Page 27 - Haematologica May 2022

sions of the SNHG15 lncRNA. In keeping with its negative prognostic impact, overexpression of the wild-type SNHG15 associated with higher proliferation rate of leukemic blasts when compared with the cytosine-to-thymidine variant. We conclude that recurrent genetic variants of lncRNA that are expressed in the leukemic blasts of CN-AML patients have prognostic and potential biological significance.
lncRNA variants in CN-AML
Introduction
Acute myeloid leukemia (AML) is heterogeneous with regard to the patients’ clinical course and the underlying molecular lesions that drive the disease.1-2 Research efforts of the past four decades have identified a growing list of genetic alterations associated with clinical outcome that could be used as biomarkers for the risk stratification of the patients’ treatment. These alterations include chromosomal abnormalities,3-5 gene mutations,6-11 and aberrant expression of RNA transcripts.12-16 The advent of next-generation sequencing has revealed that AML displays notable hetero- geneity at the level of isolated cases; leukemic blasts of indi- vidual AML patients represent, in many instances, the sum of distinct clonal subpopulations, within which mutations in different genes co-exist and co-operate.17,18 While such sequencing efforts continue to expand our understanding of AML pathogenesis, the majority of them are focused on the protein-coding fraction of the genome, which is, compara- tively, its smallest part.19
The non-protein-coding part of the genome, a large frac- tion of which is actively transcribed into non-coding RNA, is gaining gradual recognition for its important regulatory role.20,21 Long non-coding RNA (lncRNA), which are tran- scripts longer than 200 nucleotides and, per definition, lack protein-coding potential, regulate many key cellular func- tions in health and disease.22-24 Deregulated expression of individual lncRNA has been demonstrated to significantly affect the cancer phenotype and patients’ clinical out- come.25-29 We and others have previously shown that aber- rant expression of small subsets of lncRNA independently associate with the clinical outcome of patients with cytoge- netically normal AML (CN-AML).30-33 With regard to varia- tions in the nucleotide sequences of lncRNA, it has previ- ously been reported that disease-associated single nucleotide polymorphisms (SNP) are enriched in the genetic loci encoding these transcripts.34-36 In addition, acquired mutations of lncRNA that are recurrently detectable in the leukemic blasts have previously been identified.37 However, to our knowledge, their prognostic and biologic significance in AML have not yet been studied.
Herein, we analyzed whole transcriptome sequencing data of younger adults with the goal to evaluate the clinical and biological relevance of lncRNA variants in CN-AML. We used a panel of currently available databases of human genomic polymorphisms to annotate recurrent genetic vari- ants located within expressed lncRNA. We show that a sub- set of these variants independently associates with clinical outcome of CN-AML patients.
Methods
Patients and treatment
Exploratory analysis was conducted in pretreatment bone mar- row (BM) or blood samples from 377 younger adults (aged <60 years; range, 18-59 years) with de novo CN-AML. Patients were treated with intensive, first-line chemotherapy on Cancer and
Leukemia Group B (CALGB)/Alliance for Clinical Trials in Oncology (Alliance) trials. Confirmatory analyses were conducted in a set of 135 CN-AML patients (75 of whom were younger than 60 years and 60 were older) enrolled on clinical trials of the German AML Cooperative Group (AMLCG).38,39
All patients provided written informed consent regarding the research use of their specimens. All study protocols were in accor- dance with the Declaration of Helsinki and approved by Institutional Review Boards at each center.
Cytogenetic and molecular analyses
Cytogenetic analyses of CALGB/Alliance patients were per- formed in CALGB/Alliance-approved institutional laboratories and results were confirmed by central karyotype review.40 The diagnosis of normal karyotype was based on analysis of ≥20 metaphases obtained from BM specimens subjected to short-term (24- or 48-hour) unstimulated cultures.40
Mutational analyses of patient samples were conducted with
Sanger sequencing (for the CEBPA gene), fragment analysis (for
detection of FLT3-internal tandem duplications [FLT3-ITD]) and
targeted amplicon sequencing (for all other prognostic gene muta-
tions), as reported previously.31,41-43 Molecular and cytogenetic pro-
filing of the AMLCG cohort were obtained as described previous- ly.38,39
Transcriptome analyses
RNA samples of the patients treated on CALGB/Alliance proto- cols were analyzed with total RNA sequencing (RNA Seq) after depletion of ribosomal and mitochondrial RNA using the Illumina HiSeq 2500 platform. The results of the RNA Seq analysis have been deposited in the functional genomics data repository GEO and are publicly available under the accession number GSE137851. Patients in the AMLCG cohort were analyzed with RNA Seq following selection for poly-adenylated transcripts (poly-A RNA Seq) as previously described.16
For exploratory analyses, after quality control, adaptor-trimmed 50 base-pair-long paired-end reads were mapped to the human reference genome and variant calling was performed following the Genome Analysis Toolkit best practice recommendations for RNA Seq datasets.44 A two-pass variant calling approach was applied to ensure variant detection and depth of coverage (Figure 1). Unique variant positions were identified on non-coding transcripts that do not overlap with coding exons and are located in low-complexity regions of the genome (i.e., excluding repeat masked regions and segmental duplications). These variants were further evaluated for associations with clinical outcome and the expression levels of other RNA transcripts.
Statistical analyses
Clinical endpoint definitions are provided in the Online Supplementary Appendix. For each examined lncRNA variant, only patients with detectable expression of the lncRNA and adequate coverage of the variant position (i.e., depth of coverage >8) were analyzed. The estimated probabilities of disease-free (DFS), over- all (OS) and event-free (EFS) survival were calculated using the Kaplan–Meier method, and the log-rank test evaluated differences between survival distributions. Cox proportional hazard models were used to calculate hazard ratios for DFS, OS and EFS.45
haematologica | 2022; 107(5)
1035