Optimal treatment strategy with nilotinib for patients with newly diagnosed chronic‐phase chronic myeloid leukemia based on early achievement of deep molecular response (MR4.5): The phase 2, multicenter N‐Road study

Abstract For patients who have chronic myeloid leukemia (CML), one of the primary treatment options is administration of nilotinib 300 mg twice daily (BID). In previous studies which compared outcomes associated with nilotinib or imatinib treatment, nilotinib achieved a higher rate of deep molecular response (MR). We conducted a phase II, open‐label, multicenter study to investigate an intrapatient nilotinib dose‐escalation strategy for patients with newly diagnosed chronic‐phase (CP) CML based on early MR4.5 achievement. The primary study endpoint was achievement of MR4.5 by 24 months following the initiation of nilotinib 300 mg BID. Fifty‐three patients were enrolled, 51 received nilotinib, and 37 completed the treatment. An increase in the nilotinib dose (to 400 mg BID) was allowed when patients satisfied our criteria for no optimal response at any time point. The median (range) dose intensity was 600 (207‐736) mg/day. Of 46 evaluable patients, 18 achieved an optimal response and 28 did not. Of the latter, nine patients underwent dose escalation to 400 mg BID, and none achieved MR4.5. The remaining 19 patients could not undergo dose escalation, 12 (63%) because of adverse events (AEs), and 7 (37%) for non‐AE related reasons. Four of these patients achieved MR4.5. The MR4.5 rate by 24 months was 45.7%. The progression‐free, overall and event‐free survival were each 97.6%. No new safety concerns were observed. Our findings support the use of continuous nilotinib at a dose of 300 mg BID for newly diagnosed patients with CML‐CP.


| INTRODUCTION
The myeloproliferative disorder chronic myeloid leukemia (CML) is characterized by the acquisition of the Philadelphia chromosome (t [9;22][q34.1;q11.2]) in multipotent hematopoietic stem cells. 1 Approximately 15% to 20% of leukemia cases are CML, that is, one out of 100,000 leukemia cases. The BCR-ABL1 fusion gene in the Philadelphia chromosome encodes and produces the BCR-ABL tyrosine kinase; in CML, this tyrosine kinase is constantly activated. This leads to leukemia cell proliferation and disease progression. [2][3][4] Without any treatment, CML in the chronic-phase (CP) progresses to the accelerated phase (AP) or blast phase (BP). Both AP and BP are usually associated with a lethal outcome within 5 years. Initial treatment of CML-CP consists of the use of one of the following three tyrosine kinase inhibitors (TKIs) as frontline therapy: imatinib, dasatinib, or nilotinib. 5 The first approved TKI for the treatment of newly diagnosed CML-CP was imatinib, based on clinical trial results in which imatinib and interferon-α plus cytarabine were compared; the study data showed that imatinib had superior efficacy without any significant changes in the safety profile. 6 Subsequent to the approval of imatinib, [6][7][8][9] the second-generation TKIs dasatinib and nilotinib 10,11 were also approved for therapeutic use. Compared with imatinib, nilotinib has been shown to have a higher binding affinity and greater selectivity for the ABL kinase. 12 In terms of its inhibitory activity, in imatinib-sensitive CML cell lines, nilotinib is 20-50 times more inhibitory than imatinib, and in imatinib-resistant cell lines, nilotinib is 3-7 times more inhibitory. [13][14][15][16][17] For patients with newly diagnosed CML-CP, the dose of nilotinib recommended in the package insert is 300 mg twice daily (BID). 18 The ENESTnd study compared the efficacy and safety of nilotinib at two doses (300 mg BID and 400 mg BID) and of imatinib 400 mg QD, in newly diagnosed CML-CP patients. 19 At 2 years, frontline nilotinib at a dose of 300 mg BID resulted in a higher rate of deep molecular response (MR) compared with imatinib (MR 4.5 log reduction [MR 4.5 ] by 2 years: nilotinib, 26% vs imatinib, 10%; P < .0001). 19 By 5 years, more than half of all patients in each nilotinib arm (300 mg BID, 54%; 400 mg BID, 52%) achieved MR 4.5 compared with 31% of patients in the imatinib arm. 19 A phase II study in untreated CML patients evaluated 73 early CP individuals to whom nilotinib was administered at a dose of 400 mg BID. The report from that study indicated that patients were able to achieve a complete cytogenetic response (CCyR) rate of 96% by 1 year and a major MR (MMR) rate of 85%. 20 The objective of this phase II, open-label, multicenter study was to investigate the optimal nilotinib treatment strategy after having started at an initial dose of nilotinib 300 mg BID in patients with newly diagnosed CP-CML based on early achievement of MR 4.5 .

