Chronic myeloid leukemia approvals have rested on surrogate response endpoints or progression-free survival, not on overall survival. Which surrogate a regulator will accept depends on disease phase, line of therapy and the maturity of follow-up. For regulatory and clinical teams planning a CML program, the endpoint and trial design accepted for earlier tyrosine kinase inhibitors shape sample size, comparator choice, response timepoints and whether an accelerated pathway is realistic.
The analysis below traces the primary endpoints and pivotal trial designs behind FDA and EMA approvals of CML therapies across newly diagnosed chronic phase, resistant or intolerant chronic phase, and advanced-phase disease. It covers hematologic, cytogenetic and molecular response definitions and the timepoints used. It also compares randomized active-controlled trials with single-arm designs and shows where the two agencies' assessments diverged.
CML approval endpoints: what FDA and EMA have accepted for MMR, CCyR, MCyR and hematologic response
Regulators have accepted chronic myeloid leukemia (CML) approvals on response endpoints or progression-free survival, not on overall survival. Which response they accept depends on three things: disease phase, line of therapy, and how much follow-up is available. Over the past two decades the accepted primary endpoint has moved from hematologic and cytogenetic response toward molecular response. The tyrosine kinase inhibitor (TKI) approvals below show each step of that shift. FDA has usually granted accelerated approval on early response data and converted it to regular approval once longer follow-up confirmed durability. EMA's CHMP has generally accepted the same endpoints, with its own views on timepoints and comparators.
Key takeaways
- Newly diagnosed chronic phase (CP): the primary endpoint moved from progression-free survival or time to progression (imatinib, December 2002) to major molecular response (MMR) at 12 months (nilotinib, June 2010) or confirmed complete cytogenetic response (CCyR) within 12 months (dasatinib, October 2010), then to MMR at 12 months (bosutinib) and MMR at 48 weeks (asciminib) 783128273627598. Nilotinib's MMR-based first-line approval preceded dasatinib's CCyR-based approval. Every first-line approval in this group used a randomized, open-label, active-controlled trial.
- Resistant or intolerant CP: major cytogenetic response (MCyR) from single-arm trials supported most approvals (imatinib after interferon, dasatinib, nilotinib, bosutinib, ponatinib, omacetaxine) 93278133631384313. Asciminib moved this setting to a randomized trial with MMR at 24 weeks as the primary endpoint 346603.
- Accelerated and blast phase: major hematologic response (MaHR, defined as complete hematologic response plus no evidence of leukemia) has been the standard primary endpoint 278388319. FDA removed "return to chronic phase" from the composite when it reviewed nilotinib 506507.
- Current guidance: FDA guidance describes durable MMR as a surrogate endpoint used for traditional approval in CML 686. EMA's 2016 haematological malignancies appendix recommends MMR at 18 months as the primary endpoint in superiority trials 760761.
Regulatory framework: FDA and EMA guidance
FDA
The 2020 final guidance Hematologic Malignancies: Regulatory Considerations for Use of Minimal Residual Disease in Development of Drug and Biological Products for Treatment 739 defines MMR as BCR-ABL(IS) below 0.1%. Where an RT-qPCR(IS) assay is not available, it defines MMR as a reduction of more than 3 logs in BCR-ABL1 mRNA from the standardized baseline 684. The guidance says that achieving MMR predicts better long-term PFS/EFS and that "durable MMR can be a measure of clinical benefit" 685. It calls for International Scale assays with the standardized baseline set at 100%, names RT-qPCR(IS) as the preferred method 791, and notes that MRD can be used to select and monitor patients for TKI discontinuation 792.
The FDA guidance Clinical Trial Endpoints for the Approval of Cancer Drugs and Biologics states that "durable major molecular response is a surrogate endpoint used for traditional approval for therapies in chronic myelogenous leukemia" 686.
Product reviews show how FDA applies this in practice. In the asciminib first-line review, the Division described MMR with at least 12 months of follow-up as accepted for accelerated approval in newly diagnosed CP-CML, and MMR with at least five years of follow-up as a basis for regular approval 484.
