Reply to Mcmullin et al

Abstract

I thank McMullin et al for the opportunity to allay their concerns and thereby avoid perpetuating suboptimal polycythaemia vera (PV) management. First, with respect to bone marrow examination as a diagnostic test for PV, bone marrow morphology is no surrogate for clonality, and marrow morphology cannot distinguish PV from its companion myeloproliferative disorders. Furthermore, a cytogenetic abnormality will be present in only 25% of PV patients at diagnosis (Najfeld et al, 2002). As there are many potential causes for erythrocytosis other than PV, one can estimate that bone marrow examinations on 10 patients would be needed to obtain one with a cytogenetic abnormality, hardly a cost-effective endeavour. Furthermore, some PV patients have normal marrow morphology at diagnosis (Ellis et al, 1986). Finally, contrary to the contention of McMullin et al, the presence of myelofibrosis at diagnosis, like a cytogenetic abnormality, has no prognostic significance (Ellis et al, 1986). Second, the specious reasoning employed by McMullin et al to justify treating asymptomatic thrombocytosis in PV supports my contention that no other medical condition has caused more otherwise astute clinicians to suspend disbelief than thrombocytosis. The haematocrit, not the platelet count, is the major determinant of blood viscosity and the high thrombosis rate in the phlebotomy arm of the Polycythemia Vera Study Group (PVSG) trial (PVSG-01) was the direct result of inadequate phlebotomy therapy. Importantly, neither the PVSG (Berk et al, 1986) nor others (i.e. Wehmeier et al, 1991) have found a correlation between the platelet count and thrombosis in PV, and antiplatelet therapy was no panacea against thrombosis either. It is also bogus to cite survival data for untreated PV patients to justify platelet count reduction. There are no published data demonstrating that platelet count reduction prolongs survival in PV or even in essential thrombocytosis. Third, with respect to phlebotomy causing myelofibrosis, McMullin et al unfortunately have citation amnesia. Myelofibrosis is part of the natural history of PV and there are five studies, including one from the UK (Messinezy et al, 1985), that refute the contention that it is a phlebotomy stigma; certainly hydroxyurea could not prevent it. McMullin et al should also ponder why phlebotomy would cause myelofibrosis in a disease in which marrow function is autonomous but not in a disease such as haemochromatosis, where it is not. Fourth, with respect to hydroxyurea, I would remind McMullin et al that the introduction of cytotoxic agents to control bone marrow function in PV has inevitably been associated with recognition of their leukaemogenicity. There is a growing literature that suggests hydroxyurea may be no different (Weinfeld et al, 1994; Sterkers et al, 1998) and in the absence of proof of its safety, its use should be restricted to situations where its benefits outweigh its risks. Fifth, with respect to the use of epsilon aminocaproic acid (EACA), McMullin et al should not be averse to learning new treatment strategies for PV. Extreme thrombocytosis is associated with a reduction in large von Willebrand factor multimers (Budde et al, 1993). EACA has been used for decades to reduce mucous membrane bleeding in von Willebrand's disease and is useful for controlling bleeding in PV, for the same reason. Sixth, I fail to understand what could be controversial about keeping the haematocrit less than 0·36 during pregnancy in PV. Normally during pregnancy, the plasma volume expands disproportionately to the red cell mass, producing the so-called ‘physiological anaemia’. Therefore, a normal haematocrit in a pregnant PV patient must indicate the presence of an expanded red cell mass. To ignore this is to place both the patient and her fetus at risk of thrombosis, and ignoring that risk, not the removal of it, is what should be controversial. Finally, I am pleased that McMullin et al and I have a common ground for agreement as I never suggested that anagrelide should be employed during pregnancy.