- Email: [email protected]
What is the role of novel thrombopoietic agents in the management of acute leukemia?
Best Practice & Research Clinical Haematology 29 (2016) 372e378
Contents lists available at ScienceDirect
Best Practice & Research Clinical Haematology journal homepage: www.elsevier.com/locate/beha
What is the role of novel thrombopoietic agents in the management of acute leukemia? David J. Kuter a, b, * a b
Massachusetts General Hospital, Boston, MA 02114, USA Harvard Medical School, Boston, MA 02114, USA
a b s t r a c t Keywords: Eltrombopag Leukemia rHuMGDF Romiplostim Thrombopoietin Supportive care TPO MDS
The role of novel thrombopoietic agents in the management of acute leukemia is a tale of two molecules, romiplostim and eltrombopag. Both are thrombopoietin (TPO) receptor agonists with somewhat different mechanisms of action. Romiplostim is a peptide TPO receptor agonist that activates the TPO receptor by binding to it just like TPO. Eltrombopag is a nonpeptide TPO receptor agonist that activates the TPO receptor by binding to the transmembrane domain. Both TPO receptor agonists increase platelet counts in healthy humans and in those with immune thrombocytopenia. This review focuses on the potential these agents may have in supportive care of patients with acute leukemia. © 2016 Elsevier Ltd. All rights reserved.
Introduction Three questions arise when considering supportive care for patients with acute myeloid leukemia (AML). First, might a thrombopoietic agent help raise platelet counts in patients during induction or consolidation therapy thereby improving platelet recovery and decreasing bleeding? Second, in patients not receiving chemotherapy for myelodysplastic syndrome (MDS) or AML, can supportive care of this type decrease thrombocytopenia and minimize bleeding and transfusions? And third, do these drugs have an effect on AML blasts? The discussion below revolves around whether the two currently available TPO receptor agonists, romiplostim and eltrombopag, could play a role in the supportive care of patients with AML.
* Massachusetts General Hospital,, Boston, MA 02114, USA. Fax: þ1 617 724 6801. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.beha.2016.10.013 1521-6926/© 2016 Elsevier Ltd. All rights reserved.
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
373
Romiplostim Romiplostim (AMG 531, Nplate®, Romiplate®) is known to increase platelet counts in healthy individuals and in patients with immune thrombocytopenia. It consists of an IgG heavy chain into which 4 identical peptides have been added [1]. The peptide contains 14 amino acids that bind the thrombopoietin (TPO) receptor and activate it. Romiplostim binds the distal cytokine homology region of the receptor just like TPO, where it activates the JAK and STAT pathways, the MAP kinase pathway, and a wide range of anti-apoptotic pathways, thereby increasing megakaryocyte growth and viability. Romiplostim is a very potent thrombopoietic agent, with a half-life of 140 h. Romiplostim should be considered identical to recombinant TPO but with two minor exceptions. First, it binds to the TPO receptor at exactly the same site as TPO (Fig. 1) but with about one quarter of the avidity of recombinant TPO. Second, romiplostim has a half-life about 3-fold greater than recombinant TPO. Neither of these properties has been shown to affect any clinical outcomes and indeed they may compensate for each other. In healthy humans, romiplostim administration has no effect on the platelet count for the first 5 days after treatment [2]. During this time period, megakaryocytes increase in number, size, and ploidy and then on day 5 start to shed platelets in a very rapid fashion. Romiplostim is a potent stimulator of platelet production and in healthy volunteers platelet counts of 1 or 2 million can easily be attained [2]. So can this molecule stimulate early recovery of megakaryocyte precursors and raise the platelet count in patients with acute leukemia or late stage MDS? Romiplostim and induction/consolidation chemotherapy in acute leukemia No study has examined whether romiplostim can raise platelet counts in AML patients, but given its virtual identify with TPO it pays to review prior studies with recombinant TPO molecules. A number of studies with recombinant forms of TPOdpegylated recombinant human megakaryocyte growth and development factor (PEG-rHuMGDF) and recombinant TPO (rhTPO)dhad been conducted many years
Fig. 1. TPO receptor agonist mechanism of action. Romiplostim and eltrombopag bind the TPO receptor at different sites. Romiplostim binds the TPO receptor at the same distal cytokine homology region as does TPO and eltrombopag binds in the transmembrane region.
374
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
ago in leukemia patients undergoing induction and consolidation chemotherapy [3]. Patients with AML who received induction therapy for 7 days were given various combinations of doses of PEG-rHuMGDF and compared to placebo. The drug did not enhance recovery of platelets to 20,000/mL, did not decrease platelet transfusions, and did not result in earlier platelet transfusion independence. The agent was well tolerated, had no effect on the chemotherapy response rate, no effect on other blood counts, and did not alter blast counts. There was, however, a rapid, dose-dependent, rebound thrombocytosis after platelet recovery. This rebound thrombocytosis was used to advantage in a different study [4]. Investigators collected the excess platelets, cryopreserved them, and transfused them in subsequent cycles into patients who were getting consolidation chemotherapy who had become alloimmunized. Why didn't PEG-rHuMGDF enhance platelet recovery in AML patients? When a patient becomes thrombocytopenic as a result of chemotherapy, the endogenous TPO levels rise quite dramatically. There is an inverse relationship between TPO levels and platelet counts, and at the platelet count nadir in myeloid leukemia patients, the TPO levels are 20e50 times above basal levels [5]. Therefore, PEGrHuMGDF might not have worked very well because of the already high levels of TPO; making them higher might not make much difference [6]. Since romiplostim functions in a way identical to PEGrHuMGDF, this reviewer suggests that use of romiplostim in AML induction and consolidation would show a similar lack of benefit.
