Abstract The application of microRNA (miRNA) profiling in respiratory biospecimens, particularly bronchial aspirate (BAS), remains underexplored. Here, we aimed to validate and refine miRNA quantification in BAS samples to establish its suitability for molecular phenotyping. This was a multicenter study including 288 COVID-19 patients on invasive mechanical ventilation. Respiratory biospecimens included BAS, tracheal aspirate, and bronchoalveolar lavage fluid samples. A predesigned miRNA panel was evaluated using RT-qPCR. Biomarker evaluation and functional assessment were subsequently conducted. An initial technical validation phase corroborated the reproducibility of miRNA profiling in BAS samples. Comparative analyses of miRNA expression profiles across respiratory samples revealed distinct miRNA patterns among biospecimens. In the biomarker analysis, two miRNA ratios, miR-34c-5p/miR-34a-5p and miR-34c-5p/miR-125b-5p, were inversely associated with intensive care unit (ICU) survival (hazard ratio [HR]: 0.18 and 0.17, respectively) during the discovery phase. Risk and survival analyses in the test phase confirmed the reproducibility of the miR-34c-5p/miR-34a-5p ratio (hazard ratio [HR] = 0.17). Functional analyses revealed the utility of miRNA profiling in BAS for identifying pathogenic pathways and developing therapeutic targets. Overall, these findings position miRNA profiling in BAS samples as a valuable approach for biomarker discovery, identification of pathophysiological mechanisms, and development of targeted pulmonary therapies. Keywords: MT: Non-coding RNAs, biomarkers, bronchial aspirate, bronchoalveolar lavage fluid, COVID-19, invasive mechanical ventilation, microRNAs, noncoding RNA, SARS-CoV-2, tracheal aspirate Graphical abstract graphic file with name fx1.jpg [75]Open in a new tab __________________________________________________________________ de Gonzalo Calvo and colleagues’ results provide useful information about the clinical potential of microRNA profiling in respiratory specimens. Specifically, microRNA analysis in bronchial aspirate samples opens up new opportunities for developing biomarkers, exploring new pathophysiological mechanisms, and identifying new therapeutic approaches targeting the lung. Introduction MicroRNAs (miRNAs) are a class of noncoding RNAs (ncRNAs) approximately 22 nucleotides in length that regulate gene expression at the posttranscriptional level.[76]^1 miRNAs play crucial roles in a wide array of biological processes, particularly in maintaining homeostasis and mediating stress responses.[77]^2 Disruptions in miRNA expression are key factors in the onset and progression of numerous diseases, including various respiratory conditions.[78]^3 Consequently, miRNA profiling has emerged as a potent tool for elucidating the mechanistic pathways underlying human disease.[79]^4^,[80]^5 Technically, the stability and detectability of miRNAs in routinely collected clinical biospecimens, combined with reliable and cost-effective quantification through standard laboratory methods, position these small ncRNAs as practical and accessible biomarkers. Multiple studies have highlighted the potential of miRNAs for molecular phenotyping[81]^6^,[82]^7 and their utility as meaningful biomarkers for enhancing clinical decision-making.[83]^8^,[84]^9 Bronchial aspirate (BAS) is a widely used clinical biospecimen obtained from bronchial airways and provides comprehensive information about both the cellular and extracellular environments within the bronchus. BAS sampling is commonly performed for diagnostic purposes, particularly in patients with suspected pulmonary infections, where it facilitates the identification of causative pathogens and guides therapeutic decisions.[85]^10 Compared with other methods, such as bronchoalveolar lavage fluid (BALF) collection, BAS collection is notably less invasive, reducing the risk of procedural complications and making it particularly suitable for critically ill or pediatric patients. In patients undergoing invasive mechanical ventilation (IMV), BAS collection provides direct access to the lower respiratory tract, aiding in the evaluation of pulmonary infections, management of complications, and clearance of accumulated airway secretions. Despite the clinical utility of BAS, only a limited number of studies implementing innovative molecular techniques to improve its diagnostic performance have been published to date.[86]^11 In this context, the COVID-19 pandemic has provided a unique opportunity to investigate respiratory biospecimens as sources of molecular data,[87]^12 enabling advancements in the management of critically ill patients and respiratory illnesses. In a previous pilot study, our group suggested the potential of miRNA quantification in BAS samples for molecular profiling.[88]^13 We identified a panel of host miRNAs that, when combined with miRNA ratios, could offer valuable prognostic information in mechanically ventilated COVID-19 patients. The current study aimed to experimentally confirm and expand these findings by evaluating the reproducibility of miRNA quantification in BAS samples, comparing miRNA profiles across different respiratory samples, assessing the potential of validated miRNAs as biomarkers, and exploring the feasibility of miRNA profiling to elucidate pathogenetic mechanisms and identify novel therapeutic targets. Results Reproducibility of microRNA profiling in bronchial aspirate samples Given the lack of prior research on miRNA quantification in BAS samples, the initial step of our study was to assess the technical reproducibility of this approach. We analyzed 18 candidate miRNAs in independent aliquots from BAS samples previously collected and analyzed in our prior pilot study (n = 57).