• Home
  • News
  • Calendar
  • About DF/HCC
  • Membership
  • Visitor Center
 

Member Resources

Publications

Cancer Epidemiology Biomarkers & Prevention

Cancer Epidemiology Biomarkers & Prevention RSS feed -- current issue
Cancer Epidemiology Biomarkers & Prevention

We describe the "tumor-based case–control" study as a type of epidemiologic study used to evaluate associations between infectious agents and cancer. These studies assess exposure using diseased tissues from affected individuals (i.e., evaluating tumor tissue for cancer cases), but they must utilize nondiseased tissues to assess control subjects, who do not have the disease of interest. This approach can lead to exposure misclassification in two ways. First, concerning the "when" of exposure assessment, retrospective assessment of tissues may not accurately measure exposure at the key earlier time point (i.e., during the etiologic window). Second, concerning the "where" of exposure assessment, use of different tissues in cases and controls can have different accuracy for detecting the exposure (i.e., differential exposure misclassification). We present an example concerning the association of human papillomavirus with various cancers, where tumor-based case–control studies likely overestimate risk associated with infection. In another example, we illustrate how tumor-based case–control studies of Helicobacter pylori and gastric cancer underestimate risk. Tumor-based case–control studies can demonstrate infection within tumor cells, providing qualitative information about disease etiology. However, measures of association calculated in tumor-based case–control studies are prone to over- or underestimating the relationship between infections and subsequent cancer risk. Cancer Epidemiol Biomarkers Prev; 23(10); 1959–64. ©2014 AACR.


The call for multilevel interventions to improve the quality of follow-up to abnormal cancer screening has been out for a decade, but published work emphasizes individual approaches, and conceptualizations differ regarding the definition of levels. To investigate the scope and methods being undertaken in this focused area of follow-up to abnormal tests (breast, colon, cervical), we reviewed recent literature and grants (2007–2012) funded by the National Cancer Institute. A structured search yielded 16 grants with varying definitions of "follow-up" (e.g., completion of recommended tests, time to diagnosis); most included minority racial/ethnic group participants. Ten grants concentrated on measurement/intervention development and 13 piloted or tested interventions (categories not mutually exclusive). All studies considered patient-level factors and effects. Although some directed interventions at provider levels, few measured group characteristics and effects of interventions on the providers or levels other than the patient. Multilevel interventions are being proposed, but clarity about endpoints, definition of levels, and measures is needed. The differences in the conceptualization of levels and factors that affect practice need empirical exploration, and we need to measure their salient characteristics to advance our understanding of how context affects cancer care delivery in a changing practice and policy environment. Cancer Epidemiol Biomarkers Prev; 23(10); 1965–73. ©2014 AACR.


Diabetes may be a risk factor for cancer and is associated with worse cancer outcomes. Metformin may reduce cancer risk; however, its effect on mortality following cancer remains less clear. EMBASE and Medline were searched through February 10, 2014, for studies reporting an adjusted risk estimate for the effect of metformin therapy on mortality following cancer among diabetic patients. Random-effects models were used to obtain summary HR for the association between metformin and all-cause and cancer-specific mortality. Twenty-one observational studies were meta-analyzed in the primary analysis. Metformin was associated with a reduction in all-cause mortality [HR, 0.73; 95% confidence intervals (CI), 0.64–0.83] and cancer-specific mortality (HR, 0.74; 95% CI, 0.62–0.88). Subgroup analyses by cancer site showed a significant reduction in mortality for colon cancer (four studies, HR, 0.65; 95% CI, 0.56–0.76) but not for breast and prostate cancers. Observational studies indicate that metformin exposure at cancer diagnosis may be associated with a reduction in mortality. However, these findings need to be interpreted with caution as methodologic limitations of individual studies may have introduced biases in these findings. Our results emphasize the need for well-designed studies to further understand the relationship between metformin and survival following cancer. Cancer Epidemiol Biomarkers Prev; 23(10); 1974–84. ©2014 AACR.


Glioblastoma multiforme (GBM) is the most common and aggressive primary central nervous system malignancy with a median survival of 15 months. The average incidence rate of GBM is 3.19/100,000 population, and the median age of diagnosis is 64 years. Incidence is higher in men and individuals of white race and non-Hispanic ethnicity. Many genetic and environmental factors have been studied in GBM, but the majority are sporadic, and no risk factor accounting for a large proportion of GBMs has been identified. However, several favorable clinical prognostic factors are identified, including younger age at diagnosis, cerebellar location, high performance status, and maximal tumor resection. GBMs comprise of primary and secondary subtypes, which evolve through different genetic pathways, affect patients at different ages, and have differences in outcomes. We report the current epidemiology of GBM with new data from the Central Brain Tumor Registry of the United States 2006 to 2010 as well as demonstrate and discuss trends in incidence and survival. We also provide a concise review on molecular markers in GBM that have helped distinguish biologically similar subtypes of GBM and have prognostic and predictive value. Cancer Epidemiol Biomarkers Prev; 23(10); 1985–96. ©2014 AACR.


Background: We estimated the prevalence and incidence of 14 human papillomavirus (HPV) types (6/11/16/18/31/33/35/39/45/51/52/56/58/59) in cervicovaginal swabs, and the attribution of these HPV types in cervical intraepithelial neoplasia (CIN), and adenocarcinoma in situ (AIS), using predefined algorithms that adjusted for multiple-type infected lesions.

