A population-based study of chronic myeloid leukemia patients treated with imatinib in first line

Transcription

1 A population-based study of chronic myeloid leukemia patients treated with imatinib in first line AJH Fausto Castagnetti, 1,2 * Francesco Di Raimondo, 3 Antonio De Vivo, 1,2 Antonio Spitaleri, 3 Gabriele Gugliotta, 1,2 Francesco Fabbiano, 4 Isabella Capodanno, 5 Donato Mannina, 6 Marzia Salvucci, 7 Agostino Antolino, 8 Roberto Marasca, 9 Maurizio Musso, 10 Monica Crugnola, 11 Stefana Impera, 12 Elena Trabacchi, 13 Caterina Musolino, 14 Francesco Cavazzini, 15 Giuseppe Mineo, 16 Patrizia Tosi, 17 Carmela Tomaselli, 18 Michele Rizzo, 19 Sergio Siragusa, 20 Miriam Fogli, 1,2 Riccardo Ragionieri, 1,2 Alessandro Zironi, 1,2 Simona Soverini, 1,2 Giovanni Martinelli, 1,2 Michele Cavo, 1,2 Paolo Vigneri, 21 Fabio Stagno, 3 Gianantonio Rosti, 1,2 and Michele Baccarani 22 Chronic myeloid leukemia (CML) treatment is based on company-sponsored and academic trials testing different tyrosine kinase inhibitors (TKIs) as first-line therapy. These studies included patients selected according to many inclusion exclusion criteria, particularly age and comorbidities, with specific treatment obligations. In daily clinical practice (real-life), inclusion exclusion criteria do not exist, and the treatment outcome does not only depend on the choice of first-line TKI but also on second- and third-line TKIs. To investigate in a real-life setting the response and the outcome on first-line imatinib, with switch to second generation TKIs in case of unsatisfying response or intolerance, we analyzed all newly diagnosed patients (N 5 236), living in two Italian regions, registered in a prospective study according to population-based criteria and treated front-line with imatinib. A switch from imatinib to second-generation TKIs was reported in 14% of patients for side effects and in 24% for failure or suboptimal response, with an improvement of molecular response in 57% of them. The 5-year overall survival (OS) and leukemia-related survival (LRS) were 85% and 93%, respectively; the 4-year rates of MR 3.0 and MR 4.0 were 75% and 48%, respectively. Cardiovascular complications were reported in 4% of patients treated with imatinib alone and in 6% of patients receiving nilotinib as second-line. Older age (70 years) affected OS, but not LRS. These data provide an unbiased reference on the CML management and on the results of TKI treatment in real-life, according to ELN recommendations, using imatinib as first-line treatment and second-generation TKIs as second-line therapy. Am. J. Hematol. 92:82 87, VC 2016 Wiley Periodicals, Inc. Additional Supporting Information may be found in the online version of this article. 1 Department of Oncology and Hematology, S.Orsola-Malpighi University Hospital, University of Bologna, Bologna, Italy; 2 Department of Experimental, Diagnostic and Specialty Medicine, S.Orsola-Malpighi University Hospital, University of Bologna, Bologna, Italy; 3 Division of Hematology, A.O.U. Policlinico Vittorio Emanuele,, University of Catania, Catania, Italy; 4 Division of Hematology and Bone Marrow Transplantation, Ospedali Riuniti Villa Sofia - Cervello, Palermo, Italy; 5 Division of Hematology, Arcispedale Santa Maria Nuova, IRCCS, Reggio Emilia, Italy; 6 Division of Hematology, Ospedale Papardo, Messina, Italy; 7 Division of Hematology, Santa Maria delle Croci Hospital, Ravenna, Italy; 8 Service of Immunohematology and Transfusion Medicine, Azienda Ospedaliera Provinciale, Ragusa, Italy; 9 Division of Hematology, Modena University Hospital, University of Modena and Reggio Emilia, Modena, Italy; 10 Division of Hematology, Clinica La Maddalena, Palermo, Italy; 11 Division of Hematology and Bone Marrow Transplantation, University Hospital, Parma, Italy; 12 Division of Hematology, ARNAS Garibaldi, Catania, Italy; 13 Division of Hematology and Bone Marrow Transplantation, Guglielmo da Saliceto Hospital, Piacenza, Italy; 14 Division of Hematology, G. Martino University Hospital, University of Messina, Italy; 15 Division of Hematology, Sant Anna University Hospital, University of Ferrara, Ferrara, Italy; 16 Division of Hematology, S.Vincenzo Hospital, Taormina, ME, Italy; 17 Division of Hematology, Ospedale degli Infermi, Rimini, Italy; 18 Department of Oncology/Division of Hematology, Ospedale Civico, Palermo, Italy; 19 Division of Hematology, Ospedale S. Elia, Caltanissetta, Italy; 20 Division of Hematology and Bone Marrow Transplantation, Azienda Ospedaliera Universitaria Policlinico P. Giaccone,, Palermo, Italy; 21 