Romiplostim in patients with MDS or acute leukemia not receiving chemotherapy No study has been conducted examining whether romiplostim can support AML patients not receiving chemotherapy. However, there have been recent publications of romiplostim use as supportive care in thrombocytopenic MDS patients, which may shed light on its potential use in AML patients [7e9]. In one study [9], MDS patients with low-risk or intermediate-1-risk disease and platelet counts <20,000/mL or 20,000/mL with bleeding were randomized to receive either romiplostim or placebo. There was a significant platelet response [37% compared with 4% for placebo (P < 0.001)] and transfusion events were significantly less with romiplostim [984 vs 1014 per 100 pt-yr (P < 0.001)]. But the primary endpoint of the study, the number of clinically significant bleeding events (CSBE) per 100 patient-year (pt-yr), was lower with romiplostim but not statistically different from control (1.47 vs 1.94; HR ¼ 0.83; CI: 0.66e1.05; P ¼ 0.13). However, prophylactic platelet transfusions can obfuscate results of studies with bleeding endpoints. Subsequent subgroup analysis stratified patients by platelet counts <20,000/mL and 20,000/ mL, since those with the lower count tend to receive more prophylactic transfusions. The investigators found that there were significantly fewer transfusions per 100 pt-yr [1251 vs 1779; RR, 0.71 (0.61e0.82); P < 0.0001] in the low platelet group receiving romiplostim, while the CSBE/100 pt-yr did not change [515 vs 501; RR, 1.03 (0.79e1.35)]. In those patients whose platelet counts were 20,000/mL, prophylactic platelet transfusions were given infrequently and the CSBE/100 pt-yr was significantly improved with romiplostim [80 vs 226; RR, 0.35 (0.21e0.59); P < 0.0001] (Fig. 2). In this subset analysis, the effect of the drug alone raised platelet counts and reduced significant bleeding events, suggesting there may be a use for romiplostim in supportive care in this context.
Romiplostim and blast counts The MDS trial described above [9] was unfortunately stopped early because in the preliminary analysis, 6.1% of the romiplostim-treated patients seemingly progressed to AML versus 2.8% in the placebo arm. However, in the final analysis there was a slight increase in progression to AML in the romiplostim arm that was not clinically significant [6.0% vs 4.9%; HR, 1.20; 95% CI, 0.38e3.84]. But circulating blasts increased to >10% in 14.9% of romiplostim-treated patients and 3.7% in placebo patients, suggesting that romiplostim does stimulate blast counts when given in this context. Follow-up data was available in 14 of 16 patients with increased blasts, and in all 14 the blasts decreased upon discontinuation of romiplostim.
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
375
Fig. 2. Subgroup analysis of MDS patients with lower or higher platelet levels receiving romiplostim or placebo. These results suggest a possible role for romiplostim in supportive care in thrombocytopenic patients not receiving chemotherapy and with platelet counts 20,000/mL. Figure created from data in Ref. [9]. Abbreviations: CSBE, clinically significant bleeding events; NS, not significant; P, placebo; R, romiplostim; Plt txf, platelet transfusions; pt-y, patient-years.
Eltrombopag Eltrombopag (SB-497115, Promacta®, Revolade®) is a small molecule TPO receptor agonist, dosed once a day orally [10], that stimulates megakaryocyte proliferation and increases platelet counts [11]. It is active in humans and chimps but in no other species [12]. Eltrombopag is also a potent iron chelator [13]. Eltrombopag is not as robust as romiplostim at increasing platelet counts in healthy subjects. After maximal dosing, eltrombopag produced a platelet count rise of ~150,000/mL over baseline [14], compared to ~1.6 million/mL with romiplostim [2]. This difference in effect is explained in part by the finding that eltrombopag binds to a different area of the TPO receptor than does romiplostim (Fig. 1). Eltrombopag binds to the transmembrane region where species-specific differences at residue 499, histidine in humans and chimps but leucine in most other species, determine its biological effect. Upon binding, eltrombopag activates downstream signal transduction pathways differently than recombinant TPO or romiplostim: there is less phosphorylation of STAT pathways and no activation of AKT pathways [15]. Nevertheless, eltrombopag increases megakaryocyte growth and platelet production. Although differing in mechanism of action and seemingly less potent than romiplostim, eltrombopag has a major attribute that distinguishes it from romiplostim: the thrombopoietic effect of eltrombopag is additive to the effects of TPO and may in some settings may be synergistic with it [16]. Using a TPO-dependent cell line (Fig. 3), cells proliferate maximally with TPO but growth with eltrombopag alone is less robust. However, when TPO and eltrombopag are combined, cell growth is increased well above that with TPO alone [16]. Eltrombopag may be considered a potentiator of TPO in terms of acting in situations in which TPO levels are already high. This unique effect of eltrombopag is best illustrated in the treatment of aplastic anemia [17], for which eltrombopag was recently approved by the FDA. Patients with aplastic anemia (AA) have extremely high TPO levels [3000e5000 pg/mL (normal values < 100 pg/mL)]. When 25 AA patients were treated with eltrombopag, 11 (44%) had responses in one cell line (9 increased platelets; 6 increased hemoglobin; 9 increased neutrophils) with 6 (24%) having trilineage responses [17]. Some responders also experienced increased bone marrow cellularity that persisted upon stopping the drug. The exact mechanism here is unclear but this data suggests an additive or synergistic effect of eltrombopag in a high TPO environment, something that would not be expected for a recombinant TPO or romiplostim. Eltrombopag and induction/consolidation chemotherapy in acute leukemia In leukemia patients, 5 different clinical trials are underway or recently completed investigating whether eltrombopag can potentiate the high levels of TPO in this setting (or via a separate
376
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
Fig. 3. Eltrombopag effect is additive to that of TPO on growth of megakaryocyte cell line. When recombinant human TPO and eltrombopag are administered together, a TPO-dependent cell line grows considerably more than when either agent is added alone. Created from data in Ref. [16].
mechanism) and amplify platelet recovery during induction or consolidation (Table 1). No preliminary data are yet available from these trials and it should be recalled (vide supra) that the addition of supraphysiological doses of a recombinant TPO failed to hasten recovery in these settings. Eltrombopag and blast counts Given the initial concern that arose that romiplostim increased the AML transformation rate and increased the blast counts in patients with MDS [9], initial studies with eltrombopag in this patient group sought first to determine if eltrombopag had a similar effect on myeloid blasts. The results were quite striking and contrary to the romiplostim effect. After addition of eltrombopag to cultures of bone marrow from healthy or AML/MDS patients, the healthy marrow cells grew but the MDS/AML cells died [18]. Addition of eltrombopag at physiologically obtainable concentrations to AML cell lines resulted in rapid cell death, and this effect was independent of the presence of the TPO receptor [19]. The inhibitory effect of eltrombopag on leukemia cells in both primary cell lines and tissue culture lines seemed to occur by arresting cells in G1 and increasing differentiation and was directly related to the ability of eltrombopag to reduce the concentration of intracellular iron [13]. Further studies have shown that by lowering iron, eltrombopag reduces reactive oxygen species in AML cells thereby
Table 1 Clinical trials of eltrombopag in AML. Trial identifier
Patient population
Trial type
Objective
NCT01656252 (Completed March 2016) NCT01550185 (Study terminated) NCT01890746
Adult AML patients 18e70 years during consolidation and in complete remission Relapsed/refractory AML patients 18 years
Phase 1, 2
Optimal tolerated dose; increase in platelet count recovery Highest safe dose, safety, efficacy
NCT02071901 NCT02446145
AML patients 18 years receiving standard induction therapy AML patients 60 years receiving induction therapy AML patients 65 years of age not eligible for intensive chemotherapy
Phase 1
Phase 2
Safety and efficacy
Phase 2
Improve platelet recovery
Phase 2, randomized, placebocontrolled, decitabine with or without eltrombopag
Safety and efficacy
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
377
resulting in their death [20]. In a murine leukemia model, eltrombopag was shown to reduce leukemia cell growth and prolong survival [13]. This potential anti-leukemia effect may also occur in humans with AML. A patient with nucleophosmin 1 mutated AML and a high bone marrow blast count was treated for 2 months with eltrombopag and had a marked, but short-lived, decrease in blast count but no effect on peripheral blood counts [21]. Eltrombopag in patients with MDS or acute leukemia not receiving chemotherapy A recent study examined the safety and platelet count effect of eltrombopag versus placebo in 98 patients with advanced MDS or AML but not receiving chemotherapy [22]. There appeared to be no effect on peripheral blood or bone marrow blast counts and no effect upon overall progression-free survival or overall survival. However there was also no significant effect on platelet counts (eltrombopag: 17,000/mL; placebo: 12,000/mL), grade 3 bleeds (eltrombopag: 16%; placebo: 26%), or RBC or platelet transfusions. Conclusion The role of the TPO receptor agonists in the treatment of AML is unclear. What is seemingly established is that romiplostim does have the potential to increase myeloid blast counts. Whether this reversible effect is clinically meaningful is untested, but there was no adverse effect on remission rates in earlier trials of recombinant TPOs in AML. In contrast, eltrombopag appears not to increase blast counts, and may actually have a modest anti-leukemia effect. In patients undergoing AML induction or consolidation treatment, endogenous TPO levels are markedly elevated and this may explain why the addition of more TPO in the form of recombinant TPOs in prior trials had no significant effect on platelet counts, platelet transfusion, or time to platelet recovery. This reviewer predicts a similar outcome with romiplostim in AML. However, eltrombopag has demonstrated an ability to raise platelet counts in the presence of high TPO levels (eg, aplastic anemia) and the results of ongoing trials will test whether this hypothesis is clinically valid. Finally, some patients with advanced MDS or leukemia are treated with supportive care alone. In such patients, severe thrombocytopenia is a common problem. So far the data seem clear as to the potential role for TPO receptor agonists. Trials with eltrombopag show that it is safe but probably not effective. In contrast, romiplostim may raise the platelet count and reduce bleeding in patients with platelet counts 20,000/mL, but with concern over an increased blast count. This narrow window of benefit for some patients may be clinically meaningful, and further studies to better define use in this patient population should be conducted. Disclosures Consultant: Amgen, GSK, Eisai, ONO, 3SBIO, Pfizer, Merck, BMS, Syntimmune, Suppremol, BRL, Protalix, Zafgen, Rigel, Ligand, Kirin, Novartis, MedImmune, Genzyme, Alnylam, Caremark. Research support BMS, Protalix, Rigel. Stock holdings None. Conflict of interest None.
378
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
References [1] Molineux G. The development of romiplostim for patients with immune thrombocytopenia. Ann N. Y Acad Sci 2011;1222: 55e63. [2] Wang B, Nichol JL, Sullivan JT. Pharmacodynamics and pharmacokinetics of AMG 531, a novel thrombopoietin receptor ligand. Clin Pharmacol Ther 2004;76:628e38. [3] Archimbaud E, Ottmann OG, Yin JA, Lechner K, Dombret H, Sanz MA, et al. A randomized, double-blind, placebo-controlled study with pegylated recombinant human megakaryocyte growth and development factor (PEG-rHuMGDF) as an adjunct to chemotherapy for adults with de novo acute myeloid leukemia. Blood 1999;94:3694e701. [4] Vadhan-Raj S, Kavanagh JJ, Freedman RS, Folloder J, Currie LM, Bueso-Ramos C, et al. Safety and efficacy of transfusions of autologous cryopreserved platelets derived from recombinant human thrombopoietin to support chemotherapyassociated severe thrombocytopenia: a randomised cross-over study. Lancet 2002;359:2145e52. [5] Usuki K, Tahara T, Iki S, Endo M, Osawa M, Kitazume K, et al. Serum thrombopoietin level in various hematological diseases. Stem Cells 1996;14:558e65. [6] Kuter DJ. What is the potential for thrombopoietic agents in acute leukemia? Best Pract Res Clin Haematol 2011;24:553e8. [7] Kantarjian HM, Giles FJ, Greenberg PL, Paquette RL, Wang ES, Gabrilove JL, et al. Phase 2 study of romiplostim in patients with low- or intermediate-risk myelodysplastic syndrome receiving azacitidine therapy. Blood 2010;116:3163e70. [8] Giagounidis A, Mufti CJ, Kantarjian HM, Fenaux P, Sekkeres MA, Szer J, et al. Treatment with the thrombopoietin (TPO)receptor agonist romiplostim in thrombocytopenic patients (Pts) with low or intermediate-1 (Int-1) risk myelodysplastic syndrome (MDS): results of a randomized, double-blind, placebo (PBO)-controlled study. Blood ASH Annu Meet 2011;118: 117 [abstr)]. [9] Giagounidis A, Mufti GJ, Fenaux P, Sekeres MA, Szer J, Platzbecker U, et al. Results of a randomized, double-blind study of romiplostim versus placebo in patients with low/intermediate-1-risk myelodysplastic syndrome and thrombocytopenia. Cancer 2014;120:1838e46. [10] Erickson-Miller C, Delorme E, Giampa L, Hopson C, Valoret E, Tian S-S, et al. Biological activity and selectivity for Tpo Receptor of the orally bioavailable, small molecule Tpo Receptor agonist, SB-497115. Blood ASH Annu Meet 2004;104:2912 [abstr)]. [11] Luengo JI, Duffy KJ, Shaw AN, Delorme E, Wiggall KJ, Giampa L, et al. Discovery of SB-497115, a small-molecule thrombopoietin (TPO) receptor agonist for the treatment of thrombocytopenia. Blood ASH Annu Meet 2004;104:2910 [abstr)]. [12] Erickson-Miller CL, Delorme E, Iskander M, Giampa L, Hopson CB, Luengo J, et al. Species specificity and receptor domain interaction of a small molecule TPO receptor agonist. Blood ASH Annu Meet 2004;104:2909 [abstr)]. [13] Roth M, Will B, Simkin G, Narayanagari S, Barreyro L, Bartholdy B, et al. Eltrombopag inhibits the proliferation of leukemia cells via reduction of intracellular iron and induction of differentiation. Blood 2012;120:386e94. [14] Jenkins JM, Williams D, Deng Y, Uhl J, Kitchen V, Collins D, et al. Phase 1 clinical study of eltrombopag, an oral, nonpeptide thrombopoietin receptor agonist. Blood 2007;109:4739e41. [15] Erhardt JA, Erickson-Miller CL, Aivado M, Abboud M, Pillarisetti K, Toomey JR. Comparative analyses of the small molecule thrombopoietin receptor agonist eltrombopag and thrombopoietin on in vitro platelet function. Exp Hematol 2009;37: 1030e7. [16] Erickson-Miller CL, Delorme E, Tian SS, Hopson CB, Landis AJ, Valoret EI, et al. Preclinical activity of eltrombopag (SB497115), an oral, nonpeptide thrombopoietin receptor agonist. Stem Cells 2009;27:424e30. [17] Olnes MJ, Scheinberg P, Calvo KR, Desmond R, Tang Y, Dumitriu B, et al. Eltrombopag and improved hematopoiesis in refractory aplastic anemia. N Engl J Med 2012;367:11e9. [18] Will B, Kawahara M, Luciano JP, Bruns I, Parekh S, Erickson-Miller CL, et al. Effect of the nonpeptide thrombopoietin receptor agonist Eltrombopag on bone marrow cells from patients with acute myeloid leukemia and myelodysplastic syndrome. Blood 2009;114:3899e908. [19] Erickson-Miller CL, Kirchner J, Aivado M, May R, Payne P, Chadderton A. Reduced proliferation of non-megakaryocytic acute myelogenous leukemia and other leukemia and lymphoma cell lines in response to eltrombopag. Leuk Res 2010; 34:1224e31. [20] Kalota A, Selak MA, Garcia-Cid LA, Carroll M. Eltrombopag modulates reactive oxygen species and decreases acute myeloid leukemia cell survival. PLoS One 2015;10:e0126691. [21] Platzbecker U, Sockel K, Schonefeldt C, Nowak D, Helas S, Rollig C, et al. Induction of short-term remission with single agent eltrombopag in refractory nucleophosmin-1-mutated acute myeloid leukemia. Haematologica 2014;99:e247e8. [22] Platzbecker U, Wong RS, Verma A, Abboud C, Araujo S, Chiou TJ, et al. Safety and tolerability of eltrombopag versus placebo for treatment of thrombocytopenia in patients with advanced myelodysplastic syndromes or acute myeloid leukaemia: a multicentre, randomised, placebo-controlled, double-blind, phase 1/2 trial. Lancet Haematol 2015;2:e417e26.
