Romiplostim (RBE), powder for injection, 165 micrograms, 375 micrograms and 625 micrograms, Nplate®

Public summary document for Romiplostim (RBE), powder for injection, 165 micrograms, 375 micrograms and 625 micrograms, Nplate®

Page last updated: 13 November 2009

PDF printable version for Romiplostim (RBE), powder for injection, 165 micrograms, 375 micrograms and 625 micrograms, Nplate® (PDF 144 KB)

Public Summary Document

Product: Romiplostim (RBE), powder for injection, 165 micrograms, 375 micrograms and 625 micrograms, Nplate®
Sponsor: Amgen Australia Pty Ltd
Date of PBAC Consideration: July 2009

1. Purpose of Application

The submission sought a Section 100 (Highly Specialised Drugs Program) listing for initial and continuing treatment of adult patients with chronic immune (idiopathic) thrombocytopenic purpura (ITP) who meet certain criteria.
Highly Specialised Drugs are medicines for the treatment of chronic conditions, which, because of their clinical use or other special features, are restricted to supply to public and private hospitals having access to appropriate specialist facilities.

2. Background

This drug had not previously been considered by the PBAC.

3. Registration Status

Romiplostin was TGA registered on 8 August 2008 for the treatment of thrombocytopaenia in adult patients with chronic immune (idiopathic) thrombocytopaenic purpura (ITP)

who are non-splenectomised and have had an inadequate response, or are intolerant, to both corticosteroids and immunoglobulins;

who are splenectomised and have had an inadequate response to splenectomy.

 

4. Listing Requested and PBAC’s View

Section 100 (Highly Specialised Drugs Program)
Private Hospital Authority Required
Initiation:
For the treatment of thrombocytopenia in adult patients with chronic immune (idiopathic) thrombocytopenic purpura (ITP) who are:

Not splenectomised and who have had an inadequate response or are intolerant to both corticosteroids and immunoglobulins.

Splenectomised and who have had an inadequate response to splenectomy.

Inadequate response is defined as a persistent platelet count of ≤ 30 x 109/L.

Continuation:
Patients should continue treatment if they display a sustained platelet response and have not required more than one occasion of IVIg use during the first episode of treatment. A sustained platelet response is defined as:

A weekly platelet count ≥ 50 x 109/L on at least four (4) occasions; or

A platelet count > 30 x 109/L AND a doubling of baseline platelet count, on at least four (4) occasions.


The PBAC considered there were a number of uncertainties about the requested restriction, even though the proposed continuation criteria based on attaining weekly platelet counts of > 50 x 109/L on four occasions or > 30 x 109/L and doubling of the baseline platelet count on 4 occasions were consistent with the secondary outcome used in the clinical trials. The PBAC indicated that the most appropriate patients to target would be patients with severe symptomatic ITP or patients with chronic severe ITP at high risk of bleeding despite all other therapies. The majority of these patients would have had a splenectomy, and have failed steroids and immunoglobulin, usually post splenectomy; or those who require high steroid doses (>25mg/d for months) to maintain responses post splenectomy.
 

5. Clinical Place for the Proposed Therapy

Chronic immune (idiopathic) thrombocytopenic purpura (ITP) is a long-term autoimmune disorder characterised by persistently low platelet counts (thrombocytopenia) and cutaneous and mucosal bleeding. Romiplostim stimulates platelet production for long-term treatment of adult chronic ITP patients.

6. Comparator

The submission nominated placebo as the comparator. The PBAC acknowledged that placebo would be the appropriate comparator provided that the requested restriction explicitly specified that the non-splenectomised patients are unsuitable for splenectomy. Otherwise, splenectomy should also be included as a comparator.

For further comment see Recommendations and Reasons

7. Clinical Trials

The submission presented two randomised trials comparing romiplostim with placebo in adult patients with chronic idiopathic thrombocytopenic purpura (ITP) – non-splenectomised and post-splenectomised populations respectively, over a 24-week period. The studies published at the time of submission are as follows:

