neuroma* (n = 73)Controls* (n = 553)Total, sent for genotyping (n = 1,176)Total, no. blood samples (n = 405)
Sex       
    Male 211 (54.4) 37 (22.8) 26 (35.6) 254 (45.9) 528 (44.9) 194 (47.9) 
    Female 177 (45.6) 125 (77.2) 47 (64.4) 299 (54.1) 648 (55.1) 211 (52.1) 
Race/ethnicity       
    White, non-Hispanic 354 (91.2) 133 (82.1) 66 (90.4) 495 (89.5) 1,048 (89.1) 363 (89.6) 
    Hispanic 20 (5.2) 12 (7.4) 6 (8.2) 39 (7.1) 77 (6.6) 23 (5.7) 
    Black 7 (1.8) 9 (5.6) 0 (0.0) 11 (2.0) 27 (2.3) 11 (2.7) 
    Other 7 (1.8) 8 (4.9) 1 (1.4) 8 (1.5) 24 (2.0) 8 (2.0) 
Age at interview (y)       
    18-29 46 (11.9) 2 (1.2) 3 (4.1) 75 (13.6) 126 (10.7) 40 (9.9) 
    30-49 143 (36.9) 60 (37.0) 31 (42.5) 225 (40.7) 459 (39.0) 139 (34.2) 
    50-69 134 (34.5) 73 (45.1) 30 (41.1) 182 (32.9) 419 (35.6) 148 (36.5) 
    70-90 65 (16.8) 27 (16.7) 9 (12.3) 71 (12.8) 172 (14.6) 78 (19.3) 
Educational level       
    <High school 38 (10.1) 21 (13.0) 4 (5.6) 72 (13.3) 135 (11.5) 63 (15.6) 
    High school graduate 204 (54.4) 102 (63.4) 33 (45.8) 334 (61.9) 673 (57.2) 232 (57.3) 
    College or professional 133 (35.5) 38 (23.6) 35 (48.6) 134 (24.8) 340 (28.9) 99 (24.4) 
    Missing 13 13 28 11 
CharacteristicGlioma* (n = 388)Meningioma* (n = 162)Acoustic neuroma* (n = 73)Controls* (n = 553)Total, sent for genotyping (n = 1,176)Total, no. blood samples (n = 405)
Sex       
    Male 211 (54.4) 37 (22.8) 26 (35.6) 254 (45.9) 528 (44.9) 194 (47.9) 
    Female 177 (45.6) 125 (77.2) 47 (64.4) 299 (54.1) 648 (55.1) 211 (52.1) 
Race/ethnicity       
    White, non-Hispanic 354 (91.2) 133 (82.1) 66 (90.4) 495 (89.5) 1,048 (89.1) 363 (89.6) 
    Hispanic 20 (5.2) 12 (7.4) 6 (8.2) 39 (7.1) 77 (6.6) 23 (5.7) 
    Black 7 (1.8) 9 (5.6) 0 (0.0) 11 (2.0) 27 (2.3) 11 (2.7) 
    Other 7 (1.8) 8 (4.9) 1 (1.4) 8 (1.5) 24 (2.0) 8 (2.0) 
Age at interview (y)       
    18-29 46 (11.9) 2 (1.2) 3 (4.1) 75 (13.6) 126 (10.7) 40 (9.9) 
    30-49 143 (36.9) 60 (37.0) 31 (42.5) 225 (40.7) 459 (39.0) 139 (34.2) 
    50-69 134 (34.5) 73 (45.1) 30 (41.1) 182 (32.9) 419 (35.6) 148 (36.5) 
    70-90 65 (16.8) 27 (16.7) 9 (12.3) 71 (12.8) 172 (14.6) 78 (19.3) 
Educational level       
    <High school 38 (10.1) 21 (13.0) 4 (5.6) 72 (13.3) 135 (11.5) 63 (15.6) 
    High school graduate 204 (54.4) 102 (63.4) 33 (45.8) 334 (61.9) 673 (57.2) 232 (57.3) 
    College or professional 133 (35.5) 38 (23.6) 35 (48.6) 134 (24.8) 340 (28.9) 99 (24.4) 
    Missing 13 13 28 11 
*

Characteristics for individuals sent for genotyping.

Percent based on nonmissing values.

Table 3 summarizes the main effects of the 12 candidate SNPs involved in cell cycle control and apoptosis, for each brain tumor type. We observed no significant association between genotype and risk of any brain tumor (glioma, meningioma, or acoustic neuroma) for the following SNPs: CCND1 Ex5+852C>G (rs678653), CDKN1A Ex2+98C>A (rs1801270), CDKN2A Ex3-16G>A (rs3731249), CHEK1 Ex13+76A>G (rs506504), CHEK2 IVS9-200G>A (rs2267130), PTEN 1,515 bp 3′of STP C>T (rs701848), or TP53 Ex4+119C>G (rs1042522).

Table 3.

ORs for variants of apoptosis/cell cycle–related SNPs in the NCI adult brain tumor study, 1994 to 1998 (models adjusted for study matching factors age, sex, race/ethnicity, study site, and residential proximity to treatment hospital)

Gene polymorphismChromosomal locationGenotypeControls (n = 553)
Glioma (n = 388)
Meningioma (n = 162)
Acoustic neuroma (n = 73)
nnOR (95% CI)nOR (95% CI)nOR (95% CI)
CASP8 (rs13113) 2q33-q34 TT 155 129 1.0 (reference) 60 1.0 (reference) 22 1.0 (reference) 
  AT 258 166 0.8 (0.6-1.1) 78 0.8 (0.5-1.2) 30 0.8 (0.5-1.6) 
  AA 98 75 0.9 (0.6-1.3) 18 0.5 (0.3-0.9)* 17 1.1 (0.6-2.3) 
     Ptrend = 0.4  Ptrend = 0.03*  Ptrend = 0.8 
  AT or AA   0.8 (0.6-1.1)  0.7 (0.5-1.1)  0.9 (0.5-1.6) 
          
