Identification of candidate genes for prostate cancer-risk SNPs utilizing a normal prostate tissue eQTL data set.

Multiple studies have identified loci associated with the risk of developing prostate cancer but the associated genes are not well studied. Here we create a normal prostate tissue-specific eQTL data set and apply this data set to previously identified prostate cancer (PrCa)-risk SNPs in an effort to identify candidate target genes. The eQTL data set is constructed by the genotyping and RNA sequencing of 471 samples. We focus on 146 PrCa-risk SNPs, including all SNPs in linkage disequilibrium with each risk SNP, resulting in 100 unique risk intervals. We analyse cis-acting associations where the transcript is located within 2 Mb (±1 Mb) of the risk SNP interval. Of all SNP-gene combinations tested, 41.7% of SNPs demonstrate a significant eQTL signal after adjustment for sample histology and 14 expression principal component covariates. Of the 100 PrCa-risk intervals, 51 have a significant eQTL signal and these are associated with 88 genes. This study provides a rich resource to study biological mechanisms underlying genetic risk to PrCa.

Experimental designs for array comparative genomic hybridization technology.

Array comparative genomic hybridization (aCGH) technology is commonly used to estimate genome-wide copy-number variation and to evaluate associations between copy number and disease. Although aCGH technology is well developed and there are numerous algorithms available for estimating copy number, little attention has been paid to the important issue of the statistical experimental design. Herein, we review classical statistical experimental designs and discuss their relevance to aCGH technology as well as their importance for downstream statistical analyses. Furthermore, we provide experimental design guidance for various study objectives.

Validity of family history data on PD: evidence for a family information bias.

OBJECTIVE: To study the validity of information provided by case and control subjects (or their proxies) about PD among their first-degree relatives. METHODS: Secondary cases of PD were assessed both through a single informant (family history method) and through the study of each relative (family study method). The family study method was considered as the standard for comparison, and the sensitivity and specificity of the family history method were studied. RESULTS: A total of 133 population-based case subjects and their 655 relatives were recruited, and 119 population-based control subjects and their 511 relatives. Sensitivity was 68% (95% CI = 47 to 85) for cases and 45% (95% CI = 17 to 77) for controls. Specificity was 99% (95% CI = 98 to 99) for cases and 100% (95% CI = 99 to 100) for controls. The odds ratio (OR) for family history of PD was 4.34 (95% CI = 1.63 to 11.58, p = 0.003) using the family history method and 1.86 (95% CI = 0.78 to 4.44, p = 0.16) using the family study method. The former significant OR more than doubled the latter not significant OR (relative bias = 133%). Bias was more pronounced for proxy interviews and for women informants, and when the relatives were siblings, were living, and were examined or had medical record documentation. CONCLUSIONS: Case subjects with PD (or their proxies) are more aware of PD among their first-degree relatives than control subjects (or their proxies); however, they overreport PD in relatives who are not affected. This causes a substantial family information bias.

Head trauma preceding PD: a case-control study.

OBJECTIVE: To investigate the association of PD with preceding head trauma using a case-control study design. METHODS: The medical records-linkage system of the Rochester Epidemiology Project was used to identify 196 subjects who developed PD in Olmsted County, MN, from 1976 through 1995. Each incident case was matched by age (+/-1 year) and sex to a general population control. The complete medical records of cases and controls in the system were reviewed to detect preceding episodes of head trauma. RESULTS: The frequency of head trauma overall was significantly higher in cases than in controls (odds ratio [OR] = 4.3; 95% CI = 1.2 to 15.2). Compared with subjects who never experienced a trauma, subjects who experienced a mild head trauma with only amnesia had no increased risk; however, subjects who experienced a mild head trauma with loss of consciousness or a more severe trauma had an OR of 11.0 (95% CI = 1.4 to 85.2). Although not significant, head trauma resulting in hospitalization was more frequent in cases than in control subjects (OR = 8.0; 95% CI = 1.0 to 64.0). Whereas the OR was higher for men than women and for patients with later onset of PD than for patients with earlier onset, these differences were not significant. CONCLUSIONS: These results suggest an association between head trauma and the later development of PD that varies with severity. Although the OR is high (4.3), the population attributable risk is only 5% because head trauma is a relatively rare event.

