White blood cell counts and thrombosis in polycythemia vera: a subanalysis of the CYTO-PV study

Tiziano Barbui, Arianna Masciulli, Maria Rosa Marfisi, Giovanni Tognoni, Guido Finazzi, Alessandro Rambaldi and Alessandro Vannucchi

To the editor:

The pathogenesis of thrombosis in patients with polycythemia vera (PV) results from a complex interplay of patient- and disease-related variables. According to age and previous thrombosis, patients are traditionally stratified as “high risk” or “low risk.” However, recently, novel disease-related determinants such as leukocyte (white blood cell [WBC]) levels and JAK2V617F mutational burden have been proposed as new contributing predictors of vascular events and are now under active investigation.1 In the largest epidemiologic study (European Collaboration on Low-dose Aspirin), baseline leukocytosis emerged as an independent risk factor of arterial thrombosis,2 and since then, no confirmatory data have been produced. The Italian randomized trial Cytoreductive Therapy in Polycythemia Vera (CYTO-PV) tested different intensities of cytoreductive therapy (phlebotomy and cytoreductive drugs) to prevent thrombotic events and showed that the arm maintained at hematocrit (HCT) target <45% had a significant 4 times lower rate of cardiovascular death and major thrombosis than did the arm with a HCT target of 45% to 50%. Analysis also showed that during the follow-up, patients in the high-HCT group had significantly higher WBC counts than did those in the low-HCT group.3 As expected, the intensity of therapy (phlebotomy and hydroxyurea) was higher in the low-HCT than in high-HCT arm.3 During the follow-up (median, 28.9 ± 10.9 months), the median HCT level in the low-HCT group was 44.4% compared with 47.5% in the high-HCT group. Fatal and nonfatal cardiovascular events were registered in 28 patients (7.7%), corresponding to an incidence rate of 3.4 per 100 person-years. Platelet levels did not differ in the 2 arms, whereas WBC count persisted at significantly higher levels in the high-HCT group than in the low-HCT group (P < .001). To discern the relative merits of more stringent HCT control from the role of a lower WBC count to reduce the cardiovascular events, we carried out a multivariable time-dependent analysis4 by considering the level of WBC categorized into approximate quartiles and recorded in the last clinical visit before the thrombotic event. The study was approved by each Institutional Review Board. Results indicate that the risk of thrombosis was clearly increased in patients with WBC count >7 × 109/L, becoming statistically significant when WBC count was >11 × 109/L (Table 1).

Table 1

Time-dependent multivariable analysis on the risk of major thrombosis in CYTO-PV study (N = 365)

Therefore, this analysis discerns the thrombogenic role of WBC from that of HCT and corroborates other prospective and retrospective studies in myeloproliferative neoplasms. In an analysis of 21 887 serial full blood counts in the prospective Primary Thrombocythemia 1 cohort, which included 776 essential thrombocythemia patients, one-third of them previously treated with cytoreductive drugs, with a median follow-up of 36 months (range, 2-87), the risk of thrombosis, both arterial and venous, was not significantly associated with platelet count or hemoglobin level, whereas there was a significant association between WBC count and risk of thrombosis (P = .03).5 In another study of patients with World Health Organization–defined essential thrombocythemia, WBC >11 × 109/L was associated with arterial thrombosis only (hazard ratio [HR], 1.7; 95% CI, 1.01-2.72; P = .044).6 In primary myelofibrosis patients, the combination of a WBC count >15 × 109/L and the presence of the JAK2V617F mutation was associated with an increased risk of arterial and venous thrombosis (HR, 3.13; 95% CI, 1.26-7.81).7 Likewise, in early myelofibrosis, WBC count >11.2 × 109/L was associated with an increased risk of arterial thrombosis (HR, 1.12; 95% CI, 1.00-1.25).8

In conclusion, WBC count is implicated in the process of thrombogenesis in PV and should be evaluated in the therapy response.


Acknowledgments: This work was supported by a grant from Associazione Italiana per la Ricerca sul Cancro (AIRC-Milano) “Special Program Molecular Clinical Oncology 5x1000” to AIRC-Gruppo Italiano Malattie Mieloproliferative (T.B., A.V., A.R.).

Contribution: T.B., A.V., A.R., G.F., and G.T. designed the research, contributed patients, participated in data analysis and interpretation, and wrote the paper; and A.M. and M.R.M. performed statistical analysis.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Tiziano Barbui, Hematology and Research Foundation, Ospedale Papa Giovanni XXIII, Piazza OMS 1, 24127 Bergamo, Italy; e-mail: tbarbui{at}