Platelet count in essential thrombocythemia: the more the better?

To the editor:

To date, drug therapy in essential thrombocythemia (ET) has not been shown to positively influence the risk of leukemic or fibrotic transformation. In contrast, an often-cited controlled study¹  has demonstrated the antithrombotic value of hydroxyurea therapy in high-risk patients. In this particular study, “high-risk” disease was appropriately defined by the presence of advanced age (> 60 years) and/or history of thrombosis and did not include patients with a platelet count of more than 1500 × 10⁹/L.¹  There is no other “no-treatment arm” controlled study in ET that either supports the benefit of cytoreductive therapy or indicates a higher incidence of thrombotic complications in otherwise “low-risk” patients with extreme thrombocytosis. Similarly, there is limited information, in terms of hard evidence, that supports the theoretically appealing²  association between extreme thrombocytosis and non–aspirin associated bleeding.³  Instead, a recent retrospective study of low-risk ET patients with extreme thrombocytosis showed no benefit for cytoreductive therapy in terms of either thrombotic or hemorrhagic complications.³ 

Despite the above facts, the presence of extreme thrombocytosis in ET is considered by many as a recipe for vascular catastrophe and hence as an indication for cytoreductive therapy. This unsubstantiated concept has also been used to promote the therapeutic use of anagrelide, a platelet-specific cytoreductive agent. These are some of the reasons why the recent Blood paper by Carobbio et al has major relevance to current clinical practice.⁴  Among 1063 patients with ET, the authors demonstrated a significantly lower risk of thrombosis in patients presenting with extreme thrombocytosis, an effect that was independent of both age and thrombosis history. However, while I am comforted by this additional evidence for the lack of a direct correlation between extreme thrombocytosis and the risk of thrombosis in ET, I believe that the authors' claim for an inverse correlation requires further clarification.

To begin with, the authors clearly state in their paper that low-risk versus high-risk disease was distinguished by the absence in the former and presence in the latter of age 60 years or older and/or thrombosis history, and that low-risk patients were followed without chemotherapy. Although I personally subscribe to this management strategy, I find the authors' statement contrary to their previously published guidelines for the treatment of ET, where they recommended the use of cytoreductive drugs in low-risk patients with a platelet count of either more than 1500 × 10⁹/L, or 1000 to 1500 × 10⁹/L in the presence of cardiovascular symptoms.⁴  This is an important point that needs to be clarified, because the latter scenario would lead to a higher likelihood for “low-risk” patients with extreme thrombocytosis to be on cytoreductive treatment. Adjusting thrombosis risk for chemotherapy in such a setting does not necessarily address the potential confounding effect of treatment; it is often difficult to determine in a retrospective study whether or not treatment was instituted for primary or secondary prevention of thrombosis. Furthermore, from a therapeutically relevant standpoint, it would be most useful if the authors shared the outcome of their analysis restricted to “low-risk” patients.