Identification of circulating microRNA profiles associated with pulmonary function and radiologic features in survivors of SARS-CoV-2-induced ARDS

Figures & data

Table 1. Characteristics of the study population.

Sociodemographic characteristics
N= 154
Age (years)	61.0 [53.2;66.8]	154
Sex		154
Female	47 (30.5%)
BMI (kg/m^2)	29.0 [26.1;33.3]	152
Smoking history	151
Former	79 (52.3%)
Non-smoker	65 (43.0%)
Current	7 (4.64%)
Clinical characteristics
Hypertension	76 (50.0%)	152
Type II Diabetes Mellitus	30 (19.7%)	152
Obesity	61 (40.1%)	152
Cardiovascular disease	11 (7.24%)	152
Previous Chronic Lung Disease	14 (9.21%)	152
Asthma	12 (7.89%)	152
Chronic kidney disease	3 (1.97%)	152
Chronic liver disease	7 (4.61%)	152
Baseline characteristics in hospital admission
Oxygen saturation (%)	92.0 [89.0;94.0]	141
PaO2/FiO2	233 [155;286]	142
SaO2/FiO2	345 [180;438]	140
Hospital stay
Worst PaO2/FiO2	134.0 [91.0;186.3]	154
ARDS classification		154
Mild (201-300 mmHg)	33 (21.4%)
Moderate (101-200 mmHg)	70 (45.5%)
Severe (≤100 mmHg)	51 (33.1%)
Hospital stay (days)	18.0 [11.0;31.2]	152
ICU admission	127 (82.5%)	154
ICU stay (days)	11.0 [5.00;25.0]	125
High-flow nasal cannula	94 (61.0%)	154
Invasive mechanical ventilation (IMV)	64 (42.4%)	151
IMV duration (days)	17.0 [10.0;25.5]	63
Non-IMV	85 (56.7%)	150
Non-IMV duration (days)	3.00 [2.00;5.00]	84
Prone positioning	60 (40.0%)	150
Prone positioning duration (hours)	39.0 [23.0;72.0]	56
Antibiotics	122 (81.3%)	150
Hydroxychloroquine	68 (45.0%)	151
Tocilizumab	82 (53.9%)	152
Corticoids	129 (85.4%)	151
Remdesivir	29 (19.2%)	151
Interferon beta	20 (16.8%)	119
Lopinavir/ritonavir	65 (43.0%)	151
Corticoids at hospital discharge	36 (26.1%)	138
Post-COVID parameters
DLCO	66.2 [56.4;76.1]	154
<60	49 (31.8%)
<80	77 (50.0%)
≥80	28 (18.2%)
TSS	5.00 [2.00;8.00]	151

Continuous variables are expressed as median [P₂₅;P₇₅]. Categorical variables are expressed as n (%). ARDS: acute respiratory distress syndrome. BMI: body mass index. D_LCO: carbon monoxide diffusing capacity. FiO₂: fraction of inspired oxygen. ICU: intensive care unit. IMV: invasive mechanical ventilation. PaO₂: oxygen partial pressure. SaO₂: arterial oxygen saturation. TSS: total severity score.

Table 2. Characteristics of patients who entered in the inclusion criteria based on D_LCO tertiles.