| Study design, treatments, and blinding
The phase II N-Road study was an open-label, multicenter clinical study which included individuals who were newly diagnosed with CML-CP. Each patient received nilotinib 300 mg BID (600 mg/day) for a period of 24 months. In this study, increasing the nilotinib dose to 400 mg BID (800 mg/ day) was allowed if patients satisfied the criteria for lack of by 24 months following the initiation of nilotinib 300 mg BID. Fifty-three patients were enrolled, 51 received nilotinib, and 37 completed the treatment. An increase in the nilotinib dose (to 400 mg BID) was allowed when patients satisfied our criteria for no optimal response at any time point. The median (range) dose intensity was 600 (207-736) mg/day. Of 46 evaluable patients, 18 achieved an optimal response and 28 did not. Of the latter, nine patients underwent dose escalation to 400 mg BID, and none achieved MR 4.5 . The remaining 19 patients could not undergo dose escalation, 12 (63%) because of adverse events (AEs), and 7 (37%) for non-AE related reasons.
Four of these patients achieved MR 4.5 . The MR 4.5 rate by 24 months was 45.7%. The progression-free, overall and event-free survival were each 97.6%. No new safety concerns were observed. Our findings support the use of continuous nilotinib at a dose of 300 mg BID for newly diagnosed patients with CML-CP.

K E Y W O R D S
chronic myeloid leukemia, early deep molecular response, nilotinib optimal response at any time point. The criteria for lack of optimal response in this study were as follows: BCR-ABL1 IS (BCR-ABL1/ABL1 ratio on the International Scale [IS]) >10% after 3 months; BCR-ABL1 IS >1% or Philadelphia chromosome-positive rate >0% via chromosome banding analysis (G-band) or peripheral blood neutrophil-fluorescence in situ hybridization (FISH) method after 6 months; BCR-ABL1 IS > 0.1% after 12 months; two consecutive losses of MMR (BCR-ABL1 IS > 0.1%); BCR-ABL1 IS > 0.0032% after 18 months; or two consecutive losses of MR 4.5 (BCR-ABL1 IS > 0.0032%). Subjects were observed closely to detect hematotoxicity (decreased neutrophil, platelet count, and anemia), other toxicities, and QT prolongation. The dose of nilotinib could be reduced or discontinued, depending on the toxicity severity, according to dose adjustment criteria. The following four nilotinib dose reduction levels were set: Level 4, 400 mg BID (800 mg/day); Level 3, 300 mg BID (600 mg/day); Level 2, 400 mg QD (400 mg/day); and Level 1, 200 mg QD (200 mg/day).
This research was approved by the institutional review boards at Jikei University Kashiwa Hospital and other participating institutions and was conducted in accordance with the ethical guidelines for medical and health research involving human subjects. This study also adhered to the ethical principles of the Declaration of Helsinki. All patients provided written informed consent for participation prior to initiation of any study procedures. This trial was registered in the University Hospital Medical Information Network Clinical Trials Registry (UMIN000008565).

