Somatic mutation-associated risk index based on lncRNA expression for predicting prognosis in acute myeloid leukemia

Figures & data

Table 1. Clinical characteristics of AML patients in TCGA and GEO cohorts.

Clinical characteristics	Type	TCGA (n = 142)	GEO (n = 246)
Age (years)	< 60	77 (54.23%)	131 (53.25%)
	≥ 60	65 (45.77%)	115 (46.75%)
Sex	Female	64 (45.07%)	132 (53.66%)
	Male	78 (54.93%)	114 (46.34%)
Survival status	Alive	53 (37.32%)	97 (39.43%)
	Dead	89 (62.68%)	149 (60.57%)
Cytogenetic risk	Favorable	31 (21.83%)	65 (26.42%)
	Intermediate	77 (54.23%)	127 (51.63%)
	Poor	32 (22.53%)	46 (18.70%)
	unknown	2 (1.41%)	8 (3.25%)
FAB classification	M0	14 (9.86%)	17 (6.91%)
	M1	32 (22.54%)	59 (23.98%)
	M2	34 (23.94%)	58 (23.58%)
	M3	15 (10.56%)	0 (0.00%)
	M4	28 (19.72%)	62 (25.20%)
	M5	15 (10.56%)	25 (10.16%)
	M6	2 (1.41%)	4 (1.63%)
	M7	1 (0.70%)	0 (0.00%)
	unknown	1 (0.70%)	21 (8.54%)
FLT3-TKD	Yes	14 (9.86%)	22 (8.94%)
	No	128 (90.14%)	224 (91.06%)
FLT3-ITD	Yes	30 (21.13%)	49 (19.92%)
	No	112 (78.87%)	197 (80.08%)
DNMT3A mutation	Yes	34 (23.94%)	78 (31.71%)
	No	108 (76.06%)	168 (68.29%)
NPM1 mutation	Yes	38 (26.76%)	82 (33.33%)
	No	104 (73.24%)	164 (66.67%)
TP53 mutation	Yes	10 (7.04%)	19 (7.72%)
	No	132 (92.96%)	227 (92.28%)

Abbreviations: TCGA, the cancer genome atlas. GEO, gene expression omnibus.

Figure 1. Genomic instability characteristics-based hierarchical clustering of TCGA samples. (a) Kaplan–Meier survival analysis of genomic stable (GS) and genomic unstable (GU) groups. (b) Hierarchical clustering of TCGA AML samples. Samples were classified into GU-like and GS-like groups. (c) Boxplot for somatic mutation counts. (d) Co-expression network of lncRNAs and protein-coding genes. (e) Functional enrichment analysis of GO terms for protein-coding genes.

Figure 2. Construction of SMRI of AML patients in TCGA. Kaplan–Meier survival plots of high SMRI and low SMRI groups in the training (a), A (d), and B (g) cohorts. ROC curve analysis of SMRI and clinical characteristics in the training (b), A (e), and B (h) cohorts. Distribution of gene expression, SMRI, and vital status in the training (c), A (f), and B (i) cohorts.

Figure 3. Validation of SMRI in GEO. (a) Kaplan–Meier survival analysis of SMRI subtypes. (b) The area under the ROC curve based on SMRI, age, sex, and cytogenetic risk. (c) The variation trends of lncRNA expression, SMRI, and vital status of the risk subgroups. (d) Univariate and Multivariate analysis of SMRI and other prognostic factors in GEO dataset.

Table 2. Univariate and Multivariate analysis of SMRI and other prognostic factors in TCGA dataset.

Variables	Univariate analysis			Multivariate analysis
	HR	95%CI	P	HR	95%CI	P
Training cohort
Age (< 60/≥ 60)	3.11	2.02-4.77	<0.001	2.66	1.65-4.14	<0.001
Sex (Female/Male)	1.02	0.67-1.54	0.944
Risk (Poor/Favorable + Intermediate)	1.87	1.15-3.03	0.012	1.75	1.02-3.02	0.043
DNMT3A mutation (Yes/No)	0.46	0.29-0.75	0.002	0.88	0.52-1.49	0.636
FLT3-TKD (Yes/No)	2.19	1.18-4.06	0.013	2.09	0.99-4.38	0.052
FLT3-ITD (Yes/No)	1.36	0.81-2.26	0.243
NPM1 mutation (Yes/No)	0.77	0.49-1.22	0.261
SMRI (High/Low)	3.19	2.05-4.95	<0.001	2.73	1.66-4.50	<0.001
A cohort
Age (< 60/≥ 60)	3.70	1.94-7.06	<0.001	2.30	1.06-4.99	0.035
Sex (Female/Male)	1.14	0.63-2.07	0.668
Risk (Poor/Favorable + Intermediate)	1.94	1.02-3.71	0.044	1.96	0.87-4.39	0.104
DNMT3A mutation (Yes/No)	0.46	0.23-0.91	0.026	1.78	0.66-4.80	0.257
FLT3-TKD (Yes/No)	2.48	1.17-5.26	0.018	2.21	0.72-6.76	0.165
FLT3-ITD (Yes/No)	0.90	0.40-2.02	0.796
NPM1 mutation (Yes/No)	0.78	0.40-1.52	0.467
SMRI (High/Low)	1.71	1.40-2.09	<0.001	1.59	1.12-2.25	0.009
B cohort
Age (< 60/≥ 60)	2.72	1.49-4.96	0.001	2.64	1.27-5.51	0.010
Sex (Female/Male)	0.91	0.50-1.65	0.757
Risk (Poor/Favorable + Intermediate)	1.85	0.86-4.01	0.118
DNMT3A mutation (Yes/No)	0.48	0.25-0.92	0.026	0.77	0.37-1.60	0.481
FLT3-TKD (Yes/No)	1.88	0.58-6.12	0.294
FLT3-ITD (Yes/No)	1.87	0.95-3.67	0.070
NPM1 mutation (Yes/No)	0.75	0.40-1.42	0.376
SMRI (High/Low)	2.10	1.57-2.80	<0.001	2.18	1.57-3.03	<0.001

Abbreviations: SMRI, somatic mutation-associated risk index; HR, hazard ratio; CI, confidence interval.

Figure 4. Stratified survival analysis and differential distribution of SMRI in TCGA clinical subgroups. Kaplan–Meier survival curves of SMRI subtypes in clinical subgroups, samples were classified according to age (a), cytogenetic risk (b), and sex (c). (d-i) Differences of the SMRI in clinical subgroups.

Figure 5. Distribution of clinical features and GSVA of risk subgroups in training cohort. (a) Differences in clinicopathological characteristics and gene mutation status. (b) Heatmap of subgroup-specific biological pathways.

Figure 6. Mutation profiles and immune landscapes of risk subgroups in the training cohort. (a) Boxplot of TMB between risk subgroups (high SMRI and low SMRI groups). (b) Kaplan–Meier survival curves of SMRI subtypes in high TMB and low TMB groups. (c) Boxplot of immune score between risk subgroups. (d) Boxplot for immune cell infiltration. (e) Boxplot for immune checkpoint expression. *p < 0.05, **p < 0.01, ***p < 0.001.

Data availability statement

All data related to this manuscript can be downloaded for free from TCGA (https://portal.gdc.cancer.gov/), GEO (https://www.ncbi.nlm.nih.gov/geo/), and UCSC Xena (https://xenabrowser.net/).