Confirmation of linkage of prostate cancer aggressiveness with chromosome 19q.

Regions on chromosomes 7 and 19 were recently reported to contain susceptibility loci that regulate tumor aggressiveness of prostate cancer. To confirm these findings, we analyzed genome scan data from 161 pedigrees affected with prostate cancer. Using the Gleason score as a quantitative measure of tumor aggressiveness, we regressed the squared trait difference, as well as the mean-corrected cross product, on the estimated proportion of alleles shared identical-by-descent at each marker position. Our results confirm the previous linkage results for chromosome 19q (D19S902, P<.00001). In addition, we report suggestive evidence for linkage on chromosome 4 (D4S403, P=.00012). The results of previous findings, together with our results, provide strong evidence that chromosome 19 harbors a gene for tumor aggressiveness.

Hysterectomy, menopause, and estrogen use preceding Parkinson's disease: an exploratory case-control study.

We studied the association of Parkinson's disease (PD) with type of menopause (natural or surgical), age at menopause, and postmenopausal estrogen replacement therapy using a case-control design. We used the medical records-linkage system of the Rochester Epidemiology Project to identify 72 women who developed PD in Olmsted County, MN, during the twenty years 1976-1995. Each incident case was matched by age (+/- 1 year) to a general population control subject. We collected exposure data through review of the complete medical records of cases and control subjects in the system. PD cases had undergone hysterectomy (with or without unilateral oophorectomy) significantly more often than control subjects (odds ratio [OR] = 3.36; 95% confidence interval [CI] = 1.05-10.77). In addition, PD cases had experienced early menopause (< or = 46 years) more commonly than control subjects (OR = 2.18; 95% CI = 0.88-5.39). Finally, PD cases had used estrogens orally or parenterally for at least 6 months after menopause less frequently (8%) than control subjects (14%; OR = 0.47; 95% CI = 0.12-1.85). However, the findings for early menopause and estrogen replacement therapy were not statistically significant. Despite the limited sample size of this exploratory study, we hypothesize that there is an increased risk of PD in conditions causing an early reduction in endogenous estrogen. This hypothesis needs to be confirmed in a larger study.

Case-Control study of the extended tau gene haplotype in Parkinson's disease.

We investigated the association of Parkinson's disease with tau gene haplotypes. In a sample of 319 unrelated Parkinson's disease patients and 196 control subjects, we observed an increased risk of Parkinson's disease for persons with the H1/H1 genotype (odds ratio = 1.5; 95% confidence interval: 0.98-2.23); however, the finding was not statistically significant. The results remained similar after adjusting for the possible misclassification of progressive supranuclear palsy patients as Parkinson's disease, but became statistically significant after restricting the analysis to nondemented subjects.

Efficacy of bilateral prophylactic mastectomy in BRCA1 and BRCA2 gene mutation carriers.

BACKGROUND: In women with a family history of breast cancer, bilateral prophylactic mastectomy is associated with a decreased risk of subsequent breast cancer of approximately 90%. We examined the association between bilateral prophylactic mastectomy and breast cancer risk in women at high risk for breast cancer who also had mutations in BRCA1 and BRCA2 genes. METHODS: We obtained blood samples from 176 of the 214 high-risk women who participated in our previous retrospective cohort study of bilateral prophylactic mastectomy. We used conformation-sensitive gel electrophoresis and direct sequence analysis of the blood specimens to identify women with mutations in BRCA1 and BRCA2. The carriers' probabilities of developing breast cancer were estimated from two different penetrance models. RESULTS: We identified 26 women with an alteration in BRCA1 or BRCA2. Eighteen of the mutations were considered to be deleterious and eight to be of uncertain clinical significance. None of the 26 women has developed breast cancer after a median of 13.4 years of follow-up (range, 5.8-28.5 years). Three of the 214 women are known to have developed a breast cancer after prophylactic mastectomy. For two of these women, BRCA1 and BRCA2 screening was negative, and no blood specimen was available for the third. Estimations of the effectiveness of prophylactic mastectomy were performed, considering this woman as both a mutation carrier and a noncarrier. These calculations predicted that six to nine breast cancers should have developed among the mutation carriers, which translates into a risk reduction, after bilateral prophylactic mastectomy, of 89.5%-100% (95% confidence interval = 41.4% to 100%). CONCLUSIONS: Prophylactic mastectomy is associated with a substantial reduction in the incidence of subsequent breast cancer not only in women identified as being at high risk on the basis of a family history of breast cancer but also in known BRCA1 or BRCA2 mutation carriers.

