The importance of replication in gene-gene interaction studies: multifactor dimensionality reduction applied to a two-stage breast cancer case-control study

A polygenic model has been proposed to explain the bulk of the genetic component of breast cancer aetiology and this is probably to include both main effects and interactions between multiple loci. However, the power to detect the interactions using traditional analytical methods is very limited for most studies. Multifactor dimensionality reduction (MDR) has been suggested to have increased power to detect interactions and is increasing being used in published studies. We applied MDR to a two-stage case-control breast cancer study conducted in Spain and Finland. In the stage 1 Spanish study of 864 cases and 845 controls, we evaluated interaction between 474 single-nucleotide polymorphisms in 120 cancer-related genes, subdivided into 34 genetic pathways and found evidence of a four-way interaction between genes in the FatiGO-defined B-cell receptor-signalling pathway (P < 0.006). However, this result was not replicated in the stage 2 Finnish study of 580 cases and 920 controls (P = 0.99). A number of technical issues in applying MDR to case-control data were identified and discussed. One of these is that the estimated sign test P-value can vary substantially at random, which raises doubts about its reliability. More generally, the present study serves as an important caution in the interpretation of results from single studies of gene-gene or gene-environment interaction in complex diseases. Just as for genetic main effects, the replication of positive findings in additional independent series is essential.     

Evaluating bias due to population stratification in epidemiologic studies of gene-gene or gene-environment interactions.

Confounding by ethnicity (i.e. population stratification) can result in bias and incorrect inferences in genotype-disease association studies, but the effect of population stratification in gene-gene or gene-environment interaction studies has not been addressed. We used logistic regression models to fit multiplicative interactions between two dichotomous variables that represented genetic and/or environmental factors for a binary disease outcome in a hypothetical cohort of multiple ethnicities. Biases in main effects and interactions due to population stratification were evaluated by comparing regression coefficients in mis-specified models that ignored ethnicities with their counterparts in models that accounted for ethnicities. We showed that biases in main effects and interactions were constrained by the differences in disease risks across the ethnicities. Therefore, large biases due to population stratification are not possible when baseline disease risk differences among ethnicities are small or moderate. Numerical examples of biases in genotype-genotype and/or genotype-environment interactions suggested that biases due to population stratification for main effects were generally small but could become large for studies of interactions, particularly when strong linkage disequilibrium between genes or large correlations between genetic and environmental factors existed. However, when linkage disequilibrium among genes or correlations among genes and environments were small, biases to main effects or interaction odds ratios were small to nonexistent.     

Genetic susceptibility to breast cancer in French-Canadians: role of carcinogen-metabolizing enzymes and gene-environment interactions

Breast cancer is the most frequent malignancy among women. Since genetic factors such as BRCA1 and BRCA2 as well as reproductive history constitute only 30% of the cause, environmental exposure may play a significant role in the development of breast cancer. Likewise, the relevant enzymes involved in the biotransformation of xenobiotics (from tobacco smoke, diet or other environmental sources) might play a role in breast carcinogenesis. Since individuals with modified ability to metabolize these carcinogens could have a different risk for breast cancer, we investigated the role of cytochromes P-450 (CYP1A1, CYP2D6), glutathione-S-transferases (GSTM1, GSTT1, GSTP1) and N-acetyltransferases (NAT1, NAT2) gene variants in breast carcinogenesis. A case-control study was conducted on 149 women with breast carcinoma and 207 healthy controls, both of French-Canadian origin. The CYP1A1*4 allele was found to be a significant risk determinant of breast carcinoma (OR = 3.3, 95% CI 1.1-9.7), particularly among post-menopausal women (OR = 4.0, 95% CI 1.2-13.8). The frequency of NAT2 rapid acetylators was increased among smokers (OR = 2.6, 95% CI 0.8-8.2), while the NAT1*10 allele conferred a 4-fold increase in risk among women who consumed well-done meat (OR = 4.4, 95% CI 1.0-18.9). These data suggest that CYP1A1*4, NAT1 and NAT2 variants are involved in the susceptibility to breast carcinoma by modifying the impact of exogenous and/or endogenous exposures.     

