Blood mNGS-Detected Epstein-Barr Virus in ICU Patients with Pneumonia: Associations with Disease Severity Markers and Inflammatory Burden Authors: Xia Wang, Wei-ping Hu, Yi-Xing Wu, Jian-Lan Hua, Zi-Xuan Cheng, Xiao Wang, Li Liu, Jing Zhang Published: 2026 ABSTRACT AbstractBackgroundEpstein-Barr virus (EBV) is frequently detected in critically ill patients, but the clinical meaning of EBV reported by blood metagenomic next-generation sequencing (mNGS) in ICU patients with pneumonia remains unclear.MethodsThis retrospective secondary analysis used data from a prospective multicenter cohort of patients with suspected sepsis. Data included demographics, clinical variables, radiological findings, blood culture and mNGS results, cytokine levels, and 30-day mortality.ResultsA total of 184 ICU patients with pneumonia were included and classified as EBV-positive (n=34) or EBV-negative (n=150) according to blood mNGS. EBV-positive patients had lower serum albumin and systolic blood pressure, higher qSOFA score distribution, more frequent sepsis at enrollment, higher C-reactive protein and procalcitonin levels, higher PaCO2, and lower serum calcium. Pneumocystis jirovecii was numerically more frequent in EBV-positive patients but was not independently associated with EBV positivity after adjustment. In exploratory cytokine analysis, EBV-positive patients showed higher TRAIL levels and more frequent IL-17 detection. In an exploratory multivariable model, lower serum albumin and sepsis at enrollment remained associated with EBV positivity. 30-day mortality did not differ significantly between groups.ConclusionIn ICU patients with pneumonia, blood mNGS-detected EBV was associated with markers of disease severity and inflammatory burden. However, these associations do not establish clinically significant EBV reactivation, and no statistically significant difference in 30-day mortality was observed between groups.Keywords: Epstein-Barr virus, metagenomic next-generation sequencing, pneumonia, intensive care unit, sepsis Introduction Pneumonia severity reflects microbial infection, host inflammatory responses, organ dysfunction, and infection-related immune dysregulation. In this setting, latent herpesviruses, including Epstein-Barr virus (EBV), are frequently detected and are associated with greater illness severity in sepsis and severe respiratory infections.1–3 EBV is a ubiquitous herpesvirus that infects approximately 90–95% of adults and establishes lifelong latency, mainly in memory B cells.4,5 In this context, EBV detected by metagenomic next-generation sequencing (mNGS) from anticoagulated whole blood may reflect latent or cell-associated viral DNA, low-level circulating EBV DNA, bystander DNA release during critical illness, or true viral reactivation. Low-read-count detections warrant particular caution because mNGS sequence reads are not validated quantitative measures of EBV viral load. Without quantitative EBV DNA testing, serial viral load monitoring, serologic evidence, or compatible EBV-related organ disease, blood mNGS detection alone cannot define clinically significant EBV reactivation.6 mNGS is increasingly used for pathogen detection in pneumonia, especially when conventional tests are inconclusive or mixed infections are suspected.7,8 However, EBV reported by blood mNGS is difficult to interpret clinically: an isolated signal may not indicate respiratory tract EBV infection and should not, by itself, establish EBV disease or prompt EBV-directed treatment without quantitative or organ-specific corroboration. Nevertheless, it may still reflect disease severity, immune dysregulation, and systemic inflammation.6,9 The clinical phenotype, co-detected pathogens, inflammatory profile, and short-term outcomes associated with mNGS-detected EBV in pneumonia remain poorly defined. This study was a retrospective secondary analysis of a prospective multicenter cohort. We aimed to characterize the clinical, microbiological, inflammatory, and outcome features associated with mNGS-detected EBV in ICU patients with pneumonia. We focused on disease severity markers and inflammatory burden, while analyses of co-detected pathogens, cytokine profiles, and 30-day mortality were considered exploratory. Materials and Methods Study Design and Population This study was a retrospective secondary analysis of a prospective multicenter cohort of patients with suspected sepsis. A total of 277 critically ill patients with severe infections were consecutively recruited from respiratory departments and ICUs of nine Chinese tertiary hospitals between December 2018 and December 2019. Of these, 184 patients admitted to the ICU and diagnosed with pneumonia were included in this study. Accordingly, the sample size was determined by the number of eligible ICU patients with pneumonia in the parent cohort; no a priori sample-size calculation was performed. The inclusion criteria were as follows: (1) age >18 and ≤90 years; (2) ICU admission; and (3) diagnosis of pneumonia. The diagnostic criteria for pneumonia were based on the American Thoracic Society (ATS)/Infectious Diseases Society of America (IDSA) guidelines for HAP/VAP and CAP, as well as the Chinese guidelines for HAP/VAP and CAP (2018 edition).10–12 A pneumonia diagnosis required both ① new or progressive pulmonary infiltrates on imaging and ② at least two of the following clinical features: a body temperature >38°C, a peripheral white blood cell count >10 × 109/L or <4 × 109/L, and purulent airway secretion. Exclusion criteria included: (1) patients aged <18 or >90 years; (2) patients unable to complete study observation or with cognitive impairment. Withdrawal criteria included participants who withdrew informed consent. For this study, patients with blood mNGS-detected EBV were classified as EBV-positive, whereas those without EBV detection were classified as EBV-negative. Clinical data and survival outcomes after a 30-day follow-up were collected. Ethics Approval The prospective cohort study, on which the current retrospective secondary analysis was based, was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Zhongshan Hospital Affiliated to Fudan University (Ethical Approval Number: B2018-182R). All participants were informed about the study before recruitment and provided written informed consent. Personal information was protected during data collection and analysis. Data Collection Clinical data were extracted from the prospective multicenter suspected sepsis cohort by one researcher and independently checked by another. Discrepancies were resolved by review of the original data. The clinical data collected in this study were organized into the following categories: (1) Demographics: age, sex, height, and weight. (2) Clinical presentations: symptoms, signs, and past medical history. Sepsis was defined according to the Sepsis-3 criteria as life-threatening organ dysfunction caused by a dysregulated host response to infection.13 qSOFA was assessed using altered mentation, systolic blood pressure ≤100 mmHg, and respiratory rate ≥22/min. (3) Laboratory results (on the days of inclusion and during the diagnosis of pneumonia): complete blood count, inflammatory markers, liver function, renal function, serum electrolytes, blood gas measurement, myocardial markers, and coagulation function. (4) Imaging features: chest X-ray and chest computed tomography (CT). (5) Microbiological tests: blood culture, sputum culture and blood mNGS. (6) Cytokine immunoassay. (7) Survival status at day 30 of follow-up. The final diagnosis of the causative pathogen was determined by the attending physician, after reviewing all available clinical data. When the diagnosis was uncertain, three experienced clinicians in the study group reviewed and adjudicated the case. mNGS Detection Anticoagulated whole-blood samples were collected at enrollment and stored at −80 °C. Within 24 h, samples were sent to BGI Medical Laboratory Institute (Wuhan, China) for mNGS detection. The following criteria were used to define