Contents lists available at ScienceDirect
Best Practice & Research Clinical Haematology journal homepage: www.elsevier.com/locate/beha
What is the role of novel thrombopoietic agents in the management of acute leukemia? David J. Kuter a, b, * a b
Massachusetts General Hospital, Boston, MA 02114, USA Harvard Medical School, Boston, MA 02114, USA
a b s t r a c t Keywords: Eltrombopag Leukemia rHuMGDF Romiplostim Thrombopoietin Supportive care TPO MDS
The role of novel thrombopoietic agents in the management of acute leukemia is a tale of two molecules, romiplostim and eltrombopag. Both are thrombopoietin (TPO) receptor agonists with somewhat different mechanisms of action. Romiplostim is a peptide TPO receptor agonist that activates the TPO receptor by binding to it just like TPO. Eltrombopag is a nonpeptide TPO receptor agonist that activates the TPO receptor by binding to the transmembrane domain. Both TPO receptor agonists increase platelet counts in healthy humans and in those with immune thrombocytopenia. This review focuses on the potential these agents may have in supportive care of patients with acute leukemia. © 2016 Elsevier Ltd. All rights reserved.
Introduction Three questions arise when considering supportive care for patients with acute myeloid leukemia (AML). First, might a thrombopoietic agent help raise platelet counts in patients during induction or consolidation therapy thereby improving platelet recovery and decreasing bleeding? Second, in patients not receiving chemotherapy for myelodysplastic syndrome (MDS) or AML, can supportive care of this type decrease thrombocytopenia and minimize bleeding and transfusions? And third, do these drugs have an effect on AML blasts? The discussion below revolves around whether the two currently available TPO receptor agonists, romiplostim and eltrombopag, could play a role in the supportive care of patients with AML.
* Massachusetts General Hospital,, Boston, MA 02114, USA. Fax: þ1 617 724 6801. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.beha.2016.10.013 1521-6926/© 2016 Elsevier Ltd. All rights reserved.
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
373
Romiplostim Romiplostim (AMG 531, Nplate®, Romiplate®) is known to increase platelet counts in healthy individuals and in patients with immune thrombocytopenia. It consists of an IgG heavy chain into which 4 identical peptides have been added [1]. The peptide contains 14 amino acids that bind the thrombopoietin (TPO) receptor and activate it. Romiplostim binds the distal cytokine homology region of the receptor just like TPO, where it activates the JAK and STAT pathways, the MAP kinase pathway, and a wide range of anti-apoptotic pathways, thereby increasing megakaryocyte growth and viability. Romiplostim is a very potent thrombopoietic agent, with a half-life of 140 h. Romiplostim should be considered identical to recombinant TPO but with two minor exceptions. First, it binds to the TPO receptor at exactly the same site as TPO (Fig. 1) but with about one quarter of the avidity of recombinant TPO. Second, romiplostim has a half-life about 3-fold greater than recombinant TPO. Neither of these properties has been shown to affect any clinical outcomes and indeed they may compensate for each other. In healthy humans, romiplostim administration has no effect on the platelet count for the first 5 days after treatment [2]. During this time period, megakaryocytes increase in number, size, and ploidy and then on day 5 start to shed platelets in a very rapid fashion. Romiplostim is a potent stimulator of platelet production and in healthy volunteers platelet counts of 1 or 2 million can easily be attained [2]. So can this molecule stimulate early recovery of megakaryocyte precursors and raise the platelet count in patients with acute leukemia or late stage MDS? Romiplostim and induction/consolidation chemotherapy in acute leukemia No study has examined whether romiplostim can raise platelet counts in AML patients, but given its virtual identify with TPO it pays to review prior studies with recombinant TPO molecules. A number of studies with recombinant forms of TPOdpegylated recombinant human megakaryocyte growth and development factor (PEG-rHuMGDF) and recombinant TPO (rhTPO)dhad been conducted many years
Fig. 1. TPO receptor agonist mechanism of action. Romiplostim and eltrombopag bind the TPO receptor at different sites. Romiplostim binds the TPO receptor at the same distal cytokine homology region as does TPO and eltrombopag binds in the transmembrane region.
374
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
ago in leukemia patients undergoing induction and consolidation chemotherapy [3]. Patients with AML who received induction therapy for 7 days were given various combinations of doses of PEG-rHuMGDF and compared to placebo. The drug did not enhance recovery of platelets to 20,000/mL, did not decrease platelet transfusions, and did not result in earlier platelet transfusion independence. The agent was well tolerated, had no effect on the chemotherapy response rate, no effect on other blood counts, and did not alter blast counts. There was, however, a rapid, dose-dependent, rebound thrombocytosis after platelet recovery. This rebound thrombocytosis was used to advantage in a different study [4]. Investigators collected the excess platelets, cryopreserved them, and transfused them in subsequent cycles into patients who were getting consolidation chemotherapy who had become alloimmunized. Why didn't PEG-rHuMGDF enhance platelet recovery in AML patients? When a patient becomes thrombocytopenic as a result of chemotherapy, the endogenous TPO levels rise quite dramatically. There is an inverse relationship between TPO levels and platelet counts, and at the platelet count nadir in myeloid leukemia patients, the TPO levels are 20e50 times above basal levels [5]. Therefore, PEGrHuMGDF might not have worked very well because of the already high levels of TPO; making them higher might not make much difference [6]. Since romiplostim functions in a way identical to PEGrHuMGDF, this reviewer suggests that use of romiplostim in AML induction and consolidation would show a similar lack of benefit.
