TO THE EDITOR:
In patients with essential thrombocythemia (ET) with extreme thrombocytosis (ExT ≥1500 × 109/L), cytoreductive therapy is recommended in the current National Comprehensive Cancer network and European LeukemiaNet treatment guidelines.1,2 A recent international physician survey disclosed marked inconsistencies regarding the management of low-risk ET with ExT ≥1500 × 109/L, arising from concerns related to an increased risk of thrombosis and/or bleeding.3,4 The prothrombotic or prohemorrhagic impact of thrombocytosis was studied in a prospective multicenter cohort of 776 patients with ET from the randomized PT-1 trial, and no association between blood counts at diagnosis and future complications was found. However, platelet count outside of the normal range (>450 × 109/L) during follow-up was associated with an immediate risk of major hemorrhage but not thrombosis.5 Conversely, some studies in ET with ExT have implicated aspirin use as the culprit for associations with major hemorrhage,6-8 whereas other studies have suggested a lower incidence of arterial thrombosis in patients with ExT.9-11 In addition, interpretation of findings from prior reports is confounded by the variability in platelet count thresholds used to define ExT.4
The objectives for the current study in ET included: (1) estimation of prevalence of ExT ≥1500 × 109/L at time of diagnosis; (2) phenotypic and genotypic characterization of patients presenting with ExT ≥1500 × 109/L; and (3) determination of impact on thrombotic and bleeding events, in the context of other risk factors and specific therapy.
Study patients were recruited from our institutional myeloproliferative neoplasms database after institutional review board approval by the Mayo Clinic institutional review board. The study included 710 patients with ET evaluated over 5 decades (1967-2021) and retrospectively reviewed and confirmed to fulfill the 2016 World Health Organization diagnostic criteria.12 The study was conducted in accordance with the Declaration of Helsinki. To minimize the inadvertent inclusion of patients with prefibrotic myelofibrosis, cases with anemia defined according to sex-adjusted hemoglobin level <11 g/dL in women and <12.5 g/dL in men were excluded.13 All cases were molecularly annotated for driver mutations, with major arterial and venous thromboses and major hemorrhage defined according to conventional criteria.14 Follow-up information for all patients, including vascular complications and disease evolution, was updated in May 2021 through either review of last clinic visit or medical records, which included access to care delivered outside the institution, or by telephone call to the patient. Analysis considered variables obtained at time of diagnosis. Comparison between categorical variables was performed by using the χ2 test and continuous variables by using Wilcoxon/Kruskal-Wallis tests. A Cox proportional hazards model was used to compute multivariable analyses. P values ≤.05 were considered significant. The JMP Pro 16.0.0 software package (SAS Institute, Inc., Cary, NC) was used for all analyses.
Among 710 consecutive Mayo Clinic patients with ET, 41 (6%) displayed ExT ≥1500 × 109/L (platelet count range, 1500-3460 × 109/L) at time of diagnosis; incidence rates were 15% (20 of 133), 5% (13 of 247), and 2% (8 of 330) for patients aged <40 years, 41 to 59 years, and ≥60 years, respectively (P < .0001). ExT ≥1500 × 109/L was present in 27 (9%) of 315 conventional “low-risk” and 14 (4%) of 395 “high-risk” patients, based on age >60 years and thrombosis history (P = .004). Similarly, the incidence of ExT ≥1500 × 109/L was 3% (25 of 986), 5% (42 of 891), and 3% (15 of 558) in 3 independent Italian cohorts from the University of Florence, University of Bergamo, and Catholic University of Rome. Incidence rates of extreme thrombocytosis >1500 × 109/L at time of ET diagnosis at their respective institutions were provided by co-authors V.D.S., T.B. and A.M.V. in August 2021 (unpublished data).
