Serious myeloproliferative reactions associated with the use of thalidomide in myelofibrosis with myeloid metaplasia

Ayalew Tefferi and Michelle A. Elliott

Current treatment in myelofibrosis with myeloid metaplasia (MMM) is inadequate: only a small proportion of patients derive benefit from bone marrow transplantation, splenectomy, or drug therapy.1 We have recently demonstrated a prognostically detrimental increase in bone marrow microvessel density in the majority of patients with MMM.2 Because thalidomide may inhibit angiogenesis,3 its therapeutic activity in MMM is worth investigating.

Before initiating a currently ongoing phase II treatment trial with thalidomide in MMM, we had piloted the drug in 6 patients (age range, 41-71 years; 3 males, 3 females) with symptomatic disease under a compassionate-use protocol approved by the institutional review board. Treatment in all patients was initiated at a daily oral dose of 200 mg, and further dose increments were not tolerated. The duration of treatment ranged from 9 days to 9 months. Of the 6 patients, 3 had concomitant treatment with a cytoreductive agent and were therefore not eligible for accurate assessment of the thalidomide effect on blood counts or spleen size. The remaining 3 patients were treated with thalidomide alone a minimum of 2 months after discontinuing previous specific therapy. The effect of treatment in these 3 patients is as follows.

One patient experienced severe upper left quadrant pain and fever that suggested a splenic infarct after only 9 days of treatment with thalidomide, without associated changes in blood count or spleen size. The second patient had a marked increase in both his previously stable peripheral platelet count (277-1082 × 109/L) and his leukocyte count (7.2-23.6 × 109/L) after only 20 days of treatment with thalidomide (200 mg/day) and required short-term therapy with hydroxyurea. The platelet count increased (152-440 × 109/L) also in the third patient during treatment with thalidomide and returned to baseline after cessation of therapy (9 months at 50 mg per day). The last-mentioned patient also had a durable increase in hemoglobin level (10.9-13.3 g/dL).

The preliminary observations from our current phase II study confirm probable drug-related steep increases in platelet and leukocyte counts in some patients. These changes occurred in the initial 4 to 8 weeks of treatment with thalidomide and were often associated with marked basophilia. We have documented the development of pericardial effusion secondary to extramedullary hematopoiesis in one patient. The mechanism of action and net effect of this biologic activity remain to be determined. But the potential for precipitation of serious disease-related complications exists, and the use of this drug outside a protocol setting is discouraged.