ROMIPLOSTIM FOR USE IN THE TREATMENT OF IDIOPATHIC THROMBOCYTOPENIC PURPURA (ITP)

Abstract

The present invention relates to a method of treating idiopathic thrombocytopenic purpura (ITP) in a patient with ITP, comprising: (a) administering romiplostim weekly to the patient; (b) increasing the weekly dose until a platelet count of at least approximately 50 to 200 x 10⁹ / L is achieved; (c) decreasing the weekly dose of romiplostim if the platelet count remains > 200 x 10⁹ / L for two consecutive weeks; (d) permanently discontinuing romiplostim if the platelet count remains > 200 x 10⁹ / L for two consecutive weeks when the weekly dose is 1 µg / kg or the platelet count is > 400 x 10⁹ / L; and (e) if a platelet count > 200 x 109 / L is achieved during the first 4 to 12 weeks of treatment, maintain a treatment-free period of at least approximately 24 weeks during which the patient does not receive romiplostim.

Description

Primary ITP is an autoimmune disorder characterized by suboptimal platelet production and accelerated platelet destruction, mediated by both antibodies and T lymphocytes (Nugent et al., 2009). ITP in adults often has a chronic course, presenting an increased risk of bruising and skin bleeding, reduced quality of life, and, rarely, severe bleeding (Cines and McMillan, 2005). The primary goal of treatment in these patients is a durable improvement in platelet count without lifelong therapy. Current treatments include corticosteroids, anti-D immunoglobulin, and intravenous immunoglobulin (IVIg) as first-line therapies, with romiplostim, eltrombopag, rituximab, and splenectomy as second-line options. These first-line treatments can produce platelet responses in most patients, but a response may only be observed for weeks or months (Provan et al., 2010). For second-line treatments such as rituximab, 15–20% of patients may have a complete response for up to 5 years. Splenectomy has a long-term response rate of approximately two-thirds (Provan et al., 2010). The benefits of these treatments must be weighed against their potential risks, as all first- and second-line therapies are associated with adverse effects. Corticosteroids have well-recognized side effects, so clinicians frequently seek approaches that exclude steroids. IVIg and anti-D immunoglobulin are associated with infusion reactions, headaches, and hemolytic anemia. Immunosuppressive agents are associated with an increased risk of infection. These adverse effects, along with low response rates, limit the usefulness of these approaches. Romiplostim is a thrombopoietin (TPO) receptor agonist that has been shown to increase and maintain platelet counts for up to 5 years and decrease the incidence of bleeding, splenectomy, treatment failure, and rescue drug use (Kuter et al., 2008, 2010, 2013). Romiplostim is approved in several regions worldwide for the treatment of chronic immune thrombocytopenia (ITP) in adults. It is approved in the United States for the treatment of chronic immune thrombocytopenia that does not respond adequately to corticosteroids, immunoglobulins, or splenectomy, and in Europe for patients who have undergone splenectomy and are refractory to other treatments or as second-line therapy in non-splenectomized patients for whom surgery is contraindicated. Romiplostim is administered via [unclear - possibly "nc / bcn / nznz / a / Y"]. -2 an initial dose of 1 pg / kg, with weekly doses adjusted by 1 pg / kg to achieve and maintain a platelet count of 50-200 x 109 / L. Data are emerging on the use of romiplostim in children with chronic immune thrombocytopenia (CT) 2. Treatment of children with CT 2 is often guided by data from studies in adults, pilot studies in children, and expert opinion. Three large randomized controlled trials in children with CT 2 have shown that treatment with TPO receptor agonists was associated with greater efficacy compared to placebo without surprising or unexpected adverse events. However, data on the long-term effects of treatment in children with CT 2 are lacking. Romiplostim can induce platelet responses in approximately 80–90% of patients with ITP (Bussel et al., 2009; Kuter et al., 2010), but long-term treatment may be required to maintain platelet responses. Long-term treatment, however, can be associated with compliance problems and substantial costs (Ghanima et al., 2012). New data suggest that certain patients, many of whom have suffered from relapsing or refractory disease, can achieve durable remission after definitive discontinuation of romiplostim (Vlachaki et al, 2011; Ghadaki et al, 2013; Thachil et al, 2013; Mahevas et al, 2014; Provan et al, 2014; Tarantino et al., 2016; Tarantino et al., ASH2017, abstract 14, available at: https: / / ash.confex.com / ash / 2017 / webprogram / Paper100126.html). Newland et al., 2016, conducted a systematic prospective evaluation of remission with romiplostim, but ultimately concluded that no statistically significant predictors of remission were discovered when other demographic or baseline characteristics were examined. However, a higher mean platelet count during the first two months of treatment was associated with remission.The field would benefit from the discovery of predictive factors for remission and treatment methods that take advantage of those predictive factors. SUMMARY OF THE INVENTION In the study underlying this specification, achieving a platelet count > 200 x 10⁹ / L within the first 4 to 12 weeks of treatment is considered a predictive factor for remission, defined in the study as a treatment-free period of at least 24 weeks during which the patient receives neither romiplostim nor any other ITP medication. Accordingly, the present invention relates to a method of treating ITP in a patient with ITP, comprising (a) administering romiplostim weekly to the patient; (b) increasing the weekly dose until a platelet count > 200 x 10⁹ / L is achieved. (a) platelet count of at least approximately 50 to 200 x 109 / L; (c) decrease the weekly dose of romiplostim if the platelet count remains > 200 x 109 / L for two consecutive weeks; (d) permanently discontinue romiplostim if the platelet count remains > 200 x 109 / L for two consecutive weeks when the weekly dose is 1 pg / kg or the platelet count > 400 x 109 / L; and (e) if a platelet count > 200 x 109 / L is achieved during the first 4 to 12 weeks of treatment, maintain a treatment-free period of at least approximately 24 weeks during which the patient does not receive romiplostim. The invention also relates to a method as described above comprising administering an initial dose of 1 pg / kg of romiplostim to the patient. The invention further relates to a method as described above comprising maintaining the weekly dose as long as the platelet count is within approximately 50 to 200 x 109 / L. The invention further relates to a method as described above in which the platelet count is > 200 x 109 / L in the first four weeks of treatment. The invention further relates to a method as described above in which the increase in the dose of romiplostim is in increments of 1 pg / kg per week. The invention further relates to a method as described above wherein the dose in step c is reduced in increments of 1 