Fruit and vegetable intake and esophageal cancer in a large prospective cohort study

Changing patterns of esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC) incidence worldwide suggest distinct etiologies. Although associations between fruit and vegetable intake and both ESCC and EAC have been found in multiple ecological and case-control studies, few prospective studies have investigated these associations. We prospectively examined these associations in 490,802 participants of the National Institutes of Health (NIH)-AARP Diet and Health Study using Cox models adjusted for age, alcohol intake, body mass index, cigarette smoking, education, physical activity and total energy intake. We present hazard ratios and 95% confidence intervals per serving per 1,000 calories. During 2,193,751 person years of follow-up, 103 participants were diagnosed with ESCC and 213 participants with EAC. We found a significant inverse association between total fruit and vegetable intake and ESCC risk (HR: 0.78, 95% CI: 0.67-0.91), but not EAC risk (0.98, 0.90-1.08). In models mutually adjusted for fruit and vegetable intake, the protective association with ESCC was stronger for fruits (0.73, 0.57-0.93) than for vegetables (0.84, 0.66-1.07). When we examined botanical subgroups, we observed significant protective associations for ESCC and intake of Rosacea (apples, peaches, nectarines, plums, pears and strawberries) and Rutaceae (citrus fruits). A significant inverse association between EAC and Chenopodiaceae (spinach) intake was observed. Results from our study suggest that the relation of fruit and vegetable intake and esophageal cancer risk may vary by histologic type.     

Urinary isothiocyanate levels and lung cancer risk among non-smoking women: a prospective investigation

Background

Aside from tobacco carcinogen metabolism, isothiocyanates (ITC) from cruciferous vegetables may induce apoptosis or steroid metabolism to reduce lung cancer risk. To separate the effect of these divergent mechanisms of action, we investigated the association between urinary ITC levels and lung cancer risk among non-smoking women.

Methods

We conducted a nested case-control within the Shanghai Women's Health Study. Subjects included 209 incident lung cancer cases who never used tobacco, and 787 individually matched non-smoking controls. Conditional logistic regression was used to calculate odds ratios (OR) and 95% confidence intervals (CI) summarizing the association between urinary ITC levels and lung cancer. Secondary analyses stratified the ITC-lung cancer analyses by menopausal status, exposure to environmental tobacco smoke, and GSTM1 and GSTT1 genotypes.

Results

Urinary ITC levels were not significantly associated with lower lung cancer risk among non-smoking women, regardless of exposure to environmental tobacco smoke or menopausal status. Furthermore, this association was not modified by GSTT1 genotype. However, an inverse association was suggested among women with a GSTM1-positive genotype (Q1: OR = 1.0 (reference); Q2: OR = 0.35 (0.14, 0.89); Q3: OR = 0.47 (0.20, 1.10); Q4: OR = 0.63 (0.35, 1.54), p-trend = 0.38). In contrast, lung cancer risk was positively associated with urinary ITC levels among women with the GSTM1-null genotype (Q1: OR = 1.0 (reference); Q2: OR = 1.67 (0.80, 3.50); Q3: OR = 1.54 (0.71, 3.33); Q4: OR = 2.22 (1.05, 4.67), p-trend = 0.06).

Conclusion

Urinary ITC levels were not associated overall with lower lung cancer risk among non-smoking women, but secondary analyses suggested an interaction between urinary ITC levels, GSTM1 genotype, and lung cancer risk.     