positive mNGS results. For bacteria (excluding mycobacteria), fungi, viruses, and parasites: a microbe detected by mNGS at the genus level was considered positive when its read count was at least three-fold higher than that of any other detected microbe. If only one microorganism was reported by the laboratory, it was considered positive after review of the clinical context. For mycobacteria: Mycobacterium tuberculosis was determined to be positive if there was more than one read mapped at the species or genus level. After these criteria were met, patient clinical characteristics were reviewed. Potential contaminants were systematically excluded by clinicians and laboratory staff. Remaining reported microorganisms were retained as mNGS-detected organisms for analysis. The mNGS report specified that detected microorganisms should be interpreted in combination with clinical findings and should not be used as the sole basis for the final clinical diagnosis. Therefore, all mNGS results were interpreted as microbiological findings rather than definitive evidence of causative infection.6,9,14 Determination of Inflammatory Cytokines Cytokine assays were performed in a subset of 48 patients with positive mNGS results, including 8 EBV-positive and 40 EBV-negative patients, with the Human Cytokine Screening 48-Plex Services Kit (Bio-Rad, catalog number 12007283; Hercules, CA, USA) according to the manufacturer’s instructions. Statistical Analysis Data were first summarized using standard descriptive statistics. Continuous variables with normal distribution were reported as means ± standard deviation, and non-normally distributed variables were reported as medians (interquartile range, IQR). Categorical variables were reported as n (%). Normality was assessed with the one-sample Kolmogorov–Smirnov test. Normally distributed continuous variables were compared with the t test, and non-normally distributed variables were compared with the Mann–Whitney U-test. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. Cytokine analyses were exploratory, and P values were not adjusted for multiple comparisons. Variables with more than 20% missing data were excluded from the regression analysis. Variables with P < 0.1 and immunosuppressive status were considered for inclusion in multivariable logistic regression analysis. Given the limited number of EBV-positive patients, the multivariable logistic regression was considered exploratory. Two-sided P < 0.05 was considered statistically significant. SPSS 26.0 software (SPSS Inc., Chicago, IL, USA) was used in all analyses. Results Demographic and Basic Clinical Information The cohort consisted of 184 ICU patients with pneumonia, including 34 EBV-positive and 150 EBV-negative patients based on blood mNGS results. Baseline demographic characteristics and major comorbidities were generally comparable between the two groups (Table 1). At enrollment, the EBV-positive group had lower serum albumin levels and systolic blood pressure (SBP) but a higher qSOFA score distribution and a higher proportion of sepsis than the EBV-negative group. In contrast, APACHE II scores and shock index did not differ significantly between groups. Table 1Baseline Characteristics and Disease Severity Markers of ICU Patients with Pneumonia Stratified by EBV PositivityEBV Positive (n=34)EBV Negative (n=150)P-valueBasic informationMale (%)29 (85.3%)107 (71.3%)0.129Age (years old)63.2±14.059.0±16.60.163Smoking history (%)17 (50.0%)59 (39.3%)0.335Nutrient statusBMI (kg/m2)21.6±2.822.1±3.80.495Serum albumin (g/L)27.1±7.031.3±6.5<0.001ComorbidityPulmonary disease5 (14.7%)21 (14.0%)1Recent surgery/trauma history3 (8.8%)30 (20.0%)0.145Hypertension9 (26.5%)25 (16.7%)0.220Diabetes7 (20.6%)28 (18.7%)0.797Cardiovascular and cerebrovascular diseases5 (14.7%)44 (29.3%)0.090Hepatic insufficiency14 (41.2%)57 (38.0%)0.731Renal insufficiency5 (14.7%)18 (12.0%)0.774Malignant tumor3 (8.8%)22 (14.7%)0.578Connective tissue disease2 (5.9%)6 (4.0%)0.627SeveritySBP (mmHg)107±27114±230.047APACHE