Search

In this issue of the Journal, Dr. Risch (Am J Epidemiol. 2020;189(11):1218-1226) posits that the use of a combination of hydroxychloroquine and azithromycin as an outpatient treatment for high-risk patients with coronavirus 19 should be increased as a way to help curtail the ongoing pandemic. However, a calculation error occurred in the original article, and new data about the studies cited have come to light. Peculiarities in the methods of data collection and reporting in those original sources must be considered when evaluating the evidence for or against hydroxychloroquine and azithromycin bitherapy.

Contains fulltext : 169777.pdf (Publisher’s version ) (Closed access) Inflammation has been implicated in ovarian carcinogenesis. However, studies investigating the association between pelvic inflammatory disease (PID) and ovarian cancer risk are few and inconsistent. We investigated the association between PID and the risk of epithelial ovarian cancer according to tumor behavior and histotype. We pooled data from 13 case-control studies, conducted between 1989 and 2009, from the Ovarian Cancer Association Consortium (OCAC), including 9,162 women with ovarian cancers, 2,354 women with borderline tumors, and 14,736 control participants. Study-specific odds ratios were estimated and subsequently combined into a pooled odds ratio using a random-effects model. A history of PID was associated with an increased risk of borderline tumors (pooled odds ratio (pOR) = 1.32, 95% confidence interval (CI): 1.10, 1.58). Women with at least 2 episodes of PID had a 2-fold increased risk of borderline tumors (pOR = 2.14, 95% CI: 1.08, 4.24). No association was observed between PID and ovarian cancer risk overall (pOR = 0.99, 95% CI: 0.83, 1.19); however, a statistically nonsignificantly increased risk of low-grade serous tumors (pOR = 1.48, 95% CI: 0.92, 2.38) was noted. In conclusion, PID was associated with an increased risk of borderline ovarian tumors, particularly among women who had had multiple episodes of PID. Although our results indicated a histotype-specific association with PID, the association of PID with ovarian cancer risk is still somewhat uncertain and requires further investigation.

There have been recent proposals advocating the use of additive gene-environment interaction instead of the widely used multiplicative scale, as a more relevant public health measure. Using gene-environment independence enhances statistical power for testing multiplicative interaction in case-control studies. However, under departure from this assumption, substantial bias in the estimates and inflated type I error in the corresponding tests can occur. In this paper, we extend the empirical Bayes (EB) approach previously developed for multiplicative interaction, which trades off between bias and efficiency in a data-adaptive way, to the additive scale. An EB estimator of the relative excess risk due to interaction is derived, and the corresponding Wald test is proposed with a general regression setting under a retrospective likelihood framework. We study the impact of gene-environment association on the resultant test with case-control data. Our simulation studies suggest that the EB approach uses the gene-environment independence assumption in a data-adaptive way and provides a gain in power compared with the standard logistic regression analysis and better control of type I error when compared with the analysis assuming gene-environment independence. We illustrate the methods with data from the Ovarian Cancer Association Consortium.

Childbearing at an older age has been associated with a lower risk of endometrial cancer, but whether the association is independent of the number of births or other factors remains unclear. Individual-level data from 4 cohort and 13 case-control studies in the Epidemiology of Endometrial Cancer Consortium were pooled. A total of 8,671 cases of endometrial cancer and 16,562 controls were included in the analysis. After adjustment for known risk factors, endometrial cancer risk declined with increasing age at last birth (P(trend) < 0.0001). The pooled odds ratio per 5-year increase in age at last birth was 0.87 (95% confidence interval: 0.85, 0.90). Women who last gave birth at 40 years of age or older had a 44% decreased risk compared with women who had their last birth under the age of 25 years (95% confidence interval: 47, 66). The protective association was similar across the different age-at-diagnosis groups and for the 2 major tumor histologic subtypes (type I and type II). No effect modification was observed by body mass index, parity, or exogenous hormone use. In this large pooled analysis, late age at last birth was independently associated with a reduced risk of endometrial cancer, and the reduced risk persisted for many years.

Previously developed models for predicting absolute risk of invasive epithelial ovarian cancer have included a limited number of risk factors and have had low discriminatory power (area under the receiver operating characteristic curve (AUC) < 0.60). Because of this, we developed and internally validated a relative risk prediction model that incorporates 17 established epidemiologic risk factors and 17 genome-wide significant single nucleotide polymorphisms (SNPs) using data from 11 case-control studies in the United States (5,793 cases; 9,512 controls) from the Ovarian Cancer Association Consortium (data accrued from 1992 to 2010). We developed a hierarchical logistic regression model for predicting case-control status that included imputation of missing data. We randomly divided the data into an 80% training sample and used the remaining 20% for model evaluation. The AUC for the full model was 0.664. A reduced model without SNPs performed similarly (AUC = 0.649). Both models performed better than a baseline model that included age and study site only (AUC = 0.563). The best predictive power was obtained in the full model among women younger than 50 years of age (AUC = 0.714); however, the addition of SNPs increased the AUC the most for women older than 50 years of age (AUC = 0.638 vs. 0.616). Adapting this improved model to estimate absolute risk and evaluating it in prospective data sets is warranted.

The survival duration for pancreatic cancer is short. Given its low lifetime risk (1.5%), established factors for the disease have insufficient specificity to identify individuals at high risk of nonfamilial cancer, and prediagnostic signs and symptoms are vague and not limited to pancreatic causes. We considered whether statistical models that incorporated both risk factors and prediagnosis symptomatology could improve prediction enough to provide practical risk estimates. We combined US Surveillance Epidemiology and End Results (SEER) incidence data from 2008 to 2010 with regression models from representative case-control data from Connecticut (2005–2009) to estimate age- and sex-specific 5-year absolute risks of pancreatic cancer diagnosis. Our risk model included current cigarette smoking (adjusted odds ratio (OR) = 3.3, 95% confidence interval (CI): 2.1, 5.0), current use of proton pump-inhibitor antiheartburn medications (OR = 6.2, 95% CI: 1.7, 23), recent diagnosis of diabetes mellitus (OR = 4.8, 95% CI: 2.2, 11), recent diagnosis of pancreatitis (OR = 19, 95% CI: 3.1, 120), Jewish ancestry (OR = 1.8, 95% CI: 1.1, 3.1), and ABO blood group other than O (OR = 1.3, 95% CI: 1.0, 1.8). In total, 0.87% of controls with combinations of these factors had estimated 5-year absolute risks greater than 5%, and for some, the risks reached more than 10%. Combining risk factors for pancreatic cancer with detectable prediagnostic symptomatology can allow investigators to begin to identify small segments of the population with risks sufficiently high enough to make screening efforts among them potentially useful.

Controversy surrounds the relations among infertility, fertility drug use, and the risk of ovarian cancer. The authors pooled interview data on infertility and fertility drug use from eight case-control studies conducted between 1989 and 1999 in the United States, Denmark, Canada, and Australia. Odds ratios and 95% confidence intervals were calculated, adjusting for age, race, family history of ovarian cancer, duration of oral contraception use, tubal ligation, gravidity, education, and site. Included in the analysis were 5,207 cases and 7,705 controls. Among nulligravid women, attempts for more than 5 years to become pregnant compared with attempts for less than 1 year increased the risk of ovarian cancer 2.67-fold (95% confidence interval (CI): 1.91, 3.74). Among nulliparous, subfertile women, neither any fertility drug use (odds ratio (OR) = 1.60, 95% CI: 0.90, 2.87) nor more than 12 months of use (OR = 1.54, 95% CI: 0.45, 5.27) was associated with ovarian cancer. Fertility drug use in nulligravid women was associated with borderline serous tumors (OR = 2.43, 95% CI: 1.01, 5.88) but not with any invasive histologic subtypes. Endometriosis (OR = 1.73, 95% CI: 1.10, 2.71) and unknown cause of infertility (OR = 1.19, 95% CI: 1.00, 1.40) increased cancer risk. These data suggest a role for specific biologic causes of infertility, but not for fertility drugs in overall risk for ovarian cancer.

The survival duration for pancreatic cancer is short. Given its low lifetime risk (1.5%), established factors for the disease have insufficient specificity to identify individuals at high risk of nonfamilial cancer, and prediagnostic signs and symptoms are vague and not limited to pancreatic causes. We considered whether statistical models that incorporated both risk factors and prediagnosis symptomatology could improve prediction enough to provide practical risk estimates. We combined US Surveillance Epidemiology and End Results (SEER) incidence data from 2008 to 2010 with regression models from representative case-control data from Connecticut (2005-2009) to estimate age- and sex-specific 5-year absolute risks of pancreatic cancer diagnosis. Our risk model included current cigarette smoking (adjusted odds ratio (OR) = 3.3, 95% confidence interval (CI): 2.1, 5.0), current use of proton pump-inhibitor antiheartburn medications (OR = 6.2, 95% CI: 1.7, 23), recent diagnosis of diabetes mellitus (OR = 4.8, 95% CI: 2.2, 11), recent diagnosis of pancreatitis (OR = 19, 95% CI: 3.1, 120), Jewish ancestry (OR = 1.8, 95% CI: 1.1, 3.1), and ABO blood group other than O (OR = 1.3, 95% CI: 1.0, 1.8). In total, 0.87% of controls with combinations of these factors had estimated 5-year absolute risks greater than 5%, and for some, the risks reached more than 10%. Combining risk factors for pancreatic cancer with detectable prediagnostic symptomatology can allow investigators to begin to identify small segments of the population with risks sufficiently high enough to make screening efforts among them potentially useful.

Lung cancer is the most common malignancy in the Western world, and the main risk factor is tobacco smoking. Polymorphisms in metabolic genes may modulate the risk associated with environmental factors. The glutathione S-transferase theta 1 gene (GSTT1) is a particularly attractive candidate for lung cancer susceptibility because of its involvement in the metabolism of polycyclic aromatic hydrocarbons found in tobacco smoke and of other chemicals, pesticides, and industrial solvents. The frequency of the GSTT1 null genotype is lower among Caucasians (10-20%) than among Asians (50-60%). The authors present a meta- and a pooled analysis of case-control, genotype-based studies that examined the association between GSTT1 and lung cancer (34 studies, 7,629 cases and 10,087 controls for the meta-analysis; 34 studies, 7,044 cases and 10,000 controls for the pooled analysis). No association was observed between GSTT1 deletion and lung cancer for Caucasians (odds ratio (OR) = 0.99, 95% confidence interval (CI): 0.87, 1.12); for Asians, a positive association was found (OR = 1.28, 95% CI: 1.10, 1.49). In the pooled analysis, the odds ratios were not significant for either Asians (OR = 0.97, 95% CI: 0.83, 1.13) or Caucasians (OR = 1.09, 95% CI: 0.99, 1.21). No significant interaction was observed between GSTT1 and smoking on lung cancer, whereas GSTT1 appeared to modulate occupational-related lung cancer. This study was partially supported by Environmental Cancer Risk, Nutrition and Individual Susceptibility (ECNIS) contract 513943.

A case-control study of reproductive factors and cancer of the ovary was conducted during 1989-1992 in metropolitan Toronto and nearby areas of Southern Ontario, Canada. In total, 450 women aged 35-79 years with histologically verified new primary epithelial ovarian cancers were interviewed concerning their reproductive histories. Over the same time period, 564 randomly selected population controls, frequency-matched to the cases according to three 15-year age groups, were also interviewed. Continuous unconditional logistic regression methods were used for analysis. It was found that childbearing and use of oral contraceptives were associated with significant decreasing trends in risk of ovarian cancer; the respective odds ratios were 0.78 for each full-term pregnancy (p < 10(-8)) and 0.92 for each year of use (p < 10(-6)). Hysterectomy was also associated with reduced risk, even after more than 20 years. Among parous women, infertility did not appear to affect risk; for nulliparous women, some evidence of increased risk was present, although fertility problems were reported by only a small fraction of nulliparae. It is suggested that the relatively lower parity of cases as compared with controls may be due to voluntary choices for having fewer children.

