Histological concordance in familial central nervous system tumors: Evidence from nationwide Swedish Family-Cancer Database
A partir des données du registre suédois des cancers familiaux portant sur 858 patients, cette étude met en évidence une concordance entre un type histologique d'une tumeur du système nerveux central et une forte proximité génétique (entre parents au premier degré)
Background : Published studies have shown that familial risks in the primary central nervous system (CNS) tumors are usually histology-specific. If genetic factors indeed determine tumor histology it would be expected that histological types would agree between affected first-degree relatives (FDRs). Material and methods : This study was conducted using the nationwide Swedish Family-Cancer Database. FDR pairs were defined where both of them had the same histological subtype of CNS tumor. The histological concordance was determined using kappa agreement test. Results : We identified 858 familial patients (333 parent–offspring pairs, 97 sibling pairs) with primary CNS tumors. Proportion of spinal hemangioblastoma out of all familial hemangioblastomas (21%) was significantly higher than that in sporadic patients (7%; P = 0.001). The highest kappa value was found for hemangioblastoma among parent–child pairs (kappa = 86%, 95% CI: 74–98%). There was a moderate agreement for concordant neurinoma among father–daughter pairs (kappa = 48% 95% CI: 15–81%). Low grade glioma showed significant agreement among mother–daughter (kappa = 33%, 95% CI: 9–57%) and father–daughter pairs (kappa = 39%, 95% CI: 11–67%), but not in mother–son (kappa = 10%, 95% CI: −13% to 32%) and father–son pairs (kappa = 9%, 95% CI: −1% to 40%). There was histological agreements for meningioma in mother–offspring (kappa range = 20–27%) but not in father–daughter (kappa = 14%, 95% CI: −8% to 35%) and father–son pairs (kappa = 9%, 95% CI: −12% to 30%). Conclusions : Our findings suggest that shared genetic risk factors between family members could lead to specific histological types in the familial CNS tumors, especially in hemangioblastoma and neurinoma. Our data may also suggest interactions between sex hormone and some genes contributing to familial meningioma and low grade glioma.