Eltrombopag,a thrombopoietin- receptor agonist in the treatment of adult chronic immune thrombocytopenia: a review of the efficacy and safety profile

Abstract

Chronic immune thrombocytopenia (ITP) is an autoimmune disorder characterized by a low platelet count that has persisted for more than 12 months. Patients may be asymptomatic but those with severe disease may have significant morbidity and require treatment. Corticosteroids and intravenous immunoglobulin are recommended as first-line treatments. Recently, two thrombopoietin-receptor agonists, romiplostim and eltrombopag have been licensed for the treatment of chronic ITP. The current indications for thrombopoietin-receptor agonists are for splenectomized adult patients with chronic ITP who are refractory to other treatments and adult nonsplenectomized patients in whom splenectomy is contraindicated. This article reviews data on the pharmacology, clinical efficacy and safety profile of eltrombopag in the treatment of ITP.

Introduction

Immune thrombocytopenia (ITP) is an immune-mediated disease characterized by transient or persistent decrease of the platelet count to less than 100 × 10⁹/liter. The term ‘newly diagnosed ITP’ is used to describe all cases at diagnosis. Persistent ITP is defined as ITP lasting between 3 and 12 months from diagnosis while chronic ITP is defined as the presence of ITP for more than 12 months. Secondary ITP includes all forms of immune-mediated thrombocytopenia except primary ITP [Rodeghiero et al. 2008].

The major therapeutic goal for ITP is to use minimal treatment to maintain platelet counts that are sufficient to reduce bleeding symptoms with the least amount of side effects. Since most fatal bleeding in adult ITP occurs with platelet counts lower than 30 × 10⁹/liter [Cohen et al. 2000; Portielje et al. 2001], current guidelines suggest that treatment should only be considered in symptomatic patients with counts less than 30 × 10⁹/liter [Rodeghiero et al. 2008; Provan et al. 2010]. Corticosteroids are the standard initial treatment [Cines and Blanchette, 2002; Provan et al. 2010]. Prednisolone 0.5–2 mg/kg/day is the common starting dose for patients with ITP. After the platelet count increases to more than 50 × 10⁹/liter, prednisolone should be tapered off to the minimal effective dose required to maintain a platelet count over 30–50 × 10⁹/liter. If no significant increase in platelet count is observed after 4 weeks of treatment with high-dose prednisolone, the drug should be tapered off quickly [Pizzuto and Ambriz, 1984; Ben-Yehuda et al. 1994].

One to four cycles of dexamethasone 40 mg/day for 4 days is the preferred corticosteroid regimen in some centers, with response rates of 50–80% in newly diagnosed adult patients with ITP [Cheng et al. 2003; Mazzucconi et al. 2007]. Some patients had prolonged response after treatment with high-dose dexamethasone. However, a recent study suggested that dexamethasone was not more effective than conventional doses of prednisolone [Bae et al. 2011]. According to the International Consensus Report [Provan et al. 2010] on investigation and management of ITP, prednisolone, dexamethasone, or methyprednisolone are all acceptable first-line treatments.

Intravenous immunoglobulin (IVIG) and anti-D are effective in raising the platelet counts but the effects are usually transient. They are recommended as first-line treatments, especially in emergency situations [Scaradavou et al. 1997; George et al. 2003; Spahr and Rodgers, 2008].

For patients with chronic ITP whose condition failed to respond to corticosteroids or who have severe adverse effects from corticosteroids, splenectomy is the second-line therapy in many centers. However, about 15–20% of patients do not respond to splenectomy and another 15–20% of responders relapse weeks, months, or years later [Schwartz et al. 2003; Kojouri et al. 2004]. In addition, many patients with chronic ITP are reluctant to have splenectomy because of fear of complications such as bleeding, infection, thrombosis, and the reported mortality rates of 0.2–1.0% [Kojouri et al. 2004].

In patients who were refractory to or relapsing after splenectomy or when splenectomy was contraindicated, a variety of immunosuppressive or cytotoxic drugs such as vincristine, cyclophosphamide, aziathioprine, dapsone, cyclosporine A, mycophenolate mofetil and rituximab were used with a response rate of 20–80% [Hernández et al. 1995; Reiner et al. 1995; Kappers-Klunne and van’t Veer, 2001; Maloisel et al. 2004; Kotb et al. 2005; Boruchov et al. 2007; Godeau et al. 2008]. However, most of the studies with immunosuppressive agents were not randomized clinical trials and patients may have serious infection complications from prolonged use.

In a prospective phase II trial patients with ITP were given rituximab at doses of 375 mg/m² weekly for 4 weeks [Godeau et al. 2008]. One-third of the patients had a platelet count of 50 × 10⁹/liter or higher without any additional treatment after 2 years of observation. Most patients with a durable (> 1 year) complete response will respond to repeat treatment if they relapse. A combination of rituximab with high-dose dexamethasone as initial therapy may result in an even higher response rate [Zaja et al. 2010]. However, after 3 years of follow up, around 25% of the patients with ITP had platelet counts over 50 × 10⁹/liter.

Rituximab is currently not registered for the treatment of chronic ITP and can cause fulminant hepatitis in hepatitis B carriers [Fianchi et al. 2007]. Therefore, rituximab is contraindicated in patients with active hepatitis B infection and prophylaxis with lamuvidine is required in patients who carry hepatitis B without active infection. Also, more than 50 cases of progressive multifocal leukoencephalopathy associated with rituximab treatment have been reported in patients with lymphoma and systemic lupus erythematosus [Carson et al. 2009]. Additional long-term safety data are required before rituximab can be recommended as a frontline therapy.

Recently, impaired platelet production was observed in many patients with ITP [Hoffman et al. 1985; Chang et al. 2003; Houwerzijl et al. 2004; McMillan et al. 2004]. Therefore, stimulation of megakaryopoiesis by thrombopoietin or thrombopoietin-mimetic agents may be useful in the treatment of ITP. Recombinant thrombopoietin had been shown to increased platelet counts in patients with ITP [Nomura et al. 2002; Zhao et al. 2004] but was associated with production of autoantibodies that crossreact with and neutralize endogenous thrombopoietin, leading to severe thrombocytopenia [Li et al. 2001; Basser et al. 2002]. Recently, two thrombopoietin-receptor (TPO-R) agonists, romiplostim (AMG-531, Nplate; Amgen, Thousand Oaks, CA, USA) and eltrombopag (Revolade, Promacta; GlaxoSmithKline, Brentford, UK) have been licensed for the treatment of chronic ITP [Wang et al. 2004; Erickson-Miller et al. 2005]. They have no sequence homology with native thrombopoietin and should not stimulate production of antithrombopoietin antibodies. The current indications of TPO-R agonists are for relapsed splenectomized adult patients with chronic ITP who are refractory to other treatments or adult nonsplenectomized patients in whom splenectomy is contraindicated [Provan et al. 2010].

This article reviews data on the pharmacology, clinical efficacy, and safety profile of eltrombopag in the treatment of ITP.

Mechanism of action and pharmacokinetics

Eltrombopag is a low molecular weight, synthetic nonpeptide TPO-R agonist [Erickson-Miller et al. 2005]. It has excellent oral bioavailabilty with a peak concentration occurring 2–6 h after oral administration and a half life of 21–32 h [GlaxoSmithKline, 2011]. It is highly bound to human plasma proteins (>99%). Unlike native thrombopoietin which binds to the extracellular domain of the thrombopoietin receptor, eltrombopag selectively binds to the trans-membrane domain of the receptor. It stimulates megakaryocytopoiesis through the Janus kinase/signal transducer and activator of transcription (JAK/STAT) signaling pathway. Eltrombopag did not affect agonist-induced platelet aggregation or activation in in vitro and ex vivo studies using platelet samples from healthy volunteers and patients with chronic ITP [Erhardt et al. 2004].

