NCCN Categories of Evidence and Consensus
Category 1: Based upon high-level evidence, there is uniform NCCN consensus that the intervention is appropriate.
Category 2A: Based upon lower-level evidence, there is uniform NCCN consensus that the intervention is appropriate.
Category 2B: Based upon lower-level evidence, there is NCCN consensus that the intervention is appropriate.
Category 3: Based upon any level of evidence, there is major NCCN disagreement that the intervention is appropriate.
All recommendations are category 2A unless otherwise noted.
Clinical trials: NCCN believes that the best management for any cancer patient is in a clinical trial. Participation in clinical trials is especially encouraged.
Chronic myeloid leukemia (CML) accounts for 15% of adult leukemias. The median age of disease onset is 67 years; however, CML occurs in all age groups (SEER statistics). In 2018, an estimated 8,430 people will be diagnosed with CML in the United States, and 1,090 people will die of the disease.1
CML is defined by the presence of Philadelphia chromosome (Ph) in a patient with a myeloproliferative neoplasm (MPN). Ph results from a reciprocal translocation between chromosomes 9 and 22 [t(9;22] that gives rise to a BCR-ABL1 fusion gene; the product of this fusion gene is a protein with deregulated tyrosine kinase activity (p210) that plays a central role in the pathogenesis of CML.2 Another fusion protein, p190, is also produced, usually in the setting of Ph-positive acute lymphoblastic leukemia. p190 is detected only in 1% of patients with CML.3
CML occurs in 3 different phases (chronic, accelerated, and blast phases) and is usually diagnosed in the chronic phase. Untreated chronic phase CML (CP-CML) will eventually progress to advanced phase in 3 to 5 years.4 Gene expression profiling has shown a close correlation of gene expression between accelerated phase CML (AP-CML) and blast phase CML (BP-CML). The bulk of the genetic changes in progression occur in the transition from CP-CML to AP-CML.5 The activation of beta-catenin signaling pathway in CML granulocyte-macrophage progenitors (which enhances the self-renewal activity and leukemic potential of these cells) may also be a key pathobiologic event in the evolution to BP-CML.6
The NCCN Guidelines for CML discuss the clinical management of CML in all 3 phases (chronic, accelerated, and blast). Evaluation for diseases other than CML, as outlined in the NCCN Guidelines for MPN, is recommended for all patients with BCR-ABL1–negative MPN (to view the most recent version of these guidelines, visit NCCN.org).
Diagnosis and Workup
Initial evaluation should consist of a history and physical exam, including palpation of spleen,
Bone marrow cytogenetics should be done at initial workup to detect additional chromosomal abnormalities in Ph-positive cells (ACA/Ph+), also known as clonal cytogenetic evolution.7 The prognostic significance of ACA/Ph+ is related to the specific chromosomal abnormality and often other features of accelerated phase.8–12 The presence of “major route” ACA/Ph+ (trisomy 8, isochromosome 17q, second Ph, and trisomy 19) at diagnosis may have a negative prognostic impact on survival and disease progression to accelerated or blast phase.13–15 However, in a more recent analysis that evaluated the outcomes of patients with CP-CML (with or without ACA) treated with tyrosine kinase inhibitors (TKIs) in prospective studies, the presence of ACA/Ph+ at the time of diagnosis was not associated with worse prognosis.16 Patients with ACA/Ph+ at diagnosis should be watched carefully for evidence of therapy failure. Clonal cytogenetic evolution in Ph-negative cells has also been reported in a small subset of patients during the course of imatinib therapy.17–22 The most common abnormalities include trisomy 8 and loss of Y chromosome. Previous work suggested that the overall prognosis of Ph-negative CML with clonal evolution is good and is dependent on response to imatinib therapy.21 Recently, however, the presence of chromosome abnormalities other than loss of Y chromosome has been associated with decreased survival in patients with CP-CML treated with various TKIs, suggesting that closer follow-up is indicated until definitive data are available.23 Progression to myelodysplastic syndromes (MDS) and acute myeloid leukemia have been reported in patients with monosomy 7.24,25
If bone marrow evaluation is not feasible, fluorescence in situ hybridization (FISH) on a peripheral blood specimen with dual probes for BCR and ABL1 genes is an acceptable method to confirm the diagnosis of CML. Interphase FISH is performed on peripheral blood but is associated with a background level of 1%–5% depending on the specific probe used in the assay.26 Hypermetaphase FISH is more sensitive and can analyze up to 500 metaphases at a time, but it is applicable only to dividing cells in the bone marrow.27 Double-fusion FISH is also associated with low false-positive rates and can detect all variant translocations of the Ph-chromosome.28
Quantitative RT-PCR (qPCR) should be performed at initial workup to establish the presence of quantifiable BCR-ABL1 mRNA transcripts at baseline. qPCR, usually performed on peripheral blood, is the most sensitive assay available for the measurement of BCR-ABL1 mRNA and it can detect 1 CML cell in a background of ≥100,000 normal cells. qPCR results can be expressed in various ways, for instance as the ratio of BCR-ABL1 transcript numbers to the number of control gene transcripts.29 An international scale (IS) has been proposed to standardize molecular monitoring with qPCR across different laboratories with the use of 1 of 3 control genes (BCR, ABL1, or GUSB) and a qPCR assay with a sensitivity of at least 4-log reduction from the standardized baseline.30 In recent years, IS has become the gold standard of expressing qPCR values. More details on qPCR monitoring using IS are provided on MS-10 (in these guidelines, at NCCN.org).
BCR-ABL1 transcripts in the peripheral blood at very low levels (1–10 of 108 peripheral blood leukocytes) can also be detected in approximately 30% of normal individuals, and the incidence of BCR-ABL1 transcripts increases with advancing age in healthy individuals.31,32 TKI therapy is not indicated, as the risk of developing CML for these individuals is extremely low.
