Multi-Sibling Families Seen as Brain Cancer Risk

HEIDELBERG, Germany, Dec. 12 -- Older children in big families may be at an elevated risk of brain and other nervous system tumors, according to epidemiologists here.

Their findings suggest that infections -- related to the number of siblings -- may play a role in central nervous system cancer, said Andrea Altieri, D.Sc., of the German Cancer Research Center, and colleagues, in the Dec. 12 issue of Neurology.

The investigators analyzed the Swedish Family-Cancer Database, which links census, death, and family information to a compulsory national cancer registry that includes 13,613 nervous system cancer cases. Sweden has the world's highest childhood rate of astrocytoma.

Individuals with three or more younger siblings had a significantly increased overall incidence rate ratio of nervous system cancer (RR 1.22, 95% confidence interval 1.16 to 1.29) compared with those who had no younger sibling.

"The strong effect for increasing number of younger siblings suggests that exposures to infective agents later in childhood may be more likely to be involved in the etiology of the disease," the researchers wrote.

Conversely, there was a non-significant protective effect for having older siblings (overall RR 0.86, 95% CI, 0.81 to 0.92), which is likely related to an increased risk of infections early in childhood.

"The decreased risks with the number of older siblings, although not significant, suggests that exposure to infections in the perinatal period may be beneficial, in a similar way as for neuroblastoma, childhood leukemia, and Hodgkin's lymphomas," they added.

Among the cancer cases in the study, the most common subtypes were astrocytoma (44%), meningioma (19%), and neurinoma (12%), all of which are classified as brain tumors. The other subtypes accounted for less than 5% each. The mean ages at diagnosis ranged from 10.0 for neuroblastoma to 47.0 for meningiomas.

Cancer risk was significantly elevated for several types of brain cancer. The findings were (four or more younger siblings compared with only one):

• For hemangioblastoma, rate ratio was 1.68 (95% CI 1.22 to 2.31) versus 1.18 (95% CI 0.89 to 1.56, P=0.007 for trend).
• For childhood neuroblastoma, the rate ratio was 2.01 (95% CI 1.48 to 2.73) versus 0.98 (95% CI 0.77 to 1.25, P<0.0001 for trend).
• For childhood ependymoma, the rate ratio was 1.83 (95% CI 1.12 to 3.00) versus 0.83 (95% CI 0.55 to 1.25, P=0.003 for trend).

Risks were systematically higher for each type for diagnosis during childhood, and the strongest associations were found for cancers diagnosed at age 15 or younger. These rate ratios for three or more younger siblings compared with none were:

• 2.30 (95% CI 1.62 to 3.28) for medulloblastoma,
• 2.61 (95% CI 1.67 to 4.07) for ependymoma,
• 3.71 (95% CI 1.88 to 7.34) for meningioma, and
• 2.13 (95% CI 1.66 to 2.74) for neuroblastoma.

Interestingly, the researchers found that the increases in risk they found could not be fully explained by known risk factors, such as high doses of ionizing radiation, family history, and some rare genetic syndromes.

"The two- to fourfold increased risks for individuals with a high number of younger siblings are stronger than most established risk factors for [nervous system tumors]," Dr. Altieri and colleagues wrote.

When the researchers analyzed the database for the number of older siblings, they found a significant decreased risk for astrocytoma (0.87, 95% CI 0.78 to 0.96) among those with three or more older siblings. All the other subtypes showed only trends to decreased risk with increasing number of older siblings. For the older sibling analysis, risk estimates were systematically lower for cancers diagnosed at age 15 or younger.

Including only individuals with no older siblings in the number of younger sibling analyses did not eliminate the increased risk of nervous system cancer except for neurinoma and hemangioblastoma. The associations also persisted after adjustment for potential confounders including sex, birth cohort, parental history of cancer, socioeconomic index, a stricter allowance for age at diagnosis, and parental age.

The researchers said the study provides the first reliable quantification of the effects of number of siblings on the risk of nervous system tumors. Previous studies found no overall association of nervous system cancer with birth order or small increased risks, but most of them did not have a large sample size, consider tumor types separately, or stratify by age at diagnosis.

"Some of the novel findings on rare malignancies were only possible because of the large database with histopathology-specific information," they wrote.

The investigators concluded that their data are compatible with the hypothesis that increased exposure to common infections may raise the risk of hemangioblastoma and neuroblastoma.

"However," they wrote, "any hypothesis remains speculative since molecular studies have so far failed to identify a specific agent."

Further studies to identify specific agents and patterns of infections possibly involved in childhood brain cancer warrants further investigation, they said.

The Swedish Family-Cancer Database is supported by the Deutsche Krebshilfe, the Swedish Cancer Society, and the European Union. The authors reported no conflicts of interest.