Methods: A total of 10,656 women ages 15 to 26 years and 1,858 women ages 24 to 45 years were enrolled in the placebo arms of one of three clinical trials of a quadrivalent HPV vaccine. We estimated the cumulative incidence of persistent infection and the proportion of CIN/AIS attributable to individual carcinogenic HPV genotypes, as well as the proportion of CIN/AIS lesions potentially preventable by a prophylactic 9-valent HPV6/11/16/18/31/33/45/52/58 vaccine.

Results: The cumulative incidence of persistent infection with ≥1 of the seven high-risk types included in the 9-valent vaccine was 29%, 12%, and 6% for women ages 15 to 26, 24 to 34, and 35 to 45 years, respectively. A total of 2,507 lesions were diagnosed as CIN or AIS by an expert pathology panel. After adjusting for multiple-type infected lesions, among women ages 15 to 45 years, these seven high-risk types were attributed to 43% to 55% of CIN1, 70% to 78% of CIN2, 85% to 91% of CIN3, and 95% to 100% of AIS lesions, respectively. The other tested types (HPV35/39/51/56/59) were attributed to 23% to 30% of CIN1, 7% to 14% of CIN2, 3% to 4% of CIN3, and 0% of AIS lesions, respectively.

Conclusions: Approximately 85% or more of CIN3/AIS, >70% CIN2, and approximately 50% of CIN1 lesions worldwide are attributed to HPV6/11/16/18/31/33/45/52/58.

Impact: If 9-valent HPV vaccination programs are effectively implemented, the majority of CIN2 and CIN3 lesions worldwide could be prevented, in addition to approximately one-half of CIN1. Cancer Epidemiol Biomarkers Prev; 23(10); 1997–2008. ©2014 AACR.


Background: HIV-infected (HIV+) men face cancer treatment disparities that impact outcome. Prostate cancer treatment and treatment appropriateness in HIV+ men are unknown.

Methods: We used electronic chart review to conduct a retrospective cohort study of 43 HIV+ cases with prostate cancer and 86 age- and race-matched HIV-uninfected (HIV) controls with prostate cancer, ages 40 to 79 years, from 2001 to 2012. We defined treatment appropriateness using National Comprehensive Cancer Network guidelines and the Charlson comorbidity index (CCI) to estimate life expectancy.

Results: Median age was 59.5 years at prostate cancer diagnosis. Median CD4+ T-cell count was 459.5 cells/mm3, 95.3% received antiretroviral therapy, and 87.1% were virally suppressed. Radical prostatectomy was the primary treatment for 39.5% of HIV+ and 71.0% of HIV men (P = 0.004). Only 16.3% of HIV+ versus 57.0% of HIV men received open radical prostatectomy (P < 0.001). HIV+ men received more radiotherapy (25.6% vs. 16.3%, P = 0.13). HIV was negatively associated with open radical prostatectomy (OR = 0.03, P = 0.007), adjusting for insurance and CCI. No men were undertreated. Fewer HIV+ men received appropriate treatment (89.2% vs. 100%, P = 0.003), due to four overtreated HIV+ men. Excluding AIDS from the CCI still resulted in fewer HIV+ men receiving appropriate treatment (94.6% vs. 100%, P = 0.03).

Conclusion: Prostate cancer in HIV+ men is largely appropriately treated. Under- or overtreatment may occur from difficulties in life expectancy estimation. HIV+ men may receive more radiotherapy and fewer radical prostatectomies, specifically open radical prostatectomies.

Impact: Research on HIV/AIDS survival indices and etiologies and outcomes of this prostate cancer treatment disparity in HIV+ men are needed. Cancer Epidemiol Biomarkers Prev; 23(10); 2009–18. ©2014 AACR.


Background: To evaluate the relationship between methylation status of blood leukocyte DNA and risk of gastric cancer, a population-based study was conducted in Linqu County.

Methods: Methylation levels of IGFII and N33 were determined by quantitative methylation-specific PCR. The temporal trend of methylation levels during gastric cancer development was investigated in 133 gastric cancer cases from two cohorts with pre– and/or post–gastric cancer samples. As the references of pre-GCs, 204 intestinal metaplasia (IM) or dysplasia (DYS) subjects who did not progress to gastric cancer during the follow-up period were selected. Meanwhile, 285 subjects with superficial gastritis/chronic atrophic gastritis (SG/CAG) were also selected as controls.

Results: IGFII median methylation level was significantly higher in gastric cancer cases than those with SG/CAG (61.47% vs. 49.73%; P < 0.001). IGFII and N33 methylation levels were elevated at least 5 years ahead of clinical gastric cancer diagnosis comparing with SG/CAG (63.38% vs. 49.73% for IGFII, 9.12% vs. 5.70% for N33; all P < 0.001). Furthermore, the frequency of hypermethylated IGFII was markedly increased in IM or DYS subjects who progressed to gastric cancer in contrast to those who remained with IM and DYS, and adjusted ORs were 12.52 [95% confidence interval (CI), 3.81–41.15] for IM and 10.12 (95% CI, 2.68–38.22) for DYS. Similar results were also found for N33 in subjects with IM (OR, 3.77; 95% CI, 1.20–11.86).

Conclusions: Our findings suggested that hypermethylated IGFII and N33 in blood leukocyte DNA were associated with risk of gastric cancer in a Chinese population.