Efficacy of contralateral prophylactic mastectomy in women with a personal and family history of breast cancer.

PURPOSE: To estimate the efficacy of contralateral prophylactic mastectomy in women with a personal and family history of breast cancer. PATIENTS AND METHODS: We followed the course of 745 women with a first breast cancer and a family history of breast and/or ovarian cancer who underwent contralateral prophylactic mastectomy at the Mayo Clinic between 1960 and 1993. Family history information and cancer follow-up information were obtained from the medical record, a study-specific questionnaire, and telephone follow-up. Life-tables for contralateral breast cancers, which consider age at first breast cancer, current age, and type of family history, were used to calculate the number of breast cancers expected in our cohort had they not had a prophylactic mastectomy. RESULTS: Of the 745 women in our cohort, 388 were premenopausal (age < 50 years) and 357 were post- menopausal. Eight women developed a contralateral breast cancer. Six events were observed among the premenopausal women, compared with 106.2 predicted, resulting in a risk reduction of 94.4% (95% confidence interval [CI], 87.7% to 97.9%). For the 357 postmenopausal women, 50.3 contralateral breast cancers were predicted, whereas only two were observed, representing a 96.0% risk reduction (95% CI, 85.6% to 99.5%). CONCLUSION: The incidence of contralateral breast cancer seems to be reduced significantly after contralateral prophylactic mastectomy in women with a personal and family history of breast cancer.

Time trends in the incidence of parkinsonism in Olmsted County, Minnesota.

OBJECTIVE: To investigate time trends in the incidence of parkinsonism and PD over a 15-year period (1976 to 1990). METHODS: The authors used the medical records-linkage system of the Rochester Epidemiology Project to identify incidence cases of parkinsonism in Olmsted County, MN, over three 5-year periods, 1976 to 1980, 1981 to 1985, and 1986 to 1990. PD and other types of parkinsonism were classified using defined criteria. Population denominators were derived from census data and were corrected by removing prevalent cases of parkinsonism. RESULTS: Over the 15 years of the study, 364 cases of parkinsonism were identified; 154 (42%) of them had PD. The incidence of parkinsonism remained stable over the three 5-year periods for the age classes 0 to 39, 40 to 59, and 60 to 69 years. For the age class 70 to 99 years, there was some increase over time mainly owing to an increased incidence of drug-induced parkinsonism. The incidence of PD remained stable over the three 5-year periods for all age classes. Results were similar when considering men and women separately. No birth-cohort effect was present for parkinsonism. Comparison with three previous studies in the same population did not reveal any major long-term secular trends in the incidence of parkinsonism. CONCLUSIONS: The findings for PD over 15 years and comparison of the findings with historical data for parkinsonism over half a century suggest that no major environmental risk factors for PD (e.g., environmental toxins, drugs, diet constituents, or infectious agents) were introduced or removed from this population during these periods.

The frequency of hereditary defective mismatch repair in a prospective series of unselected colorectal carcinomas.