Trial ID Protocol title/ Publication title Publication citation
Direct randomised trials
A Phase II clinical trial/Kuter et al (2004) Bussel et al (2006) A phase 2 placebo controlled study evaluating the platelet response and safety of weekly dosing with a novel thrombopoietic protein (AMG531) in thrombocytopenic adult patients with immune thrombocytopenic purpura (ITP). AMG 531, a thrombopoiesis-stimulating protein, for chronic ITP. Blood; 104:148-9. New England Journal of Medicine; 355:1672-1681
Trials 212 and 105 Kuter DJ, 2008 Gernsheimer TB, 2007 Lyons R, 2007 Efficacy of romiplostim in patients with chronic immune thrombocytopenic purpura: a double-blind randomised controlled trial. Evaluation of AMG 531 Efficacy in Nonsplenectomized Patients with Chronic Immune Thrombocytopenic Purpura (ITP) in a Randomized Placebo-Controlled Phase 3 Study. Evaluation of AMG 531 Safety in Splenectomized (S) and Nonsplenectomized (NS) Patients with Chronic Immune Thrombocytopenic Purpura (ITP) in Two Randomized Placebo-Controlled Phase 3 Studies. The Lancet; 371(9610):395-403 Blood 110:8a Blood 110:8a


8. Results of Trials

The following table summarises the results of the primary and key secondary outcomes across the direct randomised trials.
 

Trial 212-non Non-splenectomised Patients (24 weeks) Trial 105-post Splenectomised Patients (24 weeks)
Rom (n = 41) Placebo (n = 21) Rom (n = 42) Placebo (n = 21)
Primary Endpoint
No. (%) Patients with Durable Platelet Response [95%CI] a 25 (61%) [0.45, 0.76] 1 (5%) [0.00, 0.24] 16 (38%) [0.24, 0.54] 0 (0%) [0.00, 0.16]
RR [95% CI] 12.80 [1.86, 88.07] 16.88 [1.06, 268.42]
RD [95%CI] 0.56 [0.39, 0.74] 0.38 [0.22, 0.54]
NNT [95%CI] 2 [1, 3] 3 [2, 5]
p-value <0.0001 .0013
Key Secondary Endpoints
No. (%) Patients with Overall Platelet Response [95%CI] b 36 (88%) [0.74, 0.96] 3 (14%) [0.03, 0.36 33 (79%) [0.63, 0.90] 0 (0%) [0.00, 0.16]
RR [95% CI] 6.15 [2.14, 17.63] 34.28 [2.20, 533.41]
RD [95%CI] 0.74 [0.56, 0.92] 0.79 [0.65, 0.92]
NNT [95%CI] 1 [1, 2] 1 [1, 2]
p-value <0.0001 <0.0001

a Durable platelet response was defined as weekly platelet count ≥50 x 109/L for six or more times for study weeks 18-25 in the absence of rescue therapy any time during the treatment period.
b Overall platelet response includes durable responders as well as those responders achieving weekly platelet count ≥50 x 109/L for four or more times during study weeks 2-25. Patient may not be classified as having a weekly response within eight weeks after receiving rescue therapy.
RR = relative risk; RD = risk difference; NNT = number need to treat; CI = confidence interval.

The two key trials used platelet count, a surrogate outcome, rather than a health outcome such as bleeding or mortality as the measurement of efficacy. Platelet count had been accepted as a useful surrogate outcome in the assessment of treatment effects of ITP (George et al 1996). Both trials indicate that romiplostim significantly improved platelet levels compared with placebo.

The objective of treatment was to prevent symptomatic bleeding by increasing platelet count and/or improving function. Apriori analysis about the effect of romiplostim on symptomatic bleeding was not available from the trials, although the patients enrolled in the trials had very low platelet levels.

There was a higher incidence of treatment-related adverse events in the romiplostim arm than in the placebo arm in both trials. Eight severe treatment-related adverse events occurred in Trial 105-post in the romiplostim arm compared with none from the placebo arm. Among the eight severe treatment-related adverse events, two were considered serious adverse events, namely increased reticulin fibrosis of the marrow leading to study discontinuation, and right transfemoral popliteal embolectomy. (Serious adverse events were events that were fatal life-threatening, causing patient hospitalisation or prolongation of existing hospitalisation, led to persistent or significant disability/incapacity, led to congenital anomaly/birth defect, or other significant medical hazard)

The submission provided a romiplostim extension study examining the long-term safety and efficacy of romiplostim treatment in thrombocytopenic subjects with ITP (Bussel et al, Safety and Efficacy of long-term treatment with romiplostim in thrombocytopenic patients with chronic ITP. Blood 2008: 113: 2161-71). Nineteen serious adverse events considered related to treatment were reported in 13 (9.2%) patients. Of these 19 events, 2 cases of increased bone marrow reticulin and one each of vaginal haemorrhage, deep vein thrombosis, and monoclonal gammopathy led to discontinuation of the study.
 

9. Clinical Claim

The submission described romiplostim as being superior in terms of comparative effectiveness to placebo, but associated with a higher incidence of mild to moderate drug-related adverse events.

For PBAC's views see the Recommendations and Reasons.

