Most patients had only 1 mutation. In most series, DNMT3A accounts for >50% of all CHIP. ASXL1 and TET2 are likely underrepresented in this series as a result of poor coverage depth of ASXL1 exon 12 and sequencing of only exon 3 of TET2. PPM1D was not included in the series but would be expected to have a frequency less than TP53 but moderately common. This table is certainly incomplete, and other recurrent mutations will be discovered.

Variants found in <5 people included BCL11B, BCOR, BCORL1, BIRC3, BRAF, CARD11, CD58, CD79B, CNOT3, CUX1, DDX3X, EP300, ETV6, EZH2, FAM46C, FBXW7, FLT3, FOXP1, HIST1H1C, IDH2, IZKF1, JAK2, JARID2, KMT2D (MLL2), KDM6A, KIT, KLHL6, KRAS, LUC7L2, MPL, MYD88, NOTCH1, NOTCH2, PDSS2, PHF6, PIK3CA, PRDM1, PRPB40B, PTPN11, RIT1, RPS15, SETDB1, SF3A1, SMC1A, SMC3, STAG1, STAG2, STAT3, SUZ12, TBL1XR1, TET1, TNFAIP3, TNFRSF14, and ZRSR2.

Variant frequency is based on supplemental Table 2 of Jaiswal et al,¹⁰  which is derived from whole-exome sequencing of peripheral blood of 17 182 people with a median age of 58 y unselected for hematological phenotype, focused on 160 genes known to be recurrently mutated in hematological cancers. Frequency of mutations in other series might vary, especially because technical differences between targeted sequencing and whole-exome or whole-genome sequencing may yield different results; because marrow aspiration was not done, some patients may have had early undiagnosed hematological neoplasms meeting WHO criteria.