Extreme thrombocytosis in low‐risk essential thrombocythemia: Retrospective review of vascular events and treatment strategies

Abstract

Formal diagnosis of essential thrombocythemia (ET) requires a platelet count of ≥450 × 109/L that is associated with morphologic or molecular evidence of an underlying myeloproliferative neoplasm (MPN) that is not otherwise classified as polycythemia vera (PV), prefibrotic or overt primary myelofibrosis (PMF).1 Approximately 22% of patients with ET present with a platelet count of ≥1000 × 109/L, which is arbitrarily referred to as extreme thrombocytosis (ExT).2, 3 The particular presentation is quite unnerving for both patients and their physicians because of assumed increased risk for thrombotic complications. In this regard, despite the lack of supporting evidence, some physicians are inclined to institute cytoreductive therapy in the presence of ExT, in otherwise low-risk ET (defined by the absence of thrombosis history and age ≤60 years). Similarly, some investigators have in the past included4 or excluded5 such patients from clinical trials under similar assumptions. In 2006, we reported on 99 low-risk ET patients with ExT and showed similar rates of subsequent thrombotic events in the presence or the absence of prophylactic cytoreductive therapy (24% vs. 13%, respectively, after median follow-up of 12 years; p = .23).6 In the current study, we have revisited the particular issue in an expanded study population of 183 low-risk ET patients with ExT, in the context of treatment strategies undertaken at time of initial diagnosis; specific objectives included (i) comparison of thrombotic events in the presence or absence of ExT in 433 low risk patients with ET, (ii) analysis regarding the influence of initial cytoreductive therapy on such events in the subset of 183 patients with ExT, and (iii) annotation of platelet count and specific treatment at the time of a vascular event. Study patients were recruited from the Mayo Clinic ET database which is maintained under an institutional review board approved protocol and includes 1070 patients evaluated over five decades (1967–2017).2 Major arterial thrombosis included myocardial infarction, angina, cerebrovascular accidents, transient ischemic attack, peripheral arterial thrombosis, aortic thrombosis, mesenteric artery thrombosis, and central retinal thrombosis. Major venous thrombosis included deep venous thrombosis, pulmonary embolism, portal/splenic/mesenteric/hepatic vein thrombosis, and cerebral sinus thrombosis. Bleeding events that required red cell transfusions were classified as major hemorrhage. Follow up for each patient was updated up to February 2021 by means of either a phone call or electronic communication with the patient or referring physician. Efforts were made in order to collect accurate information regarding platelet count and specific therapy at the time of each specific vascular event that occurred during the post-diagnosis clinical course. Comparison between categorical variables was performed by chi-square test and continuous variables by Wilcoxon/Kruskall-Wallis tests. We utilized Cox proportional hazards model for multivariable analyses. A comparison of thrombosis-free survival was computed by the Kaplan Meir method and differences evaluated by the log-rank test. The JMP Pro 14.0.0 software package, SAS Institute, Cary, NC was utilized for all analyses. Among the 1076 ET patients listed in the Mayo Clinic database used to recruit the study population, 526 were below age 60 years at the time of their initial diagnosis. Among these patients, 433 presented without history of thrombosis, thus constituting the core study population of low-risk patients with ET (median age 44 years, range 18–59; 69% females); clinical characteristics are detailed in Table S1. Among these 433 low-risk patients, 183 (42%) presented with ExT and compared to their counterparts without ExT (n = 250), they were more likely to be younger (median age 40 vs. 47 years; p < .01), CALR-mutated (50% vs. 28%; p < .01), and present with higher leukocyte count (median 9 vs. 7.9; p < .01) and leukocytosis (24% vs. 15%; p < .01) (Table S1). Because of their lower mutant JAK2 frequency (34% vs. 59%), patients with ExT, compared to those without ExT, were more likely to present with lower IPSET score (p < .01; Table S1). On the other hand, the two groups were similar in their gender distribution (p = .6) and documentation of moderate anemia (hemoglobin <10 g/dL; p = .6), palpable splenomegaly (p = .18), abnormal bone marrow karyotype (evaluable in 315 cases; p = .23) and reporting of microvascular symptoms (p = .23) while there was a trend for lower incidence of cardiovascular (CV) risk factors (32% vs. 42%; p = .06) in patients with ExT. Median follow-up documentation for the entire group of 433 patients with low-risk ET was 15.3 years (range 1–47) and longer in those presenting with ExT (16 vs. 13.2 years; p = .01). During this period, a total of 81 thrombotic events were recorded, including 33 (18%) in patients who presented with ExT versus 48 (19%) in those presenting without Ext (p = .75). Arterial/venous thrombotic events occurred in 55/32 cases with similar distribution between patients presenting with (21/15 cases) or