First-line therapy for chronic phase CML: selecting the optimal BCR-ABL1-targeted TKI

Abstract

Patients diagnosed with chronic myeloid leukemia (CML) and treated with BCR-ABL1 tyrosine kinase inhibitors (TKIs) have long life spans. Selection of an appropriate first-line therapy can be difficult as both the unique characteristics of each TKI and patient need to be taken into account to find the optimal match. Patient characteristics include comorbidities, concomitant medications, lifestyle, risk factors, BCR-ABL1 transcript type (e.g. b2a2 or b3a2) and additional chromosomal abnormalities. Just as patients differ, side effects, drug-drug interactions, administration plans, dosing schedules and treatment-related expenses across TKIs also vary. Alignment of these characteristics with the appropriate TKI is key to successfully initiating CML treatment. Continued success relies on communication between the patient and the healthcare team, adherence and optimization of therapy once it is initiated. In this review, we discuss these factors, in addition to TKI efficacy and safety, the cost of therapy, the future of treating CML and treatment-free remission.

Keywords:

• Myeloid leukemias and dysplasias
• myeloproliferative disorders
• signaling therapies

Introduction

Before targeted therapy, patients with chronic myeloid leukemia (CML) were commonly treated with interferon-α (IFN-α) plus cytarabine, providing patients three- to five-year overall survival (OS) rates ranging from 86% to 68% [1–3]. Response rates were improved with this combination over IFN-α; however, the addition of cytarabine was believed to cause an increase in adverse events (AEs) [2]. Therapy for CML radically changed with the introduction of BCR-ABL1 tyrosine kinase inhibitors (TKIs) [4]. At the moment, three TKIs (imatinib, dasatinib and nilotinib) are approved for the first-line treatment of CML in chronic phase (CP) [5–7]. Patients with CML and treated with BCR-ABL1 TKIs can experience a life span comparable to the non-CML population, with TKIs having substantially increased the OS of patients from that with combination chemotherapy [8–10].

The variety of options available for the treatment of CML-CP can make the selection of an appropriate first-line therapy difficult. Individual patient characteristics, such as preexisting comorbidities, concomitant medications, patient age and risk factors for disease progression, all influence selection of an effective therapy that also allows for a high quality of life.

Efficacy of TKIs in first-line clinical trials

Imatinib

The first BCR-ABL1 TKI to receive approval for first-line treatment of CML-CP was imatinib in 2002 [5], after improved efficacy was observed in the IRIS trial comparing imatinib with IFN-α plus cytarabine [4]. From the 80% of patients on study at 10 years with known survival status, the imatinib arm had a best estimated 10-year OS rate of 83% and a best observed major molecular response (MMR; BCR-ABL1 ≤ .1% International Scale [IS]) rate of 93% [9,11]. Efficacy outcomes from additional first-line trials with imatinib and other BCR-ABL1 TKIs are described in Table 1 [1–3,8,10–24] and report superior OS rates compared to IFN-α plus cytarabine (68%–86%) [1–3].

Table 1. Efficacy of BCR-ABL1 TKIs in first-line clinical trials in patients with CML.

	Follow-up^a^,^b	BCR-ABL1≤10% at 3 months (%)	MMR (%)	MR^4.5 (%)	OS (%)	EFS or PFS (%)	Transformation to AP/BP (%)
IFN-α + cytarabine
Italian study [1] (N = 538)	5 years^a	–	–	–	68	–	–
MDACC [3] (N = 453)	42 months^b	–	–	–	70^c (4 years)	–	–
French study [2] (N = 745)	35 months^b	–	–	–	86^c (3 years)	–	–
Imatinib
IRIS [11] (N = 1106)	10 years^a	–	93^d	63	83^c^,e	92^c	6.9
TOPS [12] (N = 476)	42 months^b	–	76^f	–	94^c (4 years)	94^c (4 years)	–
Hammersmith [13] (N = 204)	38 months^b	–	50^f	–	83^c (5 years)	81^c (5 years)	–
CML IV^g [14,15] (N = 1551)	7 years^b	72	89^f (10 years)	72^f (10 years)	84 (10 years)	82 (10 years)	6.7
Dasatinib
DASISION [8] (N = 519)	5 years^a	84	76^f	42^f	91^c	85^c	4.6
MDACC [16] (N = 107)	12 months^a	94	90	–	95^c (5 years)	92^c (5 years)	5.0
S0324 [17] (N = 246)	3 years^b	–	47^f (1 year)	21^f (1 year)	97	93	0.8
Nilotinib
ENESTnd [10,18] (N = 846)	5 years^a	91	77^f	54^f	94^c	92^c	0.7
ENESTndChina [19] (N = 267)	22 months^b	82	68^e	–	99^c (2 years)	92^c (2 years)	–
MDACC [20] (N = 148)	59 months^b	–	93^e	83^e	88^c (7 years)	77^c (7 years)	–
GIMEMA [21] (N = 73)	6 years^a	–	–	–	96	96	–
Bosutinib
BELA [22] (N = 502)	28 months^b	86	59^e	–	97	92	1.6
Ponatinib
EPIC^h [23] (N = 307)	12 months^a	94	80^f	ND	ND	ND	ND
Radotinib
Korean study [24] (N = 241)	12 months^a	–	52	15	–	–	0

^a Minimum follow-up.

^b Median follow-up.

^c Estimated.

^d Best MMR.

^e At 10 years, 20% of patients had unknown survival status.

^f Cumulative.

^g Results from combining all five treatment arms: imatinib 400 mg, imatinib 800 mg, imatinib after interferon, imatinib 400 mg + interferon and imatinib 400 mg + cytarabine.

^h Due to early termination of the trial, long-term outcomes could not be assessed and eight out of 10 patients were assessed for MMR.

