Final 3-year Results of the Dasatinib Discontinuation Trial in Patients With Chronic Myeloid Leukemia Who Received Dasatinib as a Second-line Treatment

Open AccessPublished:March 14, 2018DOI:https://doi.org/10.1016/j.clml.2018.03.004

      Abstract

      Introduction

      We previously reported an interim analysis of the DADI (dasatinib discontinuation) trial. The results showed that 48% of patients with chronic myeloid leukemia in the chronic phase who maintained a deep molecular response (DMR) for ≥ 1 year could discontinue second- or subsequent-line dasatinib treatment safely at a median follow-up of 20 months. However, the results from longer follow-up periods would be much more useful from a clinical perspective.

      Patients and Methods

      The DADI trial was a prospective, multicenter trial conducted in Japan. After confirming a stable DMR for ≥ 1 year, dasatinib treatment subsequent to imatinib or nilotinib was discontinued. After discontinuation, the loss of DMR (even of 1 point) was defined as stringent molecular relapse, thereby triggering therapy resumption. The predictive factors of treatment-free remission (TFR) were analyzed.

      Results

      The median follow-up period was 44.0 months (interquartile range, 40.5-48.0 months). The estimated overall TFR rate at 36 months was 44.4% (95% confidence interval, 32.0%-56.2%). Only 2 patients developed a molecular relapse after the 1-year cutoff point. The presence of imatinib resistance was a significant risk factor for molecular relapse. Moreover, high natural killer cell and low γδ+ T-cell and CD4+ regulatory T-cell (CD25+CD127low) counts before discontinuation correlated significantly with successful therapy discontinuation.

      Conclusion

      These findings suggest that discontinuation of second- or subsequent-line dasatinib after a sustained DMR of ≥ 1 year is feasible, especially for patients with no history of imatinib resistance. In addition, the natural killer cell count was associated with the TFR.

      Keywords

      Introduction

      Second-generation tyrosine kinase inhibitors (TKIs), including dasatinib and nilotinib, are better at inhibiting BCR-ABL kinase activity than the first-generation TKI imatinib mesylate. Second-generation TKIs were introduced initially as a second-line treatment for the patients with disease resistant or who were intolerant to imatinib and resulted in a remarkable response rate.
      • Talpaz M.
      • Shah N.P.
      • Kantarjian H.
      • et al.
      Dasatinib in imatinib-resistant Philadelphia chromosome-positive leukemias.
      Because subsequent studies demonstrated that second-generation TKIs show superior efficacy to imatinib for newly diagnosed chronic myeloid leukemia (CML),
      • Cortes J.E.
      • Saglio G.
      • Kantarjian H.M.
      • et al.
      Final 5-year study results of DASISION: the dasatinib versus imatinib study in treatment-naive chronic myeloid leukemia patients trial.
      • Hochhaus A.
      • Saglio G.
      • Hughes T.P.
      • et al.
      Long-term benefits and risks of frontline nilotinib vs imatinib for chronic myeloid leukemia in chronic phase: 5-year update of the randomized ENESTnd trial.
      their use as a first-line treatment has increased.
      TKIs have dramatically improved the life expectancy of patients with CML, with a recent study showing that the survival of CML patients will be determined more by comorbidities than by the CML itself.
      • Saussele S.
      • Krauss M.P.
      • Hehlmann R.
      • et al.
      Impact of comorbidities on overall survival in patients with chronic myeloid leukemia: results of the randomized CML study IV.
      This situation has highlighted the adverse events associated with the long-term administration of TKIs, which include cardiovascular disease.
      • Douxfils J.
      • Haguet H.
      • Mullier F.
      • Chatelain C.
      • Graux C.
      • Dogne J.M.
      Association between BCR-ABL tyrosine kinase inhibitors for chronic myeloid leukemia and cardiovascular events, major molecular response, and overall survival: a systematic review and meta-analysis.
      To date, a number of clinical trials have been conducted to investigate the feasibility of discontinuing imatinib for patients who have achieved a durable deep molecular response (DMR). The pioneering STIM (stop imatinib)
      • Mahon F.X.
      • Rea D.
      • Guilhot J.
      • et al.
      Discontinuation of imatinib in patients with chronic myeloid leukaemia who have maintained complete molecular remission for at least 2 years: the prospective, multicentre stop imatinib (STIM) trial.
      and TWISTER
      • Ross D.M.
      • Branford S.
      • Seymour J.F.
      • et al.
      Safety and efficacy of imatinib cessation for CML patients with stable undetectable minimal residual disease: results from the TWISTER study.
      trials, in which CML patients discontinued imatinib after ≥ 2 years of molecular remission at the level of MR4.5 (BCR-ABL1 transcript levels of ≤ 0.0032% standardized to the International Scale [IS]) or deeper, showed that ∼40% of the patients maintained molecular remission (ie, entered a period of treatment-free remission [TFR]).
      Although accumulating evidence has shown that discontinuing imatinib is feasible, trials of patients receiving second-generation TKIs are lacking. Therefore, we conducted a phase II trial to investigate whether long-term TFR was achievable after discontinuing second- or subsequent-line dasatinib treatment after imatinib resistance or intolerance (the DADI trial [dasatinib discontinuation]). Dasatinib was discontinued in patients who had achieved DMR (BCR-ABL1 0.0069% IS) for ≥ 1 year. We previously reported the results of an interim analysis of a study with a median follow-up period of 20 months after dasatinib discontinuation.
      • Imagawa J.
      • Tanaka H.
      • Okada M.
      • et al.
      Discontinuation of dasatinib in patients with chronic myeloid leukaemia who have maintained deep molecular response for longer than 1 year (DADI trial): a multicentre phase 2 trial.
      Recently, the long-term follow-up results of an imatinib stop trial (STIM1) were reported, in which no molecular recurrence was observed after 22 months.
      • Etienne G.
      • Guilhot J.
      • Rea D.
      • et al.
      Long-term follow-up of the French stop imatinib (STIM1) study in patients with chronic myeloid leukemia.
      However, the long-term follow-up results of second-line TKI stop trials have scarcely been reported. Thus, it is important to examine whether long-term remission can be maintained in second-line treatment settings. We report the final, planned 3-year analysis of the DADI trial (Japan Primary Registries Network no. UMIN000005130).