[89]^13 This analysis aimed to evaluate the consistency of miRNA measurements across repeated extractions from the same biospecimen. The main baseline characteristics of the study sample are summarized in [90]Table S1. Spearman’s rank correlation test was used to assess the technical reproducibility of miRNA quantification between the aliquots obtained from each patient ([91]Figure 1A; [92]Table S2). Among the 18 candidates, four miRNAs (miR-1-3p, miR-9-5p, miR-124-3p, and miR-199a-5p) were below the threshold of reproducibility (rho < 0.8) between independent aliquots. These nonvalidated miRNAs also presented lower expression levels in the analyzed samples (Cq ≥ 33) ([93]Table S2). Additionally, two miRNAs (miR-221-3p and miR-451a) demonstrated high intralaboratory variability between the respective aliquots ([94]Figure 1A). These six miRNAs were excluded from subsequent biomarker analysis. Figure 1. [95]Figure 1 [96]Open in a new tab MicroRNA profiling of respiratory samples (A) Bland-Altman plot comparing the microRNA expression levels of independent aliquots from each patient in the technical validation sample set (n = 57). Each point represents a patient. The bold and dashed lines represent the mean difference and ±1.96 SD, respectively. In this step, the expression levels were normalized using the external reference miRNA cel-miR-39-3p to avoid technical variability: ΔCq = Cq[microRNA]−Cq[cel-miR-39-3p]. (B) Dual-axis chart including the microRNA expression levels of bronchial aspirate (BAS) samples (n = 57), tracheal aspirate (TA) samples (n = 42), and bronchoalveolar lavage fluid (BALF) samples (n = 96). The bars represent the percentage of expression of each microRNA in the different respiratory specimens (expression rate axis), and the points represent the median of the quantification cycle (Cq) of each microRNA (Cq axis). Expression rate of each miRNA was calculated as the percentage of samples with detectable expression (Cq < 35). The orange line indicates the Cq threshold of 35, above which miRNAs were considered undetectable. Respiratory biospecimen specificity Given the limited exploration of miRNA quantification in respiratory biospecimens, our next step was to compare the expression profiles of the candidate miRNAs in BAS samples with those in tracheal aspirate (TA) samples (n = 42) and BALF (n = 96) samples collected from mechanically ventilated COVID-19 patients ([97]Tables S3 and [98]S4). Although six miRNAs were ultimately excluded from the biomarker analyses due to low expression or poor reproducibility in BAS samples, they were retained in this comparative analysis to illustrate biospecimen-dependent detectability. As shown in [99]Figure 1B, the expression levels of the 18 candidates varied between biospecimens, with BAS and TA showing more comparable profiles. Additionally, when defining miRNAs as “detected” if present in at least 20% of the population, we found that miRNA profiles differed depending on the type of respiratory sample, especially between BAS and TA against BALF. As expected, all 18 miRNAs analyzed were detected in the BAS samples. miR-1-3p was not detected in the TA samples, and five miRNAs (miR-1-3p, miR-9-5p, miR-124-3p, miR-199a-5p, and miR-499a-5p) were not expressed in the BALF samples ([100]Figure 1B). Biomarker value of microRNA profiling in bronchial aspirate samples Among the 12 miRNAs (miR-16-5p, miR-27b-3p, miR-34a-5p, miR-34b-5p, miR-34c-5p, miR-92a-3p, miR-93-5p, miR-125b-5p, miR-126-3p, miR-486-5p, miR-491-5p, and miR-499a-5p) displaying acceptable median expression levels (Cq < 33) and high reproducibility between independent BAS samples, we constructed miRNA ratios and evaluated their potential to predict intra-ICU mortality in a discovery cohort of critically ill COVID-19 patients undergoing IMV (n = 61). The baseline characteristics of the study population are presented in [101]Table 1. The median age of the cohort was 63.0 [55.0;67.0] years, with 29.5% being female. Compared with survivors, nonsurvivors were significantly older and had a greater incidence of hypertension. At the initiation of IMV, nonsurvivors had lower levels of oxygen saturation and higher levels of creatinine. Table 1. Characteristics of the study population (discovery phase) All (n = 61) Survivors (n = 42) Nonsurvivors (n = 19) p value n Sociodemographic characteristics __________________________________________________________________ Age (years) 63.0 [55.0;67.0] 61.5 [52.2;66.8] 67.0 [62.5;70.0] 0.011 61 Female 18 (29.5%) 14 (33.3%) 4 (21.1%) 0.502 61 BMI (kg/m^2) 32.8 (6.37) 32.7 (6.62) 32.8 (5.96) 0.946 61 __________________________________________________________________ Comorbidities __________________________________________________________________ Smoking history: – – – 0.898 60 Former 35 (58.3%) 24 (57.1%) 11 (61.1%) – – Nonsmoker 22 (36.7%) 16 (38.1%) 6 (33.3%) – – Active smoker 3 (5.00%) 2 (4.76%) 1 (5.56%) – – Hypertension 38 (62.3%) 22 (52.4%) 16 (84.2%) 0.037 61 Type II diabetes mellitus 23 (37.7%) 13 (31.0%) 10 (52.6%) 0.183 61 Chronic cardiac disease 9 (14.8%) 5 (11.9%) 4 (21.1%) 0.441 61 Chronic respiratory disease 3 (4.92%) 1 (2.38%) 2 (10.5%) 0.226 61 Chronic kidney disease 3 (4.92%) 1 (2.38%) 2 (10.5%) 0.226 61 __________________________________________________________________ Disease chronology __________________________________________________________________ Time since ICU admission to BAS sampling (days) 5.03 (2.77) 5.33 (2.79) 4.37 (2.69) 0.208 61 Time since IMV initiation to BAS sampling (days) 3.00 [1.00;4.00] 3.00 [2.00;4.75] 2.00 [1.00;3.00] 0.147 61 Time since symptom onset to BAS sampling (days) 14.1 (5.17) 14.5 (4.88) 13.3 (5.88) 0.475 60 __________________________________________________________________ Arterial blood gases and laboratory parameters at IMV initiation __________________________________________________________________ Oxygen saturation (%) 97.0 [94.0;98.0] 97.0 [95.8;99.0] 95.0 [92.5;97.5] 0.021 61 FiO[2] (%) 80.0 [63.8;92.5] 70.0 [60.0;90.0] 80.0 [80.0;100] 0.110 60 PaO[2] (mmHg) 107 (33.1) 104 (32.2) 113 (35.2) 0.438 44 PaCO[2] (mmHg) 49.0 [39.2;57.8] 46.5 [40.0;55.2] 52.5 [37.2;66.0] 0.455 42 PaO[2]/FiO[2] 137 [114;159] 140 [111;162] 134 [117;146] 0.401 42 Glucose (mg/dL) 153 [118;230] 146 [114;233] 170 [136;227] 0.299 57 Creatinine (mg/dL) 0.77 [0.63;0.93] 0.75 [0.62;0.88] 0.88 [0.72;1.17] 0.038 56 C-reactive protein (mg/L) 42.1 [16.0;142] 38.9 [17.8;150] 76.2 [9.80;108] 0.585 55 D-dimer (ng/mL) 2.89 [2.50;3.32] 2.69 [2.49;3.03] 3.17 [2.79;3.69] 0.086 46 Leukocyte count (x10^9/L) 10.6 [6.22;14.7] 11.0 [6.29;14.5] 10.0 [6.16;15.2] 0.920 58 Neutrophil count (x10^9/L) 8.87 [4.77;12.3] 9.57 [5.10;12.5] 8.18 [4.65;11.5] 0.568 58 Lymphocyte count (x10^9/L) 0.81 [0.52;1.02] 0.79 [0.54;1.01] 0.81 [0.38;1.02] 0.631 57 Monocyte count (x10^9/L) 0.39 [0.24;0.66] 0.40 [0.28;0.64] 0.36 [0.16;0.76] 0.585 58 Platelet count (x10^9/L) 252 [197;304] 264 [202;314] 243 [189;285] 0.334 58 __________________________________________________________________ Respiratory support and treatments during ICU stay __________________________________________________________________ High-flow oxygen nasal cannula 56 (91.8%) 40 (95.2%) 16 (84.2%) 0.170 61 Noninvasive mechanical ventilation 50 (82.0%) 35 (83.3%) 15 (78.9%) 0.726 61 Invasive mechanical ventilation duration (days) 14.0 [7.00;27.0] 14.0 [7.25;26.8] 12.0 [6.50;28.0] 0.663 61 Prone positioning 57 (93.4%) 38 (90.5%) 19 (100%) 0.300 61 