Vannucchi, Alessandro M.

Background New therapies for myelofibrosis (MF) are being developed but how these compare with current therapies is limited to analyzing single outcome measures such as spleen volume reduction or quality-of-life (QoL). However, these outcome trajectories might diverge and some endpoints may not be clinically relevant. To remedy this an ELN (European LeukemiaNet) taskforce developed a composite endpoint using the Desirability-Of-Outcome-Ranking (DOOR) method which allows multiple outcomes to be simultaneously considered and therapies achieving diverging endpoints compared. Methods An Expert Panel used DOOR to divide outcome relevancy into 8 dimensions: clinical meaningfulness, modifiability, reliability, feasibility, decisional value, acceptability, frequency, and cost. Next, a weight was assigned to each dimension using the mean Likert score assigned by the panelists. Next, the panel validated which dimensions were met by each of 30 outcomes most frequently listed by clinical trials. A weighted score (WS) for each outcome was calculated based on weights assigned to a dimension and accomplishment rates assigned to outcomes. Outcomes were then distributed into quartiles. Next, operational definitions of outcomes were tested in Delphi consensus method requiring > 75% agreement. Last, a literature review indicated correlations of the outcomes ranked in the 1st 2 quartiles with survival. Results Three composite endpoints were developed and named DEMYOs (DEsirability-of-MYelofibrosis-Outcomes). DEMYO-1, the proportion of people alive and free of severe vascular events, blast transformation, severe anaemia and thrombocytopenia and very-severe symptom burden, encompassed outcomes in the 1st quartile (WS > 29.5) and is the proposed composite endpoint for trials of people with advanced JAK-inhibitor refractory MF. DEMYO-2, defined as the proportion of people alive and free of accelerated or blast phase, severe or worsening anaemia, severe and major vascular events, severe or symptomatic splenomegaly, severe or worsening symptoms and any emergent adverse genetic abnormality, included the outcomes ranked in the 1st 2 quartiles (WS > 26.5) and is proposed for trials of people with a sub-optimal response to JAK-inhibitors. DEMYO-3, defined as the proportion of people alive and free of blast transformation, vascular events, severe, symptomatic or worsening clinical features and of worsening prognostic co-variates, included outcomes in the 1st 3 quartiles (WS > 23.5) and is proposed for trials of people not yet receiving JAK-inhibitors. Interpretation DEMYO composite endpoints capture relevant clinical features of MF and represent endpoints important to patients, physicians and other stakeholders. DEMYO validation is feasible by available trial data and, once validated, the composite endpoints can be endorsed by medicines regulatory agencies. DOOR proved feasible and robust in this 1st-in-haematology application and we encourage further use. No funding

Background A JAK2 mutation is detected in almost all patients with polycythemia vera (PV) and in ⁓50% of those with essential thrombocythemia (ET). The two diseases also share characteristic clinical and bone marrow morphological traits, suggesting they exist on a disease continuum rather than representing discrete entities. The objective of the current study was to compare clinical and genetic characteristics as well as long-term outcome between formally-defined PV vs. JAK2-mutated ET vs. JAK2-wild-type (WT) ET. Methods Study patients were retrospectively recruited from Mayo Clinic, USA databases (1967-2024). Diagnostic criteria were according to the International Consensus Classification (ICC; Blood 2022;140:1200). All study patients were fully annotated for JAK2, CALR, and MPL mutations while cytogenetic and additional next-generation sequencing (NGS)-derived mutation information was available in a subset of patients. Conventional statistical methods were employed for phenotypic comparisons and survival analysis. Results Among 2,010 study patients, 1,010 had PV, 617 (63%) JAK2-mutated ET, and 383 JAK2-WT ET (CALR 26%; MPL 3%; triple-negative 9%). Phenotypic comparisons revealed older age distribution in PV compared to JAK2-mutated ET (p=0.03) and the latter compared to JAK2-WT ET (p<0.01). Females were represented more in JAK2-mutated ET (69%), compared to both PV (49%; p<0.01) and JAK2-WT ET (54%; p<0.01). The frequencies of leukocytosis, palpable splenomegaly, constitutional symptoms, diabetes, hypertension, and