Serum uric acid and risk of cancer mortality in a large prospective male cohort

Objective To examine the prognostic role of serum uric acid (SUA) for cancer mortality in apparently healthy men across a wide age range. Methods Prospective data from a large cohort of 83,683 male Austrian adults with a median follow-up of 13.6 years was analyzed. Cox proportional hazards models, adjusted for established risk factors, were calculated to evaluate SUA as a predictive marker for fatal cancer events. Results High SUA (>6.71 mg/dl) was independently associated with increased risk of mortality from all cancers, showing a clear dose–response relationship (p for trend < 0.0001); the adjusted hazard ratio for the highest versus lowest quintile of SUA was 1.41 (1.22–1.62). In subgroup analyses this hazard ratio increased to 1.53 (1.29–1.80) for participants aged <65 years. When considering the time interval between baseline SUA measurement and subsequent death, SUA levels were more predictive for “late deaths”, occurring 10 or more years after screening (HR 1.65 [1.35–2.03], p < 0.0001), in comparison to deaths within 10 years after SUA measurement. In cancer site-specific analyses, SUA was significantly associated with deaths from malignant neoplasms of digestive organs (p = 0.03) and respiratory system and intrathoracic organs (p < 0.0001). Elevated SUA was further independently related to an increased risk of all-cause mortality (p < 0.0001). Conclusions Our results are contrary to the proposed antioxidant, inhibitory effect of SUA against cancer and rather suggest high SUA to be a valuable long-term surrogate parameter, indicative for a life-style at increased risk for the development of cancer.     

Age and attachment style impact stress and depressive symptoms among caregivers: a prospective investigation

Introduction Family members begin their role as caregivers to persons with cancer with little advance notice. In this situation, the caregivers’ existing psychosocial resources, including their stage in life and the nature of their relationship with the patient, can play important roles in the extent of stress caregivers experience during this unique time. Materials and methods Family caregivers (N = 98) of diagnosed colorectal cancer patients at community hospitals participated in the study around the time of diagnosis (T1) and at 6 months post-diagnosis (T2). Results Hierarchical regression analyses revealed that younger caregivers reported greater increases in caregiving stress at T2, controlling for the level of caregiving stress at T1. This was more prominent when they had an anxious attachment orientation to the care recipient (i.e., cancer survivor), which was characterized as a strong desire for closeness to the cancer survivor, hypervigilance for cues of abandonment, and emotional upset both at separation and reunion with the survivor. The same interaction effect between age and anxious attachment was found for depressive symptoms. Conclusions Results suggest that younger caregivers are more vulnerable to the challenges imposed by their relative’s cancer, and this vulnerability is exacerbated by the quality of their relationship. Implications for cancer survivors Our findings imply that younger caregivers whose relationship with the survivor can be characterized by the features above can be identified early and might benefit from intervention or additional support to reduce the caregiving stress and depressive symptoms as they carry out their new role as a cancer caregiver.     

Cervical HPV infection and neoplasia in a large population-based prospective study: the Manchester cohort

Cytology and histology records and cervical samples for HPV assay were obtained from a prospective cohort of 49 655 women attending clinics for routine cervical cytology in or near Manchester between 1988 and 1993. The women were followed up for cytological abnormality and neoplasia through the cytology laboratory's records. HPV at entry was assayed in an age- and period-stratified random sample of 7278 women and in prevalent and incident CIN3 cases. The prevalence of newly diagnosed CIN3 increased with time since last normal smear, indicating that most cases persist for several years. CIN3 prevalence did not increase further for screening intervals exceeding 5 years, however, suggesting that CIN3 eventually regresses cytologically. CIN2 prevalence increased less steeply with screening interval, while the prevalence of lesser abnormality was almost independent of screening interval. The prevalence of oncogenic HPV at entry declined from 19% among women aged under 25 to less than 3% at age 40 or above. Oncogenic HPV infection was strongly predictive of subsequent CIN3 (OR 17.2, 95% CI 10.4-28.4), but only weakly related to CIN2 (OR 2.3, 95% CI 0.5-10.7) and lesser abnormality (OR 1.4, 95% CI 0.8-2.5). At current incidence rates, the lifetime risk of developing CIN3 will be 9% in this population. The cumulative risk of CIN3 diagnosis among cytologically normal women with oncogenic HPV detected at entry was 28% (CI 18-43%) after 14 years. Persistence of oncogenic HPV may be more sensitive and specific than cytology for early detection of CIN3 and invasive cancer.     

