Treatment patterns and deep molecular response in chronic phase – chronic myeloid leukemia patients treated with second-line nilotinib or dasatinib: a multi-country retrospective chart review study

Jorge Cortesa, Lynn Huynhb, Estella Mendelsonc, Patricia Brandtc, Darshan Dalalc, Maral DerSarkissianb,
Diego Cortinac, Sahil Narkhedea and Mei Sheng Duhb
aThe University of Texas MD Anderson Cancer Center, Houston, TX, USA; bAnalysis Group, Inc., Boston, MA, USA; cNovartis
Pharmaceuticals Corporation, East Hanover, NJ, USA

ARTICLE HISTORY
Received 18 March 2019
Revised 22 June 2019
Accepted 3 July 2019

KEYWORDS
Chronic myeloid leukemia; tyrosine kinase inhibitor; targeted therapies; nilotinib; deep molecular response

Introduction

The introduction of tyrosine kinase inhibitor (TKI) ther- apy targeted at BCR-ABL kinase dramatically changed the treatment and natural history of chronic myeloid leukemia (CML), allowing patients to achieve normal life expectancy and increased long-term survival rates to over 90%, thereby reducing the need for allogeneic stem cell transplantation for many patients [1–3]. However, approximately 20–30% of CML patients with CML-chronic phase (CML-CP) do not benefit from the first generation TKI imatinib due to intolerance or development of resistance [4,5]. Nilotinib, dasatinib, and bosutinib are second generation TKIs that are more potent and effective than imatinib. Nilotinib, dasatinib, or bosutinib may be used for first-line or second-line treatment of CML among patients with resistance or intolerance to imatinib [6–8]. In two phase II clinical tri- als evaluating CML-CP patients with resistance or intolerance to imatinib, nearly 50% of CML-CP patients achieved a complete cytogenetic response when treated with either nilotinib or dasatinib in the second line with median time on treatment of about 18.7 and 13.8 months, respectively [9,10]. Second- or third-line treatment with nilotinib, dasatinib, and bosutinib have been shown to be effective in achieving major molecu- lar response (MMR) [11–13].
Achievement of deep molecular response (i.e. MR4.5, BCR-ABL1 ≤ 0.0032% on international scale; MR4.0, BCR-ABL1 ≤ 0.01% on international scale) is an important goal of TKI treatment for patients with CML, as achievement has been suggested to be associated with higher rates of overall survival and progression- free survival [14–17]. In addition, patients achieving a sustained MR4.5 are potential candidates for treat- ment-free remission (TFR), where patients may safely discontinue TKI therapy without loss of response [18–20]. In the European Society for Medical Oncology (ESMO) guidelines, prerequisites for safely discontinu- ing treatment include identification of typical BCR-ABL1 transcripts at diagnosis, ≥5 years of TKI therapy, achievement of MR4.5, and a stability of deep molecu- lar remission (at least MR4.0) for ≥2 years [19]. Patients may be motivated to initiate TFR due to considerations of long term safety and tolerability (i.e. relief from treatment side effects) [21]. In a study by Mahon et al., 52% of patients who discontinued second-line nilotinib treatment (following treatment with imatinib) achieved treatment free survival at 144 weeks [22].
Limited information is available about treatment patterns and clinical outcomes of patients treated with second-line second generation TKIs in clinical practice, especially with respect to deep molecular response. Furthermore, treatment patterns in different regions of the world may vary and might impact clinical out- comes. This study assessed and described patient characteristics, treatment patterns, proportion of patients achieving deep molecular response, and adverse events during second-line treatment in patients with CML-CP treated with nilotinib or dasatinib.

Materials and methods

Data source
An online physician panel approach was used to recruit oncologists and hematologists for a multi-coun- try retrospective chart review study from November 2016 through February 2017. Medefield, a physician panel vendor, sent open invitations (with no mention of the study sponsor) to its global community of physicians and healthcare providers to participate in this study. Ten countries were included and catego- rized into two regions: (1) Europe and Australia, including France, Germany, Italy, the Netherlands, and Australia (since approval dates were similar for second-line TKI in Europe and Australia and as data on only 12 patients were available for Australia, data from Europe and Australia were combined), and (2) Latin America, including Argentina, Brazil, Chile, Colombia, and Mexico. The study was double-blinded (i.e. physi- cians were blinded on the study sponsor information and study sponsor were blinded on the physicians recruited to participate).