pg / kg. The invention further relates to a method as described above in which the patient's platelet count remains > 50 x 109 / L during the untreated period. The invention further relates to a method as described above in which the patient does not receive medication for ITP during the untreated period. BRIEF DESCRIPTION OF THE FIGURES Figure 1 shows a study flow diagram. Of the 21 patients shown who participated in the first extension of the study, one withdrew their consent before treatment. Of the 66 patients who proceeded to the second extension, one withdrew their consent before treatment. Figure 2 shows the distribution of patients who participated in the study. Of the three patients who definitively discontinued romiplostim according to the protocol, two achieved remission and one had a platelet count <30 x 10⁹ / L despite 10 weeks of treatment with 10 pg / kg. The 37 patients who completed romiplostim treatment received the drug until the study ended in January 2017, 12 months after the last patient was enrolled. nobcn / nznz / a / Y Figure 3A shows the romiplostim dose over time. The median dose (Q1, Q3) is shown on the y-axis with the number of weeks in the study on the x-axis. Figure 3B shows the platelet count over time. The median platelet count (Q1, Q3) is shown on the y-axis with the number of weeks in the study on the x-axis. Figure 4A shows bleeding over time for all patients and for those with grade > 2. Figure 4B shows grade 2 bleeding over time. Figures 5A–5D show data on remission. Figure 5A shows the probability of achieving remission, plotted against the time required to achieve remission, with the number of patients at risk at each time point along the x-axis. Figure 5B shows the number of patients who achieved remission, grouped by the duration of remission observed in the study. Figures 5A and 5D show the administered dose and platelet count achieved by two individual patients during the Phase 1 / 2 study period, including the first and second extensions. Figure 6 shows the instantaneous risk relationships and p-values ​​of selected remission features. Figures 7A-7M show the remission data described for Figures 50 and 5D for an additional 13 patients. DETAILED DESCRIPTION OF THE INVENTION TPO receptor agonists are not a perfect treatment option for ITP, as not all patients respond, and those who do are not always able to maintain their response. The present invention is based on the longest study to date in children with TPO receptor agonists, with over 182 patient-years of exposure, or 2–8 years per patient, for 65 patients. Importantly, approximately 1% of the patients were able to permanently discontinue ITP treatments while still maintaining hemostatic platelet counts. The data described in this manuscript are extensive, both in terms of the number of patients (n=65) and the duration of treatment (up to 7 years), and show that romiplostim has an efficacy and safety profile that appears similar to that observed in adults. The objectives of the study were to describe the safety and efficacy of long-term romiplostim use in children with ITP, with the incidence of adverse events as the primary endpoint. Secondary endpoints included the assessment of long-term platelet responses, bleeding, and reduction in the use of concomitant ITP medications (both continuation of medications from baseline and - 5 rescue medications) and a post-hoc endpoint of maintaining platelet counts >50*109 / L for 6 months without ITP medications, including romiplostim (herein defined as remission). Methods Patients Eligible patients had completed a previous ITP study with romiplostim (Bussel et al., 2011; Tarantino et al., 2016) and were <18 years of age at enrollment. Exclusion criteria included a history of a bone marrow stem cell disorder, a venous or arterial thrombotic or thromboembolic event, systemic lupus erythematosus, Evans syndrome, or other secondary causes of thrombocytopenia. Studies were conducted in compliance with all regulatory obligations and clinical research ethics committee and informed consent regulations at each research site. All patients or their legal representatives provided written informed consent and assent. Procedures Patients were recruited at 28 sites in the United States, Canada, Spain, and Australia. The study (clinicaltrials.gov identifier NCT01071954) ran from December 30, 2009 (first subject enrolled) to January 12, 2017 (last patient visit). During the study, patients received weekly subcutaneous romiplostim, starting at the same dose as the final dose of the main study or 1 pg / kg (if they were previously receiving placebo or more than 24 weeks since their last dose). The dose was titrated weekly in 1 pg / kg increments, between 1 and 10 pg / kg, until target platelet counts of 50–200 x 10⁹ / L were achieved. Patients could continue receiving other ITP medications (e.g., corticosteroids, danazol, or azathioprine) that they had been taking at a stable dose and dosage regimen prior to the start of the study.These additional medications could be reduced or discontinued at the patient's and doctor's preference after platelet counts exceeded 50x109 / L. Patients could receive rescue medications when their platelet count fell below 10 x 10⁹ / L, when bleeding or wet purpura was present, or when the investigator deemed it medically necessary (e.g., before travel or a procedure). Rescue medications were defined as any medication administered to increase the platelet count and included intravenous immunoglobulin G (IVIg), anti-D, platelet transfusions, steroids, and antifibrinolytics (e.g., epsilon-aminocaproic acid, tranexamic acid). nc / hcn / nznzn / v -6 Evaluations, results and statistics At each visit, platelet counts, concomitant medications, and adverse events were assessed. Blood samples were collected for a complete blood count and blood chemistry analysis every 4 weeks, and physical examinations were performed at week 1 and every 12 weeks thereafter. Efficacy outcomes included platelet count and platelet response (platelet count >50 x 10⁹ / L without rescue medication use in the preceding 4 weeks). Platelet count data for the four weeks following rescue medication use were excluded from continuous summaries. Subjects without weekly platelet counts due to self-administration or other reasons had their platelet count deferred as long as no rescue medication was used in the four weeks following the missed visit.Other efficacy assessments included the proportion of patients using other ITP medications and the proportion of patients requiring rescue medication. Remission was defined as a platelet count >50 x 10⁹ / L in the absence of all ITP medications, including romiplostim, for a period of at least 24 weeks. Security Safety assessments included review of adverse events (including bleeding), physical examination, vital signs, serum biochemistry, complete blood count, platelet count, and antibody status. Samples were tested for agglutinating antibodies against romiplostim and agglutinating antibodies against TPO; any sample testing positive for agglutinating antibodies against either was further tested for neutralizing antibodies. Medically significant adverse events deemed treatment-related by the investigator were followed until they resolved or the patient was considered stable.For each adverse event, the investigator assigned the following attributes: description; dates of onset and resolution; severity; assessment of relationship to the treatment, another suspected drug, or device; and actions taken (e.g., permanent discontinuation of romiplostim, treatment with another drug, hospitalization, etc.). The severity of toxicities was assessed according to the Conventional Terminology Criteria for Adverse Events (CTCAE) version 3.0, which categorizes adverse events as follows: 1 mild, 2 moderate, 3 severe, 4 life-threatening, 5 fatal. Severe adverse events included any adverse event that was fatal, life-threatening, required hospitalization, or prolonged. - 7. Existing hospitalization. Thromboembolic events were assessed as part of the overall adverse event assessment. Biopsies or bone marrow aspirations were not required at any time, but could be performed at the investigator's discretion; for example, if there were abnormalities on the peripheral blood smear, such as nucleated red blood cells or dacryocytes, or if there was a loss of response to romiplostim despite dose escalation. Blood samples for TPO antibody and romiplostim assays were collected at week 1, week 52, annually, and at the end of the study. Data analysis Statistical analyses were descriptive. Categorical endpoints were summarized by the number and percentage of patients in each category. Continuous endpoints were summarized by the number of patients, mean, standard deviation, median, and 25th (Q1) and 75th (Q3) percentiles, including minimum and maximum values. Adverse events were also summarized as the number of events and rate per 100 years of exposure. Proportional hazards models were used to assess potential predictors of remission in subjects without remission, as recorded at their final platelet count. For the univariate model, each potential factor was considered individually (analogous to a log-ordinal trial). In cases where the proportional hazards assumption was found to be violated, nonparametric tests were used instead (Fisher's exact test for categorical variables and Kruskal-Wallis test for continuous variables). For the multivariate models, a staged selection criterion was used, with significance levels for both entry and exit set at 0.05. Results Demographics and disposition The patient flow from the main studies is shown in Figure 1. Sixty-six patients participated in this extension; one withdrew consent prior to treatment, and 65 had received romiplostim for <7 years. Fifteen of these patients had previously received placebo, and this study was their first exposure to romiplostim. At baseline, the median age (min-max) was 11 (3-18) years; 56% were female; 61% were white, 14% African American, 14% Hispanic / Latino, 9% Asian, and 3% other. The median (min-max) baseline platelet count was 27.5 (2-458) x 10⁹ / L (Table 1); patients were able to participate in this study without interrupting their romiplostim dosing from their previous study. Prior ITP treatments included magnesium, anti-D, corticosteroids, and rituximab. 9% had had a previous splenectomy (Table 2). - 8The reasons for definitively discontinuing romiplostim (n=28.42%) included withdrawal of consent (n=10), need for other therapy (n=6), non-compliance (n=4), per protocol (n=3), administrative decision (n=2), adverse event (n=2), and unnecessary treatment (n=1). The adverse events were asthenia, headache, dehydration, and vomiting in one patient and anxiety in the other; according to the investigators, these adverse events were not considered treatment-related. Thirty-seven patients (56%) received romiplostim until the end of the study (Figure 2). Exposure to Romiplostim The median (min-max) duration of romiplostim treatment was 135 (5-363) weeks, or a median of 2.6 years per patient, for a total of 182 patient-years. The median (min-max) weekly romiplostim dose was 4.8 (0-1-10) pg / kg, including the step-up to a steady-state dose. The mean maximum weekly romiplostim dose was 6.9 pg / kg, and the median maximum weekly dose was 8.0 pg / kg. Twenty patients started on 1 pg / kg of romiplostim, including 15 patients (23%) who had received placebo in the previous study. All 65 patients received their doses according to protocol >90% of the time; 21 patients missed >1 dose due to non-compliance a total of 65 times. The dose over time was typically approximately 4-5 pg / kg, as shown in Figure 3A. After week 240 (n<7), the dose fluctuated. Security The most common adverse events were headache and contusion (Table 3, see below). Fifty-four serious adverse events occurred in 19 patients, but were considered treatment-related in only one patient who simultaneously had grade 4 thrombocytopenia, grade 3 epistaxis, and grade 2 anemia (full list of serious adverse events in Table Comp. 1, see below). Bleeding adverse events occurred in 57 patients; three bleeding adverse events were considered treatment-related (injection site bleeding, injection site hematoma, and epistaxis). The most frequent bleeding adverse events were contusion (51%), epistaxis (49%), petechiae (31%), and gingival bleeding (20%).There were no cases of intracranial hemorrhage; specific bleeding events reported included menorrhagia (n=7, 4%), hematuria (n=4, 2%), rectal bleeding (n=4, 2%), hemoptysis (n=3, 2%), anal bleeding (n=2, 1%), hematochezia (n=2, 1%), and hematemesis (n=1.0–6%). There were 7 patients with severe or grade 3 bleeding (Table Comp. 2, see below); 1 case of worsening epistaxis was considered treatment-related. No arterial or venous thromboembolic events were reported. It should be noted that the contusion rate would decrease from 239 to 92 per 100 patient-years if one patient who had 499 events were excluded. -9 adverse events. This child, a 7-year-old male who was in the study for 3.4 years, also had several serious adverse events: 6 of decreased platelet count and 1 each of headache, head injury, vomiting, leukopenia, hematoma, streptococcal pharyngitis and gastroenteritis. Post-dose antibodies were tested annually in 60 patients. One patient developed anti-romiplostim neutralizing antibodies after leaving the study to receive other therapy; these antibodies were not present when the tests were repeated 3 and 6 months later. She required multiple additional therapies and has remained stable on mycophenolate. No patients developed anti-TPO neutralizing antibodies, even though the annual antibody tests covered more than 200 patient-years of exposure (including time in previous romiplostim studies). Bone marrow biopsies were performed in two patients with additional cytopenias; both patients were found to have iron deficiency anemia and no fibrosis or neoplasms. One biopsy, performed after 2 years of study, was in a 17-year-old girl to evaluate her persistent anemia. With regular supplemental iron intake and lighter menstrual bleeding, her anemia improved. The second biopsy was performed after 6 weeks of study in an 11-year-old girl, as she developed neutropenia and anemia; she received iron for the anemia and had intermittent neutropenia that eventually resolved. Efficacy From week 2 onward, the median platelet count remained >50 x 10⁹ / L; the median platelet count was >100 x 10⁹ / L from week 24 to week 260 (Figure 3B). Nearly all patients (94%, 61 / 65) had a platelet response >1 (platelet counts >50 x 10⁹ / L, excluding counts <4 weeks after rescue medication). The majority of patients (72%, 47 / 65) had a platelet response >75% of the time, and more than half (58%, 38 / 65) had a platelet response >90% of the time. Sixty patients (92%) (or caregivers) self-administered romiplostim (i.e., at home, not in the clinic). Twenty-three (35%) patients received rescue medication (Table 4, see below). Use was higher in the first few months. At baseline, 5 patients were taking other medications for ITP such as aminocaproic acid, prednisolone, prednisone, and tranexamic acid.Overall, 48% (31 / 65) of patients were taking other medications for ITP (either at baseline or as rescue medication); the rate of patients taking these medications decreased over the course of the study. Bleeding, both overall and grade >2, decreased over time (Figures 4A and B). nobcn / nznz / a / Y Remission / Periods without treatment Fifteen (23%) patients maintained a platelet count >50 x 10⁹ / L for >24 weeks without ITP medication (i.e., a treatment-free period, defined herein as remission; Table 5, Figures 5A–D). At the start of the treatment-free periods, these patients (9 girls, 6 boys) had had ITP for a median (min–max) of 4 (1–12) years and had received romiplostim for 3 (1–7) years. None of these patients had undergone a prior splenectomy. The median (min–max) baseline platelet count was 14 (1–44) x 10⁹ / L. These treatment-free periods lasted a median (min–max) of 1 (0–62) years, with all but one patient still off all treatments and with a platelet count >50 x 10⁹ / L at the end of the study. In the last months of dose reduction, the median (min-max) of peak platelet counts was 299 (217-730) x109 / L.After the first month of definitive discontinuation, when the drug effect / rebound was no longer expected, the median (min-max) peak platelet count was 249 (150-450) x 10⁹ / L and the median (min-max) trough platelet count was 113 (77-311) x 10⁹ / L (as can be seen in the individual patient graphs, Figures 5A-D). All 15 patients had a platelet count >100 x 10⁹ / L for >3 months and 12 / 15 for >6 months, all without ITP medication. The median duration (min-max) with >100 x 10⁹ / L for these 15 patients was 42 (13-109) weeks. In post-hoc analyses, baseline characteristics and early treatment outcomes, such as ITP duration, prior ITP treatments, and platelet count in the first 4 weeks, were evaluated. Younger age at diagnosis, younger age at first dose, platelet count >200 x 10⁹ / L in the first 4 weeks, and a higher mean platelet count in the first 4 weeks were associated with a greater likelihood of achieving remission in univariate analyses (Table 5, Figure 3). In multivariate analyses, age at first dose (p = 0.001) and platelet count >200 x 10⁹ / L in the first 4 weeks (p = 0.004) were predictive of remission. It is worth noting that the models for prior rituximab use (p=1.0) and prior splenectomy (p=0.32) had non-proportional instantaneous risks and therefore did not have an instantaneous risk relationship.When multivariate analyses were performed by baseline age, for those <10 years of age at first dose (N=32), <3 prior ITP treatments and platelet counts >200x10⁹ / L in the first 12 weeks were predictive of remission; for those <10 years of age at diagnosis (N=49), younger age at first dose, platelet counts >200x10⁹ / L in the first 4 weeks, mean platelet count >100x10⁹ / L in the first 12 weeks, and use of rescue medication in the first 6 months were also predictive, with rescue medication use being a negative predictor. There were too few patients >10 years of age to develop models of predictive factors for remission. nobcn / nznz / a / Y - 11 Discussion Effectiveness Romiplostim treatment was frequently associated with a sustained platelet response, as demonstrated by the findings that the majority (72%) of children had a platelet response >75% of the time and more than half (58%, 38 / 65) had a platelet response >90% of the time. Furthermore, median platelet counts remained within the target range (50–200 x 10⁹ / L) from week 2 onward and were >100 x 10⁹ / L from week 24 to week 260. Patients in this study were also able to reduce their use of other medications for ITP and reported less bleeding over time, particularly clinically significant bleeding (grade >2). Security The data from this study showed that romiplostim was well tolerated, with no thrombotic events, bone marrow changes, deaths, or new safety issues, despite 182 patient-years of romiplostim exposure; that is, almost 3 years per patient. The median dose was the same as in the phase 3 study (~4-5 pg / kg). Approximately 30% of patients experienced serious adverse events, but only one patient had serious adverse events related to treatment (thrombocytopenia, epistaxis, and anemia). Despite the large number of patient-years of treatment, only one patient was identified who developed anti-romiplostim neutralizing antibodies, and no patients had neutralizing antibodies against TPO. This is consistent with what has been observed in adults treated for ITP; in an integrated database of ITP trials with romiplostim, anti-romiplostim neutralizing antibodies were discovered in 4 of 1046 patients with a total exposure of 1832 patient-years.9 All 4 continued to have platelet responses with romiplostim. Bone marrow biopsies were performed when researchers felt they were clinically indicated. Bone marrow biopsies were performed on two patients because they had additional cytopenias; both had iron deficiency anemia. No cases of thrombosis were observed. In adult studies of romiplostim, thrombosis rates have been low and not different from those of placebo / standard treatment (Cines et al, 2015).10nc t bcn / ηζηζ / α / υιλι - 12 Although only two patients permanently discontinued treatment due to adverse events, overall, 42% (28 / 66) of patients dropped out of the study before it ended, the most common reasons being withdrawal of consent (n=10) and the need for alternative therapy (n=4). Remission / treatment-free periods A pleasant surprise was that almost a quarter (15 / 65) of patients were able to permanently discontinue all ITP treatments (including romiplostim) after receiving 0.7–6 years of romiplostim and still maintaining hemostatic platelet counts (>50 x 10⁹ / L) for at least 6 months, which is defined herein as remission. Only 1 / 15 patients relapsed and required additional treatment with 1 pg / kg of romiplostim. At the last follow-up, this patient had permanently discontinued romiplostim once again, as the platelet count had reached 400 x 10⁹ / L. In a multivariate analysis, a younger age at the time of the first dose and a higher mean platelet count in the first 8 weeks were associated with a greater likelihood of achieving remission. This is of particular interest because a previous study in adults also observed that a higher mean platelet count in the first 8 weeks was associated with a greater likelihood of achieving remission, although the authors ultimately concluded that no statistically significant predictors of remission were found when other demographic or baseline characteristics were examined. (Newland et al., 2016). Regarding the remission analyses, it should be noted that although no predictive factors for remission were discovered in this study, this could be due to the small size of the populations or their being an unsuitable population for the study. Absence of evidence is not evidence of absence. Furthermore, it should be noted that the definition of remission can vary considerably, sometimes having higher platelet thresholds (e.g., 100 x 10⁹ / L instead of 50 x 10⁹ / L) and / or longer durations (e.g., one year instead of six months). For the purposes of this descriptive report, remission is defined as the maintenance of a platelet threshold of 50 x 10⁹ / L for at least six months (24 weeks) without ITP medication. The development of remission was not entirely expected, and remission was not a predetermined study outcome measure, but rather an observed phenomenon. Although remission in children is likely to be very different from that observed in adults, this finding is consistent with previous studies on romiplostim in adults with ITP, including one in which 75 patients who had ITP for <6 months were treated with - 13 romiplostim for <12 months, and then had a forced taper (Newland et al, 2016).11Remission was observed in 24 patients (32%), with no significant predictive factors. Most patients (20 / 24) had an onset of remission before the forced taper. Treatment-free periods were also observed in adults with ITP in the eltrombopag extension study (EXTEND) (Saleh et al, 2013)12, in which a prolonged response (defined as platelet counts >50 x 10⁹ / L for >12 weeks without ITP medications, including eltrombopag) was found in 13 of 325 patients (4%). The median (min-max) time since ITP diagnosis was 26 months (9-128), and the median (min-max) duration of eltrombopag treatment in the extension before the prolonged response was 160 days (14-1107). In a retrospective evaluation of adults with ITP receiving eltrombopag, eltrombopag was permanently discontinued in 80 / 201 patients with a complete response (platelet counts >100x109 / L) (González-López et al, 2015)13. Of 49 evaluable patients, 26 showed a sustained response for at least 6 months after permanently discontinuing eltrombopag without additional ITP therapy. No predictive factors for a sustained response after eltrombopag withdrawal were identified. As noted above, the definitions of response, remission, and sustained response can vary considerably. In the study reported herein, platelet counts >50 x 10⁹ / L were chosen for response, and platelet counts >50 x 10⁹ / L for >6 months without ITP medication were chosen for remission. Other studies have used different platelet thresholds for response and treatment-free periods, such as the IWG response (Rodeghiero et al.)14, in which thresholds of 30 x 10⁹ / L and 100 x 10⁹ / L were used for response and complete response, both in the absence of bleeding, or treatment-free periods of at least one year, as in a long-term rituximab study (Patel et al., 2012).15 The longer duration of ITP before remission, as observed in this study, indicates that this remission is likely not spontaneous but rather due to romiplostim, suggesting that romiplostim could be a disease-altering agent, which is unexpected since romiplostim is not known to be an immunomodulator. Without intending to commit to any particular theory, possible mechanisms of action include: • Enhanced function of regulatory T lymphocytes (Bao et al., 2010; Chong, 2010; Son et al., 2015)16-18 • Natural killer T lymphocytes (Johansson et al., 2005)19 • Activity of regulatory B lymphocytes (Liu et al., 2012)20 • Induction of crystallizable fragment of receptor IIb (FcRIIb), the inhibitory FcR (Liu et al., 2016)21 The development of remission in almost % of patients suggests that maintenance with romiplostim may not be an indefinite lifelong treatment; the response in the first 8 weeks may indicate whether a patient is more likely to enter remission. ncibcn / nznz / a / Y Table 1. Baseline Demography N=66 n (%) Female 37 (56) Race / Ethnicity White 40 (61) African American 9(14) Hispanic / Latino 9(14) Asian 6(9) Other 2(3) Median age (min, max), years 11 (3, 18) Age group, years >1-< 6 12 (18) >6-< 12 25 (38) > 12 29 (44) Median baseline platelet count χ 109 / L (min, max) * 28 (2, 458) * For the extension study described in this article (i.e., not the main studies) Table 2. Patient's ITP medication history N=66 n (%) Duration of ITP, median (min-max), years 3-0(1-7-5-6) Number of previous ITP treatments 1 7(11) 2 17 (26) 3 15 (23) >3 26 (39) Previous splenectomy 6(9) Number of patients who received specific therapies in the past Mg 60 (91) Corticosteroid 54 (82) Anti-D antibody (WinRho) 24 (36) Rituximab 24 (36) Vincristine / Vinblastine 4(6) Danazol 4(6) Azathioprine 4(6) Other* 26 (39) Others include aminocaproic acid, cyclosporine, dapsone, mercaptopurine, mycophenolate, platelets, sirolimus, and tranexamic acid. Please note that the designation of platelet transfusions as rescue medication was made by the investigator. nobcn / nznz / a / Y Table 3. Adverse events Category Adverse Event N=65 n (%) Duration-adjusted events (per 100 patient-years) All 182 patient-years Excluding patients with 499 AE 179 patient-years Most common adverse events Headache 38 (59) 151 (83) 126 (71) Contusion 33 (51) 435 (239) 164 (92) Epistaxis 32 (49) 103 (57) 98 (55) Upper respiratory tract infection 32 (49) 101 (56) 101 (57) Serious adverse events* Any 19 (29) 54 (30) 41 (23) Thrombocytopenia 4(6) 6(3) 6(3) Pyrexia 3(5) 3(2) 3(2) Epistaxis 2(3) 2(1) 2(1) Headache 2(3) 2(1) 1 (0.6) Vomiting 2(3) 2(1) 1 (0.6) * The most common serious adverse events are included; a full list can be found in the supplementary tables. Table 4. Rescue medications N=65 n (%) Events (rate per 100 year-pt) Any use 23 (35) 80 (44) Intravenous immunoglobulin 10 (15) 31 (17) Corticosteroids* 13 (20) 31 (17) Antifibrinolytic (aminocaproic acid or tranexamic acid) 6 (9) 14 (8) Azathioprine 1 (2) 1 (0-6) Red blood cell transfusion 1 (2) 1 (0-6) Platelet transfusion 1 (2) 2 (1) * Corticosteroids include prednisone / prednisolone, methylprednisolone, and dexamethasone. Table 5. Treatment-free periods of >24 weeks with platelet counts >50x109 / L (herein defined as remission) nobcn / nznz / a / Y Patient number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Age at remission onset, years 16 6 10 8 6 12 18 7 6 9 4 5 6 14 16 Sex FMMFFFFMMFMFFMF Race / Ethnicity WBWWWWWWBAWBHWW ITP, years* 7 6 5 1 4 11 12 3 1 2 3 4 3 3 5 Number of previous ITP therapies (prior to studies) 3 6 4 2 2 3 4 5 2 1 2 1 4 3 4 Prior rituximab use, years prior to remission onset 4 5 5 NNNNNNNNN 2 N 5 Baseline platelet count (x109 / L)T 12 5 9 18 7 44 15 26 28 28 25 1 11 14 4 Romiplosti m, years* 7 6 5 2 3 5 4 3 3 2 2 2 3 1 3 Maximum dose, pg / kg 10 8 9 5 10 2 2 1 3 1 2 1 9 10 4 ITP remission, years 1- 2· 1- 1- 1- 0· 0- 1- 2- 1· 0-6 0- 0- 04 0- 1 1 1 6 0 8 9 7 1 1 ★ 6 8 t 6 Data are integrated on the main study and extension study A, Asian; B, black; nc / bcn / nznz / a / Y F, female; H, Hispanic / Latino; M, male; N, no rituximab use; Rx, therapy; W, white. *AI, baseline of the main study (not the extension). *AI, baseline of remission. *Remission ended before the study ended. *This patient met remission criteria for 4 years in the study and >0-5 years post-study. Table 6. Predictive factors for remission Characteristic Patients with remission (N=15) Patients without remission (N=50) RRI RRI 95% CI p-value Sex, female, n (%) 9 (60) 27 (54) 1.19 (0.42, 3.41) 0.74 Race, white, n (%) 10 (67) 30 (60) 1.05 (0.36, 3.09) 0.93 Age at time of 1st dose 6.5 (4.0) 10.6 (4.0) 0.81 (0.71, 0.93) 0.002 Age at time of ITP diagnosis 4.8 (3.6) 7.5 (3.4) 0.83 (0.70, 0.98) 0.03 Duration of baseline ITP 2.3 (2.4) 3.6 (2.7) 0.79 (0.61, 1.03) 0.08 Baseline platelet count* 16.5 (11.8) 15.9 (9.5) 1.09 (0.66, 1.80) 0.74 NQ of previous treatments 3.1 (1.4) 3.3 (1.9) 0.77 (0.56, 1.07) 0.12 Previous rituximab, n (%) 5 (33) 19 (38) N / A / N / A 1 Splenectomized, n (%) 0(0) 6(12) N / A / N / A 0.32 Dose at first response (pg / kg) 3.1 (2.9) 4.1 (3.0) 0.87 (0.71, 1.07) 0.19 Platelet count >200x109 / L in the first 4 weeks 4(27) 3(6) 5.48 (1.63, 18.42) 0.006 Mean platelet counts in the first 4 weeks* 128(149) 57.5 (42.9) 1.09 (1.04, 1.13) <0.0001 Grade >2 bleeding in the first 6 months 4(27) 8(16) 1.44 (0.46, 4.53) 0.53 Rescue drugs in the first 6 months 4(27) 20 (40) 0.65 (0.21,2.04) 0.46 Data are median (min-max) or mean (SD) unless otherwise stated. RRI, hazard ratio; w / o, sin. * = per 10 x 10⁹ / L. N / A for RRI and 95% CI when the proportional hazards assumption was violated in the model and the model results are unreliable. P-value calculated using Fisher's exact test (categorical variables) or the Kruskal-Wallis test (continuous variables). Tables to complement them Comp. Table 1. Serious adverse events N=65 n (%) Thrombocytopenia 4(6) Pyrexia 3(5) Epistaxis 2(3) Headache 2(3) Vomiting 2(3) Anemia 1 (2) Asthenia 1 (2) Asthma 1 (2) Biliary dyskinesia 1 (2) Clostridium difficile infection 1 (2) Contusion 1 (2) Dehydration 1 (2) Depression 1 (2) Febrile neutropenia 1 (2) Gastroenteritis 1 (2) Gastrointestinal infection 1 (2) Gingivitis 1 (2) Hemangioma 1 (2) Hematoma 1 (2) Head injury 1 (2) ITP 1 (2) Infection 1 (2) Leukopenia 1 (2) Viral meningitis 1 (2) Metapneumovirus infection 1 (2) Oral hemorrhage 1 (2) Streptococcal pharyngitis 1 (2) Reduced platelet count 1 (2) Mycoplasmal pneumonia 1 (2) Post-procedure hemorrhage 1 (2) Respiratory syncytial virus infection 1 (2) Subcutaneous abscess 1 (2) Suicidal ideation 1 (2) Transfusion reaction 1 (2) Ulcer hemorrhage 1 (2) Viral infection 1 (2) Viral upper respiratory tract infection 1 (2) Comp. Table 2. Severe and / or Grade 3 hemorrhagic adverse events Age (y) or X Φ ω Week Adverse hemorrhagic event Duration (days) Grade Related Severe Actions Severe hemorrhage 7 Child 96 Hematoma 16 1 NS Hospitalized, no change in romiplostim - 20 10 4 Boy 24 Epistaxis and oral bleeding 1 2 NS Hospitalized, no change on romiplostim 86 Bleeding after tonsillectomy (previous day platelet count was 191x 109 / L) 1 2 NS Hospitalized, no change on romiplostim 1 1 Girl 40 Worsening ITP and contusion 3 1 NS Hospitalized, dropped out of study a few days later due to non-compliance 0 Grade 3 bleeding 7 Boy 12 Worsening epistaxis, concurrent grade 4 thrombocytopenia 2 3 SS Hospitalized, no change on romiplostim (later discontinued as other treatment required) 3 Girl 3 Bleeding oral sores 1 3 NS Emergency Department 1 4 Boy 295 Hematuria 5 3 NN Emergency Department, no change on romiplostim 8 Boy 94 Increased petechiae 13 3 NN No change in romiplostim (later discontinued as required by other treatment) nc t bcn / nznz / α / υιλι - 21 *This patient reported 499 adverse events. Emergency Department; N, no; S, yes. Retrospective statistical analyses were performed on two adult studies (clinical trials 20080009 and 20080435) to confirm that the results observed in the pediatric study described above correlated with the adult population. Trial '0009, published by Janssens et al. (2016), involved chronic ITP, with romiplostim treatment lasting up to three years. Trial '0435, published by Newland et al. (2015), involved early treatment of ITP with romiplostim for up to one year, with a specific attempt at tapering the dose to see if the subject achieved remission. Both studies individually showed that having a platelet count >200 x 10⁹ / L within the first four, eight, and twelve weeks of romiplostim treatment was associated with a higher remission rate. A secondary, more conservative association trial was used for study '0009.In both adult studies, achieving a platelet count > 200 x 10⁹ / L within the first eight weeks was the strongest predictor of remission. The results of the analyses are shown in Tables A–D below. nobcn / nznz / a / Y Table A. Univariate analysis with respect to clinical trial 20080009 Variable Patient with remission (N=24) Patients without remission (N=51) Instantaneous hazard ratio 95% CI for RRI p-value Platelets >200 in the first 4 weeks, Yes -n (%) 9(38) 11 (22) 2.30 (1.00, 5.27) 0.0491 Platelets >200 in the first 8 weeks, Yes -n (%) 15 (63) 18 (35) 2.83 (1.23, 6.49) 0.0140 Platelets >200 in the first 12 weeks, Yes -n (%) 16 (67) 24 (47) 2.14 (0.91, 5.00) 0.0806 The instantaneous risk ratio (RR), the 95% confidence interval (CI) and the p-value Calculated using a proportional hazards model with a single explanatory variable. An IRR >1 indicates a higher probability of remission. N / A for IRR and 95% CI when the proportional hazards assumption was violated in the model and the model results are unreliable. P-value calculated using Fisher's exact test to test for association. Table B. Patients in remission with platelet count > 200 x 109 / L at Weeks 4, 8 and 12 of Clinical Trial 20080009 With Platelets >200 With Platelets <200 n / N with remission (%) - Week 4 9 / 20 (45%) 15 / 55 (27%) n / N with remission (%) - Week 8 15 / 33 (45%) 9 / 42 (21%) n / N with remission (%) - Week 12 16 / 40 (40%) 8 / 35 (23%) Table C. Univariate analysis with respect to clinical trial 20080435 Variable Patient with Remission (N=25) Patients without remission (N=144) Instantaneous Hazard Ratio 95% CI for RRI p-value Platelets >200 in the first 4 9 (36) 25 (17) 2.54 (1.12, 5.74) 0.0257 weeks, Yes -n (%) Platelets >200 in the first 8 weeks, Yes -n (%) 16 (64) 45 (31) N / A / N / A 0.0028 Platelets >200 in the first 12 weeks, Yes -n (%) 17 (68) 55 (38) N / A / N / A 0.0079 The hazard ratio (HR), 95% confidence interval (CI), and p-value were calculated using a proportional hazards model with a single explanatory variable. An HR >1 indicates a higher probability of remission. N / A for HR and 95% CI when the proportional hazards assumption was violated in the model and the model results are unreliable. P-value calculated using Fisher's exact test to test for association. Table D. Patients in remission with platelet count > 200 x 109 / L at Week 4, 8 and 12 of Clinical Trial 20080435 With Platelets >200 With platelets <200 n / N with remission (%) - Week 4 9 / 20 (45%) 15 / 55 (27%) n / N with remission (%) - Week 8 15 / 33 (45%) 9 / 42 (21%) n / N with remission (%) - Week 12 / N / N with remission (%) -24

[0001] References Bao W, Bussel JB, Heck S, et al. (2010). Improved regulatory T-cell activity in patients with chronic immune thrombocytopenia treatedwith thromboooietic agents. Blood 116:4639-45. Bussel JB, Hsieh L, Buchanan GR, et al. (2014). Long-term use ofthe thrombopoietinmimetic romiplostim in children with severe chronic immune thrombocytopenia (ITP). Pediatr Blood Cancer 62:208–13. Bussel JB, Buchanan GR, Nugent DJ, et al. (2011). A randomized, double-blind study of romiplostim to determine its safety and efficacy in children with immune thrombocytopenia. Blood 118:28–36. Bussel, J.B., Kuter, D.J., Pullarkat, V., Lyons, R.M., Guo, M. & Nichol, J.L. (2009). Safety and efficacy of long-term treatment with romiplostim in thrombocytopenic patients with chronic ITP. 6 / ood, 113. 2161-2171. Chong B.H. (2010). ITP: Tregs come to the rescue. Blood 116: 4388-90. Cines D.B., Gernsheimer T, Wasser J, et al. (2015). Integrated analysis of long-term safety in patients with chronic immune thrombocytopaenia (ITP) treated with the thrombopoietin (TPO) receptor agonist romiplostim. Int J Hematol 102: 259-70. Cines D.B., Wasser J, Rodeghiero F, et al. (2017). Safety and efficacy of romiplostim in splenectomized and nonsplenectomized patients with primary immune thrombocytopenia. Haematologica 102:1342-51. George J.N., Woolf S.H., Raskob G.E., et al. (1996). Idiopathic thrombocytopenic purpura: a practice guideline developed by explicit methods for the American Society of Hematology. Blood 88: 3-40. Ghanima, W., Godeau, B., Cines, D.B. y Bussel, J.B. (2012). How I treat immune thrombocytopenia: the choice between splenectomy or a medical therapy as a second-line treatment. Blood, 120, 960-969. Gonzalez-Lopez, T.J., Pascual, C., Alvarez-Roman M.T. et al. (2015). Successful discontinuation of eltrombopag after complete remission in patients with primary immune thrombocytopenia. Am J Hematol 90: E40-3. Janssens A., Rodeghiero F., Anderson D., Chong B.H., Boda Z., Pabinger I., Cervinek L., Terrell D.R., Wang X., Franklin J. (2016). Changes in bone marrow morphology in adults receiving romiplostim for the treatment of thrombocytopenia associated with primary immune thrombocytopenia. Ann Hematol. Jun; 95(7): 1077-87. doi: 10.1007 / s00277-016-2682-2. Epub del 30 de abril de 2016. PubMed PMID: 27130310; PubMed Central PMCID: PMC4889627. Kuter, D.J., Bussel, J.B., Lyons, R.M., Pullarkat, V., Gernsheimer, T.B., Senecal, F.M., Aledort, L.M., George, J.N., Kessler, C.M., Sanz, M.A., Liebman, H.A., Slovick, F.T., de Wolf, J.T., Bourgeois, E., Guthrie, T.H. Jr, Newland, A., Wasser, J.S., Hamburg, S.L, Grande, C., -25Lefrere, F., Lichtin, A.E., Tarantino, M.D., Terebelo, H.R., Viallard, J.F., Cuevas, F.J., Go, R.S., Henry, D.H., Redner, R.L., Rice, L., Schipperus, M.R., Guo, D.M. y Nichol, J.L. (2008). Efficacy of romiplostim in patients with chronic immune thrombocytopenic purpura: a double-blind randomised controlled trial. Lancet, 371,395-403. Kuter, D.J., Bussel, J.B., Newland, A., Baker, R.I., Lyons, R.M., Wasser, J., Viallard, J.F., Macik, G., Rummel, M., Nie, K. & Jun, S. (2013). Longterm treatment with romiplostim inpatients with chronic immune thrombocytopenia: safety and efficacy. British Journal of Haematology, 161,411-423. Kuter, D.J., Rummel, M., Boccia, R., Macik, B.G., Pabinger, I., Selleslag, D., Rodeghiero, F., Chong, B.H., Wang, X. & Berger, D.P. (2010) Romiplostim or standardof care inpatients with immune thrombocytopenia. New Enoland Journal of Medicine, 363, 1889-1899. L¡ X., Zhong H., Bao W. etal. (2012). Detective regulatory B-cellcompartmentinpatients with immune thrombocytopenia. Blood 120: 3318-25. Liu X.G., Liu S., Feng Q. etal. (2016). Thrombopoietin receptoragonistsshiftthe balance of Fcgamma receptors toward inhibitory receptor llb on monocytes in ITP. Blood 128: 852-61. Mahevas, M., Fain, O., Ebbo, M., Roudot-Thoraval, F., Limal, N., Khellaf, M., Schleinitz, N., Bierling, P., Godeau, B. & Michel, M. (2014). The temporary use of thrombopoietin-receptor agonists may induce a prolonged remission in adult chronic immune thrombocytopenia. Results of a French observational study. British Journal of Haematoloav, 165, 865-869. Newland A., Godeau B., Priego V. etal. (2016). Remission andplateletresponses with romiplostim in primary immune thrombocytopenia: final results from a phase 2 study. Br J Haematol 172: 262-73. doi: 10.1111 / bjh. 13827. Epub de 5 de noviembre de 2015. PubMed PMID: 26537623. Nugent D., McMillan R., Nichol J.L., Slichter S.J. (2009). Pathogenesis of chronic immune thrombocytopenia: increased platelet destruction and / or decreased plateletproduction. Br J Haematol 146: 585-96. Oved J., Lee C., Bussel J. (2017). Treatment of Children with Persistent and Chronic Idiopathic Thrombocytopenic Purpura: 4 Infusions of Rituximab and Three 4-Day Cycles of Dexamethasone. Journal of Pediatrics (en prensa). Patel V.L., Mahevas M., Lee S.Y. et al. (2012). Outcomes 5 years after response to rituximab therapy in children and adults with immune thrombocytopenia. Blood Λ Λ 9: 5989-95. Provan D., Stasi R., Newland A.C. et al. (2010). International consensus report on the investigation and management of primary immune thrombocytopenia. Blood ^5:168-86. Provan, D., Taylor, L., Nandigham, R., Doobaree, U., Kalkur, P. & Newland, A. (2014). Sustained responses following treatment with romiplostim in immune thrombocytopenia: a single-centre expe henee. Journal of Hematoloav & Thromboembolic Diseases, 2, 147 nobcn / nznz / a / Y -26Rodeghiero F., Stasi R., Gernsheimer T. et al. (2009). Standardization of terminology, definitions and outcome criteria in immune thrombocytopenic purpura of adults and children: report from an international working group. BloodV\3: 2386-93. Saleh M.N., Bussel J.B., Cheng G. et al. (2013). Safety and efficacy of eltrombopag for treatment of chronic immune thrombocytopenia: results of the long-term, open-label EXTEND studv. g / ood 121: 537-45. Son B.R., Kim J.Y. (2015). Association of CD4(+)CD25(+)FoxP3(+) regulatory T cells with natural course of childhood chronic immune thrombocytopenic purpura. Korean J Pediatr 58:178-82. Tarantino et al., ASH 2017, resumen 14, disponible en: https: / / ash.confex.com / ash / 2017 / webprogram / Paper100126.html Tarantino M.D., Bussel J.B., Blanchette V.S. et al. (2016). Romiplostim in children with immune thrombocytopenia: a phase 3, randomised, double-blind, placebo-controlled study. Lancet388: 45-54. Thachil, J., Salter, I., George, J.N. (2013). Complete remission of refractory immune thrombocytopenia (ITP) with a short course of Romiplostim. European Journal of Haematoloav. 91,376-377. Wang B., Nichol J.L., Sullivan J.T. (2004). Pharmacodynamics andpharmacokinetics of AMG 531, a novel thrombopoietin receptor ligand. Clin Pharmacol Ther76‘. 628-38. Each reference cited in this specification is incorporated by reference in its entirety. It is understood that the invention disclosed herein is not limited to the specific methodology, protocols, and materials described, as these may vary. It is also understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of the appended claims. Skilled workers will recognize, or be able to determine using only conventional experiments, many equivalents to the specific embodiments of the invention described herein. It is intended that such equivalents are included in the following claims.

Claims

1. A method of treating idiopathic thrombocytopenic purpura (ITP) in a patient having ITP, comprising: a. Administering romiplostim weekly to the patient; b. Increasing the weekly dose until a platelet count of at least approximately 50 to 200 x 10⁹ / L is achieved; c. Decreasing the weekly dose of romiplostim if the platelet count remains > 200 x 10⁹ / L for two consecutive weeks; d. Permanently discontinuing romiplostim if (i) the platelet count has remained > 200 x 10⁹ / L for two consecutive weeks when the weekly dose is 1 pg / kg or (ii) the platelet count is > 400 x 10⁹ / L; and e. If a platelet count > 200 x 109 / L is achieved within the first 4 to 12 weeks of treatment, maintain a treatment-free period of at least approximately 24 weeks during which the patient does not receive romiplostim.

2. The method of claim 1, comprising administering an initial dose of 1 pg / kg of romiplostim to the patient.

3. The method of claim 1, comprising maintaining the weekly dose as long as the platelet count is within approximately 50 to 200 x 109 / L.

4. The method of claim 1, wherein the platelet count is > 200 x 109 / L in the first four weeks of treatment.

5. The method of claim 1, wherein the platelet count is > 200 x 109 / L in the first eight weeks of treatment.

6. The method of claim 1, wherein the increase in the romiplostim dose is in increments of 1 pg / kg weekly.

7. The method of claim 1, wherein the dose in step c is reduced in increments of 1 pg / kg.

8. The method of claim 1, wherein the patient's platelet count remains > 50 x 109 / L during the untreated period.

9. The method of claim 1, wherein the patient does not receive any ITP medication during the treatment-free period.