      Patients and Methods

       Study Patients

      Patients with CML in the chronic phase who were undergoing second- or subsequent-line dasatinib therapy after receiving imatinib were eligible if they had achieved a DMR (before registration). Patients were also required to be aged ≥ 15 years, with adequate performance status (World Health Organization performance score, 0-2) and no severe dysfunction of the primary organs (eg, liver, kidney, and lung). Previous treatments in addition to imatinib and apart from allogeneic hematopoietic stem cell transplantation were allowed. Patients harboring additional chromosomal abnormalities in Philadelphia chromosome-positive cells at the diagnosis and patients with a history of BCR-ABL1 mutations associated with dasatinib resistance (T315I, F317L, and V299L) were excluded from the present study because they were presumed to have a high risk of relapse after therapy discontinuation.
      • Mahon F.X.
      • Rea D.
      • Guilhot J.
      • et al.
      Discontinuation of imatinib in patients with chronic myeloid leukaemia who have maintained complete molecular remission for at least 2 years: the prospective, multicentre stop imatinib (STIM) trial.
      Patients undergoing treatment for other malignancies were also excluded. Imatinib resistance and intolerance were defined according to the judgment of the treating doctor and in accordance with European LeukemiaNet 2006 definition.
      • 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.

       Molecular Monitoring

      Real-time quantitative reverse transcription polymerase chain reaction (RQ-PCR) analyses were performed at a single central laboratory: Bio Medical Laboratories (BML Inc, Tokyo, Japan). A molecular response was standardized to the BCR-ABL1 IS using the laboratory's conversion factor (0.87).
      • Yoshida C.
      • Fletcher L.
      • Ohashi K.
      • et al.
      Harmonization of molecular monitoring of chronic myeloid leukemia therapy in Japan.
      ABL1 was used as an internal control to validate amplification of BCR-ABL1. The number of stable ABL1 copies amplified was checked using a good quality RNA sample; this equated to 87,000 copies of ABL1 per RQ-PCR reaction (200 ng of total RNA). The quality and quantity of sample RNA used for BCR-ABL1 amplification was validated as 87,000 copies of ABL1.

       Definition of DMR

      The study cutoff point for RQ-PCR analysis was set such that the decision regarding a DMR or molecular relapse after discontinuation would be reliable. The lowest concentration of BCR-ABL1 cDNA required to ensure ∼100% detection sensitivity and reliability was measured in-house. Thus, a minimum of 10 copies of BCR-ABL1 per assay (200 ng of total RNA), corresponding to 0.0069% BCR-ABL1IS, was selected. DMR was defined as a CML sample harboring < 0.0069% BCR-ABL1IS, corresponding to molecular response 4 (MR4.0).
      • Cross N.C.
      • White H.E.
      • Colomer D.
      • et al.
      Laboratory recommendations for scoring deep molecular responses following treatment for chronic myeloid leukemia.

       Study Design

      The DADI trial was a prospective, multicenter, single-arm phase II study conducted in Japan. CML patients with a confirmed DMR were preregistered. To confirm a sustained DMR, RQ-PCR analysis was performed every 3 months throughout the preregistration period (ie, a consolidation phase comprising 1 year of dasatinib treatment after preregistration). Only patients who achieved a sustained DMR at 4 subsequent consecutive data points (during the 1 year) were enrolled in the DADI study. After drug discontinuation, patients were monitored by RQ-PCR every month for the first year, every 3 months during the second year, and every 6 months during the third year. During molecular monitoring, a BCR-ABL1IS of ≥ 0.0069% detected at a minimum of 1 analysis point was defined as stringent molecular relapse, thereby triggering therapy resumption. The total follow-up duration was set at 36 months after drug discontinuation. Patients with molecular relapse resumed dasatinib immediately and at the previously effective dose. These patients were then tested using RQ-PCR at 1, 3, 6, and 12 months after reintroduction of dasatinib. Routine cytogenetic analysis during molecular monitoring after discontinuation is not common in Japan; therefore, it was not performed in the DADI trial. During the study period, dasatinib dose reductions were permitted to manage AEs. The ethics committees of the participating hospitals approved the clinical trial. All participants provided written informed consent in accordance with the Declaration of Helsinki.

