Chronic inflammation has been hypothesized to be a risk factor for prostate cancer. The Toll-like receptor 4 (TLR4) presents the bacterial lipopolysaccharide (LPS), which interacts with ligand-binding protein and CD14 (LPS receptor) and activates expression of inflammatory genes through nuclear factor-κB and mitogen-activated protein kinase signaling. A previous case-control study found a modest association of a polymorphism in the TLR4 gene [11381G/C, GG versus GC/CC: odds ratio (OR), 1.26] with risk of prostate cancer. We assessed if sequence variants of TLR4 were associated with the risk of prostate cancer. In a nested case-control design within the Health Professionals Follow-up Study, we identified 700 participants with prostate cancer diagnosed after they had provided a blood specimen in 1993 and before January 2000. Controls were 700 age-matched men without prostate cancer who had had a prostate-specific antigen test after providing a blood specimen. We genotyped 16 common (>5%) single nucleotide polymorphisms (SNP) discovered in a resequencing study spanning TLR4 to test for association between sequence variation in TLR4 and prostate cancer. Homozygosity for the variant alleles of eight SNPs was associated with a statistically significantly lower risk of prostate cancer (TLR4_1893, TLR4_2032, TLR4_2437, TLR4_7764, TLR4_11912, TLR4_16649, TLR4_17050, and TLR4_17923), but the TLR4_15844 polymorphism corresponding to 11381G/C was not associated with prostate cancer (GG versus CG/CC: OR, 1.01; 95% confidence interval, 0.79-1.29). Six common haplotypes (cumulative frequency, 81%) were observed; the global test for association between haplotypes and prostate cancer was statistically significant (χ2 = 14.8 on 6 degrees of freedom; P = 0.02). Two common haplotypes were statistically significantly associated with altered risk of prostate cancer. Inherited polymorphisms of the innate immune gene TLR4 are associated with risk of prostate cancer. (Cancer Res 2005; 65(24): 11771-8)

Human Toll-like receptor (TLR) is a type I transmembrane protein with an extracellular domain consisting of a leucine-rich repeat region and an intracellular domain homologous to that of the human interleukin (IL)-1 receptor (1). Lipopolysaccharide (LPS) interacts with ligand-binding protein and CD14, a LPS receptor, to present LPS to TLR4, which then activates the expression of inflammatory genes (e.g., IL-1, IL-6, and IL-8) through either nuclear factor-κB or mitogen-activated protein kinase signaling (2, 3). Human TLR4 is located on chromosome 9q32-q33, has four exons, and is highly expressed in lymphocytes, spleen, and the heart (4). Exposure to bacterial products or proinflammatory cytokines increases TLR4 expression in monocytes and polymorphonuclear leukocytes (2, 5). TLR4-deficient or TLR4-mutant mice showed lower response to viral and bacterial infection than did wild-type mice (68), which was related to a subsequent reduction in the innate immune response. Previous work has reported that genetic variation in TLR4 is related to the risk of atherosclerosis (9, 10), septic shock, smallpox, Chlamydia trachomatis, Chlamydia pneumoniae, and Mycobacterium tuberculosis and related to the pathogen recognition of Gram-negative bacteria and respiratory syncytial virus (1113).

An association study (14) in a Swedish population explored the relationship between TLR4 sequence variations and risk of prostate cancer. In that study, among eight single nucleotide polymorphisms (SNP) studied, a sequence variant (11381G/C) in the 3′ untranslated region (UTR) of TLR4 was associated with prostate cancer risk [GG versus CG/CC: odds ratio (OR), 1.26; 95% confidence interval (95% CI), 1.01-1.57]. The same investigators also found that other SNPs in other genes in the TLR family, which formed TLR6-TLR1-TLR10 gene cluster, were associated with the risk of prostate cancer (15). Chronic inflammation has been associated with some cancers (e.g., cervix, breast, primary liver, and bladder cancers; ref. 16). An emerging body of evidence, including studies on sexually transmitted infections, clinical prostatitis, and genetic and circulating markers of inflammation and response to infection, supports a possible link between chronic intraprostatic inflammation and risk of prostate cancer (17). Therefore, we hypothesized that genetic polymorphisms of TLR4 are associated with the risk of prostate cancer. We did both SNP and haplotype analyses to test this hypothesis.

Study population. In this nested case-control study, incident prostate cancer cases were identified from the ongoing Health Professionals Follow-up Study (HPFS) with follow-up from 1986 to 2000. A total of 51,529 U.S. men ages 40 to 75 years were enrolled in 1986. Every participant completed a mailed questionnaire on demographics, lifestyle, and medical history and a semiquantitative food frequency questionnaire at baseline. Information on exposures and diseases was updated every other year, and diet information was updated every 4 years. Deaths were identified through reports by family members, the follow-up questionnaires, or a search of the National Death Index (18). This study was approved by the institutional review board at the Harvard School of Public Health.

Blood samples were obtained from 18,018 of the participants between 1993 and 1995 and collected in tubes containing sodium EDTA. Samples were shipped by overnight courier and centrifuged; the aliquots, including plasma, erythrocytes, and buffy coat, were stored in liquid nitrogen freezers. We used a QIAamp blood extraction kit (Qiagen, Inc., Valencia, CA) for DNA extraction. All DNA samples were whole genome amplified, and the quality-control samples had 100% genotype concordance rates. Among the men who gave a blood specimen, 95% responded to the 2000 questionnaire and 18 died of prostate cancer before the end of follow-up and were included in the case series.

We identified 700 incident prostate cancer cases and 700 controls (97% were Caucasians). Each case was matched with one control who was alive, had not been diagnosed with cancer by the date of the case's diagnosis, and had a prostate-specific antigen (PSA) test after the date of blood draw. The latter criterion ensured that controls had the opportunity to have an occult prostate cancer diagnosed. All controls had a PSA test within 2.5 years of the date of diagnosis of their matched case. Cases and controls were matched on year of birth (±1 year), PSA test before blood draw (yes/no), time (midnight to before 9 a.m., 9 a.m. to before noon, noon to before 4 p.m., and 4 p.m. to before midnight), season (winter, spring, summer, and fall), and exact year of blood draw, because plasma analyses were being done on the same case-control set.