abnormal karyotype were all significantly higher in PV, compared to both groups of ET, while JAK2-mutated compared to JAK2-WT ET displayed higher leukocyte count and incidence of hypertension. The incidences of arterial thrombosis at/prior to diagnosis were 14%, 16%, and 10% in PV, JAK2-mutated and WT ET, respectively (P=0.03), and venous thrombosis 14%, 13%, and 6% (p<0.01). After a median follow-up of 8.6 years, 632 (32%) deaths, 329 (16%) fibrotic progressions, 79 (4%) leukemic transformations, and 236 (12%) arterial and 202 (10%) venous thrombotic events were recorded, among all 2,010 study patients. Median overall survival was 19.3 years for PV vs. 17.8 years for JAK2-mutated ET vs. 23 years for JAK2-WT ET (p<0.01); however, the significant difference in favor of JAK2-WT ET was no longer apparent when analysis was adjusted for age and sex (p=0.15) and subsequently for other risk factors (p=0.86). All-inclusive multivariable analysis identified older age (p<0.01), male sex (p<0.01), higher leukocyte count (p<0.01), and arterial thrombosis history (p<0.01), but not the distinction between PV, JAK2-mutated, and JAK2-WT ET (p=0.9), to be independently associated with shortened survival. Leukemia-free survival was similar between PV and either JAK2-mutated (p=0.5) or JAK2-WT (p=0.1) ET. On the other hand, myelofibrosis-free survival was shorter in PV, compared to both JAK2-mutated (HR 1.8, 1.4-2.4; p<0.01) and JAK2-WT (HR 1.4. 1.0-1.8; p=0.03) ET. In multivariable analysis, older age (p<0.01), higher leukocyte count (p<0.01), PV vs. JAK2-mutated ET (HR 1.8; p<0.01), and JAK2-WT vs. JAK2-mutated ET (HR 1.6; p=0.01) were associated with a higher risk of fibrotic progression, with no difference noted between PV and JAK2-WT ET (p=0.4). Arterial thrombosis-free survival was similar between PV and JAK2-mutated ET (p=0.8). In univariate analysis, JAK2-WT ET vs. PV was associated with superior arterial thrombosis-free survival (p=0.02), but significance was lost when analysis was adjusted for arterial thrombosis history (p=0.1). Venous thrombosis-free survival favored JAK2-WT ET vs. PV (HR 0.3; p<0.01) and JAK2-mutated ET (HR 0.6; p=0.02) and the latter vs. PV (HR 0.6; p<0.01); the significant differences between PV and both groups of ET were sustained when analysis was adjusted for age, sex, and venous thrombosis history. Conclusions The current study highlights significant demographic and phenotypic differences between PV and JAK2-mutated ET. However, risk factor-adjusted overall, leukemia-free, and arterial thrombosis-free survival was not inferior in PV, compared to ET, regardless of the presence or absence of a JAK2 mutation in the latter. JAK2-WT ET was distinctly associated with a lower risk of venous thrombosis while JAK2-mutated ET appeared to be less vulnerable to fibrotic progression.

Abstract The ACE-536-MF-001 trial enrolled patients with myelofibrosis (n = 95) into 4 cohorts: patients in cohorts 1 and 3A were non–transfusion dependent (NTD) and had anemia; patients in cohorts 2 and 3B were transfusion dependent (TD); and patients in cohort 3A/3B had stable ruxolitinib treatment before and during the study. All patients received luspatercept (1.0-1.75 mg/kg, 21-day cycles). Treatment was extended if clinical benefit was observed at day 169. The primary end point was anemia response rate (NTD, ≥1.5 g/dL hemoglobin increase from baseline; TD, transfusion-independence) over any 12-week period during the primary treatment period (weeks 1-24). Overall, 14% of patients in cohorts 1 and 3A, 10% in cohort 2, and 26% in cohort 3B met the primary end point. In cohorts 1 and 3A (NTD), 27% and 50% of patients, respectively, had mean hemoglobin increase of ≥1.5 g/dL from baseline. Among TD patients, ∼50% had ≥50% reduction in transfusion burden. Reduction in total symptom score was observed in all cohorts, with the greatest response rate seen in cohort 3A. Overall, 94% of patients had ≥1 adverse event (AE); 47% had ≥1 treatment-related AE (TRAE; 11% grade ≥3), most frequently hypertension (18%), managed with medical intervention. One patient had a serious TRAE leading to luspatercept discontinuation. Nine patients died on treatment (unrelated to study drug). In most patients, ruxolitinib dose and spleen size remained stable. In patients with myelofibrosis, luspatercept improved anemia and transfusion burden across cohorts; the safety profile was consistent with previous studies. This trial was registered at www.ClinicalTrials.gov as #NCT03194542.