       Peripheral Immunoprofiles of Patients Receiving Dasatinib

      The T- and natural killer (NK)-cell profiles in the peripheral blood were evaluated throughout the preregistration period. The blood cell counts and differentials were measured using an automated blood cell count analyzer, and the lymphocyte fraction was determined by flow cytometry using forward scatter versus side scatter gating. The average number of lymphocyte subsets was calculated across the 5 data points. Immunophenotypic examination was performed using 2- or 3-color flow cytometry with a FACSCalibur system running CellQuest, version 3.3, software (BD Biosciences, Franklin Lakes, NJ). Flow cytometry was performed at a single central laboratory (BML Inc). All antibodies used in the study were purchased from BD Biosciences. The lymphocyte subsets detected by flow cytometry were as follows: CD8+ T cells (CD3+CD8+), NK cells (CD3CD56+ and CD16+CD56+), T-large granular lymphocytes (CD57+CD3+), NK-large granular lymphocytes (CD57+CD56+), γδ+ T cells (CD8TCRγδ+ and CD8+TCRγδ+), and CD4+ regulatory T cells (Tregs; CD4+CD25+CD127low).
      • Liu W.
      • Putnam A.L.
      • Xu-Yu Z.
      • et al.
      CD127 expression inversely correlates with FoxP3 and suppressive function of human CD4+ Treg cells.
      The CD4/8 ratio was determined by 2wo-color analysis of CD4 and CD8 cell populations.

       Study Endpoints

      The primary endpoint was the proportion of patients who maintained TFR at 6 months. TFR was defined as the interval from initiation of dasatinib discontinuation to the date of stringent molecular relapse. The main secondary endpoints were the TFR rate at 12 months and the clinical effect of dasatinib resumption on patients with relapse. Age, gender, the Sokal score at the diagnosis, the reasons patients switched from imatinib to dasatinib (clinical response to previous imatinib), a history of interferon (IFN)-α therapy, total duration of both imatinib and dasatinib treatment, total dose of dasatinib from the time of drug initiation, total dose of dasatinib during the 1 year before drug discontinuation, and the mean absolute number of T and NK cell subsets in the peripheral blood during dasatinib treatment were used to identify factors predictive of TFR.

       Statistical Analysis

      To determine whether dasatinib discontinuation corresponded to the TFR in CML patients compared with data from the previous study, a sample size of ≥ 50 patients was selected to provide a power of > 80%.
      • Hochhaus A.
      • Saglio G.
      • Hughes T.P.
      • et al.
      Long-term benefits and risks of frontline nilotinib vs imatinib for chronic myeloid leukemia in chronic phase: 5-year update of the randomized ENESTnd trial.
      Using the cutoff points for each continuous variable determined by the concordance index, patients were dichotomized into high and low count groups. The TFR rates were calculated using Kaplan-Meier analysis, and the stratified groups were compared using the log-rank test. P < .05 was considered to indicate statistical significance. Statistical analysis was performed using SAS software, version 9.3. The study is registered with the Japan Primary Registries Network (registry no. UMIN000005130).

       Role of the Funding Source

      The funder of the study collected the data but played no role in the study design, data analysis, or data interpretation. The funder reviewed the manuscript before journal submission. Four of us (J.I., H.T., S.M., J.S.) were able to access raw data, and 1 of us (S.K.) had full access to all the data and bears the ultimate responsibility for the decision to publish.

      Results

       Patient Characteristics

      From April 1, 2011 to March 31, 2012, 88 patients who had achieved a DMR with second- or subsequent-line dasatinib treatment were enrolled in the consolidation phase. Of these 88 patients, 25 were excluded from the discontinuation phase: 24 patients with fluctuating BCR-ABL1 transcript levels and 1 in whom minor BCR-ABL1 transcripts were detected.
      Thus, 63 patients were included in the dasatinib discontinuation phase.
      • Imagawa J.
      • Tanaka H.
      • Okada M.
      • et al.
      Discontinuation of dasatinib in patients with chronic myeloid leukaemia who have maintained deep molecular response for longer than 1 year (DADI trial): a multicentre phase 2 trial.
      Of the 63 patients, 13 (21%) had switched from imatinib because of the development of resistance, 36 (57%) had switched because of imatinib intolerance, and 14 (22%) had switched at the patient’s request because they believed dasatinib would have a more robust clinical effect. The Sokal risk score was low in 65%, intermediate in 14%, high in 14% of patients, and missing for 5%. The best responses to previous imatinib treatment were complete a hematologic response in 10%, a complete cytogenetic response in 8%, a major molecular response (MMR) in 78%, and missing for 5%. The median duration of imatinib treatment was 62 months (range, 1-122 months). Previous treatments other than imatinib included IFN-α in 12 patients, nilotinib in 4, nilotinib after IFN-α in 1, bosutinib in 1, and hydroxyurea in 1. The median duration of dasatinib treatment at enrollment was 17 months (range, 12-72 months), and the median total duration of imatinib and dasatinib treatment was 82 months (range, 23-142 months).

