New and treatment-relevant risk stratification for thrombosis in essential thrombocythemia and polycythemia vera.

Abstract

Current treatment in polycythemia vera (PV) or essential thrombocythemia (ET) has not been shown to favorably affect survival or disease transformation into acute myeloid leukemia (AML) or myelofibrosis (MF) 1. On the other hand, controlled studies have shown significant reduction of thrombosis risk with aggressive phlebotomy in PV (hematocrit target <45%) 2, low-dose aspirin in PV 3, cytoreductive therapy in PV 4, and hydroxyurea therapy in high risk ET 5. Therefore, drug therapy in PV and ET is currently directed at minimizing the risk of thrombosis with drugs that have not been shown to increase the risk of AML (leukemogenic drugs in PV include chlorambucil 4, radiophosphorus 4, and pipobroman 6) or fibrotic transformation (anagrelide has been associated with increased risk of fibrotic transformation in ET) 7. Accordingly, conventionally used and presumably non-leukemogenic drugs in ET and PV include aspirin, hydroxyurea, interferon-α and busulfan. Regardless, with the exception of aspirin, because of concerns regarding the long-term safety of the latter drugs, they are usually offered to only those patients in whom the risk of thrombosis is believed to be significantly increased. PV is genetically characterized, almost always, by the presence of an activating JAK2 mutation, primarily JAK2V617F 8. The same mutation is also present in ∼60% of patients with ET and is associated, in ET, with a higher risk of thrombosis, compared to ET patients with CALR mutations or triple-negative mutational status (i.e. wild-type for JAK2, CALR, or MPL mutations) 9. The independent association between JAK2V617F and thrombosis was also highlighted in a recent large multicenter study of 891 patients with World Health Organization (WHO)-defined ET 10. The latter study 10 also identified cardiovascular (CV) risk factors, in addition to age >60 years, history of thrombosis, and presence of JAK2V617F, as an independent risk factor for thrombosis, which allowed the authors to develop a three-tiered risk model based on hazard ratio-derived risk scores that included age >60 years (1 point), thrombosis history (2 points), cardiovascular risk factors (1 point) and JAK2V617F (2 points): low-risk <2 points; intermediate-risk = 2 points; and high-risk ≥3 points) 10. This International Prognostic Scoring System for Thrombosis in ET (IPSET-thrombosis) was not further affected by the recently discovered CALR mutation 11, an observation that underlined the supposition that it is the absence of JAK2V617F and not the presence of CALR mutation that is responsible for the lower thrombosis risk in CALR-mutated patients, compared to JAK2-mutated cases. The prothrombotic property of JAK2V617F in ET was further queried in the context of arterial versus venous thrombosis 12; in this particular international study 12, independent predictors of arterial thrombosis were age >60 years, thrombosis history, CV risk factors (i.e. tobacco use, hypertension, or diabetes mellitus), leukocytosis, and JAK2V617F 12. The same risk factors, except leukocytosis, remained significant in JAK2-mutated ET 12. On the other hand, platelet count in excess of 1 million/μL was associated with a lower risk of arterial thrombosis whereas male gender was the only predictor of venous thrombosis 12. In another international study of PV involving 1,545 patients 13, arterial events were predicted by history of arterial thrombosis and hypertension and venous events by history of venous thrombosis and advanced age. Taken together, the above-mentioned international studies in PV and ET identify “history of thrombosis” as being the most important risk factor for recurrent vascular events and provide further clarification in terms of arterial- versus venous-specific risk factors. The studies also confirm advanced age as an independent risk factor for both arterial and venous thrombosis. It should be noted that these risk factors were evident despite the fact that the study population was receiving mostly standard therapy that included phlebotomy and low-dose aspirin in low risk PV, low-dose aspirin in low risk ET and cytoreductive therapy in high risk PV or ET. Therefore, it is reasonable to continue labeling ET or PV patients with advanced age (≥60 years) or thrombosis history as “high risk,” but with the caveat that conventional therapy might not be adequate in preventing recurrent thrombosis in such patients 10, 13. Current therapy was also relatively inadequate, as reflected by residual risk of thrombosis, in the treatment of conventionally-labeled low risk patients (i.e. age <60 years and no history of thrombosis), in the presence of JAK2 mutations or CV risk factors 10. Consistent with this observation, other studies have suggested the value of aspirin therapy in reducing the risk of arterial thrombosis in low risk ET associated with CV risk factors, but not otherwise 14. The potential value of aspirin therapy in the prevention of recurrent venous thrombosis was also suggested in the latter study 14 as well as another retrospective study, especially in combination with systemic anticoagulation 15. In this regard, it is relevant to recall non-MPN studies where aspirin therapy, after discontinuation of systemic anticoagulation, was shown to significantly reduce the risk of recurrent events in patients with first unprovoked venous thromboembolism, without increased bleeding risk 16. In a large overview of randomized trials 17, the anti-venous thrombosis effect of aspirin therapy was shown in both surgical