AbstractOverviewEssential thrombocythemia is a Janus kinase 2 (JAK2) mutation‐prevalent myeloproliferative neoplasm characterized by clonal thrombocytosis; clinical course is often indolent but might be interrupted by thrombotic or hemorrhagic complications, microcirculatory symptoms (e.g., headaches, lightheadedness, and acral paresthesias), and, less frequently, by disease transformation into myelofibrosis (MF) or acute myeloid leukemia.DiagnosisIn addition to thrombocytosis (platelets ≥450 × 109/L), formal diagnosis requires the exclusion of other myeloid neoplasms, including prefibrotic MF, polycythemia vera, chronic myeloid leukemia, and myelodysplastic syndromes with ring sideroblasts and thrombocytosis. Bone marrow morphology typically shows increased number of mature‐appearing megakaryocytes distributed in loose clusters.Genetics: Approximately 80% of patients express myeloproliferative neoplasm driver mutations (JAK2, CALR, MPL), in a mutually exclusive manner; in addition, about 50% harbor other mutations, the most frequent being TET2 (9%–11%), ASXL1 (7%–20%), DNMT3A (7%), and SF3B1 (5%). Abnormal karyotype is seen in <10% of patients and includes +9/20q−/13q‐.Survival and PrognosisLife expectancy is less than that of the control population. Median survival is approximately 18 years but exceeds >35 years in younger patients. The triple A survival risk model, based on Age, Absolute neutrophil count, and Absolute lymphocyte count, effectively delineates high‐, intermediate‐1‐, intermediate‐2‐, and low‐risk disease with corresponding median survivals of 8, 14, 21, and 47 years.Risk Factors for ThrombosisFour risk categories are considered: very low (age ≤60 years, no thrombosis history, JAK2 wild‐type), low (same as very low but JAK2 mutation present), intermediate (same as low but age >60 years), and high (thrombosis history or age >60 years with JAK2 mutation).Mutations and PrognosisMPL and CALR‐1 mutations have been associated with increased risk of MF transformation; spliceosome with inferior overall and MF‐free survival; TP53 with leukemic transformation, and JAK2V617F with thrombosis. Leukemic transformation rate at 10 years is <1% but might be higher in JAK2‐mutated patients with extreme thrombocytosis and those with abnormal karyotype.TreatmentThe main goal of therapy is to prevent thrombosis. In this regard, once‐daily low‐dose aspirin is advised for all patients and twice daily for low‐risk disease. Cytoreductive therapy is advised for high‐risk and optional for intermediate‐risk disease. First‐line cytoreductive drugs of choice are hydroxyurea and pegylated interferon‐α and second‐line busulfan.Additional ContentThe current review includes specific treatment strategies in the context of extreme thrombocytosis, pregnancy, splanchnic vein thrombosis, perioperative care, and post‐essential thrombocythemia MF, as well as new investigational drugs.

Abstract Among 281 patients with essential thrombocythemia and calreticulin (CALR) mutation, we found a variant allele frequency of ≥60% to be associated with significantly shortened myelofibrosis-free survival, mostly apparent with CALR type-1 and CALR type-indeterminate mutations.