Romiplostim in patients with MDS or acute leukemia not receiving chemotherapy No study has been conducted examining whether romiplostim can support AML patients not receiving chemotherapy. However, there have been recent publications of romiplostim use as supportive care in thrombocytopenic MDS patients, which may shed light on its potential use in AML patients [7e9]. In one study [9], MDS patients with low-risk or intermediate-1-risk disease and platelet counts <20,000/mL or 20,000/mL with bleeding were randomized to receive either romiplostim or placebo. There was a significant platelet response [37% compared with 4% for placebo (P < 0.001)] and transfusion events were significantly less with romiplostim [984 vs 1014 per 100 pt-yr (P < 0.001)]. But the primary endpoint of the study, the number of clinically significant bleeding events (CSBE) per 100 patient-year (pt-yr), was lower with romiplostim but not statistically different from control (1.47 vs 1.94; HR ¼ 0.83; CI: 0.66e1.05; P ¼ 0.13). However, prophylactic platelet transfusions can obfuscate results of studies with bleeding endpoints. Subsequent subgroup analysis stratified patients by platelet counts <20,000/mL and 20,000/ mL, since those with the lower count tend to receive more prophylactic transfusions. The investigators found that there were significantly fewer transfusions per 100 pt-yr [1251 vs 1779; RR, 0.71 (0.61e0.82); P < 0.0001] in the low platelet group receiving romiplostim, while the CSBE/100 pt-yr did not change [515 vs 501; RR, 1.03 (0.79e1.35)]. In those patients whose platelet counts were 20,000/mL, prophylactic platelet transfusions were given infrequently and the CSBE/100 pt-yr was significantly improved with romiplostim [80 vs 226; RR, 0.35 (0.21e0.59); P < 0.0001] (Fig. 2). In this subset analysis, the effect of the drug alone raised platelet counts and reduced significant bleeding events, suggesting there may be a use for romiplostim in supportive care in this context.
Romiplostim and blast counts The MDS trial described above [9] was unfortunately stopped early because in the preliminary analysis, 6.1% of the romiplostim-treated patients seemingly progressed to AML versus 2.8% in the placebo arm. However, in the final analysis there was a slight increase in progression to AML in the romiplostim arm that was not clinically significant [6.0% vs 4.9%; HR, 1.20; 95% CI, 0.38e3.84]. But circulating blasts increased to >10% in 14.9% of romiplostim-treated patients and 3.7% in placebo patients, suggesting that romiplostim does stimulate blast counts when given in this context. Follow-up data was available in 14 of 16 patients with increased blasts, and in all 14 the blasts decreased upon discontinuation of romiplostim.
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
375
Fig. 2. Subgroup analysis of MDS patients with lower or higher platelet levels receiving romiplostim or placebo. These results suggest a possible role for romiplostim in supportive care in thrombocytopenic patients not receiving chemotherapy and with platelet counts 20,000/mL. Figure created from data in Ref. [9]. Abbreviations: CSBE, clinically significant bleeding events; NS, not significant; P, placebo; R, romiplostim; Plt txf, platelet transfusions; pt-y, patient-years.
Eltrombopag Eltrombopag (SB-497115, Promacta®, Revolade®) is a small molecule TPO receptor agonist, dosed once a day orally [10], that stimulates megakaryocyte proliferation and increases platelet counts [11]. It is active in humans and chimps but in no other species [12]. Eltrombopag is also a potent iron chelator [13]. Eltrombopag is not as robust as romiplostim at increasing platelet counts in healthy subjects. After maximal dosing, eltrombopag produced a platelet count rise of ~150,000/mL over baseline [14], compared to ~1.6 million/mL with romiplostim [2]. This difference in effect is explained in part by the finding that eltrombopag binds to a different area of the TPO receptor than does romiplostim (Fig. 1). Eltrombopag binds to the transmembrane region where species-specific differences at residue 499, histidine in humans and chimps but leucine in most other species, determine its biological effect. Upon binding, eltrombopag activates downstream signal transduction pathways differently than recombinant TPO or romiplostim: there is less phosphorylation of STAT pathways and no activation of AKT pathways [15]. Nevertheless, eltrombopag increases megakaryocyte growth and platelet production. Although differing in mechanism of action and seemingly less potent than romiplostim, eltrombopag has a major attribute that distinguishes it from romiplostim: the thrombopoietic effect of eltrombopag is additive to the effects of TPO and may in some settings may be synergistic with it [16]. Using a TPO-dependent cell line (Fig. 3), cells proliferate maximally with TPO but growth with eltrombopag alone is less robust. However, when TPO and eltrombopag are combined, cell growth is increased well above that with TPO alone [16]. Eltrombopag may be considered a potentiator of TPO in terms of acting in situations in which TPO levels are already high. This unique effect of eltrombopag is best illustrated in the treatment of aplastic anemia [17], for which eltrombopag was recently approved by the FDA. Patients with aplastic anemia (AA) have extremely high TPO levels [3000e5000 pg/mL (normal values < 100 pg/mL)]. When 25 AA patients were treated with eltrombopag, 11 (44%) had responses in one cell line (9 increased platelets; 6 increased hemoglobin; 9 increased neutrophils) with 6 (24%) having trilineage responses [17]. Some responders also experienced increased bone marrow cellularity that persisted upon stopping the drug. The exact mechanism here is unclear but this data suggests an additive or synergistic effect of eltrombopag in a high TPO environment, something that would not be expected for a recombinant TPO or romiplostim. Eltrombopag and induction/consolidation chemotherapy in acute leukemia In leukemia patients, 5 different clinical trials are underway or recently completed investigating whether eltrombopag can potentiate the high levels of TPO in this setting (or via a separate
376
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
Fig. 3. Eltrombopag effect is additive to that of TPO on growth of megakaryocyte cell line. When recombinant human TPO and eltrombopag are administered together, a TPO-dependent cell line grows considerably more than when either agent is added alone. Created from data in Ref. [16].
mechanism) and amplify platelet recovery during induction or consolidation (Table 1). No preliminary data are yet available from these trials and it should be recalled (vide supra) that the addition of supraphysiological doses of a recombinant TPO failed to hasten recovery in these settings. Eltrombopag and blast counts Given the initial concern that arose that romiplostim increased the AML transformation rate and increased the blast counts in patients with MDS [9], initial studies with eltrombopag in this patient group sought first to determine if eltrombopag had a similar effect on myeloid blasts. The results were quite striking and contrary to the romiplostim effect. After addition of eltrombopag to cultures of bone marrow from healthy or AML/MDS patients, the healthy marrow cells grew but the MDS/AML cells died [18]. Addition of eltrombopag at physiologically obtainable concentrations to AML cell lines resulted in rapid cell death, and this effect was independent of the presence of the TPO receptor [19]. The inhibitory effect of eltrombopag on leukemia cells in both primary cell lines and tissue culture lines seemed to occur by arresting cells in G1 and increasing differentiation and was directly related to the ability of eltrombopag to reduce the concentration of intracellular iron [13]. Further studies have shown that by lowering iron, eltrombopag reduces reactive oxygen species in AML cells thereby
Table 1 Clinical trials of eltrombopag in AML. Trial identifier
Patient population
Trial type
Objective
NCT01656252 (Completed March 2016) NCT01550185 (Study terminated) NCT01890746
Adult AML patients 18e70 years during consolidation and in complete remission Relapsed/refractory AML patients 18 years
Phase 1, 2
Optimal tolerated dose; increase in platelet count recovery Highest safe dose, safety, efficacy
NCT02071901 NCT02446145
AML patients 18 years receiving standard induction therapy AML patients 60 years receiving induction therapy AML patients 65 years of age not eligible for intensive chemotherapy
Phase 1
Phase 2
Safety and efficacy
Phase 2
Improve platelet recovery
Phase 2, randomized, placebocontrolled, decitabine with or without eltrombopag
Safety and efficacy
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
377
resulting in their death [20]. In a murine leukemia model, eltrombopag was shown to reduce leukemia cell growth and prolong survival [13]. This potential anti-leukemia effect may also occur in humans with AML. A patient with nucleophosmin 1 mutated AML and a high bone marrow blast count was treated for 2 months with eltrombopag and had a marked, but short-lived, decrease in blast count but no effect on peripheral blood counts [21]. Eltrombopag in patients with MDS or acute leukemia not receiving chemotherapy A recent study examined the safety and platelet count effect of eltrombopag versus placebo in 98 patients with advanced MDS or AML but not receiving chemotherapy [22]. There appeared to be no effect on peripheral blood or bone marrow blast counts and no effect upon overall progression-free survival or overall survival. However there was also no significant effect on platelet counts (eltrombopag: 17,000/mL; placebo: 12,000/mL), grade 3 bleeds (eltrombopag: 16%; placebo: 26%), or RBC or platelet transfusions. Conclusion The role of the TPO receptor agonists in the treatment of AML is unclear. What is seemingly established is that romiplostim does have the potential to increase myeloid blast counts. Whether this reversible effect is clinically meaningful is untested, but there was no adverse effect on remission rates in earlier trials of recombinant TPOs in AML. In contrast, eltrombopag appears not to increase blast counts, and may actually have a modest anti-leukemia effect. In patients undergoing AML induction or consolidation treatment, endogenous TPO levels are markedly elevated and this may explain why the addition of more TPO in the form of recombinant TPOs in prior trials had no significant effect on platelet counts, platelet transfusion, or time to platelet recovery. This reviewer predicts a similar outcome with romiplostim in AML. However, eltrombopag has demonstrated an ability to raise platelet counts in the presence of high TPO levels (eg, aplastic anemia) and the results of ongoing trials will test whether this hypothesis is clinically valid. Finally, some patients with advanced MDS or leukemia are treated with supportive care alone. In such patients, severe thrombocytopenia is a common problem. So far the data seem clear as to the potential role for TPO receptor agonists. Trials with eltrombopag show that it is safe but probably not effective. In contrast, romiplostim may raise the platelet count and reduce bleeding in patients with platelet counts 20,000/mL, but with concern over an increased blast count. This narrow window of benefit for some patients may be clinically meaningful, and further studies to better define use in this patient population should be conducted. Disclosures Consultant: Amgen, GSK, Eisai, ONO, 3SBIO, Pfizer, Merck, BMS, Syntimmune, Suppremol, BRL, Protalix, Zafgen, Rigel, Ligand, Kirin, Novartis, MedImmune, Genzyme, Alnylam, Caremark. Research support BMS, Protalix, Rigel. Stock holdings None. Conflict of interest None.