Impact: IGFII and N33 methylation status may be related to gastric carcinogenesis. Cancer Epidemiol Biomarkers Prev; 23(10); 2019–26. ©2014 AACR.


Background: The androgen receptor (AR) is an essential gene in prostate cancer pathogenesis and progression. Genetic variation in AR exists, including a polymorphic CAG repeat sequence that is inversely associated with transcriptional activity. Experimental data suggest that heightened AR activity facilitates formation of TMPRSS2:ERG, a gene fusion present in approximately 50% of tumors of patients with prostate cancer.

Methods: We undertook a nested case–control study to investigate the hypothesis that shorter CAG repeat length would be associated with prostate cancer risk defined by TMPRSS2:ERG status. The study included 291 men with prostate cancer (147 ERG-positive) and 1,221 cancer-free controls. ORs and 95% confidence intervals (CI) were calculated using logistic regression.

Results: Median CAG repeat length (interquartile range) among controls was 22 (20–24). Men with shorter CAG repeats had an increased risk of ERG-positive (OR, 1.07 per 1 repeat decrease; 95% CI, 1.00–1.14), but not ERG-negative prostate cancer (OR, 0.99 per 1 repeat decrease; 95% CI, 0.93–1.05).

Conclusions: These data suggest that shorter CAG repeats are specifically associated with development of TMPRSS2:ERG–positive prostate cancer.

Impact: Our results provide supportive evidence that androgen signaling underlies the development of prostate tumors that harbor TMPRSS2:ERG. Moreover, these results suggest that TMPRSS2:ERG may represent a unique molecular subtype of prostate cancer with an etiology distinct from TMPRSS2:ERG–negative disease. Cancer Epidemiol Biomarkers Prev; 23(10); 2027–31. ©2014 AACR.


Background: The U.S. FDA has the authority to limit the nicotine content of cigarettes; however, there are concerns that reduced nicotine cigarettes will be smoked more intensely and, therefore, will increase exposure to toxic chemicals in smoke. This study examined changes in consumer behavior and exposure in response to cigarettes with substantially reduced nicotine content.

Methods: Seventy-two adult smokers completed an unblinded trial of reduced nicotine cigarettes. Participants completed a 7-day baseline period during which they smoked their usual cigarette brand, followed by consecutive 7-day periods smoking cigarettes with progressively lower nicotine levels (0.6, 0.3, and 0.05 mg emission Quest cigarettes). Nicotine dependence and withdrawal, smoking behavior, and biomarkers of exposure were assessed for each 7-day period.

Results: Significant reductions in nicotine intake were observed between usual brand smoking (~1.2 mg nicotine) and the 0.3 and 0.05 mg nicotine emission cigarettes, but not the 0.6 mg cigarette. The findings provide little evidence of compensatory smoking of Quest cigarettes, with no increases in exhaled breath carbon monoxide levels, smoking intensity, or levels of 1-hydroxypyrene across study periods. No significant differences were observed for smoking urges or measures of nicotine dependence.

Conclusions: The study adds to the evidence that cigarettes with markedly reduced nicotine content are not associated with increased smoking intensity or exposure to smoke toxicants.

Impact: The findings add to the evidence base on reduced nicotine content cigarettes and have the potential to inform FDA policy on nicotine levels. Cancer Epidemiol Biomarkers Prev; 23(10); 2032–40. ©2014 AACR.


Background: Skin pigmentation is a key factor for ultraviolet radiation exposure–related cancers and can make a significant contribution to the patterns of other diseases. For surveys, and to appropriately target cancer control activities, valid and reliable measures of skin color are required.

Methods: Validity and reliability of the Munsell Soil Color Charts were investigated for skin color assessment. The unexposed skin color of 280 university students was measured by spectrophotometer to calculate an Individual Typology Angle (ITA) value, and categorized by two independent raters according to the Munsell system (the latter was repeated after a 7-day interval).

Results: Interrater and intrarater reliability for the Munsell charts was found to be acceptable [intraclass correlation coefficients (ICC) of 0.85 and 0.86, respectively]. When ITA values were converted to the six Del Bino skin color categories, weighted for agreement between raters, within rater, and between Munsell chip and spectrophotometer were 0.63, 0.60, and 0.61, respectively. A tendency toward overestimation of the extremes of skin pigmentation was evident, particularly for the "brown" and "dark" skin types.

Conclusions: Study findings suggest that the Munsell Soil Color Charts represent a reliable and valid measurement strategy when assessing skin type.

Impact: The Munsell Year 2000 Soil Color Charts may provide a useful instrument for fieldwork contexts. Subsequent classification of individuals into skin cancer risk categories, rather than the use of precise ITA values, may be sufficient for targeting public health messages for skin cancer prevention and other health risks. Cancer Epidemiol Biomarkers Prev; 23(10); 2041–7. ©2014 AACR.


Background: Androgens acting via the androgen receptor (AR) stimulate production of PSA, which is a clinical marker of prostate cancer. Because genetic variants in the AR may have a significant impact on the risk of being diagnosed with prostate cancer, the aim was to investigate whether AR variants were associated with the risk of having PSA above clinically used cutoff thresholds of 3 or 4 ng/mL in men without prostate cancer.

Methods: Men without prostate cancer history (n = 1,744) were selected from the European Male Ageing Study cohort of 40 to 79-year-old men from eight different European centers. Using linear and logistic regression models, with age and center as covariates, we investigated whether AR variants (CAG repeat-length and/or SNP genotype) were associated with having serum PSA concentrations above 3 or 4 ng/mL, which often are set as cutoff concentrations for further investigation of prostate cancer.