Laboratory assays. We selected all common SNPs (n = 20) in TLR4 with frequencies greater than 5% from the Innate Immunity in Heart, Lung, and Blood Disease-Programs for Genomic Applications (IIPGA). These SNPs were identified by resequencing the TLR4 gene of 23 unrelated Europeans from Centre du Etude Polymorphisme Humain (CEPH) families. Resequencing of TLR4 included 2.5 kb 5′ of the gene, exons, and 1.5 kb 3′ of the gene. We included a less common nonsynonymous SNP (TLR4_13015) because it causes a change in an amino acid. Laboratory personnel were blind to case-control status. All case-control matched pairs were analyzed together using the Sequenom system. Multiplex PCR were carried out to generate short PCR products (>100 bp) containing one SNP. The details of PCR and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry are available on request. Six control DNA samples were used for optimization. Three SNPs failed the Sequenom assay design due to either high dimer potential for the forward extended primer or other SNPs that might interfere with primer annealing or extension. After genotyping, two other SNPs were removed because of low genotyping success rates (<90%). Finally, a total of 16 SNPs (Fig. 1; Table 1) were genotyped in three plexes at the Harvard Partners Center for Genetics and Genomics (Boston, MA). For each SNP, genotyping data were missing in <5% of the study participants. Sixty-eight quality-control samples were obtained from 18 external participants and each of them had two to six duplicates. These quality-control samples were genotyped together with all other samples in this study. All quality-control samples passed the quality-control test (discordance rate = 0).

Figure 1.

TLR4 gene structure and SNPs. TLR4 has four transcript isoforms (A-D). Information from SNPper-Sequence Viewer at IIPGA Web site (http://snpper.chip.org/bio/show-gene/TLR4) was used to draw this figure. Boxes, exons; gray area, UTR; black area, coding sequence.

Figure 1.

TLR4 gene structure and SNPs. TLR4 has four transcript isoforms (A-D). Information from SNPper-Sequence Viewer at IIPGA Web site (http://snpper.chip.org/bio/show-gene/TLR4) was used to draw this figure. Boxes, exons; gray area, UTR; black area, coding sequence.

Close modal
Table 1.

Characteristics of TLR4 SNPs

IIPGA nameSNP nameNucleotide changeLocationrs no.Minor allele frequency, (controls)HWE P (controls)
TLR4_851 SNP1 A→G 5′ UTR rs2770150 26.3 0.32 
TLR4_1859 SNP2 G→A 5′ UTR rs11536858 38.5 0.25 
TLR4_1893 SNP3 A→G 5′ UTR rs6478317 35.3 0.18 
TLR4_2032 SNP4 T→C 5′ UTR rs10116253 28.5 0.29 
TLR4_2437 SNP5 A→G 5′ UTR rs1927914 35.1 0.15 
TLR4_2856 SNP6 T→C 5′ UTR rs10759932 16.2 0.10 
TLR4_7764 SNP7 G→A Intron rs1927911 29.0 0.44 
TLR4_9263 SNP8 C→A Intron rs11536878 12.3 0.0001 
TLR4_11547 SNP9 A→G Intron rs5030717 12.8 0.09 
TLR4_11912 SNP10 G→T Intron rs2149356 34.1 0.02 
TLR4_13015* SNP11 A→G Exon rs4986790 5.2 0.02 
TLR4_15844* SNP12 G→C 3′ UTR rs11536889 13.8 0.53 
TLR4_16649 SNP13 G→C 3′ UTR rs7873784 17.9 0.02 
TLR4_17050 SNP14 T→C 3′ UTR rs11536891 18.0 0.06 
TLR4_17723 SNP15 G→A 3′ UTR rs11536897 5.4 0.03 
TLR4_17923 SNP16 C→A 3′ UTR rs1536898 14.8 0.26 
IIPGA nameSNP nameNucleotide changeLocationrs no.Minor allele frequency, (controls)HWE P (controls)
TLR4_851 SNP1 A→G 5′ UTR rs2770150 26.3 0.32 
TLR4_1859 SNP2 G→A 5′ UTR rs11536858 38.5 0.25 
TLR4_1893 SNP3 A→G 5′ UTR rs6478317 35.3 0.18 
TLR4_2032 SNP4 T→C 5′ UTR rs10116253 28.5 0.29 
TLR4_2437 SNP5 A→G 5′ UTR rs1927914 35.1 0.15 
TLR4_2856 SNP6 T→C 5′ UTR rs10759932 16.2 0.10 
TLR4_7764 SNP7 G→A Intron rs1927911 29.0 0.44 
TLR4_9263 SNP8 C→A Intron rs11536878 12.3 0.0001 
TLR4_11547 SNP9 A→G Intron rs5030717 12.8 0.09 
TLR4_11912 SNP10 G→T Intron rs2149356 34.1 0.02 
TLR4_13015* SNP11 A→G Exon rs4986790 5.2 0.02 
TLR4_15844* SNP12 G→C 3′ UTR rs11536889 13.8 0.53 
TLR4_16649 SNP13 G→C 3′ UTR rs7873784 17.9 0.02 
TLR4_17050 SNP14 T→C 3′ UTR rs11536891 18.0 0.06 
TLR4_17723 SNP15 G→A 3′ UTR rs11536897 5.4 0.03 
TLR4_17923 SNP16 C→A 3′ UTR rs1536898 14.8 0.26 

NOTE: TLR4_13015 (SNP11) corresponds to the nonsynonymous SNP, 8552A/G, in the report of Zheng et al. (14). This nonsynonymous amino acid substitution, D299G, is in strong linkage disequilibrium with the nonsynonymous change, T399I, which was not selected in this study. TLR4_15844 (SNP12) corresponds to the prostate cancer risk SNP, 11381 G/C, in the report of Zheng et al. (14).

Ascertainment of prostate cancer. Investigators reviewed the medical and pathology records for men with prostate cancer reported from the follow-up questionnaire or death certificate to confirm adenocarcinoma of the prostate and to document clinical presentation, stage, and Gleason sum of the tumor. Because the reporting of a prostate cancer diagnosis by these health professionals tends to be accurate, self-reported cases for which we were unable to acquire medical records (7% of this case series) were included. The clinical presentation was categorized into elevated serum PSA concentration only, abnormal digital rectal examination with or without an elevated serum PSA concentration, or other/unknown. The cases were categorized into regionally invasive or metastatic (stage ≥T3b, N1, or M1), organ-confined or minimal extraprostatic extension (T1b-T3a and N0M0), higher grade (Gleason sum ≥7), and lower grade (Gleason sum <7). Incidental microscopic focal tumors (stage T1a) were excluded because they are generally indolent and susceptible to detection bias due to differential rates of surgery for benign prostatic hyperplasia. In addition, men with a previous cancer, except nonmelanoma skin cancer, before the date of blood draw were excluded. Confirmed non-T1a tumors between blood draw and January 31, 2000 were included. In the blood subcohort, 92% of cases were confirmed by medical record and 5% by other corroborating information; only 3% were based on self-report (19). We included the self-reported cases in the analyses because the concordance between self-report and medical record confirmed cases was high (>90%) in this cohort.