Unit of Medical Oncology, A.O.U. Policlinico Vittorio Emanuele, University of Catania, Catania, Italy; 22 Department of Hematology and Oncology L. and A. Seragnoli,, University of Bologna, Bologna, Italy Conflict of interest: F. Castagnetti: consultant and honoraria from Novartis, Bristol Myers Squibb, ARIAD and Pfizer; F. di Raimondo: consultant and honoraria from Novartis, ARIAD and Pfizer, consultant, honoraria and research funds from Bristol Myers Squibb; G. Gugliotta: consultant and honoraria from Novartis and Bristol Myers Squibb; R. Marasca: consultant and honoraria from Novartis, ARIAD and Pfizer; S. Soverini: consultant and honoraria from Novartis, Bristol Myers Squibb and ARIAD; G. Martinelli: consultant and honoraria from ARIAD, BMS, Novartis and Pfizer; M. Cavo: consultant and honoraria from Celgene, Janssen, Takeda, Amgen and Bristol Myers Squibb; P. Vigneri: consultant and honoraria from ARIAD, BMS and Novartis, honoraria from Pfizer; F. Stagno: consultant and honoraria from Novartis, Bristol Myers Squibb, ARIAD and Pfizer; G. Rosti: consultant and honoraria from Novartis, Bristol Myers Squibb, ARIAD, Pfizer and Roche; M. Baccarani: consultant and honoraria from Novartis, Bristol Myers Squibb, ARIAD and Pfizer. All the other authors have nothing to disclose. *Correspondence to: Fausto Castagnetti, Institute of Hematology L. & A. Seragnoli, Department of Experimental, Diagnostic and Specialty Medicine, S.Orsola-Malpighi University Hospital, University of Bologna, Via Massarenti, 9, Bologna, Italy. Tel.: ; Fax: fausto.castagnetti@unibo.it Contract grant sponsor: European LeukemiaNet and EUTOS project (through a contract between the University of Heidelberg, Novartis Oncology Europe, the University of Bologna and GIMEMA). Received for publication: 10 August 2016; Revised: 10 October 2016; Accepted: 17 October 2016 Am. J. Hematol. 92:82 87, Published online: 22 October 2016 in Wiley Online Library (wileyonlinelibrary.com). DOI: /ajh VC 2016 Wiley Periodicals, Inc. 82 American Journal of Hematology, Vol. 92, No. 1, January 2017 doi: /ajh.24591

2 Introduction Treatment recommendations for BCR-ABL1 positive (BCR- ABL11) chronic myeloid leukemia (CML) [1 4] spread from company-sponsored and academic trials [5 22], as well as from the database of referral hematological centers [23 25]. Both sources provided high quality data but can be biased and may not reflect the socalled real-life: in daily clinical practice there are no inclusion and exclusion criteria and, although the treatment choice is based on international recommendations [1 4], the ultimate decision is up to both health care providers and patients. The treatment results do not only depend on efficacy and safety of a single first-line tyrosine kinase inhibitor (TKI) but also on the subsequent use of other available TKIs. Recording all newly diagnosed patients, according to population-based criteria, and prospectively collecting exhaustive data is a quite hard task. Population-based data are available only from Sweden, but without clinical details on treatment, response, and outcome [26,27]. Until few years ago, everyday clinical practice was based on first-line imatinib and two second-generation TKIs, namely nilotinib and dasatinib, as second-line therapy [1 4,28 30]. Everyday clinical practice undergoes continuous modifications, so that yesterday s practice will never be the same of today s and tomorrow s practice, but to evaluate the outcome of a specific treatment strategy, a period of observation is necessary: at least 4 years, considering that failures and severe toxicity mainly occur during the first 3 years [1 4]. Today we are moving toward an earlier and extensive use of second-generation TKIs, because they are more powerful than imatinib and may achieve more frequent, faster, and deeper molecular responses [28 30], but a sequential use of TKIs, as available in reallife, should be the true reference point for comparisons. Moreover, CML is becoming a chronic condition, raising problems concerning both clinical management and appropriate allocations of logistic and financial resources. Hence, it is extremely important to understand how CML is managed outside of clinical trials. In this view, we report a population-based prospective study performed in Italy, as a part of the European LeukemiaNet (ELN) sponsored European Treatment and Outcome Study of CML (EUTOS) project [31,32], providing a benchmark of the standard treatment policies that were recommended before the introduction of the second-generation TKIs in the first-line setting. Methods This study, that was an extension of an European study [31,32], was reviewed and approved by the Internal Review Board of all the participating institutions. All patients living within a specified province, with newly diagnosed, chronic phase BCR-ABL11 CML were registered before any treatment; all available information on treatment, response, and outcome were then provided yearly. The study was strictly observational, without any recommendation on treatment or monitoring, according to population-based criteria, without any exclusion criteria, apart from age (<18 years old). For logistic and financial limitations, the registration study was limited to two Italian regions, Emilia-Romagna and Sicily. Data were collected under the supervision of the University of Bologna (Emilia-Romagna) and the University of Catania (Sicily) and were stored and analyzed at the University of Bologna. Newly diagnosed patients were registered between 2008 and 2012 in Emilia-Romagna and between 2010 and 2012 in Sicily. Imatinib (Glivec, Novartis Pharma) in first-line, nilotinib (Tasigna, Novartis Pharma) and dasatinib (Sprycel, Bristol-Myers Squibb) in second-line, were available throughout the whole registration period. Nilotinib and dasatinib were also approved as first-line treatment in December 2010 and February 2011, respectively, but they became available only some months later. Registrations were accomplished at the Divisions of Hematology of each Province Hospital. The chronic, accelerated, or blast disease phase (CP, AP, BP) and the response criteria (failure and suboptimal response) were as defined by 2009 ELN recommendations [2]. Relative risk was calculated using the Sokal score [33] and the new EUTOS Long Term Survival (ELTS) score [34]. Cytogenetic data were collected but could not be fully analyzed because during the study period molecular monitoring replaced cytogenetic monitoring in an increasing number of patients. Molecular response was assessed in laboratories adhering to the GIMEMA Italian Network (LabNet), according to ELN and EUTOS recommendations [35 38]. A deep molecular response (MR 4.0 or better) was defined as stable in patients with MR 4.0 or better from month 24 to last contact, and unstable when during the same period the MR fluctuated between MR 4.0 and a major molecular response (MMR or MR 3.0, corresponding to a BCR-ABL1 0.1%). The raw CML incidence was calculated in the residents of each province, by decades and by gender. Survival probabilities were calculated using the Kaplan and Meyer method [39] and compared by the log-rank test. Overall survival (OS) was calculated from the date of diagnosis to the date of death or last contact. Leukemia-related survival (LRS) was calculated from the date of diagnosis to the date of death, when death was observed after a documented progression to AP or BP, or last contact. Results Patients A population-based study of CML treated frontline with imatinib Three hundred and thirty-seven new CML cases were registered over a resident population of 7,404,000 inhabitants 18 years old, for a raw incidence of 1.56/100,000/year in males and of 1.21/100,000/ year in females (Supporting Information Table Is). First-line treatment: 5 patients (80 years old) were treated with hydroxyurea (HU) alone, 96 patients with a second-generation TKI, 236 patients with imatinib. We report here only the results concerning these 236 patients: 202 of them (86%) were treated outside of clinical studies, while 34 (14%) were enrolled in prospective studies. The clinical trials open to enrolment in the period of interest were either companysponsored (ENESTnd and DASISION trials, enrolling from September 2007 to September 2008) [7,11] or investigator-sponsored (clinicaltrials.gov NCT and NCT , accruing from February 2009 to August 2009 and from December 2011 to November 2012, respectively) [40,41]. Nine patients (4%) were lost to follow up, 2 26 months after diagnosis. All the others were followed until May 2015, with a median follow-up of 48 months (range 2 86 months). The median age was 60 years, with 29% of patients 70 years old. The transcript type was e14a2 (b3a2) in 56%, e13a2 (b2a2) in 43%, and atypical in 1% of patients. High risk patients according to Sokal score and new ELTS score were 23% and 18%, respectively. First-line treatment Only patients treated frontline with imatinib were analyzed (N 5 236). With few exceptions, almost all of them were initially treated, for a short period of time (5 35 days), with HU. Two hundred and three patients (86%) received imatinib 400 mg once