Antibiotics 56 (91.8%) 39 (92.9%) 17 (89.5%) 0.643 61 Hydroxychloroquine 50 (82.0%) 36 (85.7%) 14 (73.7%) 0.294 61 Tocilizumab 50 (82.0%) 36 (85.7%) 14 (73.7%) 0.294 61 Remdesivir 3 (4.92%) 0 (0.00%) 3 (15.8%) 0.027 61 Lopinavir/ritonavir 50 (82.0%) 36 (85.7%) 14 (73.7%) 0.294 61 Corticosteroids 59 (98.3%) 41 (100%) 18 (94.7%) 0.317 60 __________________________________________________________________ Pulmonary coinfections during ICU stay __________________________________________________________________ Influenza virus infection 1 (1.64%) 1 (2.38%) 0 (0.0%) 1.000 61 Respiratory syncytial virus infection 0 (0.0%) 0 (0.0%) 0 (0.0%) – 61 Bacterial infection 52 (85.2%) 35 (83.3%) 17 (89.5%) 0.195 61 Viral pneumonia different from SARS-CoV-2 0 (0.0%) 0 (0.0%) 0 (0.0%) – 61 Bacterial pneumonia 13 (21.3%) 6 (14.3%) 7 (36.8%) 0.088 61 [102]Open in a new tab Continuous variables are expressed as the median [P25; P75] or mean (SD), and categorical variables are expressed as n (%). BAS, bronchial aspirate; BMI, body mass index; FiO[2], fraction of inspired oxygen; ICU, intensive care unit; IMV, invasive mechanical ventilation; PaCO[2], carbon dioxide partial pressure; PaO[2], oxygen partial pressure. p values <0.05 are presented in bold. We compared the miRNA ratios between survivors and nonsurvivors during the ICU stay. miRNA ratios with a fold change (FC) exceeding 1.5 (or lower than 0.67 for downregulated levels) and a statistically significant difference (false discovery rate [FDR] < 0.10) between the study groups were considered differentially expressed ([103]Figure 2A). Univariate analysis revealed differential expression levels of two miRNAs: miR-34c-5p/miR-34a-5p (FC = 0.528; FDR = 0.059) and miR-34c-5p/miR-125b-5p (FC = 0.478; FDR = 0.065) ([104]Figures 2B; [105]Table S5). These ratios were further examined. Initially, we assessed the individual miRNAs composing the selected ratios ([106]Figure 2C). miR-34c-5p was downregulated in nonsurvivors compared with survivors, whereas miR-34a-5p and miR-125b-5p remained stable between the groups. We then analyzed the relationship between the miRNA ratios and the risk of mortality. The miR-34c-5p/miR-34a-5p ratio exhibited a quasilinear dose-response relationship, whereas the miR-34c-5p/miR-125b-5p ratio displayed a nonlinear association with mortality risk ([107]Figure 2D). To further validate these findings, the levels of the three miRNAs that composed the selected ratios were analyzed in an external dataset including human postmortem lung biopsies (GEO: [108]GSE235130 ). As shown in [109]Figure 2E, miR-34c-5p was downregulated in COVID-19 samples compared with control samples. No differences were observed in either miR-34a-5p or miR-125b-5p. Figure 2. [110]Figure 2 [111]Open in a new tab MicroRNA profile of bronchial aspirate samples from survivors and nonsurvivors admitted to the ICU in the discovery cohort (n = 61) (A) Volcano plot showing the differences in the miRNA ratios between survivors and nonsurvivors in the discovery cohort (n = 61) after adjusting for age, sex, and batch. Each point represents a detected microRNA. The light green dots represent the significantly expressed microRNA ratios with a p value <0.05. Dark green dots represent the microRNA ratios with an FDR <0.10. The dashed lines are set at FDR = 0.10 and FC = 0.67 and 1.5. (B) Boxplot including bronchial aspirate (BAS) levels of candidate microRNA ratios between study groups. Each point represents a patient. (C) Boxplot showing the individual microRNAs that compose each differentially expressed microRNA ratio between survivors and nonsurvivors. Each point represents a patient. The gray lines connect both microRNAs for the same patient. (D) GAM modeling for risk mortality and the levels of candidate microRNA ratios. Effective degrees of freedom (edf) and p values are displayed. The odds ratio represents the risk change per 1 SD in continuous predictors. (E) Volcano plot showing differential microRNA expression in a human postmortem lung biopsy RNA-seq dataset (GEO: [112]GSE235130 ) between COVID-19 patients and non-COVID-19 patients. The differential expression criteria were set at FC ≥ 1.5 and FDR <0.05. Individual microRNAs of candidate microRNA ratios are labeled. (F) Boxplot of the candidate microRNA ratio in BAS samples measured in tracheal aspirate (TA) samples. (G) Boxplot of the candidate microRNA ratios in BAS samples measured in bronchoalveolar lavage fluid (BALF) samples. Ratios are represented as ΔCq values, calculated as the difference between the Cq of the numerator miRNA and the Cq of the denominator miRNA in each ratio (ΔCq = Cq[numerator]−Cq[denominator]). Since the miRNAs that comprised these ratios were expressed in all respiratory biospecimens, we investigated whether the selected miRNAs exhibited similar behavior in other respiratory samples. To this end, the expression of miR-34c-5p/miR-34a-5p and miR-34c-5p/miR-125b-5p was evaluated in TA and BALF samples. No significant differences between survivors and nonsurvivors were observed in either TA ([113]Figure 2F) or BALF ([114]Figure 2G). Next, we evaluated the potential of the miRNA ratios to be used as biomarkers. First, we established a cutoff point for fitted mortality risk using the maximally selected log rank statistic. K-M curves ([115]Figure 3A) revealed that patients with lower levels of miR-34c-5p/miR-34a-5p (hazard ratio [HR] = 0.18; p value = 0.0006) and miR-34c-5p/miR-125b-5p (HR = 0.17; p value = 0.0003) were at a significantly greater risk of a fatal outcome in the discovery phase. Second, we evaluated the associations between both ratios and mortality risk in the test phase ([116]Table S6). As shown in [117]Figure 3B, the miR-34c-5p/miR-34a-5p ratio was validated as a prognostic biomarker (HR = 0.17; p value = 0.045), whereas the miR-34c-5p/miR-125b-5p ratio did not reach statistical significance (HR = 0.27; p value = 0.140). Third, we analyzed miR-34c-5p/miR-34a-5p as a prognostic biomarker in combined discovery and test cohorts (n = 93). The demographic, clinical and biochemical data as a function of the median miR-34c-5p/miR-34a-5p ratio are detailed in [118]Table 2. No significant associations with the characteristics of the study population were observed, except for PaO[2] at IMV initiation. Additionally, we compared the discrimination potential for mortality risk of the ratio with that of clinical information. As displayed in [119]Figure 3C, miR-34c-5p/miR-34a-5p had the highest C-index, with significant overlap with the demographic, clinical and biochemical data. Given that several ratios close to significance