       TFR After Dasatinib Discontinuation

      The previous interim analysis was performed at a median follow-up point after dasatinib discontinuation of 20.0 months. At that time, 30 patients had TFR and 33 had developed molecular relapse, which had occurred within the first 7 months after dasatinib discontinuation.
      • Imagawa J.
      • Tanaka H.
      • Okada M.
      • et al.
      Discontinuation of dasatinib in patients with chronic myeloid leukaemia who have maintained deep molecular response for longer than 1 year (DADI trial): a multicentre phase 2 trial.
      After the interim analysis, 2 patients (both of whom had switched to dasatinib because to imatinib intolerance) experienced molecular relapse at 18 and 21 months after dasatinib discontinuation. Of the 35 patients who had developed molecular relapse, only 1 patient experienced an IS > 1%. The overall probability of TFR at 6, 12, and 36 months was 49% (95% confidence interval [CI], 31%-61%), 48% (95% CI, 35%-59%), and 44% (95% CI, 32.0%-56.2%), respectively (Figure 1A).
      Figure thumbnail gr1
      Figure 1Kaplan-Meier Estimates of Treatment-free Remission (TFR) After Discontinuation of Dasatinib in Patients With Chronic Myeloid Leukemia. (A) For All 63 Patients, the Overall Probability of TFR at 6, 12, and 36 Months Was 49% (95% Confidence Interval, 31%-61%), 48% (95% CI, 35%-59%), and 44% (95% CI, 32.0%-56.2%), Respectively. (B) TFR in Patients Who Switched From Imatinib to Dasatinib

