Trial Outcomes & Findings for BLAST MRD AML-1: BLockade of PD-1 Added to Standard Therapy to Target Measurable Residual Disease in Acute Myeloid Leukemia 1- A Randomized Phase 2 Study of Anti-PD-1 Pembrolizumab in Combination With Intensive Chemotherapy as Frontline Therapy in Patients With Acute Myeloid Leukemia (NCT NCT04214249)

BLAST MRD AML-1: BLockade of PD-1 Added to Standard Therapy to Target Measurable Residual Disease in Acute Myeloid Leukemia 1- A Randomized Phase 2 Study of Anti-PD-1 Pembrolizumab in Combination With Intensive Chemotherapy as Frontline Therapy in Patients With Acute Myeloid Leukemia

MRD will be assessed by multicolor flow cytometry at a central laboratory as an integral biomarker.

Will be defined per European LeukemiaNet 2017.

Median RFS will be estimated with Kaplan-Meier curves with 95% confidence interval calculated based on the Brookmeyer-Crowley method

Median DOR will be estimated with Kaplan-Meier curves with 95% confidence interval calculated based on the Brookmeyer-Crowley method.

Will be assessed per Common Terminology Criteria for Adverse Events version 5.0. Presented are a count of those that experienced at least one adverse event.

Will be assessed by duplex sequencing. Library preparation, sequencing, and analysis will be performed with TwinStrand's optimized workflow, and TwinStrand's bioinformatics core will perform all analyses related to assay output.

Will compare duplex sequencing and multiparameter flow cytometry for MRD detection. McNemar's Chi-squared test for paired samples (regular duplex sequencing versus multiparameter flow cytometry and ultra-deep duplex sequencing versus multiparameter flow cytometry) will be used to compare the MRD measures.

Will assess immune cell subsets and correlate with response to combine chemotherapy and anti PD-1 directed therapy using mass cytometry. All data will be analyzed and graphs generated using the DVS Cytobank software (Cytobank). Will also measure CD47 levels on leukemic blasts prior to and after chemotherapy and pembrolizumab application to see whether CD47 expression levels correlate with responders versus non-responders and whether expression levels change at different times of therapy. Statistical analyses of the frequency of CD8 positive (+), CD4+, Foxp3 T regulatory cells (Tregs), CD8+/Foxp3+ Tregs, central memory effector (TCM)/ memory cells that re-express CD45RA (TEMRA), effector memory (TEM)/TEMRA, the percentage of Ki67 and GzmB in PD-1+, Eomes+ CD8 T-cells to compare changes over time from baseline to several time-points will be performed by using mixed effects modeling with a Benjamini-Hochberg correction to control for false discovery rates.

An association of clinical response with the expression of PD-L1 acute myeloid leukemia bone marrow cells will be assessed by a Pearson Chi-square test on a 2 x 2 table of frequencies. The dependent variable will be defined as response (yes versus no), and quantitative immunofluorescence categories (negative versus positive) will be the independent variables. Will also monitor the dynamic change of PD-L1 expression over the course of treatment and its correlation with clinical response. Longitudinal measurements of PD-L1 will be examined using mixed-effects modeling.

Will use massively parallel sequencing technology to sequence the genomic DNA of tumor cells (leukemic bone marrow) and normal cells (germline) obtained from patients with acute myeloid leukemia. Mutational load by whole-exome sequencing will be correlated with clinic-pathological parameters such as response to treatment, survival and immune infiltrating profile, and T cell repertoire diversity and clonality.

Will investigate protein signatures associated with response and efficacy using the Olink inflammation biomarker panel.

Will be assessed using either the Agilent 4200 TapeStation and High Sensitivity RNA ScreenTape or the Agilent 2100 Bioanalyzer and RNA 6000 Pico Chip. Either method will employ the region analysis method to determine the percentage of RNA in the sample that is \> 200 nt for each sample to be processed.

Will perform high-throughput sequencing of the TCR V beta CDR3 regions on flow cytometrically sorted T-cell subsets to assess the effect of immunotherapy on the diversity of the T-cell repertoire and assess for correlation to clinical outcomes. TCR diversity and clonality will be calculated using a software by Adaptive Technologies. T-cell repertoire diversity and clonality will be correlated with clinic-pathological parameters such as response to treatment, survival, and immune infiltrating profile, as well as genomic profiles (total mutation burden, non-synonymous mutation burden, predicted neoantigen burden, clonal mutation burden and clonal predicted neoantigen burden). TCR profile generated from treatment-refractory tumors at the time of disease progression will be compared to data from pre-treatment tumor samples to explore the TCR repertoire evolution of these tumors under therapeutic pressure.

Analysis of microbiome communities will be performed in R. Pairwise differences in temporal variability across body sites will be made using Mann-Whitney U test, whereas pairwise differences among response groups performed using Student's t-test. Correlation between microbial/metabolome changes with immune-checkpoint expression and kinetics of immune cell subset recovery and programming in the standard of care and experimental arm will be evaluated. As a marker of mucosal immunity, changes in immune cell content within stool samples in patients that experience colitis or gastrointestinal graft versus host disease will be compared to suitable controls.