Phenotypic and genotypic characteristics of the 41 patients presenting with ExT ≥1500 × 109/L were compared with those of their counterparts presenting with a platelet count <1500 × 109/L (n = 669) (Table 1). In univariate analysis, patients with ExT ≥1500 × 109/L were more likely to be younger (median age, 40 vs 59 years; P < .0001) and to display CALR mutation (44% vs 26%; P = .001), lower hemoglobin level (median, 13.1 g/dL vs 13.9 g/dL; P = .001), leukocytosis ≥11 × 109/L (43% vs 21%; P = .001), and major hemorrhage (15% vs 4%; P = .001) at presentation.15 In multivariable analysis, the significant associations with younger age, lower hemoglobin level, leukocytosis, and major hemorrhage were maintained. No significant associations were noted for incidence of arterial (7% vs 14%; P = .25) or venous (12% vs 10%; P = .58) thrombosis or microvascular symptoms (23% vs 24%; P = .87), between patients presenting with or without ExT ≥1500 × 109/L, respectively. Instead, multivariable analysis for events at/before diagnosis identified male sex (odds ratio [OR], 1.8; P = .01), JAK2 mutation (OR, 2.4; P = .01), and cardiovascular risk factors (OR, 2.4; P = .01) as risk factors for arterial thrombosis and JAK2 mutation (OR, 2.7; P = .01) as a risk factor for venous thrombosis.
Abnormal von Willebrand factor profile at diagnosis defined by ristocetin cofactor activity <30% or loss of high-molecular-weight multimers was documented in 3 of 7 (43%) vs 19 of 57 (33%) evaluable patients with or without ExT ≥1500 × 109/L (P = .62). Next-generation sequencing performed in a subset of patients (n = 244) depicted a higher incidence of U2AF1 (7% vs 0.4%; P = .01) and EZH2 (7% vs 0.9%; P = .04) mutations among patients with ExT ≥1500 × 109/L. Transformation to myelofibrosis was documented in 10 (24%) patients with ExT ≥1500 × 109/L at diagnosis vs 86 (13%) without ExT (P = .05); this difference was accounted for by longer follow-up of patients with ExT ≥1500 × 109/L (11.2 years vs 8.5 years; P < .001). Furthermore, myelofibrosis-free survival was similar among patients with ExT ≥1500 × 109/L at diagnosis (not reached) vs those without ExT (26 years; P = .63).
A comparison of patients presenting with ExT ≥1500 × 109/L and those with platelet count between 1 and 1.49 million (n = 83) revealed the following differences: younger age (40 years vs 69 years; P < .0001), predominance of female subjects (73% vs 57%; P = .07), and CALR genotype (44% vs 28%; P = .02), with a lower incidence of cardiovascular risk factors (35% vs 76%; P < .0001) and arterial thrombosis at/before diagnosis (7% vs 21%; P = .04) but a higher rate of hemorrhage at/before diagnosis (15% vs 5%; P = .07) in patients with ExT ≥1500 × 109/L. Conversely, comparison of patients presenting with a platelet count between 1 and 1.49 million and <1 million revealed predominance of the CALR genotype with platelet count between 1 and 1.49 million (43% vs 21%; P < .0001), with no differences in age, cardiovascular risk factors, or thrombosis and hemorrhage at/before diagnosis. Given the unique clinical characteristics of patients with ExT ≥1500 × 109/L, it was chosen as the cutoff for this study.