A comprehensive analysis of somatic and germline mutations related to DNA mismatch-repair (MMR) genes can clarify the prevalence and mechanism of inactivation in colorectal carcinoma (CRC). In the present study, 257 unselected patients referred for CRC resection were examined for evidence of defective DNA MMR. In particular, we sought to determine the frequency of hereditary defects in DNA MMR in this cohort of patients. MMR status was assessed by testing of tumors for the presence or absence of hMLH1, hMSH2, and hMSH6 protein expression and for microsatellite instability (MSI). Of the 257 patients, 51 (20%) had evidence of defective MMR, demonstrating high levels of MSI (MSI-H) and an absence of either hMLH1 (n=48) or hMSH2 (n=3). All three patients lacking hMSH2, as well as one patient lacking hMLH1, also demonstrated an absence of hMSH6. DNA sequence analysis of the 51 patients with defective MMR revealed seven germline mutations-four in hMLH1 (two truncating and two missense) and three in hMSH2 (all truncating). A detailed family history was available for 225 of the 257 patients. Of the seven patients with germline mutations, only three had family histories consistent with hereditary nonpolyposis colorectal cancer. Of the remaining patients who had tumors with defective MMR, eight had somatic mutations in hMLH1. In addition, hypermethylation of the hMLH1 gene promoter was present in 37 (88%) of the 42 hMLH1-negative cases available for study and in all MSI-H tumors that showed loss of hMLH1 expression but no detectable hMLH1 mutations. Our results suggest that, although defective DNA MMR occurs in approximately 20% of unselected patients presenting for CRC resection, hereditary CRC due to mutations in the MMR pathway account for only a small proportion of patients. Of the 257 patients, only 5 (1.9%) appear to have unequivocal evidence of hereditary defects in MMR. The epigenetic (nonhereditary) mechanism of hMLH1 promoter hypermethylation appears to be responsible for the majority of the remaining patients whose tumors are characterized by defective DNA MMR.

Role of HPC2/ELAC2 in hereditary prostate cancer.

The HPC2/ELAC2 gene on chromosome 17p was recently identified as a candidate gene for hereditary prostate cancer (HPC). To confirm these findings, we screened 300 prostate cancer patients (2 affected members/family) from 150 families with HPC for potential germ-line mutations using conformation-sensitive gel electrophoresis, followed by direct sequence analysis. The minimum criteria for our families with HPC was the presence of 3 affected men with prostate cancer. A total of 23 variants were identified, including 13 intronic and 10 exonic changes. Of the 10 exonic changes, 1 truncating mutation was identified, a Glu216Stop nonsense mutation. This nonsense variant was found in 2 of 3 affected men in a single family. The remaining nine alterations included five missense, three silent, and one variant in the 3' untranslated region. To additionally test for potential associations of polymorphic variants and increased risk for disease, we genotyped two common polymorphisms, Ser217Leu and Ala541Thr, in 446 prostate cancer patients from 164 families with HPC and 502 population-based controls. The frequency of the Leu217 variant was similar for patients (32.3%) and controls (31.8%), as was the frequency of the Thr541 variant (5.4% among patients versus 5.2% among controls). In contrast to previous reports, we found no association of the joint effects of Leu271 and Thr541 (odds ratio, 1.04; 95% confidence interval, 0.57-1.89). Overall, our results did not reveal any association between these two common polymorphisms and the risk for HPC. The finding of a nonsense mutation in the HPC2/ELAC2 gene confirms its potential role in genetic susceptibility to prostate cancer. However, our data also suggest that germ-line mutations of the HPC2/ELAC2 are rare in HPC and that the variants Leu217 and Thr541 do not appear to influence the risk for HPC. Cumulatively, these results suggest that alterations within the HPC2/ELAC2 gene play a limited role in genetic susceptibility to HPC.

Regression models for linkage heterogeneity applied to familial prostate cancer.