Fruit and vegetable intake and gastric cancer risk in a large United States prospective cohort study

OBJECTIVE: Fruit and vegetable intake may protect against gastric cancer incidence. Results from case-control studies have indicated an inverse association, but results from cohort studies are inconsistent. METHODS: We prospectively investigated the association in 490,802 participants of the NIH-AARP Diet and Health Study using Cox proportional hazards models adjusted for gastric cancer risk factors. We present hazard ratios (HR) and 95% confidence intervals (CI) per increase of one daily serving per 1,000 calories. RESULTS: During 2,193,751 person years, 394 participants were diagnosed with incident gastric cancer. We observed no significant associations between total fruit and vegetable intake (1.01, 0.95-1.08), fruit intake (1.04, 0.95-1.14), or vegetable intake (0.98, 0.88-1.08) and gastric cancer risk. Results did not vary by sex or anatomic subsite (cardia versus non-cardia). All 13 botanical subgroups examined had no significant associations with either anatomic sub-site. CONCLUSION: We did not observe significant associations between overall fruit and vegetable intake and gastric cancer risk in this large prospective cohort study.     

Cigarette smoking and the risk of incident and fatal melanoma in a large prospective cohort study

Objective  

Previous studies suggest that smoking may be inversely associated with risk of melanoma. We attempted to replicate this finding using data from the Cancer Prevention Study II (CPS-II) and CPS-II Nutrition cohort, two large prospective cohort studies of cancer mortality and incidence, respectively, with long-term follow-up.     

Covalent binding of dibenzpyrenes and benzo[a]Pyrene to DNA: evidence for synergistic and inhibitory interactions when applied in combination to mouse skin

Several well-documented examples of human exposure to car cinogensinvolve complex mixtures of polycydlic aromatic hydrocarbons(PAHs). Although the biological properties of many pure PANshave been Investigated, less is known about their effects whenpresent as components of mixtures. As the ability to form DNAadducts in vivo Is generally indicative of carcinogenic activityof PAHs, we have compared the DNA binding potencles of dibenzo[a,e]pyrene(DB[a,e]P), dibenzo[a,h]pyrene (DB[a,h]P), dibenzo[a,i]pyrene(DB[a,i]P),dibenzo[a,l]pyrene (DB[a,l]P) and benzo [a]pyrene(B[a]P), when applied topically, either singly or in combination,to the skin of male Parkes mice. DNA isolated from the skinand lungs was analysed by ³²P-postlabelling. The adducts formedby each PAM exhibited markedly different chromato graphic mobifitieson polyethylenelmine-cellulose TLC plates. The relative bindingpotendies of the compounds in both skin and lungs were: DB[a,l]P>B[a]PDB > DB[a,h]P > DB[a,i]P > DB[a,e]P, in good agreementwith their reported carcinogenicitlesin mouse skin. The majorityof adducts were removed from DNA within 21 days of treat ment,but low levels of adducts were found to persist for at least3 months in both tissues. When DB[a,l]P, DB[a,e]P and B[a]Pwere applied together to mouse skin, a total binding 31% lowerthan expected was detected, while with a mixture of DB[a,e]Pand B[a]P the binding to DNA In skin was 65% higher than expectedfrom the binding Levels of the carcino gens when applied singly.Other binary combinations of these three PAils gave adduct levelssimilar to the sum of the binding levels of the individual componentswhen applied singly. The results demonstrate the usefulnessof 32 labelling for the assessment of the DNA binding potenciesof PAHs in mouse tissues, and for the detection of interactionsbetween components of mixtures of carcinogens.     