AP/BP: accelerated phase/blast phase; CML: chronic myeloid leukemia; EFS: event-free survival; IFN: interferon; MDACC: MD Anderson Cancer Center; MMR: major molecular response; MR^4.5: BCR-ABL1 (International Scale [IS]) ≤.0032%; ND: not determined; OS: overall survival; PFS: progression-free survival; TKI: tyrosine kinase inhibitor.

Dasatinib

The second-generation TKI, dasatinib, was initially approved in 2006 for treatment of patients with CML-CP or CML in accelerated or blast phase (AP/BP) who are resistant/intolerant to prior therapy (that included imatinib), followed by first-line approval in 2010 [6]. Results from the phase-3 DASISION study demonstrated high efficacy and tolerability of dasatinib in patients with newly diagnosed CML-CP [8]. One-year results from DASISION included a higher confirmed complete cytogenetic response (confirmed CCyR; 0 Ph + metaphases on two consecutive assessments ≥28 days apart [77% vs. 66%]) and cumulative one-year MMR (46% vs. 28%) for dasatinib versus imatinib [25]. The cumulative five-year MMR for dasatinib (76%) (Table 1) remained higher than with imatinib (64%); however, the five-year OS and progression-free survival (PFS) rates for dasatinib were 91% and 85%, respectively, and did not differ from that observed with imatinib [8].

A key endpoint of molecular response is being investigated in trials with first-line dasatinib [16,17,26]. In a MDACC study, patients are being treated with dasatinib 100 mg qd or 50 mg bid [16]. For patients with a minimum of 12 months of follow-up, the primary endpoint of MMR was reached by 94 of 105 (90%) patients and complete molecular response by 59 (56%) patients. In the randomized S0325 study, patients within six months of diagnosis and naive to TKIs received dasatinib 100 mg qd or imatinib 400 mg qd [17]. At one year, molecular responses of 3-, 4- and 4.5-log reductions in BCR-ABL1 levels were 59%, 27% and 21%, respectively [13]; as with DASISION, the three-year OS and PFS were comparable between arms. The SPIRIT2 study randomized patients to receive imatinib 400 mg qd or dasatinib 100 mg qd [26]. Significantly more dasatinib- versus imatinib-treated patients achieved MMR (58% vs. 43%, p < .001) or MR^4.5 (BCR-ABL1 ≤.0032% [IS]; 13% vs. 6%, p = .001) at 12 months [26]; however, no significant differences in rates of progression or OS were observed after a median of 37 months [27].

Nilotinib

Nilotinib is a second-generation TKI that received approval for the treatment of CML-CP or advanced CML resistant/intolerant to prior therapy (that included imatinib) in 2007 and was approved for the first-line treatment of CML-CP in 2010 [7]. Approval of nilotinib for front line treatment occurred after positive results from the ENESTnd study, which treated patients newly diagnosed with CML-CP with nilotinib or imatinib [10,28]. In ENESTnd, the cumulative one-year MMR was higher in patients who received nilotinib 300 mg bid versus imatinib (51% vs. 27%) [28]; a gap that was maintained long term (cumulative five-year MMR: 77% vs. 60%) (Table 1) [10].

In a trial at MDACC, patients with newly diagnosed CML were treated with nilotinib 400 mg bid and had an MMR rate of 93% at seven years [20]. In the single-arm GIMEMA study, MMR was stable in 70 of 73 patients treated with nilotinib 400 mg bid, with a six-year probability of achieving MMR of 98% [21]. Nilotinib-treated Chinese patients from ENESTnd had a 12-month MMR of 52%, and 68 of 134 (51%) patients in the nilotinib arm achieved MMR at 24 months [19].

Overall, despite differences in molecular responses comparing dasatinib or nilotinib with imatinib, the OS rates across these agents are not statistically significantly different [8–10]. Further, propensity score matching of 87 dasatinib- or nilotinib-treated patients reported no major differences in response or survival endpoints [29]. One factor confounding analysis of outcomes for the treatment of CML is that many deaths occur from causes unrelated to CML [8–10]. Additional follow-up will provide details for how TKIs affect outcomes beyond five and seven years; however, due to the chronic nature of CML and long life spans with TKI therapy, the extent of follow-up required may be difficult to implement in a trial. As these studies are developed, patient risk factors are additional considerations, as lower-risk patients have better cytogenetic responses [9], faster and deeper molecular responses [8,10], and improved PFS [9,10] and OS [10] compared to high-risk patients.

Bosutinib and ponatinib

Additional BCR-ABL1 TKIs approved for the treatment of patients with CML resistant/intolerant to first-line therapy have also been investigated in front line clinical trials (Table 1). Front line use of bosutinib was investigated in the phase-3 BELA trial [22]. After a median treatment period of 28 months, cumulative CCyR rates at 24 months were 79% for bosutinib and 80% for imatinib; as the BELA study failed to meet its primary endpoint of CCyR at 12 months, bosutinib was not registered as a first-line therapy for patients with CML [22]. Bosutinib at 400 mg qd versus imatinib is currently being investigated in the phase-3 BFORE study in patients newly diagnosed with CML-CP; the primary endpoint of MMR at 12 months was reached, bosutinib was superior to imatinib and safety outcomes were as expected [30].

High molecular response rates were reported with ponatinib throughout the first 12 months of the phase-3 EPIC study [23]; however, fewer patients were on study treatment at later time points, and final outcomes could not be assessed due to early termination of the trial after the identification of arterial occlusive events in ponatinib-treated patients [31].