included miR-34b-5p, we assessed whether combining miR-34b-5p with miR-34c-5p/miR-34a-5p could increase the predictive accuracy of fatal outcomes. This combined marker demonstrated comparable discriminatory ability to that of miR-34c-5p/miR-34a-5p ([120]Figure S1). Figure 3. [121]Figure 3 [122]Open in a new tab Biomarker potential of candidate microRNA ratios of bronchial aspirate samples to discriminate survivors and nonsurvivors admitted to the ICU (A) Kaplan-Meier estimations of candidate microRNA ratios in the discovery cohort (n = 61). The cutoff values, hazard ratios (HRs), and p values are displayed. (B) Kaplan-Meier estimations for the candidate microRNA ratio in the test population (n = 32). The cutoff, hazard ratio (HR), and p value are displayed. (C) Discrimination value for the miR-34c-5p/miR-34a-5p ratio and the main clinical predictors in the whole population (n = 93). The data are presented as the C-index. Clinical predictors are divided according to two temporal points: intensive care unit (ICU) admission and invasive mechanical ventilation (IMV) initiation. Table 2. Characteristics of the whole study population (n = 93) categorized according to the median miR-34c-5p/miR-34a-5p ratio All (n = 93) ≤ Median (n = 47) > Median (n = 46) p value n Sociodemographic characteristics __________________________________________________________________ Age (years) 62.0 [54.0;67.0] 62.0 [53.5;68.0] 63.0 [54.2;67.0] 0.954 93 Female 30 (32.3%) 19 (40.4%) 11 (23.9%) 0.139 93 BMI (kg/m^2) 30.9 [28.4;36.0] 32.3 [29.9;38.0] 30.1 [28.0;34.7] 0.107 93 __________________________________________________________________ Comorbidities __________________________________________________________________ Smoking history: – – – 0.485 90 Former 43 (47.8%) 21 (45.7%) 22 (50.0%) – – Nonsmoker 43 (47.8%) 24 (52.2%) 19 (43.2%) – – Active smoker 4 (4.44%) 1 (2.17%) 3 (6.82%) – – Hypertension 53 (57.0%) 28 (59.6%) 25 (54.3%) 0.765 93 Type II diabetes mellitus 28 (30.1%) 16 (34.0%) 12 (26.1%) 0.542 93 Chronic cardiac disease 14 (15.1%) 10 (21.3%) 4 (8.70%) 0.160 93 Chronic respiratory disease 4 (4.30%) 4 (8.51%) 0 (0.00%) 0.117 93 Chronic kidney disease 4 (4.30%) 3 (6.38%) 1 (2.17%) 0.617 93 __________________________________________________________________ Disease chronology __________________________________________________________________ Time since ICU admission to BAS sampling (days) 5.00 [3.00;6.00] 4.00 [3.00;6.00] 5.00 [3.25;7.00] 0.125 93 Time since IMV initiation to BAS sampling (days) 3.00 [2.00;5.00] 3.00 [1.50;4.00] 3.00 [2.00;5.00] 0.467 93 Time since symptom onset to BAS sampling (days) 14.1 (5.14) 13.8 (5.51) 14.4 (4.77) 0.5 59 __________________________________________________________________ Arterial blood gases and laboratory parameters at IMV initiation __________________________________________________________________ Oxygen saturation (%) 97.0 [93.8;99.0] 96.0 [92.2;98.0] 97.0 [95.2;99.0] 0.057 92 FiO[2] (%) 80.0 [63.8;100] 80.0 [65.0;97.5] 80.0 [60.0;100] 0.865 92 PaO[2] (mmHg) 94.0 [75.0;119] 83.0 [73.8;108] 104 [88.0;134] 0.025 69 PaCO[2] (mmHg) 46.5 [39.2;54.5] 46.0 [39.0;50.5] 51.0 [40.0;57.5] 0.207 66 PaO[2]/FiO[2] 135 [101;163] 127 [87.0;155] 141 [123;170] 0.112 67 Glucose (mg/dL) 149 [112;218] 154 [119;222] 144 [110;180] 0.490 89 Creatinine (mg/dL) 0.77 [0.58;0.94] 0.76 [0.55;0.87] 0.80 [0.64;0.97] 0.434 87 C-reactive protein (mg/L) 62.0 [20.8;143] 87.2 [23.2;139] 55.0 [16.5;148] 0.638 85 D-dimer (ng/mL) 2.81 [2.51;3.26] 2.68 [2.46;3.07] 2.91 [2.54;3.37] 0.141 70 Leukocyte count (x10^9/L) 8.28 [5.41;13.1] 7.42 [5.41;12.4] 10.6 [5.98;14.5] 0.141 90 Neutrophil count (x10^9/L) 7.28 [4.25;11.5] 6.34 [4.25;10.2] 9.02 [4.86;12.5] 0.109 90 Lymphocyte count (x10^9/L) 0.69 [0.39;0.96] 0.72 [0.39;0.91] 0.68 [0.40;0.96] 0.840 66 Monocyte count (x10^9/L) 0.38 [0.19;0.52] 0.32 [0.15;0.50] 0.40 [0.24;0.64] 0.175 90 Platelet count (x10^9/L) 238 [178;304] 234 [166;292] 242 [196;309] 0.438 90 __________________________________________________________________ Respiratory support and treatments during ICU stay __________________________________________________________________ High-flow oxygen nasal cannula 86 (92.5%) 43 (91.5%) 43 (93.5%) 1.000 93 Noninvasive mechanical ventilation 77 (82.8%) 40 (85.1%) 37 (80.4%) 0.747 93 Invasive mechanical ventilation duration (days) 13.0 [8.00;25.0] 13.0 [8.00;19.2] 14.0 [8.00;27.0] 0.214 92 Prone positioning 84 (91.3%) 42 (91.3%) 42 (91.3%) 1.000 92 Antibiotics 86 (92.5%) 44 (93.6%) 42 (91.3%) 0.714 93 Hydroxychloroquine 78 (83.9%) 39 (83.0%) 39 (84.8%) 1.000 93 Tocilizumab 78 (83.9%) 39 (83.0%) 39 (84.8%) 1.000 93 Remdesivir 3 (3.23%) 2 (4.26%) 1 (2.17%) 1.000 93 Lopinavir/ritonavir 78 (83.9%) 39 (83.0%) 39 (84.8%) 1.000 93 Corticosteroids 90 (97.8%) 45 (95.7%) 45 (100%) 0.495 92 __________________________________________________________________ Additional pulmonary coinfections during ICU admission __________________________________________________________________ Influenza virus infection 2 (2.15%) 2 (4.26%) 0 (0.00%) 0.601 93 Respiratory syncytial virus infection 0 (0.00%) 0 (0.00%) 0 (0.00%) – 93 Bacterial infection 80 (86.0%) 40 (85.1%) 40 (87.0%) 0.759 93 Viral pneumonia different from SARS-CoV-2 0 (0.00%) 0 (0.00%) 0 (0.00%) – 93 Bacterial pneumonia 15 (16.1%) 8 (17.0%) 7 (15.2%) 1.000 93 [123]Open in a new tab Continuous variables are expressed as the median [P25; P75] or mean (SD), and categorical variables are expressed as n (%). BAS, bronchial aspirate; BMI, body mass index; FiO[2], fraction of inspired oxygen; ICU, intensive care unit; IMV, invasive mechanical ventilation; PaCO[2], carbon dioxide partial pressure; PaO[2], oxygen partial pressure. p values <0.05 are presented in bold. Functional analysis To elucidate the hierarchical functions of the downregulated miRNAs, miR-34c-5p and its cotranscribed candidate miR-34b-5p (with several ratios approaching statistical significance) ([124]Table S5), in gene regulatory networks, an enrichment analysis of miRNA targets was conducted. To this end, we explored the predicted miRNA-gene interactions using TargetScan Release 8.0, which yielded 755 and 4,105 conserved predicted targets for miR-34c-5p and miR-34b-5p, respectively. The output represented a gene union of 4,461 target genes. The gene expression profile data of the GEO: [125]GSE210223 (human bronchial epithelial [HBE] cellular culture) and GEO: [126]GSE147507 (human postmortem lung biopsy) datasets were used to screen for miRNA targets in SARS-CoV-2 infection ([127]Figures 4A and 4B). From both datasets, we identified seven common upregulated transcripts: carbonic