Ruxolitinib Reduces Hematocrit Levels in 6-Month Analysis of Patients With Polycythemia Vera

Steffen Koschmieder

Ruxolitinib (Jakafi) led to significant reductions in hematocrit levels and the number of yearly phlebectomies among patients with polycythemia vera (PV) in a prespecified futility analysis of the phase 2b RuxoBEAT trial (NCT02577926), which was published in Annals of Hematology.¹

Compared with baseline levels, after 6 months of ruxolitinib treatment, patients with PV experienced a reduction in median hematocrit levels, from 46% (range, 34.6%-61%; interquartile range [IQR], 42.98%-47.08%) to 41.45% (range, 32%-52%; IQR, 37.25%-44.95%; P = .000957).

“Treatment with ruxolitinib in untreated patients with PV is feasible, well tolerated, and efficient,” lead study author Steffen Koschmieder, of Aachen University in Germany, and colleagues wrote in the publication.

Patients with PV have decreased survival compared with age-matched controls, mainly because of adverse effects (AEs) such as thromboembolic complications, progression to post-PV myelofibrosis, and acute leukemia. The most commonly reported symptoms of PV include fatigue, insomnia, numbness, itching/pruritus, sad mood, early satiety, and concentration problems. Many of these symptoms persist regardless of complete hematologic remission status.²

RuxoBEAT is a multicenter, open-label trial with a target enrollment of 380 patients with PV (n = 190) or essential thrombocythemia (n = 190).¹ This futility analysis aimed to evaluate the early efficacy and tolerability of ruxolitinib in 28 patients with PV who had received no prior therapies, although 6 weeks of hydroxyurea, anagrelide, or interferon therapy was allowed.

Patients 18 years and older were eligible for the PV cohort of this trial if they had PV per the World Health Organization 2008 classification and an ECOG performance status of 2 or less. Patients needed to have high-risk PV or PV with indication for cytoreductive therapy because of progressive myeloproliferation, defined as fulfilling 1 of the following: over 60 years of age; previous thromboembolism or thrombosis; platelet count over 1500 x 10⁹/L; poor phlebotomy tolerance or requiring frequent phlebotomy; progressive or symptomatic splenomegaly; severe disease-related symptoms; or progressive leukocytosis with leukocyte counts over 20 x 10⁹/L.

Patients were excluded if, within the 3 months prior to enrollment, they needed prolonged oral corticosteroid use at a dose of over 20 mg per day for over 1 month or had active splanchnic vein thrombosis.

Patients with PV were randomized 1:1 to receive ruxolitinib or best available therapy (BAT). Crossover from BAT to ruxolitinib was permitted after 6 months in eligible patients who did not achieve complete remission but had sufficient hematologic reserves, defined as platelet counts of at least 140 x 10⁹/L, hemoglobin of at least 10.0 g/dL, and an absolute neutrophil count of at least 1.5 x 10⁹/L.

In the ruxolitinib arm, patients received a starting dose of 10 mg twice daily. Upon insufficient efficacy, this dose could be increased to 20 mg twice daily.

In the BAT arm, patients received monotherapies of common agents such as hydroxyurea, interferon alpha, and anagrelide. BAT therapy changes were allowed, and the patients who changed BAT therapies during the study were censored for primary end point analysis.

This trial’s primary end point was the rate of complete clinic-hematologic response at 6 months. Key secondary end points were the use of phlebotomy, defined as the number of phlebotomies during the year prior to baseline and for the first 6 months of therapy, calculated per year; spleen size; patient-reported outcomes on myeloproliferative neoplasm (MPN)–associated symptoms using the MPN-SAF-TSS form; and overall survival.

The MPN-SAF-TSS form asked patients to score their outcomes regarding concentration, early satiety, itching, inactivity, bone pain, night sweats, fever, weight loss, and abdominal discomfort from 0 (absent) to 10 (worst imaginable) in a linear analog self-assessment scale. The German version of this form used by some patients, which included 3 additional microvascular disturbance symptoms, added a question about overall quality of life, and combined the questions about fatigue and inactivity.

The median time from the first diagnosis of PV to the initiation of RuxoBEAT trial treatment was 7.6 months (IQR, 1.5-29.2 weeks). The most common reasons for PV treatment were age over 60 years (n = 13) and previously reported thromboembolism or thrombosis (n = 13).

At data lock, 28 patients had received ruxolitinib for at least 6 months, with a median time on treatment of 173 days (range, 151-189) during the first 6 months of treatment. Per the futility analysis protocol, the 6-month efficacy variables were compared with the baseline/screening variables.