Physician and patient selection
Eligible hematologists and oncologists were required to have access to CML information for patients who met the eligibility criteria, including data on molecular response (i.e. MR4.0, MR4.5) reported on the inter- national scale. Participating physicians reviewed and completed a web-based electronic case report form (eCRF) to collect data on patient demographic and clinical characteristics, treatment patterns, and molecu- lar response outcomes for up to four patient charts. To mitigate concern of selection bias, random selec- tion of patient charts was performed. The eCRF program produced a random letter and requested that physicians review a chart for a patient with a last name corresponding to the random letter. If a phys- ician did not have a patient whose last name began with the specified letter and met the eligibility criteria, the physician would then select the option to receive another random letter to search for another patient who might meet the eligibility criteria [23].
Eligible patients were diagnosed with CML-CP, ≥18 years of age at initiation of second-line nilotinib or dasatinib between January 1, 2011 and July 1, 2015, had data collected on molecular response (reported on the international scale for at least the three most recent tests) in the medical chart, and had ≥12 months of follow-up after initiation of first- or second-line treatment of CML-CP. Patients diagnosed with concurrent malignancies or enrolled in a clinical trial at initiation of second-line therapy were excluded. Patients who received bosutinib were not included in this study as it was approved in 2013 (later than dasa- tinib and nilotinib) in Europe and was not approved in all countries included in this study.

Study design
A retrospective cohort design was used for this study. Patients who initiated either nilotinib or dasatinib as second-line therapy were included in the nilotinib and dasatinib cohorts, respectively, and the date of treat- ment initiation defined the index date. During the pre-index period, data were collected on patient demographics (age at second-line TKI initiation and sex), comorbid conditions, Sokal risk score at CML diagnosis, and first-line treatment history (i.e. drug type, year of treatment initiation, treatment duration, and reasons for first-line TKI discontinuation due to resistance, intolerance, or neither). Data were also col- lected on second-line treatment patterns (e.g. dosage, duration of therapy, reasons for discontinuation), molecular response, and adverse events during the post-index period which spanned from the index date until the date of last contact or death.

Outcomes and statistical analyses
Patient characteristics, treatment patterns and adverse events were stratified by region and second-line TKI and described using means and standard deviations (SDs) for continuous variables and frequencies and proportions for categorical variables. The proportion of patients achieving MR4.0 and MR4.5 in second-line were reported and stratified by region, Sokal risk score, reasons for first-line TKI discontinuation, and generation of the TKI used as first-line treatment for CML-CP (i.e. second generation TKIs included nilotinib and dasatinib and first generation TKI included imatinib).
A multivariate Fine and Gray’s [24] extension of the Cox proportional hazards model of sub-distribution hazards in competing risk data was used to compare time to achieve MR4.0 and MR4.5 for second-line nilo- tinib versus dasatinib. The models accounted for the competing-risk event of treatment discontinuation due to TKI resistance in second-line and included a random effect for country clustering of data. The model adjusted for age, gender, Sokal risk score at CML diag- nosis, prior hydroxyurea use, generation of first-line TKI agent (1st versus (vs.) 2nd), and reasons for first- line TKI discontinuation that were pre-determined based on clinical inputs. These variables are known confounders of the relationship between treatment and molecular response. Patients were censored at the earliest of second-line treatment discontinuation, death, or end of study follow-up. Adjusted hazard ratios (HR) and 95% confidence intervals (CIs) were reported, and statistical significance was defined based on a ¼ 0.05. The same analysis was repeated for sub-groups defined by region (i.e. separately for Europe/ Australia and Latin America). All statistical analyses were conducted in SAS 9.4 (SAS Institute Inc., Cary, NC, USA) and were performed by Analysis Group, Inc.

Results

Demographics and clinical characteristics
The study included 280 patients treated with either nilotinib (N ¼ 135 [48%]) or dasatinib (N ¼ 145 [52%]) in second-line, including 108 nilotinib and 115 dasati- nib patients from Europe/Australia (80%), and 27 nilotinib and 30 dasatinib patients from Latin America (20%). Duration of observation was similar for patients in both regions (19.3 months in Europe/Australia and 19.2 months in Latin America). Patients in Latin America were younger than those in Europe/Australia with a median age (range) of 49 years (22–77 years) compared to 58 years (18–88 years), respectively. A lower proportion of patients were female in Europe/ Australia than in Latin America (34% vs. 39%, respect- ively). In Europe/Australia, a greater proportion of nilo- tinib (21%) than dasatinib (10%) patients had a high Sokal risk score at CML diagnosis, though this was more similar in Latin America (29.6% of nilotinib and 26.7% of dasatinib patients in Latin America had high Sokal risk score) (Table 1).

Treatment patterns
Imatinib was the predominant first-line therapy used among patients treated with either nilotinib or dasati- nib in second-line (92% and 82%, respectively). A higher proportion of second-line patients in Latin America previously received imatinib as first-line ther- apy compared to those in Europe/Australia (96% vs. 84%, respectively). Resistance was reported as the rea- son for first-line TKI discontinuation in the majority of nilotinib and dasatinib patients (74%). A higher pro- portion of patients in Europe/Australia discontinued first-line therapy due to resistance compared to those in Latin America (76% vs. 65%, respectively) (Table 2). The proportion of patients discontinuing second-line treatment due to resistance for nilotinib and dasatinib was 11.1% (95% CI: 6.4. 17.7) and 6.2%, (95% CI: 2.9, 11.5), respectively. Median time on second-line treat- ment was similar for both nilotinib and dasatinib (19.1 months vs. 18.7 months, respectively).

Dosing
In the overall study sample, 39% of nilotinib patients received a starting daily dose of 600 mg (of which a majority of patients received 300 mg twice daily). When stratified by region, the average dose for Latin American patients was greater than that for European/ Australian patients (656 mg/day vs. 583 mg/day at initi- ation, respectively). A high proportion of dasatinib patients (77%) received 100 mg/day at initiation. The average daily dose for dasatinib among patients in Europe/Australia was greater than that for Latin American patients (129 mg/day vs. 100 mg/day at initi- ation, respectively).

Unadjusted analysis of molecular response
Across regions, molecular response was assessed within 3 months of second-line TKI initiation for 68% of all patients (72% for Latin America and 68% for Europe/Australia). A numerically higher proportion of patients achieved deep molecular response (MR4.5 and MR4.0) on nilotinib (MR4.5: 37.8%, 95% CI [29.6, 46.5]; MR4.0: 53%, 95% CI [43.8, 61.6]) during second-line therapy as compared to dasatinib (MR4.5: 33.1%, 95% CI [25.6, 41.4]; MR4.0: 42.2%, 95% CI [33.5, 51.2]).
In the unadjusted Cox proportional hazards models, second-line treatment with nilotinib was not signifi- cantly different in achieving MR4.5 (HR ¼ 1.08, 95% CI [0.85, 1.36]), but significantly different in achieving MR4.0 (HR ¼ 1.30, 95% CI [1.05, 1.61]) compared to dasatinib. When stratified by region, Sokal risk score, reason for first-line TKI discontinuation, or use of gen- eration of TKI in first-line, similar results were observed for both MR4.0 and MR4.5. A lower proportion of patients with high Sokal risk score had achieved MR4.5 and MR4.0 than patients with low or intermedi- ate Sokal risk score (Figures 1 and 2).

Multivariate regression analysis
To account for the differences in patient baseline char- acteristics (e.g. a higher proportion of nilotinib patients had high Sokal risk score compared to dasati- nib patients, a higher proportion of nilotinib patients were treated with and discontinued imatinib as first- line TKI due to resistance), covariates identified a priori (i.e. Sokal risk (low [reference], intermediate, and high), age (<45 years [reference], 45–64 years, 65–74 years, and 75þ years), sex (male [reference] and female), prior hydroxyurea use (vs. no prior use [reference]), generation of TKI in first-line [first generation [refer- ence] and second generation), and discontinuation

Figure 1. Unadjusted proportion of second-line CML-CP patients achieving MR4.5. TKI: tyrosine kinase inhibitor.
Notes: 1Proportion of patients achieving molecular response was reported for the entire study period which is defined as from time of initiation of second- line TKI to end of follow-up. 2Patients with neither resistance nor adverse events as a reason for discontinuation for second-line nilotinib (N ¼ 8) and disati- nib (N ¼ 18) are shown. due to resistance (vs. neither intolerance nor resist- ance [reference]) in first-line) were adjusted for in the multivariate Cox proportional hazards models. After adjustment, second-line treatment with nilotinib was better in terms of achieving MR4.5 (HR ¼ 1.37, 95% CI [1.11, 1.69]) and MR4.0 (HR ¼ 1.58, 95% CI [1.28, 1.96]) compared to dasatinib. High Sokal risk score (HR ¼ 0.32, 95% CI [0.13, 0.79]) and discontinuation of first- line TKI due to resistance (HR ¼ 0.62, 95% CI [0.42, 0.92]) were significant baseline covariates that negtively affected achievement of MR4.5, irrespective of the treatment. For the analysis of MR4.0 achievement, high Sokal risk score (HR ¼ 0.46, 95% CI [0.27, 0.80]) was the only significant baseline covariate (Figure 3(a,b)). A standard Cox proportional hazards model was performed and similar findings to the Fine and Gray’s extension model were observed.
Similar results were found when the above analyses were conducted separately for the two regions (i.e. Europe/Australia and Latin America). In Europe,Australia, nilotinib was found to be better in terms of achieving MR4.5 (HR ¼ 1.36, 95% CI [1.04, 1.77]) and MR4.0 (HR ¼ 1.65, 95% CI [1.45, 1.87]) than dasatinib. The same trend seems to be applicable for Latin America (MR4.5: HR ¼ 1.15, 95% CI [0.50, 2.66] and
MR4.0: HR ¼ 1.17, 95% CI [0.40, 3.41]), although the sample size is limited and the CIs are very wide.

Adverse events experienced
Among CML patients included in this study, the most commonly reported adverse event was cytopenia, which was observed in 13% of nilotinib patients and 10% of dasatinib patients. In addition, 10% of dasati- nib patients experienced pleural effusion, compared to 2% of those treated with nilotinib (Table 3). These were the only two adverse events reported by ≥10% of patients treated with either nilotinib or dasatinib.

Discussion

This retrospective medical chart-review analysis of two geographic regions assessed clinical outcomes in second-line CML-CP patients and found that, after adjusting for potential confounders, patients treated with nilotinib had a higher rate of achieving deep molecular response defined as MR4.5 and MR4.0 com- pared to dasatinib patients. A higher proportion of

Figure 2. Unadjusted proportion of second-line CML-CP patients achieving MR4.0. TKI: tyrosine kinase inhibitor.
Notes: 1Proportion of patients achieving molecular response was reported for the entire study period which is defined as from time of initiation of second- line TKI to end of follow-up. (a) Region. (b) Sokal Risk Score. (c) Reason for first-line TKI Discontinuation. 2Patients with neither resistance nor adverse events as a reason for discontinuation for second-line nilotinib (N ¼ 8) and disatinib (N ¼ 13) are shown. (d) First-line TKI Generation. nilotinib patients had high Sokal risk score and discon- tinuation of first-line TKI due to resistance compared to dasatinib patients. These imbalances, along with those in other covariates, may have led to similarities in observed results for molecular response. After adjustment for baseline covariates, the multivariate model suggests that nilotinib may be better in achiev- ing MR4.5. Additionally, both high Sokal risk score and discontinuation of first-line TKI due to resistance increase the rate in achieving MR4.5 irrespective of the second-line treatment received. Similar trends were observed for MR4.0.
Results from this study are consistent with previously reported studies that suggested higher response for patients treated with nilotinib than those treated with dasatinib. For instance, a retrospective chart review study on CML-CP patients using nilotinib and dasatinib as second-line therapies found that nilotinib was associ- ated with a significantly lower risk of progression and longer preogression-free survival than dasatinib [25]. Another real-world study found that second-line niloti- nib had improved life expectancy and quality of life among patients with CML-CP compared with dasatinib [26]. The trends observed for MR4.5 in the current study may also extend to front-line settings, as demonstrated by findings from clinical trials: in the ENESTnd trial, 54% of nilotinib and 31% of imatinib patients achieved MR4.5 by 5 years in first-line while in the DASISION trial only 42% of dasatinib patients and 33% of imatinib patients achieved MR4.5 in the first line [27,28]. In the absence of a direct comparison, parallel review of these studies provides a plausible representation of any pos- sible differences between the two drugs.
Few studies have been conducted to assess deep molecular response outcomes among second-line CML-CP patients. Such studies are important as achievement of deep molecular response is associated with improved long-term clinical outcomes and the potential for initiating TFR [29,30]. TFR is a therapeutic goal for some patients, but criteria for safely discontin- uing TKIs are still being defined. ESMO guidelines have provided the following minimum requirements for initiating TFR: (1) institutional requirements for safe supervision of TFR and (2) green criteria, which sup- port TFR [19]. Institutional requirements include avail- ability of high quality internationally standardized, accurate, and sensitive qRT-PCR laboratory, rapid turn- around of PCR test results (within 4 weeks), capacity to provide PCR tests every 4–6 weeks when required, and structured follow-up established to enable rapid

