Introduction
History of Thrombopoietin Use in Humans
TPO Receptor Agonists: A New Era in Therapy
TPO Receptor Agonists & ITP
Romiplostim in Splenectomized Patients
Romiplostim in Nonsplenectomized Patients
Long-term Dosing of Romiplostim
Long-term Safety Data of Eltrombopag
Final Thoughts

Introduction 
Back to top

It has been over 50 years since Harrington made the pivotal observation that women with immune thrombocytopenic purpura (ITP) frequently bear thrombocytopenic children. This insightful discovery prompted Harrington years later to be infused himself with a unit of whole blood from an ITP patient after convincing his colleague, Hollingsworth, to perform the procedure; his goal was to prove that such an infusion will recapitulate the disease.¹ This classic experiment demonstrated that a blood component, subsequently shown to be immunoglobulin, was the essential factor for the immune mediated platelet destruction that defines ITP. Since platelet production is assumed to be increased in patients with this disorder, traditional therapy has focused on moderating this immune response, which leads to accelerated platelet destruction. 

Although the vast majority of ITP patients do have a compensatory increase in megakaryopoiesis, curiously, plasma derived from some ITP patients can actually inhibit platelet production. This has prompted a re-evaluation of whether impaired megakaryopoiesis is another cause for the thrombocytopenia in this disease. Thrombopoietin (TPO) levels are not markedly elevated in ITP, implying that there are normal or increased numbers of megakaryocytes in the bone marrow of these patients. With this perspective, thrombopoietin levels are actually not inappropriately low for the degree of thrombocytopenia. However, it also suggests that supplemental thrombopoietin could help increase platelet production by overdriving the system and thereby correcting the thrombocytopenia.

History of Thrombopoietin Use in Humans
Back to top

Cloned almost simultaneously by several research groups in 1994, human thrombopoietin is composed of 353 amino acids. Thrombopoietin is a potent megakaryocyte colony-stimulating factor, and, along with other cytokines, increases the size and number of megakaryocytes. When administered to humans and other primates, thrombopoietin increases the number of marrow megakaryocytes and circulating platelets up to 10-fold. A truncated form of thrombopoietin called megakaryocyte growth and development factor (MGDF) coupled to polyethylene glycol (PEG) has also been administered to a variety of thrombocytopenic patients. Both recombinant thrombopoietin and PEG-MGDF clearly shortened the period required for platelet counts to return to normal following systemic chemotherapy. Although initial reports implied that recombinant thrombopoietin had minor toxicity, one subject given this drug developed non-neutralizing antibodies. In addition, several patients receiving PEG-MDGF developed neutralizing antibodies against their endogenous thrombopoietin, resulting in profound and persistent thrombocytopenia.² Largely because of this toxicity, both recombinant thrombopoietin and PEG-MGDF were withdrawn from clinical trials.

TPO Receptor Agonists: A New Era in Therapy
Back to top

By screening peptide libraries, several groups were able to identify peptides that bind the TPO receptor with high affinity. Because these peptides bear no structural resemblance to TPO but still bind and activate the TPO receptor, these compounds are called TPO receptor agonists. Several peptides have been identified, and they have been further modified to both prolong their half-life in plasma, as well as to increase their efficiency in activating the TPO receptor. The theoretical advantage of these compounds over standard recombinant thrombopoietin is that they bear little structural similarity with native thrombopoietin, and should not trigger autoimmune anti-thrombopoietin antibodies like PEG-MDGF. Romiplostim (AMG 531) is the most developed pharmaceutical in the peptide TPO receptor agonist category. It is composed of several copies of the TPO receptor-binding peptide spliced into a recombinant antibody. This peptide agonist competes with thrombopoietin for binding to the TPO receptor and activates the receptor in an identical fashion to endogenous thrombopoietin. When administered subcutaneously in humans, romiplostim produces a dose-dependent increase in platelet counts.

A similar approach has been used to identify other small molecules capable of binding and activating the TPO receptor, yet bearing little structural similarity to thrombopoietin. Screening small molecule libraries for compounds that have TPO-like activity has identified these so called TPO receptor nonpeptide agonists. The most developmentally advanced of this category is eltrombopag. It is an orally available drug that activates the TPO receptor by binding to the receptor’s transmembrane region. Therefore, unlike romiplostim, eltrombopag does not compete with endogenous thrombopoietin for binding to the TPO receptor. Like subcutaneously administered romiplostim, oral eltrombopag also produces a dose-dependent increase in the platelet count of healthy volunteers. Although not as far along in clinical trials, several other oral TPO receptor nonpeptide agonists, including AKR-501 and LGD-4665 are also currently being developed, and appear to show good thrombopoietic activity in humans.