       Factors Associated with TFR or Molecular Relapse

      In the final analysis, the most significant risk factor associated with loss of TFR (ie, loss of a DMR, even at a single analysis point) was the presence of imatinib resistance before second- or subsequent-line dasatinib therapy. Although the TFR at 36 months was only 8% (95% CI, 0%-29%) in imatinib-resistant patients, it was 55.6% (95% CI, 38.05%-69.94%) and 50.0% (95% CI, 22.86%-72.21%) in imatinib-intolerant patients and patients who had decided to switch to dasatinib, respectively (Figure 1B).
      Interestingly, 3 of the 12 relapse events in the imatinib-resistant patients (25%) and 1 of the 23 relapse events in the nonresistant patients (0.23%) occurred as loss of a MMR (IS > 0.1%). To confirm the differences in the kinetics of relapse according to the presence or absence of imatinib resistance, we calculated the BCR-ABL1 doubling time by comparing the BCR-ABL1 levels between the last confirmed DMR and molecular relapse. The median BCR-ABL1 doubling time in imatinib-resistant patients was significantly shorter than that in non-resistant patients (16.3 days, range, 7.0-98.7 days; vs. 52.3 days, range, 7.3-341.4 days; P = .0493). Neither the Sokal score nor the duration of preceding TKI treatment, both of which have been associated with an increased risk of molecular relapse in previous imatinib discontinuation trials,
      • Etienne G.
      • Guilhot J.
      • Rea D.
      • et al.
      Long-term follow-up of the French stop imatinib (STIM1) study in patients with chronic myeloid leukemia.
      were identified as risk factors in the present analysis. Of the 5 patients who had received nilotinib as a previous treatment and switched to dasatinib because of intolerance, 4 achieved a sustained TFR.
      To examine the relationship between the immune response to dasatinib treatment and subsequent TFR, we assessed the immune profiles in 59 patients (for whom complete data sets were available) during the dasatinib consolidation phase. In general, these patients had high lymphocyte counts during dasatinib treatment, with some fluctuations. After discontinuation of dasatinib, the lymphocyte counts decreased rapidly (within ∼3 months).
      We found it interesting that, in line with the data from the interim analysis, the present univariate analysis revealed that a greater number of total NK cells (CD3CD56+) and cytolytic NK cells (CD16+CD56+) and a lower number of γδ+ T cells (Supplemental Figure 1; available in the online version) and CD4+ Tregs were associated with successful maintenance of TFR (Table 1, Figure 2). Multivariate analysis revealed that only the absence of imatinib resistance and a greater number of cytolytic NK cells (CD16+CD56+) were independently associated with maintenance of TFR. When the number of total NK cells (CD3CD56+) rather than the number of cytolytic NK cells was entered into multivariate analysis, the number of total NK cells was no longer a significant factor.
      Table 1Risk Factors for Molecular Relapse by Univariate and Multivariate Analysis
      Risk Factors for Molecular RelapseUnivariate AnalysisMultivariate Analysis
      TFR at 36 mo (%, 95% CI)HR (95% CI)P ValueHR (95% CI)P Value
      Age, y
       ≥4550.0 (34.9-63.3)0.60 (0.30-1.20).1490.80 (0.39-1.62).526
       <4529.4 (10.7-51.1)1Ref1Ref
      Sex
       Female63.6 (40.2-79.8)0.51 (0.23-1.13).0990.64 (0.27-1.53).317
       Male34.1 (20.2-48.5)1Ref1Ref
      Sokal risk score
       Low46.3 (30.7-60.5)0.90 (0.34-2.37).826
       Intermediate22.2 (3.37-51.3)1.93 (0.61-6.11).263
       High44.4 (13.5-71.9)1Ref
      Reasons for switch from imatinib to dasatinib
       Imatinib resistant7.7 (0.48-29.2)3.05 (1.17-7.96).023
       Imatinib intolerant55.6 (38.1-69.9)0.83 (0.34-2.02).68
       Patient choice50.0 (22.9-72.2)1Ref
      Imatinib resistant7.7 (0.48-29.2)3.48 (1.66-7.29).00012.84 (1.22-6.59).015
      Not imatinib resistant54.0 (39.3-66.6)1Ref1Ref
      History of IFN-α therapy
       Yes53.8 (24.8-76.0)0.75 (0.31-1.79).512
       No42.0 (28.3-55.1)1Ref
      Total TKI duration
      Both imatinib and dasatinib.
      treatment, mo
       ≥5048.9 (33.7-62.4)0.61 (0.30-1.23).167
       <5033.3 (13.7-54.5)1Ref
      CD4/8 ratio
       ≥1.0333.3 (13.7-54.5)1.39 (0.69-2.81).36
       <1.0346.3 (30.7-60.6)1Ref
      Mean CD8+ T (CD3+CD8+) cell cutoff
       ≥664/mL33.3 (13.7-54.5)1.48 (0.73-2.99).280
       <664/mL46.3 (30.7-60.6)1Ref
      Total NK (CD3CD5+) cell cutoff
       ≥539/mL51.3 (34.8-65.6)0.50 (0.25-0.99).048
       <539/mL25.0 (9.10-44.9)1Ref
      Cytolytic NK (CD16+CD56+) cell cutoff
       ≥506/mL54.1 (36.9-68.4)0.45 (0.23-0.88).0200.44 (0.22-0.90).025
       <506/mL22.7 (8.27-41.5)1Ref1Ref
      Mean T-LGLs (CD57+CD3+) cutoff
       ≥509/mL35.0 (15.7-55.2)1.37 (0.68-2.73).378
       <509/mL46.2 (30.2-60.7)1Ref
      NK-LGL (CD57+CD56+) cutoff
       ≥528/mL52.8 (35.5-67.4)0.52 (0.27-1.03).060
       <528/mL26.1 (10.6-44.7)1Ref
      γδ+ T cell cutoff
       ≥120/mL16.7 (4.12-36.5)2.51 (1.26-5.03).0092.05 (0.94-4.51).073
       <120/mL53.7 (37.4-67.4)1Ref1Ref
      Regulatory T (CD4+CD25+CD127low) cell cutoff
       ≥88/mL16.7 (2.65-41.3)2.30 (1.09-4.84).0291.83 (0.83-4.00).132
       <88/mL48.9 (34.1-62.2)1Ref1Ref
      Abbreviations: CI = confidence interval; HR = hazard ratio; IFN-α = interferon- α; LGL = T-large granular lymphocytes; NK = natural killer; Ref = reference; TFR = treatment-free remission; TKI = tyrosine kinase inhibitor.
      a Both imatinib and dasatinib.
      Figure thumbnail gr2
      Figure 2Kaplan-Meier Estimates of Treatment-free Remission After Discontinuation of Dasatinib in Patients With Chronic Myeloid Leukemia According to the Number of Cytolytic Natural Killer (NK) and Regulatory T Cells (Tregs) During Dasatinib Treatment. T- and NK-cell Profiles in the Peripheral Blood Were Evaluated Throughout the Preregistration Period. The Average Number of Lymphocyte Subsets Was Calculated Across the 5 Data Points. Graphs Show TFR Remission Periods Stratified by (A) CD16+CD56+ NK cells and (B) Tregs

       Outcomes of Patients With Molecular Relapse

      Overall, 33 of the 35 patients with relapse restarted dasatinib. Of the 2 patients who did not restart dasatinib, 1 preferred treatment with nilotinib and 1 did not wish to restart dasatinib. Subsequently, the latter experienced fluctuating BCR-ABL1 levels but never lost the MMR during the observation period.
      All patients who restarted dasatinib or nilotinib demonstrated a rapid molecular response: 30 (88%) returned to MR4.0 within 3 months and 4 had regained MR4.0 by 6 months. No loss of hematologic response or progression to advanced-phase disease was observed.

