Page 112 - 2021_09-Haematologica-web

E. Oppliger Leibundgut et al.
A
B
Figure 2. Distribution of molecular response and driver mutations among patients with or without additional mutations. (A) Patients without additional mutations (n=8). (B) Patients with additional mutations (n=8). MMR: major molecular remission; PMR: partial molecular remission; NR: no remission; JAK2: JAK2 V617F; CALR: CALR mutation; MPL: MPL mutation.
With regard to phenotypic driver mutations, nine (50%) patients had a JAK2 V617F mutation, five (28%) patients had CALR mutations (type 1, n=3; type 2, n=1; a novel 33 bp deletion at position 1092, n=1) and two patients had MPL mutations (1 with W515L, 1 with W515K). Two patients (11%) were triple negative. Overall, there was a significant reduction of driver mutant allele burden, with a median decrease of 69% at best response during treat- ment (P<0.001). In detail, of 16 patients with a phenotypic driver mutation, ten (63%) reached a MMR (8 JAK2- mutated, 1 CALR-mutated, 1 MPL-mutated), three (19%) reached a PMR (3 CALR-mutated), and three (19%) patients did not reach a PMR (1 JAK2-mutated, 1 CALR- mutated, 1 MPL-mutated).
Additional mutations at study entry
At study entry, a total of 18 different additional somatic mutations (11 missense, 3 frameshift, 2 nonsense, 2 splice site) were identified in nine patients (50%), affecting the DNMT3A (n=6), TET2 (n=3), ASXL1 (n=2), TP53 (n=2), SF3B1 (n=2), CBL (n=1), EZH2 (n=1) and U2AF1 (n=1) genes (Figure 1). Details on mutations and variant allele frequencies at diagnosis and best response are given in Online Supplementary Table S2. Among the patients with any driver mutation, 40-56% carried up to five additional mutations (5/9 patients with JAK2 V617F, 2/5 with CALR mutation, 1/2 with MPL mutation), and of two triple-neg- ative patients one had an additional mutation.
Impact of additional mutations on molecular response and dynamics of mutant clones
Patients with or without additional mutations had simi- lar molecular responses to imetelstat therapy with five (63%) patients reaching MMR in each group; one and two patients without and with additional mutations reached PMR, respectively (Figure 2). All patients with additional mutations who reached MMR had a JAK2 V617F driver mutation. Regarding the reduction in mutant allele burden, phenotypic driver mutation response was significantly deeper in patients without additional mutations (P=0.04) (Figure 3).
Different dynamics of mutations in response to imetel- stat were observed in individual patients (Figure 4). In five patients (#1, #2, #9, #10, #17), additional mutant allele bur- dens decreased with the driver mutation. In contrast, in two patients with three and five additional mutations (#4, #14), differential responses to imetelstat treatment were observed; i.e., the allele burden of some additional muta- tions decreased in parallel with the driver mutation while others persisted or increased (i.e., mutations in TP53, SF3B1, and DNMT3A) despite driver mutation response, suggesting the presence of at least two clones or subclones. Lack of response was observed in two patients: in patient #18, a MPL mutation did not respond while the DNMT3A- mutated clone expanded, and in patient #7 without a driver mutation (triple negative), a known hotspot mutation in SF3B1 persisted at a high level. In total, non-responsive mutations were detected in TP53, DNMT3A and SF3B1 genes.
Three patients (#2, #14, #17) acquired additional muta- tions in TP53, IDH1 and TET2 with low allele burden (mean 5%) after best molecular response of the driver mutation, at 10, 9 and 13 months of imetelstat treatment, respectively (Table 1). All three patients already had one to three preexisting additional mutations in other genes at study entry.
Clinical outcome in relation to additional mutations
Hematologic and molecular responses were equally reached independently of the presence of additional muta- tions (Table 2). Loss of response was, however, more fre- quent in patients with additional mutations. Namely, four patients with additional mutations lost their molecular response, including three patients with DNMT3A muta- tions and one patient with a TET2 mutation, but none of the patients without additional mutations lost their molecular response (P=0.025). Patients with a higher bur- den of additional mutations at study entry had a shorter duration of clinical response compared to patients with no or a lower burden of additional mutations (10.2 vs. 22.1 months, median; cut-off at 10% mutant allele burden, P=0.053).
2400
haematologica | 2021; 106(9)