Risk score

Prognostic scoring of patients uses patient age, spleen size and blood cell counts to assess risk of progression and can assist physicians in selecting a TKI [32]. Currently, there are three commonly used scoring systems: Sokal [33], EURO (Hasford) [34] and EUTOS [35]. The Sokal and EURO methods were developed in patients who mainly received single-agent chemotherapy [33] or patients treated with IFN [34], respectively, and utilize several factors in their calculations [33,34]. In contrast, the EUTOS method uses only basophil count and spleen size in its calculations, and was developed using data from patients treated with TKIs [33].

Risk scores can be used to stratify patients and evaluate their potential association with response rates [8–10,22,23]. A trend between lower risk score and improved cytogenetic or molecular response was observed in the IRIS [9], DASISION [8], ENESTnd [10,28], BELA [22] and PACE [23] trials (Table 2). Overall, data support that low-risk patients are likely to respond well to any of the approved front line TKIs and patients in intermediate- and high-risk categories benefit more from first-line treatment with dasatinib or nilotinib [8,10,36]. For example, retrospective analysis of data from DASISION reported more patients with a high EURO risk score to have achieved BCR-ABL1≤1% at three and six months when treated with dasatinib (33% and 60%) versus imatinib (5% and 29%) [37]. In this same study, transformation-free survival was associated with the achievement of BCR-ABL1≤10% at three and six months for both TKIs [37]. In ENESTnd, nilotinib- versus imatinib-treated patients with intermediate (93% vs. 88%) or high (86% vs. 83%) Sokal risk scores had improved five-year PFS [10].

Table 2. Response according to risk score assessment.

			Baseline assessment (%)			Cytogenetic response (%)			Molecular response (%)			Progression or PFS (%)
Study (treatment)	Assessment method	Time of assessment	Low	Int	High	Low	Int	High	Low	Int	High	Low	Int	High
IRIS [4,9] (imatinib)	Sokal	5 years	52	29	19	89	82	69		–		3^a	8	17
DASISION [8,21] (dasatinib)	EURO	5 years	33	48	19	94	78	78	90^b 55^c	71 43	67 31		–
ENESTnd [10,28] (nilotinib)	Sokal	5 years	37	36	28		–		53^c	60	45	96^d	93	86
BELA [17] (bosutinib)	Sokal	24 months	35	47	18	69	55	47	58^b	42	38		–
EPIC [18] (ponatinib)	Sokal	12 months	41	41	17		–		44^b 10^c	39 13	37 30		–

^a Risk of disease progression.

^b MMR.

^c MR^4.5.

^d PFS.

Int: intermediate; MMR: major molecular response; MR4.5: BCR-ABL1 (International Scale [IS]) ≤.0032%; PFS: progression-free survival.

Determining the appropriate scoring method for patients is difficult and evolving. Two analyses using the EUTOS method were unable to significantly predict patient outcomes [38,39]. The EUTOS long-term survival (ELTS) score was recently developed based on patients in the original EUTOS registry with the additional criteria of having b2a2 and/or b3a2 transcripts and imatinib treatment within six months of diagnosis with CML [40]. The ELTS score specifically considers long-term survival based on CML-related death and strongly differentiates between risk groups. The predictive capabilities of the ELTS method is supported by analyzes in patients treated with nilotinib [41] or imatinib [42].

BCR-ABL1 transcript type

The location of the translocation breakpoints in chromosomes 9 and 22 can generate various BCR-ABL1 transcripts [43], which could become another prognostic factor for consideration during TKI selection. The two most common transcripts are e13a2 (b2a2) and e14a2 (b3a2), expressed independently or together and occur in >95% of patients with CML.

Significantly longer event-free survival (EFS) and transformation-free survival, in addition to improved cytogenetic and molecular responses, have been observed in TKI-treated patients expressing only b3a2 transcripts versus those who expressed b2a2 transcripts [43]. Study data have indicated that characterization of BCR-ABL1 transcripts may be useful for minimizing false-negative assessment of patients with minimal residual disease, despite the limited number of patients with atypical transcripts [44]. The role of transcript type in efficacy of TKI therapy is in the nascent stages, and more data are needed to clarify their role in assessing patient disease status.

Additional chromosomal abnormalities (ACAs)

Chromosomal changes, in addition to BCR-ABL1 generation, occur in approximately 5% of patients diagnosed with CML [45]. It is believed that the accumulation of ACAs promotes progression to CML-AP and CML-BP [46] and they are included as ELN criteria to define advanced stages of CML [47]. In an analysis of patients from the CML IV Study, it was reported that patients with major ACAs (e.g. trisomy 8, a second Ph chromosome, isochromosome 17q10) had a significantly poorer prognosis than patients without these genetic variations and that this may be an important prognostic factor in identifying patients who should be considered for early stem cell transplant (SCT) [46].

Safety aspects of BCR-ABL1 TKIs

Safety profiles

Each of the BCR-ABL1 TKIs has its own unique structure, resulting in non-BCR-ABL1 targets and associated AEs (Table 3) [5–7,31,48]. The type and severity of AEs, preexisting comorbidities, and how these may affect each individual patient are considerations when prescribing a TKI [5–7,31,48]. For example, a strong association between number of comorbidities and OS was observed in imatinib-treated patients from the CML IV study; as a patient’s Charlson Comorbidity Index (CCI) increased, their probability for OS at eight years decreased [49].

Table 3. Frequently experienced (in ≥10% of study patients) nonhematologic AEs while on first-line TKI therapy.