Statistical analysis. The Hardy-Weinberg equilibrium (HWE) test was done for each SNP among controls. Haplotype block structure (Fig. 2) was determined by using Haploview (http://www.broad.mit.edu/mpg/haploview/index.php) and Locusview (http://www.broad.mit.edu/mpg/locusview/). The partition-ligation-expectation-maximization algorithm was applied to estimate haplotype frequencies in each block by using the tagSNP program (20). We calculated haplotype frequencies by using both progressive-ligation (as implemented in SAS PROC HAPLOTYPE) and partition-ligation-expectation-maximization algorithms (21, 22). We arbitrarily broke the 16 SNPs into partitions of three to four SNPs each (e.g., SNP1-4, SNP5-8, etc.). In regions of high linkage disequilibrium and limited haplotype diversity, these algorithms yield quite similar results to each other and to other algorithms (e.g., PHASE; ref. 23). In particular, choice of partition does not noticeably change estimates of haplotype frequency (22). Haplotype frequency estimates in TLR4 using partition-ligation and progressive-ligation differed by ≤0.001. Conditional logistic regression models were used to estimate ORs for disease in participants carrying either 1 or 2 versus 0 copies of the minor allele of each SNP and each multilocus haplotype; haplotype trend regression (24) was used to test global association between TLR4 haplotypes and prostate cancer. The type I error rate is controlled by the single multiple-degree-of-freedom test of association between TLR4 haplotypes and prostate cancer. Given a significant global test, haplotype- and SNP-specific tests can provide some guidance as to which variant(s) contributes to the significant global test, although the nominal Ps we present do not control the family-wise error rate for these post hoc comparisons.

Figure 2.

TLR4 linkage disequilibrium plot. This plot was generated by Haploview and Locusview programs. Using the Gabriel et al. approach, the 15 SNPs formed one block. The rs number (top, from left to right) corresponded to the SNP name (e.g., SNP1, SNP2, etc). The level of pair-wise D′, which indicates the degree of linkage disequilibrium between two SNPs, was shown in the linkage disequilibrium structure in red. Six common haplotypes (frequency > 0.05) were identified.

Figure 2.

TLR4 linkage disequilibrium plot. This plot was generated by Haploview and Locusview programs. Using the Gabriel et al. approach, the 15 SNPs formed one block. The rs number (top, from left to right) corresponded to the SNP name (e.g., SNP1, SNP2, etc). The level of pair-wise D′, which indicates the degree of linkage disequilibrium between two SNPs, was shown in the linkage disequilibrium structure in red. Six common haplotypes (frequency > 0.05) were identified.

Close modal

Age and family history are known risk factors for prostate cancer (25, 26); body mass index (BMI) is related to risk of prostate cancer although not consistently (27). We evaluated how these factors modified the association between TLR4 SNPs or haplotypes and the risk of prostate cancer by comparing a model with terms for main effects and interaction terms to the model with terms for main effects only using the likelihood ratio test. Because of the role of TLR4 in the innate immune response, aggressiveness of prostate cancer may relate to genetic variations of TLR4. We tested the association between TLR4 haplotypes and aggressiveness of prostate cancer among patients with prostate cancer by using two definitions for tumor aggressiveness (aggressiveness 1: stage T3b or T4 or N1 or M1 or death due to prostate cancer; aggressiveness 2: stage T3b or T4 or N1 or M1 or death due to prostate cancer or Gleason sum ≥7). Aggressiveness 1 is useful for participants lacking information for Gleason sum and indicates how far a cancer has progressed independent of grade. Aggressiveness 2 may indicate the potential of the tumor to progress by considering grade information. All analyses were conducted with SAS release 9.0 (SAS Institute, Cary, NC). All statistical tests were two sided.

Sixteen SNPs in TLR4 were genotyped. In the IIPGA data set, the minor allele frequency of SNP11, a nonsynonymous SNP, was slightly less than 5% in the resequencing data of 23 unrelated Europeans from CEPH families but was slightly greater than 5% in our study population. SNP8 was out of HWE among controls (P = 0.0001) and was therefore dropped from all the analyses (Table 1). SNP10, SNP11, SNP13, and SNP15 were out of HWE (P = 0.02-0.03), but we retained them in the analyses as these Ps were marginal and may be chance findings and the internal blinded quality-control specimens did not show evidence of genotyping error.

Only self-reported Caucasians (97% of participants) were included in our analyses, minimizing the likelihood of false-positive findings due to population stratification (28, 29). However, inclusion of non-Caucasians does not change the HWE results among controls.

The study population included 700 incident prostate cancer cases and 700 matched controls. Age and BMI distributions were similar for cases and controls (Table 2). Family history of prostate cancer was significantly different between cases and controls (P = 0.009). The mean age started smoking, lifetime average number of cigarettes per day, and alcohol consumption were similar for cases and controls. Among cases, 80% were in tumor stage T1b to T3a, 73% had Gleason grade 5 to 7, 8% had aggressive prostate cancer (based on aggressiveness 1 definition), and 37% had aggressive prostate cancer (based on aggressiveness 2 definition). Eighteen cases died of prostate cancer before January 31, 2000.

Table 2.

Characteristics of study participants in the HPFS

VariableCases (n = 700), n (%)Controls (n = 700), n (%)
Age   
    ≤65 313 (45) 318 (45) 
    >65 387 (55) 382 (55) 
BMI   
    ≤25 434 (62) 417 (60) 
    25-30 231 (33) 234 (33) 
    >30 35 (5) 49 (7) 
Family history of prostate cancer   
    No 559 (80) 596 (85) 
    Yes 141 (20) 104 (15) 
Age started smoking 23.1 ± 5.2 22.9 ± 5.3 
Lifetime average cigarettes per day 10.1 ± 6.3 10.9 ± 6.7 
Alcohol (g/d) 11.3 ± 14.8 10.4 ± 14.7 
PSA (ng/mL) 11.4 ± 20.9 (n = 498) NA 
Stage   
    T1b-T3a 562 (80) NA 
    T3b or T4 or N1 or M1 or death due to prostate cancer 56 (8)  
    Missing 82 (11)  
Gleason sum   
    2-4 47 (7) NA 
    5-7 512 (73)  
    8-10 59 (8)  
    Missing 82 (12)  
Aggressiveness 1   
    No 557 (80) NA 
    Yes 56 (8)  
    Missing 87 (12)  
Aggressiveness 2   
    No 442 (63) NA 
    Yes 258 (37)  
    Missing  
Death due to prostate cancer   
    No 682 (97) NA 
    Yes 18 (2)  
VariableCases (n = 700), n (%)Controls (n = 700), n (%)
Age   
    ≤65 313 (45) 318 (45) 
    >65 387 (55) 382 (55) 
BMI   
    ≤25 434 (62) 417 (60) 
    25-30 231 (33) 234 (33) 
    >30 35 (5) 49 (7) 
Family history of prostate cancer   
    No 559 (80) 596 (85) 
    Yes 141 (20) 104 (15) 
Age started smoking 23.1 ± 5.2 22.9 ± 5.3 
Lifetime average cigarettes per day 10.1 ± 6.3 10.9 ± 6.7 
Alcohol (g/d) 11.3 ± 14.8 10.4 ± 14.7 
PSA (ng/mL) 11.4 ± 20.9 (n = 498) NA 
Stage   
    T1b-T3a 562 (80) NA 
    T3b or T4 or N1 or M1 or death due to prostate cancer 56 (8)  
    Missing 82 (11)  
Gleason sum   
    2-4 47 (7) NA 
    5-7 512 (73)  
    8-10 59 (8)  
    Missing 82 (12)  
Aggressiveness 1   
    No 557 (80) NA 
    Yes 56 (8)  
    Missing 87 (12)  
Aggressiveness 2   
    No 442 (63) NA 
    Yes 258 (37)  
    Missing  
Death due to prostate cancer   
    No 682 (97) NA 
    Yes 18 (2)  