daily, while 33 patients (14%) received 400 mg twice daily. The molecular response rates are shown in Supporting Information Table IIs. The 3-month early molecular response (EMR, BCR- ABL1 10%) was not achieved in 20% of evaluable patients, but it should not be overlooked that, regrettably, the EMR data were available only in 61% of patients. At 12 months, 55% of all evaluable patients were in MR 3.0 or better. At last contact, 75% of all evaluable patients were in MR 3.0 or better and 48% were in MR 4.0 or better. From first-to second-line treatment Imatinib was changed to another TKI or HU (the so called switch ) in 97/236 patients (41%): for failure or suboptimal response in 57/236 (24%), for side effects in 34/236 (14%), and for protocol reasons in 6/236 (3%) (Supporting Information Table IIIs). The switches for side effects occurred earlier (median time to switch 8 months) than the switches for failure or suboptimal response (median time to switch 20 months) (Fig. 1). Only four patients were switched at 3 months. The estimated 5-year OS after switch was 89% in patients who switched for side effects and 87% for patients switching for failure or suboptimal response. The effect of switching on molecular response in patients who switched for failure or suboptimal response was evaluable in 61/63 patients (Table IA): in 35/61 (57%) patients the molecular response was improved, in 23/61 (38%) it was not modified, and in 3/61 (5%) doi: /ajh American Journal of Hematology, Vol. 92, No. 1, January

3 Castagnetti et al. RESEARCH ARTICLE Figure 1. Time to treatment switch according to the reported cause of switching. The switch from imatinib to another TKI or HU was performed for failure (F) or suboptimal response (SOR) in 57 patients and for side effects (SE) in 34 patients; other 6 patients (not in the graph) switched from imatinib to nilotinib for protocol reasons (enrolled in clinical studies). Median time to switch was 8 months for side effects and 20 months for failure or suboptimal response. TABLE I. (A) Molecular Response Improvement After Switching from Imatinib to Second-Generation TKIs for Failure or Suboptimal Response; (B) Patients Achieving a MR 4.0 (A) Patients switched from imatinib to second-generation TKIs for failure or suboptimal response N. total 63 N. evaluable for a change of molecular response 61 N. who improved by 1 log 14 (23%) N. who improved by 2 log 11 (18%) N. who improved by 3 4 log 10 (16%) N. who improved, total 35 (57%) (B) Number of patients who achieved a stable MR 4.0 With imatinib first-line only, no switch 52/236 (22%) With imatinib first-line, including switch 78/236 (33%) Seventy-eight patients achieved a stable MR 4.0 (MR 4.0 or better from month 24 to last contact, with a median follow-up of 48 months), 52 with imatinib alone and 26 after switching from imatinib to second-generation TKIs. the response worsened by at least 1 log. Overall, the treatment switch was associated with a remarkable increase of the number of patients in MR 3.0 or better, from 10% at the time of switching to 52% at last quantitative polymerase chain reaction (qpcr). With a median follow-up of 48 months, 78/236 patients (33%) achieved a stable MR 4.0 or better, of whom 52 (22%) with imatinib alone and 26 (11%) after switching to a second-generation TKI (Table IB). Outcome The estimated 5-year OS and LRS were 85% and 93%, respectively (Fig. 2). A progression to BP was reported in 14/236 patients (6%). Twenty-seven patients (11%) died, of whom 13 (6%) of leukemia, after disease progression and 14 (6%) of other causes, without any evidence of progression (7 of them were in stable MR 3.0 or better). The other causes of death were clearly identified in 9 patients; myocardial infarction 1, stroke 2, cardiac failure 2, pneumonitis 2, lung cancer 1, and melanoma 1. In the remaining five patients, who were years old, the cause of death could not be specified, but no evidence of progression was reported. Median age at the time of death was 69 years for those who died after progression and 83 years for those who died of other causes. A third-line treatment was reported in only three patients, of whom two died after progression and one was alive after allogeneic stem cell transplantation. Only 3/236 patients were enrolled in a trial of treatment discontinuation, not surprisingly, since median follow-up was 48 months. Figure 2. OS and LRS of all 236 patients. The 5-year OS was 85%; the 5- year LRS was 93%; the progression-free survival is not shown because it was close to OS. Prognostic factors OS was not affected