Statin use and colorectal cancer risk according to molecular subtypes in two large prospective cohort studies

Use of statins is hypothesized to reduce colorectal cancer risk but the evidence remains inconsistent. This may be partly explained by differential associations according to tumor location or molecular subtypes of colorectal cancer. We examined the association between statin use and colorectal cancer risk according to tumor location, KRAS mutation status, microsatellite instability (MSI) status, PTGS2 (COX-2) expression, or CpG island methylator phenotype (CIMP) status in two large prospective cohort studies, the Nurses' Health Study and Health Professionals Follow-up Study. We applied Cox regression to a competing risks analysis. We identified 1,818 colorectal cancers during 1990 to 2006. Compared with nonusers, current statin use was not associated with colorectal cancer [relative risk (RR) = 0.99, 95% CI = 0.86-1.14] or colon cancer (RR = 1.10, 95% CI = 0.94-1.29) but was inversely associated with rectal cancer (RR = 0.59, 95% CI = 0.41-0.84, P(heterogeneity) < 0.001). When we examined the association within strata of KRAS mutation status, we found no association with KRAS-mutated cancers (RR = 1.20, 95% CI = 0.87-1.67) but did observe a possible inverse association among KRAS wild-type cancers (RR = 0.80, 95% CI = 0.60-1.06, P(heterogeneity) = 0.06). The association did not substantially differ by PTGS2 expression, MSI status, or CIMP status. Current statin use was not associated with risk of overall colorectal cancer. The possibility that statin use may be associated with lower risk of rectal cancer or KRAS wild-type colorectal cancer requires further confirmation.     

A prospective study on habitual duration of sleep and incidence of breast cancer in a large cohort of women

Mounting evidence suggests habitual sleep duration is associated with various health outcomes; both short and long sleep duration have been implicated in increased risk of cardiovascular disease, diabetes, and all-cause mortality. However, data on the relation between sleep duration and cancer risk are sparse and inconclusive. A link between low levels of melatonin, a hormone closely related to sleep, and increased risk of breast cancer has recently been suggested but it is unclear whether duration of sleep may affect breast cancer risk. We explored the association between habitual sleep duration reported in 1986 and subsequent risk of breast cancer in the Nurses' Health Study using Cox proportional hazards models. During 16 years of follow-up, 4,223 incident cases of breast cancer occurred among 77,418 women in this cohort. Compared with women sleeping 7 hours, covariate-adjusted hazard ratios and 95% confidence intervals for those sleeping < or =5, 6, 8, and > or =9 hours were 0.93 (0.79-1.09), 0.98 (0.91-1.06), 1.05 (0.97-1.13), and 0.95 (0.82-1.11), respectively. A moderate trend in risk increase towards longer sleep duration was observed when analyses were restricted to participants who reported same sleep duration in 1986 and 2000 (P(trend) = 0.05). In this prospective study, we found no convincing evidence for an association between sleep duration and the incidence of breast cancer.     

Neuromarkers of fatigue and cognitive complaints following chemotherapy for breast cancer: a prospective fMRI investigation

The aim of this study is to use functional magnetic resonance imaging (fMRI) to prospectively examine pre-treatment predictors of post-treatment fatigue and cognitive dysfunction in women treated with adjuvant chemotherapy for breast cancer. Fatigue and cognitive dysfunction often co-occur in women treated for breast cancer. We hypothesized that pre-treatment factors, unrelated to chemotherapy per se, might increase vulnerability to post-treatment fatigue and cognitive dysfunction. Patients treated with (n = 28) or without chemotherapy (n = 37) and healthy controls (n = 32) were scanned coincident with pre- and one-month post-chemotherapy during a verbal working memory task (VWMT) and assessed for fatigue, worry, and cognitive dysfunction. fMRI activity measures in the frontoparietal executive network were used in multiple linear regression to predict post-treatment fatigue and cognitive function. The chemotherapy group reported greater pre-treatment fatigue than controls and showed compromised neural response, characterized by higher spatial variance in executive network activity, than the non-chemotherapy group. Also, the chemotherapy group reported greater post-treatment fatigue than the other groups. Linear regression indicated that pre-treatment spatial variance in executive network activation predicted post-treatment fatigue severity and cognitive complaints, while treatment group, age, hemoglobin, worry, and mean executive network activity levels did not predict these outcomes. Pre-treatment neural inefficiency (indexed by high spatial variance) in the executive network, which supports attention and working memory, was a better predictor of post-treatment cognitive and fatigue complaints than exposure to chemotherapy per se. This executive network compromise could be a pre-treatment neuromarker of risk, indicating patients most likely to benefit from early intervention for fatigue and cognitive dysfunction.     

