Clonal Architecture Analysis of TET2 Identified Distinct Origins in Myelodysplastic Syndromes

Next-generation sequencing (NGS) has deepened our understanding of myelodysplastic syndromes (MDS). Genetic mutations of TET2 were common in MDS and were also detected in asymptomatic elder persons, which were defined as aged related clonal haematopoiesis (ARCH) and considered as a myelodysplastic syndrome precursor state. In this study, we analyzed the mutation profiles of TET2 in 770 newly-diagnosed MDS subjects. 73 mutations were found in 67 of 770(8.7%) subjects. 14.9% were frame shifts, 29.9% were nonsense, 44.8% were missense and 10.7% harbored 2 TET2 mutations. TET2^MT subjects were older than TET2^WT subjects (P=0.024) and the variant allele frequency (VAF) of TET2 was positively correlated with age (r=0.318, P=0.009). Clonal architecture was determined using a copy number-adjusted VAF difference between 2 mutation events in 49 TET2^MT subjects with 2 or more mutations. 19 (38.3%) were TET2^ancestral and 30(61.2%) were TET2^subclonal. TET2^ancestral subjects were significantly older than TET2^WT subjects (P=0.013) while no difference was found between TET2^subclonal subjects and TET2^WT subjects (P=0.509). The frequency of cytosine-to-thymine (C→T) transition was significantly higher in TET2^ancestral subjects compared with TET2^subclonal subjects (P=0.029), which was considered to be a somatic mutational signature of aging, indicating that MDS driven by TET2^ancestral was likely derived from TET2^MT ARCH. The most common upstream mutations in TET2^subclonal subjects were U2AF1 (16.7%) and ASXL1 (13.3%), and the most common downstream mutations in TET2^ancestral subjects were ASXL1 (21.1%) and RUNX1 (15.8%). TET2 mutations rarely existed alone in TET2^ancestral subjects (14.3%) and were always accompanied by another mutations like U2AF1, SF3B1 and ASXL1, suggesting that the acquisition of another mutation led to the progress of ARCH to MDS in TET2^ancestral subjects. TET2^MT had no impact on survivals in overall cohort (P=0.218) while predicted poorer survivals in IPSS-R lower risk group (P=0.004). Additional ASXL1, U2AF1 and RUNX1 mutations in TET2^MT subjects indicated poorer prognosis compared with TET2^WT subjects (P=0.005; P=0.04; P<0.001) .There was no significant difference in OS of TET2^MT subjects with different clonal architecture (P=0.76). Subjects with TET2 VAF≥30% had poorer survivals compared with TET2^WT subjects (P=0.057). In conclusion, TET2^MT MDS subjects with different clonal architectures may have different origins. TET2^ancestral subjects may be derived from ARCH and progressed to MDS after secondary hits. The effects of TET2 mutation on the prognosis depended on the accompanying mutations and clonal burdens.