Safety and Efficacy of Fedratinib in Patients With Primary or Secondary Myelofibrosis:A Randomized Clinical Trial
Animesh Pardanani, MD; Claire Harrison, MD; Jorge E. Cortes, MD; Francisco Cervantes, MD; Ruben A. Mesa, MD; Donald Milligan, MD; Tamás Masszi, MD; Elena Mishchenko, MD; Eric Jourdan, MD; Alessandro M. Vannucchi, MD; Mark W. Drummond, MD; Mindaugas Jurgutis, MD; Kazimierz Kuliczkowski, MD, PhD; Emanuil Gheorghita, MD; Francesco Passamonti, MD; Frank Neumann, MD; Abhay Patki, PhD; Guozhi Gao, PhD; Ayalew Tefferi, MD
IMPORTANCE Myelofibrosis (MF) is a BCR-ABL–negative myeloproliferative neoplasm characterized by anemia, splenomegaly, debilitating constitutional symptoms, and shortened survival. Fedratinib, a JAK2-selective inhibitor, previously demonstrated clinically beneficial activity in patients with MF in early-phase trials.
Invited Commentary page 651
Supplemental content at jamaoncology.com
OBJECTIVE To evaluate the efficacy and safety of fedratinib therapy in patients with primary or secondary (post–polycythemia vera or post–essential thrombocythemia) MF.
DESIGN, SETTING, AND PARTICIPANTS Double-blind, randomized, placebo-controlled phase 3 study in 94 sites in 24 countries in which 289 adult patients (≥18 years of age) with intermediate-2 or high-risk primary MF, post–polycythemia vera MF, or post–essential thrombocythemia MF were randomly assigned between December 2011 and September 2012 to once-daily oral fedratinib, at a dose of 400 mg or 500 mg, or placebo, for at least 6 consecutive 4-week cycles.
MAIN OUTCOMES AND MEASURES The primary end point was spleen response (≥35% reduction in spleen volume from baseline as determined by magnetic resonance imaging or computed tomography) at week 24 and confirmed 4 weeks later. The main secondary end point was symptom response (≥50% reduction in total symptom score, assessed using the modified Myelofibrosis Symptom Assessment Form).
RESULTS The primary end point was achieved by 35 of 96 (36% [95% CI, 27%-46%]) and 39
of 97 (40% [95% CI, 30%-50%]) patients in the fedratinib 400-mg and 500-mg groups, vs 1 of 96 (1% [95% CI, 0%-3%]) in the placebo group (P < .001). Symptom response rates at week 24 were 33 of 91 (36% [95% CI, 26%-46%]), 31 of 91 (34% [95% CI, 24%-44%]), and 6
of 85 (7% [95% CI, 2%-13%]) in the fedratinib 400-mg, 500-mg, and placebo groups, respectively (P < .001). Common adverse events with fedratinib treatment were anemia, gastrointestinal symptoms, and increased levels of liver transaminases, serum creatinine, and pancreatic enzymes. Encephalopathy was reported in 4 women who received fedratinib 500 mg/d. A diagnosis of Wernicke encephalopathy was supported by magnetic resonance imaging in 3 cases and suspected clinically in 1 case.
CONCLUSIONS AND RELEVANCE Fedratinib therapy significantly reduced splenomegaly and symptom burden in patients with MF. These benefits were accompanied by toxic effects in some patients, the most important being encephalopathy of unknown mechanism. Clinical development of fedratinib was subsequently discontinued.
TRIAL REGISTRATION clinicaltrials.gov identifier: NCT01437787
JAMA Oncol. 2015;1(5):643-651. doi:10.1001/jamaoncol.2015.1590
Published online June 18, 2015.
Author Affiliations: Author affiliations are listed at the end of this article.
Corresponding Author: Ayalew Tefferi, MD, Division of Hematology, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (tefferi.ayalew @mayo.edu)
Copyright 2015 American Medical Association. All rights reserved.
Copyright 2015 American Medical Association. All rights reserved.
643
Downloaded From: http://oncology.jamanetwork.com/ by a Central Michigan University User on 11/01/2015
M
yelofibrosis (MF) is a BCR-ABL–negative myelopro- liferative neoplasm that can present de novo (pri- mary MF) or after transformation of polycythemia
vera (PV) or essential thrombocythemia (ET).1 Myelofibrosis is characterized by anemia, splenomegaly, debilitating constitu- tional symptoms, and shortened survival.2 Gain-of-function mu- tations resulting in constitutive activation of the Janus kinase (JAK) signal transducer and activator of transcription signal- ing pathway are common in patients with MF, with mutations in JAK2 (eg, JAK2 V617F) and myeloproliferative leukemia (MPL) occurring most frequently.3,4 Recently, calreticulin mutations have been found in a high proportion of JAK2-negative, MPL- negative patients.5,6 Currently, the only potentially curative treatment for patients with MF is allogeneic stem cell transplantation.7 However, most patients are not candidates for such treatment and alternative conventional drug treatments are palliative and often of limited benefit.8 Ruxolitinib phos- phate, a JAK1/2 inhibitor, was recently approved for the treat- ment of intermediate- or high-risk MF. However, not all pa- tients respond to ruxolitinib, and others may lose response over time.9-11 Alternative treatments are therefore needed.
Fedratinib (SAR302503/TG101348) is a JAK2-selective in- hibitor that demonstrated clinical benefit in patients with MF in early-phase clinical trials.12-14 Compared with ruxolitinib, fedratinib is a more specific inhibitor of JAK2 (vs other JAK fam- ily kinases) and appears to have activity against a broader fam- ily of kinases and kinase mutants.15 To further evaluate the ef- ficacy and safety of fedratinib, we conducted a randomized, double-blind, placebo-controlled phase 3 trial (JAKARTA) in patients with intermediate-2 or high-risk MF.