anhydrase XII (CA12), epithelial stromal interaction 1 (EPSTI1), guanylate binding protein 4 (GBP4), interferon-induced protein 44-like (IFI44L), interferon-induced protein with tetratricopeptide repeats 3 (IFIT3), 2′-5′-oligoadenylate synthetase 2 (OAS2), and XIAP-associated factor 1 (XAF1). Figure 4. [128]Figure 4 [129]Open in a new tab Functional analysis of miR-34c-5p and miR-34b-5p targets (A) Volcano plot representing the differential expression of miR-34b-5p and miR-34c-5p targets between SARS-CoV-2-infected and noninfected samples in an external dataset (GEO: [130]GSE210223 ) of human bronchial epithelial (HBE) cell cultures. The dashed lines are set at FDR = 0.05 and FC = 0.67 and 1.5. Each dot represents a target. The green dots represent the differentially expressed targets. The common targets of the GEO: [131]GSE210223 and GEO: [132]GSE147507 datasets are labeled. The analysis was based on the intersection (I = 2) of the total number of target genes of miR-34c-5p and miR-34b-5p. (B) Volcano plot representing the differential expression of miR-34b-5p and miR-34c-5p targets between SARS-CoV-2-infected and noninfected samples in an external dataset (GEO: [133]GSE147507) of human postmortem lung biopsies. The dashed lines are set at FDR = 0.05 and FC = 0.67 and 1.5. Each dot represents a target. The green dots represent the differentially expressed targets. The common targets of the GEO: [134]GSE210223 and GEO: [135]GSE147507 datasets are labeled. The analysis was based on the intersection (I = 2) of the total number of target genes of miR-34c-5p and miR-34b-5p. (C) STRING protein-protein interaction network. The analysis included the seven upregulated transcripts common in the GEO: [136]GSE210223 and GEO: [137]GSE147507 external databases. Edges indicate both physical and functional associations (interaction score cutoff is set at 0.90). The US Food and Drug Administration (FDA)-approved drugs for the selected targets are displayed. (D) Functional analysis of the top 25 terms ranked by the FDR from the gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Reactome databases. The Rich factor is the ratio of differentially expressed gene numbers annotated in this pathway term to all gene numbers annotated in this pathway term. (E) Enrichment analysis of lung cell types on the basis of single-cell RNA-seq data from the GTEx Project database. Each column shows a cell type, and each row shows a gene. The point size indicates the number of cells where the gene was detected, and the color represents the expression level. PPI analysis revealed that these targets interact to form a high-confidence interaction network ([138]Figure 4C). Gene ontology (GO) biological process, Reactome, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses elucidated the underlying mechanisms of the seven upregulated genes in COVID-19 ([139]Figure 4D). GO functional analyses revealed that the biological processes associated with the seven targets were associated mainly with host defense against pathogens, including viruses. The results of the Reactome and KEGG pathway analyses also support these findings, revealing enrichment of the interferon (IFN) type I alpha/beta and type II gamma signaling pathways and the NOD-like receptor signaling pathway, which are involved in viral defense and the immune response. Additionally, the enrichment of target genes revealed enrichment in processes related to metabolism, including the nitrogen (N) and carbon dioxide (CO[2]) metabolism signaling pathways, as well as chloride (Cl^−) ion homeostasis. An enrichment analysis of lung cell types on the basis of single-cell RNA sequencing (scRNA-seq) data from the Genotype-Tissue Expression (GTEx) Project database revealed that this gene functional module is constitutively expressed in lung cell populations ([140]Figure 4E). Notably, the seven genes presented high expression levels in immune cells, including alveolar macrophages, B cells, dendritic cells, and macrophages. To assess the potential for modifying this functional gene module, drug repositioning was evaluated using the DGIdb to identify existing marketed drugs that could be repurposed for the targets. A total of five US Food and Drug Administration (FDA)-approved drugs (acetazolamide sodium, acetazolamide, ethoxzolamide, dichlorphenamide, and sulthiame) were identified as inhibitors of CA12 ([141]Figure 4C). No drugs were identified for the remaining targets. Discussion This study addresses a significant gap in existing research by evaluating the reproducibility and specificity of miRNA patterns in respiratory biospecimens and exploring their potential for use in molecular phenotyping. In a clinical setting involving COVID-19 ICU patients on mechanical ventilation with severe pulmonary disease from four tertiary hospitals in Spain, we first validated the reproducibility of miRNA quantification in BAS samples. We then demonstrated that BAS, TA, and BALF exhibit distinct miRNA expression patterns. Additionally, we established that miRNA profiles in BAS samples serve as biomarkers and identified mechanistic pathways associated with the severity of the disease and potential therapeutic targets. To our knowledge, this study represents the most comprehensive investigation of molecular phenotyping in BAS to date and significantly advances the use of this biospecimen for miRNA profiling beyond our previous findings ([142]Figure 5).[143]^13 Figure 5. [144]Figure 5 [145]Open in a new tab Study design and main findings The incorporation of biological markers into the clinic demands standardization and reproducibility, which is a challenge for miRNA-based biomarkers due to significant technical variability, particularly during the preanalytical and analytical stages.[146]^14 Given the limited number of reports on the detection of miRNAs in BAS samples, our initial focus was on establishing technical reproducibility. We observed considerable variability in specific miRNAs across independent aliquots from the same patient, evaluated at different time periods, especially in those with low expression levels. These results highlight the need for rigorous quality control measures, including spike-ins during quantification, and suggest prioritizing miRNAs with robust expression levels to increase the reliability of the results. Accordingly, we identified a set of miRNAs with consistently acceptable expression levels and high reproducibility across samples. To test whether the miRNA expression profiles are comparable across biospecimens from different areas of the respiratory tract, we quantified the miRNAs studied in the BAS samples in both the TA and BALF samples. Our investigation revealed a discrepancy between the miRNA expression profiles of BAS, TA, and BALF samples. While all the miRNAs analyzed were detected in BAS, several miRNAs were absent in TA and BALF samples. Notably, the miRNA ratio with potential as a biomarker in BAS samples, the miR-34c-5p/miR-34a-5p ratio, lacked prognostic value in the TA or BALF. This finding supports the notion that miRNA-based biomarkers may exhibit strong biospecimen specificity and that results obtained in one respiratory fluid cannot be directly extrapolated to others. Such discrepancies likely reflect differences in cellular composition, extracellular milieu, and spatial distribution of inflammatory or infectious processes across various compartments of the respiratory tract.