All grade AEs, %	Imatinib [5] 400 mg qd^a (n = 551)	Dasatinib [6] 100 mg qd^b (n = 258)	Nilotinib [7] 300 mg bid^c (n = 279)
Gastrointestinal disorders
Abdominal pain	37	11	15
Abdominal pain, upper			18
Constipation	11	NR	20
Diarrhea	45	22	19
Dyspepsia	19	NR	10
Nausea	50	10	22
Vomiting	23	5	15
General/other disorders
Asthenia			14
Fatigue	39	11	23
Pyrexia	18	NR	14
Fluid retention	62	38	NR
Pleural effusion	NR	28	NR
Superficial edema	60	14	NR
Hypertension	NR	NR	10
Hemorrhage	29	8	NR
Rash and related terms^d	40	14	38
Weight increase	16	NR	NR
Infections and infestations
Influenza	14	NR	13
Nasopharyngitis	31	NR	27
Sinusitis	11	NR	NR
Upper respiratory tract infection	21	NR	17
Musculoskeletal and connective tissue disorders
Arthralgia/joint pain	31	7	22
Back pain	NR	NR	19
Bone pain	11	NR	NR
Muscle cramps	49	NR	NR
Muscle spasms	NR	5	12
Musculoskeletal pain	47	14	NR
Myalgia	24	7	19
Pain in extremity	NR	NR	15
Nervous system disorders	NR	NR	NR
Dizziness	19	NR	12
Headache	37	14	32
Psychiatric disorders
Depression	15	NR	NR
Insomnia	15	NR	11
Respiratory disorders
Cough	20	NR	17
Dyspnea	NR	NR	11
Pharyngolaryngeal/oropharyngeal pain	18	NR	12

^a Adverse reactions reported in ≥10% of imatinib-treated patients, regardless of relationship to study drug, in a clinical trial of imatinib versus IFN + Ara-C [5].

^b Adverse reactions reported in ≥10% of dasatinib-treated patients with newly diagnosed chronic phase CML (minimum of 60 months’ follow-up) [6].

^c Most frequently reported (≥10%) nonhematologic adverse reactions (regardless of relationship to study drug) in patients with newly diagnosed Ph+ CML-CP and treated with nilotinib 300 mg twice daily [7].

^d Includes related terms (e.g. erythema, rash, skin exfoliation) for imatinib and dasatinib and rash only for nilotinib [5–7].

AE: adverse event; Ara-C: cytarabine; CML: chronic myeloid leukemia; CML-CP: chronic myeloid leukemia in chronic phase; IFN: interferon; NR: not reported; TKI: tyrosine kinase inhibitor.

Use of dasatinib is associated with an increased incidence of pleural effusion [6]; patients with known pleural effusion at baseline have been excluded from clinical trials with dasatinib [8]. Patients with risk factors for developing pleural effusion (e.g. hypertension, prior cardiac history, bid dasatinib dosing, prior autoimmune disease) [50,51] should be closely monitored. Also, a different second-generation TKI may be preferential in patients with a history of lung disease or risk of pleural effusion [36]. Any-grade drug-related pleural effusion occurred in 28% (grade-3/grade-4: 3%) of dasatinib-treated patients in DASISION and were largely managed with dose modifications, diuretics and/or corticosteroids; 6% of patients discontinued dasatinib due to pleural effusion [8]. Pulmonary hypertension (PH) was reported in 5% of dasatinib-treated patients; 2% of all dasatinib-treated patients discontinued due to PH [6]. More deaths due to infection occurred in dasatinib- versus imatinib-treated patients in DASISION; the majority being over two months after dasatinib was discontinued [8]. A risk of infection with dasatinib has not been previously reported [16,27,52].

The five-year report of the ENESTnd trial described an increased risk of cardiovascular events (including arterial vascular events) in patients newly diagnosed with CML-CP who received nilotinib (9%) versus imatinib (3%) [7]. A black box warning is provided in the prescribing information for nilotinib associating its use with long QT syndrome [7]. Patients with any known cardiovascular risk factors should be closely monitored and drug dose modifications and interruptions managed as directed. Additionally, it is recommended that patients with hepatic impairment taking nilotinib should be prescribed a reduced dose [7]. A TKI other than nilotinib may be preferred in patients with a history of arrhythmias, heart disease, pancreatitis or hyperglycemia [36].

Imatinib-related chronic AEs

Assessment of chronic AEs in patients treated with imatinib for approximately two years was investigated in the ILTE study [53]. In this study, the proportion of patients with a nonserious imatinib-related AE increased over time, from 34% at six years to 53% at eight years of treatment. Women tended to experience a higher proportion of these AEs [53], which corresponds to findings that imatinib-treated women (and patients 18–39 years old) report greater quality of life impairments than their counterparts [54] and that front line therapy selection should also consider how sex and age may impact a patient’s quality of life.

The concept of switching from imatinib to a second-generation TKI to improve or resolve chronic, low-grade imatinib-related AEs was recently investigated in two separate studies. The majority of chronic, grade-1/grade-2 imatinib-related AEs were reported to resolve/improve in the DASPERSE (77%) [55] and ENRICH (69%) [56] studies, which switched patients from imatinib to dasatinib or to nilotinib, respectively. Importantly, the molecular responses of patients who switched TKIs were either maintained or improved, suggesting there is the potential for both improved efficacy and quality of life for patients who take second-generation TKIs following imatinib intolerance.

Adherence

Patients who adhere to their TKI medication plan, versus nonadherent patients, can achieve improved early responses and long-term outcomes and avoid higher healthcare costs [57]. High medication adherence correlates with meeting cytogenetic [58] and molecular [59] milestones defined by NCCN Guidelines^®, which recommend evaluating patient compliance throughout treatment if defined milestones are not met [36]. Nonadherence has been shown to be significantly associated with higher medical bills (e.g. more/longer hospital visits, more concomitant medications [60].

A straightforward treatment regimen can promote adherence and improve patient quality of life [61]. The guidelines across the BCR-ABL1 TKI family vary in their administration and dosing recommendations [5–7,31,48]. All of the first- or second-line approved BCR-ABL1 TKIs are qd administrations [5,6,31,48], except for nilotinib, which is bid [7]. Dasatinib is administered qd with no meal requirements [6], and it is recommended imatinib be taken with a meal [5]. Nilotinib doses should be taken 12 hours apart, fasting two hours before and one hour after each dose [7]. Conflicts between TKI administration and a patient’s lifestyle or concomitant medications should be determined prior to initiating therapy. A consistent and simplified therapeutic plan will promote adherence, and, in effect, improve treatment responses and patient quality of life.

