We found that the ALAD2 allele of the G177C polymorphism was associated with increased risk of meningioma, especially in males. Confirmation of our findings will require replication in other studies with a larger number of meningioma cases. If risk of meningioma is truly increased in individuals with the ALAD2 allele, the question arises as to whether the effect depends upon exogenous chemical exposures that act on the heme synthesis pathway or is independent of such exposures. A direct effect of the ALAD2 polymorphism might be indicated if the ALAD2 allele has lower enzyme activity than the ALAD1 allele, given that the precursor ALA is thought to be neurotoxic and genotoxic (Silbergeld 2003
). However, ALAD enzyme activity does not appear significantly different for the two alleles (Battistuzzi et al. 1981
). Alternatively, it is possible that the increased risk of meningioma in ALAD2 individuals arises in the presence of chemicals that influence the heme synthesis pathway. Several chemicals have been shown to inhibit ALAD enzyme activity, including lead, trichloroethylene, bromobenzene, and styrene (Fujita et al. 2002
). Polymorphic differences in enzyme binding or chemical uptake have been examined most extensively for lead, and individuals with the ALAD2 allele are generally reported to have higher blood lead levels than are individuals with the ALAD1 allele (Alexander et al. 1998
; Bergdahl et al. 1997
; Fleming et al. 1998
; Hsieh et al. 2000
; Shen et al. 2001
; Wetmur et al. 1991
; Ziemsen et al. 1986
The observation that the relationship between ALAD2 and risk of meningioma was stronger in men than in women could be due to biologic differences or differential exposure to a chemical agent modified by ALAD genotype. Given the small number of male meningioma cases with the variant allele, we also cannot rule out the possibility that the observed effect modification is due to chance.
The specific question of ALAD genotype and brain tumor risk has not been addressed previously in the literature. In a previously published analysis of this same data set, we found elevated risk of meningioma in individuals who had worked in military occupations or as autobody painters, designers and decorators, industrial production supervisors, teachers, or managers (Rajaraman et al. 2004
). Aside from teachers and managers, all of these occupations have potential exposure to lead, suggesting that lead might be implicated in meningioma risk. Our observation of increased risk with the ALAD2 variant for meningioma, but not for glioma or acoustic neuroma, parallels the observation that reports of increased risk of brain tumor with lead have been more consistent for meningioma (Cocco et al. 1999
; Hu et al. 1999
; Navas-Acien et al. 2002
) than for glioma or for all brain tumors combined. However, the role of lead in the observed association between ALAD genotype and meningioma can be meaningfully addressed only when data on both ALAD genotype and individual lead exposure are available.
In a study with hospital controls, study results can be biased if the exposure under study is associated with conditions enrolled in the control series (Miettinen 1985
). To assess for such a bias, we conducted a sensitivity analysis by excluding one major control subgroup at a time from the analysis. Systematically excluding control subgroups did not change observed ORs appreciably and resulted, if anything, in slightly stronger evidence of an association between ALAD2 and meningioma when circulatory or digestive disorders were excluded. If use of hospital controls introduced a bias in the observed OR, therefore, the likely direction of the bias was toward the null.
The relatively low concordance rate for study duplicates is another potential concern. However, the concordance for duplicates from meningioma cases was 100%, so our observed association for meningioma was probably not affected by genotyping concerns. Moreover, if we assume that 10% nondifferential misclassification did occur, this would have biased our findings toward the null, making our observed association a conservative estimate.
We chose to study the ALAD polymorphism based on a priori biologic and functional considerations and not by screening a large number of associations. Nonetheless, the possibility that our findings are due to chance cannot be ruled out. These findings should be viewed as hypothesis generating and need to be confirmed by replication in other studies.
Although our study had limited power for evaluating risk with respect to subtypes of tumor, it remains one of the largest case–control studies of brain tumors to date. Aside from a small percentage of brain tumors that can be explained by familial syndromes or exposure to ionizing radiation, very little is known about the etiology of brain tumors (Preston-Martin and Mack 1996
; Wrensch et al. 2002
). In order to clarify the role of lead (or other chemicals) in the observed relationship between ALAD
genotype and risk of meningioma, it will be important to conduct a detailed exposure assessment and evaluate the joint effect of exposure and ALAD
genotype in this, or another, study population.