On August 22, 2008, the US Food and Drug Administration (FDA) granted marketing approval (licensure) to romiplostim (Nplate, Amgen Inc) for the treatment of thrombocytopenia in patients with chronic immune (idiopathic) thrombocytopenic purpura (ITP) who have had an insufficient response to corticosteroids, immunoglobulins, or splenectomy. Romiplostim is a recombinant protein, thrombopoietin (TPO) receptor agonist that stimulates bone marrow megakaryocytes to produce platelets. The marketing approval was based on demonstration of a favorable risk-benefit profile, where the major benefit pertained to demonstration of a sustained increase in blood platelets, an accepted surrogate for clinical benefit in the proposed clinical setting.
Clinical and Regulatory Background
Chronic ITP is an acquired autoimmune disorder that primarily affects adults, with a female-to-male ratio of approximately 2:1 and is characterized by a low platelet count, mainly secondary to accelerated platelet destruction.[1-3] In addition to accelerated platelet destruction, some patients appear to have impaired platelet production. Nevertheless, blood TPO levels are generally not increased in patients with chronic ITP. The diagnosis of ITP is based on the detection of thrombocytopenia and the exclusion of “secondary” causes of the abnormally low platelet count. Hemorrhage is the major clinical complication of chronic ITP.
In general, no treatment is necessary for asymptomatic patients with platelet counts in excess of 30 × 109/L unless there are mitigating factors such as a planned surgical procedure or participation in contact sports that may increase the risk of bleeding. For patients requiring treatment, initial options include glucocorticoids, intravenous immunoglobulins (IVIG), and anti-D immunoglobulin. When ITP does not respond to these therapies, splenectomy results in improved platelet counts in approximately two-thirds of patients. Failure to respond to splenectomy and available medications may result in fatal hemorrhage if the patient continues to experience severe thrombocytopenia.
The initial clinical experience with recombinant TPO molecules was complicated in some patients by the development of neutralizing antibodies that resulted in thrombocytopenia. Cross-reactivity between the drug-induced antibodies and intrinsic TPO was implicated in these reactions, and further development of TPO-sequence homologous molecules was halted. Because romiplostim lacks sequence homology with TPO, investigators proposed that it may have less potential for immunologic complications. Nevertheless, the detection of antibody formation to romiplostim and intrinsic TPO was a major focus throughout the product’s clinical development program.
The initial clinical study of romiplostim showed that the product produced dose-related increases in platelet counts in healthy subjects, and clinical development proceeded in patients with chronic ITP. Because of the relative rarity of chronic ITP, the FDA designated romiplostim as an orphan drug (a product that treats a disease affecting fewer than 200,000 Americans).
Subsequently, FDA reviewed the protocols for the two major romiplostim clinical studies under the Special Protocol Assessment process. By this process, if the protocol’s description of the study design is found to be appropriate, FDA provides an explicit agreement that the design and planned analyses of the study adequately address the objectives in support of a regulatory submission. FDA agreed that the primary endpoint for the two controlled romiplostim studies—durable platelet response—was an established surrogate endpoint for clinical benefit in chronic ITP. This determination was based on the well-accepted role of thrombocytopenia as the pathophysiologic basis for hemorrhagic complications in chronic ITP. Additionally, the endpoint was defined to establish that romiplostim therapy resulted in both the achievement as well as the maintenance of a prespecified platelet count goal.
On October 23, 2007, the manufacturer submitted a Biologics License Application for romiplostim seeking market approval. Following a detailed review of the clinical data and an FDA Advisory Committee recommendation, the FDA approved romiplostim with a Risk Evaluation and Mitigation Strategy (REMS). A REMS is an FDA-required risk management plan that utilizes tools beyond the approved package insert to mitigate serious risk(s). The Food and Drug Administration Amendments Act of 2007 provides the FDA with the authority to require a REMS if FDA determines that additional postmarketing strategies (beyond routine) are necessary to ensure that the benefits of a drug outweigh the risks. Serious risks identified during the romiplostim clinical development program (worsened thrombocytopenia after drug discontinuation), as well as signals of potentially serious risks (hematologic malignancy and progression of disease in patients with myelodysplastic syndromes [MDS], bone marrow reticulin formation leading to fibrosis, as well as thrombotic/thromboembolic complications), prompted the need to require a REMS.
Phase III Clinical Studies
The approval of romiplostim was based predominantly on findings of its efficacy and safety in two double-blind, placebo-controlled clinical studies.