Results: Carriers of the SNP rs1204038 A-allele (16% of the men) were more likely to have PSA>3 and 4 ng/mL (OR; 95% confidence intervals, 1.65; 1.13–2.40 and 1.87; 1.18–2.96, respectively) than G-allele carriers. They also had shorter CAG repeats (median 20 vs. 23, P < 0.0005), but CAG repeat length per se did not affect the PSA concentrations.

Conclusion: The A-allele of the SNP rs1204038 gives a 65% higher risk of having PSA above 3 ng/mL than the G-allele in men without prostate cancer, and thereby an increased risk of being referred for further examination on suspicion of prostate cancer.

Impact: Serum PSA as a clinical marker could be improved by adjustment for AR-genotype. Cancer Epidemiol Biomarkers Prev; 23(10); 2048–56. ©2014 AACR.


Background: It is unknown whether the risk for obesity-related cancers differs between metabolically unhealthy and healthy overweight/obese adults.

Methods: Data on body mass index (BMI), waist circumference (WC), waist-to-height ratio (WHtR), and random blood glucose in Framingham Heart Study adults (n = 3,763) ages 55 to 69 years were used to estimate risks of obesity-related cancers (n = 385), including postmenopausal breast, female reproductive, colon, liver, gallbladder, pancreas, and kidney cancers, as well as esophageal adenocarcinomas. Multivariable-adjusted Cox proportional hazards models were used to estimate risk for obesity-related cancers associated with body fat and metabolic health (as defined by glucose levels) among subjects in three risk groups (vs. referent group with normal weight/normal glucose): normal weight/elevated glucose, overweight/normal glucose, and overweight/elevated glucose.

Results: Overweight adults [BMI ≥ 25 or WHtR ≥ 0.51 (men) and ≥0.57 (women)] with elevated glucose (≥125 mg/dL) had a statistically significant 2-fold increased risk of developing obesity-related cancer, whereas overweight adults with normal glucose had a 50% increased risk. Normal-weight adults with elevated glucose had no excess cancer risk. The effects of BMI and WHtR were independent of one another. Finally, overweight women with elevated blood glucose had a 2.6-fold increased risk [95% confidence interval (CI), 1.4–4.9] of female reproductive (cervical, endometrial, uterine cancers) and postmenopausal breast cancers, whereas overweight women with normal glucose levels had only a 70% increased risk (95% CI, 1.1–2.5).

Conclusion: These results suggest that cancer risk may be lower among metabolically healthy overweight/obese older adults than among overweight/obese adults with metabolic dysfunction.

Impact: Metabolic dysfunction and obesity act synergistically to increase cancer risk. Cancer Epidemiol Biomarkers Prev; 23(10); 2057–65. ©2014 AACR.


Background: The World Cancer Research Fund (WCRF) and the American Institute for Cancer Research (AICR) published eight recommendations for cancer prevention, but they are not targeted at prostate cancer prevention. We investigated whether adherence to the WCRF/AICR recommendations and a prostate cancer dietary index is associated with prostate cancer risk.

Methods: We conducted a nested case–control study of 1,806 prostate-specific antigen (PSA)–detected prostate cancer cases and 12,005 controls in the ProtecT trial. We developed a prostate cancer dietary index by incorporating three dietary factors most strongly associated with prostate cancer. Scores were computed to quantify adherence to the WCRF/AICR recommendations and the prostate cancer dietary index separately.

Results: The prostate cancer dietary index score was associated with decreased risk of prostate cancer [OR per 1 score increment: 0.91; 95% confidence interval (CI): 0.84–0.99; Ptrend = 0.04] but the WCRF/AICR index score was not (OR: 0.99; 95% CI: 0.94–1.05; Ptrend = 0.71). There was no heterogeneity in association by prostate cancer stage (P = 0.81) or grade (P = 0.61). Greater adherence to recommendations to increase plant foods (OR per 0.25 index score increment: 0.94; 95% CI: 0.89–0.99; Ptrend = 0.02) and tomato products (OR adherence vs. nonadherence: 0.82; 95% CI: 0.70–0.97; P = 0.02) was inversely associated with overall prostate cancer risk.

Conclusions: Adherence to the prostate cancer–specific dietary recommendations was associated with decreased risk of prostate cancer. High intake of plant foods and tomato products in particular may help protect against prostate cancer.

Impact: Meeting the WCRF/AICR recommendations alone is insufficient for prostate cancer prevention. Additional dietary recommendations should be developed. Cancer Epidemiol Biomarkers Prev; 23(10); 2066–77. ©2014 AACR.


Background: Breast cancer has a complex etiology that includes genetic, biologic, behavioral, environmental, and social factors. Etiologic factors are frequently studied in isolation with adjustment for confounding, mediating, and moderating effects of other factors. A complex systems model approach may present a more comprehensive picture of the multifactorial etiology of breast cancer.

Methods: We took a transdisciplinary approach with experts from relevant fields to develop a conceptual model of the etiology of postmenopausal breast cancer. The model incorporated evidence of both the strength of association and the quality of the evidence. We operationalized this conceptual model through a mathematical simulation model with a subset of variables, namely, age, race/ethnicity, age at menarche, age at first birth, age at menopause, obesity, alcohol consumption, income, tobacco use, use of hormone therapy (HT), and BRCA1/2 genotype.