[147]^15^,[148]^16 Indeed, although lower respiratory tract fluids may share clinical indications, their molecular content and diagnostic yield can differ, with implications for clinical decision-making and therapeutic guidance.[149]^17^,[150]^18 Additionally, our results are consistent with those of previous studies that demonstrated specific patterns of extracellular miRNAs and other small RNAs between different biofluids.[151]^19 Indeed, we detected miR-34b-5p and miR-34c-5p in respiratory biospecimens but not in plasma samples from critically ill COVID-19 patients.[152]^20 These variations may be attributed to the cell and tissue expression profiles of both miRNAs, which are enriched in lung tissues.[153]^21 Overall, the current findings suggest that careful consideration of biospecimens is crucial for selecting appropriate miRNA candidates. To further explore the clinical utility of miRNA profiling in BAS, we focused on evaluating the prognostic potential of miRNA ratios. Our decision to prioritize ratio-based analyses over individual miRNA expression levels was driven by both technical and biological considerations related to the characteristics of BAS biospecimens, i.e., their high heterogeneity. miRNA ratios offer a normalization strategy that inherently corrects for sample-to-sample variability, thereby improving reproducibility and enhancing the biological interpretability of the data.[154]^22 Our findings align with those of previous studies that support the potential of miRNA-based biomarkers across various biospecimens, including respiratory samples.[155]^23 In the discovery phase, we identified significant differences in two miRNA ratios, miR-34c-5p/miR-34a-5p and miR-34c-5p/miR-125b-5p, between ICU survivors and nonsurvivors. Subsequent analysis not only confirmed these differences in miR-34c-5p/miR-34a-5p levels but also demonstrated a quasilinear association between this ratio and mortality risk. The HRs were consistent across both the discovery (HR = 0.18) and test phases (HR = 0.17), further reinforcing the reproducibility of our findings. The analysis of the ratio composition indicated that a reduction in the level of miR-34c-5p relative to that of miR-34a-5p is associated with increased mortality risk. Decomposition of the ratio revealed that the prognostic signal was primarily driven by reduced miR-34c-5p levels in nonsurvivors, while miR-34a-5p remained relatively stable. In a comprehensive analysis of the entire study cohort, the miR-34c-5p/miR-34a-5p ratio showed prognostic performance comparable to conventional clinical variables. Interestingly, when the cohort was stratified by the median value of this ratio, no significant associations emerged with key demographic, clinical, or biochemical parameters, except for arterial oxygenation. This limited overlap suggests that the miRNA ratio captures distinct and complementary biological information not reflected by standard clinical indicators. Rather than replacing existing risk markers, the miR-34c-5p/miR-34a-5p ratio may augment current prognostic frameworks, offering a host-derived molecular signal that reflects disease processes in the lower respiratory tract. This suggests its potential utility as part of a multimodal decision-making tool, integrating molecular and clinical data to improve outcome prediction in COVID-19 patients undergoing IMV. The concept of combining miRNAs with other biomarker classes has already been proposed as a strategy to improve diagnostic and prognostic accuracy through a more comprehensive characterization of patient status.[156]^24 Several ratios that included reductions in miR-34b-5p approached statistical significance. Since miR-34b-5p and miR-34c-5p are cotranscribed from a bicistronic transcript on chromosome 11, their similar behavior aligns with our findings. Unfortunately, the correlation between miR-34b-5p and miR-34c-5p impacts their potential as biomarkers, as both miRNAs provide similar information; thus, combining them did not increase the predictive power beyond that of individual miRNAs. In contrast, miR-34a-5p, which is part of the optimal ratio identified, is located on chromosome 1. This chromosomal difference may explain its distinct behavior compared with that of miR-34b-5p and miR-34c-5p in terms of the ratios. Although mechanistic interpretation was not the primary focus of this study, our findings suggest that miRNA profiling in BAS may also provide useful mechanistic information. In this sense, the downregulation of the expression of miR-34b-5p and miR-34c-5p aligns with the host cellular and molecular mechanisms linked to adverse outcomes in mechanically ventilated patients with COVID-19. The upregulated miRNA targets constitute a gene functional module that forms an interaction network and displays key mechanistic pathways in COVID-19. Pathway enrichment analysis of the seven genes revealed that the IFN type I and II signaling pathways, which have antiviral and immunomodulatory effects, enhance immunity against infection or reinfection.[157]^25^,[158]^26 The enrichment of the NOD-like receptor signaling pathway, which is responsible for mediating the initial innate immune response to cellular injury and stress, has also been reported.[159]^27 Additionally, the results revealed enrichment in metabolic processes, including the CO[2] metabolism signaling pathway. Patients affected by severe COVID-19 exhibit a dysregulated acid-base status, potentially influenced by the activity of carbonic anhydrase (CA), which is markedly elevated in patients with SARS-CoV-2 infection.