Methods
Eligibility Criteria
Eligible patients were at least 18 years old with a diagnosis of primary MF, post-PV MF, or post-ET MF with high-risk or in- termediate-2 risk disease according to 2008 World Health Organization16 and modified International Working Group for Myelofibrosis Research and Treatment (IWG-MRT) criteria,17,18 life expectancy at least 6 months, and Eastern Cooperative On- cology Group performance status of 0 to 2. Patients had pal- pable splenomegaly (≥5 cm below the left costal margin) and platelet count at least 50 × 103/μL (to convert to billions per li- ter, multiply by 1.0).
Study Design
The trial was conducted at 94 sites in 24 countries. Patients were randomized (1:1:1) to receive oral fedratinib 400 mg, 500 mg, or matched placebo once daily for at least 6 consecutive 4-week cycles. Randomization and treatment allocation was performed centrally via an interactive voice response sys- tem, which assigned each patient a unique 9-digit patient iden- tification number. The interactive voice response system gen- erated separate patient randomization lists, allocated the treatment arm, and instructed the study personnel as to which kit and treatment number to dispense to the patient. Patients continued to receive their assigned treatment until they ex-
perienced disease progression or relapse (modified IWG- MRT criteria18), excess toxic effect, or other criteria outlined in the protocol. Dosing was interrupted following the occur- rence of certain adverse events (AEs), as defined in the study protocol, with the recommendation that dosing be restarted at a level 100 mg/d below that at which the event was ob- served (see Trial Protocol in Supplement 1). Crossover from pla- cebo to fedratinib was permitted after 24 weeks, or before if the patient experienced progressive disease as predefined in the study protocol. Crossover patients were randomized (1:1) to 1 of the 2 fedratinib doses.
The protocol was approved by institutional review boards at each study site. The study was conducted according to the principles of the Declaration of Helsinki. All patients provided written informed consent. Data were collected by the inves- tigators and analyzed by the study sponsor with oversight, feedback, and approval from the investigators and a data- monitoring committee.
Study End Points
The primary end point was the proportion of patients with at least a 35% reduction in spleen volume (measured by mag- netic resonance imaging [MRI] or computed tomography) from baseline to week 24 (end of cycle 6) and confirmed 4 weeks later. The proportion of patients with at least a 35% reduction in spleen volume from baseline to 24 weeks regardless of con- firmation was a secondary end point. The main secondary end point was the proportion of patients with at least a 50% re- duction in total symptom score (TSS) (calculated as the weekly mean value of the daily total score for 6 key symptoms [night sweats, pruritus, abdominal discomfort, early satiety, pain un- der ribs on left side, and bone or muscle pain]) from baseline at 24 weeks, assessed using the modified Myelofibrosis Symp- tom Assessment Form19 e-diary. Change in JAK2 V617F allele burden was an exploratory end point. Adverse events were coded using National Cancer Institute Common Terminology Criteria for Adverse Events, version 3.
Statistical Analysis
Assuming a 15% drop-out rate, 75 patients per arm provided 90% power to detect a treatment difference in spleen re-
644 JAMA Oncology August 2015 Volume 1, Number 5 jamaoncology.com
Figure 1. Patient Disposition
a One patient in the placebo group was randomized but died before taking the first dose of medication.
sponse at a 2-sided α level of .05, assuming response rates of 30% in either fedratinib arm and 5% in the placebo arm. Un- less stated, analyses were performed using the intention-to- treat population. For analysis of spleen and symptom re- sponse, patients with missing valid assessments at week 24 or who experienced disease progression before 24 weeks were considered nonresponders. Safety was assessed throughout the study period in all patients who received at least 1 dose of study medication. Comparison of the primary end point and key sec- ondary end point of symptom response between placebo and each drug arm was performed using the χ2 test at a 2-sided α level of .025. Summary and descriptive statistics were gener- ated for all other end points.
Results
Characteristics of the Patients
A total of 289 patients were enrolled from December 2011 to September 2012, with 96, 97, and 96 patients randomly assigned to fedratinib 400 mg, 500 mg, and placebo, respec- tively. Baseline patient characteristics are shown in eTable 1 in Supplement 2. At the analysis cutoff date (March 21, 2013), 64 (67%), 59 (61%), and 1 (1%) patients in the fedratinib 400- mg, 500-mg, and placebo groups were still receiving the
originally assigned treatment. In the placebo group, 24 patients discontinued treatment and 10 patients crossed over to fedratinib before the end of week 24. After this point, 60 additional patients crossed over to fedratinib; an addi- tional patient was randomized but died before receiving study treatment. Detailed patient disposition is shown in Figure 1.
