LAUSR

Myelodysplastic syndrome (MDS) consists of a heterogeneous group of clonal haematopoietic stem cell disorders that are characterized by peripheral cytopenia, hypercellular bone marrow (BM), and increased mortality due to the progression to acute myeloid leukaemia (AML). Higher risk MDS is generally progressive in nature, and hypomethylating agents (HMAs) are still the standard of care for those who are not suitable for allogeneic haematological stem cell transplantation (allo-HSCT). Clinical data showed that only approximately 30% of cases achieved complete remission (CR) with decitabine, and the response was sustained for 6–24 months (median 11 months). Mutations in the nucleophosmin-1 (NPM1) gene are among the most common somatic mutations identified in de novo AML and are usually associated with a favourable prognosis in the absence of unfavourable clinical parameters or FMS-like tyrosine kinase 3 (FLT3) comutations. NPM1 mutations have also been found in 1% to 5% of MDS patients, although with different mutation loci (L287fs) when compared to those in AML patients (W288fs). To date, the response to treatment and survival of NPM1-mutated MDS patients treated with HMA remains unknown. Here, we retrospectively analyzed several higher-risk MDS patients who accepted decitabine therapy in our centre. From December 2009 to July 2018, a total of 194 patients received decitabine induction treatment with 20 mg/m intravenously for five consecutive days every 4–6 weeks. The median therapy course was four cycles. Among them, twelve patients (6.2%) harboured the NPM1/L287fs mutation (Table I). The median decitabine therapy cycle for the 12 NPM1-mutated patients was also four cycles. Interestingly, patients harbouring NPM1 mutations achieved a relatively higher CR rate (6 of 12 cases, 50%) when compared to patients without NPM1 mutations (53 of 182 cases, 29.1%) (P = 0.191). Moreover, when the most common comutated gene, DNMT3A (6 of 12 cases, 50%) (Fig 1A), was wild type, patients harbouring the NPM1 mutation (DNMT3A wild type) achieved a CR rate of 83.3% (five out of six), which was significantly higher than that of patients without the NPM1 mutation (P = 0.011) (Fig 1B). Of note, when the paired sequencing was analyzed, the six patients who achieved CR with decitabine showed the loss of mutated NPM1. One patient who achieved haematological improvement (HI) showed decreased variant allele frequency (VAF) of NPM1, whereas two patients with no response (NR) showed unchanged NPM1 (Fig 1D). Notably, unlike patients harbouring TP53 mutations (who also achieved a higher CR rate), most of whom relapsed in 12 months even after receiving continued decitabine therapy,, a significantly longer relapse-free survival (RFS) period was observed in CR patients with NPM1 DNMT3A even without any subsequent therapy (Fig 1C). The median RFS of CR patients with NPM1 DNMT3A was 66 months, which is significantly longer than that in patients without NPM1 mutations (13.5 months, P = 0.006) (Fig 1E). A markedly prolonged median survival was also shown in patients harbouring NPM1 DNMT3A (median survival of 80 months), which is significantly longer than that of patients without NPM1 mutations (18 months) (P = 0.012) (Fig 1F). Interestingly, except for DNMT3A and PTPRD comutations, the response to treatment and survival of patients harbouring NPM1 mutations treated with decitabine were favourable even when there were comutations in IDH2, NRAS and FLT3. Montaban-Bravo et al. recently reported significantly higher CR rates by intensive chemotherapy when compared to HMA therapy in patients with NPM1 mutations. However, in this study, three out of the five patients (with NPM1 DNMT3A) did not show any response to at least three cycles of subsequent chemotherapy after unsuccessful decitabine therapy. We considered that the older age, the second-line choice, and the coexisting DNMT3A mutations may have been the reason for poor response to chemotherapy. Prospective multicentre studies taking into consideration the mutation state of DNMT3A would be needed. In summary, NPM1 mutation with wild-type DNMT3A defines a specific subgroup of MDS patients with a good response to treatment and prolonged survival on decitabine therapy, even when they also had some comutations that generally indicate a poor prognosis and did not receive subsequent treatment. An enlarged sample of randomized controlled studies is needed to confirm our preliminary findings.