II (score)18.6±8.616.3±8.60.120Shock index1.01±0.350.91±0.300.109qSOFA (score)0.04104 (11.7%)14 (9.3%)15 (14.7%)47 (31.3%)217 (50.0%)76 (50.7%)38 (23.5%)13 (8.7%)Sepsis27 (79.4%)83 (55.3%)0.011Note: Bold values indicate P < 0.05.Abbreviations: BMI, body mass index; SBP, systolic blood pressure; APACHE II, the Acute Physiologic and Chronic Health Evaluation II score. Baseline Characteristics and Disease Severity Markers of ICU Patients with Pneumonia Stratified by EBV Positivity Note: Bold values indicate P < 0.05. Abbreviations: BMI, body mass index; SBP, systolic blood pressure; APACHE II, the Acute Physiologic and Chronic Health Evaluation II score. Pre-enrollment treatment exposures were broadly similar between groups, including corticosteroid, immunosuppressive, cytotoxic chemotherapy, and overall antibiotic exposure (Table E1). But EBV-positive patients were less likely to have received arterial or venous catheterization (14.7% vs 35.5%, P = 0.024), tetracycline use (0% vs 18.0%, P = 0.005), and other antibiotic use (20.6% vs 41.3%, P = 0.024) before enrollment. Laboratory Tests, Symptoms, and Radiological Findings Laboratory results are summarized in Table 2. Compared with EBV-negative patients, EBV-positive patients had higher C-reactive protein (CRP) and procalcitonin (PCT) levels. EBV-positive patients also had lower serum calcium levels and higher PaCO2 levels. D-dimer levels were numerically higher in EBV-positive patients but did not reach statistical significance (P = 0.059). Other routine hematologic, hepatic, renal, and oxygenation-related indices were broadly comparable between groups. Table 2Laboratory Findings of ICU Patients with Pneumonia Stratified by EBV PositivityEBV Positive (n=34)EBV Negative (n=150)P-valueHemoglobin (g/L)101±3096±240.161Platelet (×10^9/L)167±135187±1220.209WBC (×10^9/L)12.0±7.611.3±9.50.497Neutrophil (×10^9/L)10.2±7.29.8±8.30.710Lymphocyte (×10^9/L)1.3±2.61.5±6.60.830NLR17.1±18.519.3±27.80.918Monocyte (×10^9/L)0.4±0.40.5±0.40.183CRP (mg/L)101.6±87.775.8±89.50.009PCT (ng/mL)12.5±23.36.3±13.90.013ALT (U/L)39.4±37.247.8±64.50.848AST (U/L)58.6±49.666.0±98.40.457Serum creatinine (μmol/L)111.5±125.9135.7±150.80.289D-dimer (mg/L)5.71±5.094.42±6.050.059Serum calcium (mmol/L)1.93±0.222.02±0.230.014PaO2(mmHg)88.0±37.8100.8±48.90.252PaCO2(mmHg)45.3±14.239.3±11.60.011pH7.40±0.097.43±0.070.151PaO2/FiO2205±174235±1620.252Note: Bold values indicate P < 0.05.Abbreviations: WBC, white blood cell; NLR, Neutrophil-to-lymphocyte ratio; CRP, C-reactive protein; PCT, procalcitonin; ALT, Alanine aminotransferase; AST, Aspartate transaminase; PaO2, Partial pressure of arterial oxygen; PaCO2, Partial pressure of arterial carbon dioxide. Laboratory Findings of ICU Patients with Pneumonia Stratified by EBV Positivity Note: Bold values indicate P < 0.05. Abbreviations: WBC, white blood cell; NLR, Neutrophil-to-lymphocyte ratio; CRP, C-reactive protein; PCT, procalcitonin; ALT, Alanine aminotransferase; AST, Aspartate transaminase; PaO2, Partial pressure of arterial oxygen; PaCO2, Partial pressure of arterial carbon dioxide. Clinical symptoms and radiological findings are shown in Tables E2 and E3. EBV-positive patients had higher frequencies of bloody sputum (11.7% vs 0.7%, P = 0.004) and hemoptysis (14.7% vs 4.7%, P = 0.048). On chest imaging, pulmonary cavitation was more frequent in EBV-positive patients than in EBV-negative patients (14.7% vs 2.1%, P = 0.002). Microbiological Findings Microbiological findings from blood mNGS and culture are summarized in Table 3. Overall mNGS detection of bacteria, fungi, and non-EBV viral co-detections did not differ significantly between EBV-positive and EBV-negative patients. Culture positivity, including blood culture positivity, was also comparable between the two groups. Table 3Microbiological Findings Detected by Blood mNGS and Culture Stratified by EBV PositivityPathogenEBV Positive (n=34)EBV Negative (n=150)OR95% CIP-valuemNGSBacteriaTotal14 (41.2%)54 (36.0%)1.240.58–2.660.57Gram positive3 (8.8%)20 (13.3%)0.630.18–2.250.58Gram negative9 (26.5%)36 (24.0%)1.140.49–2.670.76Mycobacteria1 (2.9%)1 (0.7%)4.510.28–74.100.34Anaerobic bacteria2 (5.9%)10 (6.7%)0.880.18–4.191Atypical pathogen0 (0)4 (2.7%)0.470.02–8.981FungiTotal8 (23.5%)21 (14%)1.890.76–4.730.17Candida1 (2.9%)6 (4.0%)0.730.08–6.251Aspergillus spp.3 (8.8%)10 (6.7%)1.360.35–5.220.71Pneumocystis jirovecii6 (17.6%)11 (7.3%)2.710.92–7.930.06VirusCMV10 (29.4%)39 (26%)1.190.52–2.700.68Other herpesviruses7 (20.6%)20 (13.3%)1.680.65–4.380.28Torque teno virus1 (2.9%)6 (4.0%)0.720.08–6.251Parvovirus0 (0)6 (4.0%)0.320.02–5.860.59No non-EBV organism detected11 (32.4%)33 (22%)1.70.75–3.840.2CultureTotalPositive16 (47.1%)86 (57.3%)0.660.31–1.400.28Fungi8 (23.5%)27 (18.0%)1.400.57–3.430.47BloodPositive9 (26.5%)29 (19.3%)1.50.63–3.560.35Abbreviation: CMV, cytomegalovirus. Microbiological Findings Detected by Blood mNGS and Culture Stratified by EBV Positivity Abbreviation: CMV, cytomegalovirus. Among individual organisms, Pneumocystis jirovecii (P. jirovecii) was numerically more frequent in EBV-positive patients than in EBV-negative patients, although the difference did not reach statistical significance (17.6% vs 7.3%; OR 2.71, P = 0.06). The sequence counts of EBV and P. jirovecii in the six EBV-positive patients with P. jirovecii co-detection are shown in Table E4. In two patients with high EBV sequence counts (Patients 1 and 4), the sequence counts of P. jirovecii were also high. Given the small number of cases and the absence of clinical adjudication for Pneumocystis pneumonia, this finding should be interpreted as exploratory co-detection rather than evidence of an increased risk of clinically confirmed Pneumocystis pneumonia. Exploratory Cytokine Profiles Plasma cytokine immunoassays were performed in a subset of 48 patients, including 8 EBV-positive and 40 EBV-negative patients (Table 4). Because cytokine data were available only in a limited subset of the cohort, these analyses were considered exploratory. TRAIL levels were higher in EBV-positive patients than in EBV-negative patients. IL-17 detection was also more frequent in EBV-positive patients. MCP-3 showed a borderline difference between groups but did not reach statistical significance (P = 0.051). Given the small sample size and multiple comparisons, these findings should be interpreted as hypothesis-generating, not confirmatory. Table 4Exploratory Cytokine Profiles Stratified by EBV PositivityEBV Positive (n=8)EBV Negative (n=40)P-valueTRAIL (pg/mL)16(12–20)8(2–12)0.016MCP-3 (pg/mL)8.4±7.66.0±6.40.051IL-175 (62.5%)7 (17.5%)0.017IL-68340.57IFN-alpha28340.57IFN-gamma230.189IL-1ra8320.33MIF7370.53IL-1beta7341Basic FGF4130.429PDGF-BB8371IL-13260.605IL-10490.187GRO-alpha280.666IL-1alpha3260.236Notes: Data are presented as plasma concentrations or as n (%) of patients with detectable cytokine levels, according to the format available for each cytokine. Cytokine analyses were exploratory, and P values were not adjusted for multiple comparisons. Bold values indicate P < 0.05.Abbreviations: TRAIL, TNF-related apoptosis-inducing ligand; MCP, Monocyte chemoattractant Protein; IL, interleukin; IFN, Interferon; MIF, Macrophage migration inhibitory factor; Basic FGF, Basic fibroblast growth Factor; PDGF-BB, Platelet derived growth factor BB; GRO, Growth-related oncogene. Exploratory Cytokine Profiles Stratified by EBV Positivity Notes: Data are presented as plasma concentrations or as n (%) of patients with detectable cytokine levels, according to the format available for each cytokine. Cytokine analyses were exploratory, and P values were not adjusted for multiple comparisons. Bold values indicate P < 0.05. Abbreviations: TRAIL, TNF-related apoptosis-inducing ligand; MCP, Monocyte chemoattractant Protein; IL, interleukin; IFN, Interferon; MIF, Macrophage migration inhibitory factor; Basic FGF, Basic fibroblast growth Factor; PDGF-BB, Platelet derived growth factor BB; GRO, Growth-related oncogene. Exploratory Factors Associated with EBV