Spleen Volume Reduction
The proportion of patients with at least a 35% reduction in spleen volume (spleen response) at 24 weeks, and confirmed 4 weeks later, was significantly higher (P < .001 for the com- parison of each dose with placebo) in the fedratinib 400-mg (35 [36%]) and 500-mg (39 [40%]) groups compared with pla-
cebo (1 [1%]) (Table 1). Response rates at week 24 (without con-
firmation) were 47% (45 patients at 400 mg), 49% (48 pa-
tients at 500 mg), and 1% (1 patient with placebo; P < .001) (Table 1). Information regarding spleen volume reduction for patients without confirmation of spleen response at 24 weeks is shown in eTable 2 in Supplement 2. Among patients with available data, all except 3 patients in the fedratinib 400-mg group and 2 patients in the fedratinib 500-mg group experi- enced some degree of spleen volume reduction at 24 weeks of up to 70%, whereas a majority of patients (44 of 59 [75%]) in the placebo group had an increase in spleen volume at this
Crossover after week 24
24 Discontinued by data cutoff (excludes crossover)
8 Adverse events
4 Disease progression
12 Other reasons
1 Continued placebo
jamaoncology.com JAMA Oncology August 2015 Volume 1, Number 5 645
Splenic Responsea Fedratinib 400 mg
(n = 96)
Week 24
Confirmed Week 24 Fedratinib 500 mg
(n = 97)
Week 24
Confirmed Week 24 Placebo (n = 96)
Week 24 Confirmed Week 24
Table 1. Spleen Responses
All patients, No. (%) [95% CI] 45 (47) [37-57]b 35 (36) [27-46]b 48 (49) [40-59]b 39 (40) [30-50]b 1 (1) [0-3] 1 (1) [0-3]
Baseline platelet count, ×103/μL, proportion (%)
≥100 40/82 (49) 32/82 (39) 42/82 (51) 34/82 (42) 1/77 (1) 1/77 (1)
<100 5/14 (36) 3/14 (21) 6/15 (40) 5/15 (33) 0/18 0/18
Disease subtype, proportion (%)
Primary MF 29/62 (47) 21/62 (34) 32/63 (51) 25/63 (40) 1/58 (2) 1/58 (2)
Post-ET MF 6/10 (60) 5/10 (50) 4/9 (44) 4/9 (44) 0/11 0/11
Post-PV MF 10/24 (42) 9/24 (38) 12/25 (48) 10/25 (40) 0/27 0/27
Risk status, proportion (%)
Intermediate-2 30/57 (53) 23/57 (40) 26/47 (55) 21/47 (45) 1/46 (2) 1/46 (2)
High risk 15/39 (38) 12/39 (31) 22/50 (44) 18/50 (36) 0/50 0/50
JAK2 mutational status, proportion (%)
Wild type 10/30 (33) 9/30 (30) 8/20 (40) 7/20 (35) 1/32 (3) 1/32 (3)
Mutant 34/62 (55) 25/62 (40) 37/72 (51) 31/72 (43) 0/59 0/59
Abbreviations: ET, essential thrombocythemia; MF, myelofibrosis; PV, polycythemia vera.
SI conversion factor: To convert platelet count to billions per liter, multiply by 1.0.
a Defined as at least a 35% reduction in spleen volume.
b P < .001 vs placebo (χ2 test).
Figure 2. Change in Spleen Volume in Individual Patients at 24 Weeks
Placebo
80
Change in Spleen Volume From Baseline, %
60
40
20
0
–20
–40
–60
point (Figure 2). Spleen response rates were higher in the fe- dratinib groups compared with placebo regardless of base- line platelet count, disease subtype, risk status, or JAK2 mu- tational status (Table 1).
Symptom Response
In patients with available baseline TSS, the proportions of pa- tients with a reduction of at least 50% in the TSS from base- line to week 24 were 33 of 91 (36% [95% CI, 26%-46%]), 31 of
91 (34% [95% CI, 24%-44%]), and 6 of 85 (7% [95% CI, 2%-
–80
80
Change in Spleen Volume From Baseline, %
60
40
20
0
–20
–40
–60
–80
80
Change in Spleen Volume From Baseline, %
60
40
20
0
–20
–40
–60
–80
Patient
Fedratinib 400 mg
Patient
Fedratinib 500 mg
Patient
12%]) in the 400-mg, 500-mg, and placebo groups, respec- tively. Both active treatment groups showed clinically and sta- tistically significant differences vs placebo on the basis of a step-down procedure for controlling multiplicity of statisti- cal comparison (P ≤ .001 at a significance level of .025 for each comparison).
In fedratinib-treated patients with baseline TSS greater than 0, median TSS decreased from baseline at 4 weeks, with the benefit continuing to 24 weeks (Figure 3A). The changes in TSS for individual patients at week 24 (baseline TSS > 0) are shown in Figure 3B. Twenty-six patients in the placebo group, vs 14 and 9 in the fedratinib 400-mg and 500-mg groups, showed no improvement in TSS.
JAK2 Allele Burden
At baseline, 62 of 96 (65%), 72 of 97 (74%), and 59 of 96 (61%) of patients in the 400-mg, 500-mg, and placebo groups were JAK2 V617F positive. No meaningful changes were observed in allele burden during treatment (median [range] changes at 24 weeks: +0.4% [−99% to 196%], +0.8% [−77% to 115%], and
+2%[−74% to 96%] in the 400-mg, 500-mg, and placebo groups,
respectively).
646 JAMA Oncology August 2015 Volume 1, Number 5 jamaoncology.com
Figure 3. Symptom Responses
Change in TSS by group
Placebo Fedratinib 400 mg
Fedratinib 500 mg
20
Median Change in TSS From Baseline, %
0
-20
-40
–60
–80
0 4 8 12 16 20 24
Time, wk
Change in TSS for individual patients
Placebo
200
Change in MF-SAF TSS From Baseline, %
180
160
140
120
100
80
60
40
20
0
–20
–40
–60
–80
–100
Patient
200
Change in MF-SAF TSS From Baseline, %
180
160
140
120
100
80
60
40
20
0
–20
–40
–60
–80
–100
Fedratinib 400 mg
Patient
200
180
Change in MF-SAF TSS at End of Cycle 6 Relative to Baseline, %
160
140
120
100
80
60
40
20
0
–20
–40
–60
–80
–100
Fedratinib 500 mg
Patient
A, Median percentage change in total symptom score (TSS) to week 24 in the dose groups. Ranges around median values are wide, so error bars are omitted for clarity. B, Percentage change in total symptom score from baseline at week 24 in individual patients, according to dose group.
Data are shown for patients with TSS greater than 0 at baseline and TSS data at week 24. MF-SAF indicates Myelofibrosis Symptom Assessment
Form.