Management of Chronic Phase CML
Sokal and Euro scoring systems have been used for the risk stratification of patients into 3 risk groups (low, intermediate, and high) in clinical trials evaluating TKIs (see CML-A; available online, in these guidelines, at NCCN.org).33,34 The Sokal score is based on the patient's age, spleen size, platelet count, and percentage of blasts in the peripheral blood.33 The Euro score includes eosinophils and basophils in the peripheral blood in addition to the same clinical variables used in the Sokal score.34
European Treatment and Outcome Study (EUTOS) score is based only on the percentage of basophils in the blood and spleen size. The predictive value of EUTOS score was validated in a cohort of 2,060 patients enrolled in studies of first-line treatment with imatinib-based regimens.35 EUTOS score was better than Sokal and Euro score in predicting the probability of achieving a complete cytogenetic response (CCyR) at 18 months and 5-year progression-free survival (PFS). However, the predictive value of EUTOS score has not been confirmed in subsequent studies by other investigators, and additional studies are needed to validate the EUTOS score.36–38
Determination of risk score using either the Sokal or Hasford (Euro) scoring systems before initiation of TKI therapy is recommended for patients diagnosed with CP-CML (see CML-1; page 1110).
Long-term efficacy data from randomized phase III studies for first-line TKI therapy in patients with newly diagnosed CP-CML are summarized in Table 1.39–42 In summary, (1) all TKIs are highly effective in newly diagnosed CP-CML, with long-term overall survival (OS) approaching that of aged-matched controls; (2) second-generation TKIs, compared with imatinib, generally result in faster cytogenetic and molecular responses, with less progression to advanced phase CML; and (3) yet, in randomized clinical trials, there are no differences in OS between imatinib and second-generation TKIs.
The selection of first-line TKI therapy (bosutinib, dasatinib, imatinib, or nilotinib) in a given patient should be based on the risk score, toxicity profile of TKI, patient's age, ability to tolerate therapy, and the presence of comorbid conditions. Allogeneic hematopoietic cell transplantation (HCT) is no longer recommended as a first-line treatment option for patients with CP-CML.
Imatinib, 800 mg, is not recommended as initial therapy, given the recent data showing superior efficacy of second-generation TKIs (dasatinib, nilotinib, and bosutinib) in newly diagnosed CP-CML. Data from randomized phase III studies that have evaluated high-dose imatinib as first-line therapy for CP-CML suggest that imatinib, 800 mg, was not associated with lower rates of disease progression than imatinib, 400 mg, in any of these studies, despite improved early responses (Table 2).43–45 Imatinib, 800 mg, was also associated with higher rates of dose interruption, reduction, or discontinuation
First-Line TKI Therapy for CP-CML: Long-Term Follow-Up Data From Phase III Studies
High-Dose Imatinib as First-Line Therapy for CP-CML: Long-Term Follow-Up Data From Phase III Studies
The prospective studies evaluating imatinib, 800 mg, daily found that increased toxicity of that dose forced decreasing dose to approximately 600 mg, daily when considering the actually administered dose intensity.43–45 Additionally, the French SPIRIT trial reported superior major molecular response (MMR) rates in patients treated with imatinib, 600 mg daily compared with 400 mg daily.46 These data suggest that imatinib, 600 mg, daily may be closer to the optimal dose than 400 mg.
Clinical Considerations for The Selection of First-Line Therapy
Risk Stratification: Imatinib (400 mg daily) and second-generation TKIs (dasatinib, 100 mg once daily; nilotinib, 300 mg twice daily; and bosutinib, 400 mg daily) are all appropriate options for first-line TKI therapy for patients with CP-CML across all risk scores (see CML-2; page 1112).39–42
Disease progression is more frequent in patients with intermediate- or high-risk score, and prevention of disease progression to AP-CML or BP-CML is the primary goal of TKI therapy in patients with CP-CML. Second-generation TKIs are associated with lower risk of disease progression than imatinib and are therefore preferred for patients with an intermediate- or high-risk Sokal or Euro score.
Second-generation TKIs also result in quicker molecular responses and higher rates of deep molecular responses (MMR [BCR-ABL1 ≤0.1% IS] and MR4.5 [≥4.5-log reduction in BCR-ABL1 transcripts from baseline]) in patients with CP-CML across all risk scores (Table 3), which may facilitate subsequent discontinuation of TKI therapy in selected patients.40–42 Therefore, second-generation TKIs may be preferred over imatinib for younger patients, particularly women, because the achievement of a deep and rapid molecular response may allow eventual discontinuation of TKI therapy for fertility purposes. Imatinib may be preferred for older patients with comorbidities, especially cardiovascular.
Toxicity Profile: All of the TKIs are fairly well tolerated. Because bosutinib, dasatinib, and nilotinib have very good efficacy in the upfront setting, differences in their potential toxicity profiles may inform the selection of either of these TKIs as initial therapy. Nilotinib or bosutinib may be preferred for patients with a history of lung disease or deemed to be at risk of developing pleural effusions. Dasatinib or bosutinib may be preferred in patients with a history of arrhythmias, heart disease, pancreatitis, or hyperglycemia.
Adverse events of first-line TKI therapy in patients with CP-CML reported in phase III randomized
First-Line TKI Therapy for CP-CML: MR Rates According to Sokal or Euro Risk Score
Imatinib: Chronic fatigue (mostly correlated with musculoskeletal pain and muscular cramps) is a major factor reducing quality of life.47 Hypophosphatemia and decrease in bone mineral density has been noted in a small group of patients, suggesting that monitoring bone health should be considered for patients taking imatinib.48,49 Skin hypopigmentation has also been reported as a side effect of imatinib and is reversible on discontinuation or dose reduction.50,51
Dasatinib: In the DASISION study, the incidences of grade 3/4 hematologic toxicities (anemia, neutropenia, and thrombocytopenia) were higher for dasatinib than imatinib. Nonhematologic adverse events such as muscle spasms, peripheral edema, and hypophosphatemia were more frequent with imatinib. Discontinuation of therapy because of drug-related adverse events occurred in 16% and 7% of patients in the dasatinib and imatinib arms, respectively.40 Dasatinib is also associated with significant but reversible inhibition of platelet aggregation that may contribute to bleeding in some patients, especially if accompanied by thrombocytopenia.52
Pleural effusion was more common with dasatinib (28%) than with imatinib (<1%).40 The occurrence of pleural effusion is significantly reduced with dasatinib, 100 mg once daily compared with 70 mg twice daily.53 Patients with prior cardiac history, hypertension, and those receiving twice-daily dosing of dasatinib at 70 mg are at increased risk of developing pleural effusions. Close monitoring and timely intervention are necessary for patients at risk of developing pleural effusions.
Reversible pulmonary arterial hypertension has been reported as a rare but serious side effect of dasatinib.54,55 In the DASISION study, pulmonary hypertension was reported in 5% of patients compared with 0.4% of patients treated with imatinib.40 Evaluation for signs and symptoms of underlying cardiopulmonary disease before starting and during treatment with dasatinib is recommended. If pulmonary arterial hypertension is confirmed, dasatinib must be permanently discontinued.
The recommended starting dose of dasatinib is 100 mg once daily for patients with CP-CML. Limited data available from small cohorts of patients suggest that lower doses of dasatinib may potentially have similar efficacy.56,57 Treatment interruption of dasatinib at standard dose and reintroduction of dasatinib at a lower dose of 40 mg twice daily also resolved all pulmonary complications without recurrence.58 However, the minimum effective dose has not been established in randomized clinical trials. Reintroduction of dasatinib at 50 mg (20 mg with careful monitoring in selected patients) should be considered for patients with clinically significant intolerance to dasatinib at 100 mg once daily to avoid serious adverse events necessitating the discontinuation of dasatinib (eg, pleural effusion, myelosuppression).
Adverse Events of First-Line TKI Therapy in CP-CML
Nilotinib: In the ENESTnd study, nonhematologic adverse events such as nausea, diarrhea, vomiting, muscle spasm, and peripheral edema of any grade were higher for patients receiving imatinib.41 Conversely, rash and headache were higher with nilotinib. Grade 3 or 4 neutropenia was more frequent in the imatinib group, whereas thrombocytopenia and anemia were similar in both groups. Electrolyte abnormalities and elevations in lipase, glucose, and bilirubin were more frequent with nilotinib than with imatinib. Patients with a previous history of pancreatitis may be at greater risk of elevated serum lipase levels. The overall incidences of adverse events leading to discontinuation of therapy were comparable in the nilotinib, 300 mg, twice daily arm and imatinib arms (12% and 14%, respectively) and slightly higher in the nilotinib, 400 mg, twice daily arm (20%).
Nilotinib labeling contains a black box warning regarding the risk of QT interval prolongation, and sudden cardiac death has been reported in patients receiving nilotinib. QT interval prolongation could be managed with dose reduction. Electrolyte abnormalities should be corrected before start of treatment with nilotinib, and electrolytes should be monitored periodically. Drugs that prolong QT interval should be avoided. Electrocardiogram should be obtained to monitor the QT interval at baseline, 7 days after start of nilotinib, and periodically thereafter, and after any dose adjustments. Patients with cardiovascular risk factors should be referred to a cardiologist.
Nilotinib is associated with an increased risk of peripheral arterial occlusive disease (PAOD).59–61 Patients should be evaluated for pre-existing PAOD and vascular risk factors before starting and during treatment with nilotinib. If PAOD is confirmed, nilotinib should be permanently discontinued.
Bosutinib: In the BFORE study, diarrhea, increased alanine aminotransferase (ALT), and aspartate aminotransferase (AST) were more common with bosutinib, whereas muscle spasms and peripheral edema were more common with imatinib. Grade 3/4 thrombocytopenia was higher with bosutinib and grade 3/4 neutropenia was higher with imatinib. Grade 3/4 anemia was similar in both groups. Discontinuation of therapy because of drug-related adverse events occurred in 14% of patients in the bosutinib group compared with 11% in the imatinib group. Increased ALT (5%) and increased AST increase (2%) were the most common adverse events leading to discontinuation of bosutinib. However, no hepatotoxicity-related fatalities occurred during the study.42
Management of Hematologic Toxicities of TKI Therapy: Cytopenias (anemia, neutropenia, and thrombocytopenia) should be managed with transient interruptions of TKI therapy and dose modifications. Please see the package insert for full prescribing information, available at www.fda.gov, for the recommended dose modifications of specific TKI therapy.
Assessment of reticulocyte count, ferritin, iron saturation, vitamin B12, and folate and correction of nutritional deficiencies if present is recommended for patients with grade 3/4 anemia. Red blood cell transfusions are indicated in symptomatic patients. Myeloid growth factor support can be used in combination with TKI therapy for the management of neutropenia.62,63 The use of erythropoiesis-stimulating agents (ESAs) did not impact survival or cytogenetic response rate, but was associated with a higher thrombosis rate in patients with CP-CML.64 Recent guidelines from the Centers for Medicare & Medicaid Services (CMS) and the FDA do not support the use of ESAs in patients with myeloid malignancies.
Monitoring Response to TKI Therapy
Response to TKI therapy is determined by the measurement of hematologic (normalization of peripheral blood counts), cytogenetic (decrease in the number of Ph-positive metaphases using bone marrow cytogenetics), and molecular responses (decrease in the amount of BCR-ABL1 chimeric mRNA using qPCR). The criteria for hematologic, cytogenetic, and molecular response are summarized in CML-D (page 1115).