Results: In simulating incidence for California in 2000, the separate impact of individual variables was modest, but reduction in HT, increase in the age at menarche, and to a lesser extent reduction in excess BMI >30 kg/m2 were more substantial.

Conclusions: Complex systems models can yield new insights on the etiologic factors involved in postmenopausal breast cancer. Modification of factors at a population level may only modestly affect risk estimates, while still having an important impact on the absolute number of women affected.

Impact: This novel effort highlighted the complexity of breast cancer etiology, revealed areas of challenge in the methodology of developing complex systems models, and suggested additional areas for further study. Cancer Epidemiol Biomarkers Prev; 23(10); 2078–92. ©2014 AACR.


Background: Hyperinsulinemia resulting from obesity and insulin resistance is associated with increased risk of many cancers, but the biology underlying this risk is unclear. We hypothesized that increased mRNA levels of the insulin-like growth factor I receptor (IGFIR) versus the insulin receptor (IR) or elevated ratio of IR-A:IR-B isoforms in normal rectal mucosa would predict adenoma risk, particularly in individuals with high body mass index (BMI) or plasma insulin.

Methods: Biopsies from normal rectal mucosa were obtained from consenting patients undergoing routine colonoscopy at University of North Carolina Hospitals (Chapel Hill, NC). Subjects with colorectal adenomas were classified as cases (n = 100) and were matched to adenoma-free controls (n = 98) based on age, sex, and BMI. IGFIR and IR mRNA levels were assessed by qRT-PCR, and IR-A:IR-B mRNA ratios by standard PCR. Plasma insulin and crypt apoptosis were measured by ELISA and terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL), respectively. Logistic regression models examined relationships between receptor mRNAs, BMI, plasma insulin, and adenoma risk.

Results: Unexpectedly, cases were significantly more likely to have lower IGFIR mRNA levels than controls. No overall differences in total IR mRNA or IR-A:IR-B ratios were observed between cases and controls. Interestingly, in patients with high plasma insulin, increased IR-A:IR-B ratio was associated with increased likelihood of having adenomas.

Conclusions: Our work shows novel findings that reduced IGFIR mRNA and, during high plasma insulin, increased IR-A:IR-B ratios in normal rectal mucosa are associated with colorectal adenoma risk.

Impact: Our work provides evidence supporting a link between IGFIR and IR isoform expression levels and colorectal adenoma risk. Cancer Epidemiol Biomarkers Prev; 23(10); 2093–100. ©2014 AACR.


Background: Despite the established link between oral human papillomavirus (HPV) infection and a subset of oropharyngeal squamous cell carcinoma (OSCC), little is known about the epidemiology of oral HPV infection among healthy adults in China.

Methods: Oral swab specimens and questionnaires were collected from 5,410 individuals (ages 25–65 years). HPV DNA in oral exfoliated cells was tested by general primer-mediated (SPF1/GP6+) PCR and sequencing. Univariate and multivariate analyses were performed to assess the associations between exposure factors and oral infection.

Results: Alpha mucosal HPV types were detected in 0.67% [95% confidence interval (CI), 0.47%–0.93%] of 5,351 β-globin–positive specimens, and cutaneous HPV in 5.46% (95% CI, 4.86%–6.10%). HPV 16 and 3 were the most prevalent types of α mucosal (0.43%; 95% CI, 0.27%–0.64%) and cutaneous HPV (4.17%; 95% CI, 3.65%–4.74%), respectively. The prevalence of α mucosal HPV decreased with increasing age (25–65 years) from 0.93% to 0.36% (Ptrend = 0.033), and was associated with self-reported history of oral disease [adjusted odds ratio (OR), 4.78; 95% CI, 1.07–21.41]. In 1,614 heterosexual couples, cutaneous HPV in one partner was found to increase the other partner's risk of cutaneous HPV infection (adjusted OR, 2.33; 95% CI, 1.22–4.48).

Conclusions: Oral HPV infection, particularly with α mucosal types, is rare among healthy adults in China. A younger age and a history of oral disease imply higher risk of α mucosal HPV infection.

Impact: This study addresses the paucity of epidemiological data on oral HPV infection among healthy population in China. Cancer Epidemiol Biomarkers Prev; 23(10); 2101–10. ©2014 AACR.


Background: Given the conflicting results from observational studies, we assessed whether the use of metformin after a prostate cancer diagnosis is associated with a decreased risk of cancer-specific and all-cause mortality.

Methods: This study was conducted linking four databases from the United Kingdom. A cohort of men newly diagnosed with nonmetastatic prostate cancer with a history of treated type II diabetes, between April 1, 1998 and December 31, 2009, was followed until October 1, 2012. Nested case–control analyses were performed for cancer-specific mortality and all-cause mortality, in which exposure was defined as use of metformin during the time to risk-set. Conditional logistic regression was used to estimate adjusted rate ratios (RR) of each outcome with 95% confidence intervals (CI).

Results: The cohort consisted of 935 men with prostate cancer and a history of type II diabetes. After a mean follow-up of 3.7 years, 258 deaths occurred, including 112 from prostate cancer. Overall, the post-diagnostic use of metformin was not associated with a decreased risk of cancer-specific mortality (RR, 1.09; 95% CI, 0.51–2.33). In a secondary analysis, a cumulative duration ≥938 days was associated with an increased risk (RR, 3.20; 95% CI, 1.00–10.24). The post-diagnostic use of metformin was not associated with all-cause mortality (RR, 0.79; 95% CI, 0.50–1.23).

Conclusion: The use of metformin after a prostate cancer diagnosis was not associated with an overall decreased risk of cancer-specific and all-cause mortality.

Impact: The results of this study do not support a role for metformin in the prevention of prostate cancer outcomes. Cancer Epidemiol Biomarkers Prev; 23(10); 2111–8. ©2014 AACR.


Background: Pancreatic cancer is a leading cause of cancer-related mortality in the United States and both incidence and mortality are highest in African Americans. Obesity is also disproportionately high in African Americans, but limited data are available on the relation of obesity to pancreatic cancer in this population.

Methods: Seven large prospective cohort studies pooled data from African American participants. Body mass index (BMI) was calculated from self-reported height and weight at baseline. Cox regression was used to calculate HRs and 95% confidence intervals (CI) for levels of BMI relative to BMI 18.5–24.9, with adjustment for covariates. Primary analyses were restricted to participants with ≥5 years of follow-up because weight loss before diagnosis may have influenced baseline BMI in cases who died during early follow-up.

Results: In follow-up of 239,597 participants, 897 pancreatic cancer deaths occurred. HRs were 1.08 (95% CI, 0.90–1.31) for BMI 25.0 to 29.9, 1.25 (95% CI, 0.99–1.57) for BMI 30.0 to 34.9, and 1.31 (95% CI, 0.97–1.77) for BMI ≥35.0 among those with ≥5 years of follow-up (Ptrend = 0.03). The association was evident among both sexes and was independent of a history of diabetes. A stronger association was observed among never-smokers (BMI ≥30 vs. referent: HR = 1.44; 95% CI, 1.02–2.03) than among smokers (HR = 1.16; 95% CI, 0.87–1.54; Pinteraction = 0.02).

Conclusion: The findings suggest that obesity is independently associated with increased pancreatic cancer mortality in African Americans.

Impact: Interventions to reduce obesity may also reduce risk of pancreatic cancer mortality, particularly among never-smokers. Cancer Epidemiol Biomarkers Prev; 23(10); 2119–25. ©2014 AACR.


Background: Little is known about physicians' human papillomavirus (HPV) vaccine recommendations for males while the Advisory Committee on Immunization Practices' (ACIP) permissive guidelines for male vaccination were in effect. The purpose of this study was to examine and explore factors associated with U.S. physicians' HPV vaccine recommendations to early (ages 11–12), middle (13–17), and late adolescent/young adult (18–26) males.

Methods: Nationally representative samples of family physicians and pediatricians were selected in 2011 (n = 1,219). Physicians reported the frequency with which they recommended HPV vaccine to male patients ["always" (>75% of the time) vs. other] for each age group. Statistically significant predictors of vaccine recommendation were identified using multivariable logistic regression.

Results: The prevalence of physicians reporting they "always" recommended HPV vaccination for males was 10.8% for ages 11 to 12, 12.9% for ages 13 to 17, and 13.2% for ages 18 to 26. Pediatrician specialty and self-reported early adoption of new vaccines were significantly associated with recommendation for all patient age groups. In addition, physician race and patient payment method were associated with physician recommendations to patients ages 11 to 12, and patient race was associated with recommendations to ages 13 to 17 and 18 to 26.

Conclusions: Less than 15% of physicians surveyed reported "always" recommending HPV vaccine to male patients following national guidelines for permissive vaccination. Vaccine financing may have affected physicians' vaccine recommendations.

Impact: If these recommendation practices continue following the ACIP's routine recommendation for males in October 2011, then interventions designed to increase recommendations should target family physicians and possibly use early adopters to encourage support of HPV vaccination guidelines. Cancer Epidemiol Biomarkers Prev; 23(10); 2126–35. ©2014 AACR.


Background: A diet rich in folate is associated with a reduced colorectal cancer risk, whereas the role of circulating levels is less clear. The aim of this study was to relate prediagnostic plasma folate, vitamin B12, and homocysteine concentrations to the risk of colorectal cancer.

Methods: This was a prospective case–control study of 331 cases and 662 matched controls nested within the population-based Northern Sweden Health and Disease Study. Median follow-up time from recruitment to diagnosis was 10.8 years.

Results: Plasma folate concentrations were positively related to colorectal cancer risk; multivariate odds ratios were 1.62 [95% confidence intervals (CI), 1.08–2.42] and 1.42 (95% CI, 0.94–2.21) for the middle and highest versus lowest tertile, respectively. In subjects with follow-up <10.8 years, a statistically significant doubled risk was observed for the middle and highest versus lowest tertile, whereas findings for longer follow-up times were null. A positive risk relationship was also observed for tumor stage III–IV but not I–II. Plasma vitamin B12 concentrations were inversely associated with rectal cancer risk. Homocysteine was not significantly related to colorectal cancer risk.

Conclusions: In this population-based, nested case–control study, low plasma folate concentrations were associated with a reduced colorectal cancer risk. This protective role was mainly observed in subjects with higher tumor stage or shorter follow-up time between recruitment and diagnosis. Low circulating folate status may protect against colorectal cancer or suppress progression of preneoplastic or neoplastic lesions.

Impact: These findings may have relevance for the ongoing debate about mandatory folic acid fortification of flour. Cancer Epidemiol Biomarkers Prev; 23(10); 2136–44. ©2014 AACR.


