Janus kinase inhibitor ruxolitinib in combination with nilotinib and prednisone in patients with myelofibrosis (RuNiC study): A phase Ib, multicenter study

Abstract This phase Ib, non‐randomized, open‐label study evaluates the safety and tolerability of ruxolitinib in combination with nilotinib and prednisone in patients with naïve or ruxolitinib‐resistant myelofibrosis (MF). A total of 15 patients with primary or secondary MF received the study treatment; 13 patients had received prior ruxolitinib treatment (86.7%). Eight patients completed seven cycles (53.3%) and six patients completed twelve cycles of treatment (40%). All the patients experienced at least one adverse event (AE) during the study (the most common AEs were hyperglycemia, asthenia, and thrombocytopenia), and 14 patients registered at least one treatment‐related AE (the most common treatment‐related AEs were hyperglycemia (22.2%; three grade 3 cases). Five treatment‐related serious AEs (SAEs) were reported in two patients (13.3%). No deaths were registered throughout the study. No dose‐limiting toxicity was observed. Four out of fifteen (27%) patients experienced a 100% spleen size reduction at Cycle 7, and two additional patients achieved a >50% spleen size reduction, representing an overall response rate of 40% at Cycle 7. In conclusion, the tolerability of this combination was acceptable, and hyperglycemia was the most frequent treatment‐related AE. Ruxolitinib in combination with nilotinib and prednisone showed relevant clinical activity in patients with MF. This trial was registered with EudraCT Number 2016‐005214‐21.


INTRODUCTION
Myelofibrosis (MF) is characterized by the clonal expansion of myeloid cells due to the acquisition of driver mutations in the JAK2, CALR, or MPL genes, and/or other mutations such as those in ASXL1, SRSF2, and U2AF1 [1,2].
Ruxolitinib is a reversible and selective inhibitor of Janus kinase-1 (JAK1) and JAK2/signal transducer and activator of transcription (STAT) signaling that has been approved in the European Union for the treatment of disease-related splenomegaly or symptoms in adult patients with de novo MF, MF arising from previous polycythemia vera, or essential thrombocythemia (MF) [3][4][5][6][7][8][9][10][11]. Ruxolitinib achieves spleen reduction and symptom control in a significant proportion of patients, leading to improved quality of life and possible survival benefits.
A recent in vitro study investigated the administration of ruxolitinib in combination with nilotinib and prednisone to treat patients with MF [12] and showed that ruxolitinib had a synergistic effect with nilotinib and prednisone in MF patient samples and cell lines, partly mediated through significant inhibition of the STAT5 and protein kinase B (AKT/PKB) pathways. Moreover, the combination of ruxolitinib and nilotinib has proven efficacy in chronic myeloid leukemia, and nilotinib has shown antifibrotic effects in the liver and lung in animal models [13][14][15][16].
The present phase Ib clinical trial was a two-arm study designed to obtain and assess the maximum tolerated dose (MTD) and/or recommended phase III dose (RP3D) of ruxolitinib in combination with nilotinib and prednisone in two MF separate populations: patients naïve to JAK-inhibitor treatment and patients not responding to or relapsed after JAK-inhibitor treatment. The study was also designed to provide preliminary efficacy and safety data for ruxolitinib when administered in combination with nilotinib and prednisone in MF patients.

Patient population
This study consisted of two cohorts: patients naïve to JAK-inhibitor treatment and patients who did not respond or relapsed after JAK-

Study design
RuNiC is a phase Ib/II, open-label, multicenter, dose-finding study on ruxolitinib in combination with nilotinib (300 mg twice a day (BID)) and prednisone (50 mg every other day) for patients with MF [17][18][19]. The dose-escalation phase (first six cycles, 28-day cycles) initially planned with ruxolitinib starting from 10 mg BID was based on phase II studies [13], according to the classic (non-Bayesian) model [20,21]. Next, a dose-expansion phase has been planned (12−24 cycles, 1−2 years), whereby an additional 10 patients in each arm (20 patients in total) will be treated according to their respective arm ( Figure 1). For patients with prior ruxolitinib, no washout phase was needed.
A dose-limiting toxicity (DLT) was defined as any treatment-related toxicity occurring within the first 28 days (from Cycle 1 on Day 1 to Cycle 1 on Day 28) of treatment with ruxolitinib, nilotinib, and prednisone and that met any of the DLT criteria (Table S1).