by age in the interval between 18 and 69 years (5-year OS 92%, ranging from 89% to 95% in the different decades), but 5-year OS probabilities of elderly (70 79 years) and very elderly patients (80 years) were significantly lower (78%, P , and 34%, P < , respectively) (Fig. 3A). However, the probability of dying of leukemia (LRS) was not affected by age (Fig. 3B). Similarly, the Sokal score predicted OS (Fig. 3C) but not LRS (Fig. 3D). The ELTS score clearly separated low, intermediate, and high risk patients for OS (Fig. 3E) and also for LRS (low vs. intermediate, P ; low vs. high, P , intermediate vs. high, P ) (Fig. 3F). The EURO risk score [42] was not associated with differences in OS and LRS (data not shown). The OS and LRS were not influenced by gender or transcript type (data not shown). The prognostic value of 3-month and 6-month EMR was not evaluable because of too many missing data. The depth of MR at 1 year and at 2 years did not correlate with OS or with LRS (Supporting Information Fig. 1s). Side effects, complications No major or life-threatening complications were observed, apart from grade 3 or grade 4 cardiovascular adverse events (CVAEs): 3 cases of myocardial infarction (1 on imatinib first-line, 2 on nilotinib second-line); 4 cases of symptomatic ischemic heart disease (2 on imatinib first-line, 2 on nilotinib second-line); 2 cerebrovascular events (on imatinib first-line). The age of the patients with grade 3 4 CVAEs ranged between 61 and 94 years (median 73 years). Overall, grade 3 or grade 4 CVAEs were reported in 9/236 patients (3.8%), of whom 5/142 (3.5%) in patients treated only with imatinib, and 4/63 (6.3%) in patients treated with nilotinib second-line. Discussion This population-based prospective study was not designed to identify the best CML treatment but to describe the treatment outcome when clinical practice was mainly based on imatinib in first-line and nilotinib or dasatinib in second-line, according to the most important trial reports and to ELN and NCCN recommendations [1 4,28 30]. It is acknowledged that, because of differences in health care, drug cost, life style, education, and so on, one country cannot be fully representative of all countries. In Italy, the cost of all TKIs is fully covered by Regional Health Systems, under the indications approved by the European Medicines Agency. During the registration period, TKI prescriptions were guided by the ELN recommendations, version 2009 [2], based on general consensus, but without any specific obligation. It is noticeable that during the same period ( ), the 84 American Journal of Hematology, Vol. 92, No. 1, January 2017 doi: /ajh.24591

4 A population-based study of CML treated frontline with imatinib Figure 3. (A) OS by age; (B) LRS by age; (C) OS by Sokal score; (D) LRS by Sokal score; (E) OS by the new ELTS score; (F) LRS by ELTS score. (A) The 5-year OS by age was different: 92%, 78%, and 34%, respectively (P < between and 70 79; P between and 80; P < between and 80); (B) the 5-year LRS by age was not different: 94%, 92%, and 88%, respectively (P ); (C) the 5-year OS by Sokal score was different: 92%, 86%, and 71%, respectively (P between low and intermediate, P between intermediate and high, P between low and high); (D) the 5-year LRS by Sokal score was not different: 93%, 94%, and 89%; (E) the 5-year OS by ELTS score was different: 97%, 84%, and 47%, respectively (P between low and intermediate; P between intermediate and high; P < between low and high); the 5-year LRS by ELTS score was different: 96%, 94%, and 77%, respectively (P between low and intermediate, P between intermediate and high, P < between low and high). According to the age, patients were divided into three groups: years, 167 patients (71%); years, 45 patients (19%); 80 years, 24 patients (10%); the distribution according to Sokal score was as follows: low risk 71 patients (30%), intermediate risk 111 patients (47%), high risk 54 patients (23%); the distribution according to the new ELTS score was as follows: low risk, 115 patients (49%); intermediate risk, 79 patients (33%); high risk, 42 patients (18%). same TKIs were available in many countries, with the same indications, and the 2009 ELN recommendations had a general consensus, also outside Europe. We have reported molecular response data of 236 patients treated frontline with imatinib, but it should be noticed that 94/236 patients (40%) were switched to a second-generation TKI (Supporting Information Table IIIs), with a significant benefit in terms of both depth and stability of molecular