without (34/17 cases) ExT (p = .5/.6), respectively. Considering all thrombotic events, thrombosis-free survival (TFS) was similar between patients presenting with or without ExT (p = .32; Figure 1(A)); results were similar when arterial (p = .29) and venous (p = .9) TFS were analyzed separately. The similarity in TFS in patients presenting with and without ExT was confirmed by multivariable analysis that included age, JAK2/CALR mutation status, leukocyte count ≥11 × 109/L and CV risk factors; significance in the particular analysis was apparent only for older age (p = .04). Additional details regarding peri-event platelet count were available in 39 patients with arterial thrombosis and 23 with venous thrombosis; only six (15%) of the 39 patients with arterial events and three (13%) of the 23 with venous events displayed a platelet count of ≥1000 × 109/L at the time of their thrombotic incident, while a normal platelet count (i.e., <450 × 109/L) was documented during 36 and 43% of the incidents, respectively. Next, we examined the influence of specific therapy, instituted at time of diagnosis, on the occurrence of post-diagnosis thrombotic events, among the 183 patients presenting with ExT; specific therapy was documented in 124 of 137 evaluable cases and included aspirin therapy in 78 patients, cytoreductive therapy in 91, cytoreductive plus aspirin therapy in 45 and aspirin therapy alone in 21. In univariate analysis, TFS was significantly longer in the presence of aspirin therapy (p = .03; Figure 1(B)), but not affected by cytoreductive therapy (p = .9; Figure 1(C)), instituted at time of diagnosis. The salutary effect of aspirin therapy on TFS was confirmed by multivariable analysis that included age, JAK2/CALR mutation status, leukocyte count ≥11 × 109/L and CV risk factors; only age (p = .02) and aspirin therapy (p = .03) remained significant during the particular analysis. The protective effect of aspirin therapy was apparent for both arterial and venous TFS, although the number of events was too small to garner statistical significance. Finally, we collected treatment details at the time of vascular events, where available; during 49 arterial events with documented treatment details, 30 (61%) patients were receiving cytoreductive therapy and 16 (33%) aspirin therapy, at the time of the event. Similarly, during 26 venous events, 10 (39%) patients were receiving cytoreductive therapy and nine (34%) aspirin therapy. The current study, which considered platelet count and specific treatment at the time of both initial diagnosis and time of thrombotic event, not only confirms the lack of an association between ExT and thrombosis, in otherwise low-risk patients with ET, but also contests the benefit from cytoreductive therapy in reducing thrombotic risk in such patients.6 Instead, the study suggests that the previously well-established anti-thrombotic value of aspirin therapy in polycythemia vera7 might also apply to low-risk ET with ExT. In this regard, it is important to closely monitor the latter group of patients for bleeding associated with acquired von Willebrand syndrome, which might be exacerbated by aspirin therapy8; in the current patient cohort, major hemorrhage before or after diagnosis was documented in 15% of the patients presenting with ExT and 10% of those without ExT (p = .16). The current study complements our recent report on 192 young patients (age <40 years) with ET where ExT was associated with CALR mutation, lower likelihood of presentation with thrombosis and shortened overall and leukemia-free survival.9 In both the current and the aforementioned study,9 myelofibrosis-free survival was not affected by ExT, thus arguing against the concern that our observations might have been influenced by inadvertent inclusion of patients with prefibrotic myelofibrosis.10 In conclusion, our observations from the current study do not support the consideration of ExT as a risk factor for thrombosis, in otherwise low-risk ET, and question the value of prophylactic cytoreductive therapy in the particular scenario; a similar observation was recently communicated regarding the use of hydroxyurea in ET patients age 40–59 years and without risk factors for thrombosis or platelet count ≥1500 × 109/L11; a prospective controlled study is needed for further clarification and confirmation of the observed benefit from aspirin therapy in the current retrospective study. None of the authors disclosed any conflict of interest. Ayalew Tefferi and Natasha Szuber designed the study, collected data, performed analyses and wrote the paper. Natasha Szuber participated in collection and analysis of data. Curtis A. Hanson reviewed bone marrow morphology. Animesh Pardanani provided study patients and participated in study design and analysis of data. Tiziano Barbui and Alessandro M. Vannucchi participated in study design and review of data analysis. All authors reviewed the final draft of the paper. Data can be accessed with permission from the corresponding author. Table S1 Clinical and laboratory characteristics of 433 low-risk patients with essential thrombocythemia, stratified by presence or absence of extreme thrombocytosis (platelet count ≥1000 × 109/L) Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts