Dasatinib discontinuation in patients with chronic-phase chronic myeloid leukemia and stable deep molecular response: the DASFREE study.

Abstract Treatment-free remission (TFR) in patients with chronic myeloid leukemia in chronic phase (CML-CP) is considered a feasible option, especially with the ability of second-generation tyrosine kinase inhibitors to induce higher rates of sustained deep molecular response (DMR). DASFREE is an open-label, single-arm, multicenter phase II trial assessing TFR after dasatinib discontinuation in patients with CML-CP (N = 84). At 2 years, TFR was 46% in all patients. Multivariate analyses revealed statistically significant associations between 2-year TFR and duration of prior dasatinib (≥median; p = .0051), line of therapy (first line; p = .0138), and age (>65 years; p = .0012). No disease transformation occurred, and the most common adverse events experienced off treatment were musculoskeletal (observed in 30 patients); however, dasatinib withdrawal events were reported in nine patients (11%) by the investigator. Overall, these findings support the feasibility of discontinuing dasatinib for patients with CML-CP in sustained DMR in the first line and beyond.


Introduction
Tyrosine kinase inhibitors (TKIs) have vastly improved long-term outcomes for patients with chronic myeloid leukemia in chronic phase (CML-CP). In particular, second-generation TKIs have been shown to induce higher rates of deep molecular responses (DMR) and a more rapid decline in BCR-ABL1 transcript levels compared with imatinib in newly diagnosed patients [1][2][3]. Consequently, DMR is now increasingly being considered as being correlative of disease burden reduction and is also the most relevant clinical endpoint for patients wishing ultimately to stop treatment [4][5][6]. A reduction in leukemic clone burden is associated with reduced rates of transformation to accelerated phase (AP) or blast crisis (BC) CML; however, it is not clear if the absence of detectable BCR-ABL1 transcripts is synonymous with a cure for CML [4][5][6]. Despite the persistence of the leukemic clone in such cases, a functional cure might still be possible, and the need for lifelong TKI treatment is now being challenged [7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22]. Furthermore, patients might be motivated to stop therapy for reasons that include toxicity, a desire for improvement in quality of life, and the financial burden associated with indefinite TKI treatment [23][24][25].
The consensus on the use of TFR in clinical practice, however, is still evolving. Results from ongoing TFR trials will help to provide more confirmatory data on long-term outcomes. Furthermore, many clinical trials on TFR have been unable to identify strong predictors of successful TKI discontinuation consistently. The optimal duration of DMR before entering TFR also remains to be determined. DASFREE (NCT01850004) is the largest clinical trial to date of patients with CML-CP and sustained dasatinib-induced DMR (specifically MR 4.5 ) who discontinued dasatinib across all lines of therapy. In this ongoing, open-label, single-arm phase 2 study, the primary objective was to assess the rate of TFR (defined herein as the maintenance of MMR following treatment cessation) at 1 year. Herein we report findings for this primary objective and additional pre-specified analyses for patients with a minimum of 2 years of follow-up after stopping dasatinib.

Study design and eligibility
Patients aged 18 years with CML-CP who received dasatinib treatment (as first-or subsequent-line therapy) for a minimum of 2 years at the time of enrollment and had confirmed dasatinib-induced MR 4.5 for at least 1 year prior to study entry were eligible. Patients were required to have achieved a 1-log reduction in BCR-ABL1 transcript levels relative to baseline as determined by local standards or BCR-ABL1 10% on the IS at 3 months with current dasatinib therapy. Additional eligibility criteria information are listed in the Online Supplementary Material.