Safety
Among fedratinib-treated patients, 148 of 193 (77%) com-
pleted 24 weeks of treatment (400 mg, 79 of 96 [82%]; 500 mg,
69 of 97 [71%]), compared with 62 of 95 (65%) of placebo- treated patients. Details regarding drug exposure are pro- vided in eTable 3 in Supplement 2. A total of 24 (25%) (400 mg),
32 (33%) (500 mg), and 38 (40%) (placebo) patients had dis- continued treatment at the analysis cutoff date. The percent- ages of patients discontinuing because of AEs at any time dur- ing the first 24 weeks were 14% (13 of 96), 25% (24 of 97), and 8% (8 of 95) in the 400-mg, 500-mg, and placebo groups, re- spectively. Thrombocytopenia (7 patients), cardiac failure,
vomiting, and diarrhea (4 patients each) were the most com- mon AEs leading to fedratinib discontinuation (eTable 4 in Supplement 2).
The most common hematologic AE was anemia (Table 2). Fedratinib treatment was associated with a decrease in hemoglobin levels, with nadir reached after 12 to 16 weeks (eFigure 1 in Supplement 2). Thereafter, mean hemoglobin levels showed partial recovery in the 400-mg group but not in the 500-mg group. Among 19 patients who were red blood cell transfusion dependent at baseline, 15 became transfusion independent during fedratinib treat- ment (7 of 8 [400 mg], 5 of 5 [500 mg], and 3 of 6 [placebo]).
jamaoncology.com JAMA Oncology August 2015 Volume 1, Number 5 647
Table 2. Adverse Events Observed in at Least 10% of Patients in Any Treatment Group
Fedratinib 400 mg Fedratinib 500 mg Placebo
(n = 96) (n = 97) (n = 95)
Adverse Events, No. (%) All Grades Grade 3 or 4 All Grades Grade 3 or 4 All Grades Grade 3 or 4
Any TEAE 96 (100) 52 (54) 95 (98) 68 (70) 89 (94) 30 (32)
TEAE leading to treatment discontinuation to week 24 13 (14) 12 (13) 24 (25) 15 (16) 8 (8) 4 (4)
Serious TEAE 26 (27) 17 (18) 30 (31) 23 (24) 22 (23) 14 (15)
Nonhematologica
Diarrhea 63 (66) 5 (5) 54 (56) 5 (5) 15 (16) 0
Vomiting 40 (42) 3 (3) 53 (55) 9 (9) 5 (5) 0
Nausea 61 (64) 0 49 (51) 6 (6) 14 (15) 0
Constipation 10 (10) 2 (2) 17 (18) 0 7 (7) 0
Asthenia 9 (9) 2 (2) 15 (16) 4 (4) 6 (6) 1 (1)
Abdominal pain 14 (15) 0 12 (12) 1 (1) 15 (16) 1 (1)
Fatigue 15 (16) 6 (6) 10 (10) 5 (5) 9 (10) 0
Dyspnea 8 (8) 0 10 (10) 1 (1) 6 (6) 2 (2)
Neutropenia 27 (28) 8 (8) 42 (44) 17 (18) 14 (15) 4 (4)
Infections and infestationsb 40 (42) 2 (2) 38 (39) 12 (12) 26 (27) 4 (4)
Laboratory parameter elevation
Alanine transaminase 51 (53) 3 (3) 44 (46) 3 (3) 16 (17) 0
Aspartate transaminase 58 (60) 2 (2) 46 (48) 2 (2) 27 (29) 1 (1)
Hyperbilirubinemia 30 (31) 2 (2) 27 (28) 1 (1) 38 (40) 2 (2)
Creatinine 52 (54) 3 (3) 60 (63) 0 28 (30) 1 (1)
Amylase 25 (26) 2 (2) 22 (23) 3 (3) 7 (7) 0
Lipase 43 (45) 12 (13) 34 (36) 9 (9) 6 (6) 2 (2)
Anemia 95 (99) 41 (43) 94 (98) 58 (60) 86 (91) 24 (25)
Thrombocytopenia 60 (63) 16 (17) 55 (57) 26 (27) 48 (51) 9 (9)
Lymphopenia 54 (57) 20 (21) 63 (66) 26 (27) 50 (54) 19 (21)
Leukopenia 45 (47) 6 (6) 51 (53) 15 (16) 18 (19) 3 (3)
Abbreviation: TEAE, treatment-emergent adverse event.
a Nonhematologic and hematologic adverse events that occurred in at least 10% of patients in any group. Hematologic events are assessed by means of laboratory analysis.
b System organ class.
Thirty of 178 (17%) fedratinib patients and 21 of 92 (23%) placebo patients who were transfusion independent at baseline became transfusion dependent during treatment. Rates of thrombocytopenia, leukopenia, and neutropenia were higher in both fedratinib groups vs placebo (Table 2). Fedratinib discontinuation due to thrombocytopenia was more frequent among patients with baseline platelet levels less than 100 × 103/μL (31% [400 mg, 5 of 14; 500 mg, 4 of
15] vs <1% [1 of 164] for those with baseline platelet levels
≥100 × 103/μL).
The most common nonhematologic AEs with fedratinib treatment were diarrhea, vomiting, and nausea (Table 2). Gas- trointestinal toxic effects led to fedratinib dose reductions or interruptions in 38 patients (20%) (14 of 96 [15%] 400 mg and
24 of 97 [25%] 500 mg), and permanent discontinuation of fe- dratinib therapy in 16 patients (8%) (7 [7%] and 9 [9%]). The incidence of gastrointestinal toxic effects was highest during the first cycle of treatment and decreased thereafter (eFigure 2 in Supplement 2).
Infections occurred in 40 patients (42%) (grade 3 or 4, 2 patients [2%]) and 38 patients (39%) (12 patients [12%]) in the 400-mg and 500-mg groups, respectively, compared with 26 patients [27%] (4 patients [4%]) in the placebo group. Urinary tract infections were the most commonly reported (8 patients in each fedratinib group), although no specific class of infection was observed in 10% or more of patients.
The incidence of alanine aminotransferase and aspartate aminotransferase elevations was more frequent with fedra- tinib treatment than with placebo (Table 2). These elevations were generally mild to moderate, asymptomatic, and revers- ible with fedratinib dose reduction or interruption. Similarly, increased levels of serum creatinine, amylase, and lipase were also more commonly noted in fedratinib-treated patients (Table 2).
Transformation to acute myeloid leukemia was reported in 1 patient in each of the fedratinib groups and in 2 patients in the placebo group. A total of 24 patients died during the first
648 JAMA Oncology August 2015 Volume 1, Number 5 jamaoncology.com
24 weeks (4 [4%], 10 [10%], and 10 [11%] in the 400-mg, 500- mg, and placebo groups, respectively); 9 of these deaths were attributed to AEs (1 [1%], 4 [4%], and 4 [4%]) (eTable 5 in Supplement 2). No analyses of overall or progression-free sur- vival were conducted because there were too fewevents at data cutoff.
One case of Wernicke encephalopathy (WE) and 1 case of encephalopathy of unknown origin were reported before data cutoff. Following database lock and notification of investiga- tors of the 2 prior cases, 2 additional cases of WE were re- ported in patients during the study, 1 of which was reported after fedratinib trials had been halted. All 4 cases were in women who received fedratinib 500 mg (eTable 6 in Supplement 2). Three of these cases were confirmed by an in- dependent expert safety panel, which included a neuro- oncologist and a neuroradiologist, as WE on the basis of both clinical features and MRI. Diagnosis of WE in the fourth case was confirmed by the panel on the basis of clinical symptoms alone in the absence of specific MRI findings. Symptoms de- veloped 6 to 44 weeks after initiation of fedratinib treatment and led to permanent treatment discontinuation. Pharmaco- kinetic analysis of 2 patients indicated above-mean fedra- tinib levels for the patients compared with the study popula- tion; the other 2 patients were placebo crossovers and had not been assessed for pharmacokinetic exposure. Cognitive symp- toms developed in the absence of concurrent alcohol use as a potential contributing factor but in the context of persistent vomiting, malnutrition, and cachexia in 1 patient, vomiting and hyponatremia in 1 patient, and renal failure with mild hypo- natremia in 1 patient. Intravenous thiamine was adminis- tered after recognition of the clinical syndrome, and, whereas all patients showed some response to treatment, at the time of this report, some persistent cognitive defects remain in all patients.
Discussion
These results demonstrate that fedratinib treatment is asso- ciated with significant improvements in splenomegaly and symptom burden in patients with intermediate-2 or high- risk MF, as compared with placebo. The primary end point was achieved by 36% and 40% of patients in the fedratinib 400-mg and 500-mg groups, respectively, whereas only 1 patient in the placebo group achieved this response. In addi- tion to those patients meeting the predefined primary end point, an additional 10 patients (5 in each active treatment arm) achieved spleen volume reductions of at least 35% at 24 weeks that were not confirmed 4 weeks later; in all cases, spleen volume reductions at the confirmation date were at least 25%. Spleen responses with fedratinib treatment were not appreciably affected by JAK2 mutational status, MF dis- ease subtype, or risk status. Symptom burden was reduced following 4 weeks of treatment in both fedratinib dose groups and remained at this reduced level to the 24-week analysis time point. In contrast, most patients in the placebo group showed no improvement in symptoms during a period of 24 weeks.
Given the key role played by JAK2-dependent signaling in normal hematopoiesis, it is not surprising that cytopenias are a common adverse effect of JAK inhibitor therapy.9,10,13,14 Accordingly, hemoglobin levels decreased following initia- tion of fedratinib therapy, and anemia was the most com- mon hematologic toxic effect observed in this study. How- ever, similar to the experience with ruxolitinib wherein the incidence of treatment-emergent anemia is also relatively high,10 hemoglobin levels in the present study showed par- tial recovery after 16 weeks of fedratinib treatment, albeit in the 400-mg group only, and only 1 patient from each active treatment group discontinued therapy because of anemia. Gastrointestinal toxic effects were frequent, although only 8% of patients discontinued treatment on this basis. Lack of protocol-mandated guidelines for prophylaxis, early recog- nition, and treatment of gastrointestinal toxic effects at the time of study initiation may have contributed to the high incidence of such events.
During the analysis period of the present study, 1 case of WE and 1 case of encephalopathy of unknown origin were reported. Following database lock, the study sponsor received 2 additional reports of WE, 1 in the present study and 1 in another fedratinib trial. These reports prompted a full clinical hold of the fedratinib program. Subsequently, reports were received of a fourth suspected case in the pre- sent study (later confirmed) and 3 additional cases (2 con- firmed) in other fedratinib trials. Wernicke encephalopathy is underrecognized in the general population,20 as well as in patients with myeloproliferative neoplasms,21 given that only a minority of patients present with the classic triad of encephalopathy, oculomotor dysfunction, and gait ataxia. Predisposing factors for WE resulting from thiamine defi- ciency in nonalcoholic patients include malnutrition, recur- rent vomiting and/or diarrhea, chemotherapy, renal disease,
and magnesium depletion.22 Vomiting (with or without nau- sea and/or diarrhea) was common among patients who devel- oped encephalopathy while receiving fedratinib in the pre- sent study, although similar gastrointestinal events were frequent in patients who received fedratinib across all studies and cannot be considered the dominant contributing factor. Relevant comorbidities observed in some affected patients included weight loss and/or cachexia, hyponatremia, and malnutrition. All 7 confirmed cases of WE noted in the fedra- tinib program occurred in women, and in every case the event occurred at the 500-mg dose level. Furthermore, in several patients with WE, plasma fedratinib levels when mea- sured early in the study were higher than the mean for the overall study population. Despite this, it was difficult to establish a conclusive link between fedratinib dose level and WE, given that the overall demographic and clinical risk char- acteristics were not unique to the patients who developed WE and given the variability in drug exposure at a given dose level due to pharmacokinetic differences. Because the pri- mary mechanism for development of WE was not identified, there were insufficient data to conclude that patients receiv- ing fedratinib at doses of 400 mg or less were not at risk of developing WE, particularly given that this AE may not be fully reversible even with timely thiamine therapy. Fedra-
jamaoncology.com JAMA Oncology August 2015 Volume 1, Number 5 649
tinib has been shown to cross the blood-brain barrier in ani- mal models, and consequently a direct effect of fedratinib on neurologic function, either through JAK inhibition or off- target effects, cannot be excluded.23 To date, WE has not been specifically reported in trials with other JAK inhibitors, although it may be relevant that clinical development of the JAK2-selective inhibitor XL01924 and the JAK1/JAK2 inhibitor AZD148025 was discontinued as a result of neurologic toxic effects. Furthermore, treatment-emergent peripheral neu- ropathy has been reported in up to 44% of patients with MF treated with the JAK1/2 inhibitor momelotinib.26 Following a thorough risk-benefit analysis, once the cases of WE were recognized, the sponsor decided to permanently discontinue clinical development of fedratinib.
