Eltrombopag for Low-Risk Myelodysplastic Syndromes With Thrombocytopenia: Interim Results of a Phase II, Randomized, Placebo-Controlled Clinical Trial (EQOL-MDS)

PURPOSE In myelodysplastic syndromes (MDS), severe thrombocytopenia is associated with poor prognosis. This multicenter trial presents the second-part long-term efficacy and safety results of eltrombopag in patients with low-risk MDS and severe thrombocytopenia. METHODS In this single-blind, randomized, placebo-controlled, phase-II trial of adult patients with International Prognostic Scoring System low- or intermediate-1-risk MDS, patients with a stable platelet (PLT) count (<30 × 103/mm3) received eltrombopag or placebo until disease progression. Primary end points were duration of PLT response (PLT-R; calculated from the time of PLT-R to date of loss of PLT-R, defined as bleeding/PLT count <30 × 103/mm3 or last date in observation) and long-term safety and tolerability. Secondary end points included incidence and severity of bleeding, PLT transfusions, quality of life, leukemia-free survival, progression-free survival, overall survival and pharmacokinetics. RESULTS From 2011 to 2021, of 325 patients screened, 169 patients were randomly assigned oral eltrombopag (N = 112) or placebo (N = 57) at a starting dose of 50 mg once daily to maximum of 300 mg. PLT-R, with 25-week follow-up (IQR, 14-68) occurred in 47/111 (42.3%) eltrombopag patients versus 6/54 (11.1%) in placebo (odds ratio, 5.9; 95% CI, 2.3 to 14.9; P < .001). In eltrombopag patients, 12/47 (25.5%) lost the PLT-R, with cumulative thrombocytopenia relapse-free survival at 60 months of 63.6% (95% CI, 46.0 to 81.2). Clinically significant bleeding (WHO bleeding score ≥ 2) occurred less frequently in the eltrombopag arm than in the placebo group (incidence rate ratio, 0.54; 95% CI, 0.38 to 0.75; P = .0002). Although no difference in the frequency of grade 1-2 adverse events (AEs) was observed, a higher proportion of eltrombopag patients experienced grade 3-4 AEs (χ2 = 9.5, P = .002). AML evolution and/or disease progression occurred in 17% (for both) of eltrombopag and placebo patients with no difference in survival times. CONCLUSION Eltrombopag was effective and relatively safe in low-risk MDS with severe thrombocytopenia. This trial is registered with ClinicalTrials.gov identifier: NCT02912208 and EU Clinical Trials Register: EudraCT No. 2010-022890-33.


INTRODUCTION
3][4][5] Treatment is still generally limited to PLT transfusions 6 since no effective drug is currently available for these patients, representing an unmet need in clinical practice. 7,8rombopoietin is the key regulator of PLT production by binding to its specific receptor thrombopoietin (TPO) receptor (TPO-R), on the megakaryocytic surface. 9A randomized, double-blind study with the TPO-R agonist, romiplostim, versus placebo for lower-risk MDS was stopped early because of an apparent increased risk of AML progression, which was not confirmed with long-term follow up. 10,11trombopag is an orally bioavailable, small molecule acting as a TPO-R agonist, approved for the treatment of thrombocytopenia of chronic immune thrombocytopenic purpura, chronic hepatitis C virus infection, and for acquired severe aplastic anemia. 12,13In higher-risk MDS, the addition of eltrombopag to azacitidine resulted in worse PLT recovery and increased progression to AML. 14 We have previously reported on the short-term outcome of the first 90 cases enrolled in the EQOL-MDS trial, a phase-II, randomized study designed to assess eltrombopag efficacy and safety compared with placebo in patients with lower-risk MDS with severe persistent thrombocytopenia. 15he primary end points of the first part of the trial demonstrated encouraging safety and superiority of eltrombopag in inducing PLT response (PLT-R) compared with placebo, making this drug a promising approach for the management of thrombocytopenia in low-risk MDS. 16We now report the predefined interim results on the entire sample enrolled in the EQOL-MDS trial with follow-up of at least 3 months (169 cases).

Trial Design
EQOL-MDS is an international, multicenter, randomized, single-blind, placebo-controlled, superiority trial (additional information on trial design is provided in the Data Supplement [online only]).
The trial Protocol (online only) was approved by an independent ethics committee at each participating institution, and all patients provided written informed consent.