	T1 [24.4, 60.1]	T2 (60.1, 72.3]	T3 (72.3, 100]	p-value	N
	N=51	N=50	N=53
Sociodemographic characteristics
Age (years)	64.0 [58.5;69.5]	56.5 [48.0;64.5]	59.0 [54.0;64.0]	0.026	154
Sex				0.175	154
Female	11 (21.6%)	18 (36.0%)	18 (34.0%)
BMI (kg/m^2)	28.7 [26.2;32.4]	29.2 [25.5;34.2]	29.3 [26.5;33.2]	0.497	152
Smoking history				0.678	151
Former	29 (56.9%)	23 (47.9%)	27 (51.9%)
Non-smoker	20 (39.2%)	22 (45.8%)	23 (44.2%)
Current	2 (3.92%)	3 (6.25%)	2 (3.85%)
Clinical characteristics
Hypertension	29 (56.9%)	23 (46.9%)	24 (46.2%)	0.280	152
Type II Diabetes Mellitus	10 (19.6%)	9 (18.4%)	11 (21.2%)	0.843	152
Obesity	20 (39.2%)	22 (44.9%)	19 (36.5%)	0.779	152
Cardiovascular disease	7 (13.7%)	2 (4.08%)	2 (3.85%)	0.055	152
Chronic lung disease	6 (11.8%)	5 (10.2%)	3 (5.77%)	0.294	152
Asthma	2 (3.92%)	6 (12.2%)	4 (7.69%)	0.484	152
Chronic kidney disease	1 (1.96%)	0 (0.00%)	2 (3.85%)	0.489	152
Chronic liver disease	2 (3.92%)	2 (4.08%)	3 (5.77%)	0.655	152
Baseline characteristics in hospital admission
Oxygen saturation (%)	91.5 [89.0;93.0]	93.0 [90.0;95.0]	91.0 [89.0;93.0]	0.998	141
PaO2/FiO2	236 [163;295]	238 [173;305]	227 [134;276]	0.300	142
SaO2/FiO2	413 [244;438]	358 [171;438]	275 [178;429]	0.093	140
Hospital stay
Worst PaO2/FiO2	122.0 [90.5;188.0]	134.0 [95.5;175.0]	134.0 [80.0;188.0]	0.984	154
ARDS classification				0.918	154
Mild (201-300 mmHg)	11 (21.6%)	9 (18.0%)	13 (24.5%)
Moderate (101-200 mmHg)	24 (47.1%)	25 (50.0%)	21 (39.6%)
Severe (≤100 mmHg)	16 (31.4%)	16 (32.0%)	19 (35.8%)
Hospital stay (days)	27.0 [15.5;44.5]	15.0 [10.0;31.0]	15.0 [10.0;23.8]	0.001	152
ICU admission	41 (80.4%)	43 (86.0%)	43 (81.1%)	0.928	154
ICU stay (days)	16.0 [7.25;33.8]	9.00 [5.00;21.0]	7.50 [5.00;18.2]	0.014	125
High flow nasal cannula	31 (60.8%)	32 (64.0%)	31 (58.5%)	0.261	154
IMV	27 (52.9%)	17 (35.4%)	20 (38.5%)	0.140	151
IMV duration (days)	18.0 [10.0;31.0]	18.0 [11.8;32.2]	13.5 [9.75;18.2]	0.077	63
Non-IMV	33 (64.7%)	22 (46.8%)	30 (57.7%)	0.480	150
Non-IMV duration (days)	3.00 [2.00;6.00]	3.00 [2.00;4.00]	2.50 [1.25;3.00]	0.037	84
Prone positioning	27 (52.9%)	17 (35.4%)	16 (31.4%)	0.027	150
Prone positioning duration (hours)	34.5 [15.8;57.8]	41.5 [21.8;85.5]	37.5 [26.8;62.8]	0.451	56
Antibiotics	39 (76.5%)	43 (89.6%)	40 (78.4%)	0.800	150
Hydroxychloroquine	22 (43.1%)	24 (49.0%)	22 (43.1%)	1.000	151
Tocilizumab	30 (58.8%)	27 (55.1%)	25 (48.1%)	0.275	152
Corticoids	44 (86.3%)	43 (89.6%)	42 (80.8%)	0.427	151
Remdesivir	10 (19.6%)	9 (18.4%)	10 (19.6%)	1.000	151
Interferon beta	7 (17.9%)	7 (17.9%)	6 (14.6%)	0.691	119
Lopinavir/ritonavir	22 (43.1%)	21 (42.9%)	22 (43.1%)	1.000	151
Corticoids at hospital discharge	14 (30.4%)	11 (23.4%)	11 (24.4%)	0.708	138
Post-COVID parameters
DLCO	51.5 [46.0;56.3]	66.0 [63.4;70.2]	80.4 [75.4;86.7]	<0.001	154
DLCO				<0.001	154
<60	49 (96.1%)	0 (0.00%)	0 (0.00%)
<80	2 (3.92%)	50 (100%)	25 (47.2%)
≥80	0 (0.00%)	0 (0.00%)	28 (52.8%)
TSS score	9.00 [5.00;12.0]	4.50 [2.00;7.00]	3.50 [1.00;6.00]	<0.001	151

Continuous variables are expressed as median [P₂₅;P₇₅]. Categorical variables are expressed as n (%). ARDS: acute respiratory distress syndrome. BMI: body mass index. D_LCO: carbon monoxide diffusing capacity. FiO₂: fraction of inspired oxygen. ICU: intensive care unit. IMV: invasive mechanical ventilation. PaO₂: oxygen partial pressure. SaO₂: arterial oxygen saturation. TSS: total severity score.