Each film-coated tablet contains 25 or 50 mg of eltrombopag olamine. Administration of eltrombopag concomitantly with polyvalent cations was found to significantly reduce eltrombopag absorption by up to 60% [GlaxoSmithKline, 2011], therefore eltrombopag should be taken at least 4 h before or after antacids, dairy products, other food products, or mineral supplements containing polyvalent cations. Eltrombopag is metabolized in the liver with cytochrome P450 isoenzymes CYP1A, CYP2C8, and uridine diphosphate glucuronosyltransferase (UGT 1A1 and UGT1A3) as major pathways. Eltrombopag did not inhibit or induce cytochrome P450 enzymes. However, clinical drug interactions between eltrombopag and cytochrome P450 substrates, inducers, or inhibitors have not been extensively investigated. Eltrombopag is not a substrate for P-glycoprotein or organic anion transporting polypeptide (OATP) 1B1. However, it is an inhibitor of OATP 1B1and coadministration of eltrombopag with rosuvastatin (an OATP1B1 substrate) was accompanied by a twofold increase in plasma rosuvastatin. Therefore, in patients with ITP taking eltrombopag, the dose of concomitant statins (rosuvastatin, pravastatin, simvastatin, and lovastatin) should be reduced and statin side effects should be monitored carefully. In total, 59% of a dose of eltrombopag is excreted in feces (20% as unchanged drug) and 31% in urine (0% as unchanged drug). [GlaxoSmithKline, 2011].

Systemic exposure to the drug is increased in people of East Asian descent (based on estimates from a population-based pharmacokinetic model). As a consequence, an initial dose decrease to 25 mg/day is recommended in these patients.

Clinical efficacy studies

The clinical efficacy of eltrombopag in chronic ITP has been examined in a 6-week dose-finding trial [Bussel et al. 2007], a 6-week phase III trial [Bussel et al. 2009], a 6-month phase III trial [Cheng et al. 2011b], an extension study [Saleh et al. 2011] and one intermittent therapy trial [Bussel et al. 2008].

Dose response study (TRA100773A)

The objective of the TRA100773A dose-finding study was to determine the optimal dose of eltrombopag. A total of 118 patients with chronic ITP were randomized in a 1:1:1:1 ratio into four groups receiving 30, 50, 75 mg of eltrombopag, or placebo daily for 6 weeks [Bussel et al. 2007]. The study patients must have had chronic ITP for at least 6 months and a platelet count less than 30 × 10⁹/liter at baseline. These patients had not responded to at least one prior therapy, including splenectomy, or had relapsed within 3 months of previous therapy. Within each treatment arm, patients were also stratified according to use of concomitant ITP therapy, platelet count less than 15×10⁹/liter or more than 15 × 10⁹/liter, and splenectomy status. The primary endpoint was a platelet count greater than 50 × 10⁹/liter or more at the day 43 visit. If a patient’s platelet counts were greater than 200 × 10⁹/liter, treatment was discontinued, but the patient would continue to be followed up over the full trial duration and counted as a responder. The incidence and severity of bleeding events were assessed using the World Health Organization (WHO) Bleeding Scale. Significantly far more patients in the eltrombopag 50 mg group (n = 30) and 75 mg group (n = 28) had platelet counts greater than 50 × 10⁹/liter compared with patients (n = 29) receiving placebo (70% and 81% respectively versus 11%, p values < 0.001 for both groups). The response rate of recipients of eltrombopag 30 mg daily (n = 30) did not significantly differ from that of subjects receiving placebo (28% versus 11%). The median day 43 platelet counts in subjects on eltrombopag 50 or 75 mg/day were 128 × 10⁹/liter and 183 × 10⁹/liter respectively compared with only 16 × 10⁹/liter in the placebo group [Bussel et al. 2007].

This improvement in platelet counts was accompanied by a significant reduction in bleeding symptoms in patients in the eltrombopag 50 or 75 mg/day groups. Thrombopoietin levels were within the normal range at baseline in all four groups (54-–57 ng/ liter) and were unaffected by the response to eltrombopag therapy [Bussel et al. 2007].

