AUTHORS
Maela Jan-Kronborg, Sophie Bislev2, Camilla Poulsen3, Lene Karlsson4, Jonas Abrahamsson5, NIRA Arad-Cohen, MD6, Daniel Cheuk7, Barbara De Moerloose8, Jose Marie Fernandez Navarro9, Bianca Goemans10, Kirsi Jahnukainen11, Maarja Karu12, Gertjan Kaspers13, Zhanna Kovalova14, Monica Munthe-Kaas15, Ulrika Norén-Nyström16, Josefine Palle17, Ramune Pasauliene18, Cornelis Jan Pronk19, Henrik Hasle, MD20, Kristian Løvvik Juul-Dam21
AFFILIATIONS
1 Department of Pediatrics and Adolescent Medicine, Aarhus University Hospital, Aarhus, Denmark
2 Department of Pediatrics and Adolescent Medicine, Aarhus University Hospital, Aarhus, Denmark
3 Department of Pediatrics and Adolescent Medicine, Aarhus University Hospital, Aarhus, Denmark
4 Department of Pediatrics, Institution for Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
5 Department of Pediatrics, Institution for Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
6 Department of Pediatric Hemato-Oncology, Rambam Health Care Campus, Haifa, Israel
7 Department of Pediatrics and Adolescent Medicine, Hong Kong Children's Hospital and The University of Hong Kong, and Hong Kong Pediatric Hematology and Oncology Study Group (HKPHOSG), Hong Kong, China
8 Department of Pediatric Hematology-Oncology, Ghent University Hospital, Gent, Belgium
9 Department of Pediatric Hemato-Oncology, Hospital Universitario y Politécnico La Fe, Valencia, Spain
10 Princess Máxima Center for Pediatric Oncology, Utrecht, Netherlands
11 New Children's Hospital, Pediatric Research Center, University of Helsinki and Helsinki University Hospital, Helsinki, Finland
12 Department of Pediatrics, SA Tallinna Lastehaigla, Tallinn, Estonia
13 Princess Máxima Center for Pediatric Oncology, Utrecht, Netherlands
14 Department of Pediatric Oncology/Hematology, Children's Clinical University Hospital, Riga, Latvia
15 Pediatric Department of Hematology and Oncology, Oslo, Norway
16 Department of Clinical Sciences, Pediatrics, Umeå University, Umea, Sweden
17 Department of Women's and Children's Health, Uppsala University, Uppsala, Sweden
18 Center of Oncology and Hematology, BMT Unit, Vilnius University Children's Hospital, Vilnius, Lithuania
19 Childhood Cancer Center, Skåne University Hospital, Lund, Sweden
20 Department of Pediatrics and Adolescent Medicine, Aarhus University Hospital, Aarhus, Denmark
21 Department of Pediatrics and Adolescent Medicine, Aarhus University Hospital, Aarhus, Denmark
ABSTRACT
Relapse remains a major barrier to cure in children with acute myeloid leukemia (AML). Increased treatment intensity and refined use of stem cell transplantation for high-risk patients in recent primary AML protocols may lead to more treatment-resistant relapses and higher treatment-related mortality (TRM). We investigated risk factors for treatment failure and death after first relapse in a large pediatric AML cohort.
We included children treated on the NOPHO-AML 2004, DB-AML01 (2004/DB01: 2004-2013), and NOPHO-DBH 2012 (2012: 2012-2021) protocols within the Nordic Society of Pediatric Hematology and Oncology (NOPHO) and Dutch Belgian Hong Kong (DBH) AML study groups who experienced first relapse between July 2004 and December 2022. Non-response (>5% bone marrow [BM] blasts) to first reinduction course and second relapse (collectively, termed R/R), secondary malignancy and death from any cause were considered events in event-free survival (EFS) after relapse and evaluated only in patients with hematological relapse (>5% BM blasts). Overall survival (OS), EFS, cumulative incidence of second relapse (CIR2), and TRM were compared at 3 years after first relapse using log-rank test, pseudo-observation regression and multivariate Cox regression.
A total of 373 children with relapse were included (hematological relapse: 323, 87%; immunophenotypic relapse: 16, 4%; molecular relapse: 19, 5%, and isolated extramedullary relapse: 15, 4%). Relapses of KMT2A-rearranged (KMT2A-r) AML other than KMT2A::MLLT3 were more frequent after 2012 therapy (37/188, 20%) than after 2004/DB01 therapy (19/185, 10%; p<0.01). Other baseline characteristics were comparable between the two time periods. Fludarabine/high-dose cytarabine-based salvage was the most common regimen applied at relapse (247/348 with detailed treatment, 71%).
In the total cohort, OS was 44% (95% confidence interval [CI]:39–49) and CIR2 was 29% (CI:24-34). EFS was 34% (CI:29–40) in patients with hematological relapse. Patients with core binding factor (CBF) AML had superior OS (69%, CI:56-79) compared to non-CBF AML (39%, CI:33-45; p<0.01), high EFS (46%, CI:30-60 vs. 32%, CI:26-38; p=0.06) and low CIR2 (19%, CI:11-30 vs. 31%, 26-36; p=0.09). Patients with KMT2A-r had inferior OS (33%, CI:23-43 vs. 48%, CI:42-53; p<0.01), EFS (24%, CI:14-35 vs. 37%, CI:30-43; p=0.03), and higher CIR2 (36%, CI:25-46 vs. 27%, CI:22-33; p=0.1) compared to AML without KMT2A-r. Early relapse (<12 months from diagnosis) was associated with lower OS (31%, CI:24-37 vs. 61%, CI:53-68; p<0.01), lower EFS (21%, CI:15-28 vs. 50%, CI:40-58; p<0.01), and a higher CIR2 (34%, CI:27-41 vs. 23%, CI:17-30; p=0.03).
OS was only 18% (CI:12-24) in 178/323 patients (55%) with hematological relapse and R/R disease with no difference between time periods. R/R patients with CBF AML had superior OS (34%, CI:15-55 vs. 15%, CI:10-22; p=0.02). Among R/R non-CBF patients, KMT2A-r AML had comparable OS to AML without KMT2A-r (18%, CI:8-30 vs. 14%, CI:8-22; p=0.7). Patients with R/R disease after early relapse had lower OS than R/R disease after late relapse (14%, CI:8-21 vs. 26%, CI:15-38; p<0.01).
TRM for the entire cohort was 8% (CI:6-12) with no difference between time periods (8%, CI:5-13 for both time periods) nor early vs. late relapse (8%, CI:4-13 vs. 9%, CI:5-14).
Time period had an independent impact on OS only among CBF patients, who showed improved OS over time (2004/DB01: 57%, CI:39-71 vs. 2012: 85%, CI:65-94; adjusted hazard ratio [HR] 0.3, CI:0.1–0.9; p=0.03) driven by an OS of 100% in 11 CBF patients from the 2012 time period pre-emptively treated for molecular relapse. Relapse of KMT2A-r AML (OS: HR 1.7, CI:1.3–2.4; p<0.01) and early relapse (OS: HR 3.1, CI:2.3–4.2; p<0.01 and EFS: risk ratio 2.4, CI:1.6–3.5; p<0.01) were independent predictors of adverse outcomes. Among patients with R/R disease after first relapse, both CBF AML and late relapse retained their independent favorable prognostic impact on OS (CBF: HR 0.6, CI:0.3-1.0; p=0.04 and late relapse: HR 0.4, CI:0.3-0.6; p<0.01).
OS in relapsed pediatric AML remains low, although stable over the past decades despite a relative increase of high-risk genetics. Outcome is encouraging in patients with CBF AML and molecular relapses, but remains dismal in children with KMT2A-r, early relapse or R/R disease after reinduction therapy. Safe, effective targeted therapies in clinical management of relapsed pediatric AML are needed to improve outcomes.
DISCLOSURES
Maela Jan-Kronborg: No relevant conflicts of interest to declare.