EMA
EMA's recommendations have changed across versions of the anticancer guideline:
| Guideline version | Recommended primary endpoint, CP-CML | Notes |
|---|---|---|
| Draft, 19 November 2008 | PFS preferred as the primary endpoint for non-inferiority trials | Draft only 766767 |
| Appendix 2 (confirmatory studies in haematological malignancies), 3 March 2010 | CCyR at one year for first-line superiority trials; at least 5 years of follow-up | Molecular response accepted as a secondary endpoint only 765 |
| Appendix 4, Revision 2, 1 July 2016 | MMR at 18 months for superiority trials against a licensed comparator; CCyR at 12 months, PFS and OS as secondary endpoints | Other timepoints acceptable if justified; at least 8 years of follow-up expected; non-inferiority trials need MCyR assessed after at least two years 760761 |
Newly diagnosed chronic-phase CML
| Product / trial | Design and comparator | Primary endpoint | Result | FDA pathway | EMA/CHMP position |
|---|---|---|---|---|---|
| Imatinib, Study 0106 | Randomized, open-label phase 3 vs interferon-alfa plus cytarabine, 553 per arm, crossover allowed 77978078178282 | PFS, with progression including loss of CHR or MCyR; MCyR was the main endpoint for the first interim analysis 78378 | 30-month PFS 87.8% vs 68.3% (p<0.0001) 86 | Accelerated approval December 20, 2002; commitments fulfilled with long-term Study 0106 data in 2009 944947948 | CPMP: CCyR not yet validated as a marker of clinical benefit, although an association could reasonably be assumed 546 |
| Dasatinib, DASISION | Randomized, open-label vs imatinib 400 mg, n=519 273 | Confirmed CCyR (two assessments at least 28 days apart) within 12 months 273 | 76.8% vs 66.2% (P=0.007) 274 | Accelerated approval October 28, 2010, with a 60-month follow-up requirement; regular approval August 12, 2015 908909910913 | CHMP endorsed cCCyR at 12 months, considered CCyR validated, and said MMR could also have been acceptable 455456 |
| Nilotinib, ENESTnd (CAMN107A2303) | Randomized 1:1:1, open-label phase 3 vs imatinib 55126 | MMR (BCR-ABL ≤0.1% IS) at 12 months 12856 | 44% vs 22% (P<0.0001) 129 | Accelerated approval June 17, 2010; Subpart H required a 24-month interim report and at least 60 months of follow-up, fulfilled in 2015 725726735736 | CHMP accepted 12-month MMR as a surrogate primary endpoint when supported by secondary endpoints 6364 |
| Bosutinib, BELA (3160A4-3000-WW) | Randomized, open-label phase 3 vs imatinib 251252 | CCyR at one year 251 | Primary objective failed 253575 | Not approved on this trial; FDA did not allow secondary endpoints to be treated as primary after the primary failure 575576 | CHMP: one-year CCyR acceptable as a primary objective for a superiority trial, not a non-inferiority trial 254 |
| Bosutinib, BFORE (AV001) | Randomized 1:1, open-label vs imatinib 626237238 | MMR at 12 months 627239 | 47% vs 37% (p=0.0200) 627628 | Accelerated approval December 19, 2017, with a 60-month PMR; regular approval May 14, 2021 866867869870871872 | CHMP accepted 12-month MMR as a surrogate and noted that updated guidance recommends superiority on MMR at 18 months 242243 |
| Asciminib, ASC4FIRST | Randomized 1:1, open-label vs investigator-selected TKI (imatinib, nilotinib, dasatinib or bosutinib) 597352353 | MMR at week 48, overall and in the imatinib stratum 598354 | 68% vs 49% overall; 69% vs 40% in the imatinib stratum 598599 | Accelerated approval October 29, 2024; regular approval requires five years of efficacy follow-up 970971822 | One pivotal study for the first-line extension; MMR at week 96 is the key secondary endpoint 352356 |
Two design points come out of the first-line reviews:
- Open-label design is acceptable. For both dasatinib and nilotinib, CHMP accepted open-label designs because most of the endpoints are objectively measured 45545765.