The main limitation of the study was its early termina- tion before sufficient follow-up to allow analysis of progression- free survival or overall survival and other long-term efficacy and safety outcomes. There were also slight imbalances be- tween the treatment groups with regard to baseline disease risk status and median platelet counts.
Conclusions
Fedratinib treatment demonstrated significant clinical benefit in patients with intermediate-2 or high-risk MF. These benefits were offset by a risk of developing encephalopathy, the mecha- nism and reversibility of which remain to be confirmed.
ARTICLE INFORMATION
Accepted for Publication: April 23, 2015.
Published Online: June 18, 2015. doi:10.1001/jamaoncol.2015.1590.
Author Affiliations: Department of Medicine, Mayo Clinic, Rochester, Minnesota (Pardanani, Tefferi); Guy’s and St Thomas’ NHS Foundation Trust, St Thomas’ Hospital, London, England (Harrison); University of Texas MD Anderson Cancer Center, Houston (Cortes); Hospital Clinic, IDIBAPS University of Barcelona, Barcelona, Spain (Cervantes); Division of Hematology-Oncology, Mayo Clinic, Scottsdale, Arizona (Mesa); Centre for Haematology and Stem Cell Transplantation, Heartlands Hospital, Birmingham, England (Milligan); Department of Haematology and Stem Cell Transplantation, St István and St László Hospital, Budapest, Hungary (Masszi); Hematology-Oncology Unit, Bnai-Zion Medical Center, Haifa, Israel (Mishchenko); Groupe Hospitalo-Universitaire Nimes, Nimes, France
(Jourdan); Department of Experimental and Clinical
Medicine, University of Florence, Florence, Italy (Vannucchi); Beatson West of Scotland Cancer Centre, Gartnavel General Hospital, Glasgow, Scotland (Drummond); Hematology Department, Klaipeda Seamen’s Hospital, Klaipeda, Lithuania (Jurgutis); Department of Hematology, Blood Neoplasms and Bone Marrow Transplantation, Wroclaw Medical University, Wroclaw, Poland (Kuliczkowski); Brasov Country Hospital, Brasov, Romania (Gheorghita); University Hospital Ospedale di Circolo e Fondazione Macchi, Varese, Italy (Passamonti); Sanofi Oncology, Cambridge, Massachusetts (Neumann, Patki, Gao).
Author Contributions: Drs Pardanani and Tefferi had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Harrison, Cortes, Mesa, Mishchenko, Jurgutis, Passamonti, Neumann, Patki, Gao, Tefferi.
Acquisition, analysis, or interpretation of data: Pardanani, Harrison, Cortes, Cervantes, Mesa, Masszi, Jourdan, Vannucchi, Drummond, Kuliczkowski, Gheorghita, Neumann, Patki, Gao, Tefferi.
Drafting of the manuscript: Pardanani, Harrison, Mesa, Milligan, Jurgutis, Neumann, Gao, Tefferi. Critical revision of the manuscript for important intellectual content: Pardanani, Harrison, Cortes, Cervantes, Mesa, Milligan, Masszi, Mishchenko, Jourdan, Vannucchi, Drummond, Kuliczkowski,
Gheorghita, Passamonti, Neumann, Patki, Gao, Tefferi.
Statistical analysis: Gao.
Administrative, technical, or material support: Mesa, Milligan, Masszi, Vannucchi, Kuliczkowski, Neumann, Patki.
Study supervision: Pardanani, Harrison, Cortes, Cervantes, Mesa, Milligan, Masszi, Drummond, Gheorghita, Passamonti, Neumann, Patki, Tefferi.
Conflict of Interest Disclosures: Dr Pardanani has received clinical trial support from Sanofi, Bristol- Myers Squibb, PharmaMar, and JW Pharma. Dr Harrison has received personal fees from CTI, Gilead, Sanofi, Shire, and YM Bioscience and grants and personal fees from Novartis. Dr Cortes has received grants and personal fees from Sanofi and Incyte. Dr Cervantes has received personal speaker fees from Novartis and CTI-Baxter and advisory board consultancy fees from Novartis, Sanofi, and CTI- Baxter. Dr Mesa has received consultancy fees from Novartis and grant support from Sanofi, Incyte, Gilead, CTI, Genentech, NS Pharma, and Promedior. Dr Milligan has received personal fees from Sanofi. Dr Vannucchi has received personal fees from Sanofi. Dr Drummond has received personal fees from Sanofi. Dr Jurgutis has received personal fees from Roche and Novartis. Drs Neumann, Patki, and Gao were employees of Sanofi while the trial was being conducted and during the analysis of the data. No other disclosures are reported.
Funding/Support: The study was sponsored by Sanofi.