Positivity Exploratory logistic regression was performed to evaluate factors associated with EBV positivity (Table 5). In the exploratory multivariable model, lower serum albumin (OR 0.89, P = 0.004) and sepsis at enrollment (OR 2.83, P = 0.042) were independently associated with higher odds of EBV positivity. Although P. jirovecii detection was numerically more frequent among EBV-positive patients in univariable analysis, it was not independently associated with EBV positivity after adjustment (adjusted OR 2.22, P = 0.187). Table 5Factors Associated with EBV Positivity in Exploratory Logistic RegressionVariablesUnivariateMultivariateP-valueOR95% CIP-valueSBP0.047///Serum albumin<0.0010.890.82–0.960.004CRP0.009///PCT0.013///PaCO20.011///D-dimer0.059///Serum calcium0.014///Cardiovascular and cerebrovascular diseases0.090///Immunosuppressive status0.267///Pneumocystis jirovecii0.0602.220.68–7.220.187Sepsis0.0102.831.04–7.720.042Note: /, no adjusted OR, 95% CI, or P value was reported for this variable in the final multivariable model.Abbreviations: SBP, systolic blood pressure; CRP, C-reactive protein; PCT, procalcitonin; PaCO2, Partial pressure of arterial carbon dioxide. Factors Associated with EBV Positivity in Exploratory Logistic Regression Note: /, no adjusted OR, 95% CI, or P value was reported for this variable in the final multivariable model. Abbreviations: SBP, systolic blood pressure; CRP, C-reactive protein; PCT, procalcitonin; PaCO2, Partial pressure of arterial carbon dioxide. 30-Day Mortality During the 30-day follow-up, mortality was 52.9% in EBV-positive patients and 48.7% in EBV-negative patients. No statistically significant difference in 30-day mortality was observed between the two groups (18/34 vs 73/150, P = 0.706). Discussion In this multicenter cohort of ICU patients with pneumonia, EBV positivity by blood mNGS was associated with markers of greater clinical severity and inflammation. EBV-positive patients had lower serum albumin and systolic blood pressure, higher qSOFA scores, a higher proportion of sepsis at enrollment, higher CRP and PCT levels, higher PaCO2, and lower serum calcium. In contrast, APACHE II score, shock index, overall microbial detection categories, culture positivity, and 30-day mortality did not differ significantly between groups. These findings suggest that EBV positivity may identify ICU pneumonia patients with greater systemic illness severity and inflammatory burden. EBV positivity in blood mNGS is difficult to interpret clinically. EBV persists lifelong, mainly in memory B cells; therefore, EBV detected from anticoagulated whole blood may reflect latent or cell-associated viral DNA, low-level circulating EBV DNA, illness-related bystander DNA release, or true viral reactivation.4,5 In previous sepsis and ICU studies, EBV reactivation was usually assessed by quantitative PCR in whole blood or plasma, sometimes with serial viral-load monitoring. By contrast, mNGS reports microbial sequence reads rather than validated EBV viral load.6,14 Because our study did not include EBV qPCR, serology, serial viral testing, or systematic evaluation for EBV-related organ disease, EBV positivity should not be interpreted as confirmed clinically significant EBV reactivation. Prior studies provide context for these results. Latent herpesvirus reactivation, including EBV, has been frequently reported in sepsis and critical illness and has been associated with organ dysfunction, prolonged ICU stay, and immunosuppressed host-response profiles.1–3 Respiratory-sample studies in severe pneumonia have also reported EBV and other herpesvirus reactivation.15,16 These studies make EBV detection during severe infection biologically plausible. However, they do not fully resolve whether EBV is a driver of tissue injury, a bystander of critical illness, or a marker of impaired antiviral immune control. Our results are most consistent with the interpretation that EBV positivity is a marker of disease severity and immune dysregulation and systemic