Figure 3. Multivariate Fine and Gray’s extension of the Cox model comparing second-line Nilotinib versus Dasatinib all patients. intervention if BCR-ABL is rising [19]. The green criteria include meeting the institutional criteria described, low/intermediate Sokal score at diagnosis, typical b2a2- or b3a2-BCR-ABL1 transcripts, or atypical tran- scripts which can be quantified over a 4.5 log dynamic range, chronic phase disease, optimal response to first-line therapy, duration of TKI therapy >5 years, MR4.5 reached, and duration of deep molecular response (MR4.5 or MR4.0) >2 years [19]. Current research suggests that sustained molecular response is important for successful initiation of TFR, though the necessary duration of sustained response has not been determined [29]. Discontinuation of TKI for initi- ation of TFR is still an evolving therapeutic goal; clin- ical trials have small patient cohorts and do not use standardized definitions of molecular response [29,31]. Real-world evidence is needed to confirm whether these clinical trial results can be achieved in clinical practice settings, especially since the narrow eligibility criteria of clinical trials may not reflect real-world study populations.
Subgroup analyses for Latin America were con- ducted to assess differing treatment patterns in the region compared to Europe/Australia. Recent studies have analyzed treatment patterns in Europe and the United States, however, limited information is available in other regions [32]. In addition, differing treatment patterns in low-income regions such as Latin America may impact clinical outcomes [33,34]. For instance, Cortes et al. found that there were deviations from current treatment recommendations for CML in Latin America, and highlighted the importance of under- standing treatment patterns of patients with CML in broad population studies in order to identify unmet needs and improve patient care [33]. In the current study, similar findings for the analysis of MR4.5 and MR 4.0 were observed for Europe/Australia and Latin America, though results for Latin America should be generalized with caution given the small sample size. The small sample size in Latin America is consistent with the findings of Cortes et al. who have shown that 14% of physicians in Latin America did not have
CML: chronic myeloid leukemia; TKI: tyrosine kinase inhibitor.
access to dasatinib and 44% did not have access to nilotinib [33].
In addition to examining clinical outcomes, assess- ing differences in regional treatment practice is important for identifying barriers to treatment access and areas of unmet need. 39.3% of nilotinib and 77.2% of dasatinib patients received doses that are indicated for newly diagnosed CML patients (i.e. 600 mg/day and 100 mg/day, respectively) instead of the recommended doses for patients who are resistant or intolerant to previous treatment as specified in pre- scribing information (i.e. 800 mg/day and 140 mg/day, respectively). While the overall daily initiation dose was lower than the recommended dose for both treat- ments, mean nilotinib dose at treatment initiation was lower in Europe/Australia compared to Latin America (583 mg/day vs. 656 mg/day, respectively), whereas mean dasatinib dose at treatment initiation was higher in Europe/Australia compared to Latin America (129 mg/day vs. 100 mg/day, respectively). This is important as patients may achieve improved response outcomes with higher doses and individualized dose optimization during treatment (though the latter was not assessed in the current study). In previous clinical trials, Hughes et al. found that nilotinib dose escal- ation was well tolerated and improved response out- comes in about 30% of CML-CP patients with suboptimal response [35,36]. The ENESTxtnd study showed that individualized dose optimization for nilo- tinib patients resulted in achievement of MMR [36].
These findings exemplify the variation of treatment patterns in clinical practice settings across regions, which may translate to differences in clinical outcomes among patients.
This study was subject to a number of limitations inherent to retrospective physician-administered med- ical chart review research. First, molecular monitoring practices of patients in the real-world setting may be based on heterogeneous criteria that are not uni- formly applied by physicians. Examples of heterogen- eity may include molecular response assessments being performed at different time points or laboratory assays varying by clinical sites. Second, molecular response outcomes may be subject to measurement error. BCR-ABL1 transcript levels were not collected in the study; rather, physician interpretation and report- ing of molecular response was relied upon. Although definitions for MR4.5 and MR4.0 were provided on the eCRF for reference, physician reporting may be subject to reading and interpretation errors. Additionally, data on molecular response assessed on the international scale were collected in order to ensure consistency across patient outcome assessment; however, only the three most recent response assessments for the patient were required to be on the international scale. As a result, patients may have achieved deep molecular response prior to the three most recent response assess- ments. Third, physicians with higher rates of success in treating CML-CP (i.e. higher rates of patients achieving deep molecular response) may have been more likely to the invitation to participate in the study; as such, the study sample may not be representative of the source CML-CP population in each country, and the reported molecular response outcomes in this study may be higher as a result. In order to mitigate concerns about selection bias, an algorithm for random selection of patients was implemented, but this algorithm does not ameliorate the potential responder bias. Fourth, it is likely that adverse events were underreported in this study, as it is possible that only severe adverse events are reported by patients and recorded in their electronic medical records. For instance, the proportion of second- line patients with pleural effusion was higher in a previ- ous study from a Polish tertiary care center (26% in dasatinib patients and 2% in nilotinib patients) [37] and a study by Ghorab et al. (15.2% of dasatinib patients with various cancers) [38]. Fifth, in observational studies, estimated associations may be impacted by uncon- trolled confounding due to lack of data on certain varia- bles for adjustment in the multivariate analyses.