Study endpoints and assessments
Once all eligibility criteria were met, dasatinib was discontinued. The primary objective was the rate of TFR (proportion of subjects who maintained MMR [BCR-ABL < 0.1%]) at 1 year following dasatinib discontinuation without restarting treatment. If MMR was lost, patients resumed dasatinib at the last dose received prior to stopping treatment. Key secondary endpoints include event-free survival (EFS; survival with no loss of MMR), BCR-ABL1 kinetics, rate of transformation to CML-AP/BC, and progression-free survival (PFS). Patients who discontinued from the study were censored on the date of their last molecular assessment. PFS was defined as survival without progression to CML-AP/BC or death due to any cause. Time in prior MR 4.5 was assessed retrospectively; consent for 10 patients could not be collected, thus information is available for only 74 of the 84 patients enrolled in the trial. Both intolerance and resistance were determined by the investigator. Additional study endpoints and further information on assessments and molecular testing are shown in the Online Supplementary Material.
Key exploratory endpoints included the frequency of adverse events (AEs) and serious AEs (SAEs) occurring off treatment after discontinuing and after restarting on dasatinib, the rate of MMR recapture after reinitiating dasatinib, and identification of prognostic factors in relation to TFR maintenance. AEs and SAEs were assessed according to the National Cancer Institute's Common Terminology Criteria for Adverse Events version 4.03 [29]. Withdrawal events were defined as any AE occurring and/or worsening due to treatment cessation, as determined by the investigator.
The primary objective was presented as a percentage with an exact 95% Clopper-Pearson confidence interval (CI). Additional statistical analyses are provided in the Online Supplementary Material. Data for the primary endpoint are based on a cutoff date of October 2017, at which time all patients who discontinued dasatinib had 1 or more years of follow-up. Data presented herein are based on a cutoff date of December 2018, at which time patients had a minimum follow-up of 2 years.

Ethics
All patients provided written informed consent in accordance with the Declaration of Helsinki and local guidelines before study entry. The study protocol was approved by all Institutional Review Boards of each participating center, as well as each center's Ethics Committee and competent national authority.

Accrual and patient characteristics
A total of 84 patients enrolled between February 2014 and June 2016 at 22 study centers in six countries discontinued dasatinib treatment. Baseline patient characteristics are presented in Table 1. Thirty-seven patients (44%) who were receiving first-line dasatinib and 47 patients (56%) who were receiving dasatinib in subsequent lines prior to stopping treatment were enrolled. Of those on later lines, 53% were deemed resistant to their first-line TKI and 38% were considered intolerant of a prior TKI by the investigator, including two patients (2%) on prior imatinib and nilotinib. Four patients (9%) on later lines of dasatinib were not classified as resistant to or intolerant of prior therapy by the investigator; the status of two of these patients is unknown, while one patient switched to dasatinib following closure of a first-line ponatinib trial and the other switched from imatinib to dasatinib due to out-of-pocket costs. In 74 patients with available information on duration of prior time in MR 4.5 , the median duration of MR 4.5 prior to discontinuation was 28 months (range, 13-116 months) and was similar regardless of line of therapy (Supplementary Figure  S1). The median time from CML diagnosis to discontinuation was 69 months (range, 29-244 months), and the median dose received prior to stopping dasatinib was 100 mg (range, 20-150 mg) once daily. With a minimum follow-up of 2 years, 11 patients (13%) discontinued the study (Supplementary Table S1). Two patients discontinued the study (while maintaining MMR) due to relocation, one patient was non-compliant with study visits and discontinued the study before undergoing molecular assessments, and one patient discontinued after being diagnosed with a metastatic malignancy unrelated to CML.
At 2 years, PFS was 99% (95% CI: 96, 100) in all patients, 100% (95% CI: 100, 100) in first-line patients, and 98% (95% CI: 93, 100) in subsequent-line patients. No patients progressed to CML-AP/BC. However, one patient was diagnosed with ovarian cancer, discontinued the study, and died 1 month later. This patient was categorized as having discontinued due to death for this database lock and is thus reflected in the survival results.
Associations  (Figure 3). Median duration of prior dasatinib was 55 months in patients aged 65 years compared with 45 months in patients aged <65 years. Time on prior TKIs (including prior dasatinib) was further assessed in patients who lost or maintained MMR after discontinuing first-or subsequent-line treatment, and was found to be numerically shorter in patients who lost MMR than in patients who maintained MMR (Supplementary Table S2). Seven patients had a prior history of pleural effusion while receiving dasatinib before discontinuation; all episodes were resolved at the time of study entry. None of these patients have lost MMR.

Safety
AEs of any grade and cause occurred in 67% of patients off treatment and 76% of patients after restarting treatment. All-causality AEs reported in 4% of patients off treatment following dasatinib discontinuation or on treatment after restarting dasatinib are  Loss and recovery of MMR and MR 4.5 at 2 years. a A 52-year-old male with low-risk CML treated with first-line dasatinib 100 mg once daily for 40 months discontinued the study after 33 months in MR 4.5 . This patient maintained BCR-ABL1 transcript levels between 0.0017 and 0.01 for 3 years and had an increase to 0.1% in month 39, which was confirmed on a second occasion in month 42 (0.11%). b One patient lost MMR and restarted treatment. This patient discontinued the study after only one followup molecular assessment and therefore was not considered evaluable for molecular response. CML: chronic myeloid leukemia; MMR: major molecular response; MR 4.5 : BCR-ABL1 0.0032% on the International Scale.  AE: adverse event; TKI: tyrosine kinase inhibitor. a Withdrawal events were defined as any adverse event occurring and/or worsening due to treatment cessation, as determined by the investigator. described in Table 2. Dasatinib-related AEs of any grade and grade 3/4 were experienced in 22 patients (49%) and 4 patients (9%) on treatment after restarting dasatinib, respectively, with 2 of 45 patients (4%) experiencing SAEs. Pleural effusion (any grade) was reported in one patient (1%) off treatment and in three patients (7%) after restarting dasatinib; the effusion was grade 3 in one patient. After restarting treatment, three patients (7%) experienced an AE leading to treatment discontinuation; two of these patients discontinued due to pleural effusion, while a metastatic ovarian cancer unrelated to CML led to death in one patient. As noted earlier, withdrawal events in this study were defined as any AE occurring and/or worsening after dasatinib discontinuation and determined by the investigator to be a result of stopping dasatinib. In total, study investigators deemed 15 events in nine patients (11%) to be related to dasatinib withdrawal ( Table 2). After stopping treatment, 30 musculoskeletal events were reported; 11 of these events were attributed to dasatinib withdrawal. Two arterial hypertension episodes (observed in one patient) were also considered withdrawal events. At the time of this data cutoff, six events remain unresolved, although all are grade 1 and have not required therapy.