response (Table I). At 3 months, qpcr data were missing in many patients, not surprisingly, because the importance of EMR was not fully understood until 2010 and the frequency of molecular monitoring is lower in real life than in clinical trials. The early switches were uncommon. The high number of missing molecular data at 3 and 12 months, the practice of switching, as it was carried out, and the free access to three TKIs (imatinib, nilotinib, dasatinib) may explain why the prognostic value of the 3-month EMR and 12-month MMR, widely considered as important milestones [3,4,7,11,13,17,22,25,28 30,43 46], was not confirmed (Supporting Information Figure 1s). It is difficult to assess the role of first- and second-line treatment, because it is impossible to know what would have happened without doi: /ajh American Journal of Hematology, Vol. 92, No. 1, January

5 Castagnetti et al. switching, or switching at different times. This consideration highlights the difficulty of interpreting the results of clinical trials only on the basis of first-line treatment. In prior company-sponsored and academic studies, the proportion of patients who discontinued the firstline treatment was always much higher than the proportion of patients who progressed or died, underscoring the importance of second-line therapies. Regrettably, in many reports, details and outcome of second-line treatment were rarely mentioned, and even more rarely specified. The results of this prospective population-based study emphasize the importance of second-line treatment in the clinical practice: the CML management with the three available TKIs in the period , according to current treatment recommendations for switching and without any predefined obligation, resulted in response rates and outcomes, that are in the high range of prospective interventional studies investigating the efficacy and the safety of a single TKI in selected patients. This finding confirms the efficacy of TKIs and the importance of the availability of more than one TKI. In our unselected population-based cohort, the age of imatinibtreated patients (median 60 years) was much higher (10 21 years higher) than in several company-sponsored or academic or singlecentre reports (Supporting Information Table IVs). In spite of such a big difference, the 2- and 5-year probability of achieving MMR, and the 5-year probability of achieving MR 4.0 were comparable (Supporting Information Table IVs). In contrast, 5-year OS (85%) was the lowest reported so far, because in this population-based cohort elderly patients were much more numerous; however, they did not die more often of leukemia, but of other age-related causes (Fig. 3A,B). Elderly patients were under-represented in prospective studies, so that the issue if elderly patients should receive a TKI was still a matter for discussion, and there was some concern about a limited use of TKIs in RESEARCH ARTICLE the elderly [1 4,26,27,30]. In a population-based study from Sweden, it was reported that only 18% of patients more than 80 years old diagnosed between 2001 and 2008 were given a TKI [26], but in the same country that proportion grew to 79% in the period [27]. Now, we report that 87% of patients more than 80 years old, diagnosed between 2008 and 2012, were given a TKI; they had the same LRS of other patients, but a lower OS. Therefore, patient age should no longer be considered as an argument limiting the use of TKIs, although the effect on survival is limited by age itself. Noticeably, the new ELTS score [34] was independently tested in a prospective cohort of unselected patients treated in clinical practice, confirming a difference in OS among low, intermediate, and high risk patients, and also a difference in LRS between low and intermediate compared to high risk patients (Fig. 3E,F). Interestingly, the Sokal score was able to predict OS, but not LRS (Fig. 3C,D). The international boards of CML experts recommended to monitor the response to TKIs also with conventional cytogenetics and not only with qpcr [3,4]. This was the case in the patients who were newly diagnosed between 2008 and 2010, but in 2011 and 2012 the frequency of cytogenetic testing was so reduced, that it was not possible to analyze the value of cytogenetic response. This finding clearly shows that, despite international treatment recommendations, the medical community is more and more qpcr-oriented in real life. This population-based study, performed in a period when first-line therapy was mainly based on imatinib, while second generation TKIs were available in second-line