Background: We previously reported a significant association between higher UV radiation exposure before diagnosis and greater survival with melanoma in a population-based study in Connecticut. We sought to evaluate the hypothesis that sun exposure before diagnosis was associated with greater survival in a larger, international population-based study with more detailed exposure information.

Methods: We conducted a multicenter, international population-based study in four countries—Australia, Italy, Canada, and the United States—with 3,578 cases of melanoma with an average of 7.4 years of follow-up. Measures of sun exposure included sunburn, intermittent exposure, hours of holiday sun exposure, hours of water-related outdoor activities, ambient ultraviolet B (280–320 nm) dose, histologic solar elastosis, and season of diagnosis.

Results: Results were not strongly supportive of the earlier hypothesis. Having had any sunburn in 1 year within 10 years of diagnosis was inversely associated with survival; solar elastosis—a measure of lifetime cumulative exposure—was not. In addition, none of the intermittent exposure measures—water-related activities and sunny holidays—were associated with melanoma-specific survival. Estimated ambient UVB dose was not associated with survival.

Conclusion: Although there was an apparent protective effect of sunburns within 10 years of diagnosis, there was only weak evidence in this large, international, population-based study of melanoma that sun exposure before diagnosis is associated with greater melanoma-specific survival.

Impact: This study adds to the evidence that sun exposure before melanoma diagnosis has little effect on survival with melanoma. Cancer Epidemiol Biomarkers Prev; 23(10); 2145–52. ©2014 AACR.


Background: Previous epidemiologic research suggests a protective role of one-carbon nutrients in carcinogenesis. Folate, however, may play a dual role in neoplasms development: protect early in carcinogenesis and promote carcinogenesis at a later stage. We prospectively examined associations between intake of total folate, methionine, riboflavin, vitamin B6, and risk of lymphoid and myeloid neoplasms (including subtypes) and investigated whether alcohol modified the effects of folate.

Methods: The Netherlands Cohort Study consists of 120,852 individuals who completed a baseline questionnaire in 1986, including a 150-item food-frequency questionnaire. After 17.3 years of follow-up, 1,280 cases of lymphoid and 222 cases of myeloid neoplasms were available for analysis.

Results: Intakes of folate, methionine, and riboflavin were not associated with lymphoid or myeloid neoplasms. For vitamin B6, a statistically significantly increased myeloid neoplasms risk was observed (highest vs. lowest quintile: HR = 1.87; 95% confidence intervals, 1.08–3.25). When analyzing by lymphoid and myeloid neoplasms subtypes, no clear associations were observed for most subtypes, with just a few increased risks for some subtypes and nutrients. Some risks became nonsignificant after excluding early cases. No interaction between alcohol and folate was observed.

Conclusions: We observed a few significant positive associations; however, some of these would be expected to arise due to chance alone. Furthermore, some risks became nonsignificant after excluding early cases. Therefore, we conclude that there is no association between one-carbon nutrient intake and risk of lymphoid and myeloid neoplasms.

Impact: This study contributes substantially to the limited and inconclusive evidence on the association with one-carbon nutrients. Cancer Epidemiol Biomarkers Prev; 23(10); 2153–64. ©2014 AACR.


Background: Smoking experimentation in Mexican American youth is problematic. In light of the research showing that preventing smoking experimentation is a valid strategy for smoking prevention, there is a need to identify Mexican American youth at high risk for experimentation.

Methods: A prospective population-based cohort of 1,179 adolescents of Mexican descent was followed for 5 years starting in 2005–06. Participants completed a baseline interview at a home visit followed by three telephone interviews at intervals of approximately 6 months and additional interviews at two home visits in 2008–09 and 2010–11. The primary endpoint of interest in this study was smoking experimentation. Information about social, cultural, and behavioral factors (e.g., acculturation, susceptibility to experimentation, home characteristics, and household influences) was collected at baseline using validated questionnaires.

Results: Age, sex, cognitive susceptibility, household smoking behavior, peer influence, neighborhood influence, acculturation, work characteristics, positive outcome expectations, family cohesion, degree of tension, ability to concentrate, and school discipline were found to be associated with smoking experimentation. In a validation dataset, the proposed risk prediction model had an area under the receiver operating characteristic curve (AUC) of 0.719 (95% confidence interval, 0.637–0.801) for predicting absolute risk for smoking experimentation within 1 year.

Conclusions: The proposed risk prediction model is able to quantify the risk of smoking experimentation in Mexican American adolescents.

Impact: Accurately identifying Mexican American adolescents who are at higher risk for smoking experimentation who can be intervened will substantially reduce the incidence of smoking and thereby subsequent health risks. Cancer Epidemiol Biomarkers Prev; 23(10); 2165–74. ©2014 AACR.