10. Economic Analysis

A stepped economic evaluation was presented. The type of economic evaluation presented was a cost-utility analysis.

The outcome of the model was in terms of cost per quality-adjusted life year (QALY). The model was a micro-simulation, multi-state Markov model. The extrapolation of treatment effects beyond the trial period was the key driver of the model.

The incremental cost per additional QALY gained over patient lifetime (romiplostim versus placebo) for non-splenectomy was in the range of $45,000 – 75,000, while the incremental cost per additional QALY gained over patient lifetime (romiplostim versus placebo) for post-splenectomy was in the range $15,000 - 45,000.

For PBAC’s views see Recommendations and Reasons.

11. Estimated PBS Usage and Financial Implications

The estimated financial cost per year to the PBS was in the range of $30 – 60 million per year in Year 5 of listing.

12. Recommendation and Reasons

The PBAC considered there were a number of uncertainties about the requested restriction, even though the proposed continuation criteria based on attaining weekly platelet counts of

>

50 x 109/L on four occasions or

>

30 x 109/L and doubling of the baseline platelet count on 4 occasions were consistent with the secondary outcome used in the clinical trials.

The PBAC noted the statement in the Guideline published in the British Journal of Haematology which suggested that treatment should only be used when absolutely required, for example patients with severe symptomatic ITP. In general patients with platelet counts exceeding 30 x 109/L required no treatment unless they are undergoing a procedure likely to induce blood loss. The requested restriction defined an inadequate response as a persistent platelet count <30 x 109/L.

An inconsistency between the continuation rule which referred to a “sustained platelet response”, the trial outcome was “durable response” and the economic model which used “overall response” was noted.

Although somewhat reassured about the estimation of risk associated with low platelet counts (4 x mortality rate for platelet counts between 10 and 30 x 109/L) by the clinician who presented at the hearing, the PBAC was of the view that the clinical place of romiplostim had not been clearly established and further work would be required to define an appropriate, and more narrow, population for treatment with the drug to target patients most in need. In particular, the population, who should qualify for continuing treatment, needed to be more appropriately defined.

The PBAC indicated that the most appropriate patients to target would be patients with severe symptomatic ITP or patients with chronic severe ITP at high risk of bleeding despite all other therapies. The majority of these patients would have had a splenectomy, and have failed steroids and immunoglobulin, usually post splenectomy; or those who require high steroid doses (>25mg/d for months) to maintain responses post splenectomy. A suggested eligible platelet count that could be considered would be patients with a platelet count of less than 20 x 109/L or with bleeding below a count of 30 x 109/L. Another possible patient group would be the problematic group of uncommon patients who are refractory to steroids and immunoglobulin and have an absolute contraindication to splenectomy. Although defining contraindication to splenectomy will be difficult, this could be dealt with via a risk sharing agreement.

The PBAC acknowledged that placebo would be the appropriate comparator provided that the requested restriction explicitly specified that the non-splenectomised patients are unsuitable for splenectomy. Otherwise, splenectomy should also be included as a comparator.

The Pre-Sub-Committee Response reiterated the submissions argument that currently only 24% of ITP patients were ever splenectomised and that rates had been declining over time, the sponsor claimed that the “vast majority of patients” were treated with placebo and therefore it was the appropriate main comparator. However, the PBAC considered that the issue of the appropriate comparator remained contentious. Splenectomy is historically the most effective treatment for recurrent ITP severe enough to warrant intervention and so is indicated in a minority of patients. The majority of patients with chronic ITP only need watching and occasional intervention with steroids. The trial entry requirements which were designed to demonstrate sustained improvements in platelet counts therefore did not necessarily reflect clinical need. The Australian ITP audit included patients who were not eligible for the trial, so the 24% rate of splenectomy was not an informative figure.

Most chronic severe ITP patients will have had or been considered for splenectomy. The main reason for non-elderly patients not to have splenectomy in recent times was availability of alternatives such as rituximab or romiplostim. If romiplostim was to be approved with the requested restriction, then splenectomy is the most likely effective therapy to be replaced in practice. It was not clear whether patients with less severe disease should be exposed to an expensive new therapy, when long-term safety had yet to be elucidated.

The PBAC noted that the two key trials used platelet count, a surrogate outcome, rather than a health outcome such as bleeding or mortality as the measurement of efficacy. Platelet count had been accepted as a useful surrogate outcome in the assessment of treatment effects of ITP (George JN et al. Idiopathic thrombocytopenic purpura: a practice guideline developed by explicit methods for the American Society of Hematology. Blood 1996; 88: 3-40). Both trials indicated that romiplostim significantly improved platelet levels compared with placebo. However, there was uncertainty about the long-term efficacy of romiplostim, given the efficacy data from placebo-controlled trials only related to a 24-week period.