Conclusions

Patient characteristics and treatment patterns for second-line CML-CP vary between Europe/Australia and Latin America. This chart review study suggests that patients treated with a TKI in second-line may be able to achieve molecular response after first-line TKI therapy for CML-CP, and that second-line nilotinib treatment for CML-CP may be associated with a better rate of deep molecular response over time compared to second-line dasatinib in routine clinical practice.

Acknowledgments
The authors thank Giuliana Zaccardelli for her analytical sup- port and Camara Sharperson and Jessica Marden for their writing and editorial support to the study.

Author contributions
LH, MD, and SN made substantial contributions to the ana- lysis and interpretation of data and drafting the manuscript. MSD made substantial contributions to study conception and design, analysis and interpretation of data, and revising the manuscript critically for important intellectual content. EM, PB, DD, and DC made substantial contributions to study conception and design, interpretation of data, and revising the manuscript critically for important intellectual content. JC made substantial contributions to study conception and design, interpretation of data, and revising the manuscript critically for important intellectual content. All authors have given final approval of the version to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Disclosure statement
JC has received consulting fees from Novartis Pharmaceuticals Corporation. LH, MD, SN, and MSD are employees of Analysis Group, Inc., a consultancy that has received research funding from Novartis Pharmaceuticals Corporation for this and other studies. PB, DD, DC, and EM are employees and stock holders of Novartis Pharmaceuticals Corporation.
Funding
This research was funded by Novartis Pharmaceuticals Corporation, East Hanover, NJ, US.

