Validation of a Classifier for the Prediction of Risk of Relapse Using Single Cell Network Profiling (SCNP) Assays for Childhood AML
- To validate the accuracy of the My Profile Pediatric Assay to predict risk of relapse
in pediatric patients with non-M3 acute myeloid leukemia (AML) who responded to
cytarabine-based induction chemotherapy.
- To validate the continuous score from the pre-specified classifier as a predictor of
relapse after response to induction chemotherapy, while controlling for the
simultaneous effects of the following clinical and laboratory variables: age, WBC,
ethnicity, cytogenetics, Flt3-ITD, NPM1, MRD, and bone marrow (BM) donor availability.
- To validate the accuracy of the prespecified low versus high relapse indicator variable
(I_L/H) as a predictor of relapse after induction therapy.
- To validate the accuracy of the prespecified I_L/H as a predictor of relapse after
induction therapy, while controlling for the simultaneous effects of the following
clinical and laboratory variables: MRD, age, WBC, ethnicity, cytogenetics, Flt3-ITD,
NPM1, and BM donor availability.
OUTLINE: Archived bone marrow mononuclear cells are analyzed by single cell network profile
assay, the My Profile™ AML Risk of Relapse Assay. Molecular markers analyzed include
Flt3-ITD, NPM1, and MRA. Results are then correlated with each patient's clinical data
including patient's age, race/ethnic background, gender, treatment received, and outcomes.
Area under the receiver operating characteristic (ROC) curve of continuous score from the pre-specified classifier in predicting relapse after complete response to cytarabine- based induction chemotherapy
A logistic regression model and area under the ROC curve for the model will be tested for significance against a null value of 0.5.
Up to 3 years
Norman J. Lacayo, MD
United States: Institutional Review Board