[160]^28 Therefore, conventional carbonic anhydrase inhibitors (CAIs), such as the retrieved acetazolamide drug, which reduces CA enzyme activity, may warrant consideration as adjunctive pharmacological interventions. In addition, these results may have implications beyond critical COVID-19, indicating potential utility in other respiratory diseases and critical illnesses. Other viral respiratory infections and lung conditions involve similar mechanisms of immune response and tissue damage. For example, reduced miR-34b/c-5p levels have been linked to increased CXCL10 secretion and macrophage chemotaxis in respiratory syncytial virus (RSV) infection.[161]^29 Similarly, RSV infection induces mucus secretion by downregulating miR-34b/c-5p expression in airway epithelial cells.[162]^30 The overexpression of miR-34c-5p has been proposed as a therapeutic strategy to inhibit silica-induced pulmonary fibrosis.[163]^31 While these findings are hypothesis-generating, they highlight the potential for innovative treatments for COVID-19 patients on IMV. Advances in miRNA mimics and inhibitors over the past decade could also lead to new therapeutic approaches. Recent research involving various animal models and patient samples has demonstrated that anti-miR-93-5p therapy enhances sepsis survival by modulating both innate and adaptive immunity.[164]^32 miRNA mimic therapies have demonstrated potential in the prevention or treatment of respiratory diseases such as bronchopulmonary dysplasia[165]^33 or idiopathic pulmonary fibrosis.[166]^34 The intratracheal administration of nucleic acid drugs represents a promising therapy for pulmonary diseases,[167]^35 including acute lung injury.[168]^36 Nonetheless, the safety of miRNA restoration therapies remains an area of ongoing debate.[169]^37 Our study has several limitations that should be considered. First, the variability introduced by collecting BAS, TA, and BALF samples from different patients and time points may affect comparative analyses. We addressed the significant heterogeneity in the nomenclature of respiratory biospecimens by detailing our collection processes to increase the reproducibility of our findings. Second, our focus on fatal outcomes in miRNA biomarker assessment excludes other clinically relevant endpoints. Additionally, our study specifically targeted critically ill COVID-19 patients on IMV. Third, we evaluated a predefined panel of miRNAs previously associated with potentially altered molecular pathways in COVID-19; other miRNA ratios may provide additional clinical insights. Fourth, the primary aim of the functional analysis was not to establish mechanistic causality, but rather to provide biological context for the miRNAs identified in BAS samples. By integrating predicted miRNA-mRNA interactions with publicly available datasets from lung tissues and airway epithelial cells, we aimed to highlight plausible molecular pathways potentially influenced by these miRNAs. However, to confirm these interactions and elucidate their functional consequences, dedicated in vitro and in vivo studies will be required. Fifth, BAS is a complex biospecimen containing a heterogeneous mix of cells from across the lower respiratory tract, which complicates the differentiation between airway and alveolar pathogenetic processes. Furthermore, the identified miRNAs are not cell specific, which makes it challenging to determine their precise biological origin. In conclusion, our study offers valuable insights into the clinical potential of miRNA profiling in respiratory biosamples. miRNA analysis of BAS opens new opportunities for developing biomarkers, exploring novel pathophysiological mechanisms and identifying new lung-targeted therapeutic approaches. Continued research in this field could significantly impact clinical decision-making. Materials and methods Study design and population We conducted a multicenter and ambispective study, including BAS samples from 150 critically ill COVID-19 adults undergoing IMV admitted to ICUs between March 2020 and January 2022 at the Hospital Arnau de Vilanova (Lleida, Spain), the Hospital Universitario Son Espases (Palma, Spain), and Hospital Son Llatzer (Palma, Spain). The inclusion criteria included patients aged 18 years or older who had a positive nasopharyngeal swab RT-qPCR test result for SARS-CoV-2, who were diagnosed with acute respiratory distress syndrome (ARDS) according to the Berlin definition[170]^38 secondary to COVID-19, and who were receiving IMV via orotracheal intubation or tracheostomy while hospitalized in the ICU. BAS samples were obtained with the support of the IRBLleida Biobank (B.000682), Biomodels Platform ISCIII PT23/00032, IdISBa Biobank B.0000527 ([171]www.idisba.es), and CIBERES Pulmonary Biobank Consortium B.0000471, a network formed by 12 tertiary Spanish hospitals ([172]www.ciberes.org) and integrated into the Spanish National Biobanks Network. The samples were processed following standard operating procedures with the appropriate approval of the Ethics and Scientific Committees (CEIC 2273, s007-BBCOV). To compare the miRNA patterns across respiratory biospecimens, two additional sets of respiratory samples were incorporated into the study: (1) TA samples were collected from 42 critically ill COVID-19 patients undergoing IMV who were admitted to the ICU between March 2020 and January 2021 at the Complejo Hospitalario Universitario A Coruña (A Coruña, Spain). Samples and data were provided by the Biobanco A Coruña—Área Sanitaria A Coruña y Cee—SERGAS—INIBIC (B.0000796, PT17/0015/0032) and integrated into the Spanish National Biobanks Network with the appropriate approval of the Ethics and Scientific Committees; (2) BALF samples were collected from 96 critically ill COVID-19 patients on IMV who were admitted to the ICU between March 2020 and February 2022 at the Hospital Universitario de Jerez (Jerez de la Frontera, Spain). The study was approved by the Ethics Research Committee of Cádiz (SAM-COVUCI-2020-01). Comprehensive demographic, clinical and pharmacological data, as well as information regarding hospital and ICU stays and outcomes, were manually extracted from electronic medical records by specialized clinical research assistants, as previously described.[173]^11^,[174]^13 The identities of the patients were disclosed only after the miRNA quantification and data preprocessing were completed. The studies were conducted in compliance with ethical requirements per the Declaration of Helsinki. Prior to the use of their biological samples and clinical information in the study, the participating patients or their relatives were informed about the research and provided written informed consent. Bronchial aspirate collection BAS samples were obtained through bronchoscopy during the ICU stay and after the initiation of IMV using standardized procedures as described previously.