Comparing the safety and efficacy of ruxolitinib in patients with Dynamic International Prognostic Scoring System low‐, intermediate‐1‐, intermediate‐2‐, and high‐risk myelofibrosis in JUMP, a Phase 3b, expanded‐access study

Abstract Ruxolitinib, a potent Janus kinase 1/2 inhibitor, has demonstrated durable improvements in patients with myelofibrosis. In this analysis of the Phase 3b JUMP study, which included patients aged ≥18 years with a diagnosis of primary or secondary myelofibrosis, we assessed the safety and efficacy of ruxolitinib in patients stratified by Dynamic International Prognostic Scoring System (DIPSS) risk categories. Baseline characteristic data were available to assess DIPSS status for 1844 of the 2233 enrolled patients; 60, 835, 755, and 194 in the low‐, intermediate (Int)‐1‐, Int‐2‐, and high‐risk groups, respectively. Ruxolitinib was generally well tolerated across all risk groups, with an adverse‐event (AE) profile consistent with previous reports. The most common hematologic AEs were thrombocytopenia and anemia, with highest rates of Grade ≥3 events in high‐risk patients. Approximately, 73% of patients experienced ≥50% reductions in palpable spleen length at any point in the ≤24‐month treatment period, with highest rates in lower‐risk categories (low, 82.1%; Int‐1, 79.3%; Int‐2, 67.1%; high risk, 61.6%). Median time to spleen length reduction was 5.1 weeks and was shortest in lower‐risk patients. Across measures, 40%–57% of patients showed clinically meaningful symptom improvements, which were observed from 4 weeks after treatment initiation and maintained throughout the study. Overall survival (OS) was 92% at Week 72 and 75% at Week 240 (4.6 years). Median OS was longer for Int‐2‐risk than high‐risk patients (253.6 vs. 147.3 weeks), but not evaluable in low‐/Int‐1‐risk patients. By Week 240, progression‐free survival (PFS) and leukemia‐free survival (LFS) rates were higher in lower‐risk patients (PFS: low, 90%; Int‐1, 82%; Int‐2, 46%; high risk, 15%; LFS: low, 92%; Int‐1, 86%; Int‐2, 58%; high risk, 19%). Clinical benefit was seen across risk groups, with more rapid improvements in lower risk patients. Overall, this analysis indicates that ruxolitinib benefits lower‐risk DIPSS patients in addition to higher risk.


| INTRODUCTION
Ruxolitinib is a potent Janus kinase (JAK) 1/JAK2 inhibitor that has demonstrated durable improvements in splenomegaly, myelofibrosis (MF)-related symptoms, and quality-of-life measures in patients with International Prognostic Scoring System (IPSS) 1

intermediate (Int)-2-
and high-risk MF in the Phase 3 COMFORT studies. 2,3 In COMFORT I and II, 42% and 28% of patients in the ruxolitinib groups, respectively, achieved a reduction in spleen volume of 35% or more at 24 weeks, compared with 0.7% and 0% in the placebo groups, respectively. 2,3 In post hoc analyses of the COMFORT studies, ruxolitinib was associated with improved survival rates versus placebo 2 and best available therapy 3,4 and also when compared with historical matched cohorts. 5 In primary MF, the IPSS is used for initial diagnosis of patients. 1 In this system, increasing numbers of five independent predictors of inferior survival (age >65 years, hemoglobin <10 g/dl, leukocyte count >25 �10 9 /L, circulating blasts more than 1%, and the presence of constitutional symptoms) define risk categories of low-, Int-1-, Int-2-, and high-risk disease. Subsequently, the Dynamic International Prognostic Scoring System (DIPSS) model was developed and utilizes the same prognostic variables as in the IPSS, but with differences in scoring (e.g., DIPSS assigns 2 rather than 1 point for hemoglobin <10 g/dl). 6 The DIPSS can be applied at any time during the disease course and includes risk categories of low (0 adverse points), Int-1 NCT01493414). 7 In the current analysis, we explored the safety and efficacy of ruxolitinib in patients in the JUMP study who were stratified according to the DIPSS. As only the IPSS was available when the JUMP and COMFORT study protocols were developed, this subgroup analysis was performed to give insight into any safety or efficacy differences of ruxolitinib across DIPSS risk categories and to allow comparisons with future studies using these criteria. By considering this time-dependent risk classification (DIPSS) that more accurately reflects the dynamic enrollment period of the JUMP trial, this post hoc analysis allows the impact of disease-risk group on outcomes in the JUMP trial to be appreciated in a manner that more closely reflects clinical practice, where patients can receive treatment both at the time of diagnosis and afterward.