References

[1] Jabbour EJ, Cortes JE, Kantarjian HM. Tyrosine kinase inhibition: a therapeutic target for the management of chronic-phase chronic myeloid leukemia. Expert Rev Anticancer Ther. 2013;13:1433–1452.
[2] Quintas-Cardama A, Cortes JE. Chronic myeloid leuke- mia: diagnosis and treatment. Mayo Clin Proc. 2006; 81:973–988.
[3] Sasaki K, Strom SS, O’Brien S. Relative survival in patients with chronic-phase chronic myeloid leukae- mia in the tyrosine-kinase inhibitor era: analysis of patient data from six prospective clinical trials. Lancet Haematol. 2015;2:e186–e193.
[4] Bixby D, Talpaz M. Mechanisms of resistance to tyro- sine kinase inhibitors in chronic myeloid leukemia and recent therapeutic strategies to overcome resist- ance. ASH Educ. Program Book. 2009;2009:461–476.
[5] Hughes TP, Branford S. Monitoring disease response to tyrosine kinase inhibitor therapy in CML. ASH Educ Program Book. 2009;2009:477–487.
[6] Abruzzese E, Breccia M, Latagliata R. Second-gener- ation tyrosine kinase inhibitors in first-line treatment of chronic myeloid leukaemia (CML). BioDrugs. 2014; 28:17–26.
[7] Cortes JE, Khoury HJ, Kantarjian HM, et al. Long-term bosutinib for chronic phase chronic myeloid leukemia after failure of imatinib plus dasatinib and/or nilotinib. Am J Hematol. 2016;91:1206–1214.
[8] Gambacorti-Passerini C, Cortes JE, Lipton JH, et al. Safety and efficacy of second-line bosutinib for chronic phase chronic myeloid leukemia over a five- year period: final results of a phase 1/2 study. Haematologica. 2018;103:1298.
[9] Hochhaus A, Baccarani M, Deininger M, et al. Dasatinib induces durable cytogenetic responses in patients with chronic myelogenous leukemia in chronic phase with resistance or intolerance to imati- nib. Leukemia. 2008;22:1200–1206.
[10] Giles FJ, le Coutre PD, Pinilla-Ibarz J, et al. Nilotinib in imatinib-resistant or imatinib-intolerant patients with chronic myeloid leukemia in chronic phase: 48-month 10 J. CORTES ET AL. follow-up results of a phase II study. Leukemia. 2013; 27:107–112.
[11] Shah NP, Guilhot F, Cortes JE, et al. Long-term out- come with dasatinib after imatinib failure in chronic- phase chronic myeloid leukemia: follow-up of a phase 3 study. Blood. 2014;123:2317–2324.
[12] Cortes JE, Kim DW, Kantarjian HM, et al. Bosutinib ver- sus imatinib in newly diagnosed chronic-phase chronic myeloid leukemia: results from the BELA trial. J Clin Oncol. 2012;30:3486–3492.
[13] Kantarjian HM, Giles FJ, Bhalla KN, et al. Nilotinib is effective in patients with chronic myeloid leukemia in chronic phase after imatinib resistance or intolerance: 24-month follow-up results. Blood. 2011;117: 1141–1145.
[14] Etienne G, Dulucq S, Nicolini F-E, et al. Achieving deeper molecular response is associated with a better clinical outcome in chronic myeloid leukemia patients on imatinib front-line therapy. Haematologica. 2014; 99:458–464.
[15] Falchi L, Kantarjian HM, Wang X, et al. Significance of deeper molecular responses in patients with chronic myeloid leukemia in early chronic phase treated with tyrosine kinase inhibitors. Am J Hematol. 2013;88: 1024. 10.1002/ajh.23560.
[16] Patel AB, Wilds BW, Deininger MW. Treating the chronic-phase chronic myeloid leukemia patient: which TKI, when to switch and when to stop? Expert Rev Hematol. 2017;10:659–674.
[17] Kantarjian H, Cortes JE. Complete cytogenetic response, not deep molecular response, is associated with survival in chronic myeloid leukemia. J Clin Oncol. 2014;32:3077.
[18] Mahon FX, Etienne G. Deep molecular response in chronic myeloid leukemia: the new goal of therapy? Clin Cancer Res. 2014;20:310–322.
[19] Hochhaus A, Saussele S, Rosti G, et al. Chronic mye- loid leukaemia: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2017; 28:iv41–iv51.
[20] National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology. Chronic Myeloid Leuk. 2018;4:2018.