Phase III study (TRA100773B)

Based on the results of the dose-finding TRA100773A study, another 114 patients were randomized in a 2:1 ratio to receive eltrombopag 50 mg/day or placebo [Bussel et al. 2009]. Inclusion criteria and primary response were similar to the phase II dose-finding TRA100773A study [Bussel et al. 2007]. In this trial, if the platelet count was still less than 50 × 10⁹/liter after 3 weeks of treatment with eltrombopag 50 mg/day, the dosage would be increased to 75 mg/day for the remaining 3 weeks. Again, within each treatment arm, patients were stratified according to use of concomitant ITP therapy, platelet count less than 15 × 10⁹/liter or more than 15 × 10⁹/liter and splenectomy status. Subjects on concomitant ITP medications (corticosteroids, azathioprine, danazol, cyclosporine A or mycophenolate mofetil) could be enrolled provided that the dose had been stable for at least 1 month prior to enrollment. These subjects were allowed to continue their concomitant ITP medication during the study. Significantly more subjects on eltrombopag (n = 73) demonstrated a response with platelet counts greater than 50 × 10⁹/liter on day 43 compared with the placebo group (n = 37) (59% versus 16%; p < 0.0001). In 34 patients receiving eltrombopag, the dosage was increased to 75 mg/day on or after day 22 of the study because of no response to a 50 mg/day dosage. Of these, 10 patients (29%) achieved a platelet count greater than 50 × 10⁹/liter on day 43. Patients receiving eltrombopag had lower risks of bleeding compared with those on placebo [odds ratio (OR) 0.27; 95% confidence interval (CI) 0.09–0.88; p = 0.029] [Bussel et al. 2009]. Severe bleeding events during the 6 weeks occurred only in patients who did not respond to eltrombopag, or in those receiving placebo. No clinically significant bleeding (WHO grade 2–4) occurred while patients had platelet counts greater than 50 × 10⁹/liter.

In both the phase II TRA 100773A and phase III 100773B studies, similar response rates to eltrombopag were observed irrespective of the splenectomy status, whether the patients were on concomitant ITP medications and whether the baseline counts were less than 15 × 10⁹/liter. In patients responding to treatment, the platelet counts started to increase after 1 week of treatment, peaked around the second week and were maintained throughout the remaining weeks. For most patients, the platelet count returned to baseline levels within 2 weeks of stopping eltrombopag therapy. Transient decreases in platelet counts to less than 10 × 10⁹/liter and at least 10 × 10⁹/liter below baseline (rebound thrombocytopenia) were observed in 11% and 13% of patients receiving eltrombopag and placebo respectively following treatment discontinuation in this phase III trial [Bussel et al. 2009]. In the 6-week post-treatment period of the phase III 100773B trial, the eltrombopag and placebo groups had similar incidences of bleeding. Therefore, increased incidence of rebound thrombocytopenia with worsening of bleeding symptoms following withdrawal of eltrombopag was not observed.

Study TRA102537 (RAISE)

RAISE (Randomized Placebo-controlled ITP Study with Eltrombopag) [Cheng et al. 2011b] was a double-blind phase III study addressing the safety and efficacy of prolonged 6-month treatment of patients with ITP with eltrombopag. Adult patients with ITP of more than 6 months’ duration, whose condition had previously responded to at least one ITP treatment and had platelet counts less than 30 × 10⁹/liter at the time of enrollment were eligible for the study. Concomitant ITP medications could be continued if the dosage had been stable for at least 1 month. Again, within each treatment arm, patients were stratified according to use of concomitant ITP therapy, baseline platelet counts less than 15 × 10⁹/liter or greater than 15 × 10⁹/liter and splenectomy status. A total of 197 patients were randomized in a 2:1 ratio to treatment with eltrombopag (n = 135) or placebo (n = 62). Study patients were started with 50 mg/day of eltrombopag or matching placebo. At the end of 3 weeks of treatment, if the platelet counts were less than 50 × 10⁹/liter, the dose could be increased to 75 mg/day. However, if subjects’ platelet counts were greater than 200 × 10⁹/liter, the study medication would be reduced to 25mg per day. After 6 weeks of therapy, if the platelet counts were more than 100 × 10⁹/liter on two successive visits, concomitant ITP medications could be reduced or discontinued. The primary endpoint was the odds of responding with platelet counts between 50 and 400 × 10⁹/liter at least once during the 6-month study period. Bleeding symptoms were prospectively evaluated using the WHO Bleeding Scale. Other secondary endpoints included median platelet counts, reduction of baseline ITP medication, use of rescue medication, health-related quality of life (HRQOL), and safety.