Figure 1. Molecular mechanisms associated with pulmonary function in survivors of SARS-CoV-2-induced ARDS. A) Volcano plot representing the p-value versus the fold change for each microRNA after comparison of D_LCO tertiles. The red dot indicates significant differences. B) Prediction model based on Random Forest. On the left, the importance of the contribution of each microRNA to the model. On the right, the best combination of microRNAs selected by the algorithm to reduce the error. C) Correlation between microRNAs that composed the signature. D) Linear or nonlinear relationship between the levels of each microRNA that composed the signature and D_LCO. The expression levels are expressed as log₁₀ (2^−ΔCq) for statistical purposes. E) Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis (selected). F) Gene Ontology (GO) analysis (selected). The p-value denotes the significance of the molecular pathway or the biological process and the size of the points represents the number of genes involved. The false discovery rate adjusted p-value cutoff was 0.05.

Table 3. Characteristics of patients who entered in the inclusion criteria based on TSS tertiles.

	T1 [0,3]	T2 (3,7]	T3 (7,20]	p-value	N
	N=56	N=53	N=42
Sociodemographic characteristics
Age (years)	56.0 [48.0;62.0]	62.0 [56.0;67.0]	66.0 [60.2;71.0]	<0.001	151
Sex				0.285	151
Male	37 (66.1%)	37 (69.8%)	32 (76.2%)
Female	19 (33.9%)	16 (30.2%)	10 (23.8%)
BMI (kg/m^2)	29.2 [26.5;34.7]	29.1 [25.6;32.9]	28.7 [27.0;32.7]	0.583	149
Smoking history				0.109	148
Former	28 (50.9%)	24 (47.1%)	26 (61.9%)
Non-smoker	21 (38.2%)	27 (52.9%)	15 (35.7%)
Current	6 (10.9%)	0 (0.00%)	1 (2.38%)
Clinical characteristics
Hypertension	21 (38.2%)	32 (61.5%)	22 (52.4%)	0.125	149
Type II Diabetes Mellitus	8 (14.5%)	13 (25.0%)	9 (21.4%)	0.358	149
Obesity	23 (41.8%)	22 (42.3%)	15 (35.7%)	0.566	149
Cardiovascular disease	3 (5.45%)	3 (5.77%)	5 (11.9%)	0.249	149
Chronic lung disease	5 (9.09%)	5 (9.62%)	4 (9.52%)	0.938	149
Asthma	7 (12.7%)	3 (5.77%)	2 (4.76%)	0.139	149
Chronic kidney disease	1 (1.82%)	0 (0.00%)	2 (4.76%)	0.361	149
Chronic liver disease	3 (5.45%)	1 (1.92%)	3 (7.14%)	0.769	149
Baseline characteristics in hospital admission
Oxygen saturation (%)	92.0 [89.2;93.8]	92.0 [89.0;94.0]	91.5 [89.0;94.0]	0.675	139
PaO2/FiO2	220 [134;263]	248 [179;294]	229 [134;300]	0.413	139
SaO2/FiO2	325 [174;433]	410 [234;443]	332 [172;432]	0.482	138
Hospital stay
Worst PaO2/FiO2	126.0 [87.5;175.0]	140.0 [118.0;176.0]	113.0 [85.2;206.0]	0.664	151
ARDS classification				0.463	151
Mild (201-300 mmHg)	11 (19.6%)	10 (18.9%)	12 (28.6%)
Moderate (101-200 mmHg)	22 (39.3%)	32 (60.4%)	13 (31.0%)
Severe (≤100 mmHg)	23 (41.1%)	11 (20.8%)	17 (40.5%)
Hospital stay (days)	15.0 [10.0;24.5]	17.5 [10.8;31.2]	26.5 [17.0;45.8]	0.001	149
ICU admission	46 (82.1%)	43 (81.1%)	37 (88.1%)	0.465	151
ICU stay (days)	6.00 [3.00;12.8]	6.00 [3.75;17.5]	15.0 [5.50;32.5]	0.003	148
High flow nasal cannula	31 (55.3%)	38 (71.7%)	24 (57.2%)	0.849	151
IMV	17 (30.9%)	23 (45.1%)	24 (57.1%)	0.010	148
IMV duration (days)	0.00 [0.00;5.00]	0.00 [0.00;13.0]	8.00 [0.00;25.0]	0.002	147
Non-IMV	29 (53.7%)	27 (52.9%)	28 (66.7%)	0.226	147
Non-IMV duration (days)	1.00 [0.00;3.00]	1.00 [0.00;3.00]	2.50 [0.00;4.75]	0.061	146
Prone positioning	16 (29.6%)	19 (37.3%)	25 (59.5%)	0.004	147
Prone positioning duration (hours)	0.00 [0.00;5.00]	0.00 [0.00;19.5]	17.5 [0.00;47.5]	<0.001	143
Antibiotics	46 (85.2%)	40 (78.4%)	36 (85.7%)	0.991	147
Hydroxychloroquine	31 (57.4%)	19 (36.5%)	18 (42.9%)	0.125	148
Tocilizumab	26 (47.3%)	31 (59.6%)	23 (54.8%)	0.417	149
Corticoids	44 (80.0%)	46 (90.2%)	36 (85.7%)	0.380	148
Remdesivir	7 (13.0%)	10 (19.2%)	10 (23.8%)	0.170	148
Interferon beta	5 (12.5%)	6 (14.6%)	9 (24.3%)	0.172	118
Lopinavir/ritonavir	28 (51.9%)	19 (36.5%)	18 (42.9%)	0.330	148
Corticoids at discharge	9 (18.8%)	16 (32.7%)	10 (26.3%)	0.295	135
Post-COVID parameters
DLCO	71.3 [65.0;81.6]	70.0 [56.6;78.1]	54.7 [48.1;62.5]	<0.001	151
DLCO				<0.001	151
<60	7 (12.5%)	14 (26.4%)	26 (61.9%)
<80	32 (57.1%)	29 (54.7%)	15 (35.7%)
≥80	17 (30.4%)	10 (18.9%)	1 (2.38%)
TSS score	2.00 [0.00;2.25]	5.00 [5.00;7.00]	11.0 [10.0;13.0]	<0.001	151