      Discussion

      Although increasing evidence has suggested the feasibility of discontinuing TKIs in CML patients showing long-term molecular remission, several important clinical questions should be addressed: (1) the requirement for different approaches for different TKIs; (2) the optimal duration of molecular remission before discontinuation; (3) the risk factors associated with molecular relapse, and (4) the optimal level of BCR-ABL1 transcripts that should trigger reinitiation of TKI therapy. The findings from an interim analysis of the DADI study results have provided valuable information for answering several of these questions. The present final analysis has provided even more clarity.
      First, we have confirmed that the TFR rate in the patients who discontinued second-line dasatinib after maintaining MR4.0 for 1 year was comparable with that reported in the imatinib discontinuation trial during first-line treatment. In line with the previous imatinib discontinuation trial, most of the relapse cases occurred within 6 months after dasatinib discontinuation, with only 2 (6%) developing > 12 months after dasatinib discontinuation.
      Second, we found that the presence of imatinib resistance is a significant risk factor for molecular relapse. In addition, we found that imatinib-resistant patients developed a relapse more quickly, with a significantly shorter doubling time of BCR-ABL1 transcript levels. These observations suggest that leukemia cells in imatinib-resistant patients had already gained aggressive characteristics before the patients had started second-line dasatinib therapy. This was likely due to alterations in gene expression patterns,
      • Radich J.P.
      • Dai H.
      • Mao M.
      • et al.
      Gene expression changes associated with progression and response in chronic myeloid leukemia.
      although additional chromosomal abnormalities or BCR-ABL1 mutations in CML cells, which have been associated with TKI resistance, were absent.
      Third and, more interestingly from a clinical perspective, we have confirmed the results of the interim analysis in that the immunologic response during dasatinib consolidation was significantly associated with TFR.
      • Collins Jr., R.H.
      • Shpilberg O.
      • Drobyski W.R.
      • et al.
      Donor leukocyte infusions in 140 patients with relapsed malignancy after allogeneic bone marrow transplantation.
      CML has long been known to be susceptible to an immunologic antileukemic response. In contrast, CML patients will have quantitative and functional defects within the NK cell compartment at the diagnosis.
      • Chen C.I.
      • Koschmieder S.
      • Kerstiens L.
      • et al.
      NK cells are dysfunctional in human chronic myelogenous leukemia before and on imatinib treatment and in BCR-ABL-positive mice.
      Moreover, Kreutzman et al
      • Kreutzman A.
      • Juvonen V.
      • Kairisto V.
      • et al.
      Mono/oligoclonal T and NK cells are common in chronic myeloid leukemia patients at diagnosis and expand during dasatinib therapy.
      showed that dasatinib induces persistent expansion of clonal cytotoxic T and NK cells, which has been associated with a favorable response. Furthermore, patients who maintained TFR after imatinib discontinuation have persistent leukemia cells, suggesting that the immunologic response plays a role in disease control.
      • Ross D.M.
      • Branford S.
      • Seymour J.F.
      • et al.
      Patients with chronic myeloid leukemia who maintain a complete molecular response after stopping imatinib treatment have evidence of persistent leukemia by DNA PCR.
      In the present study, univariate analysis revealed that a greater number of total NK cells (CD3CD56+) and cytolytic NK cells (CD16+CD56+) and a lower number of γδ+ T cells and CD4+ Tregs were associated with successful maintenance of TFR. Multivariate analysis identified the absence of imatinib resistance and a greater number of cytolytic NK cells (CD16+CD56+) as the only factors independently associated with maintenance of TFR. When the number of total NK cells (CD3CD56+) rather than the cytolytic NK cells was entered in the multivariate analysis, the number of total NK cells did not remain a significant factor. These results are in line with those from a recent report, which showed that a greater proportion of naive NK cells (CD56bright cells) within the NK cell population is associated with an increased risk of molecular relapse.
      • Ilander M.
      • Olsson-Stromberg U.
      • Schlums H.
      • et al.
      Increased proportion of mature NK cells is associated with successful imatinib discontinuation in chronic myeloid leukemia.
      Hughes et al
      • Hughes A.
      • Clarson J.
      • Tang C.
      • et al.
      CML patients with deep molecular responses to TKI have restored immune effectors and decreased PD-1 and immune suppressors.
      conducted a detailed analysis of the effector and suppressor immune responses in patients at each phase (diagnosis, TKI treatment, and sustained TFR) and suggested that achievement of TFR is associated with both restoration of effector responses and reduced immune suppression.
      • Hughes A.
      • Clarson J.
      • Tang C.
      • et al.
      CML patients with deep molecular responses to TKI have restored immune effectors and decreased PD-1 and immune suppressors.
      The present study had several limitations. First, when considering the optimal duration of molecular remission before dasatinib discontinuation, the data suggest that 1 year is sufficient for patients without imatinib resistance. In contrast, the data from the present study were insufficient to examine the correlation between DMR duration and TFR rate. The present trial was not a prospective trial of patients who were taking dasatinib as first-line treatment but a prospective trial of patients who were in DMR and receiving dasatinib as second-line treatment. When the present trial started to enroll patients, the methods of BCR-ABL1 gene transcript quantification had not yet been standardized. Thus, it was difficult to determine the exact DMR duration. Nonetheless, we believe the findings of the present trial will have significance in offering most patients who receive TKIs outside of clinical trials the opportunity to experience TFR.
      In the first-line imatinib discontinuation trials, imatinib was discontinued after 2 years of ≥ MR4.5.
      • Mahon F.X.
      • Rea D.
      • Guilhot J.
      • et al.
      Discontinuation of imatinib in patients with chronic myeloid leukaemia who have maintained complete molecular remission for at least 2 years: the prospective, multicentre stop imatinib (STIM) trial.
      • Ross D.M.
      • Branford S.
      • Seymour J.F.
      • et al.
      Safety and efficacy of imatinib cessation for CML patients with stable undetectable minimal residual disease: results from the TWISTER study.
      To examine whether the use of dasatinib enables a shorter consolidation period, we have initiated another study of dasatinib discontinuation (registry no. UMIN000011099) in which we will investigate the safety and efficacy of discontinuing first-line dasatinib in CML patients after ≥ 3 years of dasatinib treatment, including ≥ 1 year of DMR (the final year).
      In addition, the optimal level of BCR-ABL1 transcript that triggers the restart of dasatinib remains unestablished. Although we defined the trigger as the loss of DMR, a recently reported imatinib discontinuation study (A-STIM trial) defined the trigger as the loss of a MMR and reported the comparable DMR regaining rate.
      • Rousselot P.
      • Charbonnier A.
      • Cony-Makhoul P.
      • et al.
      Loss of major molecular response as a trigger for restarting tyrosine kinase inhibitor therapy in patients with chronic-phase chronic myelogenous leukemia who have stopped imatinib after durable undetectable disease.
      It might be more reasonable to define the trigger as the “loss of a MMR” in future trials.