Role of the Funder/Sponsor: The study sponsor participated with the academic authors in the design and conduct of the study. Data were collected by the investigators and analyzed by the study sponsor with oversight, feedback, and approval from the investigators and a data-monitoring committee. The manuscript was written by the authors with medical writing assistance from Neil Anderson, PhD, Adelphi Communications Ltd, who was funded by the study sponsor. The manuscript was reviewed by the study sponsor, but the authors had the final decision to submit the manuscript for publication.
Additional Contributions: We thank all the study investigators and their clinical teams for their contribution to this study, and the patients for agreeing to participate.
REFERENCES
⦁ Tefferi A. Myelofibrosis with myeloid metaplasia.
N Engl J Med. 2000;342(17):1255-1265.
⦁ Tefferi A, Lasho TL, Jimma T, et al. One thousand patients with primary myelofibrosis: the Mayo Clinic experience. ⦁ Mayo⦁ ⦁ Clin⦁ ⦁ Proc⦁ .⦁ ⦁ 2012;87(1):25-33.
⦁ Jatiani SS, Baker SJ, Silverman LR, Reddy EP. Jak/STAT pathways in cytokine signaling and myeloproliferative disorders: approaches for targeted therapies. ⦁ Genes⦁ ⦁ Cancer⦁ .⦁ ⦁ 2010;1(10):979-993.
⦁ Tefferi A. Novel mutations and their functional and clinical relevance in myeloproliferative neoplasms: JAK2, MPL, TET2, ASXL1, CBL, IDH and IKZF1. ⦁ Leukemia⦁ .⦁ ⦁ 2010;24(6):1128-1138.
⦁ Nangalia J, Massie CE, Baxter EJ, et al. Somatic CALR mutations in myeloproliferative neoplasms with nonmutated JAK2. ⦁ N⦁ ⦁ Engl⦁ ⦁ J⦁ ⦁ Med⦁ .⦁ ⦁ 2013;369 ⦁ (25):2391-2405.
⦁ Klampfl T, Gisslinger H, Harutyunyan AS, et al. Somatic mutations of calreticulin in myeloproliferative neoplasms. ⦁ N Engl J Med⦁ . ⦁ 2013; ⦁ 369(25):2379-2390.
⦁ Kroeger N, Holler E, Kobbe G, et al.
Dose-reduced conditioning followed by allogeneic stem cell transplantation in patients with myelofibrosis: results from a multicenter prospective trial of the chronic leukemia working party of the European Group for Blood and Marrow Transplantation (BMT). Blood. 2007;110(suppl): abstr 683.
⦁ Tefferi A. Primary myelofibrosis: 2013 update on diagnosis, risk-stratification, and management. ⦁ Am ⦁ J⦁ ⦁ Hematol⦁ .⦁ ⦁ 2013;88(2):141-150.
⦁ Harrison C, Kiladjian JJ, Al-Ali HK, et al. JAK inhibition with ruxolitinib versus best available therapy for myelofibrosis. ⦁ N⦁ ⦁ Engl⦁ ⦁ J⦁ ⦁ Med⦁ .⦁ ⦁ 2012;366 ⦁ (9):787-798.
⦁ Verstovsek S, Mesa RA, Gotlib J, et al.
A double-blind, placebo-controlled trial of ruxolitinib for myelofibrosis. N Engl J Med. 2012; 366(9):799-807.
⦁ Verstovsek S, Mesa RA, Gotlib J, et al. Efficacy, safety and survival with ruxolitinib in patients with myelofibrosis: results of a median 2-year follow-up of COMFORT-I. ⦁ Haematologica⦁ .⦁ ⦁ 2013;98(12):1865-1871.
⦁ Pardanani A, Gotlib JR, Jamieson C, et al. Safety and efficacy of TG101348, a selective JAK2 inhibitor, in myelofibrosis. ⦁ J⦁ ⦁ Clin⦁ ⦁ Oncol⦁ .⦁ ⦁ 2011;29(7):789-796.
⦁ Talpaz M, Jamieson C, Gabrail N, et al. Modulation of plasma cytokines and its association with clinical response to treatment with the
JAK2-selective inhibitor SAR302503 in
650 JAMA Oncology August 2015 Volume 1, Number 5 jamaoncology.com
myelofibrosis (MF). J Clin Oncol. 2013;31(suppl 15):abstr 7110.
⦁ Gotlib J, Pardanani A, Jamieson C, et al. Long-term follow up of a phase 1/2 study of SAR302503, an oral JAK2 selective inhibitor, in
patients with myelofibrosis (MF). Haematologica. 2012;97(s1):145.
⦁ Zhou T, Georgeon S, Moser R, Moore DJ, Caflisch A, Hantschel O. Specificity and mechanism-of-action of the JAK2 tyrosine kinase inhibitors ruxolitinib and SAR302503 (TG101348). ⦁ Leukemia⦁ .⦁ ⦁ 2014;28(2):404-407⦁ .
⦁ Vardiman JW, Thiele J, Arber DA, et al. The 2008 revision of the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia: rationale and important changes. ⦁ Blood⦁ .⦁ ⦁ 2009;114(5):937-951.
⦁ Cervantes F, Dupriez B, Pereira A, et al. New prognostic scoring system for primary myelofibrosis based on a study of the International Working Group for Myelofibrosis Research and Treatment. ⦁ Blood⦁ .⦁ ⦁ 2009;113(13):2895-2901.
⦁ Barosi G, Mesa RA, Thiele J, et al; International Working Group for Myelofibrosis Research and Treatment (IWG-MRT). Proposed criteria for the diagnosis of post-polycythemia vera and
post-essential thrombocythemia myelofibrosis: a consensus statement from the International Working Group for Myelofibrosis Research and Treatment. Leukemia. 2008;22(2):437-438.