Dose optimization of TKIs

Once a front line TKI is initiated, each patient should be closely monitored to ensure the appropriate response milestones are met [36] and any emerging AEs or worsening comorbid conditions are controlled and/or alleviated. The recommended starting dose of each TKI, which is based on data from first-line clinical trials [5–7], can be fine-tuned and optimized through dose modifications. Efforts to improve responses through dose optimization of TKIs are currently being investigated in clinical trials.

Both NCCN Guidelines^® and recommendations by ELN consider BCR-ABL1≤10% at three months an optimal molecular response [36,47]. Statistically significant improvements in five-year survival outcomes have been observed for patients who achieved BCR-ABL1≤10% versus >10% at three months treated with dasatinib (PFS: 89% vs. 72%, p = .0014; OS: 94% vs. 81%, p = .0028) or imatinib (PFS: 93% vs. 72%, p < .001; OS: 95% vs. 81%, p < .001) from DASISION [8]. A similar trend was observed for five-year survival in ENESTnd with nilotinib 300 mg bid for patients who achieved BCR-ABL1>1%–≤10% versus >10% at three months (PFS: 95% vs. 78%, p < .0014; OS: 98% vs. 82%, p < .0001) and imatinib (PFS: 99% vs. 79%, p < .0001; OS: 99% vs. 79%, p < .0001) [10]. Additional testing up to and at six months is recommended to thoroughly assess success or failure of the starting TKI dose [47,62]. A significantly higher three-year PFS (94% vs. 66%, p < .0001) and OS (97% vs. 84%, p =.0091) were observed in dasatinib-treated patients [63] and an improved four-year PFS (95% vs. 75%, p = .0633) and OS (96% vs. 75%, p = .0277) in nilotinib-treated patients who achieved BCR-ABL1 ≤ 10% versus >10% at six months [18].

In addition to preventing disease progression, a main goal of TKI therapy is for the patient to achieve CCyR by 12 months [36]. Specifically, the NCCN Guidelines^® state that, at three months, patients with a >50% reduction in BCR-ABL1 transcript levels from baseline (despite BCR-ABL1>10%) or with BCR-ABL1 minimally above the 10% cutoff may continue receiving the same dose of dasatinib or nilotinib for another three months [36].

Determining if dose adjustments/TKI switching is necessary may also be affected by whether a patient is receiving imatinib or a second-generation TKI as front line therapy [47,62]. Treatment milestones at MDACC are defined at six and 12 months, with indications for switching treatment being BCR-ABL1 [IS]>1% at six months, BCR-ABL1>1% at 12 months and patients who have lost a response at any time [62].

Phase-2 clinical trials with front line imatinib have investigated if doses higher than 400 mg qd would promote faster and deeper responses. In the RIGHT trial, 12-month MMR rates were faster in patients who received imatinib 400 mg bid (54%) compared to patients who received imatinib 400 mg qd in the IRIS trial (39%); a somewhat higher rate of grade-3/grade-4 AEs was also observed in this high-dose group [4,64]. The MMR rate in patients who received imatinib 600 mg in the TIDEL trial (47%) was comparable to that of patients from IRIS (40%) at 12 months, but higher at 24 months (TIDEL: 73%; IRIS: 55%) [4,65]. In TIDEL, grade-3/grade-4 AEs were uncommon past six months [65]. In support of these efficacy results, landmark analysis of data from the CML IV study demonstrated that patients who were treated with an optimized high dose of imatinib (400 mg bid) achieved MR^4.5 sooner than patients who received imatinib 400 mg qd (48 months and 62 months, respectively) [66]. The frequency of grade-3/grade-4 AEs was comparable across the treatment groups of the CML IV study [67].

In a phase-3 dose optimization study, patients who received imatinib 400 mg bid versus qd had significantly higher MMR rates at three and six months [68]; there was no statistically significant difference in 42-month MMR between arms [12]. The bid arm did not have improved long-term EFS, PFS or OS, possibly a result of more dose reductions (61% vs. 18%) for patients receiving a higher versus lower dose of imatinib in the first 12 months of treatment [12,68].

The phase-2 INTERIM study investigated responses and survival outcomes in patients alternating being on/off imatinib treatment every week for the first month and then on/off treatment alternating months [69]. Of 13 patients who lost CCyR and MMR, 12 (one was lost to follow-up) were able to regain these responses after resuming nonintermittent imatinib treatment. OS and PFS were not affected by intermittent treatment, suggesting that further investigation into alternative treatment plans may be warranted.

In DASISION, dasatinib-treated patients were able to achieve cytogenetic and molecular responses after undergoing dose reductions from 100 mg qd to improve AEs; patients who experienced pleural effusion underwent dose interruptions (62%) and dose reductions (41%), and achieved CCyR (96%), MMR (82%) and MR^4.5 (50%) [8]. Pharmacokinetics were used in the OPTIM trial to adjust the starting 100 mg dose of dasatinib in patients with higher plasma levels of dasatinib (assessed every 15 days) [70]. After a median of 32 months, the study reported that dose optimization via pharmacokinetic analyses was an effective method to reduce AEs and discontinuation rates while maintaining a deep molecular response [70]. A new phase-4 trial (NCT02689440) is investigating the efficacy of dasatinib at 50 mg qd to determine if it is sufficient to control CML-CP [71].

Dose optimization strategies were evaluated in patients receiving nilotinib 300 mg bid in the ENESTxtnd study [72]. Dose escalations and reductions of nilotinib did not appear to affect response, as 64% and 76% of patients who underwent these modifications, respectively, achieved MMR by 24 months.