Conventional bone marrow cytogenetics is the standard method for monitoring cytogenetic responses, and clinical trial response analyses are most often based on conventional bone marrow cytogenetics. If conventional bone marrow cytogenetics showed no analyzable metaphases, cytogenetic response can be evaluated by FISH; however, it has a false-positive rate of 1% to 10%.65,66 Although some investigators have reported that interphase FISH can be used to monitor CCyR, end points for TKI failure have not been defined on the basis of FISH analysis.67,68 The panel feels that FISH has been inadequately studied for monitoring response to TKI therapy. Therefore, FISH is not generally recommended for monitoring response if conventional cytogenetics or qPCR are available.
qPCR is the only tool capable of monitoring responses after the patient has experienced CCyR, because BCR-ABL1 transcripts typically remain detectable after CCyR is achieved. A major advantage of qPCR is the strong correlation between the results obtained from the peripheral blood and the bone marrow, allowing molecular monitoring without bone marrow aspirations.69,70
Standardization of Molecular Monitoring Using the IS: In the IS, the standardized baseline (defined as the average expression of BCR-ABL1 transcripts in 30 patients treated on the IRIS trial) is set to 100%. Molecular response is expressed as log-reduction from 100%. For example, ≥3-log reduction (≤0.1% BCR-ABL1 IS) is referred to as MMR or MR3.0).30,71,72 A 2-log reduction generally correlates with CCyR (≤1% BCR-ABL1 IS).
The sensitivity of a qPCR assay depends not only on the performance of the assay, but also on the quality of a given sample. As such the term “complete molecular response” to denote undetectable BCR-ABL1 transcripts (a negative qPCR test) should be abandoned, because it may refer to very different levels of response, dependent on the quality of the sample. Laboratories can use their individual assays, but the BCR-ABL1 transcripts obtained in a given laboratory should be converted to the IS by applying a laboratory-specific conversion factor.30,73
Recommendations for Monitoring Response to TKI Therapy: qPCR (IS) is the preferred method to monitor response to TKI therapy. qPCR assays with a sensitivity of ≥4.5-log reduction from the standardized baseline are recommended for the measurement of BCR-ABL1 transcripts. In patients with prolonged myelosuppression who may not be in complete hematologic response due to persistent cytopenias or unexplained drop in blood counts during therapy, bone marrow cytogenetics is indicated to confirm response to TKI therapy and exclude other pathology, such as MDS or the presence of chromosomal abnormalities other than Ph.
Monitoring with qPCR (IS) every 3 months is recommended for all patients after initiating TKI therapy, including those who meet response milestones at 3, 6, and 12 months (≤10% BCR-ABL1 IS at 3 and 6 months, ≤1% BCR-ABL1 IS at 12 months, and ≤0.1% BCR-ABL1 IS at >12 months). After CCyR (≤1% BCR-ABL1 IS) has been achieved, molecular monitoring is recommended every 3 months for 2 years and every 3 to 6 months thereafter (see CML-C; page 1114).
Frequent molecular monitoring with qPCR (IS) can help to identify nonadherence to TKI therapy early in the treatment course.74 Because adherence to TKI therapy is associated with better clinical outcomes, frequent molecular monitoring is essential if there are concerns about the patient's adherence to TKI therapy after CCyR has been achieved. In patients with deeper molecular responses (MMR and better) and who are adherent with TKI therapy, the frequency of molecular monitoring can be reduced, though the optimal frequency is unknown.
Prognostic Significance of Cytogenetic and Molecular Response
Early molecular response (≤10% BCR-ABL1 IS at 3 and 6 months) after first-line TKI therapy has emerged as an effective prognosticator of favorable long-term PFS and OS, regardless of TKI used (Table 5).40,41,45,75 Some reports suggest that early molecular response at 3 months has a superior prognostic value and support the use of early intervention strategies based on the BCR-ABL1 transcript level at 3 months.76,77 However other studies yielded partially conflicting results regarding the predictive value of BCR-ABL1 transcript levels at 3-months.78 From a practical perspective, it is important to consider these data points within the clinical context. For instance, if BCR-ABL1 transcript level is minimally above the 10% cutoff (11% at 3 months), it is reasonable to reassess at 6 months before considering major changes to the treatment strategy.
Recently, studies have suggested that the rate of decline in BCR-ABL1 transcripts correlates with longer-term response.79–82 Among patients with >10% BCR-ABL1 IS after 3 months of treatment with imatinib, those with a faster decline in BCR-ABL1 (BCR-ABL1 halving time <76 days) had a superior outcome compared with those with a slower decline (4-year PFS rate was 92% vs 63%, respectively).79 A rapid initial BCR-ABL1 decline also identifies a subgroup of Sokal high-risk patients with outcomes similar to those of Sokal low-risk patients.80 Among Sokal high-risk patients, a BCR-ABL1 halving time of ≤11 days was associated with significantly improved FFS (4-year FFS rate was 79% for patients with halving time of ≤11 days vs 53% for those with halving time of >11 days; P=.03). In the German CML IV study, lack of a half-log reduction of BCR-ABL1 transcripts at 3 months was associated with a higher risk of disease progression on imatinib therapy.81 The results of the D-First study also showed that in patients treated with dasatinib, BCR-ABL1 halving time of ≤14 days was a significant predictor of MMR by 12 months and deep molecular response (BCR-ABL1 <0.01% IS) by 18 months.82
Achievement of CCyR (≤1% BCR-ABL1 IS) within 12 months after first-line TKI therapy is an established prognostic indicator of long-term survival.83,84 In the IRIS study, the estimated 6-year PFS rate was 97% for patients achieving a CCyR at 6 months compared with 80% for patients with no cytogenetic response at 6 months.83 In an analysis of patients with newly diagnosed CP-CML treated with imatinib or second-generation TKIs, the 3-year event-free survival and OS rates were 98% and 99% for patients who experienced CCyR at 12 months compared with 67% and 94% in patients who did not experience a CCyR.84
The prognostic significance of MMR (0.1% BCR-ABL1 IS) after first-line imatinib has also been evaluated in several studies.69,85–89 In all of these studies, the analyses were done for different outcomes measures at multiple time points, but failed to adjust for multiple comparisons, thereby reducing the validity of the conclusions. The synoptic conclusion from these studies is that MMR is moderately superior to CCyR in predicting long-term PFS and OS. However, with longer
Early Molecular Response (≤10% BCR-ABL1 IS at 3 mo) After First-Line TKI Therapy and Survival Outcomes
Response Milestones after First-Line TKI Therapy
The goal of TKI therapy is to achieve a CCyR (≤1% BCR-ABL1 IS) within 12 months after first-line TKI therapy and to prevent disease progression to AP-CML or BP-CML. The guidelines emphasize that achievement of response milestones must be interpreted within the clinical context, before making drastic changes to the treatment strategy.