Although all 28 patients began the study treatment at 10 mg twice daily, within the first 6 months, 7 dose reductions occurred in 6 patients, and 11 dose escalations were reported. None of the 28 patients randomized to the PV arm discontinued the study within the first 6 months, and no protocol violations were reported.

Six months of ruxolitinib treatment reduced the median number of yearly phlebotomies from 4.0 (IQR, 1.00-6.00) to 0 (missing 3; IQR, 0-2.0; P = .000493). In addition, ruxolitinib decreased the median JAK2 V617F allele burden from 44% (missing 5; range, 16%-93%) to 34% (missing 5; range, 9%-91%; P = .000973).

Ruxolitinib treatment decreased the percentage of physician-assessed pruritis and night sweats, from 36% to 14% (P = .07) and 29% to 11% (P = .063), respectively. The MPN-SAF survey results showed a decrease in median patient-reported outcome points for pruritis with ruxolitinib, from 2 (n = 26; missing = 2; range, 0-10; IQR, 0-5) to 1 (range, 0-5; IQR, 0-2; P = .006). Treatment with ruxolitinib also trended toward a reduction in median patient-reported outcome points for night sweats, from 1.5 (missing = 2; range, 0-10; IQR, 0-5.25) to 0 (range, 0-7; IQR, 0-2.75; P = .101).

Ruxolitinib treatment did not change the median patient-reported outcome points for bone pain, which were 0 (missing = 2; range, 0-10; IQR, 0-4) at baseline and 1 (range, 0-9; IQR, 0-2; P = .343) after 6 months. However, the median MPN symptom score, defined as the mean of the point values from all symptom-oriented questions in the MPN-SAF survey, decreased from 1.5 (missing = 2; range, 0.09-7; IQR, 0.87-3.0) to 1.14 (range, 0-6; IQR, 0.48-2.55) after 6 months of ruxolitinib (P = .061).

During the first 6 months of treatment, patients experienced a median of 3.0 AEs. In total, 109 AEs occurred across 24 patients in the first 6 months of therapy, 11% (n = 12) grade 3, 28% (n = 31) grade 2, and 61% (n = 66) grade 1. No grade 4 or 5 AEs occurred. Eleven patients had at least 1 grade 3 AE, 8 patients had only grades 1 and 2 AEs, and 5 patients had only grade 1 AEs. In total, 68.8% of all AEs occurred during the first 3 months of treatment.

The most frequent AEs were general disorders, including fatigue; respiratory disorders; dermatologic disorders; renal and urinary tract disorders; gastrointestinal disorders; nervous system disorders; cardiovascular disorders; infections; and musculoskeletal, connective tissue, or metabolic disorders. The grade 3 AEs were liver enzyme increase (n = 2), compression fracture of the spine with spinal canal stenosis (n = 1), creatinine phosphokinase elevation (n = 1), suspected coronary syndrome (n = 1), urinary retention (n = 1), herpes zoster (n = 1), hot flashes (n = 1), choroidal melanoma (n = 1), macula edema (n = 1), diabetic retinopathy (n = 1), and hypertension (n = 1).

Of the 109 reported AEs, the treating physicians determined 50 to be unrelated to the study treatment. Of the remaining AEs, relative to ruxolitinib treatment, 11 were related, 3 were likely related, 35 were possibly related, and 10 were not likely related.

Two patients interrupted study treatment because of AEs, 1 from irritability and 1 from back pain. Both patients were re-exposed and subsequently remained on study treatment with no dose reductions. AEs caused dose reductions in 4 patients, 1 each from transaminase elevation, hypoglycemia, thoracic pain, and headache.

Five patients experienced 8 severe AEs. One serious AE, herpes zoster, was clearly related to ruxolitinib. Another, thoracic pain, was possibly related to ruxolitinib, and the treating physicians determined all the other serious AEs to be unrelated.

“Our results show that ruxolitinib is effective in reducing hematocrit levels, thereby potentially decreasing to risk for thromboembolic events and that this was possible without phlebotomy support,” the study authors concluded.

The RuxoBEAT trial is currently ongoing.

References

1. Koschmieder S, Isfort S, Wolf D, et al. Efficacy and safety of ruxolitinib in patients with newly-diagnosed polycythemia vera: futility analysis of the RuxoBEAT clinical trial of the GSG-MPN study group. Ann Hematol. 2023;102(2):349-358. doi:10.1007/s00277-022-05080-7
2. Grunwald MR, Burke JM, Kuter DJ, et al. Symptom burden and blood counts in patients with polycythemia vera in the United States: an analysis from the REVEAL study. Clin Lymphoma Myeloma Leuk. 2019;19(9):579-584.e1. doi:10.1016/j.clml.2019.06.001