LONG NON-CODING RNA PROFILING PROVIDES IMPORTANT PROGNOSTIC INFORMATION AND BIOLOGIC INSIGHTS IN YOUNGER PATIENTS WITH CYTOGENETICALLY NORMAL ACUTE MYELOID LEUKEMIA (ALLIANCE/CALGB 8461, 9665, 20202)
(Abstract release date: 05/19/16)
EHA Library. PAPAIOANNOU D. 06/12/16; 135304; S810
Dr. Dimitrios PAPAIOANNOU
Contributions
Contributions
Abstract
Abstract: S810
Type: Oral Presentation
Presentation during EHA21: On Sunday, June 12, 2016 from 08:30 - 08:45
Location: Hall C13
Background
Aberrant expression levels of mRNA and microRNA (miR) transcripts have been shown to associate with clinical outcome of cytogenetically normal acute myeloid leukemia (CN-AML) patients (pts). Recently, long non-coding RNA (lncRNA) expression was found to be an independent prognostic marker in older CN-AML pts (Garzon et al. PNAS 2014;111:18679-84). However, the prognostic value and biological implications of lncRNA expression in younger CN-AML patients are unknown.
Aims
The aims were to determine whether lncRNA expression is associated with clinical features, recurrent mutations and prognosis of younger CN-AML patients, and to obtain biological insights into the function of lncRNAs in this disease setting.
Methods
We performed whole transcriptome profiling (RNA-seq) in 377 younger (<60 years) de novo CN-AML adult pts and investigated the associations of lncRNA expression with recurrent, prognostically-relevant gene mutations and clinical features. All patients were treated on frontline Alliance/Cancer and Leukemia Group B (CALGB) protocols. Analysis of gene mutations was performed using targeted amplicon sequencing (NPM1, DNMT3A, TET2, WT1, FLT3-TKD, IDH2, IDH1, RUNX1 and ASXL1) and Sanger sequencing (FLT3-ITD, double CEBPA). Small RNA sequencing was performed for miR profiling. The 377 patients were randomly divided into a training set for exploratory analyses (n=263) and a validation set (n=114).
Results
We derived gene mutation-related lncRNA signatures in the training set and tested their accuracy in detecting the pts’ mutational status in the validation set. Of the mutations tested, only double CEBPA mutations, NPM1 mutations and FLT3-ITD associated with distinct lncRNA signatures. These signatures were able to predict mutated or wild-type gene status in the validation set with high accuracy (sensitivity and specificity >72% in all cases). Using the training set, we identified 24 lncRNAs associated with event-free survival (P<10-6). A small number of these lncRNAs associated with prognostic gene mutations (1 with FLT3-ITD, 4 with double CEBPA mutations). Linear combination of the weighted expression values of the 24 lncRNAs yielded a prognostic lncRNA score. Using the median value as a cut-off, the lncRNA score divided the training set into two groups with very different prognoses; pts with high lncRNA scores had shorter disease-free (DFS, P<.001) and overall survival (OS, P<.001) than pts with low lncRNA scores. In the validation set, pts with high lncRNA scores were more likely to have higher white blood cell counts (P=.009), FLT3-ITD (P=.007) and high miR-155 expression (P<.001) at diagnosis. There was no significant difference in complete remission rates between pts with high and low lncRNA scores (84% v 89%). After a median follow-up of 6.2 years, pts with high lncRNA scores had shorter DFS (P<.001; 5-year DFS, 17% v 51%) and OS (P=.002; 5-year OS, 26% v 52%) than those with low lncRNA scores. In multivariable analyses, the association of high lncRNA score with shorter DFS (HR: 2.16, P=.01) and OS (HR: 1.75, P=.04) remained significant after adjusting for other covariates. We also correlated the lncRNA score with mRNA and miR expression and identified several oncogenes (FOSB, JUN, ETS2, SRC, RET) and cell-cycle regulators (PLK2, PLK3), which were highly expressed in pts with high lncRNA scores. Gene Ontology analysis revealed enrichment for genes implicated in apoptosis, programmed cell death and inflammation in the high lncRNA score pts. Similarly, miRs that associate with aggressive malignant phenotype (e.g., miR-155, miR-500, miR-362) were overexpressed in pts with high lncRNA scores.
Conclusion
We conclude that lncRNA profiling provides meaningful prognostic information and identifies potentially targetable oncogenic pathways in younger CN-AML adult pts.
Session topic: Treatment in Specific AML Subgroups
Type: Oral Presentation
Presentation during EHA21: On Sunday, June 12, 2016 from 08:30 - 08:45
Location: Hall C13
Background
Aberrant expression levels of mRNA and microRNA (miR) transcripts have been shown to associate with clinical outcome of cytogenetically normal acute myeloid leukemia (CN-AML) patients (pts). Recently, long non-coding RNA (lncRNA) expression was found to be an independent prognostic marker in older CN-AML pts (Garzon et al. PNAS 2014;111:18679-84). However, the prognostic value and biological implications of lncRNA expression in younger CN-AML patients are unknown.
Aims
The aims were to determine whether lncRNA expression is associated with clinical features, recurrent mutations and prognosis of younger CN-AML patients, and to obtain biological insights into the function of lncRNAs in this disease setting.
Methods
We performed whole transcriptome profiling (RNA-seq) in 377 younger (<60 years) de novo CN-AML adult pts and investigated the associations of lncRNA expression with recurrent, prognostically-relevant gene mutations and clinical features. All patients were treated on frontline Alliance/Cancer and Leukemia Group B (CALGB) protocols. Analysis of gene mutations was performed using targeted amplicon sequencing (NPM1, DNMT3A, TET2, WT1, FLT3-TKD, IDH2, IDH1, RUNX1 and ASXL1) and Sanger sequencing (FLT3-ITD, double CEBPA). Small RNA sequencing was performed for miR profiling. The 377 patients were randomly divided into a training set for exploratory analyses (n=263) and a validation set (n=114).
Results
We derived gene mutation-related lncRNA signatures in the training set and tested their accuracy in detecting the pts’ mutational status in the validation set. Of the mutations tested, only double CEBPA mutations, NPM1 mutations and FLT3-ITD associated with distinct lncRNA signatures. These signatures were able to predict mutated or wild-type gene status in the validation set with high accuracy (sensitivity and specificity >72% in all cases). Using the training set, we identified 24 lncRNAs associated with event-free survival (P<10-6). A small number of these lncRNAs associated with prognostic gene mutations (1 with FLT3-ITD, 4 with double CEBPA mutations). Linear combination of the weighted expression values of the 24 lncRNAs yielded a prognostic lncRNA score. Using the median value as a cut-off, the lncRNA score divided the training set into two groups with very different prognoses; pts with high lncRNA scores had shorter disease-free (DFS, P<.001) and overall survival (OS, P<.001) than pts with low lncRNA scores. In the validation set, pts with high lncRNA scores were more likely to have higher white blood cell counts (P=.009), FLT3-ITD (P=.007) and high miR-155 expression (P<.001) at diagnosis. There was no significant difference in complete remission rates between pts with high and low lncRNA scores (84% v 89%). After a median follow-up of 6.2 years, pts with high lncRNA scores had shorter DFS (P<.001; 5-year DFS, 17% v 51%) and OS (P=.002; 5-year OS, 26% v 52%) than those with low lncRNA scores. In multivariable analyses, the association of high lncRNA score with shorter DFS (HR: 2.16, P=.01) and OS (HR: 1.75, P=.04) remained significant after adjusting for other covariates. We also correlated the lncRNA score with mRNA and miR expression and identified several oncogenes (FOSB, JUN, ETS2, SRC, RET) and cell-cycle regulators (PLK2, PLK3), which were highly expressed in pts with high lncRNA scores. Gene Ontology analysis revealed enrichment for genes implicated in apoptosis, programmed cell death and inflammation in the high lncRNA score pts. Similarly, miRs that associate with aggressive malignant phenotype (e.g., miR-155, miR-500, miR-362) were overexpressed in pts with high lncRNA scores.
Conclusion
We conclude that lncRNA profiling provides meaningful prognostic information and identifies potentially targetable oncogenic pathways in younger CN-AML adult pts.
Session topic: Treatment in Specific AML Subgroups
Abstract: S810
Type: Oral Presentation
Presentation during EHA21: On Sunday, June 12, 2016 from 08:30 - 08:45
Location: Hall C13
Background
Aberrant expression levels of mRNA and microRNA (miR) transcripts have been shown to associate with clinical outcome of cytogenetically normal acute myeloid leukemia (CN-AML) patients (pts). Recently, long non-coding RNA (lncRNA) expression was found to be an independent prognostic marker in older CN-AML pts (Garzon et al. PNAS 2014;111:18679-84). However, the prognostic value and biological implications of lncRNA expression in younger CN-AML patients are unknown.
Aims
The aims were to determine whether lncRNA expression is associated with clinical features, recurrent mutations and prognosis of younger CN-AML patients, and to obtain biological insights into the function of lncRNAs in this disease setting.
Methods
We performed whole transcriptome profiling (RNA-seq) in 377 younger (<60 years) de novo CN-AML adult pts and investigated the associations of lncRNA expression with recurrent, prognostically-relevant gene mutations and clinical features. All patients were treated on frontline Alliance/Cancer and Leukemia Group B (CALGB) protocols. Analysis of gene mutations was performed using targeted amplicon sequencing (NPM1, DNMT3A, TET2, WT1, FLT3-TKD, IDH2, IDH1, RUNX1 and ASXL1) and Sanger sequencing (FLT3-ITD, double CEBPA). Small RNA sequencing was performed for miR profiling. The 377 patients were randomly divided into a training set for exploratory analyses (n=263) and a validation set (n=114).
Results
We derived gene mutation-related lncRNA signatures in the training set and tested their accuracy in detecting the pts’ mutational status in the validation set. Of the mutations tested, only double CEBPA mutations, NPM1 mutations and FLT3-ITD associated with distinct lncRNA signatures. These signatures were able to predict mutated or wild-type gene status in the validation set with high accuracy (sensitivity and specificity >72% in all cases). Using the training set, we identified 24 lncRNAs associated with event-free survival (P<10-6). A small number of these lncRNAs associated with prognostic gene mutations (1 with FLT3-ITD, 4 with double CEBPA mutations). Linear combination of the weighted expression values of the 24 lncRNAs yielded a prognostic lncRNA score. Using the median value as a cut-off, the lncRNA score divided the training set into two groups with very different prognoses; pts with high lncRNA scores had shorter disease-free (DFS, P<.001) and overall survival (OS, P<.001) than pts with low lncRNA scores. In the validation set, pts with high lncRNA scores were more likely to have higher white blood cell counts (P=.009), FLT3-ITD (P=.007) and high miR-155 expression (P<.001) at diagnosis. There was no significant difference in complete remission rates between pts with high and low lncRNA scores (84% v 89%). After a median follow-up of 6.2 years, pts with high lncRNA scores had shorter DFS (P<.001; 5-year DFS, 17% v 51%) and OS (P=.002; 5-year OS, 26% v 52%) than those with low lncRNA scores. In multivariable analyses, the association of high lncRNA score with shorter DFS (HR: 2.16, P=.01) and OS (HR: 1.75, P=.04) remained significant after adjusting for other covariates. We also correlated the lncRNA score with mRNA and miR expression and identified several oncogenes (FOSB, JUN, ETS2, SRC, RET) and cell-cycle regulators (PLK2, PLK3), which were highly expressed in pts with high lncRNA scores. Gene Ontology analysis revealed enrichment for genes implicated in apoptosis, programmed cell death and inflammation in the high lncRNA score pts. Similarly, miRs that associate with aggressive malignant phenotype (e.g., miR-155, miR-500, miR-362) were overexpressed in pts with high lncRNA scores.
Conclusion
We conclude that lncRNA profiling provides meaningful prognostic information and identifies potentially targetable oncogenic pathways in younger CN-AML adult pts.
