Ruxolitinib Reduces Spleen Size in Patients With Myelofibrosis and Low-Platelet Counts

Ruxolitinib induced clinically meaningful reductions in both spleen size and symptoms for patients with myelofibrosis, including those with low platelet counts.

Ruxolitinib (Jakafi), the JAK1/2 inhibitor, induced significant spleen responses in patients with myelofibrosis (MF), both with general disease and in patients with low platelet counts, according to results from the phase 3B JUMP study (NCT01493414).

Overall, ruxolitinib induced clinically meaningful reductions in both the spleen size and symptoms for patients with MF, including those with platelet counts of <100 x 109/l.

“Findings from this study not only confirm findings from the COMFORT studies (NCT00952289, NCT00934544) but also provide additional evidence of the efficacy and safety of ruxolitinib in the treatment of patient groups that have not been as extensively studied as those evaluated in the COMFORT studies,” the study authors, led by Haifa Kathrin Al-Ali, MD, University Hospital of Halle, Germany, wrote. “However, it is important to interpret our findings with caution given the limitations associated with a single-arm study and the small number of patients in the intermediate-1-risk and low-platelet cohorts.”

This is the most extensive study in patients with MF, including the largest cohort to date of patients with MF treated with ruxolitinib. Ruxolitinib was approved as treatment of this patient population based on the data from the pivotal phase 3 COMFORT clinical trials. The agent was evaluated in the JUMP study to determine the efficacy and safety of this JAK inhibitor in patients with low platelet counts.

The primary end point of the study was safety and tolerability of ruxolitinib by the frequency, duration, and severity of adverse events (AEs). Secondary end points included the proportion of patients with a ≥50% reduction in palpable spleen length; patient-reported outcomes, progression-free survival (PFS); survival without transformation to acute myeloid leukemia (AML-free survival); and overall survival (OS).

Overall, 2233 patients were treated across 279 clinical sites in 26 countries, including Europe (n = 1831), Latin America (n = 190), North America (n = 53), and other countries (n = 159). The median age of patients was 67 years (range, 18-89), and the median time since initial diagnosis was 25.8 months (range, <.01-45.6). Investigators noted that 38.3% of patients had hemoglobin levels of <100 g/l at baseline and the mean spleen length was 13.3 cm. Most patients were classified as either intermediate-1 (37.4%) or intermediate-2 (33.8%) risk.

Patients with lower platelet counts had baseline characteristics similar to those with more advanced disease compared with patients with platelet counts of ≥100 x 109/l, including hemoglobin levels of <100 g/l (37.5% vs. 52.9%) and larger mean palpable spleen length (13.2 cm vs. 14.9 cm), respectively.

Most patients (57.5%) had completed treatment per the study protocol, and the primary reason for discontinuation was AEs (18.1%), disease progression (9.1%), death (4.5%), physician’s decision (4.2%), withdrawal of consent (3.5%), protocol deviation (1.2%), administrative problems (1.1%), and loss of follow-up (0.7%). Among the patients with low platelet counts, 43.5% completed the study protocol, and the primary reasons for discontinuation of treatment were AEs (27.5%), disease progression (12.3%), death (5.8%), physician decision (5.1%), loss of follow-up (2.9%), and withdrawal of consent (2.2%).

The median duration of follow-up was 13.8 months (range, <0.1-60.6), and the median exposure to ruxolitinib was 12.4 months (range, <0.1-59.7). Approximately 51% of patients had >1 year of exposure to ruxolitinib, 30% had >2 years, and 13% had >3 years. The median duration of exposure was lower in the low-platelet cohort, which was about 83 months (range, 0.5-41.4).

The mean dose of daily ruxolitinib was 28.7 mg in the overall population versus 13.2 mg in the low-platelet cohort. Most patients had started treatment at 5, 15, or 20 mg twice daily. The majority of patients (67.4%) had dose reductions, and 17.2% of patients had dose increased. Dose interruptions were required in 27.2% of patients. In the low-platelet cohort, 58.7% required dose reductions, 38.4% had dose increased, and 32.6% had ≥1 dose interruption.

