Page 57 - 2021_03-Haematologica-web
P. 57

NGS in the management of myeloid malignancies
System (CPSS),17 DIPSS9 and the MYSEC prognostic model18 were used for standard prognosis assessment of MDS, MPN/MDS, CMML and primary or secondary myelofibrosis. Molecular mark- ers considered to indicate a poor prognosis for MDS or MPN/MDS, excluding CMML, were TP53, EZH2, ETV6, RUNX1 and ASXL1 mutations, as reported by Bejar et al.6 The detection of an SF3B1 mutation was considered a good prognostic sign.19 The Itzykson8 and/or CPSS-Mol20 scores, which respectively integrate ASXL1 mutation only or RUNX1, NRAS, SETBP1 and ASXL1, were used for the prognostic assessment of CMML. SRSF2, ASXL1, IDH1, IDH2 and EZH2 mutations were indicative of a worse prognosis in the context of myelofibrosis.21 The ASXL1 mutation was also considered to be associated with poor progno- sis in patients with AA.16,22,23
Results
Patients’ characteristics
One hundred and seventy-seven patients (100 males and 77 females), originating from ten hospitals were included. Their median age was 60 years old (range, 10- 87). These patients had ICUS or MDS (n=77), MPN/MDS (n=38), AA or a suspicion of hMDS (n=19) and MPN or suspected MPN (n=43) (Table 1).
NGS was performed for 94 patients in group A, within a median of 4 months of hematologic detection of anom- alies. Group B consisted of 95 patients, with a median fol- low-up from molecular assessment to last news of 11 months. Twelve patients belonged to both groups.
Within group A, the most frequent blood count anom- alies were cytopenia (68%: anemia [56%], thrombocy- topenia [28%], neutropenia [30%] or pancytopenia [16%]) followed by thrombocytosis (16%), and monocytosis (13%). Rare patients had thrombosis with a suspicion of MPN but no blood count anomalies (3%). Bone marrow smears and/or biopsy evidenced no significant dysplastic anomalies in 53% of the cases. Cytogenetic abnormalities not specific for MDS, such as chromosome Y loss or chro- mosome 20 deletion, were observed in 8% of the cases. Karyotyping was normal in 72% or failed in 5% of the cases. No karyotyping was performed in seven cases of thrombocytosis, two cases of erythrocytosis, one cases of thrombosis with normal blood count and one case of hypereosinophilic syndrome. Before molecular assess- ment, the context was ICUS (with or without morpholog- ical evidence of dysplasia) (39%), suspicion of MPN/MDS (17%) or AA/hMDS (17%) (Table 1, Online Supplementary Table S2). All suspected cases of MPN (27%) were nega- tive for the three classic driver mutations.
The majority of patients in group B were diagnosed
with MDS (42%), followed by MPN/MDS (31%), MPN (21%) or hypoplasia (6%). Most patients with MDS, MPN/MDS and MPN were classified as low risk (60%) according to the respective scoring systems (Table 1).
Impact of mutations detected by next-generation sequencing on diagnosis (group A)
Thirty-three percent of the patients (31/94) had at least one mutation in favor of clonal hematopoiesis (range for the whole study cohort, 0-7). Twenty-six of them had a normal karyotype and five had a non-specific chromoso- mal abnormality (i.e., del(20q) or -Y). The most frequently mutated genes were ASXL1 (12%), TET2 (11%) and DNMT3A (9%) (Figure 1A, Online Supplementary Table S2). Patients with mutations were significantly older (median: 67 years vs. 48 years in those without mutations; P<0.001) but no difference in frequency of mutations was seen according to the context of suspected disease. Patients with clonal hematopoiesis also had significantly more myelemia (1.6% vs. 0.2%; P=0.005) and a higher red cell distribution index (16% vs. 14%; P=0.01). The signs of dysplasia observed on bone marrow smears and/or biopsy were not associated with the presence of a detected somatic mutation (P=not significant).
The detection of clonal hematopoiesis by NGS allowed us to retain the diagnosis of myeloid malignancy for 17 patients (18%) (3 MDS, 7 MPN/MDS and 7 MPN), eight cases of CCUS and six cases of bone marrow hypoplasia with clonal hematopoiesis (Table 2). In most patients with AA/hMDS (n=10/16), the absence of clonal hematopoiesis favored a diagnosis of idiopathic AA, while clonal hematopoiesis detected in six cases supported a diagnosis of hMDS. Thirteen mutations were observed in seven patients with suspected MPN. Among those, four had MPL mutations, respectively two subclonal hotspot p.Trp515Leu (not previously detected by Sanger sequenc- ing) and three non-canonical mutations (p.Trp515Ser, p.His499Valfs*46, p.Tyr591Asp).
Considering the patients without detected clonal hematopoiesis (n=63), the initial suspected diagnosis of myeloid malignancy was ruled out in 47/63 (75%) of the cases including 25/26 suspected MDS (96%), 8/9 suspect- ed MPN/MDS (89%), all suspected hMDS (10/10) and 4/18 suspected MPN (22%). In these patients, cytopenias were ultimately mainly considered as of peripheral origin or ICUS and monocytosis as reactive. In the absence of clonal hematopoiesis, no diagnosis could be ruled out for most suspected cases of MPN (14/18) (Table 2, Online Supplementary Table S2) because of morphological anom- alies on bone marrow biopsy (data not shown).
VUS were found in 26 cases (28%) (Figure 1A, Online
Table 1. Partition of the patients according to diagnosis or suspected diagnosis.
ICUS n=37
Age, years: median (range) 53 (18-84)
MDS n=40
63 (10-87)
MPN/MDS or suspicion n=38
63(32-73)
AA/hMDS or suspicion n=19
58 (14-78)
MPN
or suspicion n=43
57 (16-80)
Total n=177
60 (10-87)
GroupA,n 37 NA 16 16 25 94
GroupB,n 0 40 29 6 20 95
Bothgroups,n 0 NA 7 3 2 12
ICUS: idiopathic cytopenia of undetermined significance without dysplasia; MDS: myelodysplastic syndrome; MPN: myeloproliferative neoplasm; AA: aplastic anemia; hMDS: hypoplastic myelodysplasia.
haematologica | 2021; 106(3)
703


































































































   55   56   57   58   59