Patients
Inclusion criteria were patients age 18 years and older with diagnosis of low or intermediate-1 International Prognostic Scoring System (IPSS) 2 risk MDS with stable PLT count (<30 3 10 3 /mm 3 without exceeding >200 Gi/L) confirmed by blinded central evaluation, and refractoriness or ineligibility to receive, or relapsed while receiving treatment with alternative medications.
Exclusion criteria were (1) prior chemoradiotherapy or previous treatment with TPO-R agonists; (2) peripheral monocytosis > 310 3 /mm 3 or leukocytosis ≥ 310 3 /mm 3 ; (3) marrow fibrosis with an inability to aspirate marrow; (4) Eastern Cooperative Oncology Group performance status 17 >3; (5) serum creatinine >2 times the upper limit of normal (ULN), AST or ALT >3 times the ULN or bilirubin >1.5 times the ULN; and (6) pre-existing cardiovascular disease or arrhythmia associated with an increased risk of thromboembolic event.Cases with >5% bone marrow blasts were excluded in France.Erythropoiesis-stimulating agents or granulocyte colony-stimulating factor was permitted during the trial, as per accepted standards.Additional information is provided in the Data Supplement.

Random Assignment and Masking
Participants were randomly assigned (2:1) to either eltrombopag or matching placebo as previously described. 15dditional information is provided in the Data Supplement.

Laboratory Assessments
Peripheral blood and bone marrow assessments were performed during screening before random assignment and preselected time points throughout the trial (Data Supplement).

Eltrombopag Administration
Oral eltrombopag or matching placebo was administered at an initial dose of 50 mg once daily, titrated in 50-mg increments every 2 weeks up to 300 mg to achieve a complete PLT-R, defined as a PLT count ≥100 3 10 3 /mm 3 without bleeding (Data Supplement).

Assessment of Quality of Life
Change in quality-of-life (QoL) scores were measured at baseline and at subsequent time points using the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire C30 questionnaire 18 and MDS-specific QOL-E questionnaire 19 (Data Supplement).

Eltrombopag Pharmacokinetic Analysis
Data on the methodology used to process plasma samples for eltrombopag pharmacokinetic analysis are provided in the Data Supplement.

Study End Points
The first part of the trial determined short-term efficacy and safety at 50% accrual, as previously described 15 ; the second part, reported in the present predefined interim analysis, evaluates the long-term response and safety (Data Supplement [Fig S1 ]).A PLT-R was defined as achieving the following increases in PLT count from baseline levels: for patients with baseline PLTs at least 20 3 10 3 /mm 3 , an increase of at least 30 3 10 3 /mm 3 from baseline; and for patients with baseline PLTs <20 3 10 3 /mm 3 , an increase of more than 20 3 10 3 /mm 3 and an increase of at least 100%, not because of PLT transfusions, in the absence of bleeding (Data Supplement).
Part 2 primary end points include duration of PLT-R and long-term safety and tolerability.Secondary end points include (1) difference in time to response (time from starting treatment to time of achievement PLT-R); (2) frequency of PLT transfusions during the treatment and follow-up periods; (3) duration of PLT transfusion independence; (4) incidence and severity of bleeding using the WHO Bleeding Scale 20 ; (5) progression-free survival (PFS), leukemia-free survival (LFS), and overall survival (OS); (6) changes in QoL scores; and (7) eltrombopag population pharmacokinetic parameters and plasma concentration data.

