PONdering insecticides and brain tumors

 

Several studies have found an increased risk of childhood brain tumors (CBT) with exposure to pesticides, farm residence, or parental occupation in agriculture. The relationship is plausible since organophosphorus insecticides (OP) affect the nervous system. Paraoxonase 1 (PON1) is an enzyme present in the liver and blood that detoxifies OPs. Two polymorphisms of PON1 have been identified that seem to affect transcription of the gene product or enzymatic activity. The first is a C to T substitution at position -108 in the promoter region that appears to account for ~25% of the variation in PON1 expression. The second is a coding region polymorphism that results in a Q to R amino acid change at codon 192 near the catalytic region of the PON1 protein. This variant has been shown to reduce the rate of detoxification of one common OP, chlorpyrifos, but not diazinon.

In the current study, Nielsen et al . [Env Health Perspectives 2005; 113: 909-913] retrieved archived neonatal blood spots from the Washington State Department of Health for participants in a previous case-control study of CBT in the Puget Sound area [Gurney J et al. Am J Epidemiol 1996; 143: 120-128] . In the original study, 134 cases were identified by the local cancer registry and 281 controls were identified by random digit dialing. Mothers were interviewed using a questionnaire that included questions about home pesticide use; however, DNA was not collected. Children who were born in Washington after 1977, when archiving of blood spots began, qualified for the present study. Blood spots were retrieved for 66/70 (94%) and 137/160 (86%) such children. A second set of 100 randomly chosen, anonymous bloodspots were used to supplement the control group. PON1 C-108T and Q192R genotypes were determined and merged with participants' interview data. Odds ratios (OR) and 95% confidence intervals (CI) were calculated using logistic regression.

There was a near significant linear trend (p = 0.07) in risk of CBT with each copy of the inefficient T allele. The ORs were 1.4 (95% CI: 1.0-2.2) and 2.1 (95% CI: 0.9-4.7) comparing T allele heterozygotes and homozygotes, respectively, to homozygotes for the C allele. There was no significant trend in risk associated with Q192R genotype (p = 0.36). Interestingly, an

increased risk of CBT with each T allele was present only among children exposed to pesticides in the home. The OR per T allele was 2.6 (95% CI: 1.2-5.5) among exposed children and 0.9 (95% CI: 0.5-1.6) among unexposed children, which represented significant gene-environment interaction (p = 0.03).

 

COMMENT: This study was small but informative. The finding that a polymorphism with likely functional significance was associated with risk only in the presence of pesticide exposure enhances credibility. Future studies will need to refine pesticide exposure assessment and, of course, include more participants. This study was also interesting for having used archived blood spots to obtain DNA. The original purpose of collecting blood spots was to screen infants for metabolic diseases. Since an increasing number of governments are archiving spots, they may also be put to a variety of uses in researching childhood cancer. First, as in the present study, they may be a source of DNA. More importantly, since most current studies collect DNA up front, they may provide a means of measuring levels of growth factors, antibodies, or harmful chemicals present at birth and prior to development of disease. Lastly, blood spots can serve as a source of population-based controls for genetic association studies.

Logan G. Spector

 

The air that we breath

 

The in utero origin of at least some balanced translocations associated with childhood leukemias is now generally accepted. The etiology of these initiating events remains under intense investigation. In a new study, Bocskay KA et al [Cancer Epidemiol Biomarkers Prev 2005;14:506-11] have investigated the association between the presence of chromosome aberrations in cord blood and prenatal exposure to carcinogenic polycyclic aromatic hydrocarbons (PAH). PAHs are pervasive environmental toxicants, largely coming from the burning of fossil fuels. The study included a subset of 60 mother/infant pairs enrolled in the 600 mother/infant pair Columbia Center for Children's Environmental Health Cohort Study. The population included 30 African American and 30 Dominican mothers and their offspring. Questionnaires were administered to the women, in particular to elicit history of active and passive smoking, and basic demographics. Prenatal air monitoring was used to measure maternal exposure to PAHs and HPLC was used to detect levels of PAH-DNA adducts in the cord blood of infants. Biomarkers were measured in cord blood collected from the children at delivery. FISH analysis was used to identify the presence of stable chromosomal aberrations. Stable aberrations include balanced and unbalanced translocations, breaks, deletions and insertions. Unstable aberrations i.e. fragments of chromosomes were also quantitated. Women selected for this study were all non-smokers.

PAH-DNA adducts were measurable in 46% of cord blood samples. There was significant inter-individual variability in PAH air exposure and in PAH adducts, ranging over several orders of magnitude. There was a positive and significant correlation between stable aberration frequencies and total PAHs in air. In addition, stable aberration frequencies for African American newborns were almost 50% greater than in Dominican newborns, while exposure levels of PAHs in air and PAH-DNA adducts did not significantly differ between African American and Dominican mothers and their children. There were no correlations with unstable aberrations.