CASP8 (rs1045485) 2q33-q34 GG 426 284 1.0 (reference) 117 1.0 (reference) 50 1.0 (reference) 
  GC 118 95 1.2 (0.8-1.6) 38 1.4 (0.9-2.1) 20 1.5 (0.8-2.6) 
  CC 0.7 (0.1-2.8) 3.6 (1.0-13.1)* 5.6 (1.2-26.6)* 
     Ptrend = 0.6  Ptrend = 0.04*  Ptrend = 0.04* 
  GC or CC   1.1 (0.8-1.5)  1.5 (1.0-2.3)*  1.6 (0.9-2.8) 
          
CCND1 (rs678653) 11q13 GG 198 138 1.0 (reference) 72 1.0 (reference) 24 1.0 (reference) 
  GC 243 188 1.1 (0.8-1.5) 55 0.6 (0.4-1.0) 33 1.2 (0.7-2.1) 
  CC 74 46 0.9 (0.6-1.4) 26 0.9 (0.5-1.6) 13 1.5 (0.7-3.3) 
     Ptrend = 1.0  Ptrend = 0.3  Ptrend = 0.3 
  GC or CC   1.1 (0.8-1.4)  0.7 (0.5-1.0)  1.2 (0.7-2.2) 
          
CCND1 (rs603965) 11q13 GG 183 105 1.0 (reference) 47 1.0 (reference) 25 1.0 (reference) 
  AG 245 196 1.4 (1.0-1.9)* 63 1.1 (0.7-1.7) 36 1.2 (0.7-2.1) 
  AA 100 73 1.3 (0.8-1.9) 41 1.5 (0.9-2.5) 10 0.7 (0.3-1.6) 
     Ptrend = 0.2  Ptrend = 0.2  Ptrend = 0.6 
  AG or AA   1.3 (1.0-1.8)*  1.2 (0.8-1.8)  1.0 (0.6-1.8) 
          
CCNH (rs2266690) 5q13.3-q14 TT 345 238 1.0 (reference) 97 1.0 (reference) 38 1.0 (reference) 
  CT 166 117 1.0 (0.7-1.3) 52 1.2 (0.8-1.8) 28 1.8 (1.0-3.1)* 
  CC 16 21 2.0 (1.0-3.9)* 1.1 (0.4-3.2) 2.2 (0.6-7.8) 
     Ptrend = 0.2  Ptrend = 0.5  Ptrend = 0.02* 
  CT or CC   1.1 (0.8-1.4)  1.2 (0.8-1.8)  1.8 (1.1-3.2)* 
          
CDKN1A (rs1801270) 6p21.2 CC 434 316 1.0 (reference) 125 1.0 (reference) 55 1.0 (reference) 
  AC 90 50 0.8 (0.6-1.2) 24 0.8 (0.5-1.4) 12 1.1 (0.5-2.2) 
  AA 2.4 (0.7-7.8) 0.5 (0.1-3.3) 4.8 (0.5-43.4) 
     Ptrend = 0.9  Ptrend = 0.4  Ptrend = 0.4 
  AC or AA   0.9 (0.6-1.3)  0.8 (0.5-1.4)  1.2 (0.6-2.3) 
CDKN2A (rs3731249) 9p21 GG 523 359 1.0 (reference) 154 1.0 (reference) 69 1.0 (reference) 
  AG 24 24 1.4 (0.8-2.5) 0.7 (0.2-1.9) 1.1 (0.3-3.4) 
  AA — — — — — — — 
     Ptrend = 0.3  Ptrend = 0.5  Ptrend = 0.9 
  AG or AA   1.4 (0.8-2.5)     
          
CHEK1 (rs506504) 11q24-q24 GG 474 336 1.0 (reference) 143 1.0 (reference) 66 1.0 (reference) 
  AG 42 34 1.1 (0.7-1.8) 0.8 (0.4-1.7) 1.1 (0.4-2.7) 
  AA 4.8 (0.5-47.4) — — 
     Ptrend = 0.3  Ptrend = 0.5  Ptrend = 0.9 
  AG or AA   1.2 (0.8-1.9)     
          
CHEK2 (rs2267130) 22q12.1 AA 167 106 1.0 (reference) 48 1.0 (reference) 21 1.0 (reference) 
  AG 240 202 1.3 (0.9-1.8) 76 1.4 (0.9-2.2) 31 1.0 (0.5-1.8) 
  GG 104 63 0.9 (0.6-1.4) 32 1.2 (0.7-2.2) 15 1.1 (0.5-2.2) 
     Ptrend = 1.0  Ptrend = 0.4  Ptrend = 0.9 
  AG or GG   1.2 (0.9-1.6)  1.3 (0.9-2.1)  1.0 (0.6-1.8) 
          
MDM2 (rs769412) 12q14.3-q15 AA 463 345 1.0 (reference) 126 1.0 (reference) 64 1.0 (reference) 
  AG 70 32 0.6 (0.4-0.9)* 28 1.4 (0.8-2.3) 0.9 (0.4-2.0) 
  GG 0.9 (0.1-10.3) — — 
     Ptrend = 0.03*  Ptrend = 0.3  Ptrend = 0.7 
  AG or GG   0.6 (0.4-0.9)*    0.9 (0.4-2.0) 
          
PTEN (rs701848) 10q23.3 TT 190 138 1.0 (reference) 71 1.0 (reference) 20 1.0 (reference) 
  CT 262 184 0.9 (0.7-1.3) 63 0.8 (0.5-1.1) 39 1.4 (0.8-2.5) 
  CC 93 57 0.9 (0.6-1.3) 23 0.7 (0.4-1.3) 14 1.5 (0.7-3.2) 
     Ptrend = 0.5  Ptrend = 0.2  Ptrend = 0.2 
  CT or CC   0.9 (0.7-1.2)  0.8 (0.5-1.1)  1.4 (0.8-2.5) 
          
TP53 (rs1042522)
 