⦁ Mesa RA, Kantarjian H, Tefferi A, et al. Evaluating the serial use of the Myelofibrosis Symptom Assessment Form for measuring symptomatic improvement: performance in 87 myelofibrosis patients on a JAK1 and JAK2 inhibitor (INCB018424) clinical trial. ⦁ Cancer⦁ .⦁ ⦁ 2011;117(21): ⦁ 4869-4877⦁ .
⦁ Torvik A, Lindboe CF, Rogde S. Brain lesions in alcoholics: a neuropathological study with clinical correlations. ⦁ J⦁ ⦁ Neurol⦁ ⦁ Sci⦁ .⦁ ⦁ 1982;56(2-3):233-248.
⦁ Wu J, Zhang L, Vaze A, Lin S, Juhaeri J. Risk of Wernicke’s encephalopathy and cardiac disorders in patients with myeloproliferative neoplasm. ⦁ Cancer ⦁ Epidemiol⦁ .⦁ ⦁ 2015;39(2):242-249.
⦁ Sechi G, Serra A. Wernicke’s encephalopathy: new clinical settings and recent advances in diagnosis and management. ⦁ Lancet Neurol⦁ .⦁ ⦁ 2007;6(5):442-455.
⦁ Zhang Q, Zhang Y, Diamond S, et al. The Janus kinase 2 inhibitor fedratinib inhibits thiamine uptake: a putative mechanism for the onset of Wernicke’s encephalopathy. ⦁ Drug Metab Dispos⦁ . ⦁ 2014;42(10):1656-1662.
⦁ Quintás-Cardama A, Kantarjian H, Cortes J, Verstovsek S. Janus kinase inhibitors for the treatment of myeloproliferative neoplasias and beyond. ⦁ Nat⦁ ⦁ Rev⦁ ⦁ Drug⦁ ⦁ Discov⦁ .⦁ ⦁ 2011;10(2):127-140.
⦁ Plimack ER, Lorusso PM, McCoon P, et al. AZD1480: a phase I study of a novel JAK2 inhibitor in solid tumors. ⦁ Oncologist⦁ .⦁ ⦁ 2013;18(7):819-820.
⦁ Abdelrahman RA, Begna KH, Al-Kali A, et al. Momelotinib treatment-emergent neuropathy: prevalence, risk factors and outcome in 100 patients with myelofibrosis. ⦁ Br⦁ ⦁ J⦁ ⦁ Haematol⦁ .⦁ ⦁ 2015;169(1):77-80.
Invited Commentary
JAK/STAT Pathway Inhibitors and Neurologic Toxicity
Above All Else Do No Harm?
Bart L. Scott, MD; Pamela S. Becker, MD, PhD
Oncologists long ago recognized the fallacy of primum non no- cere. The truth is that some harm may be deemed acceptable as long as the perceived benefits have the potential to out-
weigh the risks. Janus-asso- ciated kinase (JAK)/signal
icity. Four patients (2% of those receiving fedratinib) devel- oped Wernicke encephalopathy (WE) in this trial, and an additional 3 patients developed WE in other fedratinib stud- ies. The clinical triad of WE includes ocular signs, cerebellar dysfunction, and confusion, and it is typically due to thia-
Related article page 643
transducers and activators of transcription (STAT) path-
mine deficiency, with higher prevalence in alcoholics. How- ever, treatment with thiamine did not reverse the symptoms
way inhibitors have demonstrated efficacy in the treatment of myelofibrosis, resulting in reduction of splenomegaly and constitutional symptoms, and improved survival.1-3 The COMFORT-I and COMFORT-II trials led to Food and Drug Ad- ministration approval for the JAK1/JAK2 inhibitor ruxolitinib for patients with myelofibrosis, and the RESPONSE trial, to ap- proval for patients with polycythemia vera who are unrespon- sive or intolerant to hydroxyurea therapy. The principal ad- verse effects of ruxolitinib in the treatment of myelofibrosis include anemia, thrombocytopenia, ecchymosis, dizziness, and headache. A distinct neurological adverse effect, peripheral neuropathy, was reported in up to 44% of patients treated with the JAK1/JAK2 inhibitor momelotinib. In fact, neurologic toxic effects including dizziness, ataxia, aphasia, dysarthria, and am- nesia led totermination of the clinical development of the JAK1/ JAK2 inhibitor AZD1480.4
In this issue of JAMA Oncology, Pardanani and colleagues5 describe the results of a double-blind, placebo-controlled, ran- domized, international multicenter clinical trial with a JAK2- selective inhibitor, fedratinib, in patients with primary or sec- ondary myelofibrosis. Although the drug was shown to be effective in reducing spleen size, it was also noted to occa- sionally exhibit substantial central nervous system (CNS) tox-
in all the cases reported in the randomized trial, unlike the more classic presentation. Other conditions associated with WE listed in the European Federation of Neurological Societies guide- lines include cancer, bariatric surgery, hyperemesis grav- idarum, and fasting and/or starvation. Although magnetic reso- nance imaging findings are highly specific and can confirm the diagnosis of WE, the majority of cases are only recognized post mortem.
In summary, the JAK/STAT pathway inhibitors are associ- ated with a myriad of neurological adverse effects, including dizziness, peripheral neuropathy, ataxia, aphasia, dysarthria, amnesia, and now, with fedratinib, Wernicke encephalopa- thy. Perhaps the relative differences in CNS adverse effects with these medications have more to do with the ability of the agent to cross the blood-brain barrier than any intrinsic differences in the metabolic pathways or mechanisms of action. How- ever, there is 1 report of a JAK2/STAT3 inhibitor causing cell cycle arrest and apoptosis of neural stem cells derived from glioblastoma,6 suggesting that an active JAK/STAT pathway does exist in the CNS that might be directly affected by such inhibition.
The investigators attempted to identify factors that placed patients at higher risk for development of WE. They observed
jamaoncology.com JAMA Oncology August 2015 Volume 1, Number 5 651