Status - BLA approved in Aug. 2008; EU approval in Feb. 2009

Organizations involved:

Amgen Inc. - Manuf.; R&D; Tech.; World mark.

Description: Nplate is a lyophilized (freeze-dried) formulation of recombinant romiplostim expressed by Escherichia coli (E. coli) bacteria. Romiplostim is a dimer (fusion protein or peptibody) of a human immunoglobulin IgG1 Fc domain linked to a peptide chain. Each peptide chain within the dimer contains two thrombopoietin receptor-binding domains (TRBD). Romiplostim increases platelet production through activation of the thrombopoietin (TPO) receptor c-Mpl, mimicking the action of endogenous thrombopoietin (eTPO).

Romiplostim is a thrombopoietin mimetic peptibody, a dimeric (2-chain) fusion protein with the N-terminus of an IgG1 antibody constant region (framework, not immunogen-binding, region) of human immunoglogulin (antibody; single, not double chain) bound to and used as a carrier for a 41-amino acid peptide sequence with affinity for binding to Mpl, the receptor for thrombopoietin (TPO) on the surface of blood platelets. The peptibody molecule contains two identical 269 amino acid single-chain subunits, each consisting of a human immunoglobulin IgG1 Fc domain covalently linked at the C-terminus to a peptide containing two thrombopoetin receptor binding domains, of which each binding domain consists of a 14-amino acid sequence. This type of recombinant protein is referred as a "peptibody" since it contains an active peptide region and an antibody Fc region.

Romiplostim, a member of the TPO mimetic class, is an Fc-peptide fusion protein (peptibody) that activate intracellular transcriptional pathways leading to increased platelet production via the TPO receptor (also known as cMpl). The peptibody molecule contains two identical single-chain subunits, each consisting of human immunoglobulin IgG1 Fc domain, covalently linked at the C-terminus to a peptide containing two thrombopoietin receptor-binding domains. Romiplostim has no amino acid sequence homology to endogenous TPO. Thus, the molecule is essentially two identical Fc portion, each consisting of human immunoglobulin IgG1 Fc domain, with heavy and kappa light chain constant regions with internal disulfide bonds, interlinked covalently covalently linked at the C-terminus to a peptide containing two thrombopoietin receptor-binding domains. The active peptide portion contains four repeats of an Mpl-binding peptide, i.e, has four TPO receptor-binding sites. Romiplostim has no amino acid sequence homology to endogenous TPO. Romiplostim has a molecular weight of 60 kDa, a caculated molecular formulat of C2634H4086N72O790S18 has disulfide bridges at 7-7', 10-10', 42-102, 42'-102', 148-206, and 148'-206'.

Attaching the Fc immunoglobulin portion to the active peptide increases the half-life of the peptide. The Fc region binds to FcRn salvage receptors, facilitating endothelial recirculation of the antibody-like molecule.

Single-use vials contain 250 or 500 µg of deliverable romiplostim as a sterile, lyophilized, solid white powder for subcutaneous injection and chronic use. When reconstituted with the appropriate volume of sterile water for injections (WFI), romiplostim is at a concentration of 0.5 mg/mL. Nplate has a shelf-life of 36 months at 2-8¼C (refrigerated)

Two vial presentations are available, each containing a sufficient amount of active ingredient to provide either 250 µg or 500 µg of deliverable romiplostim, respectively. Prior to administration, Nplate is reconstituted in the vial to 0.5 mg/mL (0.72 mL or 1.2 ml of sterile WFI is added to vials containing 250 µg or 500 µg of romiplostim respectively), and drawn into a syringe for subcutaneous injection. Each single-use 250 µg vial of Nplate contains the following: 375 µg romiplostim, 30 mg mannitol, 15 mg sucrose, 1.2 mg L-histidine, 0.03 mg polysorbate 20 (Tween 20), and sufficient hydrochloric acid (HCl) to adjust the pH to a target of 5.0. Each single-use 500 mcg vial of Nplate contains 625 µg romiplostim, 50 mg mannitol, 25 mg sucrose, 1.9 mg L-histidine, 0.05 mg polysorbate 20, and sufficient HCl to adjust the pH to a target of 5.0. The excipients used are commonly used in lyophilized products, and the concentrations employed are in the ranges commonly used.

Nomenclature: Thrombopoietin peptibody, rDNA [BIO]; Nplate [TR]; romiplostim [USAN; INN]; methionyl, 227 amino acid C-terminal Immunoglobulin G1 (human Fc fragment) fusion protein with 41 amino acid peptide, dimer [CAS]; L-methionyl[human immunoglobulin heavy constant gamma 1-(227-C-terminal residues)-peptide (Fc fragment) fusion protein with 41 amino acids peptide, (7-7'; 10,10')-bisulfide dimer [CAS]; 267639-76-9 [CAS RN]; thrombopoietin mimetic peptibody, recombinant [SY]; AMG-531 [SY]; Amgen Megakaryopoiesis Protein 531 [SY]

Biological: Romiplostim's mode of action is via the thrombopoietin (TPO) receptor and involves the same signaling pathways as native TPO. However, romiplostim has no homology with native TPO, which reduces the possibility of anti-TPO antibodies being generated. By stimulating platelet production, romiplostim represents a potentially new approach to the management of chronic ITP. Other TPO receptor-interacting therapeutics for ITP are in development, but these are primarily small molecule drugs.

Romiplostim acts as a ligand and binds and activates TPO cellular receptors without inducing cross-reacting, neutralizing antibodies. The peptide component mimics thrombopoietin (TPO), binding to and stimulating the TPO receptor, which is necessary for growth and maturation of bone marrow cells, megakaryocytes, that produce platelets. The active binding peptide fragment shares no sequence homology with endogenous TPO, preventing the production of neutralizing antibodies

Affinity of romiplostim for c-Mpl (thrombopoietin receptor) has been assessed in different species and in comparison with different c-Mpl ligands. In studies, direct assessment methods using Biacore 2000 and Biacore 3000 analysis were use. The KD of romiplostim to human c-Mpl was in a range between 0.51-14 nM, while affinity of thrombopoietin for the same receptor was calculated to be 33 nM. The KD of romiplostim to murine c-Mpl was calculated to be 3.5 nM, while affinity of thrombopoietin for the same receptor was calculated to be 66 nM.

The peptide sequence used in romiplostim was selected by screening of peptide libraries for sequences that bind to Mpl (TPO receptor) but lack homology to native human TPO (so they do not induce antibodies, as does recombinant TPO.

Peptibodies are engineered therapeutic molecules that can bind to human drug targets and contain peptides linked to the constant domains of antibodies. The ligand thrombopoeitin (TPO) and its interaction with its receptor (c-mpl) are important in platelet production. Romiplostim works similarly to TPO, a natural protein in the body, binding to the TPO receptor, which activates the pathway necessary for growth and maturation of bone marrow megakaryocyte cells, resulting in increased platelet production. Romiplostim appears to directly stimulate platelet production and may offer physicians a way to shift the treatment focus from preventing platelet destruction to boosting platelet production in patients with ITP.

Earlier, first-generation TPO-acting therapeutics were unsuccessful primarily due to adverse effects. Recombinant human TPO and pegylated recombinant megakaryocyte growth and development factor (PEG-rhMGDF) were studied by Amgen in clinical trials and elevated platelet counts in some recipients. However, PEG-rhMGDF was immunogenic and TPO induced neutralizing antibodies, leading to worsening of thrombocytopenia.

Companies: Nplate was developed and is manufactured by Amgen, with Amgen holding worldwide marketing rights.

Manufacture: The romiplostim gene was constructed by combining (fusing) the sequence of the Fc portion of the human IgG1 heavy chain from the first amino acid of the hinge domain to the carboxyl terminus with the sequence encoding a peptide chain containing two thrombopoietin receptor-binding domains (TRBD.

Romiplostim fusion protein is expressed intracellularly in E. coli as inclusion bodies using high cell density fed batch fermentation. The fermentation process begins with thawing and expansion of the working cell bank vial in a seed flask. This is followed by further cell expansion and induction of expression in a production fermentor. After completion of production fermentation, the broth is chilled and cell processing is initiated. The cells are harvested and lysed to release the inclusion bodies. These are recovered by centrifugation, washed and recovered as a slurry which can be stored frozen prior to further purification. Purification begins with thawing the slurry which is then solubilized. Romiplostim is purified using a series of chromatography, concentration and diafiltration steps to produce purified bulk active substance. The final medicinal product manufacturing process consists of thawing the active substance, a formulation step, followed by aseptic filtration, filling of the vials, partial stoppering, lyophilization and capping. Formulation consists of a dilution of the active substance in a formulation buffer.