17p13.1
 
GG 300 213 1.0 (reference) 82 1.0 (reference) 44 1.0 (reference) 
  CG 209 146 1.0 (0.7-1.3) 64 1.1 (0.8-1.7) 24 0.7 (0.4-1.3) 
  CC 38 27 1.0 (0.6-1.7) 13 1.1 (0.5-2.3) 1.0 (0.3-2.8) 
     Ptrend = 0.9  Ptrend = 0.6  Ptrend = 0.4 
  CG or CC   1.0 (0.8-1.3)  1.1 (0.8-1.7)  0.8 (0.4-1.3) 
Gene polymorphismChromosomal locationGenotypeControls (n = 553)
Glioma (n = 388)
Meningioma (n = 162)
Acoustic neuroma (n = 73)
nnOR (95% CI)nOR (95% CI)nOR (95% CI)
CASP8 (rs13113) 2q33-q34 TT 155 129 1.0 (reference) 60 1.0 (reference) 22 1.0 (reference) 
  AT 258 166 0.8 (0.6-1.1) 78 0.8 (0.5-1.2) 30 0.8 (0.5-1.6) 
  AA 98 75 0.9 (0.6-1.3) 18 0.5 (0.3-0.9)* 17 1.1 (0.6-2.3) 
     Ptrend = 0.4  Ptrend = 0.03*  Ptrend = 0.8 
  AT or AA   0.8 (0.6-1.1)  0.7 (0.5-1.1)  0.9 (0.5-1.6) 
          
CASP8 (rs1045485) 2q33-q34 GG 426 284 1.0 (reference) 117 1.0 (reference) 50 1.0 (reference) 
  GC 118 95 1.2 (0.8-1.6) 38 1.4 (0.9-2.1) 20 1.5 (0.8-2.6) 
  CC 0.7 (0.1-2.8) 3.6 (1.0-13.1)* 5.6 (1.2-26.6)* 
     Ptrend = 0.6  Ptrend = 0.04*  Ptrend = 0.04* 
  GC or CC   1.1 (0.8-1.5)  1.5 (1.0-2.3)*  1.6 (0.9-2.8) 
          
CCND1 (rs678653) 11q13 GG 198 138 1.0 (reference) 72 1.0 (reference) 24 1.0 (reference) 
  GC 243 188 1.1 (0.8-1.5) 55 0.6 (0.4-1.0) 33 1.2 (0.7-2.1) 
  CC 74 46 0.9 (0.6-1.4) 26 0.9 (0.5-1.6) 13 1.5 (0.7-3.3) 
     Ptrend = 1.0  Ptrend = 0.3  Ptrend = 0.3 
  GC or CC   1.1 (0.8-1.4)  0.7 (0.5-1.0)  1.2 (0.7-2.2) 
          
CCND1 (rs603965) 11q13 GG 183 105 1.0 (reference) 47 1.0 (reference) 25 1.0 (reference) 
  AG 245 196 1.4 (1.0-1.9)* 63 1.1 (0.7-1.7) 36 1.2 (0.7-2.1) 
  AA 100 73 1.3 (0.8-1.9) 41 1.5 (0.9-2.5) 10 0.7 (0.3-1.6) 
     Ptrend = 0.2  Ptrend = 0.2  Ptrend = 0.6 
  AG or AA   1.3 (1.0-1.8)*  1.2 (0.8-1.8)  1.0 (0.6-1.8) 
          
CCNH (rs2266690) 5q13.3-q14 TT 345 238 1.0 (reference) 97 1.0 (reference) 38 1.0 (reference) 
  CT 166 117 1.0 (0.7-1.3) 52 1.2 (0.8-1.8) 28 1.8 (1.0-3.1)* 
  CC 16 21 2.0 (1.0-3.9)* 1.1 (0.4-3.2) 2.2 (0.6-7.8) 
     Ptrend = 0.2  Ptrend = 0.5  Ptrend = 0.02* 
  CT or CC   1.1 (0.8-1.4)  1.2 (0.8-1.8)  1.8 (1.1-3.2)* 
          
CDKN1A (rs1801270) 6p21.2 CC 434 316 1.0 (reference) 125 1.0 (reference) 55 1.0 (reference) 
  AC 90 50 0.8 (0.6-1.2) 24 0.8 (0.5-1.4) 12 1.1 (0.5-2.2) 
  AA 2.4 (0.7-7.8) 0.5 (0.1-3.3) 4.8 (0.5-43.4) 
     Ptrend = 0.9  Ptrend = 0.4  Ptrend = 0.4 
  AC or AA   0.9 (0.6-1.3)  0.8 (0.5-1.4)  1.2 (0.6-2.3) 
CDKN2A (rs3731249) 9p21 GG 523 359 1.0 (reference) 154 1.0 (reference) 69 1.0 (reference) 
  AG 24 24 1.4 (0.8-2.5) 0.7 (0.2-1.9) 1.1 (0.3-3.4) 
  AA — — — — — — — 
     Ptrend = 0.3  Ptrend = 0.5  Ptrend = 0.9 
  AG or AA   1.4 (0.8-2.5)     
          
CHEK1 (rs506504) 11q24-q24 GG 474 336 1.0 (reference) 143 1.0 (reference) 66 1.0 (reference) 
  AG 42 34 1.1 (0.7-1.8) 0.8 (0.4-1.7) 1.1 (0.4-2.7) 
  AA 4.8 (0.5-47.4) — — 
     Ptrend = 0.3  Ptrend = 0.5  Ptrend = 0.9 
  AG or AA   1.2 (0.8-1.9)     
          
CHEK2 (rs2267130) 22q12.1 AA 167 106 1.0 (reference) 48 1.0 (reference) 21 1.0 (reference) 
  AG 240 202 1.3 (0.9-1.8) 76 1.4 (0.9-2.2) 31 1.0 (0.5-1.8) 
  GG 104 63 0.9 (0.6-1.4) 32 1.2 (0.7-2.2) 15 1.1 (0.5-2.2) 
     Ptrend = 1.0  Ptrend = 0.4  Ptrend = 0.9 
  AG or GG   1.2 (0.9-1.6)  1.3 (0.9-2.1)  1.0 (0.6-1.8) 
          
MDM2 (rs769412) 12q14.3-q15 AA 463 345 1.0 (reference) 126 1.0 (reference) 64 1.0 (reference) 
  AG 70 32 0.6 (0.4-0.9)* 28 1.4 (0.8-2.3) 0.9 (0.4-2.0) 
  GG 0.9 (0.1-10.3) — — 
     Ptrend = 0.03*  Ptrend = 0.3  Ptrend = 0.7 
  AG or GG   0.6 (0.4-0.9)*    0.9 (0.4-2.0) 
          
PTEN (rs701848) 10q23.3 TT 190 138 1.0 (reference) 71 1.0 (reference) 20 1.0 (reference) 
  CT 262 184 0.9 (0.7-1.3) 63 0.8 (0.5-1.1) 39 1.4 (0.8-2.5) 
  CC 93 57 0.9 (0.6-1.3) 23 0.7 (0.4-1.3) 14 1.5 (0.7-3.2) 
     Ptrend = 0.5  Ptrend = 0.2  Ptrend = 0.2 
  CT or CC   0.9 (0.7-1.2)  0.8 (0.5-1.1)  1.4 (0.8-2.5) 
          
TP53 (rs1042522)
 
17p13.1
 
GG 300 213 1.0 (reference) 82 1.0 (reference) 44 1.0 (reference) 
  CG 209 146 1.0 (0.7-1.3) 64 1.1 (0.8-1.7) 24 0.7 (0.4-1.3) 
  CC 38 27 1.0 (0.6-1.7) 13 1.1 (0.5-2.3) 1.0 (0.3-2.8) 
     Ptrend = 0.9  Ptrend = 0.6  Ptrend = 0.4 
  CG or CC   1.0 (0.8-1.3)  1.1 (0.8-1.7)  0.8 (0.4-1.3) 
*

Significant at P < 0.05 level.

The Ex14-271A>T (rs13113) and Ex13+51G>C (D285H; rs1045485) polymorphisms in CASP8 were both associated with risk of meningioma: the AT and AA variants of CASP8 Ex14-271A>T were associated with significantly decreased risk of meningioma (ORAT, 0.8; 95% CI, 0.5, 1.2 and ORAA, 0.5; 95% CI, 0.3-0.9; Ptrend = 0.03), whereas the GC and CC variants of CASP8 Ex13+51G>C were associated with increased risk of meningioma (ORGC, 1.4; 95% CI, 0.9, 2.1 and ORCC, 3.6; 95% CI, 1.0, 13.1; Ptrend = 0.04) as well as acoustic neuroma (ORGC, 1.5; 95% CI, 0.8, 2.6 and ORCC, 5.6; 95% CI, 1.2, 26.6; Ptrend = 0.04). No association was observed between either of the CASP8 polymorphisms and glioma. The Ex14-271A>T (rs13113) and Ex13+51G>C (rs1045485) polymorphisms of CASP8 were found to be in linkage disequilibrium (D′ = 1.0, r2 = 0.12). Haplotype analyses indicated that the CT haplotype of the two SNPs was associated with significantly increased risk of meningioma (OR, 1.7; 95% CI, 1.1, 2.64), but not glioma or acoustic neuroma (Table 4).

Table 4.

ORs and 95% CIs for the association between common CASP8 haplotypes and risk of glioma, meningioma and acoustic neuroma, NCI adult brain tumor study, 1994 to 1998

Haplotype*Glioma (n = 354)
Meningioma (n = 133)
Acoustic neuroma (n = 66)
Controls (n = 495)
%OR%OR%OR%
G-A 0.44 1.0 (reference) 0.39 1.0 (reference) 0.46 1.0 (reference) 0.46 
G-T 0.42 1.0 (0.8-1.3) 0.44 1.3 (0.9-1.7) 0.35 0.8 (0.5-1.3) 0.42 
C-T 0.14 1.1 (0.8-1.5) 0.17 1.7 (1.1-2.6) 0.19 1.5 (0.9-2.5) 0.12 
Global test omnibus  0.67  0.048  0.07  
Haplotype*Glioma (n = 354)
Meningioma (n = 133)
Acoustic neuroma (n = 66)
Controls (n = 495)
%OR%OR%OR%
G-A 0.44 1.0 (reference) 0.39 1.0 (reference) 0.46 1.0 (reference) 0.46 
G-T 0.42 1.0 (0.8-1.3) 0.44 1.3 (0.9-1.7) 0.35 0.8 (0.5-1.3) 0.42 
C-T 0.14 1.1 (0.8-1.5) 0.17 1.7 (1.1-2.6) 0.19 1.5 (0.9-2.5) 0.12 
Global test omnibus  0.67  0.048  0.07  
*

Order of SNPs is CASP8 Ex13+51G>C and CASP8 Ex14-271A>T.

ORs and global test are adjusted for age and sex. Analyses restricted to non-Hispanic Whites.

The Ex8+49T>C variant of CCNH (rs2266690) was associated with increased risk of glioma (ORCT, 1.0; 95% CI, 0.7-1.3 and ORCC, 2.0; 95% CI, 1.0-3.9; Ptrend = 0.2) and acoustic neuroma (ORCT, 1.8; 95% CI, 1.0-3.1 and ORCC, 2.2; 95% CI, 0.6-7.8; Ptrend = 0.02), but was not associated with increased risk of meningioma. These trends were more strongly apparent in analyses restricted to non-Hispanic whites for both glioma (ORCT, 1.1; 95% CI, 0.8-1.5 and ORCC, 2.3; 95% CI, 1.1-4.7; Ptrend = 0.07) and acoustic neuroma (ORCT, 2.2; 95% CI, 1.2-3.9 and ORCC, 3.2; 95% CI, 0.9-11.9; Ptrend = 0.004).

Individuals with the variant allele of the CCND1 Ex4-1G>A polymorphism (rs603965) also showed some indication of increased risk for glioma (ORAG/GG, 1.3; 95% CI, 1.0-1.8). Although the rare C-A haplotype (frequency = 0.01) was associated with decreased risk of glioma (OR, 0.5; 95% CI, 0.3-0.7), the global test was not statistically significant (P = 0.15). No CCND1 haplotype associations were observed for meningioma or acoustic neuroma. Significantly reduced risk of glioma was observed for individuals with the G variant of the MDM2 Ex12+162A>G (rs769412) polymorphism (ORAG/GG, 0.6; 95% CI, 0.4-0.9). We observed no significant gene-gene interaction between the TP53 Ex4+119C>G nonsynonymous SNP and the MDM2 Ex12+162A>G variant (results not shown). Risk estimates from unadjusted analyses were very similar to adjusted risk estimates. ORs for all statistically significant associations remained very similar when restricted to non-Hispanic whites (Table 5). Although gender-specific estimates were not as stable given smaller numbers, results were generally consistent in males and females (Table 6). None of the observed associations passed the false discovery rate criterion of 15% chance of false-positive findings.

Table 5.

Adjusted ORs for variants of selected apoptosis/cell cycle–related SNPs in the NCI adult brain tumor study, Non-Hispanic Whites only

Gene polymorphismChromosomal locationGenotypeGlioma
Meningioma
Acoustic neuroma
nOR (95% CI)nOR (95% CI)nOR (95% CI)
CASP8 (rs13113) 2q33-q34 TT 112 1.0 (reference) 45 1.0 (reference) 19 1.0 (reference) 
  AT 153 0.8 (0.6-1.1) 66 0.8 (0.5-1.2) 28 0.8 (0.4-1.6) 
  AA 71 0.9 (0.6-1.4) 17 0.5 (0.3-1.0) 15 1.1 (0.5-2.3) 
    Ptrend = 0.4  Ptrend = 0.04*  Ptrend = 0.9 
CASP8 (rs1045485) 2q33-q34 GG 257 1.0 (reference) 91 1.0 (reference) 44 1.0 (reference) 
  GC 89 1.1 (0.8-1.6) 35 1.4 (0.9-2.3) 19 1.5 (0.8-2.7) 
  CC 0.7 (0.2-2.8) 4.0 (1.1-14.8)* 5.4 (1.2-26.3)* 
    Ptrend = 0.6  Ptrend = 0.03*  Ptrend = 0.03* 
CCND1 (rs603965) 11q13 GG 93 1.0 (reference) 36 1.0 (reference) 22 1.0 (reference) 
  AG 179 1.3 (1.0-1.9)* 54 1.0 (0.6-1.7) 34 1.2 (0.6-2.1) 
  AA 69 1.2 (0.8-1.9) 35 1.4 (0.8-2.5) 0.6 (0.3-1.5) 
    Ptrend = 0.2  Ptrend = 0.3  Ptrend = 0.4 
CCNH (rs2266690) 5q13.3-q14 TT 211 1.0 (reference) 75 1.0 (reference) 32 1.0 (reference) 
  CT 112 1.1 (0.8-1.5) 46 1.4 (0.9-2.2) 27 2.2 (1.2-3.9)* 
  CC 21 2.3 (1.1-4.7)* 1.4 (0.5-4.4) 3.2 (0.9-11.9) 
    Ptrend = 0.07  Ptrend = 0.1  Ptrend = 0.004* 
MDM2 (rs769412)
 
12q14.3-q15 AA 313 1.0 (reference) 106 1.0 (reference) 57 1.0 (reference) 
  AG 30 0.6 (0.4-1.0) 20 1.3 (0.7-2.3) 1.0 (0.5-2.4) 
  GG 0.8 (0.1-9.6) — — 
    Ptrend = 0.06  Ptrend = 0.5  Ptrend = 0.9 
Gene polymorphismChromosomal locationGenotypeGlioma
Meningioma
Acoustic neuroma
nOR (95% CI)nOR (95% CI)nOR (95% CI)
CASP8 (rs13113) 2q33-q34 TT 112 1.0 (reference) 45 1.0 (reference) 19 1.0 (reference) 
  AT 153 0.8 (0.6-1.1) 66 0.8 (0.5-1.2) 28 0.8 (0.4-1.6) 
  AA 71 0.9 (0.6-1.4) 17 0.5 (0.3-1.0) 15 1.1 (0.5-2.3) 
    Ptrend = 0.4  Ptrend = 0.04*  Ptrend = 0.9 
CASP8 (rs1045485) 2q33-q34 GG 257 1.0 (reference) 91 1.0 (reference) 44 1.0 (reference) 
  GC 89 1.1 (0.8-1.6) 35 1.4 (0.9-2.3) 19 1.5 (0.8-2.7) 
  CC 0.7 (0.2-2.8) 4.0 (1.1-14.8)* 5.4 (1.2-26.3)* 
    Ptrend = 0.6  Ptrend = 0.03*  Ptrend = 0.03* 
CCND1 (rs603965) 11q13 GG 93 1.0 (reference) 36 1.0 (reference) 22 1.0 (reference) 
  AG 179 1.3 (1.0-1.9)* 54 1.0 (0.6-1.7) 34 1.2 (0.6-2.1) 
  AA 69 1.2 (0.8-1.9) 35 1.4 (0.8-2.5) 0.6 (0.3-1.5) 
    Ptrend = 0.2  Ptrend = 0.3  Ptrend = 0.4 
CCNH (rs2266690) 5q13.3-q14 TT 211 1.0 (reference) 75 1.0 (reference) 32 1.0 (reference) 
  CT 112 1.1 (0.8-1.5) 46 1.4 (0.9-2.2) 27 2.2 (1.2-3.9)* 
  CC 21 2.3 (1.1-4.7)* 1.4 (0.5-4.4) 3.2 (0.9-11.9) 
    Ptrend = 0.07  Ptrend = 0.1  Ptrend = 0.004* 
MDM2 (rs769412)
 
12q14.3-q15 AA 313 1.0 (reference) 106 1.0 (reference) 57 1.0 (reference) 
  AG 30 0.6 (0.4-1.0) 20 1.3 (0.7-2.3) 1.0 (0.5-2.4) 
  GG 0.8 (0.1-9.6) — — 
    Ptrend = 0.06  Ptrend = 0.5  Ptrend = 0.9 
*

Models adjusted for study matching factors age, sex, study site, and residential proximity to treatment hospital.

