Transcriptome study in patients with acute lymphoblastic leukemia reveals a higher frequency of single nucleotide variants and insertions/deletions in relapsed patients

RUIZ MARÍA SOL; VAZQUEZ ELBA; ABBATE MARÍA MERCEDES; COTIGNOLA JAVIER; RICCHIERI MARÍA CECILIA

Honolulu

Congreso; 53rd Congress of the International Society of Paediatric Oncology (virtual); 2021

Background and aims: Patients diagnosedwith acute lymphoblastic leukemia (ALL) are stratified into risk groups based on biochemical, cytogenetic and molecular parameters, and early response totherapy. Although considerable progress on survival rates has been achieved, some patients still relapse and/or die in all risk groups. The identification of new biomarkers, such as single nucleotide variants (SNVs) and small Insertion/Deletions (InDels) with biological and clinical relevance might contribute to improve the initial stratification based on molecular profiles. Methods: We collected bone marrow samples at diagnosis from 39 patients with ALL from three hospitals in Argentina. We performed paired-end transcriptome sequencing (RNAseq). Clinico-pathological characteristics were recorded (mean follow-up: 32 months; range: 5-73). We used the RNAmut software to evaluate SNVs/InDels in 114 selected genes and 79 selected fusion genes. All found variants were analyzed and filtered based on: 1) mutant reads>=15 and variant allele fraction >= 20%; 2) variant frequency in 1000 genomes database <0.01; 3) annotation in ClinVar; 4) association with leukemia phenotype; 5) prediction of pathogenicity (Polyphen); 6) oncogenicity predictions (OncoKB); and 7) previously reports in patients with ALL (COSMIC). Results:We found a total of 9,594 variants in the 39 transcriptomes analyzed. After annotation and filtering, we identified 17 SNVs/InDels in 12 patients and 4 fusion genesin 5 patients. We observed a significant association between the presence of SNVs/InDels with disease relapse (OR=33; 95%CI=1.63-669.99; p=0.023) compared to patients without SNVs/InDels. No associations were found with MRD at day 15 nor Risk Group. Conclusions: The associations observed warrant further analysis after a longer follow-up, and in a validation cohort. Our results show that it is possible to detect oncogenic mutations at diagnosis using RNA samples from bone marrow aspirates. Their assessment might contribute to the detection of patients with an increased risk of relapse.