Welcome to SIOP 2022 Interactive Programme
The Congress will officially run on CET time zone (Central European Time, Barcelona)
- André Baruchel (France)
- Stephen P. Hunger (United States of America)
OUTCOMES IN CHILDREN, ADOLESCENTS, AND YOUNG ADULTS WITH DOWN SYNDROME AND ACUTE LYMPHOBLASTIC LEUKEMIA: A REPORT FROM THE CHILDREN’S ONCOLOGY GROUP
- Karen Rabin (United States of America)
Abstract
Background and Aims
Patients with Down syndrome (DS) and B-acute lymphoblastic leukemia (B-ALL) experience increased rates of relapse, toxicity, and death. We report results for DS B-ALL patients enrolled on Children’s Oncology Group trials conducted from 2003-2018.
Methods
We analyzed data for patients with DS (n = 743) and without DS (n = 20,067), age 1-30 years, enrolled on four B-ALL standard-risk (SR) and high-risk (HR) trials.
Results
Patients with DS exhibited more frequent minimal residual disease (MRD) >0.01% at end-induction (30.8% vs 21.5%, p <0.001) compared to non-DS B-ALL. Within patients with NCI HR B-ALL, this difference persisted at end-consolidation (34.0% vs 11.7%, p <0.0001). Five-year event-free survival (EFS) and overall survival (OS) were significantly poorer for patients with vs without DS overall (EFS 79.0+2.0% vs 87.3+0.3%, p <0.0001; OS 86.8+1.7% vs 93.5+0.2%, p <0.0001), and within NCI SR and HR subgroups. Multivariable Cox regression analysis for risk factors associated with inferior EFS in DS B-ALL identified age >10 years, initial white blood count ≥50,000/microliter, and end-induction MRD positivity, but not cytogenetics or flow cytometric CRLF2 overexpression. Patients with DS demonstrated higher 5-year cumulative incidence of relapse (11.7+1.4% vs 9.3+0.2%, p = 0.0198), induction death (3.4% vs 0.8%, p <0.0001), and death in remission (4.8+0.8% vs 1.7+0.1%, p <0.0001). Mucositis, infections, and hyperglycemia were significantly more frequent in all patients with DS, while seizures were more frequent in patients with DS on HR trials (4.1% vs 1.7%, p = 0.001).
Conclusions
Patients with DS B-ALL exhibit increased rates of treatment toxicity, death during treatment, and relapse. Novel therapeutic strategies are needed to improve these outcomes.
COMPARISON OF CURRENT AND ENHANCED RISK STRATIFICATION OF 21,199 CHILDREN, ADOLESCENTS, AND YOUNG ADULTS WITH ACUTE LYMPHOBLASTIC LEUKEMIA: A CHILDREN’S ONCOLOGY GROUP REPORT
- Natalie Delrocco (United States of America)
Abstract
Background and Aims
Children’s Oncology Group (COG) therapeutic risk stratification algorithms use categorical (yes/no) variables. We investigated whether using continuous variables assigned with different weights would improve relapse prediction.
Methods
We retrospectively classified patients (N=21,199 from COG trials AALL0331/0232; AALL0932/1131) using COG’s current risk stratification algorithm. We next developed and validated a multivariable Cox model for relapse free survival (RFS) using peripheral blood minimal residual disease (MRD) at induction day 8, marrow MRD at induction day 29, white blood cell count, nonlinear age at diagnosis, CNS status, and favorable/unfavorable cytogenetics. We compared the predictive ability of this model to machine-learning (ML) alternatives: survival random forest (RF), boosted Cox, and linear support vector machine (SVM). We identified four risk groups by selecting cutoffs of the COG Prognostic Index (PICOG) model that maximized the discrimination of the predictive model and compared the model-based risk groups to COG’s current gold-standard risk groups.
Results
Increased PICOG is strongly associated with increased risk of relapse (testing hazard ratio = 2.55) and has good predictive discrimination (testing concordance index (c-index) 0.74). ML models explored did not sufficiently outperform the PICOG, with best performance from the RF model (training c-index 0.74). For those with moderate relapse risk in current COG risk classification, the PICOG identifies subgroups with variable 5-year RFS. Within current standard risk average (SR-Avg) patients, there is an ultra-low risk group (RFS = 0.96) and a group with higher risk (0.85) than typical SR-Avg patients. Similarly, amongst COG NCI high risk (HR) patients, PICOG identifies 4 groups ranging from 0.96 to 0.65, providing additional discrimination for future treatment stratification.
Conclusions
The PICOG can identify patients for whom therapeutic intensification may not result in significantly better outcomes while improving the discrimination of HR patients to allow randomized interventions with achievable hazard ratios.
OUTCOMES OF PATIENTS WITH ETV6-RUNX1 AND HYPERDIPLOID B-ACUTE LYMPHOBLASTIC LEUKEMIA TREATED IN THE ST. JUDE TOTAL 15 AND 16 STUDIES
- Hiroto Inaba (United States of America)
Abstract
Background and Aims
Children with ETV6-RUNX1 and hyperdiploid (> 50 chromosomes) B-acute lymphoblastic leukemia (B-ALL) have favorable outcomes. St. Jude classification considers these patients (with DNA index ≥ 1.16 if hyperdiploid) provisional low-risk (LR), regardless of age and presenting leukocyte count (National Cancer Institute [NCI] risk), and upgrades them if minimal residual disease (MRD) is ≥1% on day 15 or ≥ 0.01% on day 43 or there is testicular or central nervous system (CNS3) involvement.
Methods
We analyzed outcomes of children (1-18.99 years) with these genotypes in the St. Jude Total 15 and 16 studies (2000-2017) based on NCI and St. Jude risk.
Results
Patients with ETV6-RUNX1 (n = 222) or hyperdiploid (n = 296) B-ALL had 5-year event-free survival (EFS) of 97.7% ± 1.1% and 94.7% ± 1.4%, respectively. For ETV6-RUNX1, EFS did not differ between “NCI” standard-risk (SR) (n = 182; 97.8% ± 1.2%) and high-risk (HR) (n = 40; 97.5% ± 2.6%) patients (P = 0.917) or between “St. Jude” LR (n = 195; 97.4% ± 1.2%) and SR (n = 27; 100.0%) patients (P = 0.360). Notably, 37 of 40 “NCI” HR ETV6-RUNX1 patients who received low-intensity “St. Jude” LR therapy had excellent EFS (97.3% ± 2.8%). For hyperdiploid B-ALL, EFS was worse for “NCI” HR (n = 59; 87.6% ± 4.5%) than for “NCI” SR (n = 237; 96.4% ± 1.3%) patients (P = 0.016) but did not differ between “St. Jude” LR (n = 220; 96.1% ± 1.4%) and “St. Jude” SR/HR (n = 76; 90.6% ± 3.6%) patients (P = 0.133). Although EFS was similar for “NCI” SR (n = 188; 96.0% ± 1.5%) and HR (n = 32; 96.9% ± 3.2%) hyperdiploid patients classified as “St. Jude” LR (P = 0.719), EFS was worse for “NCI” HR (n = 27; 77.4% ± 8.2%) than for “NCI” SR (n = 49; 98.0% ± 2.2%) patients among those receiving “St. Jude” SR/HR intensified therapy (P = 0.004). The results were similar when only hyperdiploid patients with DNA indices ≥ 1.16 were evaluated.
Conclusions
Contemporary MRD-directed therapy provides excellent outcomes except for NCI HR hyperdiploid patients with slow early MRD response, who require new approaches. Among NCI HR patients, 93% with ETV6-RUNX1 and 54% with hyperdiploid ALL experienced excellent outcomes with a low-intensity regimen.
IAMP21 PREDICTS POOR OUTCOMES IN B-ACUTE LYMPHOBLASTIC LEUKEMIA: A REPORT FROM THE CHILDREN’S ONCOLOGY GROUP AALL1131
- Wanda Salzer (United States of America)
Abstract
Background and Aims
Background: Overall survival (OS) is >90% for pediatric acute lymphoblastic leukemia (ALL). However, patients with B-ALL and intrachromosomal amplification of chromosome 21 (iAMP21) are at very high risk (VHR) of relapse.
Methods
Methods: Children’s Oncology Group (COG) AALL1131 enrolled patients 1-30 years old with newly diagnosed High-Risk (HR) B-ALL. Following induction, iAMP21+ patients were risk-stratified to the VHR cohort, which also included Standard-Risk (SR) B-ALL patients initially enrolled on AALL0932, who crossed over to AALL1131 post-induction. In total, 330 patients with iAMP21 (142 SR; 188 HR) were enrolled and treated on AALL1131.
Results
Results: At end of induction (EOI), 37.6% of iAMP21+ patients had flow cytometry minimal residual disease (MRD) >0.01% versus 21.6% of iAMP21- patients. The 5-year event-free survival (EFS) and OS for NCI HR B-ALL iAMP21+ were 69.7% (95% Confidence Interval (CI), 56.0%, 79.8%) and 83.9% (95% CI, 70.5%, 91.6%), respectively. Both 5-year EFS/OS were significantly worse for NCI HR iAMP21+ patients who were EOI MRD-positive: 55.7% (95% CI, 29.2%, 75.7%) and 71.2% (95% CI, 38.5%, 88.6%) versus EOI-MRD negative 78.3% (95% CI, 62.0%, 88.2%) and 90.9% (95% CI, 75.6%, 96.8%) (p=0.002; p=0.007), respectively. The 5-year EFS/OS for NCI SR iAMP21+ patients were 76.1% (95% CI, 64.6%, 84.3%) and 89.1% (95% CI, 79.6%, 94.3%). Similar findings were reported in this NCI SR cohort based on EOI MRD where both EFS/OS were significantly worse in patients with EOI MRD >0.01%: 63.9% (95% CI, 44.1%, 78.3%) and 81.2% (95% CI, 62.9%, 91.1%) versus EOI MRD-negative 85.8% (95% CI, 71.3%, 93.3%) and 96.0% (95% CI, 83.6%, 99.1%) (p=0.001; p=0.009), respectively.
Conclusions
Conclusion: Despite post-induction VHR therapy, outcome remains poor for NCI SR and HR BALL iAMP21+ patients. Patients with iAMP21 who are EOI MRD-positive have significantly inferior EFS/OS, warranting new treatment approaches for these patients.
NEW PROGNOSTIC BIOMARKERS IN ACUTE LYMPHOBLASTIC LEUKEMIA: LONG NON-CODING RNAS
- Elixabet Lopez-Lopez (Spain)
Abstract
Background and Aims
Acute Lymphoblastic Leukemia (ALL) remains an important cause of death from disease in children. Therefore, the identification of new outcome predictors and therapeutic targets is still a need. In this context, long non-coding RNAs (lncRNAs), which represent an important fraction of the functional genome that has been poorly studied to date, could be novel candidates with great potential. This work aimed to identify new biomarkers of ALL relapse, analyzing expression differences in lncRNAs through a high throughput RNA sequencing approach.
Methods
Total RNA was extracted from tumor samples at diagnosis of 52 pediatric ALL patients treated following the same protocol (SEHOP-PETHEMA 2013) in three Spanish hospitals. RNA libraries (stranded and without rRNA) were generated and sequenced with a depth of 150 million paired-reads using Illumina technology. Different approaches were evaluated for alignment, quantification and differential expression analysis. Kaplan-Meier and Cox Proportional Hazard Model methods were used to perform 5-year event-free survival (EFS) analyses.
Results
A total of 25,783 expressed lncRNAs were identified in our patient population from 56,846 lncRNAs annotated in LNCipedia. Among them, 18 lncRNAs were differentially expressed (padj<0.01; log2FC>2/<-2) comparing relapsed (n=7) vs. non-relapsed patients (n=45). Patients were divided into three groups for survival analyses: “High expression”, “Low expression”, and “No expression”. From the 18 significant lncRNAs, 14 showed significant differences (p<0.05) in Kaplan-Meier survival analyses. These results point to an important role of lncRNA expression in treatment response, and warrant further downstream studies to understand their role in the pathogenesis of the disease and as potential new therapeutic targets.
Conclusions
Our high throughput approach identified 14 lncRNAs associated with relapse and EFS, suggesting that lncRNA expression pattern could serve as prognostic predictor in pediatric ALL.
RACIAL AND ETHNIC OUTCOME DISPARITIES AMONG PATIENTS WITH ACUTE LYMPHOBLASTIC LEUKEMIA ARE NOT FULLY ATTENUATED BY DISEASE PROGNOSTICATORS OR SOCIOECONOMIC STATUS: A COG STUDY
- Sumit Gupta (Canada)
Abstract
Background and Aims
Previous studies identified racial and ethnic disparities in childhood acute lymphoblastic leukemia (ALL) survival. We determined whether disparities persist in contemporaneous cohorts and if present, are attributable to differences in leukemia biology or socioeconomic status (SES).
Methods
Patients with newly diagnosed ALL, 0-30 years of age, enrolled on Children’s Oncology Group (COG) trials between 2004-2019 were included. Race/ethnicity was categorized as non-Hispanic White, Hispanic, non-Hispanic Black, non-Hispanic Asian, and non-Hispanic Other. Insurance status served as a proxy for SES. Event-free and overall survival (EFS, OS) were compared across race/ethnicity. The relative contribution of clinical and biologic disease prognosticators and SES was examined.
Results
The cohort included 21,152 patients. Five-year EFS was 87.4%±0.3% among non-Hispanic White patients vs. 82.8%±0.6% [HR 1.37, 95% confidence interval (95CI) 1.26-1.49; p<0.0001] among Hispanic patients and 81.9%±1.2% (HR 1.45, 95CI 1.28-1.56; p<0.0001) among non-Hispanic Black patients. Inferior EFS among Hispanic patients was substantially attenuated by disease prognosticators and SES (HR decreased from 1.37 to 1.11; p=0.045). The increased risk among non-Hispanic Black patients was minimally attenuated (HR 1.45 to 1.32). Disparities in OS were wider than EFS. Disparities were restricted to B-ALL patients as no differences in EFS or OS were seen in T-ALL based on race/ethnicity. Among patients with B-ALL, Hispanic, non-Hispanic Black, and non-Hispanic Other patients experienced higher cumulative incidences of relapse, isolated bone marrow relapse, relapses involving the CNS, testicular relapse, and death in remission. Relapses in these three ethnic/racial groups occurred earlier than relapses in non-Hispanic White or non-Hispanic Asian patients.
Conclusions
Substantial disparities in outcome for B-ALL persist by race/ethnicity, but are not observed in T-ALL. Underlying mechanisms may vary between disadvantaged groups. Future studies of relapsed patients, and of access to and quality of care are warranted to inform interventions aimed at dismantling racial and ethnic disparities.