The two phase III investigations shared many design features. Both trials were designed as multicenter, double-blind, placebo-controlled studies in which patients were randomly assigned (in a 2:1, active:placebo ratio) and received treatment for 24 weeks, followed by 12 weeks when the study drug was not administered. Patients were evaluated at the end of 36 weeks of study. Eligible patients were adults with a diagnosis of ITP according to American Society of Hematology guidelines. The patients also had to have completed at least one previous treatment for ITP and have a mean of three platelet counts during screening and pretreatment that was ≤ 30 × 109/L, with no individual count > 35 × 109/L.
One study enrolled patients who had undergone splenectomy, while the other study enrolled nonsplenectomized patients. Patients with splenectomy were required to have had the splenectomy procedure performed at least 4 weeks before study entry. Exclusion criteria also helped select for patients who were receiving only stable doses of corticosteroids, azathioprine(Drug information on azathioprine), or danazol(Drug information on danazol). Patients receiving IVIG or anti-D immunoglobulin were excluded if they received the products less than 2 weeks prior to screening. Also excluded were patients treated with rituximab(Drug information on rituximab) (Rituxan) within 14 weeks prior to screening, as well as those using various other platelet-stimulating agents prior to screening.
In both studies, romiplostim was administered subcutaneously once per week at a starting dose of 1 μg/kg. Dose adjustment was allowed throughout the 24-week treatment period to allow subjects to maintain platelet counts in the target range of 50 to 200 × 109/L. The maximum permitted dose was 15 μg/kg. Romiplostim was administered by health-care providers. After 24 weeks of treatment, romiplostim was withdrawn and the platelet count was monitored. Participation was completed once platelet counts were ≤ 50 ×109/L or the subject reached week 36 with a platelet count > 50 ×109/L, whichever occurred first.
The primary endpoint in both studies was a comparison of the rates of “durable platelet response,” defined as at least six weekly platelet counts ≥ 50 × 109/L during the last 8 weeks of study drug treatment, in the absence of “rescue medications” at any time during the 24-week treatment period. The major secondary endpoints involved comparisons of platelet count “responses” (defined as any weekly platelet count ≥ 50 × 109/L) and the use of thrombocytopenia “rescue medications.”
Patients were assessed weekly for major study outcomes (including platelet counts) during the 24-week treatment period. Patients were also monitored for 12 weeks after the study medication had been discontinued. Patients in both arms were eligible to receive rescue medication throughout the study. Rescue medication was permitted for bleeding or wet purpura, or if the patient was at immediate risk of bleeding. Concurrent ITP therapies could be reduced during the first 12 weeks of treatment once platelet counts were > 100 × 109/L prior to dosing. Predose sampling for pharmacokinetic studies was performed once the patient reached a dose ≥ 10 μg/kg/wk.
Overall, 125 patients underwent randomization in the two phase III clinical studies; 41 were in the placebo group and 84 in the romiplostim group. Major baseline characteristics are summarized in Table 1. In general, baseline characteristics were relatively balanced between the groups within each study. All patients had failed at least one prior ITP therapy, and most (77%) had received three or more different types of prior ITP therapies. In general, patients who had undergone splenectomy had slightly lower baseline platelet counts as well as a more extensive history of prior ITP medication use compared to patients who had not undergone splenectomy.
Disposition within the two studies is summarized in Table 2. Overall, a relatively small number of patients (8%) failed to complete the studies. The major efficacy results are described in Table 3.
Overall, the romiplostim group experienced increased platelet counts compared to the placebo group, with statistical success for all the platelet count endpoints. Additionally, fewer patients in the romiplostim group received “rescue medications” because of bleeding or marked decreases in platelet counts. In general, platelet count response rates were higher for patients who had not undergone splenectomy, compared to those who had undergone the procedure.
Major safety findings were similar between the two phase III studies. The pooled, summarized results are shown in Table 4. The incidence of patients who experienced any bleeding event was similar in the romiplostim and placebo groups. Serious hemorrhages were uncommon and too few to definitively detect differences between the study groups. The most common adverse event in the studies was headache, which occurred among 35% of patients receiving romiplostim and 32% of those receiving placebo. The majority of adverse events were of mild to moderate severity. As summarized below, most of the serious risks for romiplostim were detected in the supportive studies and/or integrated analyses of the overall safety database.