Safety and efficacy of self‐administered romiplostim in patients with immune thrombocytopenia: Results of an integrated database of five clinical trials

Abstract Romiplostim self‐administration by patients or caregivers may offer time/cost savings to healthcare professionals (HCPs) and convenience for patients who avoid weekly clinic visits. We performed an integrated analysis of five clinical trials to evaluate the efficacy and safety of romiplostim self‐administration. Data were analyzed from adults with immune thrombocytopenia (ITP) who received weekly romiplostim via self‐administration or from an HCP. Patients who achieved a stable romiplostim dose for ≥3 weeks (HCP group ≥5 weeks to provide an appropriate index date to enable comparisons with the self‐administration group) with platelet counts ≥50 × 109/L were eligible. In the self‐administration (n = 621) vs HCP (n = 133) groups, respectively, median age was 53 vs 58 years, median time since primary ITP diagnosis was 3.7 vs 2.5 years, and median baseline platelet count at ITP diagnosis was 19.0 vs 20.0 × 109/L. In the self‐administration and HCP‐dosed groups, median romiplostim treatment duration was 89 vs 52 weeks and median total number of doses was 81 vs 50, respectively. In the self‐administration and HCP groups, respectively: 95.0% and 100.0% of patients achieved ≥1 platelet response (defined as weekly platelet count ≥50 × 109/L without rescue medication in previous 4 weeks); the median percentage of weeks with a response was 94.5% and 95.9%; and rescue medication was used in 36.7% and 39.8% of patients. Self‐administration did not adversely affect safety; duration‐adjusted rates for all treatment‐emergent adverse events (TEAEs) and bleeding TEAEs were numerically lower with self‐administration. Romiplostim self‐administration appears effective and well tolerated in eligible patients with ITP.


| Study design
Data were integrated from five romiplostim trials in patients with ITP aged ≥18 years in which self-administration of romiplostim was permitted (Table 1). [12][13][14]17,18 All patients received weekly subcutaneous injections of romiplostim with dose adjustments based on platelet count. Eligibility for self-administration in all trials, at the investigatorsʼ discretion, required platelet counts ≥50 × 10 9 /L to be achieved without romiplostim dose adjustment for ≥3 or ≥4 consecutive weeks (Table 1). Patients could continue with self-administration provided their romiplostim dose remained stable, as assessed by platelet counts every 4 weeks. Where dose adjustment was needed, patients were required to return to weekly clinic visits until the dose stabilized. All studies were conducted with institutional review board or independent ethics committee approval and all patients provided written informed consent.

| Study sample
The analysis set comprised the self-administration group and the HCP comparator group. The self-administration group included any patients who self-injected romiplostim in any of the studies included in the integrated analysis. For these patients, the index date (day 1) was the first day a subject self-administered romiplostim across all studies. The HCP group included patients who had received a stable dose of romiplostim for ≥5 consecutive weeks and who had achieved a platelet count ≥50 × 10 9 /L. Patients in this group never self-injected and continued to attend the clinic for administration of romiplostim.
The index date for these patients was the first instance of a fifth consecutive stable dose across all studies. For the HCP group, the requirement for ≥5 consecutive stable doses was specifically chosen to provide an appropriate index date to enable comparisons with the self-administration group. Although patients in the HCP group did not self-administer, for those who had stable doses for 4 weeks (stipulated as the minimum for self-administration in three of the five studies) the fifth dose would represent the equivalent time point at which self-administration could have started.

| Objectives
The primary objectives of the integrated analysis were to examine the efficacy and safety of romiplostim self-administration.
The efficacy analysis evaluated platelet response to romiplostim, defined as a weekly platelet count ≥50 × 10 9 /L without the use of any rescue medication within the 4 weeks preceding platelet measurement. This was assessed at 4-week intervals and excluded platelet counts measured following on-study splenectomy. Rescue medication was defined as any medication or transfusion used for the purposes of increasing platelet counts. Analysis of platelet response included evaluation of the incidence of platelet response, percentage of time with a platelet response, and percentage of patients with platelet counts <50 and ≥50 × 10 9 /L. Also included in the efficacy analysis was the incidence of rescue medication use. In addition, the total number of romiplostim doses was assessed. Finally, the number of dose adjustments (defined as observing different stable doses following ≥1 dose change in between) within 8 weeks and 6 months of the index date, as well as the duration-adjusted rate of dose adjustment (number of dose adjustments per 100 subject-weeks) were evaluated.
Dosing algorithms used for each study are described in Table S1.
Safety data were collected throughout the entirety of each study and comprised the nature, frequency, severity, seriousness and relationship to treatment of treatment-emergent adverse events (TEAEs). The TEAEs that led to withdrawal from romiplostim treatment or study withdrawal were recorded. Events of special interest included hypersensitivity, cardiac disorders, drug-related hepatic disorders, thrombotic/thromboembolic events, hemorrhages, immunogenicity, malignancies, myelofibrosis, pulmonary disorders, renal disorders, thrombocytosis, cytopenias, anemia, and leukocytosis. Also, TEAEs were coded according to Medical Dictionary for Regulatory Activities (MedDRA) version 20.1 by Standardized MedDRA Query (SMQ) and their preferred term. Bleeding events were identified using a pre-defined list of preferred terms from the "Hemorrhages" (narrow search) SMQ.

| Data management and analytical methods
Data were analyzed using descriptive statistics with no hypothesis testing. Continuous variables were described by number (n), mean, SD, median, interquartile range (IQR; Q1-Q3), minimum and maximum. Numbers and percentages were used for categorical variables.
Safety parameters were adjusted for differences in romiplostim treatment duration in the self-administration and HCP groups and presented as duration-adjusted event rates (calculated as: [total number of events/total duration in patient-years] × 100).

| Patients
The analysis set comprised 754 patients, of which 621 patients were in the self-administration group and 133 were in the HCP group. Baseline demographics and patient characteristics are shown in Table 2. While statistical testing was not performed, there was a numerical trend for patients in the self-administration group to be younger than those in the HCP group (median age of 53 vs 58 years, respectively) and with a longer median time since primary ITP diagno- Overall, the median number of visits for assessment of platelet counts was 47 vs 53 in the self-administration group and HCP group, respectively.

| Efficacy
During weeks −4 to −1, 86.9% (95% exact binomial confidence inter- Rescue medication was most commonly given during the first 6 months of the study period ( Figure 1B)

| DISCUSSION
This post-hoc analysis, integrating data from five clinical trials of romiplostim, suggests that romiplostim self-administration has comparable efficacy and safety to that of romiplostim administered by an HCP.
Most patients responded well to romiplostim treatment and attained their target platelet counts in the 4 weeks preceding selfadministration, with response rates generally exceeding 90%. These high platelet response rates were maintained after the commencement of self-administration and were similar to those in the HCP group. These response rates are comparable to those observed in a study of patients undergoing long-term treatment with romiplostim for ITP, which reported platelet responses in 87% of patients (N = 142), two-thirds of whom self-administered their medication. 11 Additionally, the use of rescue medication was similar in both groups of self-administration. 16 Overall, these data suggest that dose adjustments are not negatively impacted by self-administration of romiplostim. The numerically higher rate of dose adjustments in the HCP arm may also be reflective of the possibility that patients who tend to require dose adjustments would be less likely to be allowed to self-administer, thus receiving HCP administration only, making them eligible to be included in the HCP population. Additionally, patients in the HCP group attended numerically more visits for platelet assessments than self-administration patients (53 vs 47, respectively), giving more opportunity for dose adjustments in the former group.
Self-administration did not appear to affect the safety profile of romiplostim compared with patients in the HCP group. The most commonly observed TEAEs in the self-administration group -headache, nasopharyngitis, and arthralgia -are congruent with those reported in open-label extension studies of extended romiplostim treatment. 11,16 The rates of SAEs (34.6% in both self-administration and HCP-dosed patients) are also comparable to those seen in the long-term dosing Additionally, while statistical testing was not performed, there was a trend for treatment-related TEAEs leading to drug withdrawal, study discontinuation and fatal events to be numerically less frequent in the self-administration group compared with the HCP group when adjusted for treatment duration. These apparent reductions in TEAE incidences in the self-administration group may be due to several reasons. Firstly, due to the non-randomized allocation of patients to the self-administration and HCP dosing groups, clinicians may have offered self-administration to "fitter" patients; indeed, median age was lower in the self-administration group. This may also explain the lower percentage of patients requiring dose adjustments in the self-administration group. Additionally, an important factor to consider is the reduced number of clinic visits attended by the selfadministration group, at which monitoring could be carried out and where TEAEs would be reported. For example, patients in the selfadministration group attended numerically fewer visits for platelet counts than the HCP group. Accuracy could have been affected by patientsʼ recall of events that may have occurred several weeks before. Although this may be a contributing factor to the lower rates of TEAEs and dose adjustments observed in the self-administration group, efficacy of self-administration and overall safety were not largely affected and still support the use of romiplostim self-administration. Overall, the TEAE profile of romiplostim in self-administration patients from this study was similar to that from the earlier integrated analysis of self-administration in three clinical trials, 9 and no new safety concerns were identified when compared with published product information for romiplostim. 12,13 Patients with ITP are known to visit their physician more often and to have poorer work and productivity scores than healthy age-and sexmatched controls. 3  with self-administration kits has been evaluated as part of a multicenter observational study. 19 Among 40 patients/caregivers enrolled across 12 study centers, 35 (87.5%) administered romiplostim correctly following use of HAT materials and no injection errors were recorded. 19 This supports the use of HAT resources as a risk minimization tool. 19 In clinical studies, 11,12,[14][15][16] between 3%-24% of patients initiating self-administration of romiplostim changed to HCP-administration. In the majority of these cases, this was due to an administrative/investigator decision. Other reasons included patient request and in a small minority of cases (~1% of patients who started self-administration) due to non-adherence. 11 Amgen.

DATA SHARING STATEMENT
There is a plan to share data. This may include de-identified individual patient data for variables necessary to address the specific research question in an approved data-sharing request; also related data dictionaries, study protocol, statistical analysis plan, informed consent form, and/or clinical study report. Data sharing requests relating to data in this manuscript will be considered after the publication date and (a) this product and indication (or other new use) have been granted marketing authorization in both the US and Europe, or (b) clinical development discontinues and the data will not be submitted to regulatory authorities. There is no end date for eligibility to submit a data sharing request for these data. Qualified researchers may submit a request containing the research objectives, the Amgen product(s) and Amgen study/studies in scope, endpoints/outcomes of interest, statistical analysis plan, data requirements, publication plan, and qualifications of the researcher(s). In general, Amgen does not grant external requests for individual patient data for the purpose of re-evaluating safety and efficacy issues already addressed in the product labeling. A committee of internal advisors reviews requests. If not approved, requests may be further arbitrated by a Data Sharing Independent Review Panel.
Requests that pose a potential conflict of interest or an actual or potential competitive risk may be declined at Amgenʼs sole discretion and without further arbitration. Upon approval, information necessary to address the research question will be provided under the terms of a data sharing agreement. This may include anonymized individual patient data and/or available supporting documents, containing fragments of analysis code where provided in analysis specifications. Further details are available at the following: https://wwwext.amgen.com/ science/clinical-trials/clinical-data-transparency-practices/.