378
D.J. Kuter / Best Practice & Research Clinical Haematology 29 (2016) 372e378
References [1] Molineux G. The development of romiplostim for patients with immune thrombocytopenia. Ann N. Y Acad Sci 2011;1222: 55e63. [2] Wang B, Nichol JL, Sullivan JT. Pharmacodynamics and pharmacokinetics of AMG 531, a novel thrombopoietin receptor ligand. Clin Pharmacol Ther 2004;76:628e38. [3] Archimbaud E, Ottmann OG, Yin JA, Lechner K, Dombret H, Sanz MA, et al. A randomized, double-blind, placebo-controlled study with pegylated recombinant human megakaryocyte growth and development factor (PEG-rHuMGDF) as an adjunct to chemotherapy for adults with de novo acute myeloid leukemia. Blood 1999;94:3694e701. [4] Vadhan-Raj S, Kavanagh JJ, Freedman RS, Folloder J, Currie LM, Bueso-Ramos C, et al. Safety and efficacy of transfusions of autologous cryopreserved platelets derived from recombinant human thrombopoietin to support chemotherapyassociated severe thrombocytopenia: a randomised cross-over study. Lancet 2002;359:2145e52. [5] Usuki K, Tahara T, Iki S, Endo M, Osawa M, Kitazume K, et al. Serum thrombopoietin level in various hematological diseases. Stem Cells 1996;14:558e65. [6] Kuter DJ. What is the potential for thrombopoietic agents in acute leukemia? Best Pract Res Clin Haematol 2011;24:553e8. [7] Kantarjian HM, Giles FJ, Greenberg PL, Paquette RL, Wang ES, Gabrilove JL, et al. Phase 2 study of romiplostim in patients with low- or intermediate-risk myelodysplastic syndrome receiving azacitidine therapy. Blood 2010;116:3163e70. [8] Giagounidis A, Mufti CJ, Kantarjian HM, Fenaux P, Sekkeres MA, Szer J, et al. Treatment with the thrombopoietin (TPO)receptor agonist romiplostim in thrombocytopenic patients (Pts) with low or intermediate-1 (Int-1) risk myelodysplastic syndrome (MDS): results of a randomized, double-blind, placebo (PBO)-controlled study. Blood ASH Annu Meet 2011;118: 117 [abstr)]. [9] Giagounidis A, Mufti GJ, Fenaux P, Sekeres MA, Szer J, Platzbecker U, et al. Results of a randomized, double-blind study of romiplostim versus placebo in patients with low/intermediate-1-risk myelodysplastic syndrome and thrombocytopenia. Cancer 2014;120:1838e46. [10] Erickson-Miller C, Delorme E, Giampa L, Hopson C, Valoret E, Tian S-S, et al. Biological activity and selectivity for Tpo Receptor of the orally bioavailable, small molecule Tpo Receptor agonist, SB-497115. Blood ASH Annu Meet 2004;104:2912 [abstr)]. [11] Luengo JI, Duffy KJ, Shaw AN, Delorme E, Wiggall KJ, Giampa L, et al. Discovery of SB-497115, a small-molecule thrombopoietin (TPO) receptor agonist for the treatment of thrombocytopenia. Blood ASH Annu Meet 2004;104:2910 [abstr)]. [12] Erickson-Miller CL, Delorme E, Iskander M, Giampa L, Hopson CB, Luengo J, et al. Species specificity and receptor domain interaction of a small molecule TPO receptor agonist. Blood ASH Annu Meet 2004;104:2909 [abstr)]. [13] Roth M, Will B, Simkin G, Narayanagari S, Barreyro L, Bartholdy B, et al. Eltrombopag inhibits the proliferation of leukemia cells via reduction of intracellular iron and induction of differentiation. Blood 2012;120:386e94. [14] Jenkins JM, Williams D, Deng Y, Uhl J, Kitchen V, Collins D, et al. Phase 1 clinical study of eltrombopag, an oral, nonpeptide thrombopoietin receptor agonist. Blood 2007;109:4739e41. [15] Erhardt JA, Erickson-Miller CL, Aivado M, Abboud M, Pillarisetti K, Toomey JR. Comparative analyses of the small molecule thrombopoietin receptor agonist eltrombopag and thrombopoietin on in vitro platelet function. Exp Hematol 2009;37: 1030e7. [16] Erickson-Miller CL, Delorme E, Tian SS, Hopson CB, Landis AJ, Valoret EI, et al. Preclinical activity of eltrombopag (SB497115), an oral, nonpeptide thrombopoietin receptor agonist. Stem Cells 2009;27:424e30. [17] Olnes MJ, Scheinberg P, Calvo KR, Desmond R, Tang Y, Dumitriu B, et al. Eltrombopag and improved hematopoiesis in refractory aplastic anemia. N Engl J Med 2012;367:11e9. [18] Will B, Kawahara M, Luciano JP, Bruns I, Parekh S, Erickson-Miller CL, et al. Effect of the nonpeptide thrombopoietin receptor agonist Eltrombopag on bone marrow cells from patients with acute myeloid leukemia and myelodysplastic syndrome. Blood 2009;114:3899e908. [19] Erickson-Miller CL, Kirchner J, Aivado M, May R, Payne P, Chadderton A. Reduced proliferation of non-megakaryocytic acute myelogenous leukemia and other leukemia and lymphoma cell lines in response to eltrombopag. Leuk Res 2010; 34:1224e31. [20] Kalota A, Selak MA, Garcia-Cid LA, Carroll M. Eltrombopag modulates reactive oxygen species and decreases acute myeloid leukemia cell survival. PLoS One 2015;10:e0126691. [21] Platzbecker U, Sockel K, Schonefeldt C, Nowak D, Helas S, Rollig C, et al. Induction of short-term remission with single agent eltrombopag in refractory nucleophosmin-1-mutated acute myeloid leukemia. Haematologica 2014;99:e247e8. [22] Platzbecker U, Wong RS, Verma A, Abboud C, Araujo S, Chiou TJ, et al. Safety and tolerability of eltrombopag versus placebo for treatment of thrombocytopenia in patients with advanced myelodysplastic syndromes or acute myeloid leukaemia: a multicentre, randomised, placebo-controlled, double-blind, phase 1/2 trial. Lancet Haematol 2015;2:e417e26.