[175]^13 Briefly, bronchoaspiration was conducted using a flexible bronchoscope equipped with an Ambu aScope 4 Broncho Large 5.8/2.8 bronchoscope (Ambu, Ballerup, Denmark) connected to a vacuum. Bronchial washing with a 0.9% saline solution (Braun, Melsungen, Germany) was performed when necessary. Samples were collected within the first 10 days following IMV initiation. BAS samples collected at the Hospital Arnau de Vilanova (Lleida, Spain) were aliquoted, immediately frozen, and stored at −80°C at the IRBLleida Biobank. The BAS aliquots from IdISBa Biobank B.0000527 and CIBERES Pulmonary Biobank Consortium B.0000471 were shipped frozen on dry ice to the Biomedical Research Institute of Lleida (Lleida, Spain) and stored at −80°C. Tracheal aspirate collection TA samples were obtained through endotracheal suctioning using an open secretion aspiration system. Following at least 1 minute of ensuring patient oxygenation, the invasive mechanical ventilator was disconnected, and a probe was inserted up to the tracheal carina in alignment with the endotracheal tube. The suction orifice was subsequently occluded, and the TA samples were intermittently suctioned for 10–15 s with gentle rotary movements while the probe was withdrawn. In cases where the procedure was repeated, the patient was oxygenated for at least 1 minute prior to continuing the procedure, and a new probe was utilized. Samples were collected within the first 18 days of intubation. TA samples were aliquoted, immediately frozen, and stored at −80°C. The frozen aliquots were shipped on dry ice to the Biomedical Research Institute of Lleida (Lleida, Spain). Bronchoalveolar lavage fluid collection Diagnostic fibrobronchoscopy was conducted within the initial 24 h of IMV, employing an Ambu aScope 4 Broncho Regular 5.0/2.2 bronchoscope (Ambu). BALF was obtained from 150 mL of a physiological saline solution and divided into three aliquots. The initial 20 mL of BALF was discarded to eliminate potential contaminants, and the remaining fluid was preserved for microbiological analysis. To prevent aerosol generation, a specialized adapter valve was utilized during bronchoscopy. The procedure was performed in pressure control ventilation mode while maintaining optimal previous end-expiratory pressure (PEEP) levels and a fraction of inspired oxygen of 100%. The patient was not repositioned to the supine position during the diagnostic procedure. The continuous infusion sedative used to sedate the patient was not changed. BALF samples were aliquoted, immediately frozen, and stored at −80°C. The frozen aliquots were shipped on dry ice to the Biomedical Research Institute of Lleida (Lleida, Spain). microRNA quantification miRNA quantification was centrally conducted at the Biomedical Research Institute of Lleida (Lleida, Spain) in a controlled laboratory environment by trained personnel. All the procedures utilized RNase-free and DNase-free reagents and materials. Respiratory biospecimens (BAS, TA, or BALF) were thawed at room temperature before 100 μL was collected for total RNA extraction using the miRNeasy Mini Kit (Qiagen, Hilden, Germany). To monitor RNA isolation, the Caenorhabditis elegans miRNA cel-miR-39-3p (1.6 × 10^8 copies/μL) (Qiagen) and the RNA Spike-In Kit (synthetic UniSp2, UniSp4, and UniSp5) (Qiagen) were added. Additionally, 1 μg of MS2 carrier RNA (Roche, Merck, Darmstadt, Germany) was used to increase the extracellular RNA yield. All reagents were spiked into samples during RNA isolation after incubation with the denaturing solution. The RNA was eluted into 30 μL of nuclease-free water and stored at −80°C. Reverse transcription (RT) was conducted using the miRCURY LNA RT Kit (Qiagen), with UniSp6 included as an RT control. The RT reactions were carried out in a 10 μL volume under the following conditions: 60 min at 42°C, followed by 5 min at 95°C and immediate cooling to 4°C. The resulting cDNA was stored at −20°C. Eighteen candidate miRNAs were selected based on prior results from a pilot study conducted by our group, which identified multiple miRNA candidates in BAS samples associated with ICU mortality in critically ill COVID-19 patients.[176]^13 The selection of miRNAs was based on two criteria: (1) miRNA ratios ranked by area under the curve (AUC), applying a p value threshold of < 0.10 and ensuring that each miRNA appeared only once across different ratios to avoid redundancy and (2) miRNA ratios were filtered by selecting those with an AUC ≥ 0.70 ([177]Table S7). The final set of selected miRNAs corresponded to the individual components of these optimized ratios. miRNAs were quantified via qPCR using the miRCURY LNA SYBR Green PCR Kit (Qiagen) in 384-well miRCURY LNA miRNA Custom PCR Panels (Qiagen) with precoated primers, according to the manufacturer’s protocols. Synthetic UniSp3 served as an interplate calibrator and qPCR control. The qPCR was performed using a QuantStudio 7 Flex Real-Time PCR System (Thermo Fisher Scientific, Massachusetts, United States) under the following conditions: 2 min at 95°C, 40 cycles of 10 s at 95°C, 60 s at 56°C, and, finally, melting curve analysis. qPCR data analysis and miRNA ratios QuantStudio Software v1.3 (Thermo Fisher Scientific) was used to determine the quantification cycle (Cq) and conduct melting curve analysis. The spike-in RNA templates were analyzed to monitor the RNA extraction procedure and the efficiency of RT-qPCR. Cq was defined as the fractional cycle number at which the fluorescence exceeded a given threshold. Cq values ≥ 35 cycles were considered undetectable and were censored at a Cq level of 35 cycles. Expression rate was defined as the percentage of samples in which each individual miRNA showed a detectable signal, with Cq values of 35 considered as the threshold for detection. To define reliably expressed miRNAs, a detectability threshold was set at 20%, i.e., a given miRNA had to be detected (Cq < 35) in at least 20% of samples. Normalization is crucial in miRNA quantification.[178]^39^,[179]^40 While exogenous oligonucleotides can monitor the efficiency of the analytical process, they do not account for sampling bias or sample quality. The use of endogenous controls is generally preferred, as their expression is influenced by the same variables affecting target miRNAs. However, no suitable endogenous controls have been established for respiratory biospecimens. To address this, we employed miRNA ratios, an easy-to-apply method that minimizes variability by avoiding the need for external references and reducing the impact of sample