ALBERT

Description: Background: This year, germline predisposition to haematological malignancy (HM) debuts in the World Health Organization classification of myeloid neoplasms and acute leukemia (Blood, 2016;127:2391). It has been 17 years since germline mutations in RUNX1 were found to lead to familial platelet disorder (FPD) with predisposition to myelodysplastic syndrome and acute myeloid leukaemia (MDS/AML) (Nat Genet. 1999;23:166). Now, nearly 80 families have been reported with damaging germline mutations or deletions affecting RUNX1 function, associated with FPD, making it an increasingly significant clinical presence. Although thrombocytopenia and platelet dysfunction are present in almost all RUNX1 mutant carriers, we and others have observed that the predisposition to HM varies between family members, with respect to age at diagnosis and the type of malignancy, and in some cases RUNX1 mutation carriers have no apparent HM development over their lifespan. The reasons for this heterogeneity are currently unknown. Aims: We are conducting an international collaborative study examining RUNX1 mutated families. The aim of the research project is to classify the range of phenotypes correlated with RUNX1 mutations comprehensively (including non-malignant phenotypes such as skin disorders) and to determine if the type of RUNX1 mutation and the presence of other germline and acquired mutations in relevant HM genes correlate with the likelihood of HM development, or the type of HM that develops. Across all of our data we aim to analyse clinically relevant information that will be used to inform prognosis and clinical management in germline RUNX1 mutation carriers. Results:From a review of the literature for previously characterised RUNX1 mutant families most mutations are predicted to be loss-of-function, with the combination of frameshift, stopgain, splicing and deletion accounting for the majority of alterations (57, 70%) compared to missense mutations (22, Figure 1). The most common sites of mutation are R201 and R204, affected by both missense and stopgain (10 total), which lie within the nuclear localisation signal at the end of the RUNT domain (Figure 1). We also surveyed in detail 12 RUNX1 pedigrees with both novel and previously described missense, frameshift, stopgain and deletion mutations and found that, while all families developed myeloid malignancies, 6 families also had individuals who developed lymphoid malignancy (most often Acute lymphoblastic leukemia (ALL)) which was heritable in sub-families, and subject to anticipation (e.g see IV-5 and V-5 in Figure 2). Consistent with population genome wide association studies identifying RUNX1 as a susceptibility locus for psoriasis (J Autoimmun. 2015;64:66), we find that skin conditions (psoriasis, eczema) are common, and present in germline RUNX1 carriers in 50% of our families; most commonly observed in families with stopgain and frameshift mutations. Genomic analysis of selected samples confirms that mutation of the other RUNX1 allele is the most commonly acquired mutation in germline RUNX1 mutation carriers developing HM. Alterations of chromosomes 21 and 7 are also common. DNMT3A and PHF6 acquired mutations were the next most frequently observed in tumors and mutations in U2AF1 and ASXL1 in the blood of RUNX1 carriers without HM were observed, suggestive of pre-HM clonal expansion. Finally, in a family with a novel R169I RUNX1 mutation, a rare germline ASXL1 variant (E1102D, 1.0% in ExAC) was found in two RUNX1 carriers who developed early onset AML. This variant is also significantly enriched in an MDS cohort unselected for family history compared to the general population (HR 1.3, p=0.02), as well as ASXL1 N986S (0.1% in ExAC, HR 3.3, p=0.0002) suggesting they operate as germline HM risk modifiers. Interestingly RUNX1 and ASXL1 acquired mutations often co-occur in sporadic MDS/AML and our data suggests this collaboration may also occur at the germline level. Conclusions:Annotation of skin phenotypes co-existent with a family history of haematological malignancy may assist in identifying RUNX1 mutant families. Both acquired and germline mutations in known HM genes may modify germline RUNX1 driven HM penetrance and phenotype. Our data suggest that screening of RUNX1 germline mutation carriers for germline and acquired variants in other HM genes could provide an important tool for defining risk and requires further investigation. Disclosures Owen: Pharmacyclics: Research Funding; Janssen: Honoraria; Roche: Honoraria, Research Funding; Novartis: Honoraria; Gilead: Honoraria, Research Funding; Lundbeck: Honoraria, Research Funding; Celgene: Honoraria, Research Funding; Abbvie: Honoraria. Godley:UpToDate: Honoraria; Onconova, Inc.: Research Funding.