Introduction: Anemia is a cardinal feature of myelofibrosis (MF), along with splenomegaly and constitutional symptoms. Red blood cell (RBC) transfusions, often used to manage anemia in patients (pts) with MF, are a negative prognostic factor for overall survival and are associated with diminished quality of life and increased healthcare-related economic burden. Momelotinib (MMB), a Janus kinase (JAK) 1/JAK2/activin A receptor type 1 inhibitor, has demonstrated clinically meaningful and durable improvements in anemia, splenomegaly, and symptoms. Prespecified anemia endpoints in MMB phase 3 studies prioritized achievement of a strict transfusion independence response, defined as no transfusions for ≥12 wk immediately before the end of wk 24, with all hemoglobin levels ≥8 g/dL. Longitudinal analysis of transfusion intensity (units per 28 days) in a phase 2 trial of MMB in pts with transfusion-dependent MF demonstrated a reduction in transfusion burden in most (85%) pts (Harrison C, et al. EHA 2023). Here, we further characterize the impact of MMB and comparators on transfusion burden in pts with JAK inhibitor-naive and -experienced MF from the phase 3 SIMPLIFY-1 and MOMENTUM trials. Methods: This analysis evaluated time-dependent transfusion burden in pts enrolled in the phase 3 SIMPLIFY-1 (NCT01969838) and MOMENTUM (NCT04173494) trials. SIMPLIFY-1 included pts with JAK inhibitor-naive MF randomized (1:1) to receive MMB or ruxolitinib (RUX). In MOMENTUM, pts with symptomatic (Myelofibrosis Symptom Assessment Form Total Symptom Score ≥10), anemic (hemoglobin &lt;10 g/dL), JAK inhibitor-experienced MF were randomized (2:1) to receive MMB or danazol (DAN). All RBC units transfused were collected during the 24-wk randomized treatment period of each study and during the 84 days before receipt of randomized treatment on study for baseline assessment. Time-dependent transfusion burden was quantified by number of RBC units administered, in a tabular display of baseline- and treatment-period intensity. Pts were grouped jointly based on the baseline- and treatment-period intensity of RBC units per 28 days into ordinal bins: exactly zero units, &gt;0 to 1 unit, &gt;1 to 2 units, &gt;2 to 3 units, &gt;3 to 4 units, and &gt;4 units. Results: In SIMPLIFY-1 (JAK inhibitor-naive pts), 150 of 213 evaluable pts (70%) in the MMB arm and 163 of 216 evaluable pts (76%) in the RUX arm required zero units of RBC transfusion per 28 days at baseline. Of pts who required zero units of transfusion at baseline, a higher proportion in the MMB arm (142 of 150 [95%]) maintained a requirement of zero RBC transfusions during randomized treatment of 24 wks vs the RUX arm (93 of 163 [57%]). Overall, the mean RBC transfusion burden per 28 days declined 0.10 units (SD, 0.701) from baseline to randomized treatment in the MMB arm and increased 0.39 units (SD, 1.016) in the RUX arm. Using ordinal bins, 87% of pts in the MMB arm maintained (144 [67.6%]) or improved (41 [19.2%]) RBC transfusion intensity during randomized treatment compared with baseline vs 54% in the RUX arm (maintained, 94 [43.5%]; improved, 23 [10.6%]) (Figure). In MOMENTUM (JAK inhibitor-experienced pts), most pts had some transfusion requirement at baseline, with 26 of 130 (20%) in the MMB arm and 11 of 65 (17%) in the DAN arm requiring zero units of RBC transfusion per 28 days. During randomized treatment, a higher proportion of pts in the MMB arm (46 of 130 [35%]) required zero units of RBC transfusion vs the DAN arm (11 of 65 [17%]), including a higher proportion of those who had zero RBC transfusion requirement at baseline who maintained zero RBC transfusion requirement during randomized treatment (MMB, 92%; DAN, 64%). The mean RBC transfusion burden per 28 days declined 0.86 units (SD, 1.748) from baseline to randomized treatment in the MMB arm and declined 0.28 units (SD, 1.584) in the DAN arm. Using ordinal bins, 85% of pts in the MMB arm maintained (25 [19.2%]) or improved (85 [65.4%]) RBC transfusion intensity compared with baseline vs 63% in the DAN arm (maintained, 7 [10.8%]; improved, 34 [52.3%]). Conclusions: These data demonstrate that MMB was associated with better maintenance of RBC transfusion intensity and zero RBC transfusion status vs RUX in pts with MF who were JAK inhibitor naive and showed greater reduction in RBC transfusion burden from baseline vs DAN in pts with MF who were JAK inhibitor experienced. Across both trials, ≥85% of pts treated with MMB maintained or improved transfusion intensity.