inflammation. EBV-positive patients had lower blood pressure, higher qSOFA scores, and more frequent sepsis at enrollment. In the exploratory regression model, sepsis at enrollment was associated with higher odds of EBV positivity; however, the model outcome was EBV positivity, not incident sepsis. This study therefore cannot show that EBV increases the risk of sepsis. A more cautious interpretation is that severe systemic infection may impair host immune control and allow EBV DNA to become detectable in blood.1–3 The albumin result supports this interpretation. Serum albumin was lower in EBV-positive patients and remained associated with EBV positivity after adjustment. In critical illness, hypoalbuminemia may reflect systemic inflammation, capillary leakage, reduced hepatic synthesis, malnutrition, increased catabolism, or overall physiological stress.17,18 Higher CRP and PCT levels in EBV-positive patients are consistent with a greater inflammatory burden. These findings do not prove EBV-mediated inflammation, but they suggest that EBV positivity may occur in patients with a more inflamed and clinically vulnerable host profile. Several additional differences should be interpreted cautiously. Higher PaCO2 may reflect ventilatory failure, respiratory muscle fatigue, airway obstruction, or more severe parenchymal disease. Lower serum calcium may partly reflect hypoalbuminemia and critical illness-related metabolic disturbance. Bloody sputum, hemoptysis, and cavitation may indicate complicated pulmonary infection, tissue destruction, vascular injury, or coexisting bacterial or fungal disease. These features are not EBV-specific manifestations, but they are consistent with the pattern that EBV-positive patients had more complex pulmonary and systemic illness at enrollment. In the limited cytokine subset, TRAIL levels were higher and IL-17 was detected more frequently in EBV-positive patients. However, cytokine testing was available for only 48 patients, including eight EBV-positive patients, and P values were not adjusted for multiple comparisons. These findings should therefore be considered hypothesis-generating rather than mechanistic evidence. The microbiological findings should also be interpreted cautiously. Overall detection of bacteria, fungi, and viruses by mNGS did not differ significantly between groups, and neither did culture positivity. P. jirovecii was numerically more frequent in EBV-positive patients, but the difference did not reach statistical significance and was not independently associated with EBV positivity after adjustment. The six EBV and P. jirovecii co-detected cases may reflect shared host vulnerability, including impaired cellular immunity or critical illness-related immune dysfunction, rather than a direct interaction between the two pathogens.19,20 In addition, mNGS detection alone cannot reliably distinguish colonization, transient DNA detection, and clinically confirmed Pneumocystis pneumonia.21–23 No significant difference in 30-day mortality was observed between EBV-positive and EBV-negative patients. This finding argues against presenting EBV positivity as an independent prognostic marker in the current cohort. The high mortality in both groups, small EBV-positive sample size, and single-time-point EBV assessment may have limited the ability to detect outcome differences. The main clinical implication of this study is not that EBV is a therapeutic target, rather, blood mNGS-detected EBV may serve as a non-specific marker of disease severity, immune dysregulation, and systemic inflammation in ICU pneumonia. Clinicians should not automatically diagnose EBV disease or initiate EBV-directed treatment based only on blood mNGS detection.6,9,14 Instead, EBV positivity should be interpreted together with illness severity, immune status, albumin level, inflammatory markers, co-detected organisms, conventional microbiology, and evidence of organ-specific EBV disease. This study has several strengths. The analysis used a multicenter