The romiplostim active substance is routinely controlled by a range of chemical-physical and biological tests (covering appearance/color, pH, identity, purity, potency, quantity and microbial content) to assure consistent production. Routine tests and specifications include a range of chemical-physical and biological tests, covering appearance/color and moisture content for the lyophilized product and appearance/color, identity, purity, sterility, potency, quantity, pH, polysorbate 20, sub-visible particles, and osmolality for the reconstituted medicinal product. The control of the medicinal product relies to a large extent on the same analytical methods as those used to control the active substance.

 During development an improved second generation active substance manufacturing process was developed to optimize commercial production without changing the manufacturing scale. The fermentation process utilized the same basal media and overall production scheme. The purification process used the same sequence of chromatography steps and process chemistry. An unnecessary ultrafiltration step was removed from the process. These changes in manufacturing process wre considered minor by EMEA (EU), and were made to improve manufacturability and maintain product quality. Extensive comparability studies have shown a good consistency of production through the different sites and scales. No significant differences among active substance lots across processes and sites were found by EMEA/EU.

Nplate's development history includes both frozen liquid and lyophilised formulations. For the early clinical trials, a frozen liquid formulation was used. Based on clinical dosing requirements and because of the instability trends in the frozen liquid formulation at low protein concentrations a lyophilized formulation was developed. L-histidine was used to maintain pH control at a target of 5.0. Dilute hydrochloric acid was added to adjust the pH to the acceptable range. Sucrose was incorporated as a stabilizer. Mannitol was included as a bulking agent. Several formulations were tested. Polysorbate 20 was chosen as a stabilising agent to minimise aggregation at the required protein concentration and pH. Several experiments were performed showing that lyophilization did not alter the bioactivity or structural properties of the protein.

Some starting materials and reagents used in the production process of romiplostim involve material of animal origin have been identified. EMEA/EU concluded that the associated TSE risk of romiplostim is negligible.

FDA class: Biologic BLA

Approvals: Date = 20080822; BLA

Indications: [Full text of the "INDICATIONS AND USAGE" section of product insert/labeling]:

Nplate is indicated 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. Nplate should be used only in patients with ITP whose degree of thrombocytopenia and clinical condition increases the risk for bleeding. Nplate should not be used in an attempt to normalize platelet counts.

Status: In 2004, the FDA granted fast track designation for Nplate. Orphan designation for ITP was granted in 2003 by the FDA and in 2005 by EMEA/European Union.

In Oct. 2007, Amgen filed a BLA seeking approval for treatment of thrombocytopenia in adults with chronic ITP who have not undergone splenectomy (removal of the spleen) and have had an inadequate response or are intolerant to corticosteroids and/or immunoglobulins; or patients who have had their spleen removed and have an inadequate response to the procedure. The application was granted priority review, with a PDUFA (target approval) date of April 23, 2008. In April, Amgen announced the PDUFA date had been extended to July 23, 2008/

In Nov. 2007, Amgen filed a MAA seeking Nplate approval for for treatment of ITP with orphan designation.

In March 2008, the Oncology Drugs Advisory Committee, FDA, considered Nplate. Although expressing concerns about adverse effects, such as an increase in bone marrow fibers known as reticulin, an increase in blood clots and the development of resistance to the treatment, the panel voted 10-0 that the benefits outweighed the risks after Amgen proposed a risk-management programme and postmarketing studies to monitor adverse events. There was also concern about whether Nplate could cause or contribute to other blood disorders, such as leukemia, due to its mechanism of action.

In late July 2008, FDA reported that it would miss the PDUFA (targeted approval decision) date for the BLA. With FDA not having issued an "approvable" (now called "complete response") letter or other communication indicating a problem with the application, this was interpreted as a positive sign. Nplate's approval included a mandatory Risk Evaluation & Mitigation Strategy, or REMS, a new feature of the regulatory process created by legislation enacted in 2007. FDA was setting precedents with every REMS, and the agency had already extended or missed several deadlines for other products covered by that authority.