Men carrying one copy of the minor allele of SNP1 had an increased risk of prostate cancer (OR, 1.38; 95% CI, 1.10-1.73; Table 3). Men carrying one copy of the minor allele of SNP6 and SNP9, on the contrary, had lower risk of prostate cancer (OR, 0.73; 95% CI, 0.57-0.93; OR, 0.66; 95% CI, 0.51-0.86). Men carrying two copies of the minor alleles of eight SNPs were at statistically significantly lower risk of prostate cancer (SNP3: OR, 0.66; 95% CI, 0.46-0.94; SNP4: OR, 0.59; 95% CI, 0.39-0.90; SNP5: OR, 0.64; 95% CI, 0.45-0.93; SNP7: OR, 0.63; 95% CI, 0.41-0.95; SNP10: OR, 0.64; 95% CI, 0.45-0.91; SNP13: OR, 0.51; 95% CI, 0.28-0.96; SNP14: OR, 0.50; 95% CI, 0.27-0.95; SNP16: OR, 0.38; 95% CI, 0.16-0.92).

Table 3.

TLR4 SNP analysis by genotype

SNP0 copies
1 copy
2 copies
P*
Case/controlORCase/controlOR (95% CI)Case/controlOR (95% CI)
SNP1 319/374 1.00 299/254 1.38 (1.10-1.73) 55/52 1.25 (0.83-1.88) 0.02 
SNP2 273/270 1.00 323/315 1.02 (0.81-1.28) 100/110 0.90 (0.65-1.23) 0.73 
SNP3 308/295 1.00 317/297 1.02 (0.82-1.28) 64/93 0.66 (0.46-0.94) 0.04 
SNP4 393/360 1.00 262/272 0.88 (0.71-1.10) 40/62 0.59 (0.39-0.90) 0.04 
SNP5 297/290 1.00 301/288 1.02 (0.81-1.28) 60/91 0.64 (0.45-0.93) 0.04 
SNP6 511/472 1.00 155/197 0.73 (0.57-0.93) 11/12 0.84 (0.37-1.92) 0.04 
SNP7 386/352 1.00 263/275 0.87 (0.70-1.08) 43/62 0.63 (0.41-0.95) 0.06 
SNP9 547/511 1.00 115/162 0.66 (0.51-0.86) 6/6 0.92 (0.30-2.89) 0.01 
SNP10 320/305 1.00 286/275 0.99 (0.79-1.24) 61/91 0.64 (0.45-0.91) 0.04 
SNP11 588/605 1.00 66/59 1.15 (0.80-1.66) 3/5 0.61 (0.15-2.58) 0.60 
SNP12 515/513 1.00 167/159 1.04 (0.81-1.34) 10/15 0.66 (0.29-1.48) 0.54 
SNP13 475/459 1.00 178/180 0.96 (0.75-1.22) 16/30 0.51 (0.28-0.96) 0.10 
SNP14 480/466 1.00 174/186 0.91 (0.71-1.16) 15/29 0.50 (0.27-0.95) 0.08 
SNP15 596/606 1.00 57/62 0.93 (0.64-1.36) 1/5 0.20 (0.02-1.75) 0.23 
SNP16 487/478 1.00 144/158 0.90 (0.69-1.16) 7/18 0.38 (0.16-0.92) 0.06 
SNP0 copies
1 copy
2 copies
P*
Case/controlORCase/controlOR (95% CI)Case/controlOR (95% CI)
SNP1 319/374 1.00 299/254 1.38 (1.10-1.73) 55/52 1.25 (0.83-1.88) 0.02 
SNP2 273/270 1.00 323/315 1.02 (0.81-1.28) 100/110 0.90 (0.65-1.23) 0.73 
SNP3 308/295 1.00 317/297 1.02 (0.82-1.28) 64/93 0.66 (0.46-0.94) 0.04 
SNP4 393/360 1.00 262/272 0.88 (0.71-1.10) 40/62 0.59 (0.39-0.90) 0.04 
SNP5 297/290 1.00 301/288 1.02 (0.81-1.28) 60/91 0.64 (0.45-0.93) 0.04 
SNP6 511/472 1.00 155/197 0.73 (0.57-0.93) 11/12 0.84 (0.37-1.92) 0.04 
SNP7 386/352 1.00 263/275 0.87 (0.70-1.08) 43/62 0.63 (0.41-0.95) 0.06 
SNP9 547/511 1.00 115/162 0.66 (0.51-0.86) 6/6 0.92 (0.30-2.89) 0.01 
SNP10 320/305 1.00 286/275 0.99 (0.79-1.24) 61/91 0.64 (0.45-0.91) 0.04 
SNP11 588/605 1.00 66/59 1.15 (0.80-1.66) 3/5 0.61 (0.15-2.58) 0.60 
SNP12 515/513 1.00 167/159 1.04 (0.81-1.34) 10/15 0.66 (0.29-1.48) 0.54 
SNP13 475/459 1.00 178/180 0.96 (0.75-1.22) 16/30 0.51 (0.28-0.96) 0.10 
SNP14 480/466 1.00 174/186 0.91 (0.71-1.16) 15/29 0.50 (0.27-0.95) 0.08 
SNP15 596/606 1.00 57/62 0.93 (0.64-1.36) 1/5 0.20 (0.02-1.75) 0.23 
SNP16 487/478 1.00 144/158 0.90 (0.69-1.16) 7/18 0.38 (0.16-0.92) 0.06 
*

P tested the null hypothesis: OR1 copy = OR2 copies = 1.

Fifteen SNPs spanning TLR4 form one “block” using a modified version of the Gabriel et al. (30) algorithm, where blocks identified with the default settings in Haploview are merged if they have multiallelic D′ greater than 0.8 and the cumulative frequency of common (>5% frequency) haplotypes in the merged block is >80% (31). Six common haplotypes (frequency >5%) were found with an accumulated frequency of 81% in controls (Table 4). The P for the global test of the six common haplotypes was 0.02. Men carrying one copy of the minor Hap1 had a 1.40-fold increased risk of prostate cancer (95% CI, 1.12-1.75). Hap5, on the other hand, was associated with a lower risk of prostate cancer (OR, 0.59; 95% CI, 0.41-0.86).

Table 4.

ORs between TLR4 haplotypes and the risk of prostate cancer

HaplotypePrevalence among controls, % (95% CI)Global test P = 0.02
P*
0 copies
1 copy
2 copies
Case/controlORCase/controlOR (95% CI)Case/controlOR (95% CI)
Hap1: GGATATGAGAGGTGC 25.2 (22.9-27.5) 340/393 1.00 309/257 1.40 (1.12-1.75) 51/47 1.25 (0.82-1.90) 0.01 
Hap2: AAATATGAGAGGTGC 25.0 (23.1-27.7) 403/387 1.00 250/270 0.89 (0.71-1.11) 47/39 1.15 (0.74-1.80) 0.41 
Hap3: AAATATGAGACGTGC 13.1 (11.3-14.8) 526/528 1.00 164/155 1.07 (0.83-1.37) 9/14 0.64 (0.28-1.50) 0.50 
Hap4: AGGCGCAGTAGGTGC 7.3 (6.0-8.7) 618/597 1.00 79/98 0.78 (0.56-1.07) 3/2 1.46 (0.24-8.77) 0.27 
Hap5: AGGCGCAGTAGCCGA 5.6 (4.5-6.9) 651/619 1.00 49/79 0.59 (0.41-0.86) 0/0 — 0.01 
Hap6: AGGCGTAATAGCCAA 5.2 (4.0-6.3) 638/630 1.00 61/62 0.95 (0.65-1.39) 1/5 0.22 (0.03-1.73) 0.26 
HaplotypePrevalence among controls, % (95% CI)Global test P = 0.02
P*
0 copies
1 copy
2 copies
Case/controlORCase/controlOR (95% CI)Case/controlOR (95% CI)
Hap1: GGATATGAGAGGTGC 25.2 (22.9-27.5) 340/393 1.00 309/257 1.40 (1.12-1.75) 51/47 1.25 (0.82-1.90) 0.01 
Hap2: AAATATGAGAGGTGC 25.0 (23.1-27.7) 403/387 1.00 250/270 0.89 (0.71-1.11) 47/39 1.15 (0.74-1.80) 0.41 
Hap3: AAATATGAGACGTGC 13.1 (11.3-14.8) 526/528 1.00 164/155 1.07 (0.83-1.37) 9/14 0.64 (0.28-1.50) 0.50 
Hap4: AGGCGCAGTAGGTGC 7.3 (6.0-8.7) 618/597 1.00 79/98 0.78 (0.56-1.07) 3/2 1.46 (0.24-8.77) 0.27 
Hap5: AGGCGCAGTAGCCGA 5.6 (4.5-6.9) 651/619 1.00 49/79 0.59 (0.41-0.86) 0/0 — 0.01 
Hap6: AGGCGTAATAGCCAA 5.2 (4.0-6.3) 638/630 1.00 61/62 0.95 (0.65-1.39) 1/5 0.22 (0.03-1.73) 0.26 
*

P tested the null hypothesis: OR1 copy = OR2 copies = 1.

The associations between TLR4 genotypes and the risk of prostate cancer varied by age (Table 5). SNP4, SNP7, SNP13, and SNP16 variant carriers ages <65 years had a statistically significantly lower risk of prostate cancer than did noncarriers (CC and CT versus TT: OR, 0.64; 95% CI, 0.47-0.86; AA and AG versus GG: OR, 0.63; 95% CI, 0.46-0.85; CC and CG versus GG: OR, 0.66; 95% CI, 0.47-0.94; CC and CA versus AA: OR, 0.59; 95% CI, 0.41-0.86, respectively). BMI and family history were not effect modifiers for the association between TLR4 SNPs and prostate cancer. In addition, the association between TLR4 haplotypes and prostate cancer was not modified by age, BMI, or family history.

Table 5.

Effect modification by age

SNPNoncarriers
Carriers
P*
Case/controlORCase/controlOR (95% CI)
SNP1      
    Age ≤65 132/171 1.00 169/139 1.47 (1.09-2.00) 0.22 
    Age >65 187/203 1.00 185/167 1.20 (0.91-1.58)  
SNP2      
    Age ≤65 127/121 1.00 184/194 0.92 (0.68-1.24) 0.48 
    Age >65 146/149 1.00 239/231 1.04 (0.79-1.37)  
SNP3      
    Age ≤65 146/132 1.00 163/180 0.84 (0.63-1.14) 0.27 
    Age >65 162/163 1.00 218/210 1.04 (0.79-1.37)  
SNP4      
    Age ≤65 193/157 1.00 118/157 0.64 (0.47-0.86) 0.01 
    Age >65 200/203 1.00 184/177 1.02 (0.78-1.35)  
SNP5      
    Age ≤65 146/132 1.00 151/173 0.79 (0.59-1.07) 0.18 
    Age >65 151/158 1.00 210/206 1.00 (0.76-1.31)  
SNP6      
    Age ≤65 227/210 1.00 73/100 0.67 (0.47-0.94) 0.53 
    Age >65 284/262 1.00 93/109 0.79 (0.57-1.09)  
SNP7      
    Age ≤65 190/152 1.00 120/161 0.63 (0.46-0.85) 0.01 
    Age >65 198/200 1.00 186/176 1.05 (0.80-1.39)  
SNP9      
    Age ≤65 250/230 1.00 51/79 0.59 (0.40-0.87) 0.41 
    Age >65 299/281 1.00 70/89 0.72 (0.51-1.03)  
SNP10      
    Age ≤65 153/138 1.00 144/170 0.76 (0.56-1.02) 0.17 
    Age >65 167/167 1.00 203/196 1.03 (0.78-1.35)  
SNP11      
    Age ≤65 258/281 1.00 37/24 1.61 (0.94-2.74) 0.04 
    Age >65 330/324 1.00 32/40 0.78 (0.48-1.27)  
SNP12      
    Age ≤65 236/233 1.00 73/80 0.91 (0.63-1.29) 0.40 
    Age ≤65 279/280 1.00 104/94 1.12 (0.81-1.54)  
SNP13      
    Age ≤65 222/205 1.00 73/101 0.66 (0.47-0.94) 0.03 
    Age >65 253/254 1.00 121/109 1.13 (0.83-1.53)  
SNP14      
    Age ≤65 227/208 1.00 73/101 0.66 (0.46-0.93) 0.06 
    Age >65 253/258 1.00 116/114 1.00 (0.73-1.35)  
SNP15      
    Age ≤65 278/273 1.00 21/34 0.60 (0.34-1.06) 0.09 
    Age >65 318/333 1.00 37/33 1.12 (0.69-1.82)  
SNP16      
    Age ≤65 232/211 1.00 56/86 0.59 (0.41-0.86) 0.02 
    Age >65 255/267 1.00 95/90 1.05 (0.76-1.45)  
SNPNoncarriers
Carriers
P*
Case/controlORCase/controlOR (95% CI)
SNP1      
    Age ≤65 132/171 1.00 169/139 1.47 (1.09-2.00) 0.22 
    Age >65 187/203 1.00 185/167 1.20 (0.91-1.58)  
SNP2      
    Age ≤65 127/121 1.00 184/194 0.92 (0.68-1.24) 0.48 
    Age >65 146/149 1.00 239/231 1.04 (0.79-1.37)  
SNP3      
    Age ≤65 146/132 1.00 163/180 0.84 (0.63-1.14) 0.27 
    Age >65 162/163 1.00 218/210 1.04 (0.79-1.37)  
SNP4      
    Age ≤65 193/157 1.00 118/157 0.64 (0.47-0.86) 0.01 
    Age >65 200/203 1.00 184/177 1.02 (0.78-1.35)  
SNP5      
    Age ≤65 146/132 1.00 151/173 0.79 (0.59-1.07) 0.18 
    Age >65 151/158 1.00 210/206 1.00 (0.76-1.31)  
SNP6      
    Age ≤65 227/210 1.00 73/100 0.67 (0.47-0.94) 0.53 
    Age >65 284/262 1.00 93/109 0.79 (0.57-1.09)  
SNP7      
    Age ≤65 190/152 1.00 120/161 0.63 (0.46-0.85) 0.01 
    Age >65 198/200 1.00 186/176 1.05 (0.80-1.39)  
SNP9      
    Age ≤65 250/230 1.00 51/79 0.59 (0.40-0.87) 0.41 
    Age >65 299/281 1.00 70/89 0.72 (0.51-1.03)  
SNP10      
    Age ≤65 153/138 1.00 144/170 0.76 (0.56-1.02) 0.17 
    Age >65 167/167 1.00 203/196 1.03 (0.78-1.35)  
SNP11      
    Age ≤65 258/281 1.00 37/24 1.61 (0.94-2.74) 0.04 
    Age >65 330/324 1.00 32/40 0.78 (0.48-1.27)  
SNP12      
    Age ≤65 236/233 1.00 73/80 0.91 (0.63-1.29) 0.40 
    Age ≤65 279/280 1.00 104/94 1.12 (0.81-1.54)  
SNP13      
    Age ≤65 222/205 1.00 73/101 0.66 (0.47-0.94) 0.03 
    Age >65 253/254 1.00 121/109 1.13 (0.83-1.53)  
SNP14      
    Age ≤65 227/208 1.00 73/101 0.66 (0.46-0.93) 0.06 
    Age >65 253/258 1.00 116/114 1.00 (0.73-1.35)  
SNP15      
    Age ≤65 278/273 1.00 21/34 0.60 (0.34-1.06) 0.09 
    Age >65 318/333 1.00 37/33 1.12 (0.69-1.82)  
SNP16      
    Age ≤65 232/211 1.00 56/86 0.59 (0.41-0.86) 0.02 
    Age >65 255/267 1.00 95/90 1.05 (0.76-1.45)  
*

P is testing the significance of the interaction.

Case-only analysis (aggressive versus nonaggressive) was done among prostate cancer patients. None of the SNPs or haplotypes (Table 6) showed a statistically significant association with tumor aggressiveness.

Table 6.

TLR4 haplotypes and risk of aggressive prostate cancer among cases

HaplotypeNoncarriers
Carriers
P*
Aggressive/nonaggressiveORAggressive/nonaggressiveOR (95% CI)
Aggressiveness 1 (Global test P = 0.86)      
    Hap1: GGATATGAGAGGTGC 28/274 1.00 28/283 1.17 (0.81-1.69) 0.41 
    Hap2: AAATATGAGAGGTGC 35/334 1.00 21/243 0.96 (0.66-1.39) 0.83 
    Hap3: AAATATGAGACGTGC 38/426 1.00 18/131 0.98 (0.64-1.50) 0.91 
    Hap4: AGGCGCAGTAGGTGC 48/495 1.00 8/62 1.03 (0.58-1.84) 0.92 
    Hap5: AGGCGCAGTAGCCGA 53/520 1.00 3/37 1.58 (0.79-3.18) 0.20 
    Hap6: AGGCGTAATAGCCAA 54/501 1.00 2/56 0.83 (0.43-1.62) 0.58 
Aggressiveness 2 (Global test P = 0.70)      
    Hap1: GGATATGAGAGGTGC 128/213 1.00 130/229 1.06 (0.78-1.44) 0.73 
    Hap2: AAATATGAGAGGTGC 145/258 1.00 113/184 0.92 (0.68-1.26) 0.61 
    Hap3: AAATATGAGACGTGC 185/341 1.00 73/101 0.75 (0.52-1.06) 0.11 
    Hap4: AGGCGCAGTAGGTGC 226/392 1.00 32/50 0.91 (0.56-1.47) 0.69 
    Hap5: AGGCGCAGTAGCCGA 239/412 1.00 19/30 0.91 (0.50-1.65) 0.75 
    Hap6: AGGCGTAATAGCCAA 236/402 1.00 22/40 1.05 (0.60-1.84) 0.85 
HaplotypeNoncarriers
Carriers
P*
Aggressive/nonaggressiveORAggressive/nonaggressiveOR (95% CI)
Aggressiveness 1 (Global test P = 0.86)      
    Hap1: GGATATGAGAGGTGC 28/274 1.00 28/283 1.17 (0.81-1.69) 0.41 
    Hap2: AAATATGAGAGGTGC 35/334 1.00 21/243 0.96 (0.66-1.39) 0.83 
    Hap3: AAATATGAGACGTGC 38/426 1.00 18/131 0.98 (0.64-1.50) 0.91 
    Hap4: AGGCGCAGTAGGTGC 48/495 1.00 8/62 1.03 (0.58-1.84) 0.92 
    Hap5: AGGCGCAGTAGCCGA 53/520 1.00 3/37 1.58 (0.79-3.18) 0.20 
    Hap6: AGGCGTAATAGCCAA 54/501 1.00 2/56 0.83 (0.43-1.62) 0.58 
Aggressiveness 2 (Global test P = 0.70)      
    Hap1: GGATATGAGAGGTGC 128/213 1.00 130/229 1.06 (0.78-1.44) 0.73 
    Hap2: AAATATGAGAGGTGC 145/258 1.00 113/184 0.92 (0.68-1.26) 0.61 
    Hap3: AAATATGAGACGTGC 185/341 1.00 73/101 0.75 (0.52-1.06) 0.11 
    Hap4: AGGCGCAGTAGGTGC 226/392 1.00 32/50 0.91 (0.56-1.47) 0.69 
    Hap5: AGGCGCAGTAGCCGA 239/412 1.00 19/30 0.91 (0.50-1.65) 0.75 
    Hap6: AGGCGTAATAGCCAA 236/402 1.00 22/40 1.05 (0.60-1.84) 0.85 
*

P tested the null hypothesis: ORcarriers = 1.