Among patients in the low-platelet cohort receiving a starting dose of 5 mg, 54.6% had dose reductions, 29.5% had interruptions, while 76.9% of patients starting at higher doses of ruxolitinib had to reduce the dose and 46.2% had dose interruptions.

The most common hematologic AEs, of all grades and grade 3/4, respectively, included anemia (59.5% vs. 34.8%) and thrombocytopenia (53.5% vs. 19.3%). The most common hematologic AEs, of all grades and grade 3/4, respectively, in the low-platelet cohort, included thrombocytopenia (73.2% vs. 54.3%) and anemia (52.9% vs. 35.5%).

Because fewer patients discontinued treatment due to anemia (2.0%) and thrombocytopenia (3.4%), the data suggest that these events were manageable in most patients. The median hemoglobin levels decreased from baseline to a nadir at 8 (95.0 g/l) and 12 weeks (94.0 g/l), and the number of transfusion-dependent patients at baseline (7.1%) was the highest during the first 12 weeks and decreased over time.

The median platelet levels decreased from baseline (254 x 109/l) during the first 4 weeks of treatment to a nadir (153 x 109/l), which remained stable over time. Among patients in the low-platelet cohort, the median platelet levels and hemoglobin levels decreased significantly from baseline. Overall, 10.1% of patients in this cohort discontinued treatment due to thrombocytopenia and 2.2% due to anemia.

Among those in the low-platelet cohort who discontinued treatment due to thrombocytopenia, the median platelet count at the time of discontinuation was 32 x 109/l (range, 8-55 x 109/l). Additionally, 5 patients discontinued treatment because they reached the study protocol of <25 x 109/l, and 9 discontinued for platelet counts of 25 to 47 x 109/l without reaching the stopping rule.

Among the 14 patients with low platelets at baseline who discontinued due to thrombocytopenia, 2 patients had a grade 1/2 hemorrhage and 1 had a grade 3/4 upper gastrointestinal hemorrhage, which occurred after the treatment discontinuation. In total, 10 patients (7.2%) had grade 1/2 hemorrhages, and 6 patients had grade 3/4 hemorrhages, which were defined by investigators as unrelated to ruxolitinib. The grade 3 hemorrhages included 1 bleeding esophageal varices, 1 upper gastrointestinal hemorrhage, and 1 hemorrhage after a molar extraction. Grade 4 hemorrhages included 1 intercoastal left artery hemorrhage, 1 gastric hemorrhage, and 1 intestinal hemorrhage.

The most common non-hematologic AEs were primarily grade 1/2 in the overall and low-platelet cohorts and included pyrexia (16.0% vs. 17.4%), asthenia (15.4% vs. 13.0%), diarrhea (12.5% vs. 10.1%), and fatigue (10.0% vs. 5.1%), respectively. The rates of grade 3/4 AEs were low, with the exception of pneumonia, which occurred in 4.7% in the overall population and 5.8% in the low-platelet cohort and led to ruxolitinib discontinuation in 0.4% of patients overall. Rates of infection were also low, leading to treatment discontinuation in 2.6% of the overall patient population.

Overall, 6.1% of patients reported second malignancies, which included non-melanoma skin carcinoma (2.7%), lung neoplasm (0.2%), prostate cancer (0.2%), and lymphoma (0.2%). Patients who had developed a second malignancy discontinued treatment and were not followed by the investigator thereafter. Among those who developed a second malignancy, a higher proportion were male (61.3%), has post-polycythemia vera MF (34.2%), prior hydroxycarbamide exposure (65.0%), and elevated neutrophil counts of <25 x 109/l (14.6%).

Serious AEs included pneumonia (5.5%), anemia (4.2%), pyrexia (3.5%), cardiac failure (1.9%), dyspnea (1.6%), sepsis (1.4%), abdominal pain (1.3%), respiratory failure (1.2%), thrombocytopenia (1.1%), and urinary tract infection (1.0%).