TPO Receptor Agonists & ITP
Back to top

Since thrombopoietin levels are not extremely elevated in the plasma of patients with ITP, it seemed reasonable to attempt to ramp up platelet production with the administration of exogenous thrombopoietin. In 2002, S. Nomura et al. demonstrated that seven daily doses of PEG-MGDF significantly increased the platelet count in two out of four refractory ITP patients.³ Although this experiment appeared promising, as mentioned previously, PEG-MGDF development was halted because of its propensity to induce autoimmune antibodies against endogenous thrombopoietin.  Once TPO receptor agonists became available, several groups analyzed whether these compounds would also increase platelet counts in patients with ITP without inducing an immunotoxicity.

The first Phase II trial to test this hypothesis has been published.⁴ Sixteen ITP patients who had previously been treated with at least one prior form of therapy were enrolled in a double blind, placebo-controlled trial of romiplostim. Patients were treated with six weekly subcutaneous injections of 1 or 3 micrograms/kg of romiplostim. This peptide TPO receptor agonist increased the platelet count to greater than 50,000/ul in twelve patients (75%), including two patients in the 3 microgram/kg dose group who had platelet counts that exceeded 500,000/ul. There appeared to be no relationship between baseline thrombopoietin levels and platelet responses. This implied that one is able to overdrive platelet production even in patients with appropriate (i.e. normal) thrombopoietin levels. Reassuringly, no thrombotic complications occurred and the overall toxicity was low.

A placebo-controlled Phase I/II study of the oral TPO receptor agonist, eltrombopag was recently published.⁵ One hundred and eighteen patients with chronic ITP and platelet counts less than 30,000/ul were treated with one of three doses of eltrombopag. The primary end point was a platelet count greater than 50,000/ul on day 43. Eltrombopag increased the platelet count in a dose-dependent fashion. Although only 11% of the placebo-treated patients met the primary endpoint, 81% of patients treated with the highest dose of eltrombopag met the primary endpoint.  Importantly, the incidence of bleeding episodes also decreased with the use of this agent. In this study, there were no clear significant adverse events that were clearly attributed to the drug. The major side effect was headache. Thus, eltrombopag and romiplostim both appear to be capable of raising platelet counts in patients with ITP with low toxicity.

Updates from ASH 2007

At the American Society of Hematology Meeting in December, four significant abstracts were presented that further our knowledge of the use of TPO receptor agonists in ITP.

Romiplostim in Splenectomized Patients
Back to top

This Phase III study presented at ASH sought to determine the efficacy in patients who already had a splenectomy­—a common circumstance for ITP patients being considered for novel therapy.⁶ The primary objective of the study was to determine the rate of durable platelet responses during 24 weeks of treatment. A durable platelet response was defined as a platelet count greater than 50,000/ul for at least 6 of the last 8 treatment weeks. Sixty-three patients were enrolled into the study. Forty-two patients were randomized to receive weekly subcutaneous injections of romiplostim, and another twenty-one patients were randomized to receive placebo. The average platelet count was ~15,000/ul before the start of therapy; the duration of ITP was over eight years; and most patients had already received more than five other types of therapy.

Durable platelet responses were observed in 16 of the 42 patients (38%) of those treated with romiplostim, and not in any of the 21 patients (0%) treated with placebo. A total of 33 of the 42 patients (79%) treated with romiplostim had a platelet count greater than 50,000 for at least four weeks during the 24-week study, while this result was never seen in the placebo treated cohort (0%). During the 24-week study, the average number of weeks that the platelet count was over 50,000/ul was 12.3 weeks for patients who received romiplostim and only 0.2 weeks in the placebo-treated group. 

The most common side effects reported were arthralgias, myalgias, diarrhea, and fever. Two worrisome significant adverse events were reticulin fibrosis in one patient and a thrombosis in another patient. Thus, weekly romiplostim can achieve significant increases in the platelet counts in difficult ITP patients who are refractory to splenectomy and many other therapies.

Romiplostim in Nonsplenectomized Patients
Back to top

This study was designed in an identical fashion to the above study, except that it analyzed the response and toxicity rates of romiplostim in patients with ITP who did not have a splenectomy.⁷ Typically, cohorts of patients who have not had a splenectomy have more easily managed disease. This is probably because cohorts comprised solely of splenectomy failure patients are already restricted to patients with disease that is more severe and often refractory to other treatment. Consistent with this idea, patients enrolled in this trial typically were diagnosed within the past two years and had only received two to three other forms of prior therapy. Still the patients enrolled in this trial had significant thrombocytopenia (average platelet count 19,000/ul.) Forty-two patients were randomized to receive weekly subcutaneous injections of romiplostim, and another twenty-one patients were randomized to receive placebo.