TLR4 is located on the surface of the macrophage and plays an important role in signaling the innate immune response in response to bacterial or viral infection. TLR4 is the pathogen recognition receptor for most Gram-negative bacteria (LPS), many Gram-positive bacteria, mycobacteria (2), spirochetes, viruses, and other ligands (4). Chronic intraprostatic inflammation may increase risk of prostate cancer. Variation in TLR4 may reduce or even block the signaling of the immune response and therefore lower the risk of prostate cancer.

Animal studies showed that TLR4-deficient mice exposed to respiratory syncytial virus had higher levels of infectious virus in their lungs and either were unable to clear the virus or cleared it several days later than wild-type mice (6). For C3H/HeJ mice, the codominant lpsd allele corresponded to a missense mutation in the third exon of the TLR4 gene, resulting in substitution of proline with histidine at codon 712. For C57B/10ScCr mice, a null mutation of TLR4 was observed among mice homozygous for the second lps mutation. These mutations rendered the C3H/HeJ and C57B/10ScCr mice refractory to the toxic effects of lps from Gram-negative bacteria (7, 8). Experimental studies revealed that mutations in TLR4 were related to hyporesponsiveness to viral and bacterial infection and therefore to a reduction in the innate immune response and inflammation.

A previous study (14) in a Swedish population explored the association between genetic variation of TLR4 and the risk of prostate cancer. The authors found that one SNP (11381G/C, equivalent to our SNP12) was associated with the risk of prostate cancer (CC and CG versus GG: OR, 1.26; 95% CI, 1.01-1.57); this SNP was not associated with risk in our study. In addition, the authors did not observe any association between TLR4 haplotypes based on eight SNPs and prostate cancer. The disparity between the Zheng et al. study and our study might be due to differences in the study populations, differences in case detection by PSA, or extent of genotyping. We genotyped more SNPs in TLR4, which provided a more comprehensive assessment of TLR4 genetic variation than the previous study, which was based on genotyping eight common SNPs.

Age modified the association between TLR4 SNPs and prostate cancer, but we observed no effect modification by BMI and family history. Among men ages <65 years, carriers of SNP4, SNP7, SNP13, or SNP16 had a statistically significantly lower risk of prostate cancer. This indicated that TLR4 genetic polymorphisms had a greater influence among younger than older men and is consistent with a greater influence of inherited genetic risk among younger men.

TLR4 haplotypes and risk of aggressive prostate cancer among cases (Table 6) did not differ using either of the two aggressiveness definitions used in this study or the definition of aggressiveness used by Zheng et al. (14).

Men carrying one copy of variant Hap1 had a 1.40-fold increased risk of prostate cancer. In Hap1, SNP1 is the only SNP carrying the variant allele. Men carrying one copy of the minor allele of SNP1 had a 1.38-fold increased risk of prostate cancer, although the OR for homozygous variant was not significantly elevated. Hap5 was associated with a lower risk of prostate cancer, which after comparison with hap4 and hap6 was found to be mainly attributable to variants of SNP13, SNP14, and SNP16. These SNPs located in the 3′ UTR may be associated with lower risk of prostate cancer because they may influence mRNA stability and thereby reduce the innate immune response, inflammation, and subsequent carcinogenesis. SNP1, located in the 5′ UTR, where the promoter or transcription factor binding sites are located, may potentially exert regulator effects and therefore increase cancer risk. Functional assays are needed to elucidate the molecular mechanisms underlying these associations.

Four of the SNPs genotyped were significantly out of HWE in the controls but not in the cases. Because of the large number of loci tested, a conservative significance threshold (α = 0.001) is conventionally used when testing HWE to detect genotyping errors (32). None of these four loci are significant at the Bonferroni-corrected significance level of 0.003 (0.05/15). Even with the less conservative Benjamini-Hochberg step-up procedure to control the false discovery rate (<5%), the null hypothesis of HWE is rejected only for SNP8. Considering none of the SNPs that showed marginal evidence for deviation from HWE in controls (P < 0.05) showed any evidence of deviation in cases (P > 0.10), we do not believe that the marginal HWE tests in controls suggest systematic genotyping errors for these SNPs.

Furthermore, haplotype trend regression conditions on observed genotypes and hence is relatively robust to moderate departures from HWE in haplotype frequencies. We have shown (33)4

4

P. Kraft, unpublished simulation studies.

that the haplotype trend regression is closely related to the prospective likelihood (34, 35); the latter approach has been shown to yield accurate tests and parameter estimates when HWE does not hold (specifically when there is an excess of homozygotes, as is the case for these five SNPs; ref. 36). We also did haplotype analyses excluding the four SNPs showing marginal evidence of departures from HWE. Haplotype frequency and OR estimates for the 11-SNP haplotypes corresponding to the previous 15-SNP haplotypes (e.g., GGATATGA**G*T*G and GGATATGAGAGGTGC) were quite similar (differing by at most 0.005 and 0.03, respectively). The two haplotypes corresponding to the significant haplotypes in the previous analysis were also significant at the 0.05 level.

In recent years, chronic inflammation due to the innate immune response to infection has been suspected to be a risk factor for cancer at many sites (37, 38). However, research on inflammation and prostate cancer is limited. Our findings suggest that inflammation may be associated with risk of prostate cancer, and future investigation of genetic variation in innate immune genes may cast light on the etiology of this enigmatic disease.

Grant support: NIH grants UO1 CA98233 and CA55075.

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.

We thank Pati Soule and Ana-Tereza Andrade for DNA sample extraction and the Partners High-Throughput Genotyping Center (Dr. David Kwiatkowski, Alison Brown, and Maura Regan) for genotyping.