and medical patients as well as in those patients who were concurrently receiving systemic anticoagulation 17. Based on the cumulative data from the aforementioned international studies in ET, it is reasonable to expect a lower thrombosis risk in currently-assigned “low risk” ET patients (i.e. age <60 years and no history of thrombosis) without JAK2 mutations or CV risk factors, compared to their counterparts with either JAK2 mutations or CV risk factors. In other words, the former (i.e. age <60 years, no history of thrombosis, JAK2-unmutated, and no CV risk factors) can now be separated from the latter (i.e. age <60 years, no history of thrombosis but with JAK2 mutations or CV risk factors) as having “very low risk” disease, in terms of thrombosis, and may not require any form of therapy, including aspirin therapy (Fig. 1). In the IPSET-thrombosis database, the annual rate of thrombosis was 0.59% in conventionally-labeled low-risk patients who were JAK2-unmutated versus 1.55% in JAK2-mutated otherwise low-risk patients 10. Contemporary treatment algorithm for polycythemia vera and essential thrombocythemia. Aspirin is used in the absence of treatment contraindications including clinically significant acquired von Willebrand syndrome. We recommend performing ristocetin cofactor activity in patients with over 1 million platelets per microliter and holding aspirin if the activity level is below 20%. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] In our newly assigned “low risk”, as opposed to “very low risk” disease category (Fig. 1), which includes both ET and PV patients who are young (age <60 years) and without history of thrombosis, but are either JAK2-mutated or harbor CV risk factors, we recommend once- or twice-daily low-dose aspirin therapy based on the presence of one or both risk factors for arterial thrombosis (i.e. JAK2V617F and CV risk factors) (Fig. 1). Our main rationale to use aspirin therapy in such patients is to prevent arterial events, an assumption supported by a controlled study in PV 3, but it is possible that such therapy would also help prevent recurrent venous thrombosis as was elaborated in detail in a previous paragraph 14, 15, 17. The rationale in considering twice-daily aspirin dosing in low risk JAK2-mutated patients with cardiovascular risk factors is based on (i) our multicenter study 10 that disclosed residual risk of thrombosis in otherwise low risk patients receiving conventional therapy, which suggested the inadequacy of once-daily aspirin therapy in a subset of patients with multiple risk factors for arterial thrombosis, and (ii) emerging data on the inadequacy of once-daily aspirin dosing for 24-h optimal suppression of thromboxane-A2 synthesis, in the presence of high platelet turnover, and demonstration of superior biological efficacy in ET with twice-daily dosing 18, 19. In high risk patients, preventative measures should be specific to the type of thrombosis that is relevant to the individual patient (i.e. arterial vs. venous) and also address the presence of other risk factors for arterial thrombosis (Fig. 1). Cytoreductive therapy, for prevention of thrombosis, is currently recommended in high-risk (age ≥60 years or presence of thrombosis history) patients with PV or ET, regardless of the type of thrombosis (i.e. arterial vs. venous), based on evidence from controlled studies 4, 5. In this regard, hydroxyurea is the evidence-based choice for first-line therapy whereas non-controlled studies have supported the use of interferon-α and busulfan as second-line drugs of choice 1. We do not use ruxolitinib in PV, despite its recent FDA-approval in hydroxyurea refractory cases 20, unless in the presence of severe pruritus or symptomatic splenomegaly that is shown to be refractory to adequate doses of hydroxyurea, interferon-α or busulfan. Additional therapy in high risk patients depends on the history of arterial or venous thrombosis and includes systemic anticoagulation in the presence of venous thrombosis history 15 and once-daily aspirin in the presence of only one risk factor for arterial thrombosis (i.e. age >60 years or history of arterial thrombosis or presence of JAK2V617F or CV risk factors) or twice-daily aspirin in the presence of two or more risk factors for arterial thrombosis (Fig. 1). However, we currently do not recommend twice-daily aspirin therapy in patients receiving systemic anticoagulation, because of concern for increased bleeding (Fig. 1). Otherwise, as discussed above, the combination of systemic anticoagulation and once-daily aspirin therapy, in patients with history of venous thrombosis, might enhance the protection against recurrent venous thrombosis, and might be justified in the presence of at least one risk factor for arterial thrombosis (Fig. 1) 15. Finally, our modified model of risk stratification and the accompanying risk-based treatment algorithm (Fig. 1) is not wholly based on hard evidence but indirectly surmised from recent, mostly retrospective, studies and few controlled clinical trials. As such, we welcome critical viewpoints on the subject matter and look forward to additional prospective studies that would either validate or refute our expertise-based recommendations. Examples of future prospective studies in this regard include (i) once- versus twice-daily aspirin therapy in ET or PV patients with risk factors for arterial thrombosis, (ii) systemic anticoagulation alone versus in combination with once-daily aspirin therapy in ET or PV patients with history of venous thrombosis, and (iii) observation alone versus once-daily aspirin therapy in “very low risk” ET.