The median age was 52 years in the placebo arm and 47 years in the eltrombopag group. There was a higher proportion of female patients (69%) in both groups and 70% of patients were white.

Approximately half of the patients in the placebo and eltrombopag groups (50% and 47% respectively) were receiving ITP medication at randomization or had baseline platelet counts of less than 15 × 10⁹/liter (48% and 50% respectively). About one-third of the patients in each arm had been splenectomized. About 80% of the placebo and eltrombopag-treated patients had received at least two prior therapies, and more than 50% in each group had three or more prior therapies.

In the primary efficacy analysis, patients receiving eltrombopag were eight times more likely to achieve the primary endpoint of a platelet count between 50 and 400×10⁹/liter (OR 8.20, 95% CI 3.59–18.73, p < 0.001) compared with those in the placebo group. Baseline median platelet counts were around 16 × 10⁹/liter in both groups. In the placebo group, the counts never exceeded 30 × 10⁹/liter. In contrast, the median platelet counts rose to 36 × 10⁹/liter in the eltrombopag group after 1 week of treatment and stayed around 52–91 × 10⁹/liter for the remainder of the study. Similar responses were observed irrespective of splenectomy status, baseline platelet count, or baseline ITP medication use.

Like the observation in the two 6-week trials, the platelet counts started to increase by the end of the first week of treatment, peaked at the second week, and were maintained throughout the 6-month study period. After stopping eltrombopag, the platelet counts usually returned to baseline levels in 2 weeks.

Bleeding symptoms were similar in both groups at baseline. From day 15 onwards, patients in the eltrombopag group had less bleeding symptoms compared with those in the placebo group (79% versus 93% for grade 1–4 bleeding and 33% versus 53% for grade 2–4 bleeding). Three eltrombopag-treated patients (2%) had grade 3 or higher adverse bleeding events compared with four patients (7%) in the placebo group (p = 0.03). The eltrombopag and placebo groups had similar incidence of rebound thrombocytopenia (7% each). No increased incidence of serious bleeding episodes was observed in the eltrombopag group in the post-treatment period (4%; n = 6) compared with placebo (10%; n = 6).

A total of 59% (37/63) of patients receiving eltrombopag were able to reduce or discontinue concomitant ITP medications compared with 32% (10/31) in the placebo group (p = 0.02). Rescue medications were required in 18% (24/135) of patients receiving eltrombopag compared with 40% (25/62) in the placebo group (p = 0.001). Five out of eight HRQOL domains (physical function, physical role, emotional function, vitality, and social function) showed significant improvements from baseline in patients receiving eltrombopag [Cheng et al. 2011b].

EXTEND study

EXTEND is an ongoing, multicenter, open-label extension study [Saleh et al. 2011]. Patients with ITP who had not experienced any eltrombopag-related serious adverse event or drug intolerance and had completed treatment and follow up with either eltrombopag or placebo in a prior eltrombopag clinical study are eligible. They must have a washout period of at least 4 weeks. Eltrombopag treatment is initiated at 50 mg once daily. The dose of eltrombopag will be adjusted to the minimal effective dose (between 75 mg once daily to 25 mg once daily, or less) necessary to maintain platelet counts of at least 50 × 10⁹/liter in conjunction with the minimal dose of concomitant ITP medication. Patients will be followed until withdrawal or eltrombopag becomes commercially available. The primary objective is to assess the long-term safety and tolerability of eltrombopag as measured by clinical laboratory tests, and frequency of adverse events. Secondary endpoints include the proportion of patients who achieve platelet counts of at least 50 × 10⁹/liter at least once during treatment.