Despite the substantial epidemiologic evidence on the inverse association between circulating 25-hydroxyvitamin D [25(OH)D] and colorectal cancer, it remains controversial whether this relationship is causal or due to confounding by inflammation. We reevaluated the association between plasma 25(OH)D and colorectal cancer risk by additionally accounting for inflammatory markers in a prospective case–control study nested within the Nurses' Health Study and Health Professionals Follow-up Study (615 cases and 1,209 matched controls). Conditional logistic regression was used to estimate relative risk (RR) and 95% confidence interval (CI) of colorectal cancer in relation to quartiles of plasma 25(OH)D. Results were compared before and after adjusting for inflammatory markers in the multivariable model. Plasma 25(OH)D was associated with reduced risk of colorectal cancer (multivariable RR comparing extreme quartiles = 0.71; 95% CI, 0.52–0.97; Ptrend = 0.01). Additional adjustment for C-reactive protein, IL6, soluble tumor necrosis factor receptor 2, or a composite inflammatory score did not change the results [multivariable (including inflammatory score) RR = 0.72; 95% CI, 0.53–0.98; Ptrend = 0.02). Our findings suggest that confounding by inflammation, as reflected by circulating inflammatory markers, does not appear to account for the inverse association between plasma 25(OH)D and colorectal cancer. Cancer Epidemiol Biomarkers Prev; 23(10); 2175–80. ©2014 AACR.


Despite the widespread use of bras among U.S. women and concerns in the lay media that bra wearing may increase breast cancer risk, there is a scarcity of credible scientific studies addressing this issue. The goal of the study was to evaluate the relationship between various bra-wearing habits and breast cancer risk among postmenopausal women. We conducted a population-based case–control study of breast cancer in the Seattle–Puget Sound metropolitan area that compared 454 invasive ductal carcinoma (IDC) cases and 590 invasive lobular carcinoma (ILC) cases diagnosed between 2000 and 2004 with 469 control women between 55 to 74 years of age. Information on bra-wearing habits and other breast cancer risk factors was collected from study participants through in-person interviews. Multivariate adjusted odds ratios (OR) and their associated 95% confidence intervals (CI) were estimated using polytomous logistic regression. No aspect of bra wearing, including bra cup size, recency, average number of hours/day worn, wearing a bra with an underwire, or age first began regularly wearing a bra, was associated with risks of either IDC or ILC. Our results did not support an association between bra wearing and increased breast cancer risk among postmenopausal women. Cancer Epidemiol Biomarkers Prev; 23(10); 2181–5. ©2014 AACR.


Socioeconomic and environmental influences are established factors promoting cancer disparity, but the contribution of biologic factors is not clear. We report a mechanistic link between carbohydrate-derived metabolites and cancer that may provide a biologic consequence of established factors of cancer disparity. Glycation is the nonenzymatic glycosylation of carbohydrates to macromolecules, which produces reactive metabolites called advanced glycation end products (AGE). A sedentary lifestyle and poor diet all promote disease and the AGE accumulation pool in our bodies and also increase cancer risk. We examined AGE metabolites in clinical specimens of African American and European American patients with prostate cancer and found a higher AGE concentration in these specimens among African American patients when compared with European American patients. Elevated AGE levels corresponded with expression of the receptor for AGE (RAGE or AGER). We show that AGE-mediated increases in cancer-associated processes are dependent upon RAGE. Aberrant AGE accumulation may represent a metabolic susceptibility difference that contributes to cancer disparity. Cancer Epidemiol Biomarkers Prev; 23(10); 2186–91. ©2014 AACR.


Background: Farming is often a family and multigenerational business. Relatedness among farmers could bias gene–environment interaction analysis. To evaluate the potential relatedness of farmers, we used data from a nested case–control study of prostate cancer conducted in the Agricultural Health Study (AHS), a prospective study of farmers in Iowa and North Carolina.

Methods: We analyzed the genetic data for 25,009 SNPs (single-nucleotide polymorphisms) from 2,220 White participants to test for cryptic relatedness among these farmers. We used two software packages: (i) PLINK, to calculate inbreeding coefficients and identity-by-descent (IBD) statistics and (ii) EIGENSOFT, to perform a principal component analysis on the genetic data.

Results: Inbreeding coefficients estimates and IBD statistics show that the subjects are overwhelmingly unrelated, with little potential for cryptic relatedness in these data.

Conclusions: Our analysis rejects the hypothesis that individuals in the case–control study exhibit cryptic relatedness.

Impact: These findings are important for all subsequent analyses of gene–environment interactions in the AHS. Cancer Epidemiol Biomarkers Prev; 23(10); 2192–5. ©2014 AACR.


Background: PSA and PSA velocity (PSAV, rate of PSA change over time) are biomarkers for diagnosis and prognosis of prostate cancer. Men who are at high risk for prostate cancer also have associated comorbidities for which they are taking NSAIDs and statins for long periods; therefore, it is important to understand the effect of these medications on markers used to assess prostate cancer risk.

Methods: Using a population of 699 men, multiple linear regressions were used to investigate the associations between PSA and concomitant medications, and mixed-effects models were used to investigate these associations with PSAV.

Results: After adjusting for selenium use, age, race, body mass index, and pack-years of smoking, aspirin, other NSAIDs, or statins did not demonstrate statistically significant associations with PSA (P = 0.79, 0.68, and 0.79, respectively) or PSAV (P = 0.23, 0.43, and 0.84, respectively). Results were not altered upon stratifying the sample between men who developed prostate cancer during the course of the study and those who did not.

Conclusions: Results from this study indicate that chronic use of aspirin, other NSAIDs, or statins did not affect PSA levels or PSAV in men at high risk for prostate cancer. Larger prospective studies designed to investigate these relationships are needed to confirm this result.

Impact: Long-term use of NSAIDs or statins in men at high risk for prostate cancer may not interfere with the diagnosis or prognosis of this disease, and supports appropriate use of these medications with regard to prostate cancer risk. Cancer Epidemiol Biomarkers Prev; 23(10); 2196–8. ©2014 AACR.