| Patients
Major inclusion criteria were as follows: patients aged ≥16 years, diagnosed with CML (Philadelphia chromosomepositive, confirmed by chromosome analysis, or having BCR-ABL mRNA confirmed by reverse transcription-polymerase chain reaction [RT-PCR]) within 6 months of study registration; patients who did not meet any AP or BP criteria; and had Eastern Cooperative Oncology Group performance status of 0 to 2.
Key criteria resulting in exclusion from the study were as follows: patients undergoing TKI treatment other than imatinib; patients undergoing imatinib treatment for ≥2 weeks; patients undergoing interferon-α treatment; patients receiving oral cancer drugs (eg, hydroxyurea) for ≥3 months; patients receiving other study drugs for CML; patients with BCR-ABL point mutations (T315I); patients with history of hematopoietic stem cell transplantation; patients with uncontrollable cardiovascular conditions (for example, individuals with hypertension, congestive heart failure, or arrhythmia); patients with acute/chronic pancreatitis; patients with a digestive disorder; patients with a congenital/acquired bleeding disorder; patients with uncontrollable diabetes or hyperlipidemia; and patients with any other uncontrollable diseases.

| Efficacy outcomes
The primary endpoint, as specified in the study analysis plan, was MR 4.5 rate by 24 months. Secondary endpoints were as follows: cumulative incidence of MR 4.5 , MR 4 , and MMR at 12 months and at 24 months; MR 4.5 rate by 12 and 18 months; rate of MR 4 and MMR by 12, 18, and 24 months; major CCyR rate by 6 and 12 months; and survival (specifically progression-free survival [PFS], overall survival [OS], and event-free survival [EFS]).
For PFS, progression was defined in this study as disease exacerbation to AP/BP and was measured from the date of registration to the date of exacerbation or death. For EFS, events were defined as AP/BP progression or loss of response (either complete hematologic response, CCyR, or partial cytogenetic response); EFS was measured from the date of registration to the date of the event or death.

| Safety
Adverse events (AEs) were collected up to the date of study treatment completion (or 30 days after discontinuation for the patients who discontinued). The reported AEs were classified and graded according to Common Terminology Criteria for Adverse Events v4.0, Japan Clinical Oncology Group.

| Assessments
Drug exposure to nilotinib was measured in terms of treatment duration, total dose received, and dose intensity in all cases, as well as by achievement or nonachievement of optimal response. Testing of BCR-ABL1 gene levels was conducted; measurements were obtained using IS quantitative RT-PCR performed in a central laboratory using a Molecular MD One-Step qRT-PCR BCR-ABL kit (BML Inc).
Following the methodology used in a previously published nilotinib study, 21 and using standardized definitions, 20 MR 4 was defined as detectable ≤0.01% BCR-ABL1 IS or undetectable BCR-ABL1 in samples with ≥10 000 ABL1 transcripts. In the case of undetectable BCR-ABL1, samples with a mean of <10 000 ABL1 transcripts, or with a total of <10 000 ABL1 transcripts, were considered not evaluable for MR 4 . MR 4.5 was defined as detectable ≤0.0032% BCR-ABL1 IS or undetectable BCR-ABL1 in samples with ≥32 000 ABL1 transcripts. 20,21 In the case of undetectable BCR-ABL1, samples with a mean of <32 000 ABL1 transcripts, or with a total of <32 000 ABL1 transcripts, were considered unevaluable for MR 4.5 . No major MR was defined as a BCR-ABL1 IS >0.1%. MMR was defined as a BCR-ABL1 IS ≤0.1%. Loss of MMR was defined in this study as an increase in BCR-ABL1 IS above >0.1%; this was reconfirmed after Week 4 by molecular genetic testing. Loss of MR 4.5 was defined as a BCR-ABL1 IS >0.0032% and this was reconfirmed after Week 4 by molecular genetic testing. Confirmation of the Philadelphia chromosome presence was made via the G-band or the FISH method. MR 4.5 duration was defined as the longest duration without MR 4.5 loss (two successive losses, BCR-ABL1 IS >0.0032%).

| Statistical methods
Descriptive statistics were used for baseline demographic and clinical characteristics, with number (%) for categorical variables and median (range) for continuous variables. Kaplan-Meier analysis was used to assess the cumulative incidence of MR 4.5 , MR 4 , MMR, PFS, OS, and EFS over 12-and 24-month periods. The proportion of patients with CMR (considered the best response) during the follow-up period (24 months after the start of nilotinib treatment) was analyzed. MR rates were analyzed by the best response at each time point. Accordingly, point estimation for the proportions and two-sided confidence intervals (CI) and 95% confidence coefficients were calculated. Using Kaplan-Meier methodology, the cumulative survival functions (ie, OS/PFS/EFS) were each determined and shown graphically. The estimated cumulative survival rate at 6, 12, 18, and 24 months and the two-sided CI with 95% confidence coefficient were also calculated. The cumulative survival function was likewise determined using the Kaplan-Meier method and depicted graphically. The time to achieve each specified event was summarized using descriptive statistics. SAS software version 9.4 (SAS Institute) was used for statistical analysis.