Body mass index, physical activity, and bladder cancer in a large prospective study.

Increased body size and lack of physical activity are associated with increased risk of several cancers, but the relations of body mass index (BMI) and physical activity to bladder cancer are poorly understood. We investigated the associations between BMI, physical activity, and bladder cancer in the NIH-AARP Diet and Health Study, a prospective cohort of 471,760 U.S. men and women, followed from 1995 to 2003. During 3,404,642 person-years of follow-up, we documented 1,719 incident cases of bladder cancer. Compared with normal weight, obesity was associated with an up to 28% increased risk for bladder cancer. The multivariate relative risks of bladder cancer for BMI values of 18.5 to 24.9 (reference), 25.0 to 29.9, 30.0 to 34.9, and >or=35 kg/m2 were 1.0, 1.15, 1.22, and 1.28 (95% confidence interval, 1.02-1.61; P trend = 0.028). The association between BMI and bladder cancer was consistent among subgroups defined by gender, education, smoking status, and other potential effect modifiers. In contrast, physical activity showed no statistically significant relation with bladder cancer. After multivariate adjustment, including BMI, the relative risks of bladder cancer for increasing frequency of physical activity [0 (reference), <1, 1-2, 3-4, and >or=5 times a week] were 1.0, 0.85, 0.89, 0.91, and 0.87 (95% confidence interval, 0.74-1.02; P trend = 0.358), respectively. In conclusion, these findings provide support for a modest adverse effect of adiposity on risk for bladder cancer. In contrast, our results do not suggest a relation between physical activity and bladder cancer.     

Patterns of meat intake and risk of prostate cancer among African-Americans in a large prospective study

Objective  

Given the large racial differences in prostate cancer risk, further investigation of diet and prostate cancer is warranted among high-risk groups. The purpose of this study was to examine the association between type of meat intake and prostate cancer risk among African-American men.     

Couples who get closer after breast cancer: frequency and predictors in a prospective investigation.

PURPOSE: Although some couples report an improved relationship since coping with breast cancer together, little quantitative information exists about this phenomenon. We assessed extent to which both couple members report that breast cancer brought them closer and characteristics that predicted this. PATIENTS AND METHODS: This prospective study was based on all women with newly diagnosed nonmetastatic disease first treated during recruitment in four Quebec hospitals, in addition to their spouses. Participation was 87% among eligible patients and 91% among spouses of participating patients. Both couple partners were interviewed individually about quality of life at 2 weeks and 3 and 12 months after treatment start. At 12 months, each was asked whether the disease had brought them closer, distanced them, or had no effect. RESULTS: Overall, 42% of the 282 couples said breast cancer brought them closer, 6% had one or other partner reporting feeling distanced, and less than 1% of couples had both partners reporting feeling distanced. Characteristics assessed explained 31% of variance in the proportion of couples getting closer (P < .0001). After taking into account partners' prediagnosis characteristics and the woman's treatment, the spouse reporting the patient as confidant (P = .003), getting advice from her in the first 2 weeks about coping with breast cancer (P = .03), accompanying her to surgery (P = .057), the patient's reporting more affection from her spouse at 3 months since diagnosis (P = .003) predicted both partners saying the disease brought them closer. CONCLUSION: Breast cancer can be a growth experience for couples under certain conditions. This information may help reassure patients and their spouses confronting this disease.