Background: In addition to the driver mutations JAK2, CALR and MPL, patients (pts) with primary myelofibrosis (PMF) may also express mutations in a number of genes frequently associated with myeloid neoplasms. These genes encode regulators of chromatin modification and DNA methylation, RNA splicing and transcription, as well as oncogenes and signal transduction. In pts with PMF, including those with prefibrotic (pre-)-PMF, selected mutations were shown to deserve prognostic relevance, and led to the development of molecular integrated, or exclusive, scores (such as MIPSS70, MIPSS70-plus, GIPSS and others). Aim: The aim of the study was to correlate the type of mutations, clustered in relation to their known functional classes, with hematological and clinical correlates, and outcome in a large population of molecularly annotated pts with PMF. Patients and Methods: All pts with a confirmed ICC 2022 diagnosis of PMF from the database at CRIMM (Florence), for whom mutation analysis of a panel of 40 myeloid neoplasm-associated genes performed by NGS (Oncomine, ThermoFisher) was available, were included in the study. Pts were stratified in 6 categories based on known function of mutated gene: spliceosome (cat. 1: SF3B1, U2AF1, SRSF2, ZRSR2), DNA methylation gene (cat. 2: TET2, DNMT3A, IDH1, IDH2, SETBP1), histone modification gene (cat. 3: ASXL1, EZH2), RAS pathway (cat. 4: NRAS, KRAS, CBL, NF1, PTPN11), transcription factor (cat. 5: RUNX1, NFE2, TP53) and “other” gene mutations (cat. 6: LNK/SH2B3). Results: A total of 286 pts with at least 1 myeloid neoplasm-associated gene mutation were included (110 prePMF, 38.5%; 176 overt PMF, 61.5%). Of these, 69 have mutations in spliceosome genes, 81 in DNA methylation genes, 95 in histone modification genes, 19 in the RAS pathway, 13 in transcription factor genes and 9 in “other” gene category. Analysis of clinical characteristics (Table) showed that Cat. 1 pts were significantly older (80.6% aged &gt;60y, P=0.003), no difference regarding gender. There was no difference in blood counts but circulating blasts &gt;1%, that were higher in cat 3 and 1 compared to the others (P=0.009). A G3 fibrosis was more common in cat 3 and 5 (43.5% and 50%, p&lt;0.001). No significant difference was noticed as regarded driver mutations and karyotype abnormalities. Splenomegaly was more common in cat 3 and cat 6 (86.5% and 88.9%, respectively, p=0.015), while constitutional symptoms were enriched in cat 5 (69.2%, p=0.015). The proportion of pts with transfusion dependency was lower among those with mutations in DNA methylation genes (cat. 2) and “other” mutations (cat. 6) (21% and 22.2%, respectively, p=0.014). There was no significant difference among gene categories as regarded IPSS score, MIPSS70 score, thrombosis and bleeding events, progression to overt MF (in case of pre-PMF), transformation to AML and rate of secondary neoplasms. The median overall survival (OS) for all pts was 7.1 years (IC: 5.7-8.5), with categories 2, 4 and 5 showing the longest survival (10.7y (6.3-15.1), 11.5y (4.4-18.7), and 10.5y (2-18.9) respectively). In particular, pts in cat 2 showed significantly longer OS than pts in cat 1 (5.4y; HR 1.7, 95% CI, 1.1-2.6; p=0.021) and cat 3 (10.7 vs 6.1y; HR 1.5, 95% CI, 1-2.3). Considering that cat 1, 3, and 6 on one side, and cat 2, 4, 5 on the other, did not differ significant each to the other, we combined those categories to assess respective impact on survival. We found a statistically significant difference between the 2 groups, with pts harboring mutations in spliceosome, histone modifiers and “other” genes having significantly shortened survival (5.8y, CI 4.1-7.5; HR 1.57, CI 1.13-2.18) compared to the other categories (median survival, 10.8y, CI 7.2-14.3) (P=.007). Conclusion: Current analysis revealed distinctive clinical characteristics and survival outcomes associated with specific categories of mutated myeloid genes in pts with PMF, largely overlapping with the ICC defined “myelodysplasia-related gene mutations” category.