| Patient disposition and baseline characteristics
The study was conducted between August 2012 and July 2015 in 14 centers. Figure 1A shows the disposition of patients.

F I G U R E 1 A, Patient disposition, and
B, dose adjustment outcome. AE, adverse event; BID, twice daily; M, month; MR 4 , molecular response 4 log reduction; MR 4.5 , molecular response 4.5 log reduction; PCR, partial complete response A total of 53 patients were enrolled in this study. Of the 51 eligible patients, three patients with variant cases and two patients who withdrew from the study before PCR at 3 months were not included in the efficacy analysis; thus, 46 patients were evaluable for efficacy. Of these, 37 completed the 24month treatment, and nine discontinued treatment. The main reasons for treatment discontinuation were loss to follow-up (n = 3) and insufficient effect (n = 3).
The main characteristics and clinical features of the patients included in the analysis sets for efficacy and safety are shown in Table 1. In the safety analysis set, patients had a median (range) age of 50 (24-86) years and 35.3% were females. Patients in the safety analysis set had a predominantly low (41% and 43%) or intermediate risk (41% and 47%) according to the Sokal and Hasford risk scores, respectively. Most patients (76%) in the safety analysis set had low risk according to the European Treatment and Outcome Study risk score. As patients in this study were newly diagnosed with CML, per the inclusion criteria, the median disease duration was short (0.16 [0-2.5] months; safety analysis set). Complications such as diabetes mellitus and hypertension were present in 11% and 20% of patients (safety analysis set).

| Treatment exposure
Of the 46 evaluable patients, 18 achieved an optimal response and 28 did not ( Figure 1B). Of the latter, nine patients underwent dose escalation to 400 mg BID, but none achieved MR 4.5 . Of these nine patients, three required dose reduction/ interruption from 300 mg BID before the dose could be increased. In two of these three cases, the dose could be increased immediately at the time no optimal response was established, but a dose reduction/interruption was required before the determination of no optimal response. Response in the one remaining patient was judged as no optimal response at 3 months, but the dose could not be increased due to an AE, and further reduction/interruption was required. At 12 months, in that one patient, the dose was finally increased from 300 mg BID to 400 mg BID because of no optimal response. Moreover, of these nine patients, no case required further dose reduction or interruption after the dose was increased.
The dose could not be escalated in the remaining 19 patients-12 (63%) because of AEs and 7 (37%) because of non-AE related reasons ( Figure 1B). However, four of these patients achieved MR 4.5 . The main reasons for not undergoing dose escalation were nonhematological AEs at 300 mg BID or less (n = 9), hematological AEs at 300 mg BID or less (n = 3), and patient refusal (n = 3).

| Primary and secondary endpoints
The MR 4.5 rate by 24 months (the primary endpoint) was 45.7% (95% CI: 30.9-61.0). At 12 and 24 months, the    Table 2 summarizes the AEs of any grade and Grade 3/4 reported during the study.

| Safety
The most common nonhematologic AEs reported with an incidence of ≥10% were rash, headache, nausea, malaise, fatigue, myalgia, upper respiratory tract infection, and QTc prolongation. Of these, one case of Grade 3/4 headache and one of myalgia were reported.
The most common hematologic AEs with an incidence of ≥5% were anemia, lymphopenia, neutropenia, thrombocytopenia, leukopenia, and lymphocytosis. Of these AEs, three cases of Grade 3/4 neutropenia were reported. Other Grade 3/4 AEs of interest were pericardial effusion and myocardial infarction, reported in two patients each, and death in a single F I G U R E 2 A, Cumulative incidence of MR 4.5 , B, Cumulative incidence of MR 4 , and C, Cumulative incidence of MMR. CI, confidence interval; MR 4 , molecular response 4 log reduction; MR 4.5 , molecular response 4.5 log reduction; MMR, major molecular response patient. The patient who died seemingly had no health issues while under treatment but died suddenly. As no autopsy was performed, the cause of death was not confirmed. Other AEs were not unexpected.