Currently under way is the DESTINY trial, which is reducing front line TKI dose by 50% in patients with CML who had MMR or MR^4.0 (BCR-ABL1 <.01% [IS]) for three tests over the past 12 months [73]. After dose reduction, patients are monitored for the maintenance of an optimal molecular response.

While major cytogenetic and molecular responses are primary goals of TKI therapy for CML, the long-term plans for treatment will likely vary across patient age groups. For example, younger patients usually are more physically able and may be more willing to undertake dose escalations to improve molecular responses to attempt treatment-free remission (TFR) [74]. Treatment of older patients often includes more dose modifications to avoid drug–drug interactions with concomitant medication [75] and management of comorbidities than in younger patients, in addition to reducing the risk of disease progression and improving general quality of life.

The cost of TKI therapy

While successful therapeutic outcomes and long-term survival are clearly the main goal of TKI treatment, patients are also faced with lifelong management of the financial aspects of being diagnosed with CML and affording treatment.

First-line TKIs versus SCT

While not as common in the United States and Europe, SCT remains a frequent treatment option in other countries. Allogenic SCT is a curative procedure for CML-CP and is 25 to 40 times less expensive in developing countries than in the United States [62].

Cost analysis studies in Europe and Mexico have estimated that a one-time, larger payment for SCT is likely more cost-effective than lifelong TKI treatment [76], although this varies by nation and will likely be impacted by the introduction of generic formulations. Important considerations are the difficulty in finding an appropriate donor match for SCT, the serious complications that can result after low-dose whole-body irradiation and cell transfer and that transplant is typically only considered in patients <55 years due to general health issues being more common in the elderly [77].

Generic imatinib

Analysis by Padula et al. considered changes in cost of TKI therapy in the United States once generic imatinib is commercially available [78]. Their estimates suggest the total cost of imatinib therapy per patient will drop from $79,000/year with branded imatinib to an estimated $46,000/year with generic imatinib; therapy with dasatinib and nilotinib is expected to remain unchanged ($87,000–$92,000/year) [78]. Current pricing, however, suggests that the price of generic imatinib may actually only be 10% lower than branded imatinib [79], although medication pricing varies globally. In a 2015 analysis, branded imatinib was priced over two times higher in the United States as most countries in the European Union and generic imatinib was priced significantly lower in India compared to the United States [80].

For younger patients, <50 years, the early and sustained molecular responses (required for treatment discontinuation studies) achieved with second-generation TKIs may counterbalance any early financial benefit gained by using generic imatinib [62]. Studies in countries where generic imatinib is available suggest comparable efficacy and safety compared to branded imatinib [81,82] in addition to no difference in rate of switching to subsequent therapy [81]. The loss of exclusivity of dasatinib and nilotinib in the future and the introduction of new therapies for CML will affect both the cost and treatment landscape for this disease.

Compliance and cost

The total cost of CML therapy includes the CML medication, charges for checkups and hospital visits for disease-related complications and management of TKI-related AEs. Patient compliance can also affect healthcare costs. A retrospective analysis of medical claims data from 267 imatinib-treated patients reported a significant relationship between adherence to therapy and lower total healthcare and medical expenses (excluding the drug cost of imatinib) [83]. Similar results were observed across the approved front line BCR-ABL1 TKIs in a cost analysis with patients treated with imatinib, dasatinib or nilotinib. Overall, patients with higher adherence paid 6.5 times less in leukemia-related hospitalization costs than their less adherent counterparts [84].

The future of CML therapy

While CML therapy has made successful strides over the past decade due to BCR-ABL1 TKIs, new TKIs in development and combination treatments may have additional therapeutic benefits. Deep, sustained responses are also of growing interest due to the prospect of TFR.

New therapies

New therapies are being investigated to maximize molecular response and OS for patients with CML. Radotinib is a BCR-ABL1 TKI that was approved for first-line treatment of CML in Korea in 2015 [85] after demonstrating its efficacy in a phase-3 study [20]. Preliminary results include a significantly higher MMR rate for radotinib (52%) versus imatinib (30%) (Table 1) and no progression to AP/BP by 12 months [20]. Long-term efficacy and safety analysis of radotinib versus imatinib is yet to be evaluated. The efficacy and safety of radotinib as a second- or third-line agent is currently being evaluated in clinical trials (NCT02422719; NCT01602952) [86]. Since radotinib is generated in South Korea, where the annual price of approved TKIs is >4 times lower than in the United States, financial benefits are likely [87].

ABL001 is a new BCR-ABL1 inhibitor that does not compete for the ATP-binding pocket like the BCR-ABL1 TKIs [88]. Preliminary evidence indicates that ABL001 is well tolerated in previously treated patients with CML and is being tested in patients with T315I-mutation–positive CML or Ph+ ALL. Long-term results from these and other trials, with medications alone or in combination, will determine if survival outcomes can be improved upon even more than what is observed with current BCR-ABL1 TKI therapies.

New combinations with currently approved BCR-ABL1 TKIs

The currently approved BCR-ABL1 TKIs are being tested in combination with other anticancer agents to determine whether paired medications can provide benefits over TKI monotherapy. In particular, all three of the approved front line TKIs are currently being investigated for use in combination with IFN-α.

In the SPIRIT study, patients who received imatinib plus peg-IFN-α had a higher MMR rate at 12 months (57%) and 24 months (64%) compared to patients who received imatinib 600 mg (49% and 53%) or imatinib 400 mg (38% and 43%) [89]. The German CML IV study treated CML-CP patients with imatinib, and the incidences of MMR at 24 months for patients treated with imatinib 400 mg qd alone or in combination with IFN-α were the same and were a little lower than that in patients who received imatinib 400 mg bid (63% and 76%, respectively) [67]. There were no differences in OS or PFS across the three treatment arms. Peg-IFN-α is more protected from proteolytic breakdown, which could explain the difference in results between the SPIRIT and CML IV studies [90].