On Aug. 22, 2008, FDA approved Nplate for treatment of chronic ITP in adults patients not responding to other treatments. This approval was roughly about 1 months after the PDUFA target date in July. FDA required a Risk Evaluation and Mitigation Strategy (REMS) be implemented to address the risks of Nplate therapy. Under the FDA Amendments Act of 2007, FDA determined that a REMS was necessary for the benefits of Nplate to outweigh the risks of the product. The REMS included a Medication Guide for patients and required that all prescribers and patients enroll in a special program to track the long term safety of Nplate therapy.

Upon FDA approval, Recothrom had the most extensive Risk Evaluation & Mitigation Strategy yet required by FDA under its new post-marketing authorities, including registration of institutions, prescribers and patients. At 94 pages, this shows the accelerating complexity of the new FDA Amendments Act of 2007 rules for sponsors.

Regulatory filings in the European Union (EU), Canada and Australia were also filed in 2007.

On Feb. 9, 2000 Nplate received approval in the EU with orphan status for the treatment of splenectomised adult chronic immune (idiopathic) thrombocytopenic purpura (ITP) patients who are refractory to other treatments (e.g. corticosteroids, immunoglobulins). Nplate may be considered as second line treatment for adult non-splenectomised ITP patients where surgery is contraindicated. Nplate became the first and only approved platelet producing agent in Europe.

Tech. transfer: The U.S. product insert cites 6,835,809 and 7,189,827.

U.S. 7,189,827; 7,186,810; 7,169,905; and 7,166,707 are each titled, " Modified peptides as therapeutic agents." These patents concern peptibody fusion proteins, i.e.., active binding peptides appended the Fc portion of immunoglobulin.

U.S. 6,835,809, "Thrombopoietic compounds," assigned to Amgen, concerns peptides and polypeptides of useful to increase platelets or platelet precursors (e.g., megakaryocytes) in mammals.

Romiplostim may meet the definition of an immunoadhesin, and be covered by related patents held by Genentech. Immunoadhesins are recombinant fusion proteins, expressed from the joining of two genes, combining certain structural features of an antibody with a high-affinity cell surface receptor. Essentially, the antibody-binding (variable) region of a monoclonal antibody is replaced with a high affinity receptor. Genentech's immunoadhesin patent portfolio includes U.S. patents 5,116,964; 5,428,130; 5,455,165; and 5,514,382. These patents include claims covering aspects of antibody-like fusion proteins. Presuming these patents apply, Amgen either has taken a license for Genentech or will likely be prosecuted by Genentech for infringement.

Disease: Idiopathic thrombocytopenic purpura (ITP) is an autoimmune blood disorder in which there is platelet destruction mediated by autoantibodies, i.e., platelets are destroyed by the patient's own immune system. ITP has historically been considered a disease of platelet destruction. However, recent studies suggest that the body's natural platelet production processes are unable to compensate for low levels of platelets in the blood. Increasing the rate of platelet production may address low platelet levels associated with ITP.

Current therapy for ITP is unsatisfactory and relies on generalized immunosuppression using a variety of drugs; other therapies include human blood products such as immunoglobulin or splenectomy. Current medical treatment includes corticosteroids and immunoglobulin. Surgery to remove the spleen, a procedure known as a splenectomy, may help some patients. Nplate is approved only for patients with chronic ITP who do not respond sufficiently to current treatments.

There are an estimated 140,000 people with chronic ITP in the U.S., as reported by FDA, while other sources report much less, e.g., 60,000. ITP affects about twice as many adult women as men.

There are an estimated 50,000 adult patients with chronic ITP in the EU.

Trials: BLA approval was based on largely on two randomized Phase III trials, each conducted under an SPA, in a total of about 125 patients who had received at least one prior ITP treatment; and an extension study that contained patients from the Phase III studies. One study enrolled patients who still had their spleen, the other enrolled patients who did not. During six months of treatment, patients who received Nplate had significantly higher platelet counts and maintained those higher counts, compared to those who did not receive the drug. The response to Nplate was higher in those patients who still had their spleen than in those patients who had undergone a splenectomy. In those patients who did not receive Nplate, only one experienced a sustained increase in platelet counts.