- Missing-data rules matter. In ENESTnd, a missing month-12 assessment counted as non-response unless MMR was present at both month 9 and month 15 56. The key secondary endpoint was durable MMR, defined as MMR at both months 12 and 24 with no loss in between 57.
The imatinib history shows how far expectations have moved. In early meetings for the first-line program, FDA said six-month CHR was not an adequate primary endpoint. It said superiority in MCyR at 24 months might support accelerated approval, and the parties ultimately agreed on superiority in time to progression 177178179.
Resistant or intolerant chronic-phase CML
| Product / trial | Design | Primary endpoint | Key result | FDA pathway and endpoint view |
|---|---|---|---|---|
| Imatinib, Study 0110 (after interferon) | Single-arm phase 2, n=532 92174 | MCyR (0 to 35% Ph+ metaphases); unconfirmed rate was the principal outcome 528531 | 49.4% unconfirmed MCyR in the original analysis 531 | Accelerated approval May 10, 2001. FDA considered MCyR in any setting, and CHR, reasonably likely to predict clinical benefit 940166170 |
| Dasatinib, CA180013 / CA180017 / CA180034 | Single-arm; randomized 2:1 vs high-dose imatinib (MCyR at 12 weeks); randomized 2x2 dose/schedule study 416283285 | MCyR 278283286 | CP MCyR 45% in the single-arm studies; once-daily vs twice-daily difference 1.9% (95% CI −6.8% to 10.6%) 278286 | Accelerated approval in 2006 with 24-month follow-up commitments; FDA described durable MCyR as a basis for accelerated approval 899900648649 |
| Nilotinib, A2101 | Single-arm, open-label; CP and AP cohorts 124132 | MCyR in CP 133503 | 51% (95% CI 46% to 57%) 133 | Accelerated approval October 29, 2007, with 24-month follow-up required. FDA accepted bone-marrow cytogenetics, not FISH, as the validated measure 724732505 |
| Bosutinib, Study 200 | Single-arm, open-label 630 | MCyR by week 24 631570 | 40.1% (imatinib only); 25.9% (imatinib plus at least one more TKI) 632631 | Regular approval recommended. The team leader noted that FDA had previously accepted MCyR with 24 months of follow-up as evidence of clinical benefit in second-line CML 570572 |
| Ponatinib, PACE | Single-arm, open-label, cohorts by phase and T315I status 382383 | MCyR by 12 months (CP) 384 | 55% overall; 70% in the T315I cohort 385386 | Accelerated approval December 14, 2012, with a 24-month PMR; regular approval November 28, 2016 884885886887888 |
| Omacetaxine | Combined analysis of two open-label, single-arm trials after at least two TKIs 311704 | MCyR (CP); MaHR and/or MCyR (AP) 312702 | CP MCyR 18.4% 313 | Accelerated approval; the reviewer cited the lack of the 24 months of follow-up FDA expected for regular approval 706 |
| Asciminib, ASCEMBL | Randomized 2:1, open-label phase 3 vs bosutinib, after at least two TKIs 344345602 | MMR at week 24 346 | 25.5% vs 13.2% (p=0.029) 338 | Accelerated approval October 29, 2021; converted on October 12, 2022 after 96-week follow-up 965966967480 |
ASCEMBL marks the change from MCyR to MMR in previously treated disease. During design review, FDA said MMR at 24 weeks "may be acceptable" as an endpoint reasonably likely to predict clinical benefit. It recommended MMR at two years as the first key secondary endpoint to verify that benefit 474475. CHMP accepted MMR as a CML surrogate in principle but said week-24 MMR needed longer-term data, such as week-96 MMR, to support it 350. CHMP also raised concerns about bosutinib discontinuations and how robust the result was under different missing-data assumptions 351.
