Characterization of risk-stratified AML engraftment in NSGS mice. (A) Experimental design (left panel) and representative FACS analysis of the engraftment (right panels). (B) Kaplan-Meier curves representing the 18-week EFS from each risk group (n = 28 AML primary samples and 112 mice). Mice were considered “dead” when they showed disease symptoms with detectable human graft in BM or when human graft reached 70% in BM. (C) Percentage of AML engraftment in BM at euthanization for all patient samples grouped by risk (upper panel). The horizontal dotted lines indicate the cutoff for engraftment levels >0.1% (red line), between 0.1% to 5% (blue line), and >20% (green line). Each circle represents a single mouse. Gray circles depict mice with <0.01% engraftment. The mean engraftment for the 3 or 4 mice used per patient sample is represented as a horizontal bar. Percentage of engrafted mice (leukemic cells > 0.1%) and percentage of engrafted patient samples (lower panels). Based on engraftment levels, patient samples were categorized as high (leukemic graft > 20%), medium (5-20%), or low (0.1-5%). (D) In situ 3-dimensional microscopy imaging of murine BM from a representative very low engrafted sample (0.8% by FACS; supplemental Video). (E) FISH for the indicated chromosomal abnormalities in PDX cells retrieved from mice at week 18. Magnification 100×. Yellow arrowheads depict the indicated chromosome abnormalities. (F) Engraftment kinetics in BM for each AML risk group at 6, 12, and 18 weeks (or at the end point). Data are mean (± SEM) engraftment levels for all patients within each risk group. (G) As in (F) but showing each patient and risk group independently. Each point indicates the mean engraftment for each patient sample (3-4 mice). Individual mice euthanized before week 18 are indicated by “†”; the engraftment value observed at euthanization was maintained at the end point. ****P < .0001, log-rank test. NK, normal karyotype; CK, complex karyotype; D, day; IBMT, intra-BM transplantation; W, week.