Session topic: Treatment in Specific AML Subgroups
Type: Oral Presentation
Presentation during EHA21: On Sunday, June 12, 2016 from 08:30 - 08:45
Location: Hall C13
Background
Aberrant expression levels of mRNA and microRNA (miR) transcripts have been shown to associate with clinical outcome of cytogenetically normal acute myeloid leukemia (CN-AML) patients (pts). Recently, long non-coding RNA (lncRNA) expression was found to be an independent prognostic marker in older CN-AML pts (Garzon et al. PNAS 2014;111:18679-84). However, the prognostic value and biological implications of lncRNA expression in younger CN-AML patients are unknown.
Aims
The aims were to determine whether lncRNA expression is associated with clinical features, recurrent mutations and prognosis of younger CN-AML patients, and to obtain biological insights into the function of lncRNAs in this disease setting.
Methods
We performed whole transcriptome profiling (RNA-seq) in 377 younger (<60 years) de novo CN-AML adult pts and investigated the associations of lncRNA expression with recurrent, prognostically-relevant gene mutations and clinical features. All patients were treated on frontline Alliance/Cancer and Leukemia Group B (CALGB) protocols. Analysis of gene mutations was performed using targeted amplicon sequencing (NPM1, DNMT3A, TET2, WT1, FLT3-TKD, IDH2, IDH1, RUNX1 and ASXL1) and Sanger sequencing (FLT3-ITD, double CEBPA). Small RNA sequencing was performed for miR profiling. The 377 patients were randomly divided into a training set for exploratory analyses (n=263) and a validation set (n=114).
Results
We derived gene mutation-related lncRNA signatures in the training set and tested their accuracy in detecting the pts’ mutational status in the validation set. Of the mutations tested, only double CEBPA mutations, NPM1 mutations and FLT3-ITD associated with distinct lncRNA signatures. These signatures were able to predict mutated or wild-type gene status in the validation set with high accuracy (sensitivity and specificity >72% in all cases). Using the training set, we identified 24 lncRNAs associated with event-free survival (P<10-6). A small number of these lncRNAs associated with prognostic gene mutations (1 with FLT3-ITD, 4 with double CEBPA mutations). Linear combination of the weighted expression values of the 24 lncRNAs yielded a prognostic lncRNA score. Using the median value as a cut-off, the lncRNA score divided the training set into two groups with very different prognoses; pts with high lncRNA scores had shorter disease-free (DFS, P<.001) and overall survival (OS, P<.001) than pts with low lncRNA scores. In the validation set, pts with high lncRNA scores were more likely to have higher white blood cell counts (P=.009), FLT3-ITD (P=.007) and high miR-155 expression (P<.001) at diagnosis. There was no significant difference in complete remission rates between pts with high and low lncRNA scores (84% v 89%). After a median follow-up of 6.2 years, pts with high lncRNA scores had shorter DFS (P<.001; 5-year DFS, 17% v 51%) and OS (P=.002; 5-year OS, 26% v 52%) than those with low lncRNA scores. In multivariable analyses, the association of high lncRNA score with shorter DFS (HR: 2.16, P=.01) and OS (HR: 1.75, P=.04) remained significant after adjusting for other covariates. We also correlated the lncRNA score with mRNA and miR expression and identified several oncogenes (FOSB, JUN, ETS2, SRC, RET) and cell-cycle regulators (PLK2, PLK3), which were highly expressed in pts with high lncRNA scores. Gene Ontology analysis revealed enrichment for genes implicated in apoptosis, programmed cell death and inflammation in the high lncRNA score pts. Similarly, miRs that associate with aggressive malignant phenotype (e.g., miR-155, miR-500, miR-362) were overexpressed in pts with high lncRNA scores.
Conclusion
We conclude that lncRNA profiling provides meaningful prognostic information and identifies potentially targetable oncogenic pathways in younger CN-AML adult pts.
Session topic: Treatment in Specific AML Subgroups
{{ help_message }}
{{filter}}