and were approved for first-line only later, provides a robust dataset on the results of CML treatment with TKI in clinical practice, that will be important for interpreting and evaluating the results of the next prospective studies, arguably limited to selected patients. References 1. Baccarani M, Saglio G, Goldman J, et al. Evolving concepts in the management of chronic myeloid leukemia: Recommendations from an expert panel on behalf of the European LeukemiaNet. Blood 2006;108: Baccarani M, Cortes J, Pane F, et al. Chronic myeloid leukemia: An update of concepts and management recommendations of European LeukemiaNet. J Clin Oncol 2009;27: Baccarani M, Deininger MW, Rosti G, et al. European LeukemiaNet recommendations for the management of chronic myeloid leukemia: Blood 2013;122: O Brien S, Radich JP, Abboud CN, et al. Chronic myelogenous leukemia, version J Natl Compr Canc Netw 2014;12: Hughes TM, Kaeda J, Branford S, et al. Frequency of major molecular responses to imatinib or interferon a plus cytarabine in newly diagnosed chronic myeloid leukemia. N Engl J Med 2003;349: Druker BJ, Guilhot F, O Brien SG, et al. Fiveyear follow-up of patients receiving imatinib for chronic myeloid leukemia. N Engl J Med 2006; 355: Saglio G, Kim DW, Issaragrisil S, et al. Nilotinib versus imatinib for newly diagnosed chronic myeloid leukemia. N Engl J Med 2010;362: Kantarjian HM, Hochhaus A, Saglio G, et al. Nilotinib versus imatinib for the treatment of patients with newly disgnosed chronic-phase Philadelphia chromosome-positive, chronic myeloid leukaemia: 24month minimum follow up of the phase 3 randomised ENESTnd trial. Lancet Oncol 2011;123: Larson RA, Hochhaus A, Hughes TP, et al. Nilotinib vs. imatinib in patients with newly diagnosed Phialdelphia chromosome-positive chronic myeloid leukemia in chronic phase: ENESTnd 3-year follow-up. Leukemia 2012;26: Hochhaus A, Saglio G, Hughes TP, et al. Longterm benefits and risks of frontline nilotinib vs. imatnib for patients with chronic myeloid leukemia in chronic phase: ENESTnd 5-year update. Leukemia 2016;30: Kantarjian H, Neil PS, Hochhaus A, et al. Dasatinib versus Imatinib in newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med 2010;362: Kantarjian HM, Shah NP, Cortes JE, et al. Dasatinib or imatinib in newly diagnosed chronicphase chronic myeloid leukemia: 2-year followup from a randomized phase 3 trial (DASI- SION). Blood 2012;119: Jabbour E, Kantarjian HM, Saglio G, et al. Early response with dasatinib or imatinib in chronic myeloid leukemia: 3-year follow-up from a randomized phase 3 trial (DASISION). Blood 2014; 123: Cortes JE, Saglio G, Kantarjian HM, et al. Final 5-year study results of DASISION: The Dasatinib versus Imatinib study in treatment-na ıve chronic myeloid leukemia patients trial. J Clin Oncol 2016;34: Cortes JE, Baccarani M, Guilhot F, et al. Phase III, randomized, open-label study of daily imatinib mesylate 400 mg versus 800 mg in patients with newly diagnosed, previously untreated chronic myeloid leukemia in chronic phase using molecular end points: Tyrosine kinase inhibitor optimization and selectivity study. J Clin Oncol 2010;28: Baccarani M, Druker BJ, Branford S, et al. Long-term response to imatinib is not affected by the initial dose in patients with Philadelphia chromosome-positive chronic myeloid leukemia in chronic phase: Final update from the Tyrosine Kinase Inhibitor Optimization and Selectivity (TOPS) study. Int J Hematol 2014;99: Hochhaus A, Rosti G, Cross NC, et al. Frontline nilotinib in patients with chronic myeloid leukemia in chronic phase: Results from the European ENEST1st study. Leukemia 2016;30: Wang J, Shen ZX, Saglio G, et al. Phase 3 study of nilotinib vs. imatinib in Chinese patients with newly diagnosed chronic myeloid leukemia in chronic phase: ENESTchina. Blood 2015;125: Cervantes F, Lopez-Garrido P, Montero MI, et al. Early intervention during imatinib therapy in patients with newly diagnosed chronic-phase chronic myeloid leukemia: A study of the Spanish PETHEMA Group. Haematologica 2010;29: Hehlmann R, Lauseker M, Jung-Munkwitz S, et al. Tolerability-adapted imatinib 800 mg/d versus 400 mg/d versus 400 mg/d plus interferonalpha in newly diagnosed chronic myeloid leukemia. J Clin Oncol 2011;29: Preudhomme C, Guilhot J, Nicolini FE, et al. Imatinib plus peginterferon a-2a in chronic myeloid leukemia. N Engl J Med 2010;363: Castagnetti F, Gugliotta G, Breccia M, et al. Long-term outcome of chronic myeloid leukemia patients treated frontline with imatinib. Leukemia 2015;29: de Lavallade H, Apperley JF, Khorashad