CALR mutation was associated with a lower incidence of venous (0% CALR vs 22% others; P = .03) and arterial (0% CALR vs 13% others; P = .11) thrombosis at presentation; leukocytosis ≥11 × 109/L (24% vs 4% with/without; P = .07) was identified as an additional risk factor for venous thrombosis (supplemental Table 1). Neither driver mutations (P = .77) nor leukocytosis ≥11 × 109/L (P = .73) was associated with major hemorrhage at presentation. A total of 14 vascular events (5 arterial thrombosis, 2 venous thrombosis, and 7 major hemorrhage) were documented during median follow-up of 11.2 years as detailed in Table 2. Leukocytosis ≥11 × 109/L (24% vs 4% with/without; P = .06), and cardiovascular risk factors (23% vs 8% with/without; P = .11) were borderline significant for arterial thrombosis–free survival. In terms of major hemorrhage-free survival, presence of JAK2 mutation (36% JAK2 vs 8% others; hazards ratio, 5.4; P = .03) and leukocytosis ≥11 × 109/L (29% vs 9% with/without; P = .07) emerged significant/near significant, with the former retaining significance on age-adjusted analysis. Notably, major hemorrhages during follow-up occurred in 3 cases in the absence of antiplatelet/anticoagulant treatments. In the remaining 4 cases, the patients at higher risk of hemorrhage were receiving treatments: high-dose aspirin (325 mg) (n = 1), warfarin (n = 1), and warfarin plus aspirin (n = 2), suggesting caution in adopting such regimens in those patients. Among 19 patients with persistent ExT ≥1500 × 109/L, 2 arterial events, 1 venous thrombotic event, and 6 major hemorrhagic events were recorded at follow-up. On univariate analysis for hemorrhage-free survival, age >60 years (P = .03), male sex (P = .02), leukocytosis ≥11 × 109/L (P = .19), and JAK2 mutation (P = .14) emerged significant/near significant. The limited number of thrombotic events precluded analyses for thrombosis-free survival.
The presenting clinical features and vascular events for 24 low-risk patients presenting with ExT ≥1500 × 109/are provided in supplemental Table 2. Initial treatment details included observation alone (n = 5), aspirin alone (n = 5), cytoreductive therapy alone (n = 7), and aspirin plus cytoreduction (n = 7). At a median follow-up of 15.3 years, 2 arterial thrombotic events were documented; in both instances, cytoreductive therapy but not aspirin was ongoing at the time of event. A single venous thrombotic event was recorded postdiagnosis in a patient who was under observation. Of 12 patients initiated on aspirin at diagnosis, none experienced thrombosis while on therapy, and all 3 thrombotic events occurred in its absence, suggesting a protective effect of aspirin for both arterial and venous thrombosis. Two patients experienced major hemorrhage postdiagnosis; one was associated with acquired von Willebrand syndrome (ristocetin cofactor activity, 32%), in a patient with platelet count of 1520 × 109/L who was under observation. Meanwhile, none of the patients on aspirin experienced major hemorrhage postdiagnosis.
The current study provides information regarding the phenotype and genotype of patients with ET presenting with ExT at diagnosis. The prospect of controlled studies for further clarification of treatment approach in ET patients with ExT is challenged by the very low incidence of informative cases. Regardless, the information from the current retrospective study is not inconsistent with our current practice of avoiding cytoreductive therapy in otherwise low-risk ET patients with ExT. The current findings require validation in prospective multicenter series.
Contribution: N.G. and A.T. designed the study, collected data, performed analysis, and co-wrote the paper; N.S., Y.J., and F.F. collected data; C.A.H. performed review of bone marrow biopsy samples; and K.B., M.A.E., A.P.W.-S., A.D.P., V.D.S., A.M.V., and T.B. contributed patients; and all authors reviewed and approved the final draft of the paper.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
Correspondence: Naseema Gangat, Division of Hematology, Department of Medicine, Mayo Clinic, 200 First St SW, Rochester, MN 5590; e-mail: gangat.naseema@mayo.edu; and Ayalew Tefferi, Division of Hematology, Department of Medicine, Mayo Clinic, 200 First St SW, Rochester, MN 5590; e-mail: tefferi.ayalew@mayo.edu.
References
Author notes
Presented in abstract form at the 63rd annual meeting of the American Society of Hematology, Atlanta, GA, 11-14 December 2021.
Requests for data sharing may be submitted to Naseema Gangat (gangat.naseema@mayo.edu).
The full-text version of this article contains a data supplement.