-Author name in bold denotes the presenting author
-Asterisk * with author name denotes a Non-ASH member
Clinically Relevant Abstract denotes an abstract that is clinically relevant.

PhD Trainee denotes that this is a recommended PHD Trainee Session.

Ticketed Session denotes that this is a ticketed session.

1385 Revised Prognostic Index for Patients with Acute Myeloid Leukemia and Other High-Grade Myeloid Neoplasms in First Relapse

Program: Oral and Poster Abstracts
Session: 613. Acute Myeloid Leukemias: Clinical and Epidemiological: Poster I
Hematology Disease Topics & Pathways:
Research, Acute Myeloid Malignancies, AML, Clinical Research, health outcomes research, Diseases, Myeloid Malignancies
Saturday, December 10, 2022, 5:30 PM-7:30 PM

Noam E. Kopmar, MD1,2, Megan Othus, PhD3,4, Olivia Gilbert, BA1*, Kelda Schonhoff, PA-C2*, Carole M. Shaw, BA5*, Kathryn Russell, ARNP2*, Anna B. Halpern, MD6,7,8, Jacob S. Appelbaum, MD, PhD7,9, Roland B. Walter, MD, PhD, MS7,10,11, Paul C. Hendrie, MD, PhD6,7, Bart L. Scott, MD12,13 and Mary-Elizabeth M. Percival, MD2,6

1Division of Hematology, University of Washington, Seattle, WA
2Fred Hutchinson Cancer Center, Seattle, WA
3Department of Biostatistics, University of Washington School of Public Health, Seattle, WA
4Division of Public Health, Fred Hutchinson Cancer Research Center, Seattle, WA
5Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, WA
6Department of Medicine, University of Washington, Seattle, WA
7Clinical Research Division, Fred Hutchinson Cancer Center, Seattle, WA
8Clinical Research Division, Fred Hutchinson Cancer Center, Seattle, WA
9Department of Medicine, Division of Hematology, Seattle, WA
10Division of Hematology, Dept of Medicine, University of Washington Medical Center, Seattle, WA
11Department of Epidemiology, University of Washington Medical Center, Seattle, WA
12Department of Medicine, Division of Medical Oncology, University of Washington, Seattle, WA
13Fred Hutchinson Cancer Research Center, Seattle, WA


Though many patients with acute myeloid leukemia (AML) will achieve a first complete remission (CR1), most patients will relapse. A previous analysis from HOVON on 667 consecutive AML patients in first relapse identified four key predictors for achieving a second CR (CR2), which improved overall survival (OS): age at relapse, CR1 duration, cytogenetic risk, and prior receipt of hematopoietic cell transplant (HCT), thereby developing a scoring system to aid in therapeutic decision making (Breems, JCO 2005). Since this original publication, other predictors of relapse have emerged, most notably measurable residual disease (MRD), and cytogenetic/molecular risk stratification has been refined in the form of ELN2017 (Dohner, Blood 2017). We hypothesize that incorporation of MRD and ELN2017 categorization into a risk stratification model will improve our ability to predict CR2 likelihood and OS after relapse.


Eligible patients included those ≥18 years (yrs) old who received treatment (tx) at our center for AML or other high-grade myeloid neoplasms (≥10% marrow/peripheral blasts at diagnosis) between January 2008 and July 2019, who later relapsed. Using our IRB-approved institutional AML database, we obtained the following demographic and outcome parameters: age; gender; new diagnosis (ND) type (AML vs. other; de novo vs. secondary); ECOG performance status; TRM score; ELN2017 risk; CR1 date; CR1 details [CR vs. CR with incomplete hematologic recovery (CRi) and MRD status]; initial tx intensity [low being hypomethylating agents (HMA); intermediate being 7+3 or similar; high being FLAG-ida/CLAG-M or similar]; cycles to CR1; prior HMA before relapse; prior HCT before relapse; relapse date; relapse salvage tx type and intensity; response to salvage (CR vs. CRi + MRD status); CR2 date; subsequent HCT; and death date or date of last contact. We fit multivariable (MV) models for OS after first relapse (Cox) and CR2 after first relapse (logistic). A score for each respective outcome was developed by rounding the coefficients from the models. C-statistics (Cox) and area under the receiver operating characteristic curve (AUC; logistic) were also calculated.


We identified 315 eligible patients who achieved CR1 and relapsed. Among these, 266 received subsequent tx and were included in our analysis: the median age at relapse was 63 yrs (range 20-85); 40% (n=106) were female; 56% (n=148) received high-intensity tx for ND; 47% (n=125) had secondary AML at ND; 28% (n=75) had received prior HCT before relapse; 36% (n=95) had received prior HMA before relapse. For initial tx of first relapse, 52% (n=138) received high-intensity tx; 5% (n=13) received HCT; 12% (n=31) received intermediate-intensity tx; and 32% (n=84) received low-intensity tx. Overall, 60% (n=160) achieved a subsequent CR2.

Significant covariates in the CR2 MV model were CRi following ND tx (odds ratio [OR] 0.85; 95% confidence interval [CI] 0.73-0.99), ≥2 cycles to CR1 (OR 0.86; CI 0.75-0.99), CR1 7-18 m (OR 1.22; CI 1.07-1.38), CR1>18 m (OR 1.31; CI 1.12-1.53), post-relapse HCT (OR 1.43; CI 1.1-1.86), and low-intensity tx post-relapse (OR 0.77; CI 0.68-0.87). Significant covariates leading to improved OS in the MV model were CR1 duration of 7-18 m (hazard ratio [HR] 0.5; 95% CI 0.36-0.69), CR1 duration >18 m (HR 0.36; CI 0.24-0.55); and post-relapse HCT (HR 0.49; CI 0.26-0.91). We derived scoring systems from these MV models (Table 1). A positive score indicates a favorable prognosis (median OS of 13.6 m); a score of 0 (median OS 6.5 m) or below (median OS 5.6 m) indicates a poor prognosis (Figure 1). When the Breems scoring system for OS was applied to our dataset, the C-statistic was 0.55; our score improved to a C-statistic of 0.61. The AUC for the CR2 score from our model was 0.69.


The Breems system was not well-calibrated to our dataset, underscoring the need to develop an updated scoring model for OS and CR2 in relapsed AML patients. Surprisingly, factors such as CR1 MRD status and ELN2017 risk did not appear to be significant in MV analyses. Efforts are underway to update this dataset to include additional contemporary patients; we hope to enhance the prognostic scoring algorithm by capturing the effect of specific mutational profiling (e.g. TP53, FLT3, and others).

Disclosures: Othus: Merck: Consultancy; Celgene: Consultancy; Biosight: Consultancy; Glycomimetics: Consultancy; Daiichi Sankyo: Consultancy. Halpern: Karyopharm Therapeutics: Research Funding; Incyte Pharmaceuticals: Research Funding; Jazz Pharmaceuticals: Research Funding; Gilead Sciences: Research Funding; Imago Biosciences: Research Funding; Novartis: Research Funding; Tolero Pharmaceuticals: Research Funding; Bayer Pharmaceuticals: Research Funding; Abbvie: Consultancy; Notable Labs: Consultancy. Walter: New Link Genetics: Consultancy; GSK: Consultancy; Genentech: Consultancy; Celgene, Inc: Consultancy, Research Funding; Bristol Myers Squibb, Inc: Consultancy; Stemline Therapeutics: Research Funding; AbbVie: Consultancy; Selvita: Research Funding; ImmunoGen: Research Funding; Janssen Global Services, LLC: Consultancy; Janssen Research and Development: Research Funding; Jazz Pharmaceuticals: Consultancy, Research Funding; Kite Pharma, Inc: Consultancy; Kronos Bio, Inc: Consultancy; Kura Oncology: Consultancy, Research Funding; MacroGenics: Consultancy, Research Funding; Race Oncology LTD: Consultancy; Amphivena Therapeutics, Inc: Current equity holder in publicly-traded company; Boston Biomedical, Inc: Consultancy; BioLineRx, LTd: Consultancy, Research Funding; Arog Pharmaceuticals: Research Funding; Astellas Pharma US, Inc: Consultancy; Amgen: Consultancy, Research Funding; Pfizer, Inc: Consultancy, Research Funding; Aptevo Therapeutics: Consultancy, Research Funding; Agios: Consultancy, Research Funding; BerGenBio, ASA: Consultancy; Orum Therapeutics, Inc.: Consultancy. Scott: Nektar: Other: data and safety monitoring board; Jazz Pharmaceuticals: Other: Advisory Panel; Novartis: Other: Advisory Panel, Research Funding; Alexion: Consultancy; Celgene: Consultancy, Honoraria, Other: Advisor Panel; Bristol Myers Squibb: Consultancy, Honoraria, Other: Advisory Panel, Research Funding; Johnson and Johnson: Other: data and safety monitoring board; Incyte: Consultancy. Percival: Telios: Research Funding; Oscotec: Research Funding; Cardiff Oncology: Research Funding; Trillium: Research Funding; Celgene/BMS: Research Funding; Abbvie: Research Funding; Biosight: Research Funding; Ascentage: Research Funding; Glycomimetics: Research Funding; Pfizer: Research Funding.

*signifies non-member of ASH