Continuous variables are expressed as median [P₂₅;P₇₅]. Categorical variables are expressed as n (%). ARDS: acute respiratory distress syndrome. BMI: body mass index. D_LCO: carbon monoxide diffusing capacity. FiO₂: fraction of inspired oxygen. ICU: intensive care unit. IMV: invasive mechanical ventilation. PaO₂: oxygen partial pressure. SaO₂: arterial oxygen saturation. TSS: total severity score.

Figure 2. Molecular mechanisms associated with radiologic features in survivors of SARS-CoV-2-induced ARDS. A) Volcano plot representing the p-value versus the fold change for each microRNA after comparison of TSS tertiles. The red dots indicate significant differences. B) Prediction model based on Random Forest. On the left, the importance of the contribution of each microRNA to the model. On the right, the best combination of microRNAs selected by the algorithm to reduce the error. C) Correlation between microRNAs that composed the signature. D) Linear or nonlinear relationship between the levels of each microRNA that composed the signature and D_LCO. The expression levels are expressed as log₁₀ (2^−ΔCq) for statistical purposes. E) Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis (selected). F) Gene Ontology (GO) analysis (selected). The p-value denotes the significance of the molecular pathway or the biological process and the size of the points represents the number of genes implicated. The false discovery rate adjusted p-value cutoff was 0.05.

Figure 3. Molecular mechanisms associated with pulmonary function and/or radiologic features. A) Venn diagram displaying the shared and unique Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (selected). B) Venn diagram displaying the shared and unique Gene Ontology (GO) terms (selected). The size of each circle is proportional to the total number of molecular pathways or biological processes related to D_LCO and TSS. On the right, significant shared KEGG pathways and GO terms are reported for each diagram.