AUTHORS
Timothy Pardee, MD, PhD, Michael Keng, MD2, Deepa Jeyakumar, MD3, Yazan F. Madanat, MD4, Diego Adrianzen Herrara, MD MS5, Rushang D. Patel, MD PhD6, Brenda Cooper, MD7, Sravanti Rangaraju, MD8, Zeina Al-Mansour, MD9, Roberto Ferro, MD10, Ken Byrd, DO11, Hana Safah, MD12, Emily Curran, MD13, Gary J. Schiller, MD14, Karen Seiter, MD15, Sharad Khurana, MD, MSc16, Barry Anderson, MD, PhD17, Kenzo Iizuka, PhD18, Chun Zhang19, Tapan Kadia I, MD20
AFFILIATIONS
1 Wake Forest University, Winston-Salem, North Carolina, United States
2 University of Virginia, Charlottesville, Virginia, United States
3 University of California Irvine, Irvine, California, United States
4 UT Southwestern, Dallas, United States
5 University of Vermont, Burlington, Vermont, United States
6 Advent Health, Orlando, United States
7 University Hospitals, Cleveland, United States
8 University of Alabama, Birmingham, United States
9 University of Florida, Gainsville, United States
10 Avera Medical Group, Sioux Falls, United States
11 University of Kansas, Leawood, United States
12 Tulane University, New Orleans, United States
13 University of Cincinnati, Cincinnati, United States
14 University of California Los Angeles, Los Angeles, United States
15 New York Medical College, Valhalla, United States
16 University of Arizona, Tucson, United States
17 Theradex Oncology, Princeton, New Jersey, United States
18 Delta Fly Pharma, Tokushima, Japan
19 Delta Fly Pharma, Tokushima, Japan
20 MD Anderson Cancer Center, Houston, United States
ABSTRACT
Background: Patients (pts) with relapsed or refractory (R/R) acute myeloid leukemia (AML) receiving second or later salvage treatment, particularly with adverse cytogenetic or molecular features, have limited options and poor outcomes. DFP-10917 is a novel deoxycytidine analog administered by prolonged continuous infusion. We report results of a randomized phase 3 trial in second-, third-, or fourth-line salvage treatment.
Methods: In this open-label, phase 3 trial, pts with R/R AML after 2–4 prior induction regimens were randomized 1:1 to DFP-10917 6 mg/m²/day by continuous intravenous infusion on days 1–14 of each 28-day cycle or investigator-selected salvage therapy. Investigators preselected nonintensive reinduction (NIR) with low-dose cytarabine, azacitidine, decitabine, or a venetoclax-based regimen or an intensive reinduction (IR) with a high- or intermediate-dose cytarabine regimen; allocation was balanced accordingly. The primary endpoint was complete remission (CR). Secondary endpoints included composite complete remission (CRc; CR, CR with incomplete hematologic recovery [CRi], or CR with incomplete platelet recovery [CRp]), duration of remission, overall survival (OS), transition to hematopoietic stem cell transplantation (HSCT), safety, tolerability, and pharmacokinetics (PK). At the 150-patient interim analysis, the Data Safety Monitoring Board (DSMB) reviewed an unblinded CR effect estimate and could stop for superiority or futility. Exploratory intent-to-treat (ITT) analyses by control intensity and responder cytogenetic/molecular analyses used data through 20 Nov 2025.
Results: Overall, 167 pts were randomized to DFP-10917 (n=85) or control (n=82; NIR, n=48; IR, n=34); 154 were treated and 140 were efficacy-evaluable (EE; n=77 and n=63). Median age was 63 years, 59.9% had received ≥3 prior regimens, and 69.5% had received venetoclax. The DSMB stopped the trial for futility.
EE CR rates were 6.5% (5/77) with DFP-10917 and 7.9% (5/63) with pooled control (adjusted odds ratio [OR], 0.79; 95% confidence interval [CI], 0.21–3.00; P=0.7356). EE CRc rates were 19.5% (15/77) and 19.0% (12/63), respectively (adjusted OR, 0.98; 95% CI, 0.41–2.35; P=0.9620), including 24.2% (8/33) with NIR and 13.3% (4/30) with IR. ITT CRc rates were 17.6% (15/85) with DFP-10917, 14.6% (12/82) with pooled control, and 11.8% (4/34) with IR; central marrow-evaluable CRi rates were 9.1% (7/77), 6.3% (4/63), and 0% (0/30). Median OS was 3.68 months with DFP-10917 versus 4.70 months with NIR (hazard ratio [HR], 1.06; 95% CI, 0.69–1.62; P=0.7824) and 3.58 months with IR (HR, 0.72; 95% CI, 0.46–1.13; P=0.1483). Six-, 12-, and 24-month OS rates were 43%, 27%, and 7% with DFP-10917 versus 21%, 7%, and 4% with IR. HSCT followed in 3 DFP-10917 pts (3.9%) and 5 controls (7.9%).
Among DFP-10917 responders with cytogenetic data, 7/10 (70%) had intermediate or adverse cytogenetics; the most common mutations among responding pts included ASXL1 and U2AF1 (n=2 each) and SRSF2, BCOR, SF3B1, and TP53 (n=1 each). PK showed sustained 14-day exposure: mean DFP-10917 concentrations increased from 5.13 to 6.46 ng/mL and metabolite concentrations from 4.25 to 9.13 ng/mL.
Treatment-related adverse events (TRAEs) occurred in 88.8% of DFP-10917-treated and 73.0% of control pts. The most common TRAEs with DFP-10917 versus control were febrile neutropenia (40.0% versus 27.0%), diarrhea (35.0% versus 16.2%), nausea (33.8% versus 12.2%), anemia (31.3% versus 23.0%), fatigue (25.0% versus 17.6%), and hypokalemia (18.8% versus 4.1%). Grade ≥3 adverse events (AEs) occurred in 92.5% and 85.1%, serious AEs in 51.3% and 28.4%, serious TRAEs in 17.5% and 4.1%, and fatal events in 22.5% and 13.5%. AE-related discontinuation was similar (13.8% versus 13.5%).
Conclusions: The CR+CRi+CRp rate for DFP-10917 was slightly higher than IR control, but the trial was halted for futility. CR and median OS were similar in this heavily pretreated, predominantly venetoclax-exposed R/R AML population. Exploratory ITT analyses showed higher CRc and objective response rates versus pooled control, with the largest separation from intensive salvage. OS was similar versus NIR and numerically favored DFP-10917 versus IR. Responses occurred with intermediate/adverse cytogenetic or molecular features, including ASXL1, U2AF1, and TP53 mutations. These findings suggest DFP-10917 monotherapy may be superior to intensive reinduction in poor-prognosis AML. (ClinicalTrials.gov identifier: NCT03926624)
DISCLOSURES
Michael Keng, MD: No relevant conflicts of interest to declare.