Expanding the use of thrombopoietin mimetic drugs: what about chronic lymphocytic leukemia?

In 1994 the gene for thrombopoietin (TPO), the growth factor regulating megakaryocytopoiesis and platelet production, was cloned, and since then much effort has been invested in the construction of recombinant molecules with thrombopoietin-like structure and activity. The primary objective here was the development of an agent which could be effective in the treatment of chemotherapy-induced thrombocytopenia, thereby attempting to avoid the obvious side effects of platelet transfusions in these patients [1]. Initial results using the first generation of these compounds were quite disturbing, as these agents induced a paradoxical effect, causing thrombocytopenia instead of increases in platelet counts, because of the development of neutralizing antibodies cross-reacting with endogenous thrombopoietin [1].

In recent years a second generation of thrombopoietin drugs, which were less immunogenic small molecules, was developed. These included eltrombopag, an oral non-peptide molecule taken on a daily basis [2], and romiplostim, a subcutaneous peptidobody given once weekly [3]. Both agents were shown to be clinically effective treatment for patients with immune thrombocytopenic purpura (ITP) and hepatitis C-related thrombocytopenia [2,3], and were approved by the Food and Drug Administration (FDA) in 2008 [4]. These initial encouraging results led to their wider use in other settings including the myelodysplastic syndromes (MDS), inherited and neonatal thrombocytopenia, and chemotherapy-induced thrombocytopenia, and have also improved the platelet yield from healthy donors [5]. Other intriguing investigative applications include their use for ex vivo expansion of the CD34+ progenitor cell pool mobilized from the peripheral blood [6].

Based on the successful results obtained with TPO agents in idiopathic immune thrombocytopenias, it was not surprising that the use of these agents would be extended to the treatment of secondary or associated immune thrombocytopenia as encountered in patients with chronic lymphocytic leukemia (CLL). In this respect, the first report was from the M. D. Anderson group in Houston, which described the positive effect of eltrombopag on thrombocyte counts in two of three patients with CLL who had associated immune thrombocytopenia [7]. In this issue of Leukemia and Lymphoma, D'Arena and Cascavilla describe the first experience with romiplostim used in a 72-year-old patient with CLL and severe immune mediated thrombocytopenia, refractory to treatment with steroids, IV immunoglobulin, and rituximab [8]. Therapy with romiplostim achieved its target in this patient, increasing the platelet count to ≥50 000/µL during a follow-up period of 27 weeks.

These initial positive results in CLL [7,8] pose a new set of issues that will need to be addressed in the near future, and these include the first predictable questions relating to the significance of plasma thrombopoietin levels in CLL and the important issue of whether the drug affects the natural history of CLL itself. In this regard, it is of interest to note that the short follow-up in the single patient described here revealed no changes in the absolute B-CLL lymphocyte count [8]. However, it should be remembered that the follow-up was only 27 weeks. In this respect, one should also bear in mind that in one of the cases described by Koehrer et al. [7] a reduction in total lymphocyte count occurred at the same time as the platelet count increased during treatment with eltrombopag. In addition, in vitro data using leukemia cell lines treated with eltrombopag unexpectedly demonstrated reduced proliferative capacity of the cultured leukemic cells [9]. Plasma thrombopoietin levels have been analyzed in several studies involving patients with CLL. Koller et al. predicted a more aggressive clinical course in patients with CLL with higher levels of TPO, which appeared to correlate with the unmutated status of the immunoglobulin heavy chain gene [10]. In contrast, Molica et al. examined TPO levels in 71 previously untreated patients with early stage CLL and found that TPO levels did not correlate with the absolute peripheral blood lymphocyte count, clinical stage, platelet count, hemoglobin, and lactate dehydrogenase (LDH) and β₂-microglobulin levels, and concluded that there was no correlation between plasma TPO levels and other markers of disease activity [11]. So, when looking at all the data collectively, there is currently no evidence to suggest that thrombopoietin or ‘thrombopoietin mimetic agents’ affect the natural history of CLL, but obviously more experience with these agents is required before any final conclusions can be drawn.

It should also be noted that doses and frequency of administration of romiplostim may be different in different settings, and policy in this respect will have to be agreed upon in the future. In a study evaluating the effects of romiplostim in patients with MDS, higher doses (750 µg) were required, and once-weekly administration of the agent was as effective as twice-weekly [5]. Side effects associated with the use of TPO agents relate to the possible development of bone marrow fibrosis or the tendency to venous thrombosis [3]. However, to date, no association between CLL and myelofibrosis has been reported, while it is interesting to note that thrombotic complications have been reported to be less frequent in patients with CLL when compared to other malignancies [12]. Nevertheless, these aspects require further evaluation in a larger cohort of patients with a longer follow-up. Finally, there is also the issue of whether TPO affects the CLL microenvironment. In this respect lessons learned from patients with ITP indicate that treatment with thrombopoietin mimetic agents may affect the immune profile [13]. In these patients, plasma levels of transforming growth factor-β1 (TGF-β1) appeared to correlate with the degree of increase in platelet counts and with the improvement of T-cell regulatory activity [13].

In conclusion, it is clear that the development of effective TPO mimetic drugs is a meaningful advance and a large step forward in this field. We now need to expand our overall experience of when and how to use these agents. In this regard, CLL-associated immune thrombocytopenia may well be an appropriate model for a larger clinical study testing the possible benefit of these agents.