A new scoring system in cancer genetics: application to criteria for BRCA1 and BRCA2 mutation screening. Issue 2 (16th December 2013)

• Record Type: Journal Article
• Authors: Bonaïti, Bernard; Alarcon, Flora; Andrieu, Nadine; Bonadona, Valérie; Dondon, Marie-Gabrielle; Pennec, Sophie; Stoppa-Lyonnet, Dominique; Bonaïti-Pellié, Catherine; Perdry, Hervé
• Is Part Of: Journal of medical genetics. Volume 51:Issue 2(2014)
• Page Start: 114
• Page End: 121
• Publication Date: 2013-12-16
• Subjects: Cancer: Breast -- Genetic Screening/Counselling; Medical genetics -- Periodicals; 616.042
• DOI: 10.1136/jmedgenet-2013-101674
• Languages: English
• ISSNs: 1468-6244

Abstract

Abstract : Background: In hereditary forms of cancer due to mutations of genes such as BRCA1 and BRCA2, methods have been proposed to predict the presence of a mutation in a family. Methods: Relying on carriage probability computation is the most predictive, but scores are a good proxy and avoid using computer software. An empirical method, the Manchester scoring system, has been elaborated for BRCA1 and BRCA2 mutation identification. We propose a general scoring system based on a transformation of the carriage probability. Up to an approximation, the transformed carriage probability becomes an additive score. We applied this new scoring system to the diagnosis of BRCA1 -associated and BRCA2 -associated breast–ovarian cancer predisposition. Using simulations, its performance was evaluated and compared with that of the Manchester scoring system and of the exact probability. Finally, the score system was used on a sample of 4563 families screened for BRCA1 and BRCA2 mutations. Results: The performance of the new scoring system was superior to the Manchester scoring system, but the probability computation remained the most predictive. The better performance of the new scoring system was attributed to accounting for unaffected family members and for the degree of kinship of relatives with the proband. Conclusions: The new scoring system has a theoretical basis and may be applied to any cancer family syndrome and, more generally, to any disease with monogenic subentities, in which the causal gene mutations have been identified. It will be easily modified when additional predictive factors are found.