JS, et al. Imatinib for newly diagnosed patients with chronic myeloid leukemia: Incidence of sustained responses in an intention-to-treat analysis. J Clin Oncol 2008;26: Kim D, Goh HG, Kim S-H, et al. Comprehensive therapeutic outcomes of frontline imatinib mesylate in newly diagnosed chronic phase chronic myelpoid leukemia patients in Korea. Feasibility assessment of current ELN recommendations. Int J Hematol 2012;96: Sasaki K, Strom SS, O Brien S, et al. Relative survival in patients with chronic-phase chronic myeloid leukaemia in the tyrosine-kinase inhibitors era: Analysis of patients data from six prospective trials. Lancet Haematol 2015;2:e186 e Biorkholm M, Ohm L, Eloranta S, et al. Success story of targeted therapy in chronic myeloid leukemia: A population-based study of patients 86 American Journal of Hematology, Vol. 92, No. 1, January 2017 doi: /ajh.24591

6 A population-based study of CML treated frontline with imatinib diagnosed in Sweden from 1973 to J Clin Oncol 2011;29: Hoglund M, Sandin F, Hellstrom K, et al. Tyrosine kinase inhibitor usage, treatment outcome, and prognostic scores in CML: Report from the population-based Swedish CML registry. Blood 2013;122: Jabbour E, Kantarjian HM, O Brien S, et al. Front-line therapy with second-generation tyrosine kinase inhibitors in patients with early chronic phase chronic myeloid leukemia: What is the optimal response?. J Clin Oncol 2011;29: Hughes T, White D. Which TKI? An embarassment of riches for chronic myeloid leukemia patients. Hematol Am Soc Hematol Educ Program 2013;2013: Larson RA. Is there a best TKI for chronic phase CML?. Blood 2015;126: Hoffmann VS, Baccarani M, Hasford J, et al. The EUTOS population-based registry: Incidence and clinical characteristics of 2904 CML patients in 20 European Countries. Leukemia 2015;29: Hoffmann VS, Baccarani M, Hasford J, et al. Treatment and outcome of 2904 patients from the EUTOS population based registry. Leukemia 2016; Sep 23. doi: /leu [Epub ahead of print]. 33. Sokal JE, Cox EB, Baccarani M, et al. Prognostic discrimination in good-risk chronic granulocytic leukemia. Blood 1984;63: Pfirrmann M, Baccarani M, Saussele S, et al. Prognosis of long-term survival considering disease-specific death in patients with chronic myeloid leukemia. Leukemia 2016;30: Hughes T, Deininger M, Hochhaus A, et al. Monitoring CML patients responding to treatment with tyrosine kinase inhibitors: Review and recommendations for harmonizing current methodology for detecting BCR-ABL transcripts and kinase domain mutations and for expressing results. Blood 2006;108: Muller MC, Cross NCP, Erben P, et al. Harmonization of molecular monitoring of CML in Europe. Leukemia 2009;23: Cross NC, White HE, Muller MC, et al. Standardized definitions of molecular response in chronic myeloid leukemia. Leukemia 2012;26: Cross NC, White HE, Colomer D, et al. Laboratory recommendations for scoring deep molecular responses following treatment for chronic myeloid leukemia. Leukemia 2015;29: Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53: Gugliotta G, Castagnetti F, Breccia M, et al. Rotation of nilotinib and imatinib for first-line treatment of chronic phase chronic myeloid leukemia. Am J Hematol 2016;91: Castagnetti F, Breccia M, Gugliotta G, et al. Nilotinib 300 mg twice daily: an academic single-arm study of newly diagnosed chronic phase chronic myeloid leukemia patients. Haematologica 2016;101: Hasford J, Pfirrmann M, Hehlmann R, et al. A new prognostic score for survival of patients with chronic myeloid leukemia treated with interferon alfa. J Natl Cancer Inst 1998;90: Hughes TP, Hochhaus A, Branford S, et al. Long-term prognostic significance of early molecular response to imatinib in newly diagnosed chronic myeloid leukemia: An analysis from the International Randomized Study of Interferon and STI571 (IRIS). Blood 2010;116: Marin D, Ibrahim AR, Lucas C, et al. Assessment of BCR-ABL1 transcript levels at 3 months is the only requirement for predicting outcome for patients with chronic myeloid leukemia treated with tyrosine kinase inhibitors. J Clin Oncol 2012;30: Hanfstein B, Muller MC, Hehlmann R, et al. Early molecular and cytogenetic response is predictive for long-term progression-free and overall survival in chronic myeloid leukemia (CML). Leukemia 2012;26: Jain P, Kantarjian H, Nazha A, et al. Early responses predict better outcomes in patients with newly diagnosed chronic myeloid leukemia: Results with four tyrosine kinase inhibitor modalities. Blood 2013;121: doi: /ajh American Journal of Hematology, Vol. 92, No. 1, January