Sustained‐release ruxolitinib: Findings from a phase 1 study in healthy subjects and a phase 2 study in patients with myelofibrosis

Abstract Ruxolitinib is the first Janus kinase (JAK) inhibitor approved for the treatment of myelofibrosis, where its efficacy is often associated with cytopenia. It is possible that the severity of cytopenia is in part driven by C max. A once‐daily sustained‐release (SR) formulation of ruxolitinib was therefore developed to decrease the C max/C min ratio relative to twice‐daily immediate‐release (IR) ruxolitinib. An SR formulation was identified based on pharmacokinetic evaluation in a phase 1 study in healthy adults (N = 9). This was followed by an open‐label phase 2 study in patients with myelofibrosis (N = 41). Ruxolitinib SR treatment was well tolerated with blood cell counts relatively unchanged through week 16. In terms of efficacy, 7 patients (17.1%) had clinical improvement and 33 (80.5%) had stable disease. While this study has raised the possibility of an increased therapeutic index for ruxolitinib via an SR formulation, further studies are required to validate the hypothesis.


| Phase 1 study: healthy subjects
Details about enrolled subjects, study design, study endpoints, and statistical analyses for the phase 1 study of healthy subjects can be found in the Supplemental Appendix.

| Patients
All enrolled participants were adults (aged ≥18 years) diagnosed with PMF, PPV-MF, or PET-MF for which treatment was indicated as per physician assessment. Study participants had a life expectancy ≥6 months and a spleen length ≥5 cm below the costal margin (determined by palpation).

| Study design and dosing
INCB 18424-260 was a single-arm, open-label phase 2 trial of ruxolitinib SR tablets QD. Study visits occurred at screening; baseline; day 1; weeks 2, 4, 8, 12, 16, 20, and 24; and every 12 weeks thereafter. All patients began treatment with 25 mg ruxolitinib SR QD. After 8 or 12 weeks, the dose level of ruxolitinib could be titrated to 50 mg SR QD for inadequate efficacy. To address potential toxicity arising from the increased dose, the protocol was amended to include an optional titration to ruxolitinib 25 mg SR alternating with 50 mg SR every other day (QOD), depending on platelet count; this amendment occurred after the study began (patients entering the study before the amendment did not have this option). Efficacy was considered inadequate if patients had <40% reduction from baseline in palpable spleen length at the week 8 or 12 study visit.
At week 16, all patients transitioned to ruxolitinib IR BID, with the starting dose depending on platelet count: Patients with a platelet count ≥200 × 10 9 /L initiated 20 mg IR BID, those with a platelet count 100 to <200 × 10 9 /L initiated 15 mg IR BID, and those with a platelet count 75 to <100 × 10 9 /L initiated 10 mg IR BID. All patients with platelet counts between 50 and <75 × 10 9 /L were subject to a mandatory dose reduction to ruxolitinib IR 5 mg BID. Further treatment with ruxolitinib was withheld from patients with platelet counts <50 × 10 9 /L. Doses were restarted or increased after platelet counts or absolute neutrophil count (ANC) levels recovered to acceptable levels. Patients had the option to remain on ruxolitinib IR treatment until ruxolitinib IR tablets became commercially available or until the last patient completed 36 weeks of treatment, whichever occurred earlier.

| Study endpoints
The primary study endpoints were safety/tolerability and overall response (OR). Safety/tolerability assessments included monitoring adverse events (AEs), vital signs, and clinical laboratory data. Changes

| Statistical analyses
Descriptive summaries were included for continuous and categorical variables. Unless otherwise stated, all CIs were 2-sided 95% CIs, unadjusted for multiplicity. The safety population was used for all safety analyses; the intent-to-treat population was used for efficacy analyses.
Pharmacokinetic data were analyzed by population pharmacokinetic analysis using pharmacokinetic-evaluable subjects.

| Subject disposition and demographics
Nine healthy adults (6 men and 3 women) were enrolled in the phase 1 study and received 25 mg ruxolitinib IR. Eight of the 9 subjects remained on-study to receive treatment with 2 different ruxolitinib SR formulations (SR-1 and SR-2). The median (range) age of study participants was 27 (18-53) years. Eight subjects (88.9%) were white; 1 (11.1%) was African American.

| Safety
No deaths or serious AEs (SAEs) occurred during this study. Elevated blood creatine phosphokinase led to study withdrawal in 1 subject after administration of ruxolitinib IR. This treatment-emergent AE (TEAE) was classified as moderate in intensity and was considered unrelated to study medication. Overall, safety was similar to that previously observed with ruxolitinib IR in healthy subjects. 9

| Pharmacokinetics
Plasma concentrations of ruxolitinib IR and SR over time after administration in fasted subjects are presented in Figure 1. The mean C max was 1100 nM for the IR formulation, compared with 333 nM for SR-1 and 394 nM for SR-2 (Table 1). These values represent reductions in C max of 69.7% and 64.2%, respectively, versus the IR formulation.
The t ½ of ruxolitinib approximately doubled for both SR formulations compared with the IR formulation ( Table 1). The ratio of C max to plasma concentration at 12 hours after dosing (C 12h ) decreased from 40 for ruxolitinib IR to 3.0 for ruxolitinib SR-1 and 4.7 for SR-2. Compared with ruxolitinib IR, the relative bioavailability of ruxolitinib SR-1 and SR-2 was 76% and 87%, respectively.
Pharmacokinetic simulation indicated that ruxolitinib SR-1 and SR-2 could provide higher steady-state mean C min values compared with ruxolitinib IR (6.5-fold and 11.4-fold, respectively) and lower steadystate mean C max values (69% and 61%). The SR-2 formulation was ultimately selected for further development because of a slightly higher relative bioavailability compared with the SR-1 formulation (Table 1).