Safety concerns observed with Nplate include fibrous deposits in the bone marrow and the possibility that once Nplate is stopped, platelet counts could drop below what they were before beginning treatment. Additional risks include blood clots due to excessive increases in platelets and, if Nplate were given to patients with an abnormal blood condition known as myelodysplasia, a risk for a form of blood cancer known as acute leukemia. Myelodysplasia, which is associated with low platelet counts, predisposes some patients to leukemia. In a study of 44 patients who had myelodysplasia and received Nplate, four patients developed leukemia. Further clinical trials in patients with predisposing conditions for leukemia will be needed to determine whether the development of leukemia may relate to the use of Nplate.

In the Feb. 2, 2008 issue of The Lancet [371(9610):395-403], pooled results were published from two Phase III studies which evaluated the administration of Nplate on increasing and sustaining platelet counts in both splenectomized (spleen removed) and non-splenectomized patients with chronic Immune Thrombocytopenic Purpura (ITP). A durable platelet response was achieved by 16 of 42 splenectomised patients given romplostim vs. none of 21 given placebo (difference in proportion of patients responding 38% [95% CI 23.4-52.8], p=0.0013), and by 25 of 41 non-splenectomised patients given romplostim versus one of 21 given placebo (56% [38.7-73.7], p<0.0001). The overall platelet response rate (either durable or transient platelet response) was noted in 88% (36/41) of non-splenectomised and 79% (33/42) of splenectomised patients given romiplostim compared with 14% (three of 21) of non-splenectomised and no splenectomised patients given placebo (p<0.0001). Patients given romiplostim achieved platelet counts of 50x10(9)/L or more on a mean of 13.8 (SE 0.9) weeks (mean 12.3 [1.2] weeks in splenectomised group vs 15.2 [1.2] weeks in non-splenectomised group) compared with 0.8 (0.4) weeks for those given placebo (0.2 [0.1] weeks vs 1.3 [0.8] weeks). Adverse events were much the same in patients given romiplostim and placebo. No antibodies against romiplostim or thrombopoietin were detected.

The EU approval of Nplate was based on data from two separate placebo-controlled Phase I studies, demonstrating that platelet counts were raised and sustained in 83% of patients for both splenectomised and non-splenectomised groups when treated with Nplate. Patients treated with Nplate were able to reduce or discontinue concomitant medications, such as corticosteroids, which are often not well tolerated. ,Nplate patients also used far less "emergency" medications such as IVIG and Win-Rho, whose effects are transient. Upon completion of the Phase III studies, almost 90% of patients enroll into the Nplate long term extension study which showed that Nplate continued to effectively increase and sustain platelet counts. In this open label long term extension study, the average treatment period was 76 weeks and the longest duration of treatment was 204 weeks. This study showed platelet counts of Nplate-treated patients were increased from baseline by 20,000 platelets per microliter more than 80% of the time in 47% of patients and more than half the time in 67% of patients..

Medical: Only prescribers enrolled in the Nplate NEXUS (Network of Experts Understanding and Supporting Nplate and Patients) Program may prescribe Nplate. Nplate must be administered by the enrolled prescribers or healthcare providers under their direction.

Romiplostim may be useful in a wide range of thrombocytopenic states, such as chemotherapy-induced thrombocytopenia, myelodysplasia and hepatitis C-related thrombocytopenia. It may also be useful in stimulating platelet yield from normal donors. Studies involving the use of romiplostim in these settings are under way or planned.

The initial dose for Nplate is 1 mcg/kg based on actual body weight. Physicians should use the lowest dose to achieve and maintain a platelet count ³ 50 x 109/L as necessary to reduce the risk for bleeding. Nplate is administered as a weekly subcutaneous injection with dose adjustments based upon the platelet count response. Nplate should not be used in an attempt to normalize platelet counts. The prescribed Nplate dose may consist of a very small volume (eg, 0.15 mL), and only a syringe that contains 0.01 mL graduations should be used.

Physicians should use the actual body weight at initiation of therapy, then adjust the weekly dose of Nplate by increments of 1 µg/kg until the patient achieves a platelet count ³ 50 x 109/L as necessary to reduce the risk for bleeding; but not exceeding a maximum weekly dose of 10 mcg/kg. In clinical studies, most patients who responded to Nplate achieved and maintained platelet counts ³ 50 x 109/L with a median dose of 2 µg/kg..

Market.: Various analysts project peak sales of $300-$500 million/year.

Reportedly costing roughly $4 per microgram, an average 150-lb. patient at maximum dose (10 µg/kg/week) could potentially cost over $140,000 per year.