[Leukemic risk in thrombocythemia]

To define pathogenesis, incidence, clinical presentation and prognosis of leukemic transformation (LT) in patients with essential thrombocytemia (ET); to compare the incidence of LT in previously treated versus untreated patients; to search for risk factors of LT, with special reference to the iatrogenic risk of myelosuppressive drugs.

Most significant French and English-language papers published between 1981 and 1997 dealing with either ET, LT, clonality, cytogenetics, alkylating agents or hydroxyurea were reviewed. Patient characteristics, clinical presentation, bone marrow and cytogenetic findings, and survival data were extracted from each case report and review, and possible risk factors were analyzed.

Despite the existence of clonal hematopoiesis in up to 60-80% of patients with ET, suggesting an intrinsic potential for subsequent leukemia, spontaneous LT is poorly documented and the common presumption that TL occurs in excess in treated patients with ET is still questionable. The incidence of LT, probably underestimated in the past, varies from 0.7% to 5.3% in retrospective series, and from 8% to 12% in prospective studies. Recurrent cytogenetic abnormalities are significantly more frequent and complex in marrow karyotypes performed at the time of blastic crisis, as compared with those performed at diagnosis. A review of 62 reports of LT highlights some important characteristics: a) LT can occur anytime during the chronic phase of ET, especially within the first eight years, b) all subtypes of leukemia are described, notably unclassifiable cases and megakary oblastic acute leukemias, c) bone marrow fibrosis is frequent, as is a progressive onset of a myelodysplastic syndrome preceeding the blastic crisis, d) the prognosis of LT is very poor, two-third of all patients die within 6 months. Apart from the onset of a myelodysplastic syndrome, which carries a poor outcome, there is no well-established risk factor of LT. The leukemogenic risk of myelosuppressive therapies, clearly demonstrated for years in polycythemia vera, cannot be directly ascertained in patients with ET, but is suggested by various indirect observations. This putative risk of inducing secondary leukemia affects not only alkylating agents but also hydroxyurea.

LT is a highly serious and largely unpredictable complication of ET. The current debate on the specific risk of myelosuppressive treatments deserves large-scale prospective randomized studies. Until this crucial point is elucidated, any cytotoxic therapy should be reserved to elderly and/or symptomatic patients and to those having other vascular risk factors.