Relationship between additional mutations at diagnosis and treatment response in patients with essential thrombocythemia

Abstract

Patients with essential thrombocythemia (ET) have a chronic evolution with a risk of hematologic transformation associated with a dismal outcome. Because patients with resistance or intolerance have adverse prognosis, it is important to identify which patient will respond to first-line treatment. We, therefore, aim to describe the association between additional mutations and response to first-line treatment in patients with ET. In this retrospective study, we analyzed the molecular landscape of 121 ET patients first-line treated with hydroxyurea (HU; n = 86) or pegylated interferon (peg-IFN; n = 35). Patients undergoing peg-IFN therapy were younger and had higher proportion of low and very low risk of thrombosis recurrence. A total of 62 patients (51%) had ≥1 additional mutations at diagnosis. At 12 months of treatment, 75 patients (62%) achieved complete response (CR), 37 (31%) partial response, and 7 (6%) no response. The presence of at least 1 additional mutation at diagnosis was associated with not achieving CR (hazard ratio [HR], 0.65; P = .038), whereas treatment with peg-IFN was associated with higher CR (HR, 2.00; P = .002). The number of additional mutations at diagnosis was associated with hematologic progressions (P < .0001). None of the patients receiving peg-IFN therapy progressed to myelofibrosis, whereas 16 of 86 patients (19%) treated with HU developed secondary myelofibrosis. In conclusion, our results suggest that the presence of at least 1 additional mutation at diagnosis is associated with failure to achieve CR and also with an increased risk of hematologic evolution.

Graphical abstract

Figure 1.

Molecular characteristics of 121 patients at diagnosis. (A) Distribution of driver mutations in the whole cohort. (B) Distribution of the number of additional clonal mutation per patient. (C) Number of additional mutations per gene in the whole cohort. The y-axis represents the total count of mutations detected in all patients, and the x-axis represents the 32 mutated genes ranked in order of frequency. (D) Distribution of variant allele frequency (VAF). Violin plots representing the distribution of allele burden for the most frequent mutation in JAK2V617F, CALR, TET2, and DNMT3A genes.

Figure 2.

Evolution of response to treatment. Response assessment time points at 1, 2, 3, 4, and 5 years are shown with transitions between each. The total height of the columns is proportional to the sample size. The number of patients receiving treatment is shown below the columns. This Sankey diagram represents the transition between 8 states: CR, PR, no response, partial loss of response, loss of response, missing data, treatment interruption, and end of follow-up.

Figure 3.

Variables associated with CR and treatment interruption. Predicted outcomes in patients on first-line therapy were accessed by a multistate model. (A) Kaplan-Meier curves representing 4 states and their transitions: treatment interruption because of resistance (red), interruption for other reasons (orange), patients in complete response (green), and patients on first-line therapy who did not reach any of the previous 3 states (blue). (B) Multistate model analysis with proportional Cox model results for each transition. Significant factors associated with state transition are indicated, with risk factors in red and protective factors in green. Treatment discontinuations due to resistance or other reasons were merged in the analysis. HRs are presented with their 95% confidence intervals. For each transition, the following factors were tested: age >60 years, driver mutation, ≥1 additional mutation, treatment type, previous thrombosis, and leukocyte counts, followed by a backward stepwise selection.

Figure 4.

Hematologic-free survival according to mutational status. (A) Kaplan-Meier curve for hematologic evolution–free survival according to the number of additional mutations at diagnosis. The number of patients at risk for each group are shown in the table below the curves. (B) Forest plots representing the multivariable analysis including the following variables: age >60 years, gender, previous thrombosis, leukocytosis >11 G/L, and response at 12 months. (C) Kaplan-Meier curve for hematologic evolution–free survival according to both the presence of additional mutations and the type of treatment received.