Assessments
The primary objective was to determine the MTD and the RP3D of ruxolitinib in combination with nilotinib and prednisone. The

Statistical analysis
Quantitative variables are described with measures of central tendency and dispersion: mean, median, SD (standard deviation), Q1 (first quartile) and Q3 (third quartile), and minimum and maximum. Qualitative variables are described using absolute and relative frequencies (N, %). The description only concerned the descriptive analysis of the data.
No statistical modeling procedures were performed. The analysis was performed using the IBM SPSS Statistics software, Version 22.0 (IBM Corp. Armonk, NY).

Ethical approvals
This study was approved by the institutional review board of the Hos-

Study population
From

Demographic, clinical, and MF baseline characteristics
The demographic and clinical characteristics of the patients are described in Table 1.

Study treatment dosing
The maximum tolerated doses were 20 mg twice a day (BID) of ruxolitinib, 300 mg BID of nilotinib, and 50 mg EOD of prednisone, at Cycle 12.
The dose and duration for each treatment were analyzed for study

Safety
All the patients experienced at least one AE throughout the study.
Fourteen patients (93.3%) registered at least one AE related to any of the study treatments.
A summary of the main all-grade AEs and grade 3−4 AEs is shown in  Overall, 251 AEs were reported, among which 90 were treatmentrelated AEs (Table S4). The most common treatment-related AEs (≥5% of the total) were hyperglycemia (22.2%), thrombocytopenia (13.3%), anemia (8.9%), and elevated ALT (6.7%). Temporary interruption of treatment occurred for five (33.3%) patients; dose adjustment was reported for one (6.7%) patient, and permanent discontinuation of the treatment was registered for three (20%) patients.
From the start of the study, a total of five treatment-related SAEs were reported by two patients (13.3%): one patient had pericardial effusion and bilateral pleural effusion (both conditions related to nilotinib), and one patient had congestive heart failure, pleural effusion, and pulmonary hypertension (all conditions related to nilotinib). No deaths were registered throughout the study.
For the patients who discontinued due to nilotinib toxicity, both passed away 9 and 5 months later due to progression of the disease and splenic radiotherapy complications.

Spleen length reduction
The spleen reduction is shown in Figure 2B,C. The Figure

MF-SAF
The descriptive analysis of each MF-SAF individual item and total score at screening and Cycles 4, 7, and 12, are shown in Table 3.
As observed, the individual symptom values of the MF-SAF median scores improved from screening to Cycles 4, 7, and 12, for most of the items. An increase in symptoms related to concentration difficulties and itching was observed at Cycle 12. From the initial 15 patients, only six had symptoms evaluated at screening and Cycle 7 ( Figure 2D).

TA B L E 3
Median MPN-SAF TSS items and total score, at screening and cycles 4, 7 and 12. Among these six patients, four patients experienced a total symptom score reduction (one patient reported a 50% reduction, two patients reported 14.3% reductions, and one patient reported a 12.5% reduction) from baseline to the end of Cycle 7. By contrast, one patient had the same total symptom score, and two patients reported increases (20% and 50%, respectively).

Ex vivo assays pharmaflow chemosensitivity profiling
We analyzed the correlations between the treatment with ruxolitinib in combination with nilotinib and prednisone by an ex vivo PharmaFlow study in samples from nine treated patients, and the patients' clinical responses ( Figure S1). This was feasible in eight out of the fifteen cases evaluated (53.3%). The ex vivo classification correlated with the clinical response to Cycle 7, since 75% of the samples from clinical responders were classified as sensitive. By contrast, 50% of the samples from patients who did not clinically respond to Cycle 7 were classified ex vivo resistant.

