ctDNA as a disease monitoring strategy in MDS. (A) Mutations and cytogenetic abnormalities present among all patients recruited to the azacitidine + eltrombopag study. (B) MAF measured dPCR of a TP53 P177R and U2AF1 Q157P gene mutation in BM MNL, PB MNL, PB neutrophils, plasma ctDNA, and whole-blood DNA from a patient with high-grade MDS. The patient had no circulating blasts in the PB detected by morphology. BM morphology revealed multilineage dysplasia with an excess of myeloblasts (11% of nucleated cells). (C) Correlation between MAF measured by dPCR between BM and plasma ctDNA across 75 matched time points (r² = 0.84; P < .0001). (D) Case AZA004. Serial comparison of the MAF of TET2 N1890I and TET2 D1376V mutation by dPCR between BM and plasma ctDNA. The patient had MDS with a classification of RAEB-2, which responded to azacitidine and eltrombopag therapy by a reduction in BM myeloblast percentage. At various time points (day 193, day 318, and day 599), poor-quality blood-dilute aspirate samples were obtained (denoted by *). At these times, plasma ctDNA MAF was higher than BM MAF of the TET2 N1890I and TET2 D1376V mutation. There was also severe neutropenia at these time points (neutrophils = 0.13, 0.15, and 0.16 × 10⁹/L, respectively). (E) Case AZA009. Serial MAF of a NRAS G13D mutation by dPCR of BM and plasma ctDNA. The patient had RAEB-1 and stable disease after azacitidine and eltrombopag therapy represented by a persistent but stable excess of BM blasts. There was eventual progression at day 407 of therapy to acute myeloid leukemia. (F) Case AZA007. Serial MAF of a KRAS A59G and SRSF2 P95H mutation by TS of BM and plasma ctDNA, respectively. The patent had refractory cytopenia with unilineage dysplasia with severe thrombocytopenia, which responded initially to azacitidine and eltrombopag therapy. The patient eventually progressed with an increase in BM myeloblast percentage. Of note, the SRSF2 mutation, despite being reduced, still remained detectable in ctDNA at all time points sampled. At the time of disease progression, the MAF of the SRSF2 mutation in plasma had clearly increased, while the KRAS mutation remained undetectable. (G) Depth of coverage (DOC) log2 ratio plots from low-coverage whole-genome sequencing (LC-WGS) of plasma in patient AZA011. At “baseline” (top), the plot shows the presence of a loss of copy number at chromosome 9 (yellow) prior to azacitidine therapy. At “response” on day 167 (middle), there is near resolution of the copy-number alteration at chromosome 9. At day 1441, while still on therapy, at “pre-progression” (bottom) there is reemergence of the loss of copy number at chromosome 9 (yellow). (H) Serial MAF of a CBL, U2AF1, TET2, and an ASXL1 mutation of patient AZA011 throughout azacitidine and eltrombopag therapy. Response to therapy was achieved by an improvement in platelet count. The MAF of the CBL, U2AF1, and TET2 mutations reduced accordingly. At day 1441, all 3 of these MAFs increased alongside emergence of a new ASXL1 mutation. The patient subsequently progressed on day 1525 with thrombocytopenia and an increase in BM myeloblasts.