T315I-mutated disease
FDA recommended regular approval for asciminib in T315I-mutated CP-CML. The evidence came from a non-randomized, open-label cohort with 96-week MMR data 481482. That study's primary endpoint was dose-limiting toxicity, and MMR was a secondary endpoint 483. MMR by 24 weeks was 42% (19/45) and by 96 weeks 49% 608. CHMP called this an uncontrolled exploratory study with substantial uncertainties. It said a randomized comparison with ponatinib would have been more appropriate 358361.
Accelerated- and blast-phase CML
In advanced-phase disease the primary endpoint has stayed hematologic:
- Imatinib: confirmed hematologic response (CHR, NEL or return to chronic phase) in single-arm phase 2 studies; FDA granted accelerated approval on hematologic-response surrogates 9899170171.
- Dasatinib: MaHR (CHR plus NEL, confirmed after four weeks). FDA called MaHR "reasonably likely to predict clinical benefit" and considered overall hematologic response less clinically significant 279653654.
- Nilotinib: the protocol used overall hematologic response. FDA accepted CHR plus NEL but not return to chronic phase, and recommended assessing the endpoint without that component 506507.
- Ponatinib (PACE): MaHR by 6 months, 57% in AP-CML and 31% in BP-CML 388.
- Bosutinib (Study 200): confirmed CHR and overall hematologic response by week 48 were the stated efficacy endpoints 631634.
Pediatric CML
Pediatric programs have been single-arm and have used a range of endpoints:
- Nilotinib (A2203): single-arm phase 2. The primary endpoint was MMR at cycle 6 in resistant/intolerant patients. In newly diagnosed patients the co-primary endpoints were MMR by cycle 12 and CCyR at cycle 12 2559.
- Dasatinib: EMA's assessment describes on-study MCyR as the primary endpoint for the imatinib-resistant/intolerant cohort and on-study CCyR for the newly diagnosed cohort 449451.
- Ponatinib (pediatric phase 2): best MCyR within 12 months; 9/10 responded 214215.
- Bosutinib (BCHILD): the primary endpoints were safety and pharmacokinetics. FDA used MCyR, MMR and response duration as descriptive efficacy measures to support regular approval 579580. EMA likewise described the efficacy assessment as exploratory 245246.
Dose-optimization trials
The ponatinib OPTIC trial (NCT02467270) used a molecular threshold other than MMR. Its primary endpoint was BCR::ABL1 ≤1% IS at 12 months in patients with more than 1% at entry. In the 45 mg starting-dose group (reduced to 15 mg once the threshold was reached), 44% of patients achieved it 849852853. OPTIC results are in the CP-CML labeling, and FDA records refer to a five-year OPTIC follow-up requirement 855889.
Design patterns across approvals
- Accelerated approval followed by conversion is the norm at FDA. Imatinib, dasatinib, nilotinib, bosutinib (first-line), ponatinib and asciminib (third-line and first-line) all received accelerated approval first 940944908725866884965970. Conversion required 24-month follow-up in resistant/intolerant settings and 60-month follow-up for first-line indications 886900909735870.
- First-line indications require randomized active-controlled superiority trials. In the post-imatinib first-line approvals (dasatinib, nilotinib, bosutinib, asciminib), the investigational TKI was compared with imatinib or a TKI chosen by the investigator 273126626597. Imatinib itself was compared with interferon-alfa plus cytarabine.
- Later-line and advanced-phase indications accept single-arm trials. FDA's nilotinib reviewers accepted no comparator because no effective control therapy existed 508509. In the ponatinib review, FDA said a single-arm trial could support accelerated approval if responses were meaningful, durable and verifiable 15. CHMP warned that a non-comparative design might not be acceptable for some resistant patients without a BCR-ABL mutation 197.
- Assay quality affects whether a molecular endpoint can be reviewed. In PACE, FDA could not adequately evaluate MMR as a secondary endpoint because assay-performance data had not been submitted 18.
- Timepoints vary by agency and version. FDA has accepted MMR at 12 months, 48 weeks and 24 weeks as primary endpoints, with longer follow-up confirming durability. EMA's current appendix recommends 18 months for superiority, and CHMP raised this point when it reviewed bosutinib's 12-month endpoint 128598346760243.