Linkage heterogeneity frequently occurs for complex genetic diseases, and statistical methods must account for it to avoid severe loss in power to discover susceptibility genes. A common method to allow for only a fraction of linked pedigrees is to fit a mixture likelihood and then to test for linkage homogeneity, given linkage (admixture test), or to test for linkage while allowing for heterogeneity, using the heterogeneity LOD (HLOD) score. Furthermore, features of the families, such as mean age at diagnosis, may help to discriminate families that demonstrate linkage from those that do not. Pedigree features are often used to create homogeneous subsets, and LOD or HLOD scores are then computed within the subsets. However, this practice introduces several problems, including reduced power (which results from multiple testing and small sample sizes within subsets) and difficulty in interpretation of results. To address some of these limitations, we present a regression-based extension of the mixture likelihood for which pedigree features are used as covariates that determine the probability that a family is the linked type. Some advantages of this approach are that multiple covariates can be used (including quantitative covariates), covariates can be adjusted for each other, and interactions among covariates can be assessed. This new regression method is applied to linkage data for familial prostate cancer and provides new insights into the understanding of prostate cancer linkage heterogeneity.

Analysis of the prostate cancer-susceptibility locus HPC20 in 172 families affected by prostate cancer.

A recent study of hereditary prostate cancer has provided evidence for a prostate cancer-susceptibility locus, HPC20, which maps to 20q13. To assess further the potential contribution of this locus to prostate cancer susceptibility, we studied 172 unrelated families affected by prostate cancer, using 17 polymorphic markers across a 98.5-cM segment of chromosome 20 that contains the candidate region. Parametric analysis, allowing for heterogeneity, resulted in an overall HLOD score of 0.09 (P=.39) at D20S171, under the assumption of linkage in 6% of families. Mode-of-inheritance-free analysis of the entire data set resulted in a maximal Zlr score of 0.76 (LOD 0.13; P=.22) at the same location. The strongest evidence for linkage was seen in the subset of 16 black families (LOD 0.86; Zlr=1.99; P=.023) between markers D20S893 and D20S120, near the putative location of HPC20. Although some positive results were observed, our linkage study does not provide statistically significant support for the existence of a prostate cancer-susceptibility locus HPC20 at 20q13.

Smoking, alcohol, and coffee consumption preceding Parkinson's disease: a case-control study.

OBJECTIVE: To study the association of PD with preceding smoking, alcohol, and coffee consumption using a case-control design. METHODS: The authors used the medical records linkage system of the Rochester Epidemiology Project to identify 196 subjects who developed PD in Olmsted County, MN, during the years 1976 to 1995. Each incident case was matched by age (+/-1 year) and sex to a general population control subject. The authors reviewed the complete medical records of cases and control subjects to abstract exposure information. RESULTS: For coffee consumption, the authors found an OR of 0.35 (95% CI = 0.16 to 0.78, p = 0.01), a dose-effect trend (p = 0.003), and a later age at PD onset in cases who drank coffee compared with those who never did (median 72 versus 64 years; p = 0.0002). The inverse association with coffee remained significant after adjustment for education, smoking, and alcohol drinking and was restricted to PD cases with onset at age <72 years and to men. The OR for cigarette smoking was 0.69 (95% CI = 0.45 to 1.08, p = 0.1). The authors found no association between PD and alcohol consumption. Extreme or unusual behaviors such as tobacco chewing or snuff use and a diagnosis of alcoholism were significantly more common in control subjects than cases. CONCLUSIONS: These findings suggest an inverse association between coffee drinking and PD; however, this association does not imply that coffee has a direct protective effect against PD. Alternative explanations for the association should be considered.

Association of Peutz-Jeghers-like mucocutaneous pigmentation with breast and gynecologic carcinomas in women.

Most reports describe an increased risk of malignancy in Peutz-Jeghers syndrome (PJS). We identified individuals with PJS-like pigmentation but no polyposis, designated as isolated mucocutaneous melanotic pigmentation (IMMP), and 1) characterized their clinical features, 2) assessed them for cancer events, and 3) screened a sample of these subjects for mutations in LKB1, a gene responsible for a portion of PJS cases. Review of Mayo Clinic records from 1945 to 1996 identified 26 patients with IMMP. All were then interviewed or their medical records reviewed to determine if cancer had developed. Conformation-sensitive gel electrophoresis (CSGE) screening for LKB1 mutations was followed by direct sequencing. Ten of these 26 individuals (38%) developed 12 malignancies that arose in the cervix (n = 3), endometrium (n = 3), breast (n = 1), kidney (n = 1), lung (n = 2), colon (n = 1), and lymphatic tissue (n = 1). In females with IMMP, the relative risk for cancer was 3.2 (95% CI, 1.2-6.9), while that for males was not increased. The relative risk for breast and gynecologic cancers was 7.8 (95% CI, 2.5-18.1) in affected females. Of 9 individuals tested, no LKB1 mutations were detected. Classical PJS is associated with an increased cancer risk. Our results indicate that IMMP is another lentiginosis with cancer predisposition. In particular, the relative risk for cancer in females with IMMP was significantly increased, as is true in females with PJS. However, LKB1 mutations did not contribute to the development of IMMP in the patients tested.