A total of 301 patients were enrolled in the ongoing EXTEND study. Median duration of follow up was 121 weeks. More than half (58%) of the patients had been followed for more than 104 weeks, 28% (84/301) for more than 3 years, and 8% (23/301) for over 4 years. A total of 79% of the patients were white and 15% (45/301) were Asians. About one-third of the patients had splenectomy (38%, n = 115), or were on concomitant ITP medications (34%, n = 101). Overall, 88% (264/301) of patients achieved platelet counts of at least 50 × 10⁹/liter at any time during the study. Median platelet counts increased to at least 50 × 10⁹/liter by week 2, and remained consistently at 50 × 10⁹/liter or higher throughout observation. At baseline, 56% of patients reported bleeding symptoms (WHO grades 1–4) compared with 16%, 19%, and 9% of patients at 52, 104, and 156 weeks respectively. At 52, 104, and 156 weeks, the proportion of patients with clinically significant bleeding (WHO grades 2–4) was reduced from 16% at baseline to 3%, 5%, and 0% respectively. At baseline, 101/301 patients reported the use of concomitant ITP medications. Of these 101 patients, 69% attempted to reduce or discontinue concomitant ITP medication and 43% had a sustained reduction or permanently stopped at least one concomitant medication. Late resistance to eltrombopag was not observed. In some patients, the platelet counts dropped following reduction or discontinuation of concomitant medications, and required rescue therapy. It is difficult to ascertain whether this was due to resistance to eltrombopag or reduction of concomitant ITP medications

REPEAT trial

The REPEAT (Repeat Exposure to Eltrombopag in Adults with Idiopathic Thrombocytopenic Purpura) trial was a phase II, noncomparative study of the intermittent treatment of ITP with eltrombopag. Patients with previously treated chronic ITP and baseline platelet counts of 20–50 × 10⁹/liter (n = 66) were eligible. Patients received eltrombopag 50 mg/day for up to a maximum of 6 weeks for three cycles. Each cycle of treatment was followed by an off-therapy period of 4 weeks. A response to treatment was defined as a platelet count of more than 50 × 10⁹/liter and at least twice the baseline count at day 43 of the treatment cycle. If the platelet count was more than 200 × 10⁹/liter at day 43 of the treatment cycle, eltrombopag treatment was discontinued but the patient was still followed up and counted as a responder. Patients who did not respond in cycle 1 did not continue in the study. The primary endpoint of REPEAT was consistency between cycles – whether a similar proportion of patients who showed a response in cycle 1 also responded in cycles 2 or 3 [Bussel et al. 2008]. A total of 80% of patients responded in cycle 1 and were thus permitted to continue in the study. A total of 87% of cycle 1 responders also responded in cycle 2 or 3. Median platelet counts remained above 70 × 10⁹/liter after day 8 in all three cycles.

Safety issues

The safety concerns of eltrombopag are based on toxicity studies in rats, mice, rabbits, and dogs [European Medicine Agency, 2009] and those reported from clinical trials [Bussel et al. 2007, 2008, 2009; Cheng et al. 2011b; Saleh et al. 2011].

Hepatocyte degeneration, often accompanied by increased serum liver enzymes, was observed in mice, rats, and dogs at doses that were associated with morbidity and mortality [European Medicine Agency, 2009]. Approximately 10% of the patients receiving eltrombopag in the 6-week TRA100773A and TRA100773B studies [Bussel et al. 2007, 2009], in the RAISE study [Cheng et al. 2011b], and in the EXTEND study [Saleh et al. 2011] had elevations of alanine aminotransferase (ALT) at least three times the upper limit of normal compared with around 3% in the placebo group (p > 0.05). In these patients, the elevated ALT returned to normal levels while the patients continued on eltrombopag or promptly after treatment discontinuation. In the EXTEND study, all bilirubin elevations were indirect bilirubin, which is not indicative of serious liver injury. Moreover, the Hepatobilary Laboratory Abnomalities (HBLAs) observed during the EXTEND study were not always the same HBLA experienced by the patient during the previous eltrombopag study. Some patients who had experienced HBLAs during the RAISE study did not have recurrence of the HBLAs when retreated with eltrombopag in the EXTEND study. So far there is no clinical evidence that eltrombopag, at the recommended dose, would result in serious irreversible liver damage. However, liver function tests should be monitored regularly, and if there is progressive increase in serum aminotransferases, the drug should be stopped. For patients with underlying liver failure, the drug should be used with caution and a lower starting dose of 25 mg is recommended.