3.3.4
Mutational profile by NGS Mutational profiling was performed by NGS with a custom panel (Table   S5) in 10 cases. Driver mutations were detected in the JAK2 (seven cases, 46.7%), calreticulin (CALR) (two cases, 13.3%), and thrombopoietin (MPL) genes (one case, 6.7%). The mutational landscapes of the responders and nonresponders in Cycle 7 are shown in Figure S2.
Among the seven cases with mutated JAK2, five cases were responders; both cases with mutated CALR were responders; and the only case with mutated MPL was a non-responder. The most commonly comutated genes were the JAK2 gene with BCORL1 (two cases) and TET2 (two cases) and the CALR gene with DNMT3A (two cases) and TET2 (three cases). The MPL gene co-mutated with BCORL1, DNMT3A, PHF6, and SH2B3 (one case). The distribution of the most frequent driver mutations in relation to the Cycle 7 response was as follows: among the four cases with mutated TET2, three were responders to Cycle 7, and among the four cases with mutated BCORL1, two were responders to Cycle 7.

DISCUSSION
The main objectives of this phase Ib/II trial were to determine the MTD, toxicity, clinical efficacy, and biological activity of ruxolitinib in combination with nilotinib and prednisone in MF patients, naïve or previously treated with JAK inhibitors. This combination was safe and tolerable with relevant clinical activity in these patients.
Since MF is a complex disease in which tyrosine-kinase-mediated signaling pathways are involved in the production of hematopoietic clones and with the fibrosis of the bone marrow, the use of drugs in combination against different targets may be beneficial. Previous data have shown that ruxolitinib combined with prednisone and nilotinib exhibits synergistic effects in human cell lines and primary cells from myeloproliferative neoplasms (MPN) [29]. Ruxolitinib showed the ability to diminish or stabilize fibrosis. Ruxolitinib plus nilotinib showed a synergistic behavior that blocked colony formation in patients' primary cells and inhibited the phosphorylation of STAT5 and ERK1/2. In addition, the combination with prednisone increased this synergy and inhibited the synthesis of collagen in bone marrow mesenchymal cells [29].
Moreover, ex vivo PharmaFlow analyses in samples of patients treated with the combination showed a trend of correlations between the sensitive ex vivo classification and clinical responses, but a larger sample is needed to confirm these results. Regarding the influence of the mutational profile and the response at Cycle 7, five of the seven cases with a mutated JAK2 gene and the two cases with a mutated CALR gene were determined to be responders. On the contrary, the only case with a mutated MPL gene was a non-responder. Although the series is small, the mutated genes associated especially with the nonresponder group were PHF6, SRSF2, SH2B3, and ZRSR2, and we observed no difference in response between JAK2 and CALR patients.
The responses to ruxolitinib treatment are known to be transitory.
Ruxolitinib rechallenge has shown limited effectiveness after a period of treatment interruption [30,31]. JAK2-V617F leukemia cell lines could be re-sensitized to ruxolitinib after a period of withdrawal [31].
Nine of thirteen MF patients retreated with ruxolitinib after the loss of the initial response or an inadequate response registered significant spleen size reductions, and twelve of thirteen patients reported symptom improvement [30]. In this case series, ruxolitinib was ultimately discontinued in seven patients due to a loss of response, inclusion in another clinical trial, or intolerance to the treatment [30].
The main limitation of our study is that, due to the number of patients finally enrolled, it was not possible to comply with the evaluation characteristics of the MTD, and therefore, no data regarding the main objective of the trial were obtained. Another limitation is that seven intermediate-1-risk patients (46.7%) were included, making it difficult to assess the efficacy as compared with other studies. No differences in response between naïve and pretreated patients could be observed, probably due to the low number of patients. However, our study provides relevant information on the efficacy and safety of the combination, its correlations with ex vivo data, and the mutational landscape of the patients.
In conclusion, the current results provide evidence that ruxolitinib in combination with nilotinib and prednisone has therapeutic activity in MF, with a good safety profile.

DATA AVAILABILITY STATEMENT
Questions regarding data sharing should be addressed to the corresponding author. For original data and study protocol, please contact Prof. Joaquín Martinez-Lopez at jmarti01@med.ucm.es. Deidentified individual participant data that underlie the reported results will be made available 3 months after publication for a period of 5 years after the publication date.