| Patient disposition and characteristics
For weeks 1 to 16 of the phase 2 study, patients received 25 mg ruxolitinib SR QD; at week 16, all patients transitioned to 25 mg ruxolitinib IR BID. Of the 41 patients enrolled, 39 (95.1%) completed through week 16, and 24 (58.5%) completed through week 24. Reasons for study withdrawal before week 16 included AEs (n = 1) and transferring to treatment with a commercial product (n = 1). Reasons for withdrawal from weeks 16 to 24 included transferring to treatment with a commercial product (n = 11), consent withdrawn (n = 2), and AEs (n = 2).
Baseline demographics, disease characteristics, and laboratory values are presented in Table 2. Forty-one adult patients (median For most patients, platelet counts were within normal limits at baseline. The median hemoglobin level of patients was below the lower limit of normal (130 × 10 9 /L for men and 120 × 10 9 /L for women); 37 (90.2%) had decreased hemoglobin of at least grade 1 at baseline, and 15 (36.6%) had grade 2 low hemoglobin levels (ie, <100 × 10 9 /L). The median ANC was above the upper limit of normal (7.9 × 10 9 /L).
The median levels of platelets, hemoglobin, neutrophils, and leukocytes over time are shown in Figure 2A.

| Week 16 efficacy
Efficacy endpoints at week 16 are summarized in

| Pharmacokinetics
At weeks 4 and 12, the median t max after the first dose of ruxolitinib was 2.0 hours (Table 4)

| DISCUSSION
Pharmacokinetic simulations based on the single-dose phase 1 study results in healthy subjects demonstrated higher steady-state mean C min values and lower steady-state mean C max values with ruxolitinib SR formulations compared with ruxolitinib IR. In fasted subjects, ruxolitinib SR provided extended absorption of ruxolitinib, with lower maximal exposure compared with ruxolitinib IR without compromising relative oral bioavailability. The ruxolitinib SR-2 formulation was selected for further development because of a slightly higher relative bioavailability of 87% compared with SR-1.
In the phase 2 study of patients with MF, the pharmacokinetic characteristics of ruxolitinib SR were similar to those observed in the phase 1 study of healthy subjects. Ruxolitinib SR clearance and volume of distribution were lower in women compared with men.
Although smaller in magnitude than was observed with ruxolitinib SR, a similar difference was also observed between women and men receiving ruxolitinib IR. The reason for this difference remains unclear; however, it could not be explained by differences in body weight alone. No other covariates were significant predictors for pharmacokinetic parameters.
Patients receiving ruxolitinib SR exhibited reductions in spleen size compared with baseline, as well as improvement in symptoms as    Because ruxolitinib is a balanced JAK1/JAK2 inhibitor, it is unlikely that any potential differences in efficacy between the SR and IR formulations were attributable to JAK enzyme specificity. The variance is most likely due to the differences in peak, average, and minimum inhibition of JAK/signal transducer and activator of transcription signaling. However, an evaluation of isoform specificity was beyond the scope of our analysis.

| STUDY LIMITATIONS
Given the small size of the phase 2 study population, comparisons of efficacy between dose groups could not be adequately assessed.
Direct comparisons of the SR and IR formulations were not planned in this phase 2 proof-of-concept study, and post hoc comparisons are not feasible because the SR and IR formulations were administered sequentially without a washout period. In addition, several patients dropped out of the study after the week 16 visit before ruxolitinib IR was initiated.
A QOD dosing regimen that was available to some patients in the phase 2 study (25 mg SR QOD/50 mg SR QOD) was not available to the first several patients in the study. As such, dose titration was inconsistent and generally more challenging for the first few patients. Limited dosing options (ie, use of 1 25 mg tablet size) may have precluded effective dose titration and achievement of full clinical benefit for some patients. Finally, the phase 2 study was potentially limited by the exclusion of patients who were intolerant of ruxolitinib IR but may have derived clinical benefit from ruxolitinib SR.

| CONCLUSIONS
The current study results suggest that the higher peak inhibitions achieved with ruxolitinib IR may have incremental contributions to efficacy compared with ruxolitinib SR. However, there was a trend toward a lesser anemia effect with ruxolitinib SR, which could be beneficial for some patients with MF. Collectively, these data suggest that the development of effective SR formulations of ruxolitinib for treating patients with MF is feasible. However, further long-term studies with multiple ruxolitinib SR dosage strengths will be required to adequately compare efficacy and safety outcomes to those observed with ruxolitinib IR.