      Conclusion

      We have confirmed the feasibility of dasatinib discontinuation in patients without imatinib-resistant disease who maintain MR4.0 for ≥ 1 year. Moreover, we have demonstrated that both the characteristics of leukemia cells (an intrinsic factor) and the host immunologic responses (an extrinsic factor) are involved in maintenance of TFR. These findings could be useful for identifying patients who will derive the greatest benefit from TKI discontinuation and enable optimization of the algorithm used for treatment of CML patients taking TKIs. In particular, the use of the cytolytic NK cell numbers during TKI treatment as a determinant of eligibility for TKI discontinuation is an attractive option, although further investigations are needed. In contrast, the patients with imatinib-resistant disease probably require novel strategies to improve TFR. These could include concurrent use of IFN-α during the consolidation phase or after TKI discontinuation.

       Clinical Practice Points

      • Previous studies have shown that ∼40% of CML patients who have maintained an MR4.5 for ≥ 2 years with first-line imatinib treatment could successfully discontinue imatinib treatment.
      • The present study has shown that 44.4% of CML patients who maintained a DMR for ≥ 1 year with second- or subsequent-line dasatinib treatment maintained long-term TFR at 36 months.
      • The presence of imatinib resistance was identified as a significant risk factor for molecular relapse.
      • The high NK-cell and low γδ+ T-cell and CD4+ regulatory T-cell (CD25+CD127low) counts before discontinuation correlated significantly with successful therapy discontinuation.
      • These findings could be useful in identifying patients likely to derive the greatest benefit from TKI discontinuation.
      • In particular, the use of cytolytic NK cell numbers during TKI treatment as a determinant of eligibility for TKI discontinuation is an attractive option, although further investigations are needed.

      Disclosure

      H.N. reports a consultant/advisory role with Novartis and honoraria and research funding from Bristol-Myers Squibb and Novartis. T.K. reports a consultant/advisory role with, and honoraria from, Bristol-Myers Squibb and Novartis. M.O. reports honoraria from Novartis. Y.I., S. Mizuta, and T.U. reports honoraria from Bristol-Myers Squibb. K.U., S. Morita, and S.K. report honoraria from Bristol-Myers Squibb and Novartis. Y.M., H.S., T.F., and Y.O. report honoraria from Bristol-Myers Squibb, Novartis, and Pfizer. H.Y. and Y.S. report honoraria from Novartis. Y.I. reports research funding from Bristol-Myers Squibb. Y.M. reports research funding from Novartis and Kirin-Kyowa. H.Y. reports research funding from Novartis. S.K. reports research funding from Bristol-Myers Squibb and Novartis. The remaining authors have stated that they have no conflicts of interest.

      Acknowledgments

      We thank the doctors who enrolled the patients. We also thank Yumi Miyashita (ECRIN, Japan) for monitoring the clinical trial and Yukie Nakazato for technical support.

      Supplemental Data

      Figure thumbnail fx1
      Supplemental Figure 1Kaplan-Meier Estimates of Treatment-free Remission After Discontinuation of Dasatinib in Patients With Chronic Myeloid Leukaemia Stratified by T-cell Immunoprofile During Dasatinib Treatment. T-cell Profiles in the Peripheral Blood Were Evaluated Throughout the Preregistration Period. The Average Number of Lymphocyte Subsets Was Calculated Across the 5 Data Points. Graphs Showing Treatment-free Remission Period Stratified by (A) CD4/CD8 Ratio and Number of (B) CD8+ Cells, (C) T-large Granular Lymphocyte (LGL) cells, and (D) γδ T cells. LGL Phenotypes Were Defined as CD57+CD3+ (T-LGL) Cells or CD57+CD56+ (NK-LGL) Cells

