Clinical applications of thrombopoietic growth factors

INTRODUCTION — The use of hematopoietic growth factors has markedly changed the practice of medicine. Erythroid growth factors (eg, erythropoietin) and myeloid growth factors (eg, granulocyte colony-stimulating factor) have allowed the specific stimulation of the production of erythrocytes and neutrophils, respectively. (See "Introduction to recombinant hematopoietic growth factors".)

With the discovery of thrombopoietin (TPO) and the development of a variety of peptide and non-peptide TPO receptor agonists (TPO-RAs), clinically effective thrombopoietic growth factors are now part of the clinical armamentarium. This topic review will discuss the thrombopoietic growth factors that have been developed, their activity in preclinical models, and the available clinical studies. The biology of TPO and its related disorders are discussed separately. (See "Biology and physiology of thrombopoietin".)

CLINICAL UTILITY — The potential clinical applications of thrombopoietic growth factors are suggested by analyzing the use of platelet transfusions (figure 1). (See "Platelet transfusion: Indications, ordering, and associated risks", section on 'Indications for platelet transfusion'.)

●Three TPO receptor agonists (romiplostim, eltrombopag, avatrombopag) have been approved for the treatment of immune thrombocytopenia (ITP). (See "Second-line and subsequent therapies for immune thrombocytopenia (ITP) in adults", section on 'TPO receptor agonists'.)

●Two TPO receptor agonists (eltrombopag and romiplostim [Japan only]) have been approved for the treatment of aplastic anemia. (See "Treatment of aplastic anemia in adults", section on 'Eltrombopag alone'.)

●Two TPO receptor agonists (avatrombopag and lusutrombopag) have been approved for the treatment of thrombocytopenic patients with chronic liver disease about to undergo procedures.

●The utility of these agents in myelodysplastic syndromes is discussed separately. (See "Myelodysplastic syndromes/neoplasms (MDS): Management of hematologic complications in lower-risk MDS", section on 'Refractory bleeding'.)

●There are numerous uses of these agents in other thrombocytopenic disorders. (See 'Use of TPO receptor agonists' below.)

●Although interferon is no longer used to treat hepatitis C virus infection, one TPO receptor agonist (eltrombopag) was approved for the treatment of thrombocytopenia in patients with hepatitis C who are being treated with interferon.

THROMBOPOIETINS (C-MPL LIGANDS) — A number of thrombopoietin (TPO) molecules have been developed [1,2]; they define a new and growing family of molecules called the thrombopoietin (or mpl ligand) family, based upon their common ability to bind and activate the TPO receptor, c-mpl [3] (table 1). These include the recombinant thrombopoietins and two different types of TPO receptor agonists: peptide TPO receptor agonists and non-peptide TPO receptor agonists [4]. These TPO receptor agonists are functional rather than structural TPO mimetics because they act as agonists for the TPO receptor but lack sequence homology to the TPO protein.

There have been two waves of development of the thrombopoietic growth factors.

●The first-generation molecules were recombinant thrombopoietins (eg, rHuTPO, PEG-rHuMGDF), only one of which, rHuTPO, was clinically developed; this therapy is only available in China. (See 'Recombinant thrombopoietins' below.)

●The second-generation molecules are highly modified structures (eg, romiplostim, eltrombopag, avatrombopag, lusutrombopag, hetrombopag) that mimic the function of TPO. Three of these (romiplostim, eltrombopag, and avatrombopag) are approved by the US Food and Drug Administration (FDA) and are widely used to treat ITP [5]. Eltrombopag is also approved for the treatment of thrombocytopenia in hepatitis C patients receiving interferon. Eltrombopag and romiplostim (Japan only) are approved for the treatment of aplastic anemia. Avatrombopag and lusutrombopag are licensed for the treatment of thrombocytopenia in patients with liver disease undergoing surgical procedures [6-8]. Hetrombopag is under development in China for chemotherapy-induced thrombocytopenia. (See 'Clinical utility' above and 'Non-peptide TPO receptor agonists' below.)

Recombinant thrombopoietins — Two recombinant thrombopoietins were subjected to intensive clinical investigation (figure 2):

●Recombinant human thrombopoietin (rHuTPO) is a glycosylated molecule produced in Chinese hamster ovary (CHO) cells consisting of the full-length, native human amino acid sequence, which has a circulatory half-life of 20 to 40 hours. rHuTPO (TPIAO) is approved only in China for treatment of chemotherapy-induced thrombocytopenia and ITP [9].

●Pegylated, recombinant megakaryocyte growth and development factor (PEG-rHuMGDF), a nonglycosylated, truncated molecule produced in Escherichia coli, is composed of the first 163 amino acids of the native molecule, chemically coupled to polyethylene glycol. This half of the native molecule is 50 percent similar to erythropoietin, contains the receptor-binding domain, but very little biologic activity in vivo due to the absence of the remaining, carbohydrate-rich portion of the native molecule that gives it a very short circulatory half-life. Addition of the polyethylene glycol moiety serves to replace the missing carbohydrate domain and stabilize the molecule in the circulation [10]. PEG-rHuMGDF has a half-life of 30 to 40 hours. Clinical development of this molecule was stopped because of the development of antibodies, leading to thrombocytopenia, in approximately 8 percent of individuals [2,3,11-13]. (See 'Antibody formation' below and "Biology and physiology of thrombopoietin".)

Studies with both of these molecules in healthy people and primate models have demonstrated the general clinical effects of all the thrombopoietic growth factors [4,5,14]. After a single dose of rHuTPO or PEG-rHuMGDF, there is an increase in megakaryocyte size, number, and DNA ploidy followed by a dose-dependent increase in platelet count that begins on day 5, peaks at a mean of day 12, and returns to baseline by day 28 [15,16].

Peptide TPO receptor agonists — Peptide TPO receptor agonists are short peptide sequences, usually with no sequence homology with TPO, which bind and activate the TPO receptor. They were discovered by screening peptide "libraries" for sequences that activated the TPO receptor. One of these was a 14-amino acid peptide having no sequence homology with TPO, which was able to bind to and activate the TPO receptor, c-mpl [17]. Dimerization of this peptide increased its activity 10,000-fold, thereby making it as active as recombinant TPO.

A key attribute of all of the peptide TPO receptor agonists is the need for a structure that allows them to dimerize and activate the TPO receptor. These peptides have too short a half-life in the circulation to be clinically active. To enhance their circulatory half-life, peptides have been inserted into carrier proteins such as the immunoglobulin heavy chain and have marked clinical activity when administered subcutaneously or intravenously.

Romiplostim (Nplate, Romiplate) — Romiplostim (Nplate, Romiplate, AMG 531) is a "peptibody" composed of two disulfide-bonded human IgG1 kappa heavy chain constant regions (an Fc fragment), each of which is covalently bound at residue 228 with two identical peptide sequences linked via polyglycine (figure 3) [10]. This agent increased the platelet count in healthy individuals [18,19] as well as in patients with immune thrombocytopenia (ITP) [20-23] and myelodysplastic syndrome [24-26]. Romiplostim has also been evaluated in patients with chemotherapy-induced thrombocytopenia [27,28], hepatitis C-related thrombocytopenia, thrombocytopenia prior to undergoing surgery [29], and aplastic anemia [30].

Romiplostim is FDA-approved for the treatment of ITP in adults and in children age one and above. It is administered weekly as a subcutaneous injection at a dose of 1 to 10 mcg/kg. (See "Second-line and subsequent therapies for immune thrombocytopenia (ITP) in adults", section on 'TPO receptor agonists'.)

It is also FDA-approved for radiation injury, based on animal studies. (See 'Radiation injury' below.)

Romiplostim is approved in Japan for treatment of aplastic anemia [30]. (See "Treatment of aplastic anemia in adults".)

Non-peptide TPO receptor agonists — Non-peptide TPO receptor agonists were derived from small molecule screening techniques that uncovered a number of chemical structures that bound to and activated the TPO receptor. Structures such as the hydrazinonaphthalene compounds readily activate the TPO receptor [31-34]. Three of these are FDA approved for various thrombocytopenic disorders. One potential advantage they have over romiplostim is that they may be administered orally. All four bind to and activate the TPO receptor by a mechanism different from endogenous TPO, recombinant TPO, or romiplostim [5]. Their effect may be additive to that of TPO, and this may confer on them a second advantage over romiplostim.

Eltrombopag (Promacta, Revolade, SB497115) — Eltrombopag (Promacta, Revolade) has an acidic (COOH) group at one end, lipophilic (eg, methyl) groups at the other end, and a metal chelate group in the center that creates an orally available non-peptide TPO receptor agonist (figure 4) [35]. It increases the platelet count in healthy individuals and in thrombocytopenic patients with ITP, hepatitis C, or aplastic anemia [33,36-43]. In one trial, it raised the nadir platelet count when administered to patients after non-myeloablative chemotherapy [44]. It is FDA-approved for the treatment of adults and children over one year of age with ITP, patients with hepatitis C-related thrombocytopenia who are being treated with interferon, patients with aplastic anemia who are unresponsive to immunosuppressive therapy, and adults and children over two years of age as the initial therapy for aplastic anemia in combination with immunosuppressive therapy. It is available as orally administered tablets of 12.5, 25, 50, 75, and 100 mg as well as 12.5 mg and 25 mg oral suspension; lower starting doses should be used in patients of Asian ancestry or with impaired liver function.

Avatrombopag (Doptelet, E5501, AKR-501, YM-477) — Avatrombopag (Doptelet) is an orally-available non-peptide TPO receptor agonist (figure 5). It is approximately threefold more potent than eltrombopag in raising the platelet count in healthy volunteers [45]. Unlike eltrombopag, it does not alter liver function tests or require any dietary restrictions [46-48]. It has been shown to raise platelet counts in individuals with immune thrombocytopenia (ITP) and chronic liver disease [49-51]. In a phase III trial, 227 individuals with liver disease who had platelet counts <50,000/microL and required an invasive procedure were randomly assigned to receive avatrombopag or placebo daily for five days and assessed for the need for platelet transfusions up to seven days after procedure [8]. Avatrombopag increased the platelet count and reduced the need for transfusions (transfusions given in 12 to 34 percent of the avatrombopag group versus 62 to 77 percent of the placebo group). There was no difference in bleeding events, and no thrombotic events after close assessment of the portal vein system by ultrasound.

In a randomized trial in chronic ITP, avatrombopag was associated with a greater platelet count response (platelet count >50,000/microL) than placebo (12.4 weeks versus 0 weeks with placebo) [51]. Responses occurred early with 65.6 percent of responses seen after eight days. Bleeding was also decreased and use of concomitant ITP medications reduced in those on avatrombopag.

Avatrombopag is FDA approved for adults with chronic liver disease who are scheduled to undergo a procedure and those with chronic ITP who have had an insufficient response to a previous treatment. (See "Hemostatic abnormalities in patients with liver disease", section on 'Invasive procedures' and "Second-line and subsequent therapies for immune thrombocytopenia (ITP) in adults", section on 'TPO receptor agonists'.)

Lusutrombopag (Mulpleta) — Lusutrombopag is another orally available non-peptide TPO receptor agonist akin to eltrombopag and avatrombopag. It is FDA approved to increase the platelet count in adults with chronic liver disease who are scheduled to undergo a procedure [6,7]. (See "Hemostatic abnormalities in patients with liver disease", section on 'Invasive procedures'.)

Hetrombopag — Hetrombopag is an orally available TPO receptor agonist available in China for ITP and severe aplastic anemia; it is also under evaluation for chemotherapy-induced thrombocytopenia [52].

OTHER MOLECULES WITH THROMBOPOIETIC ACTIVITY — Since they are based on the only physiologically relevant platelet growth factor, thrombopoietin (TPO), the thrombopoietins have captured most of the attention. (See 'Thrombopoietins (c-mpl ligands)' above.)

However, several other molecules not related to thrombopoietin or its receptor can enhance platelet production.

Interleukins (IL)-3, IL-6, and IL-11 stimulate platelet production.

Only one (Oprelvekin; IL-11) was approved by the US Food and Drug Administration (FDA) for use in chemotherapy-associated thrombocytopenia; it was able to stimulate megakaryocyte growth and increase platelet production with a time course similar to that of thrombopoietin, in a mechanism independent of thrombopoietin release or synergism and is independent of the thrombopoietin receptor [53,54]. It modestly reduced chemotherapy-induced thrombocytopenia and the need for platelet transfusions, but it had significant side effects including fluid retention with dilutional anemia, peripheral edema, pleural effusions, and atrial arrhythmias [5,55-59]. Oprelvekin is no longer clinically available [60].

IL-11 is probably not important for normal megakaryocytopoiesis, since elimination of the gene for the IL-11 receptor in mice produced no effect on the production of platelets or any other blood cell line [61,62].

USE OF TPO RECEPTOR AGONISTS — Thrombocytopenia is a common problem in many chronic hematologic conditions, ranging from aplastic anemia to drug-induced thrombocytopenia. It is anticipated that if adequate amounts of responsive bone marrow precursor cells are present, thrombopoietin (TPO) may be helpful in many of these patients. Extensive preclinical and clinical data provide guidance as to the settings in which the thrombopoietic growth factors may be useful [5,63-65].

This section provides an overview of the thrombocytopenic conditions treated with these thrombopoietic growth factors with a focus on the current TPO receptor agonists but supported with the preclinical and early clinical data with the recombinant thrombopoietins.

Immune thrombocytopenia — The finding that TPO levels in patients with immune thrombocytopenia (ITP) were not significantly elevated [66,67] and that most patients with ITP are not producing platelets at a maximal rate [68,69], suggests a potential benefit of TPO receptor agonists in this disorder.

Romiplostim, eltrombopag, and avatrombopag are US Food and Drug Administration (FDA) approved for treating adults with ITP and have been found to be highly effective [20-23,38,40]. In general, over 80 percent of ITP patients treated with TPO receptor agonists attain a long-term rise in platelet count >50,000/microL with reduced bleeding, increased quality of life, and minimal adverse effects (table 2) [70]. Their long-term use in patients with ITP has been found to be safe, effective, and reduced the need for splenectomy [22,23,40,71]. Use of these agents in the management of ITP is discussed in detail separately [72,73]. (See "Second-line and subsequent therapies for immune thrombocytopenia (ITP) in adults", section on 'TPO receptor agonists'.)

Liver disease

Rationale — The thrombocytopenia seen in liver disease can be due to several mechanisms. These include an immune thrombocytopenia (ITP)-like process that is commonly seen in patients with chronic hepatitis C virus (HCV) infection and usually responds to the standard ITP treatments. Another mechanism is splenic pooling of platelets when portal hypertension causes splenic enlargement. Finally, thrombocytopenia can result from reduced TPO production by a damaged liver. Studies of liver transplantation have shown that TPO levels are often undetectable pre-transplantation and rise immediately post-transplantation, followed by a rise in peripheral platelet count but no change in spleen size [74,75].

Chronic HCV infection — Hepatitis C treatments rarely use interferon-based therapy. However, when interferon was used, eltrombopag was shown to be effective in raising the platelet count in a series of 74 patients with chronic HCV and platelet counts in the range of 20,000/microL to 70,000/microL. Patients were treated for four weeks with placebo or eltrombopag at doses of 30, 50, or 75 mg/day daily in a blinded, randomized phase II study [41]. In a subsequent, large, phase III study, all patients with HCV infection and platelet count under 100,000/microL were treated with eltrombopag before initiation of antiviral treatment, and 95 percent increased their platelet count to over 100,000/microL by week 9 [76]. Patients were then randomized to receive either eltrombopag or placebo during the next 24 to 48 weeks of antiviral treatment. More patients on eltrombopag maintained a platelet count >50,000/microL than those receiving placebo (69 to 81 percent versus 15 to 23 percent, respectively), and there was a higher rate of sustained virologic response in those on eltrombopag than in those on placebo (19 to 23 percent versus 13 to 15 percent). Based on these data, eltrombopag was approved for the treatment of thrombocytopenia in patients with chronic HCV to allow the initiation and maintenance of interferon-based therapy.

Advanced liver disease — In a randomized trial, administration of eltrombopag to patients with advanced chronic liver disease and cirrhosis raised platelet counts and reduced the need for platelet transfusions prior to invasive procedures [77]. However, this was associated with an apparent increase in portal vein thrombosis, resulting in premature termination of the trial. Given the high rate of asymptomatic portal vein thrombosis in this patient population, the failure to determine the pretreatment rate of thrombosis is a major limitation of this trial. However, a similar trial with avatrombopag did not find an increased rate of thrombosis. When compared with placebo, avatrombopag for five days raised the platelet count, reduced platelet transfusions, and was not associated with any increase in thrombosis; all patients underwent ultrasound evaluation of the portal vein [8]. Although most patients underwent low-risk procedures (endoscopy, liver biopsy), World Health Organization (WHO) bleeding >grade 2 occurred in under 5 percent of participants and was not related to treatment. Identical beneficial outcomes were seen with short-term administration of lusutrombopag. The use of these agents in patients with liver disease is discussed separately. (See "Hemostatic abnormalities in patients with liver disease".)

Aplastic anemia — The use of a TPO receptor agonist might be predicted to have little benefit in patients with aplastic anemia (AA), given that TPO levels are usually elevated 10- to 20-fold greater than normal values of 25 to 100 pg/mL [78]. However, both eltrombopag and romiplostim have been shown to be effective in improving hematologic parameters in AA, including neutropenia, anemia, and thrombocytopenia in adults who are unable to undergo hematopoietic cell transplantation [43]. Some patients attaining near normal blood counts were able to discontinue treatment [42].

This unexpected result has led to an improved understanding of the mechanism of TPO in aplastic anemia. It appears that interferon gamma (levels of which are highly elevated in aplastic anemia) directly binds to endogenous TPO and prevents its interaction with the TPO receptor on hematopoietic precursor cells. This block in TPO-mediated hematopoiesis leads to pancytopenia and the elevated TPO levels. Since eltrombopag and presumably romiplostim are not "neutralized" by interferon gamma, they can still activate the TPO receptor and restore hematopoiesis [79]. (See "Aplastic anemia: Pathogenesis, clinical manifestations, and diagnosis".)

A 2017 study has shown that the addition of eltrombopag to standard immunosuppression therapy from day 1 to month 6 of treatment resulted in a complete hematologic remission in 58 percent and overall response of 94 percent, compared with historical rates of 10 and 66 percent, respectively [80]. A 2019 study found that romiplostim at 10 mcg/kg weekly produced a platelet response in 70 percent of individuals treated [30]. Some had multilineage responses; responses were durable. Management of aplastic anemia is discussed separately. (See "Treatment of aplastic anemia in adults".)

OTHER POTENTIAL USES OF TPO RECEPTOR AGONISTS

Myelodysplastic syndromes — Bone marrow from some patients with myelodysplastic syndrome (MDS) can be stimulated in vitro to form megakaryocytes, suggesting that some patients might benefit from administration of TPO [81]. In one early study of 21 patients, treatment with PEG-rHuMGDF doubled the platelet count after five weeks of treatment [82]. One individual with MDS was treated for over 450 days; while the platelet count rose from <10,000/microL to 50,000/microL, there were also effects on other cell lines, with the hemoglobin rising to >13 g/dL [59].

In general, romiplostim treatment increased the platelet count and reduced the need for platelet transfusions in patients with MDS receiving chemotherapy or supportive care [24-26]. One trial randomly assigned thrombocytopenic patients with low/intermediate-1-risk MDS (platelet count <20,000/microL or history of bleeding along with platelets <50,000/microL) to receive romiplostim or placebo for 58 weeks [83]. The primary study endpoints, clinically important bleeding events, were not significantly reduced with romiplostim (hazard ratio [HR], 0.83; 95% CI 0.66-1.05), but in those with platelet counts >20,000/microL, significant reductions were seen (HR, 0.34; 95% CI 0.20-0.58). Romiplostim reduced all bleeding events (relative risk [RR], 0.92) and platelet transfusions (RR, 0.77) and increased platelet response (odds ratio [OR], 15.6). This study was stopped at an interim analysis because of a perceived increase in AML rate (HR, 2.51) with romiplostim, but the final analysis showed AML rates of 6 percent with romiplostim and 4.9 percent placebo (HR, 1.20; 95% CI 0.38-3.84) and similar survival rates. The management of MDS is discussed separately. (See "Overview of the treatment of myelodysplastic syndromes".)

Because of the concern that TPO receptor agonists might promote leukemic transformation, the benefit of eltrombopag has been carefully assessed. Paradoxically, eltrombopag was found to inhibit leukemic cell growth in tissue culture and in animal models of leukemia [84]. When studied in 98 thrombocytopenic patients with advanced MDS or AML who were not receiving chemotherapy, there was no effect on peripheral blood or bone marrow blast counts and no effect upon overall progression-free survival or overall survival [85]. However, there was also no significant effect on platelet counts (eltrombopag: 17,000/microL; placebo: 12,000/microL), grade 3 bleeds (eltrombopag: 16 percent; placebo: 26 percent), or RBC or platelet transfusions.

Guidelines for MDS from the National Comprehensive Cancer Network (NCCN) suggest consideration of TPO receptor agonists in low-risk MDS patients with severe or life-threatening thrombocytopenia [86]. (See "Myelodysplastic syndromes/neoplasms (MDS): Management of hematologic complications in lower-risk MDS", section on 'Refractory bleeding'.)

MYH9-related congenital thrombocytopenia — Of the congenital thrombocytopenias, non-muscle heavy-chain myosin-9 (MYH9)-related thrombocytopenias are associated with a number of distinct clinical syndromes and MYH9 mutations [87]. Most have mild thrombocytopenia with minimal bleeding; however, some have more severe thrombocytopenia with associated increased bleeding risk. In one study, 12 adults with MYH9-related thrombocytopenia (platelets <50,000/microL) were treated with eltrombopag 50 mg daily for at least three weeks [88]. In eight patients the platelet count either rose above 100,000/microL or three times baseline; three patients doubled their baseline platelet count; only one patient failed to respond. Bleeding decreased in 8 of 10 individuals who had bleeding at baseline. Only two patients had mild adverse events. (See "Inherited platelet function disorders (IPFDs)", section on 'MYH9-related disease'.)

HIV-associated thrombocytopenia — In a single study of six HIV-infected patients with thrombocytopenia treated with PEG-rHuMGDF, all six normalized platelet counts [89]. The reported mechanism of effect was a reduction in the rate of megakaryocyte apoptosis (and hence increased rate of platelet production) due to the PEG-rHuMGDF. Although there is only a small number of case reports demonstrating the efficacy of TPO receptor agonists in HIV-associated thrombocytopenia [90,91], this author has found major platelet count increases in multiple HIV-infected patients with platelet counts <10,000/microL after treatment with TPO receptor agonists; many of these patients were assumed to have "secondary ITP" due to HIV. The management of HIV-associated thrombocytopenia is discussed in detail separately. (See "HIV-associated cytopenias", section on 'Thrombocytopenia'.)

Chemotherapy — Chemotherapy patients use approximately 25 percent of all platelet products transfused in the United States. Chemotherapy drugs such as carboplatin and gemcitabine cause significant thrombocytopenia (platelets <25,000/microL) in up to 7 percent of patients [92]. This is a major area in which thrombopoietic growth factors might show benefit in the primary or secondary prophylaxis of thrombocytopenia [92]. In several animal models of radiation- and chemotherapy-induced thrombocytopenia, administration of recombinant thrombopoietins was associated with a slightly earlier platelet nadir, shorter duration of thrombocytopenia, and, in other animal models, a reduction in the depth of the platelet nadir (figure 6) [93]. The duration of neutropenia and anemia was also shortened in some animal studies.

Before considering TPO receptor agonists for the treatment of chemotherapy-induced thrombocytopenia, it is important to note that the TPO receptor (c-mpl) is either not present or is inactive in solid tumor cells [94,95]. It is also important to note that the thrombocytopenia in non-myeloablative chemotherapy is distinct from that of myeloablative chemotherapy or hematopoietic cell transplantation (HCT) in that the latter are associated with very high levels of endogenous TPO that may not benefit from the addition of further recombinant TPO or romiplostim. Indeed, older studies (1995 through 1998) with the recombinant thrombopoietins showed improved platelet counts in the non-myeloablative chemotherapy regimens but little or no effect in the myeloablative regimes or HCT. (See 'Hematopoietic cell transplantation' below.)

However, given the unique mechanism of action of eltrombopag, avatrombopag, and lusutrombopag and their ability to potentiate the effect of endogenous TPO, there is some enthusiasm for studies with these agents in the myeloablative/HCT settings.

Solid tumors — Following the administration of recombinant TPO to humans undergoing non-myeloablative chemotherapy, platelet recovery was enhanced but, unlike the animal models, with no change in the recovery of white or red blood cells.

●One study, for example, evaluated 53 patients with lung cancer treated with carboplatin and paclitaxel [96]. Thirty-eight received doses of PEG-rHuMGDF ranging from 0.03 to 5 mcg/kg daily after chemotherapy and 15 received placebo. Patients who received PEG-rHuMGDF had a significantly higher platelet nadir than those treated with placebo (188,000/microL versus 111,000/microL) and an earlier return of their platelet counts to baseline (14 versus 21 days) (figure 7).

●In a second trial, 41 cancer patients treated with carboplatin and cyclophosphamide were randomly assigned to receive either PEG-rHuMGDF or placebo [97]. There was no effect on the platelet nadir although the patients who received PEG-rHuMGDF had an earlier return of their platelet counts to baseline (17 versus 22 days).

Although these studies demonstrated that recombinant TPO has activity in the chemotherapy setting and is safe, as no adverse events were attributed to the PEG-rHuMGDF, they were not designed to show a reduction in the need for platelet transfusions. Few of the patients received platelet transfusions because the dose intensity of the chemotherapy was not high enough. However, in a third trial using a dose-intense chemotherapy regimen to treat patients with ovarian cancer, rHuTPO elevated nadir platelet counts, reduced the duration of thrombocytopenia, along with an over 50 percent reduction in the need for platelet transfusions (figure 8) [98].

It is not clear if eliminating treatment delays by using thrombopoietic agents and thereby maintaining the dose intensity of ICE chemotherapy would translate into better clinical outcomes. A randomized, double-blind, placebo-controlled phase I/II trial of PEG-rHuMGDF in combination with filgrastim studied 38 patients with relapsed, refractory aggressive NHL who were given ICE chemotherapy for stem cell mobilization prior to autologous HCT with the following results [99]:

●Patients given PEG-rHuMGDF had significantly less grade IV thrombocytopenia (35 versus 15 percent) and a significantly higher median nadir platelet count (49,000/microL versus 20,000/microL) compared with those given placebo.

●Patients given PEG-rHuMGDF were more likely to receive ICE on schedule (77 versus 44 percent).

●For patients who went on to receive HCT, overall survival was significantly higher in those who received their therapy on schedule compared with those whose therapy was delayed (73 versus 13 percent).

Recombinant TPO is extensively used to support chemotherapy in China [9]. However, the TPO receptor agonists have not been extensively studied in chemotherapy-associated thrombocytopenia. The following summarizes the major studies:

●A 2017 Cochrane review identified two trials comparing a TPO receptor agonist with placebo in individuals with solid tumors who were receiving chemotherapy and concluded evidence was insufficient to determine whether these drugs reduced bleeding or the need for platelet transfusions [100].

•One of the trials randomly assigned 26 patients receiving gemcitabine chemotherapy regimens to receive eltrombopag or placebo prophylaxis starting at cycle 2 and found higher nadir platelet counts in the eltrombopag group (115,000/microL versus 53,000/microL, respectively) [101].

•Another trial assigned 183 patients receiving carboplatin/paclitaxel regimens for solid tumors to eltrombopag (a series of doses) or placebo and found higher post-nadir platelet counts in the eltrombopag groups [102].

●Another Phase II study assigned patients with solid tumors receiving gemcitabine monotherapy or gemcitabine plus cisplatin/carboplatin to either eltrombopag 100 mg (n = 52) or placebo (n = 23) for five days before and after chemotherapy was started [44]. The primary endpoint (prechemotherapy [day 1] platelet count across chemotherapy cycles) was higher with eltrombopag. Platelet transfusions were fewer and platelet counts were higher in those on eltrombopag, but none reached statistical significance.

●In a case series of 20 patients treated with romiplostim for prolonged chemotherapy-induced thrombocytopenia, 19 demonstrated improvement in platelet counts; chemotherapy could be resumed at full doses in 15 (75 percent) [103]. The mean dose to achieve adequate platelet counts was 2.9 mcg/kg weekly. There were no significant toxicities. Three deep vein thromboses occurred, which was considered similar to the anticipated rates of thrombosis in patients with cancer on active chemotherapy.

●In a small randomized trial of solid tumor patients with chemotherapy-associated thrombocytopenia, 14 of 15 romiplostim-treated patients (93 percent) experienced correction of their platelet count within three weeks, compared with one of eight control patients (12.5 percent; P <0.001) [28]. The mean platelet count at two weeks of treatment was 141,000/microL versus 57,000/microL in the eight observation patients at 3 weeks.

●In a large retrospective multi-center study of romiplostim in 173 patients with chemotherapy-associated thrombocytopenia, 71 percent of patients had a platelet count response to over 100,000/microL, and the median platelet count on romiplostim was 116,000/microL versus 60,000/microL before treatment [104]. Chemotherapy dose reductions/delays were avoided in 79 percent, and 89 percent avoided platelet transfusions.

Acute leukemia — In contrast to non-myeloablative chemotherapy for solid tumors, the administration of PEG-rHuMGDF following chemotherapy for acute leukemia has failed to produce enhancement of platelet recovery when given after standard myeloablative induction regimens [105,106]. The reasons for this failure are not entirely clear. They may relate to the absence of sufficient bone marrow progenitor cells upon which to act, high endogenous TPO levels, or an inappropriate administration scheme [107,108]. Attempts to increase the dose of TPO or to alter the dosing scheme (eg, to include doses prior to, during, or after induction or consolidation chemotherapy) did not have an effect on the platelet count [109]. Indeed, neutropenia was actually prolonged when the TPO was given during chemotherapy.

The administration of eltrombopag during induction chemotherapy for acute myeloid leukemia was found not to be helpful [110]. Starting at day 4 of a standard daunorubicin/cytarabine induction course, patients received 200 mg/day eltrombopag (n = 74) or placebo (n = 74) until remission or day 42. Eltrombopag patients had an increased death rate (53 percent versus 41 percent) than those on placebo.

Hematopoietic cell transplantation — Although comprising a small number of patients, hematopoietic cell transplant (HCT) recipients, especially those who have delayed or absent engraftment of platelets, consume a disproportionately large amount of the national supply of platelets. Accordingly, stimulation of megakaryocyte and platelet engraftment after HCT is an important potential use of TPO.

Stem cell mobilization — In mice, TPO administration improved stem cell harvests and subsequent engraftment. When donor animals were treated with recombinant TPO and the TPO-stimulated marrow then transplanted into recipient animals, there was a reduction in the duration of thrombocytopenia as well as earlier recovery of erythrocytes (figure 9) [111].

In oncology patients, administration of rHuTPO efficiently mobilizes peripheral blood progenitor cells. When combined with chemotherapy and filgrastim (G-CSF), PEG-rHuMGDF produced a 250-fold increase in the number of circulating megakaryocyte colony forming cells (Meg-CFC), a 190-fold increase in granulocyte- macrophage-CFC, a 65-fold increase in erythroid burst forming cells and a 24-fold increase in CD34+ cells when compared with patients receiving only chemotherapy and filgrastim (figure 10) [97]. (See "Hematopoietic cell transplantation (HCT): Sources of hematopoietic stem/progenitor cells".)

The utility of simultaneous G-CSF (5 mcg/kg SQ or IV twice daily on days three to eight) and rHuTPO (total dose 0.6 to 2.4 mcg/kg IV) for peripheral blood stem cell mobilization was studied in 29 breast cancer patients; a concurrent group of 20 patients receiving G-CSF only served as controls [112]. CD34+ cell yields in the first apheresis (day 6) were 4.1 and 0.8 x 10⁶ cells/kg for the G-CSF/TPO and G-CSF groups, respectively. The targeted minimum yield of 3 x 10⁶ CD34+ cells/kg was procured following a single apheresis in 61 and 10 percent of subjects in the G-CSF/TPO and G-CSF groups, respectively. Use of TPO with G-CSF modestly accelerated platelet and granulocyte recovery post-chemotherapy and modestly reduced RBC transfusion requirements; no adverse toxicities were noted [112].

Stem cell engraftment — The results of administration of PEG-rHuMGDF or rHuTPO after autologous HCT in general have been disappointing, with no significant enhancement of platelet recovery and no reduction in platelet transfusions. Several approaches have been tried:

●In one report, rHuTPO was administered to patients who had received either bone marrow or peripheral blood stem cell transplants in whom platelets had failed to engraft after 30 days [113]. Only 3 of 37 patients showed increased bone marrow megakaryocytes, and only two became transfusion-independent.

●A second study evaluated patients with breast cancer undergoing autologous HCT using bone marrow stem cells (STAMP V regimen) who were given PEG-rHuMGDF immediately after the stem cell infusion [114]. There was no effect on the number of days that the patients had severe thrombocytopenia (ie, a platelet count <20,000/microL) or on the time to platelet recovery. However, there was a dose-dependent rebound in the post-transplant platelet count.

●In another series of patients with breast cancer treated with the STAMP I regimen, PEG-rHuMGDF was given before HCT (days minus 14 to minus 8) using peripheral blood stem cell infusion and, in some patients, after the infusion as well [115]. The PEG-rHuMGDF treatment increased the platelet count on the day of hematopoietic stem cell infusion (228,000 versus 107,000/microL) but had no effect on platelet recovery or the need for platelet transfusion.

The ultimate role of TPO in HCT may be to stimulate the number of progenitor cells in the marrow or peripheral stem cell population prior to harvesting. As suggested by murine models [111], it may be more important to use TPO to treat the stem cell donor rather than the recipient.

Delayed platelet engraftment — Delayed engraftment of platelets is a major complication of HCT. In one retrospective series of 86 patients with persistent thrombocytopenia after allogeneic HCT (allo-HCT), administration of either romiplostim or eltrombopag starting a median of 127 days after allo-HCT was associated with platelet recovery in 72 percent of patients after a median of 66 days of treatment [116]. A systematic review of 25 reports of using TPO receptor agonists in patients with persistent thrombocytopenia after allo-HCT [117], demonstrated a durable platelet response in 70 percent of 121 patients receiving eltrombopag and 82 percent of 49 patients receiving romiplostim. Prospective studies in this area are indicated.

Additional discussions about management of thrombocytopenia following HCT are presented separately. (See "Hematopoietic support after hematopoietic cell transplantation" and "Platelet transfusion: Indications, ordering, and associated risks", section on 'Leukemia, chemotherapy, and HSCT'.)

Surgery — Cardiac surgery, liver transplantation, or other major surgery consume about 40 percent of all platelets transfused in North America, a need that might be ameliorated by administration of a TPO receptor agonist. Additionally, some patients may be unable to benefit from platelet transfusion due to alloimmunization or religious belief.

Evidence for the efficacy and safety of preoperative administration of TPO receptor agonists is limited to observational studies.

●In a series of 18 patients with thrombocytopenia of various etiologies who received romiplostim prior to elective surgical procedures (including three Jehovah's Witnesses), administration of romiplostim was effective in raising platelet counts in all cases (from a median of 47,000/microL to a median of 144,000/microL) [118]. There were no surgical delays or cancellations due to thrombocytopenia, and there were four postoperative bleeding episodes at platelet counts between 82,000/microL and 185,000/microL. There were no thromboembolic events directly related to the drug, although one patient had a urinary bladder catheter-associated clot after prostate surgery with a platelet count of 48,000/microL. An update of this study to include 51 surgical procedures in 47 thrombocytopenic individuals reported that median (range) baseline platelet counts rose from 47,000 (9000 to 120,000)/microL to 164,000 (28,000 to 603,000)/microL at the time of surgery [29]. A romiplostim dose of 3 mcg/kg per week for two doses increased the platelet count to >100,000/microL in 79 percent of patients within 14 days.

●In a series of 35 patients with chronic hepatitis C who received romiplostim prior to elective surgical procedures, administration of romiplostim was effective in raising platelet counts in 33 (from a median of 31,000/microL to a maximum of 73,000/microL to 240,000/microL) [119]. There were no postoperative bleeding or thrombotic complications.

The optimal timing and dose of administration seems straightforward. In the studies above, romiplostim was administered for approximately two to four weeks prior to surgery, and dosing was in the range of 2 to 3 mcg/kg, administered once per week.

These reports suggest that romiplostim may be appropriate for some patients with a platelet count <50,000/microL who are unable to receive platelet transfusions. While concerns regarding the possible exacerbation of thromboembolic disease remain, especially for patients with underlying thrombophilia and/or thrombosis, few were reported in the studies above. (See 'Thrombocytosis and thrombosis' below.)

A randomized trial of thrombocytopenic ITP patients requiring a procedure showed that eltrombopag was as efficient as IVIG in elevating the platelet count [120]. Perioperative platelet count targets were achieved for 29 of 37 patients (78 percent) in the eltrombopag group and 20 of 32 (63 percent) in the IVIG group.

Transfusion medicine — In addition to its regulatory role in platelet production, TPO also has an important physiologic role in the regulation of early hematopoiesis [121,122]. As a result, there are potential applications for the use of TPO in stem cell expansion as well as stimulation of platelet production ex vivo and in vivo.

Ex vivo expansion of progenitor cells — TPO in combination with other hematopoietic growth factors (eg, flt-3 ligand, IL-3) may be used to expand cord blood or peripheral blood progenitor cells ex vivo. In one study, for example, a simple cocktail of flt-3 ligand and TPO expanded cord blood progenitors several hundred-thousand-fold after 25 weeks in culture (figure 11) [123]. These expanded progenitor populations appear to provide adequate hematopoietic reconstitution in mice [124].

Ex vivo platelet production — Platelets can be produced in vitro under the stimulation of TPO. CD34+ and other precursor cells can be grown in the presence of TPO such that most of the cells become megakaryocytes and, in turn, shed platelets [125]. These shed platelets have normal ultrastructure and function compared with their in vivo counterparts [126,127]. Although probably not an economical source of platelets, a number of "platelet bioreactors" have been developed that can produce normal human platelets [127-132].

Increasing platelet apheresis yields — rHuTPO and PEG-rHuMGDF stimulated platelet production in healthy apheresis donors and increased the apheresis yield (figure 12) [133,134]. In another study, rHuTPO was given to cancer patients prior to chemotherapy and platelets were then harvested and frozen. Subsequent administration of these platelet products to their donors successfully ameliorated the thrombocytopenia associated with their chemotherapy [135]. Although this modality is no longer in development with recombinant thrombopoietins due to concerns regarding development of thrombocytopenia due to anti-TPO antibodies, similar treatment of healthy donors with current TPO receptor agonists could be considered. (See 'Antibody formation' below.)

Thrombocytopenia with absent radius syndrome — The low platelet count in patients with the thrombocytopenia with absent radius (TAR) syndrome appears to result from failure of cell signaling events downstream from the TPO receptor [136]; the TPO receptor itself is not mutated. It is thus unlikely that exogenous TPO will be useful in this disorder; in vitro bone marrow studies did not show a response to recombinant TPO [136].

Congenital amegakaryocytic thrombocytopenia — Individuals with congenital amegakaryocytic thrombocytopenia (CAMT) have mutations in the TPO receptor that render them unable to respond to TPO [137-139]. They present with thrombocytopenia at birth and subsequently develop pancytopenia. In vitro, a few respond to recombinant TPO but most do not. However, a monoclonal antibody ("minibody") has been developed that binds the TPO receptor in a way different from TPO and has been shown in vitro to activate the mutated receptor [140]. However, no clinical studies have been conducted to confirm this finding and this antibody is no longer in clinical development.

Radiation injury — Severe thrombocytopenia may occur after therapeutic or accidental radiation exposure. TPO may have a bone marrow protective effect. When mice were given a high dose (950 rad) of total body irradiation (TBI), 83 percent of the mice died; but if they were given recombinant murine TPO (rMuTPO) two hours before the irradiation, only 25 percent died [141-143]. The protective effect on platelets was maximal from minus 2 to plus 4 hours after TBI but was still present if administered 24 hours after TBI.

Given the anti-apoptotic effect of TPO on hematopoietic stem cells [144,145], these studies suggest that there may be a narrow window in time during which administration of TPO may reverse the apoptotic effects of irradiation on stem cells. After sub-lethal TBI (5Gy) to rhesus monkeys, administration of TPO 24 hours later and continued for 21 days markedly reduced the subsequent thrombocytopenia and increased the reticulocyte count; even a single dose at 24 hours was almost as effective, and a reticulocyte response but not neutrophil response was also augmented [146,147]. The addition of GM-CSF or G-CSF daily for 14 days to the single dose of TPO at 24 hours markedly promoted a trilineage recovery [148].

Despite these animal findings, the role for TPO in the treatment of radiation exposure in humans has not been well assessed for the TPO receptor agonists. In the only relevant study, eltrombopag was given to stem cell transplant patients conditioned with TBI [149]. The drug was safe, but there were insufficient patients to document a treatment benefit.

Romiplostim was approved in 2021 by the FDA for acute radiation syndrome, based on animal studies [150]. Treatment of radiation injury is discussed separately. (See "Management of radiation injury".)

IMPACT ON PLATELET TRANSFUSION — There was much early enthusiasm for using thrombopoietin (TPO) in the acute treatment of thrombocytopenia in lieu of platelet transfusions [151]. However, since none of the thrombopoietic molecules hastens platelet formation from megakaryocytes and all take five days to start to increase platelet production [15], they will not replace platelet transfusions in the acute setting. Nevertheless, there remains interest in developing another substance (eg, SDF-1) that might stimulate platelet shedding from existing megakaryocytes and possibly reduce the need for acute platelet transfusions.

Except for nonmyeloablative chemotherapy, it is unlikely that TPO receptor agonists will have a major impact on the need for platelet transfusions.

SIDE EFFECTS AND RISKS — For any hematopoietic growth factor, potential adverse effects must be carefully assessed. For the thrombopoietic growth factors a number of actual or potential toxicities have been identified (table 3). Cost may also be limiting in some settings.

Antibody formation — Formation of antibodies to recombinant human hematopoietic growth factors (eg, granulocyte colony-stimulating factor, erythropoietin) has been relatively uncommon, but was a significant problem in the clinical development of the recombinant thrombopoietins.

In one report, approximately 8 percent of healthy volunteers developed thrombocytopenia after receiving three monthly injections of PEG-rHuMGDF. This was apparently due to the development of antibodies that cross-reacted with endogenous thrombopoietin, neutralized its biologic activity, and produced thrombocytopenia [12]. Platelet counts as low as 4000/microL were reported [12,13]. All patients eventually recovered once the antibody abated, but clinical development of PEG-rHuMGDF by Amgen was stopped in 1998 because of this side effect [2]. In comparison, administration of rHuTPO was not associated with neutralizing antibodies; post-marketing data in China have identified no neutralizing antibodies in over 700 patients treated with rHuTPO.

The TPO receptor agonists have been developed to minimize the formation of autoantibodies. The TPO non-peptide mimetics do not carry the risk of antibody formation. The antigenicity of the one peptide mimetic, romiplostim, has been extensively studied. Of 961 patients in 13 clinical trials, 60 patients developed antibodies against romiplostim (most against the carrier Fc portion) [152]; four neutralized romiplostim but were not associated with loss of platelet response. None cross-reacted with endogenous TPO.

Thrombocytosis and thrombosis

Portal vein thrombosis — Eltrombopag may be associated with a risk of thrombosis in the portal venous system in patients with advanced liver disease. One randomized trial (292 patients) of eltrombopag versus placebo in patients with chronic liver disease or cirrhosis from all causes undergoing elective invasive procedures showed an increase in thrombotic events (seven in the eltrombopag group versus three in the placebo group; odds ratio [OR]: 3.04, 95% CI 0.62-14.82) [77]. Five of six patients with thrombosis had platelet counts over 200,000/microL. It is unclear whether these asymptomatic portal vein thromboses were due to the eltrombopag or the underlying liver disease, as imaging studies were not performed pretreatment to assess for thromboses that might have been present before eltrombopag treatment. In a much larger group of hepatitis patients, interferon and ribavirin treatment was supported with either eltrombopag (in 955 patients) or placebo (in 484 patients) and all patients underwent screening for thrombosis [76]. Eltrombopag treatment was associated with an increased rate of thromboembolic events (5.8 per 100 patient-years versus 1.9 per 100 patient-years) and slightly increased rate of portal vein thrombosis (1.25 versus 0.4 percent). Since eltrombopag treatment allowed higher dosing of antivirals, it is unclear whether the eltrombopag or the interferon plus ribavirin was the causative factor [41].

An increased rate of portal vein thrombosis has not been seen in trials of avatrombopag or lusutrombopag in liver disease patients being prepared for invasive procedures. This may reflect an underlying difference with eltrombopag or, more likely, the short exposure and relatively low platelet count elevations targeted by these two therapies. (See "Hemostatic abnormalities in patients with liver disease", section on 'Invasive procedures'.)

Lack of increased thrombosis in ITP — ITP is associated with an increased baseline risk of venous and arterial thrombosis and this risk rises as the platelet count falls [153,154]. In randomized trials and other studies, the use of TPO receptor agonists did not increase the rate of thrombosis compared with placebo [21,23,38,40,155]. In those studies, thromboses occurred at both low and high platelet counts.

Potential mechanisms of thrombosis — There are several potential prothrombotic mechanisms of the TPO receptor agonists that deserve attention:

●These molecules can be extremely potent thrombopoietic growth factors and can markedly elevate the platelet count in a short period of time. In a baboon model, the deposition of platelets in an extravascular shunt, which mimics an ulcerated atheroma in humans, was directly related to the platelet count after PEG-rHuMGDF administration [16,156]. Except for its ability to elevate the platelet count, PEG-rHuMGDF did not synergize with or exacerbate platelet deposition. Nevertheless, increasing the platelet count in individuals with active arterial thrombotic disease may increase the risk of thrombosis.