Background. A mutation in calreticulin ( CALRm) is found in 30-35% of patients (pts) with Essential Thrombocythemia (ET). There are 2 main CALRm mutation types, Type 1/Type 1 like (T1) a 52bp deletion, and Type 2/Type 2 like (T2) a 5bp insertion. Other atypical mutations (T3) occur in a minority of pts. Recent study ( Guglielmelli P et al, BCJ 2023) in pts with myelofibrosis (MF) showed that higher CALR variant allele frequency (VAF) was associated with anemia at diagnosis and during FU, and to features of more advanced disease (CD34 + cell counts, ASXL1 mutation (mut), &gt;2 mutated myeloid genes, and shorter leukocytosis-free survival. Aim. To evaluate whether the CALRm VAF in pts with ET was associated with predefined major clinical outcomes: evolution to MF, transformation to acute leukemia (AML), thrombosis, major bleeding, and survival (OS). Patients and Methods. Diagnosis of ET was strictly according to 2023 WHO and ICC to avoid mis-inclusion of prefibrotic MF. CALRm VAF was determined by capillary gel electrophoresis as the ratio (%) of areas under the curve of CALRm/ CALRm+ CALRwt. A panel of 45 myeloid neoplasm-associated genes was sequenced by NGS. Results. A total of 281 CALRm ET pts were identified from CRIMM (Florence, I), Quebec MPN Research Group centers' (Canada) and Mayo Clinic (Rochester, US) databases; 152 (54%), 101 (36%) and 28 (10%) were T1, T2 and T3 CALRm, respectively. Overall FU was 8.6y (0.3-39.4), median survival was not reached. Rate of death at 10y, 20y and 30y was 8%, 15% and 25%. MF transformation occurred in 50 pts (18%), AML in 2 (0.7%), 25 pts (8.9%) died. A major thrombotic event before or at ET diagnosis occurred in 17 (6.0 %) pts (14 arterial and 3 venous), whereas 28 pts (10.0%) had &gt;1 major thrombosis during follow-up (53.6% arterial, 46.4% venous). Major bleeding occurred in 19 pts (6.8%; 16 at or before diagnosis). Ascertainment of correlation of CALRm VAF with predefined outcomes by continuous variable analysis (ROC curve) highlighted a significant correlation of VAF &gt;60% with MF-free survival (MFS), with HR of 2.85 (95%CI, 1.4-5.6; P= .002. Fig. 1A). On the contrary, there was no impact of CALRm VAF as continuous variable on AML, thrombosis, bleeding and OS. We therefore used a threshold of &gt;60% to compare main clinical and hematologic characteristics at diagnosis and outcomes during FU in pts categorized as CALR-high and CALR-low. Of all pts, 21 (7.5%) harbored a VAF &gt;60%; of these, 52%, 19% and 29% were T1, T2, T3, respectively, compared to 54%, 37%, 8% of CALR-low (P= .008). We found no difference in age, gender, constitutional symptoms, splenomegaly, arterial and venous thrombosis at diagnosis and FU, IPSET revised risk stratification, and bleeding events. Hemoglobin was lower (12.8 g/dL (8.4-14.0) versus 13.6 (10.1-16.0); P=.02), while leukocytes and platelets did not differ. The proportion of pts with additional mutations in myeloid genes was significantly greater in CALR-high pts compared to CALR-low (66.7% vs 30.2%; P=0.01), as it was the proportion of pts with &gt;2 mut myeloid genes (25% vs 11.6%, P=0.04). During FU, 52% of CALR-high pts (n=11) transformed to MF compared to 15% (n=39) of CALR-low (P&lt;.0001); AML and death not differ. Anemia (&lt;10g/dL) during the FU developed in 69% of CALR-high pts vs 18%, leukocytosis (&gt;15x10 9/L) in 46% vs 8% and splenomegaly (greater than 5 cm from LCM) in 50% vs 12% (P&lt;.0001 for all). The HR for anemia-free survival was 2.87 (95%CI, 1.3-6.3), for leukocytosis HR 3.71 (95%CI, 1.3-10.5) and splenomegaly HR 3.40 (95%CI, 1.2-9.2), all P&lt;.01. MFS was significantly shorter in T1 (HR 2.0; P=0.04) or T3 (HR 2.7; P=0.03) using T2 as reference. Finally, we compared C ALR-mut, JAK2V617F mut and triple-negative (TN) ET pts. ROC analysis indicated a JAK2VAF &gt;35% ( JAK2-high) as the best cutoff for shorter MFS. As in Fig. 1B, the MFS survival of CALR-high and JAK2-high was superimposable, and significantly shorter that in CALR/JAK2-low and TN pts. Conclusions. A CALR VAF &gt;60% in pts with ET is associated with greater risk and shorter time to MF progression, development of anemia and splenomegaly. Similar findings in JAK2-high pts reinforce that that accumulation of mutated CALR and JAK2 alleles, evident since diagnosis, is detrimental for evolution to post-ET myelofibrosis. Longitudinal studies after diagnosis might provide further information about the kinetics of allele accumulation.