A preliminary study of 40 patients suggests that dasatinib plus peg-IFN is tolerable and may increase response rates [91]. An ongoing phase-2 study of dasatinib with peg-IFN will continue to assess the efficacy and safety of this combination in patients with CML-CP [92]. New therapies used to treat other cancers are also being evaluated in combination with dasatinib. For example, dasatinib in combination with the PD-1 immune checkpoint inhibitor nivolumab is being investigated in patients with CML who are experiencing disease progression or are resistant/intolerant to other BCR-ABL1 TKIs [93]. The main objectives are to assess efficacy and safety, in addition to evaluating rate, time and duration of molecular responses.

Three studies are recruiting patients (NCT02001818, NCT01657604, NCT02201459) to examine the combination of IFN-α with nilotinib to determine whether it improves the molecular response of nilotinib monotherapy [94]. Nilotinib is also being investigated in combination with the JAK (Janus kinase) 1/2 inhibitor, ruxolitinib, as the JAK2 pathway is constitutively active in BCR-ABL1–positive cells [95]. Ruxolitinib is approved for treatment of primary myelofibrosis, leading to this study. Preliminary findings suggest that patients treated with a BCR-ABL1 TKI who responded with minimally residual disease may benefit from the addition of drugs like ruxolitinib to promote apoptosis of resistant CML stem cell/progenitor cells.

Treatment-free remission

Several studies are investigating the concept of TFR in patients who have achieved a sustained, deep molecular response on TKI therapy (Table 4) [114,115]. The European Stop Kinase Inhibitor (EURO-SKI) trial has enrolled 821 patients taking imatinib, dasatinib or nilotinib as a first- or second-line therapy; of these patients, 750 have molecular response data after achieving MR^4.0 for at least one year prior to enrollment [96,97]. Twenty-four months after discontinuation, 52% of patients were free from molecular relapse [96]. Prognostic modeling of molecular response data from imatinib-treated patients in the EURO-SKI trial identified a significant correlation (p < .0001) between duration of MR^4.0 while on TKI therapy and six-month MMR post-discontinuation [96]. Specifically, molecular relapse-free survival at six months was 66% versus 43% for patients treated with imatinib for >5.8 years versus ≤5.8 years. In the DESTINY trial, patients receive a 50% dose of an approved first-line TKI for 12 months, followed by discontinuation [73,99]. Although data on the discontinuation phase is not yet available, molecular relapse occurred in 3/125 (2.4%) patients who entered the study with stable MR^4.0 on the full dose of their prescribed TKI and then received the reduced dose for 12 months.

Table 4. Studies investigating treatment-free remission of CML with BCR-ABL1 TKIs.

Study (ID)	Phase	Total enrolled patients/no. discontinued (Region)	TKI treatment	Study-defined DMR requirement for discontinuation	Definition of relapse	Relapse free	Response after reinitiating TKI therapy	Outcome correlations
EURO-SKI (NCT01596114) [96,97]	3	821/750 (Europe)	Imatinib, dasatinib, or nilotinib (front line or second line due to intolerance)	MR^4.0 for ≥1 year	Any loss of MMR	MRFS 6 months: 62% 12 months: 56% 24 months: 52%	NR	Longer duration of MR^4.0 on imatinib and duration of imatinib prior to stop were associated with 6-month TFR (p < .0001)
LAST (NCT02269267) [98]	2	173  (USA)	Imatinib, dasatinib, nilotinib, or bosutinib (any line)	MR^4.0 for ≥2 years	Molecular recurrence	NR	NR	NR
DESTINY (NCT01804985) [99]	2	174/174  (UK)	Front line imatinib, dasatinib or nilotinib (second line if due to intolerance allowed)	MMR for 12 months on reduced (50%) TKI dose	Loss of MMR on 2 consecutive samples	NR	NR	NR
STOP-2G-TKI [100]	Observational	100/60	Dasatinib or nilotinib (front line or subsequent line)	MR^4.5 for ≥2 years	Loss of MMR	57%	26/26 regained MMR and MR^4.5	Significantly worse outcome was associated with prior suboptimal response/TKI resistance
DADI (UMIN000005130) [101]	2	88/63  (Japan)	Dasatinib  (front line or subsequent line after imatinib)	MR^4.0 for 1 year	Loss of MR^4.0	48%	29/33 by 3 months; 33/33 by 6 months	Lower TFR at 12 months was associated with imatinib resistance (p = .029), high γδ T cell (p = .0022), or high Treg (p = .011); higher TFR at 12 months was associated with high NK cell counts (p = .0053–.022)
DASFREE (NCT01850004) [102]	2	71/30  (global)	Dasatinib (front line or subsequent line)	MR^4.5 for ≥1 year	Loss of MMR	63%	10/10^a	NR
1^st-DADI (UMIN000011099) [103]	2	Targeted enrollment: 100 (Japan)	Dasatinib  (front line)	CMR for ≥1 year	Loss of CMR	NR	NR	NR
D-STOP (NCT01627132) [104]	2	65/54  (Japan)	Dasatinib  (front line or subsequent line after imatinib)	MR^4.0 for two years	Two consecutive  > MR^4.0 in 1 month	63%	20/20	Higher TFR at 12 months was associated with higher counts of cells expressing: CD4 + CD8- (p < .0001), CD3-CD56 + (p < .001), CD16 + CD56 + (p < .0001), or CD57 + CD56 + (p < .01)
TRAD (NCT02268370) [105]	2	92/75 (Canada)	Dasatinib rechallenge and discontinuation after imatinib discontinuation	MR^4.5 for ≥2 years	Loss of MR^4.0 on two consecutive occasions or MMR on one occasion	69% after imatinib discontinuation	14/14 regained MMR and MR^4.5 with dasatinib at 3 months	TFR was associated with longer duration of imatinib therapy (p = .001) and longer duration of MR^4.0 (p = .004) or MR^4.5 (p = .006) prior to discontinuation
STIM1 (NCT00478985) [106]	Observational	100/100 (France)	Imatinib  (line not specified)	CMR for ≥2 years	Two consecutive 1-log increases in BCR-ABL1 from baseline or single loss of MMR	Overall: 39% MMR: 83%	55/57	Molecular relapse was associated with age (p = .0364) and Sokal risk score (p = .0210)
TWISTER  (ACTRN 12606000118505) [107]	2	40/40 (Australia)	Imatinib  (front line or after IFN)	CMR for ≥2 years	Loss of MMR for any sample or any two consecutive positive samples	45%	22/22	TFR was associated with duration of IFN and time to CMR after switching to imatinib
ENESTFreedom (NCT01784068) [108]	2	215/190 (global)	Nilotinib  (front line)	All assessments ≥ MR^4.0,>MR^4.0–<MR^4.5 for ≤2 assessments, and MR^4.5 at last assessment over 1 year	Loss of MMR	52%	MMR: 85/85^b MR^4.5: 76/85	NR
ENESTGoal (NCT01744665) [109]	2	59/4  (USA)	Nilotinib  (after imatinib)	MR^4.5 for 1–2 years	Confirmed loss of MMR	25%	3/3	NR
ENESTPath (NCT01743989) [110]	3	619/101  (Europe)	Nilotinib  (after imatinib)	Stable MR^4.0 for ≥1 year	Loss of MR^4.0	NR	NR	NR
ENESTop (NCT01698905) [111]	2	163/126 (global)	Nilotinib  (after imatinib)	Confirmed MR^4.5 for 1 year on nilotinib	Loss of MMR or confirmed loss of MR^4.0	58%	≥MMR: 50/51 ≥MR^4.0: 48/51 MR^4.5: 47/51	NR
Nilst (UMIN000007141) [112]	2	112/87 (Japan)	Nilotinib (front line or after imatinib)	MR^4.5 for 2 years	Loss of MR^4.5	59%	32/34	No correlations observed
STAT2 (UMIN 000005904) [113]	NR	96/73  (Japan)	Nilotinib  (after imatinib)	MR^4.5 for 2 years	Confirmed loss of MR^4.5	66%	NR	NR

^a An 11th patient lost MMR but discontinued the study, restarted therapy at another site and was lost to follow-up.

^b An 86th patient lost MMR but discontinued the study.

CML: chronic myeloid leukemia; CMR: complete molecular response (undetectable BCR-ABL1 on the International Scale [IS]); DMR: deep molecular response; IFN: interferon; MMR: major molecular response; MR^4.0: BCR-ABL1(IS) < .01%; MR4.5: BCR-ABL1 (IS) ≤.0032%; MRFS: molecular recurrence-free survival; NK: natural killer; NR: not reported; TFR: treatment-free remission; TKI: tyrosine kinase inhibitor.

Additional clinical trials investigating TFR in patients treated with imatinib, dasatinib or nilotinib, and any available outcome data, can be viewed in Table 4 [96–113]. These studies will help determine if there are prognostic factors to identify candidates for TFR and optimal pairing of a patient and front line TKI to promote a sustained, deep molecular response.

Summary and conclusions

The introduction of BCR-ABL1 TKIs has revolutionized the treatment of CML, affording patients long-term survival rates. Consequently, long-term treatment emphasizes the importance of considering several factors when pairing a patient and front line TKI. A balance between the achievements of an early response, long-term outcomes, potential AEs, and a patient’s lifestyle are all key components for successful therapy. To assure appropriate TKI selection, comprehensive first-line clinical trial data need to be considered alongside patient characteristics.

First-line trials with dasatinib and nilotinib have demonstrated improved early molecular responses, a better response in intermediate- and high-risk patients with CML, and fewer chronic AEs compared with imatinib [8,10]. As new data and TKIs are generated, physicians will need to remain apprised of new information and increasingly consider prognostic factors, such as risk scores, when prescribing initial CML therapy. Also, deep, early and sustained responses are becoming of greater interest in more patients due to the desire to achieve minimally residual disease with the prospect of TFR.

Special considerations for patient-TKI pairing include comorbidities, concomitant medications, age and sex. Ideally, dose optimization will regulate any AEs or suboptimal responses from a prescribed TKI to produce both the best response and highest quality of life for the patient. Female and younger patients in particular have been described as being more affected by chronic AEs than their male and older counterparts, respectively [44]; therefore, the safety profiles of the TKIs should likely be examined more closely in certain patient subsets.

The assessment of a patient’s medical history, awareness of prognostic factors defining a patient’s disease state and a comprehensive knowledge of the response and safety profiles of TKIs will help physicians select the best front line TKI for each patient.

Potential conflict of interest

Disclosure forms provided by the authors are available with the full text of this article online at https://doi.org/10.1080/10428194.2017.1379074.

Acknowledgements

The authors take full responsibility for the content of this publication and confirm that it reflects their viewpoints and medical expertise. Professional medical writing and editorial assistance was provided by Kelly M. Fahrbach, PhD, of StemScientific, an Ashfield Company, part of UDG Healthcare plc, funded by BMS.

GS has acted as a consultant and a speaker for ARIAD, Bristol-Myers Squibb, Novartis, and Pfizer. EJ has received honoraria from ARIAD, Bristol-Myers Squibb, Novartis, and Pfizer.