Influence of strict, intermediate, and broad diagnostic criteria on the age- and sex-specific incidence of Parkinson's disease.

We studied the influence of three sets of diagnostic criteria on the age- and sex-specific incidence of Parkinson's disease (PD) among residents of Olmsted County, Minnesota, for the period 1976 to 1990. Incidence cases of parkinsonism were detected using the medical records-linkage system of the Rochester Epidemiology Project. PD was separated from other types of parkinsonism using strict, intermediate, and broad criteria. We found 154 incident cases of PD using the strict criteria, 215 using the intermediate criteria, and 266 using the broad criteria. The incidence rate was consistently higher for men across all ages with all three sets of criteria; however, sex differences were more striking at older ages when using the broad criteria. In men above age 79 years, the incidence rate of PD declined with strict criteria, remained stable with intermediate criteria, and increased with broad criteria. The impact of diagnostic criteria on the age-specific incidence curve was less striking for women. When using the broad criteria, the risk of PD increased constantly with age in both sexes, suggesting that PD is an aging-related disease. Our findings suggest that the diagnostic criteria used to separate PD from other types of parkinsonism influence the magnitude of PD incidence and its distribution by age and sex.

Anxiety disorders and depressive disorders preceding Parkinson's disease: a case-control study.

We studied the association between preceding psychiatric disorders and Parkinson's disease (PD) using a case-control design. We used the medical records-linkage system of the Rochester Epidemiology Project to identify 196 subjects who developed PD in Olmsted County, Minnesota, during the years 1976-1995. Each case was matched by age (+/-1 yr) and sex to a general population control. We reviewed the complete medical records of cases and control subjects to detect preceding psychiatric disorders. The frequency of psychiatric disorders was higher in cases than in control subjects; the odds ratio was 2.2 for anxiety disorders (95% confidence interval [95% CI] = 1.4-3.4; p = 0.0003), 1.9 for depressive disorders (95% CI = 1.1-3.2; p = 0.02), and 2.4 for both anxiety disorders and depressive disorders occurring in the same individual (95% CI = 1.2-4.8; p = 0.02). When we restricted analyses to disorders present 5 years or more before the onset of motor symptoms of PD, the association with depressive disorders lost statistical significance. However, the association with anxiety disorders remained significant for disorders present 5, 10, or 20 years before onset of motor symptoms. Our results suggest that anxiety disorders and depressive disorders are associated with PD and that the causative process or the risk factors underlying PD are present many years before the appearance of motor symptoms.

Case-control study of debrisoquine 4-hydroxylase, N-acetyltransferase 2, and apolipoprotein E gene polymorphisms in Parkinson's disease.

We investigated the association of Parkinson's disease (PD) with two genes encoding liver-detoxifying enzymes, debrisoquine 4-hydroxylase (CYP2D6) and N-acetyltransferase 2 (NAT2), and with one gene related to Alzheimer's disease, apolipoprotein E (APOE). In a sample of 139 unrelated PD cases and 113 control subjects, the NAT2 M3 allele was associated with PD (odds ratio = 7.9; 95% confidence interval = 1.7-36.3). Case-control analyses for CYP2D6, APOE, and NAT2 M1 or M2 did not show a significant association. However, the age at onset of PD was significantly earlier in cases with the APOE epsilon2/epsilon3 genotype than in cases with the epsilon3/epsilon3 genotype.