The panel has included ≤10% BCR-ABL1 IS at 3 and 6 months and ≤1% BCR-ABL1 IS at 12 and 15 months as response milestones after first-line TKI therapy (see CML-3; page 1112). Patients who experience these response milestones are considered to have TKI-sensitive disease, and continuation of the same dose of TKI and assessment of BCR-ABL1 transcripts with qPCR (IS) every 3 months is recommended for this group of patients.
In patients with a >10% BCR-ABL1 IS at 3 months and >1% BCR-ABL1 IS at 12 months, clinical judgement should be used, considering problems with adherence (which can be common given drug toxicity at start of therapy), rate of decline in BCR-ABL1 (the faster, the better), and how far from the 10% cutoff the BCR-ABL1 value falls. That being said, failure to experience ≤10% BCR-ABL1 IS at 3 months or ≤1% BCR-ABL1 IS at 12 months is associated with a higher risk for disease progression.
Patients with >10% BCR-ABL1 IS at 3 months or >1% BCR-ABL1 IS at 12 months can continue the same dose of dasatinib or nilotinib or bosutinib for another 3 months. Mutational analysis and evaluation for allogeneic HCT should considered. Bone marrow cytogenetics should be considered to assess for MCyR at 3 months or CCyR at 12 months.
Patients with >10% BCR-ABL1 IS at ≥6 months and those with BCR-ABL1 IS >1% at 15 months are considered to have TKI-resistant disease. Evaluation for allogeneic HCT (that is, a discussion with a transplant specialist, which might include HLA testing) is recommended. Alternate treatment options should be considered as described subsequently.
Early molecular response (≤10% BCR-ABL1 IS at 3 and 6 months) after second-line TKI therapy with dasatinib or nilotinib has also been reported to be a prognosticator of OS and PFS (Table 7). Patients who do not experience cytogenetic or molecular responses at 3, 6, or 12 months after second-line and subsequent TKI therapy should be considered for alternative therapies or allogeneic HCT if deemed eligible.
Second-line and Subsequent TKI Therapy for CP-CML: Long-Term Follow-Up Data From Phase II/III Studies
Management of Patients With Inadequate Response to Imatinib: Switching to an alternate TKI is recommended for patients with disease that is resistant to imatinib 400 mg daily. Dasatinib, nilotinib, and bosutinib are active against many of the imatinib-resistant BCR-ABL1 kinase domain mutants, except T315I, and are effective treatment options for patients with CP-CML intolerant to imatinib or those with CP-CML resistant to imatinib.93-95
Dose escalation of imatinib up to 800 mg daily has been shown to overcome some of the primary resistance and is particularly effective in patients with cytogenetic relapse who had achieved cytogenetic response with imatinib, 400 mg daily, although the duration of responses has typically been short.97–100 However, it is unlikely to benefit patients with hematologic failure or those who never had a cytogenetic response with imatinib 400 mg daily. Switching to nilotinib has been shown to result in higher rates of cytogenetic and molecular response than dose escalation of imatinib in patients with inadequate response to imatinib, 400 mg.101,102 In the TIDEL-II study, the cohort of patients with >10% BCR-ABL1 IS at 3 months after imatinib, 400 mg, who were switched directly to nilotinib had higher rates of MMR and CMR at 12 months (but not at 24 months) than the cohort of patients who received dose escalation of imatinib before switching to nilotinib.101 Although dose escalation of imatinib has been shown to be beneficial for patients in CCyR with no MMR, no randomized studies have shown that a change of therapy would improve PFS or event-free survival in this group of patients.103,104
Management of Patients with Inadequate Response to Dasatinib, Nilotinib or Bosutinib: Switching to an alternate TKI (other than imatinib) in the second-line setting could be considered for patients with disease that is resistant to dasatinib, nilotinib, or bosutinib. However, no clear evidence supports that switching to alternate TKI therapy would improve long-term clinical outcome for this group of patients.
Ponatinib is an option for patients with T315I mutation and for those with disease that has not responded to several TKIs.96 Long-term efficacy data from phase II/III studies evaluating bosutinib or ponatinib in patients with pretreated CP-CML are summarized in Table 6.
In the PACE trial, serious arterial occlusive events (cardiovascular, cerebrovascular, and peripheral vascular) and venous thromboembolic events occurred in 31% and 6% of patients, respectively.96 Cardiovascular occlusion, cerebrovascular occlusion, and peripheral arterial occlusive events were reported in 16%, 13%, and 14% of patients, respectively. Ponatinib labeling contains a black box warning regarding vascular occlusion, heart failure, and hepatotoxicity. Cardiovascular risk factors (eg, diabetes mellitus, hypertension, hyperlipidemia, smoking, estrogen use) should be identified and controlled before starting ponatinib. Patients should be monitored
Early Molecular Response (≤10% BCR-ABL1 IS) After Second-Line TKI Therapy and Survival Outcomes
The recommended initial dose of ponatinib is 45 mg once daily. High dose intensity of ponatinib is significantly associated with increased risk of adverse events.105 Therefore, dose modifications may be necessary for the management of adverse events. In a post hoc analysis of the PACE trial that assessed the clinical impact of dose modification and dose intensity on outcomes of patients with CP-CML, substantial responses were seen at lower dose levels and the rates of maintenance of MCyR and MMR were high irrespective of dose-reductions.96 Thus, an initial dose of 30 mg may be a safer and effective dose for patients with cardiovascular risk factors. Safety and efficacy of ponatinib at initial doses lower than 45 mg are being evaluated in a randomized clinical trial.
Omacetaxine is an option for patients with the T315I mutation and in those with CML that is resistant to ≥2 TKIs.106–108 In the CML 202 study, among 62 evaluable patients with T315I and CP-CML resistant to prior TKI therapy, MCyR, CCyR, and MMR were achieved in 23%, 16%, and 17% of patients, respectively, and the T315I clone declined to below detection limits in 61% of patients.106 After a median follow-up of 19 months, the median PFS was 8 months and the median OS had not yet been reached. In the cohort of 46 patients with CP-CML that is resistant to ≥2 TKIs (CML 203 study), MCyR and CCyR were achieved in 22% and 4% of patients, respectively. Median PFS and OS were 7 months and 30 months, respectively.107 Omacetaxine had an acceptable toxicity profile, and the most common grade 3/4 adverse events were thrombocytopenia (67%), neutropenia (47%), and anemia (37%).108
Clinical Considerations For The Selection Of Second-Line Therapy
BCR-ABL kinase domain mutation analysis (see subsequent section), evaluation of drug interactions, and compliance to therapy are recommended before the start of second-line TKI therapy.
Drug Interactions: Bosutinib, dasatinib, imatinib, and nilotinib are metabolized in the liver by cytochrome P450 (CYP) enzymes. Drugs that are CYP3A4 or CYP3A5 inducers may decrease the therapeutic plasma concentration of TKIs, whereas CYP3A4 inhibitors and drugs that are metabolized by the CYP3A4 or CYP3A5 enzyme might result in increased plasma levels of TKIs.109 In addition, imatinib is also a weak inhibitor of the CYP2D6 and CYP2C9 isoenzymes and nilotinib is a competitive inhibitor of CYP2C8, CYP2C9, CYP2D6, and UG-T1A1, potentially increasing the plasma concentrations of drugs eliminated by these enzymes.
Concomitant use of drugs that are metabolized by these enzymes requires caution, and appropriate alternatives should be explored to optimize treatment outcome. If coadministration cannot be avoided, dose modification should be considered.
Concomitant use of H2 blockers or proton pump inhibitors (PPIs) is not recommended in patients receiving dasatinib. If their use is inevitable, they should be administered 12 hours before the next dasatinib dose. Concomitant use of PPI is not recommended in patients receiving bosutinib. The use of short-acting antacids or H2 blockers should be considered instead of PPIs.
Adherence to Therapy: Treatment interruptions and nonadherence to therapy may lead to undesirable clinical outcomes.110–112 In the ADAGIO study, non-adherence to imatinib was associated with poorer response. Patients with suboptimal response missed significantly more imatinib doses (23%) than did those with optimal response (7%).110 Adherence to imatinib therapy has been identified as the only independent predictor for achieving complete molecular response (CMR) on standard-dose imatinib.111 Poor adherence to imatinib therapy has also been identified as the most important factor contributing to cytogenetic relapse and imatinib failure.112 Patients with adherence of ≤85% had a higher probability of losing CCyR at 2 years than those with adherence of >85% (27% and 2%, respectively). Poor adherence to therapy has also been reported in patients receiving dasatinib and nilotinib after imatinib failure.113,114
Patient education on adherence to therapy and close monitoring of patient's adherence is critical to achieving optimal responses. In a significant proportion of patients with TKI-induced toxicities, responses have been observed with doses well below their determined maximum tolerated doses.115 Short interruptions or dose reductions, when medically necessary, may not have a negative impact on disease control or other outcomes. Adequate and appropriate management of side effects and scheduling appropriate follow-up visits to review side effects may be helpful to improve patient adherence to therapy.116 Switching to an alternate TKI because of intolerance might be beneficial for selected patients with acute grade 3/4 nonhematologic toxicities or in those with low-grade, chronic, and persistent adverse events that are not manageable with adequate supportive care measures.117
Resistance to TKI Therapy: Aberrant expressions of drug transporters118–120 and plasma protein binding of TKI121–123 could contribute to primary resistance by altering the intracellular and plasma concentration of TKI. Monitoring imatinib plasma levels may be useful in determining patient adherence to therapy. However, there are no data to support that change of therapy based on plasma imatinib levels will affect treatment outcomes. Pretreatment levels of organic cation transporter 1 (OCT1) have been reported as the most powerful predictor of response to imatinib.124 Conversely, cellular uptake of dasatinib or nilotinib seems to be independent of OCT1 expression, suggesting that patients with low hOCT1 expression might have better outcomes with dasatinib or nilotinib than with imatinib.125–128
BCR-ABL Kinase Domain Mutation Analysis: Point mutations in the BCR-ABL1 kinase domain are a frequent mechanism of secondary resistance to TKI therapy and are associated with poor prognosis and higher risk of disease progression.129–134 Among the BCR-ABL1 kinase domain mutations, the T315I mutation confers the complete resistance to imatinib, dasatinib, nilotinib, and bosutinib.135,136
F317L and V299L mutants are resistant to dasatinib and Y253H, E255K/V, and F359V/C mutants are resistant to nilotinib.137–140 E255K/V, F359C/V, Y253H, and T315I mutants are most commonly associated with disease progression and relapse.140 Bosutinib has demonstrated activity in patients with BCR-ABL1 mutants resistant to dasatinib (F317L) and nilotinib (Y253H, E255K/V, and F359C/I/V).95 T315I, G250E, and V299L mutants are resistant to bosutinib. Ponatinib is active against other BCR-ABL1 mutants resistant to dasatinib or nilotinib, including E255V, Y253H, and F359V, in addition to T315I.96,141 Response rates based on BCR-ABL mutation status are listed in Table 8.