●Platelets produced by stimulation with any of the recombinant thrombopoietins have normal aggregation responses. However, when the recombinant thrombopoietins (but not the non-peptide TPO receptor agonists) are added directly to platelets (or persist at elevated levels in the circulation in vivo), they decrease by approximately 50 percent the threshold for activation by various platelet agonists (eg, ADP, collagen) in platelet aggregometry studies [16,156]. This is probably not clinically relevant, since studies with recombinant thrombopoietins in oncology patients, a patient group with an inherent increased rate of thrombosis, showed no increased rate of thrombosis. Furthermore, treatment of baboons with PEG-rHuMGDF did not increase fibrin deposition in extravascular shunts despite marked increases in platelet count [16,156]. Finally, in the rabbit carotid artery model, PEG-rHuMGDF did not alter "cyclic flow reduction", a sensitive test of platelet-dependent thrombosis [157].

●TPO receptor agonists may increase the rate of microparticle formation in patients with ITP [158]. Microparticles in turn increase the rate of thrombin generation and are associated with increased rates of thrombosis in many disorders of increased platelet turnover.

Bone marrow fibrosis — In several preclinical models, murine bone marrow was transfected with the thrombopoietin gene, causing thrombopoietin to be expressed at very high levels in the bone marrow of mice following transplantation. Most of these mice developed extensive marrow fibrosis similar to that seen in primary myelofibrosis [159-161]. This fibrotic response is probably secondary to the high local bone marrow concentrations of transforming growth factor beta1 (TGF-beta1) and other factors [162].

Administration of recombinant thrombopoietin to mice produces increased bone marrow reticulin, which is reversible upon discontinuation of treatment [163]. There has been little clinical evidence for bone marrow fibrosis in humans given rHuTPO or PEG-rHuMGDF. However, one case series using serial analyses of bone marrow and peripheral blood in nine patients who received rHuTPO after induction therapy for acute myeloid leukemia (AML) and in eight patients undergoing the same AML induction treatment but without rHuTPO treatment showed increased bone marrow cellularity in eight of the nine rHuTPO-treated and five of the eight control patients [164]. Eight of nine rHuTPO-treated and two of six control patients had increased bone marrow reticulin staining. These morphologic findings resolved within 42 days of the last dose of rHuTPO. Given the brevity of most exposures and the lack of routine bone marrow analysis, this potential issue was not fully explored.

Romiplostim causes a reversible increase in reticulin in rats [165]. Thus far, fewer than 5 percent of ITP subjects treated with the TPO receptor agonists have been reported to have increased reticulin. In some it was present at baseline; in others it increased from pretreatment levels, and receded upon discontinuation of treatment; in others it increased and then decreased despite continued therapy. Only a rare patient exhibited signs of collagen formation in addition to reticulin [166]. No patient developed other cytopenias or a myeloproliferative syndrome [167]. A three-year prospective bone marrow study with romiplostim in patients with ITP showed that 9 of 131 patients (7 percent) had increases in reticulin of ≥2 grades, and two of the nine also had increased collagen [167]. It was unclear if the frequency of reticulin increased with exposure time: increased reticulin occurred in 3 of 41 after one year, 1 of 38 after two years, and 5 of 52 after three years. Three of the nine patients had repeat biopsies after stopping romiplostim and all had disappearance of the reticulin and collagen.

A similar prospective bone marrow study has been done in ITP patients treated with eltrombopag [166,168]. In 93 patients who had bone marrow examinations at baseline and then after one and two years of treatment, 89 percent had no reticulin (MF-0) and 11 percent had mild reticulin (MF-1). Only one patient showed collagen deposition. These studies indicate that bone marrow biopsies need not be done on patients treated with TPO receptor agonists, unless other peripheral blood findings suggest otherwise.

Hepatotoxicity — The risk of hepatic decompensation is included in a Boxed Warning in the product information for eltrombopag when used in combination with interferon or ribavirin in individuals with chronic hepatitis C virus infection [169]. In this patient group, eltrombopag administration was associated with hyperbilirubinemia in 54 percent, compared with 25 percent in those who received placebo, and hepatic decompensation in 10 percent, compared with 5 percent in those who received placebo [76]. In ITP patients, liver function abnormalities occurred in about 11 percent; in general, these effects were mild and resolved with drug discontinuation (with many resuming the drug with no recurrence) [38].

The serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), and bilirubin should be measured prior to initiation of eltrombopag, every two weeks during dose titration, and monthly during stable dosing. Eltrombopag should be discontinued if there are progressive or persistent increases or other clinical signs of liver injury or hepatic decompensation.

Hepatotoxicity has not been reported with romiplostim or avatrombopag [45].

Rebound thrombocytopenia — A risk of TPO receptor agonists in ITP patients comes from abruptly stopping treatment. An unfortunate aspect of product labeling [169] for TPO receptor agonists in ITP is the recommendation that they be held if the platelet count rises over 400,000/microL; this is based on no identified risk but was part of the dosing algorithm of the initial clinical trials. Given the very short platelet life-span, stopping the drug results in a rapid decrease in platelet count; platelet counts may often drop from >400,000/microL to less than 5000/microL in less than one week. Rebound thrombocytopenia (defined as the platelet count dropping 10,000/microL below the prior baseline) was seen in approximately 10 percent of clinical trial patients [20].

SUMMARY

●TPO – A number of thrombopoietin (TPO) molecules have been developed; they define a new and growing family of molecules called the thrombopoietin (c-mpl ligand) family, based upon their common ability to bind and activate the TPO receptor, c-mpl (table 1). (See 'Thrombopoietins (c-mpl ligands)' above.)

●Clinical uses

•ITP – TPO receptor agonists (TPO-RAs) available for use in immune thrombocytopenia (ITP) are romiplostim, eltrombopag, avatrombopag, and, in China, hetrombopag. (See "Second-line and subsequent therapies for immune thrombocytopenia (ITP) in adults", section on 'TPO receptor agonists'.)

•Aplastic anemia – Eltrombopag and romiplostim (Japan only) are approved for use in patients with aplastic anemia for whom immunosuppressive therapy has failed or as initial therapy when combined with immunosuppression. (See "Treatment of aplastic anemia in adults".)

•Liver disease – Avatrombopag and lusutrombopag are non-peptide TPO receptor agonists approved for use in thrombocytopenic patients with liver disease undergoing an invasive procedure. (See "Hemostatic abnormalities in patients with liver disease", section on 'Invasive procedures'.)

•Other disorders – TPO receptor agonists are under evaluation for a wide range of other thrombocytopenic disorders, including those associated with chemotherapy, surgery, and myelodysplastic syndromes. (See 'Use of TPO receptor agonists' above.)

●Adverse effects – Side effects of TPO receptor agonists include thrombocytosis, possibly thrombosis, bone marrow reticulin fibrosis, and rebound thrombocytopenia. Hepatotoxicity has been reported with eltrombopag. Cost may also be an issue in some settings. (See 'Side effects and risks' above.)