In the RAISE study [Cheng et al. 2011b], three patients (2%) receiving eltrombopag reported on-therapy thromboembolic events (two with pulmonary embolism, one had deep vein thrombosis) compared with none in the placebo group. All three patients had risk factors for venous thrombosis and the platelet counts were less than 50,000/µl around the time of the thrombotic events. In the EXTEND study [Saleh et al. 2011], 18 patients (6%) experienced 25 confirmed or suspected thromboembolic events with an incidence rate of 3.02/100 patient years. Deep vein thrombosis (n =10) and cerebral vascular events (n = 7) were the most common thromboembolic events. The majority of these patients (15/18) experienced thromboembolic events at a platelet count within the normal range and lower than the maximum platelet count achieved during eltrombopag treatment. All 18 patients had at least one thromboembolic risk factor such as hypertension, smoking, or obesity. The frequency of thromboembolic events in patients treated with eltrombopag in the EXTEND study is similar to that reported for the ITP patient population [Sarpatwari et al. 2010]. The data so far do not suggest an increased risk of thromboembolism with eltrombopag treatment. However, in patients with underlying risk factors of thrombosis, eltrombopag should be used with caution, perhaps starting at a lower dose, monitoring the platelet counts very carefully, and aiming at the minimal platelet counts adequate to reduce bleeding symptoms.

TPO-R agonists may increase the risk for developing or progressing reticulin fiber deposition in bone marrow [Douglas et al. 2002]. Over 180 on-treatment bone marrow biopsies were performed on more than 100 patients in the EXTEND study [Saleh et al. 2011]. These patients had received eltrombopag for 1–4 years and 39 patients had two or more bone marrow biopsies. No significant increase in fibrosis was observed. None of the patients had an abnormal karyotype of bone marrow cells, increased bone marrow blast count greater than 3%, or any clinically relevant bone marrow abnormality other than those compatible with ITP. For patients on eltrombopag, peripheral blood smears should be examined for morphological abnormalities such as teardrop, nucleated red blood cells, leucoerythroblastic pictures, dysplastic changes, or cytopenia. If such abnormalities develop or are worsening, a bone marrow biopsy should be performed. A loss of response or failure to maintain a platelet response with eltrombopag treatment within the recommended dosing range should also prompt a search for causative factors such as myelofibrosis.

One concern about treatment with TPO-R agonists is the possibility of rebound thrombocytopenia upon withdrawal of the drug. For patients treated with eltrombopag, thrombopoietin levels were within the normal range at baseline and were unaffected by eltrombopag therapy. Transient decreases in platelet counts to less than 10 × 10⁹/liter and at least 10 × 10⁹/liter below baseline (rebound thrombocytopenia) were observed in 11% and 13% of eltrombopag and placebo recipients following treatment discontinuation in the controlled clinical trials [Bussel et al. 2009]. Post-therapy bleeding events were reported by a lower proportion of patients receiving eltrombopag (4%) compared with placebo (10%) in the RAISE study [Cheng et al. 2011b]. Nonetheless, platelet counts and bleeding symptoms should be monitored closely following discontinuation of eltrombopag.