Statistical Analysis
Data are summarized as mean 6 standard deviation (normally distributed variables), median and IQR (non-normally distributed variables), or as percent frequency (binary/categorical variables), and between-arms comparisons were performed by independent t test, Mann-Whitney U test, or Pearson's chisquare test, as appropriate.
Efficacy and safety analyses were performed on the full analysis set, that is, randomized patients who had received at least one dose of eltrombopag, according to the modified intention-to-treat (ITT) principle.
The primary end point (PLT-R) of the 24-week trial (first part) was analyzed with the use of a logistic regression model that included the trial group as an independent variable and was also presented graphically by reverse Kaplan-Meier curves and compared by log-rank test.PLT-R, defined according to International Working Group 2006 criteria, 21 was assessed at each visit.A response (PLT-R) required an absence of bleeding, with the following increases in PLT count from baseline levels: for patients with baseline PLTs ≥20 3 10 3 /mm 3 , an increase of at least 30 3 10 3 /mm 3 from baseline; for patients with baseline PLTs <20 3 10 3 /mm 3 , the achievement of >20 3 10 3 /mm 3 and an increase of at least 100%, not because of PLT transfusions.
To account for differences in follow-up time between patients of the two study arms, we also performed a time-to-event analysis by a reverse Kaplan-Meier method.
The duration of PLT-R (primary end point of second part) was calculated from the time of PLT-R to the date of loss of PLT-R, defined as bleeding or PLT count <30 3 10 3 /mm 3 or last date in observation for those who did not lose the response.The time to the loss of PLT-R was investigated by Kaplan-Meier curves.
Safety outcomes (primary end point of both parts) were summarized descriptively.Between-treatment comparisons (adverse events [AEs], progression of MDS, AML evolution, and death) were conducted using a chi-square test.Additional information is provided in the Data Supplement.

Patient Characteristics
At data cutoff, March 3, 2022, of the 325 patients screened, 169 patients were randomly assigned eltrombopag (N 5 112) or placebo (N 5 57; Fig 1 and Table 1).Of these, 165 patients received at least one dose of study drug (111 in the eltrombopag arm and 54 in the placebo arm) and were considered in the modified ITT analysis.Baseline features were similar in the two arms (Table 1).Main MDSrelated concomitant treatments were erythropoiesis-stimulating agents, steroids (16.0% for both), and deferasirox (5.9%).
In a linear-mixed model, hemoglobin was found to modify the PLT-R to eltrombopag, with the between-treatment difference in PLT count over the trial period being closely related to hemoglobin levels (Fig 3B  In the placebo arm (observation period of 83-398 weeks), the six placebo responders maintained the PLT-R until study termination.One remains in the study after 76 months of response.
Additional results on PLT-R and reasons for early study termination are provided in the Data Supplement.

Safety
At the time of data cutoff, 15

Median Time to PLT-R
The median time to PLT-R was significantly earlier in the eltrombopag arm (2.1 weeks, 1-6 weeks) versus the placebo arm (54 weeks, 6-155 weeks; P < .001).

Incidence and Severity of Bleeding
Thirty-nine subjects had at least one clinically significant bleeding event (WHO bleeding score ≥ 2) during the entire study period: 17 (31.5%)cases in the placebo group (exposureadjusted event rate [EAER], 7.3 events per 100 patients/week) versus 22 eltrombopag cases (19.8%;EAER, 3.9 events per 100 patients/week; incidence rate ratio, 0.54; 95% CI, 0.38 to 0.75; P 5 .0002).The distribution of patients according to the maximum degree of bleeding is provided in Table 2.
Overall, the combined outcome AML evolution and/or disease progression occurred in 19
• Median LFS, combined outcome (AML, disease progression, and death), and OS were not reached in the whole group.

DISCUSSION
The occurrence of severe thrombocytopenia in patients with low-risk MDS still represents a challenging condition because of the important burden of related AEs and the lack of effective treatments.In most countries, only PLT transfusions are currently offered to prevent or treat bleeding in these subjects.Despite this unmet clinical need, few investigational products have been tested.Among them, TPO mimetics are the most attractive: however, though the use of romiplostim seemed promising in a randomized trial, 10 there is still concern regarding the risk of MDS progression/AML evolution and warning in the label of all TPO mimetics. 16ltrombopag has a different mechanism of action from romiplostim.Furthermore, a trilineage marrow response is observed in severe aplastic anemia. 22eliminary results of this trial already showed high rates of PLT-R in this challenging subset of patients. 15Our present predefined interim analysis report on the long-term efficacy and safety of eltrombopag.Our previous part-1 results included 90 patients with a follow-up of 6 months and a median follow-up of 11 weeks. 15The present part-2 interim results confirm our previous results in terms of obtaining a durable and stable PLT-R and extend them further where we included 169 randomly assigned patients with a follow-up of 60 months (a median follow-up of 25 weeks).
Importantly, the findings from the present trial show a favorable safety profile in terms of long-term AEs and AML evolution/disease progression.
While the rate of PLT-R was confirmed, a complete PLT-R (27.9%) was observed in the eltrombopag arm only.4][25] The achievement of PLT-R occurred significantly earlier in the eltrombopag patients with a higher median PLT count increment in responders compared with placebo responsive patients (P < .001).PLT-R to eltrombopag was durable, with 60% of patients still maintaining response at 5 years.Moreover, although only a fraction of patients were receiving platelet transfusions at study entry, the rate of patients achieving PLT transfusion independence was higher and, importantly, the rate of incidence of bleeding was lower in the eltrombopag arm than in the placebo arm, reaching statistical significance.
Our findings also corroborate with results from a recent multicenter retrospective study performed in France to evaluate the use of eltrombopag in patients with MDS (N 5 50) and chronic myelomonocytic leukemia (N 5 11). 26LT-R occurred in 47 (77%) patients, a higher rate than what we observed (42.3%), possibly attributed to differences in baseline clinical characteristics.
The dose of eltrombopag (50 mg once daily) to obtain PLT-R was the same as that used in our trial but we observed a substantially shorter median time to reach PLT-R (2.1 v 4.3 weeks) that may be due to the higher frequency of blood counts undertaken in our trial.Disease progression was observed in 16% of patients, almost identical to the rate that we observed in both eltrombopag and placebo arms (17%).
Similar to what is observed in aplastic anemia, a sizable fraction of patients receiving eltrombopag also achieved an erythroid/granulocytic response.This particular finding deserves further evaluation in larger cohorts of cytopenic patients with low-risk MDS.
As expected, patients in the eltrombopag arm had a higher rate of nonhematologic AEs and 18 patients necessitated permanent treatment discontinuation because of persistent drug-related AEs.Most grade 3-4 AEs occurred early during treatment, within the first 24 weeks, and were reversible upon drug discontinuation.However, no differences were reported between the two arms for deaths.
The most important safety evaluation, however, concerned the risk of the increase in blasts and MDS progression/AML evolution, which was shown during romiplostim treatment in a similar trial. 10In the present trial, long-term data are encouraging, with no transient increase in circulating peripheral myeloblasts and similar 5-year LFS, PFS, and CFS in the two arms.
Finally, it should be stressed that a relatively low daily dose of eltrombopag (50 mg once daily) was identified as the optimal dose at which patients more commonly achieved a consistent PLT-R, a finding also confirmed in the real-life setting, 26 with a lower risk of AEs compared with higher doses.Besides the important clinical benefit in terms of efficacy with associated low risk of AEs over the long term, the use of low-dose eltrombopag may also translate into significant reduction in health care costs.Although there is evidence of an immunomodulatory role and the ability to mobilize intracellular iron, 27,28 the biological mechanisms between eltrombopag and the immune system still need to be elucidated, particularly at low doses.On the basis of our long-term results, we may conclude that there is no major advantage in a dose increase beyond 150 mg, limited to selected cases at high risk of bleeding.
In conclusion, eltrombopag has an acceptable toxicity profile and is effective in raising and maintaining PLT count and reducing bleeding without any associated risk of MDS progression.

SUPPORT
Associazione QOL-ONE, a nonprofit organization, sponsored and funded the trial described in this report, and paid for the services of professional medical writers, who provided editorial assistance in refining the draft manuscript, and a statistician, who carried out the statistical data analyses.In addition, Associazione QOL-ONE contributed to the design of the trial; the collection, analysis, and interpretation of data; the writing of the report; and the decision to submit the report for publication.Novartis donated eltrombopag for evaluation in the trial and provided partial and unconditional funding to Associazione QOL-ONE, but played no other role in the design or conduction of the trial, the analysis or interpretation of the data, or the development or submission of this report.The authors received no payment, from either a pharmaceutical company or any other agency, to write this report.All the authors had full access to all the trial data, and were jointly responsible for the decision to submit the report for publication.

FigFIG 1 .
FIG 1. CONSORT diagram.Four patients (one in the eltrombopag arm and three in the control arm) withdrew from the trial before receiving the first dose of eltrombopag; thus, 165 patients (111 patients of the eltrombopag arm and 54 patients of the control arm) were considered in the modified ITT.The first patient was enrolled on June 13, 2011, and the last patient was enrolled on October 1, 2021.At the time of data extraction, 45 patients of 111 of the eltrombopag group and 17 patients of 54 of the placebo group entered the second part of the trial.AE, adverse event; ITT, intention-to-treat; MDS, myelodysplastic syndrome; PLT, platelet.