1
Medzhitov R, Preston-Hurlburt P, Janeway CA, Jr. A human homologue of the Drosophila Toll protein signals activation of adaptive immunity.
Nature
1997
;
388
:
394
–7.
2
Aderem A, Ulevitch RJ. Toll-like receptors in the induction of the innate immune response.
Nature
2000
;
406
:
782
–7.
3
Chaudhary PM, Ferguson C, Nguyen V, et al. Cloning and characterization of two Toll/interleukin-1 receptor-like genes TIL3 and TIL4: evidence for a multi-gene receptor family in humans.
Blood
1998
;
91
:
4020
–7.
4
Opal SM, Esmon CT. Bench-to-bedside review: functional relationships between coagulation and the innate immune response and their respective roles in the pathogenesis of sepsis.
Crit Care
2003
;
7
:
23
–38.
5
Muzio M, Bosisio D, Polentarutti N, et al. Differential expression and regulation of toll-like receptors (TLR) in human leukocytes: selective expression of TLR3 in dendritic cells.
J Immunol
2000
;
164
:
5998
–6004.
6
Kurt-Jones EA, Popova L, Kwinn L, et al. Pattern recognition receptors TLR4 and CD14 mediate response to respiratory syncytial virus.
Nat Immunol
2000
;
1
:
398
–401.
7
Poltorak A, He X, Smirnova I, et al. Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene.
Science
1998
;
282
:
2085
–8.
8
Qureshi ST, Lariviere L, Leveque G, et al. Endotoxin-tolerant mice have mutations in Toll-like receptor 4 (Tlr4).
J Exp Med
1999
;
189
:
615
–25.
9
Vink A, de Kleijn DP, Pasterkamp G. Functional role for toll-like receptors in atherosclerosis and arterial remodeling.
Curr Opin Lipidol
2004
;
15
:
515
–21.
10
Michelsen KS, Doherty TM, Shah PK, Arditi M. TLR signaling: an emerging bridge from innate immunity to atherogenesis.
J Immunol
2004
;
173
:
5901
–7.
11
Abreu MT, Arditi M. Innate immunity and toll-like receptors: clinical implications of basic science research.
J Pediatr
2004
;
144
:
421
–9.
12
Mogensen TH, Paludan SR. Reading the viral signature by Toll-like receptors and other pattern recognition receptors.
J Mol Med
2005
;
83
:
180
–92.
13
Imahara SD, O'Keefe GE. Genetic determinants of the inflammatory response.
Curr Opin Crit Care
2004
;
10
:
318
–24.
14
Zheng SL, Augustsson-Balter K, Chang B, et al. Sequence variants of Toll-like receptor 4 are associated with prostate cancer risk: results from the Cancer Prostate in Sweden Study.
Cancer Res
2004
;
64
:
2918
–22.
15
Sun J, Wiklund F, Zheng SL, et al. Sequence variants in Toll-like receptor gene cluster (TLR6-TLR1-TLR10) and prostate cancer risk.
J Natl Cancer Inst
2005
;
97
:
525
–32.
16
Kinlen L. Infections and immune factors in cancer: the role of epidemiology.
Oncogene
2004
;
23
:
6341
–8.
17
Platz EA, De Marzo AM. Epidemiology of inflammation and prostate cancer.
J Urol
2004
;
171
:
S36
–40.
18
Stampfer MJ, Willett WC, Speizer FE, et al. Test of the National Death Index.
Am J Epidemiol
1984
;
119
:
837
–9.
19
Platz EA, Leitzmann MF, Rifai N, et al. Sex steroid hormones and the androgen receptor gene CAG repeat and subsequent risk of prostate cancer in the prostate-specific antigen era.
Cancer Epidemiol Biomarkers Prev
2005
;
14
:
1262
–9.
20
Stram DO, Leigh Pearce C, Bretsky P, et al. Modeling and E-M estimation of haplotype-specific relative risks from genotype data for a case-control study of unrelated individuals.
Hum Hered
2003
;
55
:
179
–90.
21
Clayton D. SNPHAP: a program for estimating frequencies of large haplotypes of SNPs; Genet Epidemiol 2002.
22
Qin Z, Niu T, Liu J. Partition-ligation-expectation-maximization algorithm for haplotype inference with single-nucleotide polymorphisms.
Am J Hum Genet
2002
;
70
:
157
.
23
Niu T. Algorithms for inferring haplotypes.
Genet Epidemiol
2004
;
27
:
334
–47.
24
Zaykin DV, Westfall PH, Young SS, Karnoub MA, Wagner MJ, Ehm MG. Testing association of statistically inferred haplotypes with discrete and continuous traits in samples of unrelated individuals.
Hum Hered
2002
;
53
:
79
–91.
25
Kalish LA, McDougal WS, McKinlay JB. Family history and the risk of prostate cancer.
Urology
2000
;
56
:
803
–6.
26
Cerhan JR, Parker AS, Putnam SD, et al. Family history and prostate cancer risk in a population-based cohort of Iowa men.
Cancer Epidemiol Biomarkers Prev
1999
;
8
:
53
–60.
27
Giovannucci E, Rimm EB, Liu Y, et al. Body mass index and risk of prostate cancer in U.S. health professionals.
J Natl Cancer Inst
2003
;
95
:
1240
–4.
28
Wacholder S, Rothman N, Caporaso N. Population stratification in epidemiologic studies of common genetic variants and cancer: quantification of bias.
J Natl Cancer Inst
2000
;
92
:
1151
–8.
29
Wacholder S, Rothman N, Caporaso N. Counterpoint: bias from population stratification is not a major threat to the validity of conclusions from epidemiological studies of common polymorphisms and cancer.
Cancer Epidemiol Biomarkers Prev
2002
;
11
:
513
–20.
30
Gabriel SB, Schaffner SF, Nguyen H, et al. The structure of haplotype blocks in the human genome.
Science
2002
;
296
:
2225
–9.
31
Florez JC, Burtt N, de Bakker PI, et al. Haplotype structure and genotype-phenotype correlations of the sulfonylurea receptor and the islet ATP-sensitive potassium channel gene region.
Diabetes
2004
;
53
:
1360
–8.
32
Weir BS, Hill WG, Cardon LR. Allelic association patterns for a dense SNP map.
Genet Epidemiol
2004
;
27
:
442
–50.
33
Kraft P, Cox DG, Paynter RA, Hunter D, De Vivo I. Accounting for haplotype uncertainty in matched association studies: a comparison of simple and flexible techniques.
Genet Epidemiol
2005
;
28
:
261
–72.
34
Lake S, Lyon H, Tantisira K, et al. Estimation and tests of haplotype-environment interaction when linkage phase is ambiguous.
Hum Hered
2003
;
55
:
56
–65.
35
Zhao L, Li S, Khalid N. A method for the assessment of disease associations with single-nucleotide polymorphism haplotypes and environmental variables in case-control studies.
Am J Hum Genet
2003
;
72
:
1231
–50.
36
Satten GA, Epstein MP. Comparison of prospective and retrospective methods for haplotype inference in case-control studies.
Genet Epidemiol
2004
;
27
:
192
–201.
37
Konig JE, Senge T, Allhoff EP, Konig W. Analysis of the inflammatory network in benign prostate hyperplasia and prostate cancer.
Prostate
2004
;
58
:
121
–9.
38
Kinoshita I, Dosaka-Akita H, Shindoh M, et al. Human papillomavirus type 18 DNA and E6-E7 mRNA are detected in squamous cell carcinoma and adenocarcinoma of the lung.
Br J Cancer
1995
;
71
:
344
–9.