The association between FANCD1/BRCA2 mutations and leukaemia

Patients with Fanconi Anaemia (FA) have an approximately 800-fold increased risk of leukaemia (Rosenberg et al, 2003). The rate of this complication was found to be higher in patients in the FANCG group when compared with FANCA and FANCC, and also to be higher in those with null compared with missense mutations in FANCA (Faivre et al, 2000). Further examination of genotype/leukaemia associations was recently provided by Barber et al (2005), who reviewed the rare group of FA patients with mutations in FANCD1/BRCA2, and stated that all of the cases of acute myeloid leukaemia (AML) occurred in patients with BRCA2 mutations involving the splice site for IVS7. Barber et al (2005) stated in the text that nine of 23 FANCD1/BRCA2 patients developed AML, and that all of those cases had at least one allele with an IVS7 mutation. However, in their Table I they show five cases of AML among six patients with a mutation in IVS7. In my review of 26 patients (23 from Barber's report and three additional literature cases; Howlett et al, 2002; Faivre et al, 2005; MacMillan et al, 2005), 10 cases with AML were identified, of whom only five had IVS7 mutations in one or both BRCA2 alleles. In one case with no mutation information provided in the report, T-cell acute lymphocytic leukaemia (ALL) was followed by AML (MacMillan et al, 2005). There were three additional patients reported who had ALL, two T-cell and one B-cell (listed in Table I of Barber et al, 2005). Two of the five cases with AML and mutations in IVS7 were siblings, reducing the frequency of IVS7 mutations to four unrelated families among the nine families with cases of AML. One patient with a mutation in IVS7 had neutropenia at 5 months, a Wilms tumour at 9 months, underwent nephrectomy followed by stem cell transplant and died from a pulmonary haemorrhage before 1 year of age without AML (MacMillan et al, 2005). Thus, five of six patients with mutations in IVS7 developed AML, while five of 20 without mutations in that region also had AML; the association of AML with IVS7 was still significant [OR 15, 95% confidence interval (CI) 1·1–750, P = 0·02, Fisher exact test] (stata Statistical Software, Release 9; Stata Corporation, College Station, TX, USA), but the correlation was neither as consistent nor as strong as stated by Barber et al (2005). A time-dependent analysis sheds further light on this topic. The cumulative probability of AML or ALL combined reached a plateau of 79% by 10 years of age (Fig 1A); the probability of AML alone or after ALL was 60% by age 6 years (Fig 1B). The five patients with IVS7 mutations who developed AML all did so by age 3 years (median 2 years), while the four AML patients whose mutation was not in IVS7, for whom age information was available, developed AML by age 6 years (median also 2 years); the cumulative probability of AML was 41% in the latter group (Fig 1C). The cumulative incidence by age was significantly earlier in those with mutations in IVS7 (P = 0·004 by Cox regression). Leukaemia in patients with Fanconi anaemia because of mutations in FANCD1/BRCA2. (A) Probability of all types of leukaemia. (B) Probability of acute myeloid leukaemia (AML). (C) Probability of AML according to the presence or absence of a mutation in IVS7. Black line, mutations not in IVS7 (n = 20); grey line, mutations in IVS7 (n = 6). The implication for patients with FA associated with mutations in IVS7 of BRCA2 is that the risk of AML appears to be higher, and the age at onset younger, than among patients with other mutations in BRCA2. However, it must be emphasized that the risk of both AML and ALL is still very high in patients with mutations in other regions of BRCA2. Further studies of larger numbers of patients will be required to clarify and quantify the genotype/leukaemia associations among FA patients with mutations in FANCD1/BRCA2. This research was supported by the Intramural Research Program of the National Institutes of Health and the National Cancer Institute.