TABLE 1 .
Main Demographic and Clinical Characteristics of Patients 1, 2021.At the time of data extraction, 45 patients of 111 of the eltrombopag group and 17 patients of 54 of the placebo group entered the second part of the trial.AE, adverse event; ITT, intention-to-treat; MDS, myelodysplastic syndrome; PLT, platelet.Journal of Clinical Oncology ascopubs.org/journal/jco| Volume 41, Issue 28 | 4489 Eltrombopag for Low-Risk MDS With Thrombocytopenia

TABLE 1 .
Main Demographic and Clinical Characteristics of Patients (continued) NOTE.Data are expressed as mean 6 standard deviation, median and IQR, or absolute and percentage frequency, as appropriate.Differences between two arms are expressed as Mann-Whitney U test for continuous variable and Pearson chi-square or Fisher exact test for categorical variables, as appropriate.Abbreviations: EB-1, excess blasts-1; ESA, erythropoiesis-stimulating agents; GCSF, granulocyte colony-stimulating factor; IPSS, International Prognostic Scoring System; IPSS-R, revised International Prognostic Scoring System; MDS, myelodysplastic syndromes; MLD, multilineage dysplasia; PLT, platelet; RS, ring sideroblasts; SLD, single-lineage dysplasia.FIG 2. Reverse Kaplan-Meier curves of platelet response in both treatment groups.
Box and whisker plot of PLT count over time.The horizontal line in the middle of each box represents the median, while the top and bottom of each box represent the 75th and 25th percentiles, respectively.The whiskers above and below the box plot mark the 97.5th and 2.5th percentiles, respectively.(B)Hemoglobinasan effect modifier of PLT response (log-transformed data) to eltrombopag.Data represent the ratio (and 95% CI) over the study period of the geometric mean of PLT count in eltrombopag-treated patients to the geometric mean of PLT count in patients in the placebo group.The numbers reported in parentheses (B) represent the number of observations with a hemoglobin value over time falling in a specific hemoglobin interval.LMM, linear mixed models; PLT, platelet.stoppingrulewasnot reached.Eighteen eltrombopag patients required permanent treatment discontinuation because of severe liver or persistent grade 3-4 AEs after a median time from baseline of 13 weeks (IQR,6-20), whereas no placebo cases experienced such events..15).The incidence rate of PLT transfusion was on average 16.3 per 100 patients-week (95% CI, 15.4 to 17.2) in the active arm and 16.9 per 100 patients-week (95% CI, 15.5 to 18.4) in the control arm (P 5 .49).
unrelated deaths occurred: 11 (9.9%) in the eltrombopag arm for infection in five, cardiorespiratory failure in two, hemorrhage in one, and worsening of general condition in three; and four (7.4%) in the placebo arm for infection, heart failure, hemorrhage, and worsening of difference between arms (P 5 .868).Fifty patients in the eltrombopag arm (EAIR, 2.5; 95% CI, 1.9 to 3.4) and 11 in the placebo arm (EAIR, 1.6; 95% CI, 0.8 to 2.8) experienced grade 3-4 nonhematologic AEs (EAIR, 1.62; 95% CI, 0.83 to 3.46; P 5 .14;DataSupplement[TableS3]), with a significant difference between arms (P 5 .002)favoring placebo, but the 95% CI, 9.7 to 100.0) in the placebo arm, and did not reach statistical significance.Additional results are available in the Data Supplement([Figs S4 and S5]).The number needed to treat for a PLT-R of at least 4 weeks' duration was 3.2, and the number needed to harm for grade 3-4 AEs was 4.1.The number needed to observe AML evolution was 62. Thus, the likelihood of being helped or harmed, in terms of serious AEs versus efficacy, is 1.3 and AML evolution versus efficacy is 19.4.Therefore, eltrombopag treatment is 1.3-19.4times more likely to help in terms of PLT-R than to harm.

TABLE 2 .
Distribution of Patients According to the Maximum Degree of