Table 6.

Age-adjusted ORs for variants of selected apoptosis/cell cycle–related SNPs in the NCI adult brain tumor study, Non-Hispanic Whites only, by gender

GenotypeMales
Females
nOR (95% CI)nOR (95% CI)
Glioma      
    CCND1 (rs603965) GG 54 1.0 (reference) 39 1.0 (reference) 
 AG 98 1.3 (0.8-2.0) 81 1.4 (0.8-2.2) 
 AA 37 1.6 (0.9-2.9) 32 1.0 (0.6-1.9) 
    CCNH (rs2266690) TT 125 1.0 (reference) 86 1.0 (reference) 
 CT 53 0.8 (0.5-1.3) 59 1.6 (1.0-2.4) 
 CC 12 2.0 (0.7-5.3) 2.5 (0.9-7.2) 
    MDM2 (rs769412) AA 174 1.0 (reference) 139 1.0 (reference) 
 AG 14 0.5 (0.2-0.9) 16 0.8 (0.4-1.6) 
 GG — — 
Meningioma      
    CASP8 (rs13113) TT 10 1.0 (reference) 35 1.0 (reference) 
 AT 14 0.9 (0.4-2.1) 52 0.8 (0.4-1.3) 
 AA 0.6 (0.2-2.1) 13 0.5 (0.2-1.1) 
    CASP8 (rs1045485) GG 22 1.0 (reference) 69 1.0 (reference) 
 GC 0.5 (0.2-1.6) 31 1.9 (1.1-3.3) 
 CC 5.6 (0.8-38.9) 2.2 (0.4-11.4) 
Acoustic neuroma      
    CASP8 (rs1045485) GG 18 1.0 (reference) 26 1.0 (reference) 
 GC 0.8 (0.3-2.2) 14 2.1 (1.0-4.4) 
 CC — 8.0 (1.4-46.9) 
    CCNH (rs2266690) TT 10 1.0 (reference) 22 1.0 (reference) 
 CT 11 2.1 (0.8-5.2) 16 1.6 (0.8-3.3) 
 CC 4.8 (0.8-30.5) 1.8 (0.3-9.4) 
GenotypeMales
Females
nOR (95% CI)nOR (95% CI)
Glioma      
    CCND1 (rs603965) GG 54 1.0 (reference) 39 1.0 (reference) 
 AG 98 1.3 (0.8-2.0) 81 1.4 (0.8-2.2) 
 AA 37 1.6 (0.9-2.9) 32 1.0 (0.6-1.9) 
    CCNH (rs2266690) TT 125 1.0 (reference) 86 1.0 (reference) 
 CT 53 0.8 (0.5-1.3) 59 1.6 (1.0-2.4) 
 CC 12 2.0 (0.7-5.3) 2.5 (0.9-7.2) 
    MDM2 (rs769412) AA 174 1.0 (reference) 139 1.0 (reference) 
 AG 14 0.5 (0.2-0.9) 16 0.8 (0.4-1.6) 
 GG — — 
Meningioma      
    CASP8 (rs13113) TT 10 1.0 (reference) 35 1.0 (reference) 
 AT 14 0.9 (0.4-2.1) 52 0.8 (0.4-1.3) 
 AA 0.6 (0.2-2.1) 13 0.5 (0.2-1.1) 
    CASP8 (rs1045485) GG 22 1.0 (reference) 69 1.0 (reference) 
 GC 0.5 (0.2-1.6) 31 1.9 (1.1-3.3) 
 CC 5.6 (0.8-38.9) 2.2 (0.4-11.4) 
Acoustic neuroma      
    CASP8 (rs1045485) GG 18 1.0 (reference) 26 1.0 (reference) 
 GC 0.8 (0.3-2.2) 14 2.1 (1.0-4.4) 
 CC — 8.0 (1.4-46.9) 
    CCNH (rs2266690) TT 10 1.0 (reference) 22 1.0 (reference) 
 CT 11 2.1 (0.8-5.2) 16 1.6 (0.8-3.3) 
 CC 4.8 (0.8-30.5) 1.8 (0.3-9.4) 

Although the etiology of brain and central nervous system tumors is largely unknown, several lines of evidence indicate that events that alter cell cycle control and apoptosis pathways could be of importance. Among the few confirmed risk factors for brain tumors are certain rare predisposing genetic syndromes (15-17). At least three of these syndromes, Li-Fraumeni syndrome, neurofibromatosis 1, and retinoblastoma, are characterized by germ line mutations in genes affecting progression through the cell cycle and apoptosis (TP53, NF1, and RB1, respectively). The importance of apoptosis and cell cycle control in brain tumor etiology is also suggested by mutations or loss of expression of cell cycle/apoptosis genes (PTEN, CDKN2A, MDM2, TP53, and RB) in brain tumors (3). More generally, these pathways are of importance in other cancers types, including breast, melanoma, and colon (1).

Activation of the caspase (cysteine aspartyl-specific proteases) family of intracellular cysteine proteases is central to the initiation of apoptosis. CASP8 is the apical caspase in the tumor necrosis factor family death receptor pathway (extrinsic pathway) of apoptosis (18). We observed novel associations between common genetic variants in CASP8 and brain tumors: Ex14-271A>T was associated with decreased risk of meningioma, whereas Ex13+51G>C (D285H) was associated with increased risk of meningioma and acoustic neuroma. Our haplotype analyses further implicates the chromosomal region indicated by the two CASP8 SNPs we evaluated; the significantly increased risk of meningioma observed with the CT haplotype suggests that this chromosomal region either harbors or is in linkage disequilibrium with a functional polymorphism that affects meningioma risk. Others have noted decreased risk of non–Hodgkin's lymphoma with the Ex14-271A>T variant (19). Our finding of increased risk for the D285H nonsynonymous SNP is surprising because of its association with decreased breast cancer risk in other studies, including a large international case-control consortium (20-22). Given that the functionality of this particular SNP has not been well characterized, it is premature to speculate on a mechanistic reason for the inverse direction and the possible role of chance cannot be ruled out.

Elevated risks of glioma and acoustic neuroma were observed for several cell cycle gene polymorphisms. Molecular complexes of the cyclin-dependent kinases and their associated cyclins are responsible for sending out signals from the cell cycle clock to responder molecules that effect the transition of the cell through its cycle of growth and division (23). We found evidence of increased risk of glioma with the Ex4-1G>A (G870A) polymorphism of the CCND1 gene coding for the cyclin D1 protein. Although no prior study has examined brain tumor risk with respect to the CCND1 G870A variant, increased risk has been observed with the A allele for several cancer sites, including bladder cancer, colorectal cancer, head and neck cancer, lung cancer, leukemia, and non–Hodgkin's lymphoma (24, 25). Laboratory studies indicate that the A870 allele hinders splicing of the CCND1 protein and may cause production of a variant splice product (“transcript b”; ref. 24). However, it is still unclear as to whether transcript b production is directly associated with the G/A 870 polymorphism or cancer risk.

In our study, we examined common SNPs in TP53 and MDM2, two genes central to cell cycle progression, cell survival, and genomic stability, and found no association with the well-studied exon 4 nonsynonymous SNP in TP53. This finding is consistent with null findings for this polymorphism in a Swedish study of 202 glioma and 164 meningioma cases (5), and a multicenter study of brain tumors (7), as well as one smaller study of brain tumors (4). Although a possible association with this polymorphism was reported in a previous study with adult and pediatric astrocytomas, DNA for that analysis was extracted from tumor samples, raising the possibility that the mutations occurred after tumor initiation (6). The potential importance of the TP53 gene in brain tumor risk, however, has been shown by the fact that other polymorphisms have been associated with differences in risk: an intron 6 variant (rs1625895) in TP53 has been associated with decreased risk of glioma and glioblastoma, as has the 1-2-2 haplotype of TP53 (promoter-codon72-intron; ref. 7), whereas the CC-CG-CC haplotype (promoter-exon4-intron6) of TP53 has been associated with increased risk of glioma and meningioma (5).

We observed decreased risk of glioma in individuals with the Ex12+162A>G polymorphism of the MDM2 gene. The relationship between this MDM2 polymorphism and brain tumor has not been previously reported in the literature. However, another common polymorphism in MDM2 (IVS+309T>G, rs2279744) has been associated with increased risk of gastric cancer (26), esophageal cancer (27), lung cancer (28, 29), and bladder cancer (30) and may accelerate the risk of tumor formation in patients with familial breast cancer (31), Li-Fraumeni syndrome (32, 33), colorectal cancer (34), and gastric cancer (34). Given that MDM2 is a strong negative regulator of the p-53 cascade, one might expect a gene-gene interaction between polymorphisms in these two genes. Although previous studies have reported gene-gene interactions between TP53 and the MDM2 309T>G polymorphism (34, 35), we did not detect any gene-gene interactions between the MDM2 Ex12+162A>G and the TP53 P72R polymorphisms.

This study had adequate statistical power to detect moderate to strong (but not weak) main effects of common genetic polymorphisms. Genotyping for the study was standardized; QC samples indicated high reproducibility of the genotyping results; and controls were in Hardy-Weinberg equilibrium for all polymorphisms under study.

Despite these strengths, we underscore the need for replication of our findings given the large number of false positives generated in genetic association studies (36), and the fact that our results did not meet the false discovery rate of 15%. Moreover, it is possible that the notable SNPs are actually in linkage disequilibrium with other causally relevant polymorphisms, but further work is needed to determine this. A more systematic approach toward coverage of these genes is warranted in follow-up studies. Although nonparticipation in the blood draw was higher among controls than cases, we believe that this is unlikely to be related to genotype, and thus unlikely to bias our results.

Our findings suggest that common variants in the apoptosis and cell cycle pathways may be important in brain tumor etiology. In particular, if our observations are verified, our results indicate that CASP8, CCND1, CCNH, and MDM2 may be promising candidates for brain tumor susceptibility genes. Future research in this area should include more detailed coverage of SNPs within the genes implicated in this paper and examine related genes (e.g., caspases) in the apoptosis and cell cycle control pathways.

Grant support: Intramural funds from the National Cancer Institute, NIH, Department of Health and Human Services.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Vogelstein B, Kinzler KW. Cancer genes and the pathways they control.
Nat Med
2004
;
10
:
789
–99.
2
The cell cycle and programmed cell death. In: Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P, authors. Molecular biology of the cell. New York: Garland Science; 2002.
3
Holland EC. Gliomagenesis: genetic alterations and mouse models.
Nat Rev Genet
2001
;
2
:
120
–9.
4
Biros E, Kalina I, Kohut A, Bogyiova E, Salagovic J, Sulla I. Allelic and haplotype frequencies of the p53 polymorphisms in brain tumor patients.
Physiol Res
2002
;
51
:
59
–64.
5
Malmer B, Feychting M, Lonn S, Ahlbom A, Henriksson R. p53 Genotypes and risk of glioma and meningioma.
Cancer Epidemiol Biomarkers Prev
2005
;
14
:
2220
–3.
6
Parhar P, Ezer R, Shao Y, Allen JC, Miller DC, Newcomb EW. Possible association of p53 codon 72 polymorphism with susceptibility to adult and pediatric high-grade astrocytomas.
Brain Res Mol Brain Res
2005
;
137
:
98
–103.
7
Malmer BS, Feychting M, Lonn S, et al. Genetic variation in p53 and ATM haplotypes and risk of glioma and meningioma.
J Neurooncol
2007
;
82
:
229
–37.
8
Sadetzki S, Flint-Richter P, Starinsky S, et al. Genotyping of patients with sporadic and radiation-associated meningiomas.
Cancer Epidemiol Biomarkers Prev
2005
;
14
:
969
–76.
9
Inskip PD, Tarone RE, Hatch EE, et al. Cellular-telephone use and brain tumors.
N Engl J Med
2001
;
344
:
79
–86.
10
Daly A, Steen V, Fairbrother K, Idle J. CYP2D6 multiallelism.
Methods Enzymol
1996
;
272
:
199
–210.
11
Packer BR, Yeager M, Burdett L, et al. SNP500Cancer: a public resource for sequence validation, assay development, and frequency analysis for genetic variation in candidate genes.
Nucleic Acids Res
2006
;
34
:
D617
–21.
12
Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing.
JASA
1995
;
57
:
289
–300.
13
Excoffier L, Slatkin M. Maximum-likelihood estimation of molecular haplotype frequencies in a diploid population.
Mol Biol Evol
1995
;
12
:
921
–7.
14
Schaid DJ, Rowland CM, Tines DE, Jacobson RM, Poland GA. Score tests for association between traits and haplotypes when linkage phase is ambiguous.
Am J Hum Genet
2002
;
70
:
425
–34.
15
Inskip PD, Linet MS, Heineman EF. Etiology of brain tumors in adults.
Epidemiol Rev
1995
;
17
:
382
–414.
16
Wrensch M, Minn Y, Chew T, Bondy M, Berger MS. Epidemiology of primary brain tumors: current concepts and review of the literature.
Neuro-oncol
2002
;
4
:
278
–99.
17
Schwartzbaum JA, Fisher JL, Aldape KD, Wrensch M. Epidemiology and molecular pathology of glioma.
Nat Clin Pract Neurol
2006
;
2
:
494
–503; quiz 1 p following 516.
18
Reed JC. Mechanisms of apoptosis.
Am J Pathol
2000
;
157
:
1415
–30.
19
Lan Q, Zheng T, Chanock S, et al. Genetic variants in caspase genes and susceptibility to non-Hodgkin lymphoma.
Carcinogenesis
2007
;
28
:
823
–7.
20
Breast Cancer Association C. Commonly studied single-nucleotide polymorphisms and breast cancer: results from the Breast Cancer Association Consortium.
J Natl Cancer Inst
2006
;
98
:
1382
–96.
21
Frank B, Hemminki K, Wappenschmidt B, et al. Association of the CASP10 V410I variant with reduced familial breast cancer risk and interaction with the CASP8 D302H variant.
Carcinogenesis
2006
;
27
:
606
–9.
22
MacPherson G, Healey CS, Teare MD, et al. Association of a common variant of the CASP8 gene with reduced risk of breast cancer.
J Natl Cancer Inst
2004
;
96
:
1866
–9.
23
Weinberg RA. The biology of cancer. New York: Garland Science Taylor and Francis; 2007.
24
Knudsen KE, Diehl JA, Haiman CA, Knudsen ES. Cyclin D1: polymorphism, aberrant splicing and cancer risk.
Oncogene
2006
;
25
:
1620
–8.
25
Wang SS, Cozen W, Severson RK, et al. Cyclin D1 splice variant and risk for non-Hodgkin lymphoma.
Hum Genet
2006
;
120
:
297
–300.
26
Ohmiya N, Taguchi A, Mabuchi N, et al. MDM2 promoter polymorphism is associated with both an increased susceptibility to gastric carcinoma and poor prognosis.
J Clin Oncol
2006
;
24
:
4434
–40.
27
Hong Y, Miao X, Zhang X, et al. The role of P53 and MDM2 polymorphisms in the risk of esophageal squamous cell carcinoma.
Cancer Res
2005
;
65
:
9582
–7.
28
Park SH, Choi JE, Kim EJ, et al. MDM2 309T>G polymorphism and risk of lung cancer in a Korean population.
Lung Cancer
2006
;
54
:
19
–24.
29
Lind H, Zienolddiny S, Ekstrom PO, Skaug V, Haugen A. Association of a functional polymorphism in the promoter of the MDM2 gene with risk of nonsmall cell lung cancer.
Int J Cancer
2006
;
119
:
718
–21.
30
Onat OE, Tez M, Ozcelik T, Toruner GA. MDM2 T309G polymorphism is associated with bladder cancer.
Anticancer Res
2006
;
26
:
3473
–5.
31
Wasielewski M, Nagel JH, Brekelmans C, et al. MDM2 SNP309 accelerates familial breast carcinogenesis independently of estrogen signaling. Breast Cancer Res Treat 2006 Nov 2 [Epub ahead of print].
32
Bougeard G, Baert-Desurmont S, Tournier I, et al. Impact of the MDM2 SNP309 and p53 Arg72Pro polymorphism on age of tumour onset in Li-Fraumeni syndrome.
J Med Genet
2006
;
43
:
531
–3.
33
Ruijs MW, Schmidt MK, Nevanlinna H, et al. The single-nucleotide polymorphism 309 in the MDM2 gene contributes to the Li-Fraumeni syndrome and related phenotypes.
Eur J Hum Genet
2007
;
15
:
110
–4.
34
Menin C, Scaini MC, De Salvo GL, et al. Association between MDM2-309 and age at colorectal cancer diagnosis according to p53 mutation status.
J Natl Cancer Inst
2006
;
98
:
285
–8.
35
Boersma BJ, Howe TM, Goodman JE, et al. Association of breast cancer outcome with status of p53 and MDM2 SNP309.
J Natl Cancer Inst
2006
;
98
:
911
–9.
36
Ioannidis JP, Ntzani EE, Trikalinos TA, Contopoulos-Ioannidis DG. Replication validity of genetic association studies.
Nat Genet
2001
;
29
:
306
–9.
37
Kleihues P, Cavanee WK, editors. Pathology and genetics: tumors of the nervous system. World Health Organization classification of tumors. Lyons: IARC Press; 2000.