BCR-ABL kinase domain mutational analysis is helpful in the selection of subsequent TKI therapy for patients with inadequate initial response to first- or second-line TKI therapy.142 Treatment options based on BCR-ABL1 mutation status are outlined on CML-5 (page 1113). BCR-ABL mutational analysis provides additional guidance in the selection of subsequent TKI therapy only in patients with identifiable mutations. In patients with no identifiable mutations, the selection of subsequent TKI therapy should be based on the toxicity profile of TKI, patient's age, ability to tolerate therapy, and the presence of comorbid conditions. Adverse events of second-line TKI therapy in patients with CP-CML are summarized in Table 9.
The use of an alternate second-generation TKI after treatment failure with 2 prior TKIs, including a second-generation TKI, is not associated with durable responses, except in occasional patients with CP-CML.143 The guidelines recommend BCR-ABL1 mutational analysis for patients who do not experience response milestones, for those with any sign of loss of response (hematologic or cytogenetic relapse), and if there is a 1-log increase in BCR-ABL1 level with loss of MMR.
Rising BCR-ABL1 Transcript Levels: Rising BCR-ABL1 transcript levels are associated with an increased
Responses Based on BCR-ABL1 Mutations Status
The precise increase in BCR-ABL1 transcripts that warrants a mutation analysis depends on the performance characteristics of the qPCR assay.148 Some laboratories have advocated a 2- to 3-fold range,88,147,148 whereas others have taken a more conservative approach (5- to 10-fold).146 Obviously, some common sense must prevail, because the amount of change in absolute terms depends on the level of molecular response. For example, a finding of any BCR-ABL1 after achieving a deep molecular response (MR4.5; ≤0.0032% BCR-ABL1 IS) is an infinite increase in BCR-ABL1 transcripts. However,
Adverse Events of Second-Line and Subsequent TKI Therapy in CP-CML
Currently there are no specific guidelines for changing therapy based on rising BCR-ABL1 levels as detected by qPCR. Changes of therapy based solely on rising BCR-ABL1 levels should be done only in the context of a clinical trial.
Discontinuation of TKI Therapy
The feasibility of discontinuation of TKI therapy (with close monitoring) in carefully selected patients who have experienced and maintained deep molecular response (≥MR4.0; ≤0.01% BCR-ABL1 IS) for ≥2 or more years has been evaluated in several clinical studies. Limited longer-term follow-up data from the TKI discontinuation trials are summarized in Table 10.
The possibility of treatment-free remission (TFR) after discontinuation of imatinib was first evaluated in the Stop Imatinib (STIM1) study in 100 patients with a CMR for at least 2 years (5-log
Summary of Limited Longer-Term Follow-Up Data From the TKI Discontinuation Trials
More recent studies have also confirmed the feasibility of TFR after discontinuation of dasatinib or nilotinib, in patients with CP-CML who have achieved and maintained MR4.5 for 12 months after ≥2 years of TKI therapy in the first-line or second-line setting (TFR rates ranging from 44% to 54%; Table 10).156–160 The feasibility of TFR after discontinuation of bosutinib or ponatinib has not yet been evaluated in clinical studies. In the EURO-SKI study that evaluated TFR after discontinuation of any first-line TKI therapy (imatinib, dasatinib, or nilotinib) in eligible patients, the type of first-line TKI therapy did not significantly affect molecular relapse-free survival.156 Therefore, it is reasonable to assume that the likelihood of TFR after discontinuation would be similar irrespective of TKI in patients who have experienced and maintained deep molecular response (MR4.0; ≤0.01% BCR-ABL1 IS) for ≥2 years.
The results of the RE-STIM study demonstrated the safety of a second TKI discontinuation after a first unsuccessful attempt.161 The rate of molecular relapse after the first TKI discontinuation attempt was the only factor significantly associated with outcome. The TFR rate at 24 months after second TKI discontinuation was higher for patients who remained in deep molecular response within the first 3 months after the first TKI discontinuation (72% vs 32% for other patients).