 

COMMENT: This study is a useful contribution to the ongoing debate about placental transfer of maternal exposures and their role in the generation of balanced translocations and subsequent leukemias. The increased frequency of stable aberrations in African American newborns is an unexpected finding and the reason for this is unclear. It is possible that there are additional exposures that were unmeasured in this study which contributed to the frequency aberrations or that there are differences in the ability to repair such aberrations or detoxify some exposures. This area is worthy of additional investigation.

Stella M Davies

 

Breastfeeding: Mooving Back to Pasture?

 

Childhood leukemia, particularly CD10 positive acute lymphoblastic leukemia (ALL), has been speculated to arise from an abnormal response to exposure to common infections in early childhood. Both breastfeeding and exposure to early day care have been found to be protective in several studies, although there are some inconsistencies. These events are thought to assist in modulating a child's immune system in early life, enabling the child to respond more effectively to infectious challenges later on. In this report, Kwan ML et al [Br J Cancer 2005; 93:379-384] investigated the association between breastfeeding and risk of childhood ALL in the Northern California Childhood Leukemia Study (NCCLS). Briefly, the NCCLS is an ongoing case-control study of childhood leukemia in 35 counties in Northern and Central California . The study is rigorously designed and includes in-person interviews with case and control mothers of children under the age of 15 years as well as collection of biological specimens. Controls are identified through the California Birth Registry and matched to cases on date of birth, sex, Hispanic ethnic status, and maternal race. In this report from Phase I of the study, analyses were restricted to children diagnosed with ALL between 1 and 14 years of age and included a total of 311 cases and 400 controls. Overall, the authors found no association between breastfeeding and risk of childhood ALL. After controlling for income and maternal education, the overall odds ratio (OR) was 0.99 for ever breast fed (95% Confidence Interval (CI)=0.64-1.55). There was also no significant association when cases were restricted to those diagnosed between the ages of 2 and 5 years (OR=1.49; 95% CI=0.83-2.65). Further, there was no evidence of a linear trend with increased duration of breastfeeding nor any associations with mixed feeding methods or types of formula. The authors conclude that their study provides no evidence that breastfeeding reduces the risk of childhood ALL.

 

COMMENT: The authors note the high prevalence (84%) of maternal report of ever having breast fed in their control group, which may have hindered their ability to detect an association. This high prevalence of breast feeding compared to national norms (approximately 54% from the Third National Health and Nutrition Examination Survey in the United States; Li R et al; Am J Pub Health 2002; 92:1107-1110 ]) is striking and may indicate some underlying bias. However, in other analyses these investigators have shown that their interviewed control group of mothers (identified from birth certificates) is fairly representative of the source population. Importantly, a recent meta-analysis suggests an overall protective effect for breastfeeding in childhood leukemia [Kwan ML et al; Pub Health Rep 2004; 119:521-535] . Regardless of whether the inverse association will hold up in additional studies, the benefits of breastfeeding are numerous and obviously go beyond the potential protection against childhood leukemia.

Julie A. Ross

 

More buzz about EMF

 

In the last issue of C3 [Vol 16, No 4] , we covered the England and Wales study of electromagnetic field (EMF) exposure and childhood leukemia in which Draper G and colleagues [Br Med J 2005; 220:1290-92] mapped the distance from high voltage power lines and the homes of children with and without cancer. They found that children who lived closer were at an increased risk of leukemia compared to children who lived further away. As we noted, they were circumspect in their interpretation of their data. However, many critical letters to the editor followed [BMJ 2005; 331:634-637] , which raised several interesting points including a) questioning whether the controls used in the study may be systematically different than cases; b) lack of biological plausibility; c) the likely underestimating of risk by excluding home appliances and wiring; and d) lack of consideration of possible other markers that may be linked with EMF such as air pollution. Draper and colleagues were thoughtful in their response to these letters and should be commended for trying to address a difficult and volatile finding. Because epidemiology is a necessary but inexact science, we will always have caveats to interpretation. The concern is that thoughtful criticism and discussion seems to often only take place following results people take issue with.

Julie A. Ross

 

C3 Quarterly Newsletter
Children's Cancer Research Fund
Epidemiology Research Unit
Division of Pediatric Epidemiology
Clinical Research
University of Minnesota
420 Delaware St. SE, Box 422
Minneapolis, MN 55455
pedsepi@umn.edu

Editors: 
Stella M. Davies, MD, PhD, and Julie A. Ross, PhD