ICU cohort and combined clinical, laboratory, radiological, microbiological, and exploratory cytokine data. It also addresses a practical diagnostic problem: how to interpret EBV reported by blood mNGS in ICU patients with pneumonia. The study also has important limitations. First, this was a retrospective secondary analysis, and residual confounding cannot be excluded. Second, only 34 patients were EBV-positive, limiting statistical power and the stability of the multivariable model. Third, EBV positivity was defined by blood mNGS rather than quantitative EBV DNA testing; mNGS read counts should not be interpreted as viral load.6,14 Fourth, serial EBV testing, EBV serology, and systematic evaluation for EBV-related organ disease were unavailable. Thus, latent EBV DNA, bystander detection, true reactivation, and clinically relevant EBV disease could not be distinguished. Fifth, cytokine testing was available only in a small subset and was not adjusted for multiple comparisons. Finally, P. jirovecii detection did not necessarily indicate clinically adjudicated Pneumocystis pneumonia, and the mortality analysis may have been underpowered.21,23 Future studies should prospectively evaluate EBV positivity in ICU pneumonia using quantitative EBV DNA testing, serial sampling, and paired blood and respiratory specimens. Standardized mNGS reporting thresholds are needed.9,24,25 Integration of immune profiling, including lymphocyte subsets, immunoglobulins, monocyte HLA-DR expression, and prespecified cytokine panels, may help clarify whether EBV positivity is primarily a marker, mediator, or bystander of immune dysfunction in ICU patients with pneumonia. Conclusions In summary, EBV positivity by blood mNGS was associated with disease severity markers and inflammatory burden in ICU patients with pneumonia, including lower serum albumin, a higher qSOFA score distribution, more frequent sepsis at enrollment, and higher inflammatory biomarker levels. These findings suggest that mNGS-detected EBV may be a marker of host immune dysregulation and systemic inflammation, although the observed associations do not establish a causal role for EBV. Associations with cytokine profiles and P. jirovecii co-detection were exploratory. Future studies should combine mNGS with quantitative EBV DNA testing, serial sampling, and immune profiling to clarify the clinical meaning of EBV positivity in ICU pneumonia. Funding Statement This work was supported by grants from the National Science and Technology Major Project of China (No. 2023ZD0506205), Shanghai Top-Priority Clinical Key Disciplines Construction Project (2017ZZ02013) and Outstanding Resident Clinical Postdoctoral Program of Zhongshan Hospital Affiliated to Fudan University. Data Sharing Statement All data are available in the main text or the Supplementary Materials. The datasets used and/or analyzed during the current study are available from the corresponding author Jing Zhang on reasonable request. Ethics Considerations The prospective cohort study, on which the current retrospective secondary analysis was based, was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Zhongshan Hospital Affiliated to Fudan University (Ethical Approval Number: B2018-182R). Consent to Participate Written informed consent was obtained from all the participants. Author Contributions Wei-ping Hu and Jing Zhang conceived and designed the study. Wei-ping Hu, Yi-Xing Wu, Jian-Lan Hua, Zi-Xuan Cheng, Xiao Wang, and Li Liu collected the data. Wei-ping Hu and Xia Wang analyzed the data. Xia Wang drafted the manuscript. Wei-ping Hu and Jing Zhang revised the manuscript. All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Disclosure The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. References Associated Data Data Availability Statement All data are available in the main text or the Supplementary Materials. The datasets used and/or analyzed during the current study are available from the corresponding author Jing Zhang on reasonable request.