gms | German Medical Science

BRCA1/2 sequence analysis frequently reveals unclassified DNA variants (UCV) i.e. missense mutations of unknown clinical significance. A set of different parameters i.e. co-occurrence, co-segregation, biochemical, interspecies variability and loss of heterozygosity (LOH) analysis can be analysed to categorize these UCV according to a multifactorial likelyhood-ratio (LR) model (Goldgar et al., AJHG 75:535-544, 2004 and Tavtigian et al, J Med Genet 13:Epub, 2005). Within the GCHBOC 88 unique BRCA1 and 124 unique BRCA2 UCV have been detected in 3081 independend index cases that have not been categorized (Update: Dec. 05). So far we performed co-occurrence and LOH analysis in 23 UCV. Calculation of LOH LRs is performed under the assumption that 95% of hereditary breast cancers but only 20% of sporadic breast cancers reveal LOH of the BRCA chromosomal regions. Co-occurrence is based on the assumption that two deleterious mutations in trans are either lethal (BRCA1) or associated with Fanconi anemia type D1 (BRCA2). Three BRCA1 UCV (R866C, S1512I, M1652I) and nine BRCA2 UCV (N289D, S384F, K1434I, I1446T, R2108C, A2717S, V2728I, E2856A, F3065L) could be classified as neutral. Eleven UCV (BRCA1: L668F, P1099L, V1653M, T1685A, G1738E, W1837R, BRCA2: L1019V, S1221Y, I2449I, V2969M, R3052W) revealed uncertain LRs due to either limited availability of tumour material or the inability to exclude co-occurrence in cis. These UCV will now be further evaluated by biochemical and segregation analysis. In summary, the multifactorial LR model provides a highly efficient approach for the evaluation of UCV. While LOH and segregation analysis requires additional material from several family members, co-occurrence is the most informative factor for LR calculation. Therefore, large data sets as well as the availability of tumour material strongly promote clarification of UCV.

For the German Consortium for Hereditary Breast and Ovarian Cancer (GCHBOC).