      References

        • Talpaz M.
        • Shah N.P.
        • Kantarjian H.
        • et al.
        Dasatinib in imatinib-resistant Philadelphia chromosome-positive leukemias.
        N Engl J Med. 2006; 354: 2531-2541
        • Cortes J.E.
        • Saglio G.
        • Kantarjian H.M.
        • et al.
        Final 5-year study results of DASISION: the dasatinib versus imatinib study in treatment-naive chronic myeloid leukemia patients trial.
        J Clin Oncol. 2016; 34: 2333-2340
        • Hochhaus A.
        • Saglio G.
        • Hughes T.P.
        • et al.
        Long-term benefits and risks of frontline nilotinib vs imatinib for chronic myeloid leukemia in chronic phase: 5-year update of the randomized ENESTnd trial.
        Leukemia. 2016; 30: 1044-1054
        • Saussele S.
        • Krauss M.P.
        • Hehlmann R.
        • et al.
        Impact of comorbidities on overall survival in patients with chronic myeloid leukemia: results of the randomized CML study IV.
        Blood. 2015; 126: 42-49
        • Douxfils J.
        • Haguet H.
        • Mullier F.
        • Chatelain C.
        • Graux C.
        • Dogne J.M.
        Association between BCR-ABL tyrosine kinase inhibitors for chronic myeloid leukemia and cardiovascular events, major molecular response, and overall survival: a systematic review and meta-analysis.
        JAMA Oncol. 2016; 2: 625-632
        • Mahon F.X.
        • Rea D.
        • Guilhot J.
        • et al.
        Discontinuation of imatinib in patients with chronic myeloid leukaemia who have maintained complete molecular remission for at least 2 years: the prospective, multicentre stop imatinib (STIM) trial.
        Lancet Oncol. 2010; 11: 1029-1035
        • Ross D.M.
        • Branford S.
        • Seymour J.F.
        • et al.
        Safety and efficacy of imatinib cessation for CML patients with stable undetectable minimal residual disease: results from the TWISTER study.
        Blood. 2013; 122: 515-522
        • Imagawa J.
        • Tanaka H.
        • Okada M.
        • et al.
        Discontinuation of dasatinib in patients with chronic myeloid leukaemia who have maintained deep molecular response for longer than 1 year (DADI trial): a multicentre phase 2 trial.
        Lancet Haematol. 2015; 2: e528-e535
        • Etienne G.
        • Guilhot J.
        • Rea D.
        • et al.
        Long-term follow-up of the French stop imatinib (STIM1) study in patients with chronic myeloid leukemia.
        J Clin Oncol. 2017; 35: 298-305
        • 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: 1809-1820
        • Yoshida C.
        • Fletcher L.
        • Ohashi K.
        • et al.
        Harmonization of molecular monitoring of chronic myeloid leukemia therapy in Japan.
        Int J Clin Oncol. 2012; 17: 584-589
        • Cross N.C.
        • White H.E.
        • Colomer D.
        • et al.
        Laboratory recommendations for scoring deep molecular responses following treatment for chronic myeloid leukemia.
        Leukemia. 2015; 29: 999-1003
        • Liu W.
        • Putnam A.L.
        • Xu-Yu Z.
        • et al.
        CD127 expression inversely correlates with FoxP3 and suppressive function of human CD4+ Treg cells.
        J Exp Med. 2006; 203: 1701-1711
        • Radich J.P.
        • Dai H.
        • Mao M.
        • et al.
        Gene expression changes associated with progression and response in chronic myeloid leukemia.
        Proc Natl Acad Sci U S A. 2006; 103: 2794-2799
        • Collins Jr., R.H.
        • Shpilberg O.
        • Drobyski W.R.
        • et al.
        Donor leukocyte infusions in 140 patients with relapsed malignancy after allogeneic bone marrow transplantation.
        J Clin Oncol. 1997; 15: 433-444
        • Chen C.I.
        • Koschmieder S.
        • Kerstiens L.
        • et al.
        NK cells are dysfunctional in human chronic myelogenous leukemia before and on imatinib treatment and in BCR-ABL-positive mice.
        Leukemia. 2012; 26: 465-474
        • Kreutzman A.
        • Juvonen V.
        • Kairisto V.
        • et al.
        Mono/oligoclonal T and NK cells are common in chronic myeloid leukemia patients at diagnosis and expand during dasatinib therapy.
        Blood. 2010; 116: 772-782
        • Ross D.M.
        • Branford S.
        • Seymour J.F.
        • et al.
        Patients with chronic myeloid leukemia who maintain a complete molecular response after stopping imatinib treatment have evidence of persistent leukemia by DNA PCR.
        Leukemia. 2010; 24: 1719-1724
        • Ilander M.
        • Olsson-Stromberg U.
        • Schlums H.
        • et al.
        Increased proportion of mature NK cells is associated with successful imatinib discontinuation in chronic myeloid leukemia.
        Leukemia. 2017; 31: 1108-1116
        • Hughes A.
        • Clarson J.
        • Tang C.
        • et al.
        CML patients with deep molecular responses to TKI have restored immune effectors and decreased PD-1 and immune suppressors.
        Blood. 2017; 129: 1166-1176
        • Rousselot P.
        • Charbonnier A.
        • Cony-Makhoul P.
        • et al.
        Loss of major molecular response as a trigger for restarting tyrosine kinase inhibitor therapy in patients with chronic-phase chronic myelogenous leukemia who have stopped imatinib after durable undetectable disease.
        J Clin Oncol. 2014; 32: 424-430