The objective of treatment was to prevent symptomatic bleeding by increasing platelet count and/or improving function. However, a priori analysis about the additional benefit of romiplostim on symptomatic bleeding was not available from the trials, although the patients enrolled in the trials had very low platelet levels. Hence the submission argument was based entirely on a transformation of the surrogate into a clinically relevant endpoint. Further, it was noted the selection of the target platelet count was 50 × 109/L to define a response to treatment, which was then used to build the economic analysis. In trial 212 for non-splenectomised patients, the mean platelet count in the placebo arm was 19 x 109/L with a range of 5-31 x 109/L. The trial suggested a very large treatment effect (61% responders in romiplostim arm vs. 5% in the placebo arm), but the PBAC considered that the selection of this threshold as a surrogate outcome with predictive value for the effect of the treatment on symptomatic bleeding, and hence extension of life and quality of life, was not sufficiently justified.

The PBAC noted the Pre-PBAC Response included data drawn from patients with thrombocytopenia of non-ITP origin, where bleeding risk was higher for any given platelet count, casting doubt on the applicability of these data. The outcome of most interest was significant bleeding (grade 3-5), but confidence intervals for the outcomes were not provided.

Thus the PBAC concluded that although low platelet count was a surrogate for bleeding, the magnitude of the reduction in risk of major bleeding with romiplostim was uncertain.

There was a higher incidence of treatment-related adverse events in the romiplostim arm than in the placebo arm in both trials. Eight severe treatment-related adverse events occurred in Trial 105-post in the romiplostim arm compared with none from the placebo arm. Among the eight severe treatment-related adverse events, two were considered serious adverse events, including increased reticulin fibrosis of the marrow leading to study discontinuation The PBAC noted the claim in the Pre-PBAC Response that reticulin was a diagnostic finding, characteristic of ITP, with no known clinical sequelae. However, the Committee considered that long-term safety of romiplostim remained to be determined.

The PBAC agreed with the issues of uncertainty about the economic model raised during evaluation. It was highlighted that essentially the model was driven by; the probability of bleeding, which was determined by platelet count < or > 50 × 109/L and the fact that romiplostim increased the probability of platelet count >50 × 109/L.

The extrapolation of treatment effects beyond the trial period was the key driver of the model, because the probabilities of bleeding-related events, and consequent IVIg use, as well as the probability of remission were all dependent on the platelet count. However, the assumption regarding the long-term treatment effect being continued from the trial period had not been substantiated.

The PBAC noted that the EQ-5D was administered during the trial and did not show a difference between romiplostim and placebo, but was not used in the modelled economic evaluation. Published utility data were used instead. The populations were not chronic ITP patients and therefore the utility weights measured in these patients may not be representative of the ITP patients in the requested listing. Thus there was uncertainty about the extent of quality of life improvement for patients treated with romiplostim. This was important because the model was sensitive to the application of a utility gain attributed to romiplostim as shown by an improvement in the ICER for both non-splenectomised patients and for splenectomised patients, when quality adjusted. The PBAC noted that if the target population were to be more highly restricted, there would be less concern about the acceptance of these utility values.

There was also uncertainty about the applicability of the trial results to Australian chronic ITP patients, given that the comparability of body weight distribution and the risk factors for bleeding between the Australian ITP patients and the trial population was unclear, and that the patients in the trials were only required to have completed at least one prior therapy, while the requested PBS population would have to fail both corticosteroids and immunoglobulins.

In addition to the uncertainties reported during evaluation, the PBAC also noted that the model assumes the costs of rescue medication, but not its benefits, which favours romiplostim. The PBAC also questioned the appropriateness of the length of the economic model, which would also favour romiplostim.

The PBAC therefore rejected the submission because of the uncertain place in treatment for romiplostim, uncertain clinical benefit and uncertain and unacceptable cost-effectiveness.

The PBAC noted that the submission meets the criteria for an independent review.

Recommendation
Reject

13. Context for Decision

The PBAC helps decide whether and, if so, how medicines should be subsidised in Australia. It considers submissions in this context. A PBAC decision not to recommend listing or not to recommend changing a listing does not represent a final PBAC view about the merits of the medicine. A company can resubmit to the PBAC or seek independent review of the PBAC decision.

14. Sponsor’s Comment

Amgen looks forwards to working with the PBAC to ensure a successful listing for the benefit of patients in this population

.