Overall, 56.5% of patients with splenomegaly achieved a ≥50% reduction from baseline in palpable spleen length at week 24 compared with 61.4% at week 48 and 66.5% at week 96, and in the low-platelet cohort, this was achieved in 38.4% at week 24 and 31.9% at week 48. Additionally, 25% to <50% reductions were observed in 23.3% of patients at week 24, 18.9% at week 48, and 14.3% at week 96, and this was achieved in 64.0% at week 24 and 57.4% at week 48 in the low-platelet cohort. At every assessment, at least two-thirds of patients had a ≥25% reduction from baseline in palpable spleen length.

A ≥50% reduction from baseline spleen length at any time during the study occurred in most patients (71.7%), including 43.8% in the low-platelet cohort. A total of 24.7% of patients had resolution of splenomegaly, and the median time to first ≥50% reduction in the spleen length was 5.8 weeks (range, 2.6-236.1) in the overall population versus 8.0 weeks (range, 3.3-84.6) in the low-platelet cohort.

The Kaplan-Meier estimated probability of maintaining a ≥50% reduction in spleen length at 48 and 96 weeks was 88% (95% CI, 86%-90%) and 78% (95% CI, 75%-81%) in the overall population versus 75% (95% CI, 58%-86%) and 54% (95% CI< 34%-71%) in the low-platelet cohort. Most patients with non-palpable spleen at baseline remained non-palpable throughout the study, where 94% were non-palpable at 24 weeks and 13.3% by the end-of-treatment visit.

The estimated OS was 95% (95% CI, 92%-95%) at 48 weeks and 87% (95% CI, 85%-89%) at 96 weeks. Overall, 9.2% of patients died on treatment of up to 28 days after the end-of-treatment visit, and the primary causes of death included MF (n = 38), pneumonia (n = 15), septic shock (n = 14), cardiac arrest (n = 13), cardiac failure (n = 12), sepsis (n = 11), respiratory failure (n = 9), multiple organ dysfunction syndrome (n = 9), AML (n = 7), cardiorespiratory arrest (n = 7), pulmonary embolism (n = 6), acute respiratory distress syndrome (n = 4), myocardial infarction (n = 4), cardiogenic shock (n = 4), general physical health deterioration (n = 3), and acute kidney injury (n = 3). The cause of death was undetermined in 13 patients, and all other causes were reported in 1 or 2 patients each. A total of 18 patients died of second malignancies, which included 12 patients who died of leukemia.

Among patients in the low-platelet cohort, 3 died due to septic shock, 2 to multiple organ dysfunction syndrome, and 2 to progressive MF, while all other causes occurred in 1 patient each. Four patients in this cohort developed AML.

The estimated AML-free survival probability was 85% (95% CI, 83%-87%) at 96 weeks. The estimated PFS rate was 81% (95% CI, 78%-83%) at 96 weeks. Investigators noted that patients with higher-risk disease had worse survival and a lower probability of AML-free survival and PFS.

In the global, single-arm, open-label, expanded access JUMP study, patients with a confirmed diagnosis of MF who were classified as high, intermediate-1, or interimediate-2 risk were evaluated. Starting doses of ruxolitinib were based on platelet counts at baseline, in which 5 mg was administered twice daily to patients with platelets ≥50 to <100 x 109/l, 15 mg twice daily to patients with platelets 100–200 x 109/l, or 20 mg twice daily to patients with platelets >200 x 109/l. Some patients received a 10 mg dose or another dose, although this was not per the protocol.

Patients received treatment with ruxolitinib for up to 24 months or until disease progression, unacceptable toxicity, death, or study withdrawal. Investigators followed patients for up to 28 days after the end-of-treatment visit.


Al-Ali HK, Griesshammer M, Foltz L, et al. Primary analysis of JUMP, a phase 3b, expanded-access study evaluating the safety and efficacy of ruxolitinib in patients with myelofibrosis, including those with low platelet counts. Br J Haematol. 189: 888-903. doi: 10.1111/bjh.16462