ALL-061 Delineating the Genomic and Proteomic Landscape of BCR-ABL1-like Acute Lymphoblastic Leukemia Subtype: Bench-to-Bedside Translational Research

Abstract

Baseline detection of Ph/BCR-ABL1-like-ALL cases is extremely beneficial for hematologists/physicians to design personalized medicine and to treat these patients efficiently. Worldwide, researchers use patented TaqMan low-density array (TLDA) gene expression profiling (GEP) to identify these cases. In developing countries, its detection is very challenging and expensive due to the unavailability of patented TLDA and there is a paucity of literature on the identification and characterization of Ph-like ALL cases. Stepwise identification followed by genomic and proteomic characterization of Ph-like ALL cases. This study was conducted at PGIMER, India (2016-2021). Flowcytometric-immunophenotyping and multiplex RT-PCR was performed to detect four recurrent fusion transcripts (n=1021). Transcriptome GEP was performed on 12 Ph-positive ALL cases. Ph-like-ALL cases identified using nCounter Platform (n=96) and RQ-PCR (n=536) were characterized using FISH, DD-PCR, RQ-PCR and MRD-positivity. Nine HKGs' expression was evaluated using 6 algorithms. Furthermore, we built and validated PHi-RACE classifier (n=629 cases) for its rapid detection. Lastly, we delineated the proteomic landscape of Ph-like ALL cases using LC-MS/MS. We studied 1021 treatment-naive and newly diagnosed B-ALL cases in this study cohort. Multiplex assay revealed 18.66% BCR-ABL1, 3.85% TCF3-PBX1, 5.27% ETV6-RUNX1, and 1.21% KMT2A-AFF1. Using transcriptome GEP, we identified 32.05% Ph-like-ALL cases using NanoString and 26.67% using RQ-PCR. These cases were substantially older and represented higher MRD positivity. In CRLF2 subgroup of Ph-like ALLs, we identified chimeric gene fusions including 10% JAK2 rearrangement, 8.33% CRLF2-IGH and CRLF2-P2RY8 translocation with concomitant R683S mutation. We found 8.33% IKZF1 (Δ4-7) deletion using DD-PCR. Furthermore, we developed an online logistic classifier entitled PHi-RACE (sensitivity=95.20%, specificity=83.70%, cut-off >0.28) for its detection using logistic regression. We have validated PHi-RACE classifier (n=93) and identified 56 Ph-like ALL cases. Six algorithms' findings revealed that EEF2, GAPDH and PGK1 represent the best HKGs for its detection. Lastly, from LC-MS/MS findings, we identified 3079 proteins, spliceosomal mediated mRNA splicing pathway and four miRNAs were significantly expressed in Ph-like ALL cases. The authors created and published the PHi-RACE classifier to detect Ph-like ALL cases, which has direct clinical implications. Further, we have delineated the genomic and proteomic landscape of Ph-like-ALL for the first time from developing countries.