| DISCUSSION
The N-Road study aimed to determine the optimal treatment strategy for nilotinib use in patients with newly diagnosed CML-CP. The MR 4.5 rate by 24 months was 45.7% in the present study, which is higher than the MR 4.5 achieved in the 24-month follow-up of the ENESTnd trial (namely, 25% with a nilotinib dose of 300 mg BID). 19,22 MMR achievement was reported for 71% of nilotinib 300 BID-treated patients in a 24month follow-up of the ENESTnd trial, 19,22 which is comparable to the 76.1% rate of MMR achievement by 24 months in the present study. The cumulative incidences of MR 4.5 , MR 4 , and MMR with 300 mg BID nilotinib at 24 months were higher in this study than that of the ENESTnd trial (51.4% vs 25%, 58.3% vs 39%, and 82.2% vs 71%, respectively). 19 In the ENEST1st trial, which assessed nilotinib 300 mg BID for patients with newly diagnosed CML-CP, 21 the cumulative incidences of MR 4.5 , MR 4 , and MMR by 24 months were 38.6%, 55.2%, and 80.4%, respectively. The response rate with nilotinib was similar to that in this study, but the cumulative incidence of MR 4.5 by 24 months was higher in the current study.
Compared with the 5-year follow-up of the ENESTnd trial, 19 we found that the incidence of MR 4.5 in individuals who had a high Sokal score was lower than that of individuals who had a low or intermediate Sokal score (28% ENESTnd, 11 18.1% ENEST1st, 21 and 13%, N-Road). The treatment efficacy in the N-Road study was different from other studies possibly because there were fewer patients with a high Sokal score in this study. Additionally, the clinical experience gained since the ENESTnd trial 11 was conducted has led to improved management of patients treated with nilotinib. This difference may have influenced the results in this study and the ENEST1st trial. 21 However, it is not possible to directly compare these results as the conditions (ie, selection criteria, patients' background, number of patients, and dose reduction criteria) differed across studies. Importantly, a dose increase of >300 mg BID was allowed in the N-Road study while this was not allowed in the ENESTnd and ENEST1st trials. 11,21 Our study had similar dose increase criteria (>300 mg BID) to two other nilotinib studies of patients with newly diagnosed CML-CP: the ENESTxtnd 23 and GIMEMA 24,25 trials. In the ENESTxtnd trial, 23 20.9% of patients had their nilotinib dose increased when they showed a suboptimal response to treatment and by 24 months, 63.6% of those patients were shown to have achieved MMR. A total of 17.6% of patients decreased doses because of AEs. The cumulative incidence of MMR was 81% by 24 months. 23 Conversely, in the GIMEMA trial, 25 77% of patients continued nilotinib up to the last observation. Most patients (68%) received nilotinib 300 mg BID and 3% received 400 mg BID. The cumulative incidences of MR 4 and MR 4.5 by 2 years were 51% and 24%, respectively. 25 Compared with those two trials, the N-Road study showed a deeper MR. The N-Road study aimed to achieve early response; thus, the dose increase criteria for no optimal response at 3, 6, and 12 months recommended by ELN 2013 was applied in this study. 26 The dose increase criteria applied in N-Road were, therefore, even stricter than those of ENESTxtnd 23 and GIMEMA trials. 24,25 Consequently, the cumulative rate of MR 4.5 by 24 months was high in our study. A total of 28 patients met the dose increase criteria, but only nine patients tolerated an increased dose. Of these nine patients, none achieved MR 4 Regarding patient background characteristics in the ENESTxtnd trial, 23 GIMEMA trial, 25 and the present N-Road study, the median age (range) was comparable. The rates of high Sokal score were no data, 23 21%, 25 and 13%, respectively. A higher MR 4.5 accumulative rate in this study may have been a result of the inclusion of fewer individuals who had a high Sokal score within our sample. The dose reduction criteria of this study were different from those applied in the ENESTxtnd trial. 