[21] Hochhaus A, Masszi T, Giles FJ, et al. Treatment-free remission following frontline nilotinib in patients with chronic myeloid leukemia in chronic phase: results from the ENESTfreedom study. Leukemia. 2017;31: 1525–1531.
[22] Mahon F-X, Boquimpani C, Takahashi N, et al. Long- term treatment-free remission (TFR) in patients (pts) with chronic myeloid leukemia in chronic phase (CML-CP) after stopping second-line (2L) nilotinib: ENESTop 144-wk results. J Clin Oncol. 2018;36: 7003–7003.
[23] Pal SK, Signorovitch JE, Li N, et al. Patterns of care among patients receiving sequential targeted thera- pies for advanced renal cell carcinoma: A retrospect- ive chart review in the USA. Int J Urol. 2017;24: 272–278.
[24] Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509.
[25] Griffin JD, Guerin A, Chen L, et al. Comparing nilotinib with dasatinib as second-line therapies in patients with chronic myelogenous leukemia resistant or intolerant to imatinib – a retrospective chart review analysis. Curr Med Res Opin. 2013;29:623–631.
[26] Li N, Yang X, Fan L, et al. Nilotinib versus dasatinib as second-line therapy in patients with Philadelphia chromosome-positive chronic myeloid leukemia in chronic phase who are resistant or intolerant to imati- nib: a cost-effectiveness analysis based on real-world data. J Med Econ. 2017;20:328–336.
[27] Saglio G, Kim D-W, Issaragrisil S, et al. Nilotinib versus Imatinib for newly diagnosed chronic myeloid leuke- mia. N Engl J Med. 2010;362:2251–2259.
[28] Cortes JE, Saglio G, Kantarjian HM, et al. Final 5-year study results of DASISION: the dasatinib versus imati- nib study in treatment-naive chronic myeloid leuke- mia patients trial. J Clin Oncol. 2016;34:2333–2340.
[29] Dulucq S, Mahon FX. Deep molecular responses for treatment-free remission in chronic myeloid leukemia. Cancer Med. 2016;5:2398–2411.
[30] Hehlmann R, Muller MC, Lauseker M, et al. Deep molecular response is reached by the majority of patients treated with imatinib, predicts survival, and is achieved more quickly by optimized high-dose ima- tinib: results from the randomized CML-study IV. J Clin Oncol. 2014;32:415–423.
[31] Saußele S, Richter J, Hochhaus A, et al. The concept of treatment-free remission in chronic myeloid leuke- mia. Leukemia. 2016;30:1638–1647.
[32] Kantarjian HM, Cortes J, Guilhot F, et al. Diagnosis and management of chronic myeloid leukemia: a sur- vey of American and European practice patterns. Cancer. 2007;109:1365–1375.
[33] Cortes J, De Souza C, Ayala-Sanchez M, et al. Current patient management of chronic myeloid leukemia in Latin America: a study by the Latin American Leukemia Net (LALNET). Cancer. 2010;116:4991–5000.
[34] Mendizabal AM, Garcia-Gonzalez P, Levine PH. Regional variations in age at diagnosis and overall survival among patients with chronic myeloid leuke- mia from low and middle income countries. Cancer Epidemiol. 2013;37:247–254.
[35] Hughes TP, Hochhaus A, Kantarjian HM, et al. Safety and efficacy of switching to nilotinib 400 mg twice daily for patients with chronic myeloid leukemia in chronic phase with suboptimal response or failure on front-line imatinib or nilotinib 300 mg twice daily. Haematologica. 2014;99:1204–1211.
[36] Hughes TP, Munhoz E, Aurelio Salvino M, et al. Nilotinib dose-optimization in newly diagnosed chronic myeloid leukaemia in chronic phase: final results from ENESTxtnd. Br J Haematol. 2017;179:219–228.
[37] Gora-Tybor J, Medras E, Calbecka M, et al. Real-life com- parison of severe vascular events and other non-hema- tological complications in patients with chronic myeloid leukemia undergoing second-line nilotinib or dasatinib treatment. Leuk Lymphoma. 2015;56:2309–2314.
[38] Ghorab A, Kantarjian HM, Elhalawani H, et al. Comparative risk assessment of imatinib, nilotinib and dasatinib in randomized controlled trials: a meta-ana- lysis. Blood. 2017;130:2869–2869.