Durable platelet responses defined as a platelet count greater than 50,000/ul for at least 6 of the last 8 treatment weeks were seen in 61% of those treated with romiplostim and in 4.8% of those treated with placebo. Approximately 88% of patients treated with romiplostim had a platelet count greater than 50,000 for at least 4 weeks during the 24-week study, while this result was only seen in about 14% of the placebo treated patients. The average number of weeks during the 24-week study that the platelet count was over 50,000/ul was 15.2 weeks in the group that received romiplostim and 1.3 weeks in the placebo-treated group. Another primary endpoint of this study was to determine whether patients receiving romiplostim needed any rescue medications for their ITP. The incidence of additional ITP therapy was 17.1% in the romiplostim treatment group and 61.9% of those in the placebo group.

This trial shows that in patients who have not had splenectomy, romiplostim can usually raise the platelet count. Although the true goal of ITP therapy is to decrease bleeding, it is notable that romiplostim did not decrease the incidence of either overall or severe bleeding events in this study. Most likely this is because the incidence of significant bleeding in most ITP patients is low, and therefore is not altered by TPO receptor agonist therapy. 

Long-term Dosing of Romiplostim
Back to top

In another study presented at ASH 2007, researchers evaluated the long-term dosing of romiplostim in thrombocytopenic patients with chronic ITP. ⁸ This trial was an open label extension study of patients who previously participated in another romiplostim trial. Sixty percent of the 137 patients enrolled in this study had a splenectomy. It was designed to determine the long-term safety of this compound.  The secondary endpoints were defined as the duration of the platelet response, whether anti-TPO antibodies developed, and whether concurrent forms of ITP therapy could be discontinued. Approximately 60% of patients enrolled in this study of long-term therapy with romiplostim had platelet counts greater than 50,000/ul regardless of how long they were treated (including 14 patient treated as long as 112 weeks).  This suggests that patients who respond to this therapy frequently have a sustained response.  Forty-three percent of the thirty patients who were receiving other ITP treatment before romiplostim could discontinue their supplemental therapy.  Headache was the most common adverse effect.  One patient developed an antibody against romiplostim, but it did not cross-react with endogenous TPO.  Mentioned at the meeting was an incidence of reticulin fibrosis in 8 out of the 200 romiplostim-treated patients who were examined by bone marrow biopsies.  Two of these patients had repeat bone marrow biopsies, and both had decreased but not absent reticulin fibrosis.

Long-term Safety Data of Eltrombopag
Back to top

Long-term safety data of eltrombopag in chronic ITP patients was also presented at ASH 2007.⁹ This was an open label extension study of patients who participated in a previous eltrombopag trial.  One hundred and nine patients were enrolled after at least a 4-week washout period.  Forty-four percent of patients had previously undergone a splenectomy, over 70% had a baseline platelet count of less than 30,000/ul, and 44% had a platelet count of less than 15,000/ul.

Overall, 80% of patients in this study (86/108) had a platelet count greater than 50,000/ul on at least one occasion, and 54% of patients had a platelet count above this level for at least 10 weeks (43/80).  Limiting the analysis to the 49 patients who previously responded to eltrombopag, 92% of patients (45) had a platelet count that reached 50,000/ul.  Of the 40 patients who were on other therapy for ITP, 14 (35%) were able to discontinue their other treatment.  Similar to romiplostim, headache was the most common adverse event.  One patient had a pulmonary embolism that was felt to be not clearly related to therapy.  Overall, long-term therapy with eltrombopag seems to be well tolerated and associated with sustained efficacy.

Final Thoughts
Back to top

The new TPO receptor agonists including romiplostim and eltrombopag are likely the first of many new agents that will be available to increase platelet counts. Thus far, they have response rates in ITP patients of 60-80%, and many of these responses are quite durable. It is noteworthy that these agents probably do not alter the natural history of this disease, so patients will need to continuously take these medications for a sustained effect. In addition, the long-term toxicity, potentially including thrombosis and bone marrow fibrosis, needs to be further analyzed. Therefore, when balanced against the low mortality (1% or less) and impressive sustained efficacy (66% complete remission plus another 22% partial remission) of splenectomy, it is too early to conclude that TPO receptor agonists should be routinely used as first-line therapy.

References: