TriangleHEALTH PROFESSIONAL · SOURCE READING
Risk Classification Systems
Source: Childhood Acute Myeloid Leukemia Treatment (PDQ®)–Health Professional Version, National Cancer Institute.
Source updated: April 15, 2025 · Captured 2026-09-09.
Selected source text with whitespace normalised. This Triangle page is not an NCI PDQ summary.
Context: Treatment Option Overview for Childhood AML
Risk classification for treatment assignment has been used by several cooperative groups performing clinical trials in children with AML. In the COG, stratifying therapeutic choices on the basis of risk factors is a relatively recent approach for the non-APL, non–Down syndrome patient.
Classification is most directly derived from the observations of the MRC AML 10 trial for EFS and OS.[66] Classification is further applied based on the ability of the pediatric patient to undergo reinduction to obtain a second complete remission and their subsequent OS after first relapse.[80]
The following COG trials have used a risk classification system to stratify treatment choices:
In COG AAML0531 (NCT00372593), the first COG trial to stratify therapy by risk group, patients were stratified into three risk groups on the basis of diagnostic cytogenetics and response after induction 1.[12]
Source links and citations
In COG AAML0531 (NCT00372593), the first COG trial to stratify therapy by risk group, patients were stratified into three risk groups on the basis of diagnostic cytogenetics and response after induction 1.[12]
Low-risk patients included those diagnosed with a core-binding factor AML (either t(8;21) or inv(16)).
In COG AAML0531 (NCT00372593), the first COG trial to stratify therapy by risk group, patients were stratified into three risk groups on the basis of diagnostic cytogenetics and response after induction 1.[12]
High-risk patients had either monosomy 7, monosomy 5 or del(5q), chromosome 3 abnormalities, or a poor response to induction 1 therapy with morphological marrow leukemic blasts (>15%).
In COG AAML0531 (NCT00372593), the first COG trial to stratify therapy by risk group, patients were stratified into three risk groups on the basis of diagnostic cytogenetics and response after induction 1.[12]
All other patients fell into the intermediate-risk category.
In COG AAML0531 (NCT00372593), the first COG trial to stratify therapy by risk group, patients were stratified into three risk groups on the basis of diagnostic cytogenetics and response after induction 1.[12]
This resulted in a risk distribution of 24% low risk, 59% intermediate risk, and 17% high risk.
In the subsequent COG-AAML1031 (NCT01371981) trial, the risk groups were reduced to two on the basis of the finding that those in the intermediate category could be more specifically and prognostically defined by adding the use of MRD by multiparameter flow cytometry.[31,81]
Source links and citations
In the subsequent COG-AAML1031 (NCT01371981) trial, the risk groups were reduced to two on the basis of the finding that those in the intermediate category could be more specifically and prognostically defined by adding the use of MRD by multiparameter flow cytometry.[31,81]
Patients whose cytogenetics and/or molecular genetics were noninformative (i.e., traditional intermediate risk) and were negative for MRD (<0.1%) were placed in the low-risk category.
In the subsequent COG-AAML1031 (NCT01371981) trial, the risk groups were reduced to two on the basis of the finding that those in the intermediate category could be more specifically and prognostically defined by adding the use of MRD by multiparameter flow cytometry.[31,81]
Patients who were positive for MRD (≥0.1%) were placed in the high-risk category.
In the COG-AAML1031 trial, the study stratification was further based on cytogenetics, molecular markers, and MRD at bone marrow recovery postinduction 1, with patients being divided into a low-risk or high-risk group as follows:[31]
Source links and citations
In the COG-AAML1031 trial, the study stratification was further based on cytogenetics, molecular markers, and MRD at bone marrow recovery postinduction 1, with patients being divided into a low-risk or high-risk group as follows:[31]
The low-risk group represented 78% of patients, had a 3-year OS rate from the end of induction 1 of 74.1% (±3.4%), and was defined by the following:
In the COG-AAML1031 trial, the study stratification was further based on cytogenetics, molecular markers, and MRD at bone marrow recovery postinduction 1, with patients being divided into a low-risk or high-risk group as follows:[31] › The low-risk group represented 78% of patients, had a 3-year OS rate from the end of induction 1 of 74.1% (±3.4%), and was defined by the following:
Inv(16), t(8;21), NPM1 variants, or CEBPA variants, regardless of MRD and other cytogenetics.
In the COG-AAML1031 trial, the study stratification was further based on cytogenetics, molecular markers, and MRD at bone marrow recovery postinduction 1, with patients being divided into a low-risk or high-risk group as follows:[31] › The low-risk group represented 78% of patients, had a 3-year OS rate from the end of induction 1 of 74.1% (±3.4%), and was defined by the following:
Intermediate-risk cytogenetics (defined by the absence of either low-risk or high-risk cytogenetic characteristics) with negative MRD (<0.1% by flow cytometry) at end of induction 1.
In the COG-AAML1031 trial, the study stratification was further based on cytogenetics, molecular markers, and MRD at bone marrow recovery postinduction 1, with patients being divided into a low-risk or high-risk group as follows:[31]
The high-risk group represented the remaining 22% of patients, had a 3-year OS rate from the end of induction 1 of 36.9% (± 7.6%), and was defined by the following:Where risk factors contradicted each other, the following evidence-based table was used (see Table 6).
Source links and citations
In the COG-AAML1031 trial, the study stratification was further based on cytogenetics, molecular markers, and MRD at bone marrow recovery postinduction 1, with patients being divided into a low-risk or high-risk group as follows:[31] › The high-risk group represented the remaining 22% of patients, had a 3-year OS rate from the end of induction 1 of 36.9% (± 7.6%), and was defined by the following:Where risk factors contradicted each other, the following evidence-based table was used (see Table 6).
High-allelic ratio FLT3 ITD positive with any MRD status.
In the COG-AAML1031 trial, the study stratification was further based on cytogenetics, molecular markers, and MRD at bone marrow recovery postinduction 1, with patients being divided into a low-risk or high-risk group as follows:[31] › The high-risk group represented the remaining 22% of patients, had a 3-year OS rate from the end of induction 1 of 36.9% (± 7.6%), and was defined by the following:Where risk factors contradicted each other, the following evidence-based table was used (see Table 6).
Monosomy 7 with any MRD status.
In the COG-AAML1031 trial, the study stratification was further based on cytogenetics, molecular markers, and MRD at bone marrow recovery postinduction 1, with patients being divided into a low-risk or high-risk group as follows:[31] › The high-risk group represented the remaining 22% of patients, had a 3-year OS rate from the end of induction 1 of 36.9% (± 7.6%), and was defined by the following:Where risk factors contradicted each other, the following evidence-based table was used (see Table 6).
Monosomy 5/del(5q) with any MRD status.
In the COG-AAML1031 trial, the study stratification was further based on cytogenetics, molecular markers, and MRD at bone marrow recovery postinduction 1, with patients being divided into a low-risk or high-risk group as follows:[31] › The high-risk group represented the remaining 22% of patients, had a 3-year OS rate from the end of induction 1 of 36.9% (± 7.6%), and was defined by the following:Where risk factors contradicted each other, the following evidence-based table was used (see Table 6).
Intermediate-risk cytogenetics with positive MRD at end of induction 1.
In the COG-AAML1031 trial, the study stratification was further based on cytogenetics, molecular markers, and MRD at bone marrow recovery postinduction 1, with patients being divided into a low-risk or high-risk group as follows:[31] › The high-risk group represented the remaining 22% of patients, had a 3-year OS rate from the end of induction 1 of 36.9% (± 7.6%), and was defined by the following:Where risk factors contradicted each other, the following evidence-based table was used (see Table 6).
| Risk Assignment: | Low Risk | High Risk | |||
|---|---|---|---|---|---|
| Low-Risk Group 1 | Low-Risk Group 2 | High-Risk Group 1 | High-Risk Group 2 | High-Risk Group 3 | |
| ITD = internal tandem duplications. | |||||
| aGroups are based on combinations of risk factors, which may be found in any individual patient. | |||||
| bBold indicates the overriding risk factor in risk-group assignment. | |||||
| cNPM1, CEBPA, t(8;21), inv(16). | |||||
| d"Any" indicates any status and thus the marker's presence/absence or minimal residual disease status does not impact risk classification in the particular Risk Group. | |||||
| eMonosomy 7, monosomy 5, del(5q). | |||||
| FLT3 ITD allelic ratio | Low/negative | Low/negative | High | Low/negative | Low/negative |
| Good-risk molecular markersc | Present | Absent | Anyd | Absent | Absent |
| Poor-risk cytogenetic markerse | Anyd | Absent | Anyd | Present | Absent |
| Minimal residual disease | Anyd | Negative | Anyd | Anyd | Positive |
The high-risk group of patients was guided to transplant in first remission with the most appropriate available donor. Patients in the low-risk group were instructed to pursue transplant if they relapsed.[68,82]
The COG AAML1831 (NCT04293562) trial for patients with newly diagnosed AML uses a more complex risk-stratification system. This system incorporates more genetic lesions into the high-risk group and builds on the use of MRD as a strong prognostic marker.[83] Specific prognostic factors were identified in Table 5.
Source links and citations
With this classification, the following three risk groups were described:
Low risk 1 (LR1).
Low risk 1 (LR1).
Presence of inv(16)/t(16;16) or t(8;21) cytogenetic features, if negative for MRD (<0.05%) at the end of induction 1, unfavorable risk markers, and KIT exon 17 variants.
Low risk 1 (LR1).
Presence of NPM1 or CEBPA variants, if negative for MRD (<0.05%) at the end of induction 1 and unfavorable risk markers.
Low risk 2 (LR2).
Low risk 2 (LR2).
Presence of inv(16)/t(16;16) or t(8;21) cytogenetic features or NPM1 or CEBPA variants, negative for unfavorable risk markers, and positive for MRD (≥0.05%) at the end of induction 1.
Low risk 2 (LR2).
Presence of inv(16)/t(16;16) or t(8;21) cytogenetic features, a coexisting KIT exon 17 variant, and negative for unfavorable risk markers.
Low risk 2 (LR2).
FLT3 ITD with allelic ratio greater than 0.1 with concurrent bZIP, CEBPA, or NPM1 variants and negative MRD (<0.05%) at the end of induction 1.
Low risk 2 (LR2).
Negative MRD (<0.05%) at the end of induction 1 and no favorable or unfavorable prognostic markers.
Low risk 2 (LR2).
Presence of a non-FLT3 ITD activating variant and negative MRD (<0.05%) at the end of induction 1, regardless of presence of favorable genetic markers.
High risk (HR).
High risk (HR).
FLT3 ITD with allelic ratio greater than 0.1 without bZIP, CEBPA, or NPM1 variants.
High risk (HR).
FLT3 ITD with allelic ratio greater than 0.1 with concurrent bZIP, CEBPA or NPM1 variants and MRD (>0.05%) at the end of induction 1.
High risk (HR).
Presence of RAM phenotype or unfavorable prognostic markers (other than FLT3 ITD) per cytogenetics, FISH, NGS Foundation Medicine results, regardless of favorable genetic markers, MRD status, or FLT3 ITD variant status.
High risk (HR).
AML without favorable or unfavorable cytogenetic or molecular features but with MRD (>0.05%) at the end of induction 1.
High risk (HR).
Presence of a non-FLT3 ITD variant and positive MRD (>0.05%), regardless of presence of favorable genetic markers.
All patients with HR AML were assigned to HSCT if a suitable donor was available, whereas patients with LR1 or LR2 disease received four or five cycles of chemotherapy, respectively.
It is important to recognize that factors used for stratification vary by pediatric and adult cooperative clinical trial groups. The prognostic impact of a given risk factor may vary in their significance depending on the backbone of therapy used. Other pediatric cooperative groups use some or all of these same factors, generally choosing risk factors that have been reproducible across numerous trials and sometimes including additional risk factors previously used in their risk group stratification approach.
Studies cited in this source section
Source citation is not a determination that a study applies to you.
Publication references
Read the original reference and check its publication notices.
Preserved source evidence · Not medical advice