The association between the occurrence of carcinoma of the breast and previous usage of menopausal estrogens, progestins, and oral contraceptives is examined in a record-linkage study using the Saskatchewan Health Prescription-Drug-Plan Database. Saskatchewan Health is a governmental agency that funds publicly insured health care for essentially all residents of the province. For this study, all women aged 43-49 years in 1976 resident in Saskatchewan were identified from the Saskatchewan Health master registration file. These women were linked by registration beneficiary number to the Drug-Plan Database for the period from January 1976 to June 1987 and to the Provincial Cancer Registry Database for the period from March 1960 to December 1990. The fact and date of death or emigration from the province were obtained through the annual updates of the health plan. Of the 33,003 women initially in the cohort, 213 had a breast cancer diagnosed before 1976 and were omitted from this analysis. Between 1976 and 1990, 742 new primary breast cancer cases occurred. Women taking estrogens unopposed by progestins had an elevated risk of breast cancer, the risk increasing by 7% (relative risk = 1.072, 95% confidence interval 1.02-1.13; p = 0.008) for each 252 tablets used (approximately 1 year of use). Usage of estrogens opposed by progestins showed no association with risk (p = 0.48). Women taking oral contraceptives during this follow-up period also had a higher risk, increasing by 14% (relative risk = 1.144, 95% confidence interval 1.05-1.24; p = 0.002) for every 252 tablets used. These 1-year risk elevations are small but become appreciable at longer durations. For example, at 5 years of unopposed estrogen use, the relative risk is 1.42; for 5 years of oral contraceptive use, it is 1.96.

A case-control study of male-female differences in cigarette smoking and lung cancer was conducted during 1981-1985 in Toronto, St. Catharine's, and Niagara Falls, Ontario, Canada. In total, 442 female and 403 male histologically verified cancer cases were individually matched by age and area of residence to each other and to 410 female and 362 male randomly selected population controls. Subjects were interviewed concerning their exposures to various life-style factors, and in particular, they received detailed questioning regarding their lifelong histories of usage of tobacco products. It was found that, for both sexes, a greatly elevated risk of developing lung cancer was associated with cigarette consumption, increasing with pack-years of cigarettes smoked and declining with duration of time since quitting smoking. Furthermore, the association was significantly (p = 0.010) and appreciably stronger for females than for males. At a history of 40 pack-years relative to lifelong nonsmoking, the odds ratio for women was 27.9 (95% confidence interval (CI) 14.9-52.0) and that for men was 9.60 (95% CI 5.64-16.3). Higher odds ratios for females were also seen within each of the major histologic groupings. Thus, the higher elevated risk of lung cancer currently observed in other studies for female ever smokers compared with male ever smokers, while possibly attributable in part to greater smoking cessation among males, may be due to higher susceptibility among females.

Lung cancer is the most common cancer worldwide. Polymorphisms in genes associated with carcinogen metabolism may modulate risk of disease. Glutathione S-transferase pi (GSTP1) detoxifies polycyclic aromatic hydrocarbons found in cigarette smoke and is the most highly expressed glutathione S-transferase in lung tissue. A polymorphism in the GSTP1 gene, an A-to-G transition in exon 5 (Ile105Val, 313A --> 313G), results in lower activity among individuals who carry the valine allele. The authors present a meta- and a pooled analysis of case-control studies that examined the association between this polymorphism in GSTP1 and lung cancer risk (27 studies, 8,322 cases and 8,844 controls and 15 studies, 4,282 cases and 5,032 controls, respectively). Overall, the meta-analysis found no significant association between lung cancer risk and the GSTP1 exon 5 polymorphism. In the pooled analysis, there was an overall association (odds ratio = 1.11, 95% confidence interval: 1.03, 1.21) between lung cancer and carriage of the GSTP1 Val/Val or Ile/Val genotype compared with those carrying the Ile/Ile genotype. Increased risk varied by histologic type in Asians. There appears to be evidence for interaction between amount of smoking, the GSTP1 exon 5 polymorphism, and risk of lung cancer in whites.

