AUTHORS
Omer Karrar, MD, Alex Bataller, MD, PhD2, Manuel M. Maroun, MD3, Jovitta Jacob, RN4, Silvia Rivera, MS5, Kelly Sharon Chien, MD6, Abhishek Maiti, MBBS7, Jayastu Senapati, MBBS, MD, DM8, Wei Ying Jen, BM BCh, FRCPath9, Hannah Goulart, MD10, Tapan Kadia I, MD11, Farhad Ravandi, MD12, Courtney DiNardo, MD, MSCE13, Gautam Borthakur, MD14, Hagop Kantarjian, MD15, Guillermo Garcia-Manero, MD16, Nicholas Short, MD17
AFFILIATIONS
1 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
2 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
3 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
4 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
5 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
6 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
7 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
8 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
9 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
10 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
11 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
12 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
13 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
14 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
15 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
16 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
17 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
ABSTRACT
Background: Despite advances in the treatment of acute myeloid leukemia (AML), outcomes remain suboptimal for many patients (pts), particularly those with relapsed/refractory (R/R) disease or treated secondary AML (tsAML; i.e. AML arising from previously treated myeloid neoplasm). Clinical trial participation is recommended by consensus guidelines whenever feasible, as it provides access to novel therapeutic strategies and is associated with improved outcomes in newly diagnosed pts. However, patterns of clinical trial enrollment and barriers to participation across AML populations remain poorly characterized. Here, we characterize clinical trial enrollment and barriers among pts with frontline or R/R AML at a high-volume academic cancer center with an extensive portfolio of AML clinical trials.
Methods: We prospectively collected data on AML treatment instances at our institution from December 1, 2025 to June 30, 2026. Each treatment instance was categorized as frontline or R/R and according to whether therapy was administered on a clinical trial or not. Frontline cases were further classified as tsAML or non-tsAML (i.e. de novo or secondary AML without prior myeloid-directed therapy). Reasons for lack of enrollment on a clinical trial were prospectively recorded, with more than one reason permitted per treatment instance. The reasons were obtained from research study notes, treating physician documentation, and/or direct communication with the treating physician and clinical team. Enrollment rates on trials and barriers to enrollment were compared between frontline and R/R cohorts and among frontline subgroups.
Results: Over the study period, we identified 350 AML treatment instances, 158 (45%) frontline and 192 (55%) R/R. Among the 158 frontline pts, 20 (13%) had tsAML and 138 (87%) had non-tsAML. Overall, 233 pts (67%) were treated on a clinical trial and 117 (33%) were treated off-protocol. Pts with newly diagnosed AML were significantly more likely to receive treatment on a clinical trial than those with R/R AML (72% vs. 62%, respectively; P=0.04). Among the frontline cohort, clinical trial enrollment was significantly higher in pts with non-tsAML than in those with tsAML (77% vs. 40%, respectively; P<0.001). Among pts with tsAML, the rate of trial enrollment of pts with prior exposure to a hypomethylating agent (but without prior venetoclax exposure) was 54% (7/13) vs. 14% (1/7) in those with prior hypomethylating agent + venetoclax exposure (P=0.08), likely due to exclusion of these pts from many frontline trials.
Across the entire cohort of both frontline and R/R pts, the most common documented barriers to trial enrollment were clinical-value exclusions in 30 (26%; renal, n=13, hepatic, n=7, cardiac, n=6, multiple organs, n=4), patient wish / physician decision in 22 (19%), active infection in 20 (17%), need for emergent treatment in 11 (9%), and lack of insurance coverage in 9 (8%).
Barriers to clinical trial enrollment differed according to line of therapy. Compared with pts with R/R disease, the reason for lack of trial enrollment in pts with newly diagnosed AML was more frequently due to clinical-value exclusions (41% vs. 16%; P=0.004) and requirement of emergent treatment (18% vs. 4%; P=0.02), whereas patient wish / physician decision was more frequent among R/R cases (25% vs. 9%; P=0.04).
Frontline clinical trials often exclude pts with tsAML due to prior myeloid-directed therapy, while these pts are often not considered within R/R protocol, despite their poor outcomes. Lack of an available clinical trial as the primary reason to lack of enrollment was significantly more frequent in pts with tsAML than in those with non-tsAML (33% vs. 3%; P=0.005). Other reasons for lack of clinical trial enrollment were similar in the two groups.
Conclusion: Rates of clinical trial enrollment differed by AML disease setting, with lower enrollment rates in pts with R/R AML. Clinical-value exclusions and the need for emergent treatment were the most common barriers to trial enrollment in frontline AML, whereas patient wish / physician decision was more frequent in R/R AML. Among pts with “newly diagnosed” tsAML, the lack of an available clinical trial was a major barrier, particularly for those with prior venetoclax exposure. These findings highlight the need for more inclusive protocol language and/or dedicated studies for this poor-risk tsAML population.
DISCLOSURES
Omer Karrar, MD: No relevant conflicts of interest to declare.