Treatment-related cataracts were observed in rodents and were dose and time dependent. The clinical relevance of these findings is unknown. In the RAISE Study [Cheng et al. 2011b], similar incidences of worsening cataracts were observed in the eltrombopag and placebo group (8% and 10% respectively). However, for patients on eltrombopag, routine monitoring for cataracts is recommended. Renal tubular toxicity was observed in studies of up to 14 days’ duration in mice and rats at doses that were generally associated with morbidity and mortality. In humans, fecal elimination of eltrombopag is the dominant route and only 31% of the metabolized drug is excreted in urine [GlaxoSmithKline, 2011]. Currently, no dose adjustment is needed in patients with renal impairment, but eltrombopag should be used with caution in these patients with close monitoring of serum creatinine urine analysis.

Eltrombopag was not carcinogenic in mice at doses up to 75 mg/kg/day or in rats at doses up to 40 mg/kg/day (exposures up to four times the human clinical exposure based on area under the curve). In the pre- and postnatal development study in rats, there were no undesirable effects on pregnancy [European Medicine Agency, 2009], parturition, or lactation of female rats at maternally nontoxic doses (10 and 20 mg/kg/day) and no effects on the growth, development, neurobehavioral, or reproductive function of the offspring. However, at a maternally toxic dose of 60 mg/kg/day (six times the human clinical exposure based on AUC) in rats, eltrombopag treatment was associated with embryo lethality and reduced fetal body weight [European Medicine Agency, 2009]. Eltrombopag is not recommended during pregnancy and in women of childbearing potential not using contraception.

Eltrombopag was detected in the plasma of all rat pups for the entire sampling period following administration of medicinal product to the mothers, suggesting that rat pup exposure to eltrombopag was likely via lactation [European Medicine Agency, 2009]. Therefore, eltrombopag should be used with caution in lactating mothers.

Conclusions

Eltrombopag is an orally administered small-molecule nonpeptide TPO-R agonist. It has been shown to effectively increase platelet counts and reduce bleeding symptoms in patients with chronic ITP with overall response rate of 60–80%. Eltrombopag is well tolerated and has a good safety profile. It is recommended for splenectomized patients with ITP who are refractory to other treatments (e.g. corticosteroids, immunoglobulins). It may also be considered as second-line treatment for adult nonsplenectomized patients who refused surgery or in whom surgery is contraindicated.

The recommended starting dose of eltrombopag is 50 mg once daily. For patients of East Asian ancestry, eltrombopag should be initiated at a reduced dose of 25 mg once daily. After initiating eltrombopag, if no significant increase in platelet counts is observed after 2–3 weeks of treatment, could the dose may be increased. After achieving a stable platelet count on a specific dose, the dose should be further adjusted to the lowest dose sufficient to maintain a platelet count of around 50 × 10⁹/liter with minimal bleeding symptoms. In the EXTEND study, 13 of 301 patients (4.5%) had prolonged remission (median 50 weeks) without ITP therapy following discontinuation of eltrombopag [Cheng et al. 2001a]. The median time from diagnosis of ITP was 24 months (range 6–73 months) and the median duration of eltrombopag therapy was 237 days (range 14–1014 days). Therefore, in some patients with chronic ITP, it may be possible to gradually taper and eventually discontinue eltrombopag. If the disease relapses, eltrombopag can be restarted at the previous effective dose. According to data from the REPEAT study, such patients will still be responsive to eltrombopag.

In most responding patients, platelet counts start to increase after the first week of therapy and peak at the second week. Therefore, eltrombopag may be used in the preparation of patients with chronic ITP for elective surgery. Patients can start taking eltrombopag at home 2 weeks before the scheduled surgery, thus avoiding presurgical admission for IVIG infusion, which is common practice.

Eltrombopag should be taken at least 4 h before or after any products such as antacids, dairy products (or other calcium-containing food products), or mineral supplements containing polyvalent cations (e.g. iron, calcium, magnesium, aluminum, selenium, and zinc).

Eltrombopag is not recommended for use in children and adolescents below the age of 18 years due to insufficient data on safety and efficacy.

Even though no increased incidence of thromboembolic events, myelofibrosis, or irreversible hepatic damage was reported in follow up over 4 years, patients on eltrombopag should still be monitored closely for such adverse events.