Discussion
While other studies have shown promising TFR data with dasatinib, DASFREE is the largest dasatinib trial to date to examine TFR in patients discontinuing any line of therapy. In DASFREE, 48% (95% CI: 37, 59) of all patients maintained TFR at 1 year following dasatinib discontinuation, and remission was found to be durable at 2 years (with one late relapse at month 39). Furthermore, patients who lost MMR and restarted treatment quickly regained their response (median time to regain MMR and MR 4.5 was 2 and 3 months, respectively). Three factors identified in this study as prognostic for TFR at 2 years (duration of prior dasatinib [median], line of therapy [first line], and age [>65]) might help to predict which patients are more likely to remain in TFR. Overall, these findings not only provide new information on achieving improved TFR outcomes with dasatinib but also emphasize the practicality of stopping dasatinib across all lines of therapy.
The findings from the DASFREE study are comparable to other dasatinib discontinuation trials: in D-STOP, 63% maintained MR 4 at 1 year, and the phase II Japanese Dasatinib Discontinuation (DADI) study reported an estimated overall rate of TFR was 44% at 3 years [11,14]. Findings for other second-generation TKIs have been similar: the ENESTfreedom and ENESTop trials with first-and second-line nilotinib reported TFR rates of 47% and 48%, respectively, at 144 weeks [9,12,15,19]. Finally, in STOP 2 G-TKI (dasatinib and nilotinib discontinuation), 63% of patients remained in TFR at 1 year [16]. It is important to note, however, that the capability to make an in-depth, comprehensive comparison of results across all discontinuation trials is limited by differences in eligibility criteria and varying DMR and molecular relapse definitions.
Although several TKI discontinuation trials have sought to identify prognostic factors that could help predict successful TFR [7,30], a consensus has yet to be reached. In this study, duration of prior dasatinib (median; p ¼ .0051), line of therapy (first line; p ¼ .0138), and age (>65 years; p ¼ .0012) represent three prognostic factors for 2-year TFR. Although prior line of therapy is considered a prognostic factor for TFR, these findings suggest that TFR is certainly achievable regardless of the line of discontinued therapy. Identification of these factors adds to the growing body of evidence in delineating key prognostic factors for successful TFR. Similar findings were also observed for age in the Imatinib Suspension and Validation (ISAV) study, where an inverse relationship between patient age and risk of relapse was reported (95% of patients aged <45 years relapsed vs 42% of patients aged 45 to <65 and 33% of patients aged 65 years), although this may be an effect of a smaller patient population [30]. In addition, other discontinuation trials have established the presence of imatinib resistance as a significant risk factor for molecular relapse [14,16]. This association was particularly evident in the DADI trial, in which imatinib-resistant patients relapsed more quickly and with a significantly shorter doubling time of BCR-ABL1 transcript levels [14]. In DASFREE, however, an important proportion of resistant patients (42%) were able to maintain TFR at 2 years. Although some prognostic factors were identified in this study, other ongoing trials are seeking to identify additional factors predictive of TFR outcomes, including the possible effect of pleural effusion [22].
No significant association between TFR and duration of prior TKI therapy (other than dasatinib) or prior MR 4.5 was identified in this study. However, treatment duration and duration of DMR have been identified in other studies as prognostic factors for TFR when a TKI is discontinued [11,[30][31][32][33][34]. Results from the multicenter Stop Imatinib (STIM) trial, in which 40% of patients entering TFR with MR 4.5 remained in remission at 18 months, indicated an imatinib treatment duration of 50 months as a prognostic factor of TFR [11]. In addition, prognostic analysis of patients in EURO-SKI (European Stop Tyrosine Kinase Inhibitor Study) suggested that the duration of DMR had the greatest impact on the success of stopping TKI treatment [21]. In EURO-SKI it was recommended that the patient should be in MR 4 for 3 or more years before stopping [21]. Patients in DASFREE were required to be in MR 4.5 for 1 or more years (Online Supplementary Figure S1) and on dasatinib for 2 or more years prior to enrollment, which might be considered aggressive in comparison to the treatment duration and duration of response criteria in similar discontinuation trials.
Patients considering TFR are not only concerned about the risk of relapse but also fear that their disease might not remain responsive to treatment following relapse. Across TFR trial reports, all patients who have relapsed after stopping treatment have remained sensitive to re-treatment and regained molecular responses MMR [22]. However, one patient (in ENESTfreedom) with loss of MMR was found to have a detectable BCR-ABL1 mutation, though the significance of this finding is unclear because the mutation might have been present prior to stopping nilotinib [19]. In DASFREE, PFS was high (99%) at 2 years and no cases of disease progression to CML AP/BC were observed. All patients who relapsed and were evaluable for molecular assessment after restarting dasatinib regained MMR, and 93% regained MR 4.5 .
Withdrawal events after discontinuation of TKI therapy are well described in the current literature on TFR [7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22]. In DASFREE, only nine patients (11%) off treatment reported symptoms that investigators considered related to withdrawal, most of which resolved on their own without the use of concomitant therapy. Furthermore, AEs observed in patients who restarted treatment were found to be consistent with the known safety profile of dasatinib, with no new safety signals being reported in this study. To our knowledge, the specific definition of withdrawal events utilized in DASFREE (investigator assessed and defined as any AE occurring and/or worsening in severity after discontinuation and related to dasatinib discontinuation) is unique to this trial and unlike other discontinuation trials that defined withdrawal events to be solely musculoskeletal [16,21,22].
One limitation of this study is the retrospective manner in which some data were collected, which might account for some of the differences between DASFREE and other TFR studies. This might play a particularly important role with regard to the low rate of withdrawal events in this study, which might be associated with a definition of withdrawal that requires an investigator assessment of relatedness. Conversely, most patients discontinued treatment following the first published description of TKI withdrawal syndrome, so most investigators were aware of this phenomenon. A second possible limitation of this study, which is also characteristic of other discontinuation trials, is sample size. A better understanding of the long-term durability of TFR, including identification of definitive prognostic factors and longer follow-up, is needed before fully incorporating these findings into clinical decision-making, particularly given the apparently low risk of attempting TKI discontinuation. Long-term (5-year) follow-up is planned, including an assessment of the stability of MR 4.5 achieved on dasatinib, recurrence rate, patients' quality of life off treatment, and the ability to regain a molecular response in patients who relapse after discontinuation.
In summary, DASFREE results provide clinically relevant information on TFR for many patients with CML-CP who achieve stable DMR with dasatinib. Suggestions on appropriate treatment duration and duration of response criteria prior to discontinuing, as well as newly identified prognostic factors, are novel contributions to those considering TFR. Furthermore, these data support the practicality of TFR in patients with resistance to or intolerance of prior TKI treatment and, perhaps most importantly, confirm the feasibility and safety of dasatinib discontinuation in patients stopping first-line treatment.