Approximately 40% to 60% of patients who discontinue TKI therapy after achieving deep molecular response experience recurrence within 6 months of treatment cessation, in some cases as early as 1 month after discontinuation of TKI therapy. Resumption of TKI therapy immediately after recurrence results in the achievement of undetectable disease in almost all patients.149–160 TKI withdrawal syndrome (aggravation or new development of musculoskeletal pain and/or pruritus after discontinuation of TKI therapy) has been reported during the TFR period in some TKI discontinuation studies,155,158,159 and the occurrence of imatinib withdrawal syndrome was associated with a lower rate of molecular relapse in the KID study.155
In the STIM study, molecular relapse (trigger to resume TKI therapy) was defined as positivity for BCR-ABL1 transcripts by qPCR confirmed by a 1-log increase in BCR-ABL1 transcripts between 2 successive assessments or loss of MMR at one point.149,150 The results of the A-STIM study showed that loss of MMR (≤0.1% BCR-ABL1 IS) could be used as a practical criterion for restarting therapy. The estimated probability of MMR loss was 35% at 12 months and 36% at 24 months after discontinuation of imatinib.153 Several factors may help predict the risk of relapse after discontinuation of TKI therapy (eg, a higher Sokal risk score, female sex, lower natural killer cell counts, suboptimal response or resistance to imatinib, duration of TKI therapy, and deep molecular response prior to TKI discontinuation).149,150,156–160,162 However, only the duration of TKI therapy and deep molecular response before TKI discontinuation therapy have been associated with TFR with a high level of consistency.149,156 In the EURO-SKI study, duration of treatment with imatinib (≥ 6 years) and deep molecular response duration (MR4.0 for 3 years) were significantly associated with MMR maintenance at 6 months after discontinuation of imatinib.156
Based on the available evidence from clinical studies that have evaluated the feasibility of TFR, the panel members feel that discontinuation of TKI therapy (with close monitoring) is feasible in carefully selected patients (in early CP-CML) who have achieved and maintained a deep molecular response (≥MR4.0) for ≥2 years. Clinical studies that have evaluated the safety and efficacy of discontinuation of TKI have employed strict eligibility criteria and have mandated more frequent molecular monitoring than typically recommended for patients on TKI therapy. Access to a reliable qPCR (IS) with a sensitivity of detection of at least MR4.5 (BCR-ABL1 ≤ 0.0032% IS) and the availability of test results within 2 weeks is one of the key requirements to monitor patients after TKI discontinuation and ascertain their safety.
The criteria for the selection of patients suitable for discontinuation of TKI therapy are outlined in CML-E (page 1116). The guidelines emphasize that discontinuation of TKI therapy outside of a clinical trial should be considered only if all the criteria included in the list are met. The panel acknowledges that more frequent molecular monitoring is essential after discontinuation of TKI therapy for the early identification of loss of MMR. Frequency of molecular monitoring has varied substantially among different studies, and the optimal frequency of molecular monitoring in patients with a loss of MMR after discontinuation of TKI therapy has not been established. The panel recommendations for molecular monitoring in TFR phase are outlined in CML-E (page 1116).
Individual Disclosures for Chronic Myeloid Leukemia
MedinaJKantarjianHTalpazM. Chromosomal abnormalities in Philadelphia chromosome-negative metaphases appearing during imatinib mesylate therapy in patients with Philadelphia chromosome-positive chronic myelogenous leukemia in chronic phase. Cancer2003;98:1905–1911.
HughesTDeiningerMHochhausA. 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. Blood2006;108:28–37.
BaccaraniMDrukerBJBranfordS. 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 Hematol2014;99:616–624.
DeiningerMWKopeckyKJRadichJP. Imatinib 800 mg daily induces deeper molecular responses than imatinib 400 mg daily: results of SWOG S0325, an intergroup randomized PHASE II trial in newly diagnosed chronic phase chronic myeloid leukaemia. Br J Haematol2014;164:223–232.
MuhlmannJThalerJHilbeW. Fluorescence in situ hybridization (FISH) on peripheral blood smears for monitoring Philadelphia chromosome-positive chronic myeloid leukemia (CML) during interferon treatment: a new strategy for remission assessment. Genes Chromosomes Cancer1998;21:90–100.
TestoniNMarzocchiGLuattiS. Chronic myeloid leukemia: a prospective comparison of interphase fluorescence in situ hybridization and chromosome banding analysis for the definition of complete cytogenetic response: a study of the GIMEMA CML WP. Blood2009;114:4939–4943.
AkardLPCortesJEAlbitarM. Correlations between cytogenetic and molecular monitoring among patients with newly diagnosed chronic myeloid leukemia in chronic phase: post hoc analyses of the rationale and insight for Gleevec high-dose therapy study. Arch Pathol Lab Med2014;138:1186–1192.
BranfordSFletcherLCrossNC. Desirable performance characteristics for BCR-ABL measurement on an international reporting scale to allow consistent interpretation of individual patient response and comparison of response rates between clinical trials. Blood2008;112:3330–3338.
IriyamaNFujisawaSYoshidaC. Shorter halving time of BCR-ABL1 transcripts is a novel predictor for achievement of molecular responses in newly diagnosed chronic-phase chronic myeloid leukemia treated with dasatinib: results of the D-first study of Kanto CML study group. Am J Hematol2015;90:282–287.
BranfordSRudzkiZWalshS. Detection of BCR-ABL mutations in patients with CML treated with imatinib is virtually always accompanied by clinical resistance, and mutations in the ATP phosphate-binding loop (P-loop) are associated with a poor prognosis. Blood2003;102:276–283.
SoveriniSMartinelliGRostiG. ABL mutations in late chronic phase chronic myeloid leukemia patients with up-front cytogenetic resistance to imatinib are associated with a greater likelihood of progression to blast crisis and shorter survival: a study by the GIMEMA Working Party on Chronic Myeloid Leukemia. J Clin Oncol2005;23:4100–4109.
SoveriniSGnaniAColarossiS. Philadelphia-positive patients who already harbor imatinib-resistant Bcr-Abl kinase domain mutations have a higher likelihood of developing additional mutations associated with resistance to second- or third-line tyrosine kinase inhibitors. Blood2009;114:2168–2171.
ThielenNvan der HoltBCornelissenJJ. Imatinib discontinuation in chronic phase myeloid leukaemia patients in sustained complete molecular response: a randomised trial of the Dutch-Belgian Cooperative Trial for Haemato-Oncology (HOVON). Eur J Cancer2013;49:3242–3246.
RousselotPCharbonnierACony-MakhoulP. 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 Oncol2014;32:424–430.