23 However, in both studies, the dose discontinuation criteria were the same, and the dose intensity in this study was similar to the other studies. 23,25 The curve for the cumulative rate of MMR was steeper at 6 months in this study compared with the ENESTxtnd trial. 23 Patients in this study achieved a deep MR earlier than in some prior studies. 23 The ENESTxtnd trial results imply many patients with poor early response could achieve MMR by increasing the nilotinib dose to 400 mg BID. Of those patients who did not achieve MMR or had a loss of MMR at 12 months of treatment, it was reported that 80% achieved MMR at 24 months when the dose was increased. 23 Of patients with BCR-ABL1 IS >10% at 3 months of treatment, only one-third of patients achieved MMR. 24 BCR-ABL1 IS ≤10% at 3 months of treatment is an important milestone to achieve an optimal response. [27][28][29] In the ENEST1st trial, 21 none of the patients with BCR-ABL1 IS >10% at 3 months of nilotinib treatment achieved MR 4.5 at 2 years; 14.5% and 45.8% of patients with BCR-ABL1 IS >1% to ≤10% and BCR-ABL1 IS > 0.0032% to ≤1% at 3 months of nilotinib treatment achieved MR 4.5 at 2 years, respectively. BCR-ABL1 IS >0.0032% to ≤1% at 3 months of treatment seems to be a significant factor for the achievement of early MR 4.5 . The rates of BCR-ABL1 IS >0.0032% to ≤1% at 3 months in the ENEST1st 21 and N-Road studies were 78.5% and 58.1%, respectively; those with BCR-ABL1 IS >1% to ≤10% were 18.5% and 30.2%, respectively; and those with BCR-ABL1 IS >10% were 2.9% and 11.6%, respectively. The rate of MR 4.5 at 2 years in the N-Road study was higher than that in the ENEST1st trial although the rate of BCR-ABL1 IS >10% was higher and that of BCR-ABL1 IS >0.0032% to ≤1% was lower in the N-Road study; however, the reason for this difference in results is unknown. The safety of nilotinib in this study was similar to that reported in previous studies. 19,21,25 In this study, two patients (3.9%) presented cardiovascular events, both grade 3 myocardial infarction. We did not record any cases of ischemic cerebrovascular disease. In the ENEST1st trial, 21 during 2 years of treatment, 3.4% of patients had ischemic heart disease, and 0.8% had ischemic cerebrovascular disease. In the GIMEMA trial, 25 4.6% of patients presented an arterial thromboembolic event. In this study, patients who had myocardial infarction or unstable angina pectoris within the prior 12 months immediately before study registration were excluded. After two patients presented myocardial infarction in this study, patients with congestive heart failure, uncontrollable diabetes, or hyperlipidemia were excluded from this study. After this change, no obstructive cardiovascular events were observed. Thus, we consider that cardiovascular events may be prevented by controlling risk factors such as diabetes, hyperlipidemia, blood pressure, and smoking, 30 making it possible to continue nilotinib treatment.
This study had some limitations, such as those related to the open-label nonrandomized study design. Additionally, the sample size was limited, and the study lacked a comparator; thus, the generalizability of the results may be limited. Finally, this study aimed to assess the optimal treatment strategy with nilotinib; however, dose escalations were not possible in several patients, which precluded dose-escalation analyses.
Notably, there were many cases in which an early dose increase was difficult because of AEs. In other cases, even if the dose could be increased, MR 4.5 could not be reached in the 24-month period. While the results of this study do not support an early dose increase of nilotinib for individuals who do not manage to achieve an optimal response, most patients in our study received nilotinib 300 mg BID. By optimizing the nilotinib dose in cases requiring dose increases or decreases, and by managing AEs, treatment continuation was successful at this dose. The results of the N-Road study suggest the importance of continuing nilotinib treatment at a dose of 300 mg BID, as it may lead to the achievement of early deep MR in patients with newly diagnosed CML-CP.