| METHODS AND PATIENTS
Full details of the study methodology have been published previously. 7,8 In brief, patients aged ≥18 years with a diagnosis of primary or secondary MF by World Health Organization and International Working Group for Myeloproliferative Neoplasms Research and Treatment (IWG-MRT) criteria 9,10 and classified by the treating investigator as high, Int-2, or Int-1 risk using IPSS criteria, were eligible for enrollment. Patients were stratified by DIPSS based on baseline patient characteristics into low-, Int-1-, Int-2-, and high-risk categories. Further details of inclusion criteria have also been published previously. 7,9 Patients were treated with ruxolitinib for up to 24 months after the last patient's first visit, or until the drug became commercially available, with doses titrated from starting doses based on baseline platelet counts, from 5 mg to a maximum dose of 25 mg twice daily (b.i.d.), until discontinuation criteria were met (i.e., disease progression, unacceptable toxicity, death, discontinuation from the study for any other reason, physician decision, withdrawal of informed consent) or completed treatment per protocol, whichever occurred first (for further details, see Al-Ali et al. 8 ). The median duration of exposure to ruxolitinib was 12.4 months (range, <0.1-59.7 months). 7 The study was conducted in accordance with the Declaration of Helsinki and principles of Good Clinical Practice.
Written informed consent was obtained from all participants prior to enrollment, and the protocol and its amendments were approved by the institutional review boards of the respective institutions prior to study commencement. and serious AEs, hemoglobin levels, and platelet counts). Response on the FACT-Lym TS was defined as the minimally important difference (i.e., an 11.2-point improvement from baseline). 11 On the FACIT-Fatigue scale, response was defined as the minimally important difference of a three-point improvement from baseline. 12

| RESULTS
Baseline patient disease characteristics to determine DIPSS risk status were available in 1844 of 2233 enrolled patients (low, n = 60; Int-1, n = 835; Int-2, n = 755; and high, n = 194; Table 1). Sixty lowrisk patients were enrolled in violation of the protocol, likely due to differences between investigator-assessed risk and calculated risk or timing of assessment for inclusion (screening vs. baseline). These patients have been included in the analysis. IPSS risk statuses at baseline were low, n = 1; Int-1, n = 288; Int-2, n = 355; high, n = 255; and missing, n = 945. Patients with higher-risk MF (Int-2/high risk) according to DIPSS were older, had lower hemoglobin levels, and had higher circulating blast counts (Table 1).
Most patients (72.6%) experienced ≥50% reduction from baseline in palpable spleen length at any postbaseline visit; the proportion was highest in low-risk (82.1%) and Int-1-risk (79.3%) patients compared with Int-2-risk (67.1%) and high-risk (61.6%) patients. By Week 72, 78.3%, 67.6%, 48.4%, and 51.5% of low-, Int-1-, Int-2-, and high-risk patients, respectively, showed ≥50% reduction from baseline in palpable spleen length ( Clinically meaningful symptom improvements were seen as early as 4 weeks after treatment initiation and were maintained throughout the study. In the FACT-Lym TS, 17%-55% of patients achieved a response at each time point. On the FACIT-Fatigue Scale, 43%-57% of patients achieved a response at each time point. A higher proportion of high-risk patients achieved a response in both the FACT-Lym TS and the FACIT-Fatigue Scale versus low-, Int-1-, and Int-2-risk patients. This was expected given that high-risk patients had a higher symptom burden at baseline; however, symptom burden in high-risk patients at Week 48 continued to be higher than in lower-risk patients.
Overall, 191 deaths occurred during the study treatment period.
The most common primary causes of death across the groups were MF (n = 35), pneumonia (n = 12), sepsis (n = 12), cardiac arrest (n = 11), and septic shock (n = 10). Median OS was longer for Int-2- The most common hematologic AEs across all DIPSS categories were anemia and thrombocytopenia, with the highest rates of Grade ≥3 anemia and thrombocytopenia seen in high-risk patients (Table 3).
Across the DIPSS risk groups, median hemoglobin levels decreased from baseline (106 g/L) to a nadir from Weeks 4 to 24 (94-99 g/L across the four time points), but they increased to near-baseline levels after Week 36 (Figure 2A (Table 3).

| DISCUSSION
The JUMP study included the largest cohort of patients with MF treated with ruxolitinib to date, including patients with low-and Int-   19 Similarly, a report of 70 Int-1-risk patients treated with ruxolitinib, according to standard clinical practice, showed rates of spleen and symptom response of 55% and 80%, respectively; however, rates of Grade 3 anemia were higher and rates of thrombocytopenia were lower than those observed in the current analysis (40.6% and 2.9%, respectively). 20 Finally, a Phase 2 study assessing alternative dosing regimens of ruxolitinib that included a large

| CONCLUSION
Overall, the findings from this analysis suggest that ruxolitinib is well tolerated in DIPSS lower risk patients with MF and may be of benefit to this population in addition to the previously demonstrated benefits in higher risk patients.

ETHICS STATEMENT
The study was approved by the institutional review board at each participating institution and conducted in accordance with applicable local regulations and the principles of the Declaration of Helsinki. All patients provided written informed consent before entry into the study.

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1002/hon.2898.