A case-control study of associations between dietary and reproductive factors and cancer of the ovary was conducted during 1989-1992 in metropolitan Toronto and nearby areas of southern Ontario, Canada. In total, 450 women aged 35-79 years with histologically verified new primary epithelial ovarian cancers were interviewed concerning their reproductive history and dietary practices. Over the same time period, 564 randomly selected population controls, frequency-matched to the cases according to three 15-year age groups, were also interviewed. Continuous unconditional logistic regression methods were used for analysis. It was found that childbearing and use of oral contraceptives were associated with significant decreasing trends in the risk of epithelial ovarian cancer of all principal histologic types except mucinous tumors. For each full-term pregnancy, the odds ratio was 0.76 (95% confidence interval (CI) 0.69-0.85) for nonmucinous tumors and 1.03 (95% CI 0.88-1.21) for mucinous tumors; for each year of oral contraceptive use, the odds ratio was 0.89 (95% CI 0.85-0.93) for nonmucinous tumors and 0.98 (95% CI 0.93-1.04) for mucinous tumors (p = 0.00051 and p = 0.0040, respectively, for the difference in odds ratios between mucinous and nonmucinous tumors). Saturated fat intake also appeared to convey greater increased risk for women with mucinous tumors than for women with neoplasms of other histologic types (p = 0.029). Among women with nonmucinous tumors, increasing trends in risk of invasive serous cancer (p = 0.018), and particularly endometrioid cancer (p = 0.0041), were seen with use of noncontraceptive estrogens. Otherwise, borderline-malignant neoplasms seemed to have a similar spectrum of risk factor associations as invasive cancers. On the basis of this study and a number of others, the authors suggest that mucinous ovarian tumors may be etiologically unrelated to other types of epithelial tumors, and thus should be considered separately in studies of ovarian cancer.

A case-control study of male-female differences in cigarette smoking and lung cancer was conducted during 1981-1985 in Toronto, St. Catharine's, and Niagara Falls, Ontario, Canada. In total, 442 female and 403 male histologically verified cancer cases were individually matched by age and area of residence to each other and to 410 female and 362 male randomly selected population controls. Subjects were interviewed concerning their exposures to various life-style factors, and in particular, they received detailed questioning regarding their lifelong histories of usage of tobacco products. It was found that, for both sexes, a greatly elevated risk of developing lung cancer was associated with cigarette consumption, increasing with pack-years of cigarettes smoked and declining with duration of time since quitting smoking. Furthermore, the association was significantly (p = 0.010) and appreciably stronger for females than for males. At a history of 40 pack-years relative to lifelong nonsmoking, the odds ratio for women was 27.9 (95% confidence interval (CI) 14.9-52.0) and that for men was 9.60 (95% CI 5.64-16.3). Higher odds ratios for females were also seen within each of the major histologic groupings. Thus, the higher elevated risk of lung cancer currently observed in other studies for female ever smokers compared with male ever smokers, while possibly attributable in part to greater smoking cessation among males, may be due to higher susceptibility among females.

It is well established that women with a family history of breast cancer run a higher risk of breast cancer than do women without a family history. The evidence, however, is less clear regarding a possible association between a family history of breast cancer and risk of second primaries. The purpose of this prospective study was to estimate the risk for second primary breast cancer associated with having a family history of breast, endometrial, and ovarian cancers. A cohort of 4,660 women with a first primary breast cancer diagnosed between 1980 and 1982 were interviewed as part of the Cancer and Steroid Hormone Study, a multi-center population-based case-control study, and followed through eight Surveillance, Epidemiology, and End Results (SEER) program registries for 4 to 6 years. Of these women, 136 developed a second primary breast cancer in the contralateral breast at least 6 months after diagnosis of the first primary. Cox proportional hazards modeling techniques were used to model the time to onset of second primary breast cancer while adjusting for multiple predictors. The risk of contralateral breast cancer was elevated among cohort members who reported a history of breast cancer in a first-degree relative (multivariable-adjusted rate ratio (RR) = 1.91, 95% confidence interval (CI) = 1.22-2.99). Early age at onset (< 46 years) in the relative further increased the risk of developing contralateral breast cancer (sister: multivariable-adjusted RR = 3.36, 95% CI 1.62-6.98; mother: multivariable-adjusted RR = 2.35, 95% CI 1.02-5.43). Bilateral breast cancer in mothers was also associated with more than a two and a half-fold increase in risk (multivariable-adjusted RR = 2.55, 95% CI 1.02-6.35). The association between family history of breast cancer and risk of contralateral breast cancer did not vary substantially according to age at onset of the first primary breast cancer. The age-adjusted rate ratio for development of a second primary breast cancer among women with a first-degree relative with endometrial cancer was 2.13 (95% CI 1.04-4.35), while the corresponding rate ratio among women with a family history of ovarian cancer was 1.69 (95% CI 0.42-6.83). There was little evidence that age at onset among the relatives with endometrial or ovarian cancer affected the risk. Some of these findings have not been previously reported and need replication in future studies.