Clinical application value of simultaneous plasma and bronchoalveolar lavage fluid metagenomic next generation sequencing in patients with pneumonia-derived sepsis

doi:10.21203/rs.3.rs-4949378/v1

Download PDF

Research Article

Clinical application value of simultaneous plasma and bronchoalveolar lavage fluid metagenomic next generation sequencing in patients with pneumonia-derived sepsis

https://doi.org/10.21203/rs.3.rs-4949378/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Despite the increasing use of metagenomic next-generation sequencing (mNGS) in sepsis, identifying clinically relevant pathogens remains challenging. This study was aimed to evaluate the clinical utility of simultaneous plasma and bronchoalveolar lavage fluid (BALF) detection using mNGS.

Methods

This retrospective study enrolled patients with pneumonia-derived sepsis (PDS) admitted to the intensive care unit (ICU) between October 2021 and January 2023. Patients were divided into two groups: the mNGS group and the non-mNGS group, based on whether simultaneous plasma and BALF mNGS were conducted. All patients underwent conventional microbiological tests (CMT), including bacterial/fungal culture of peripheral blood and BALF, as well as sputum culture, detection of 1, 3-beta-D- glucan in BALF and RT-PCR testing.

Results

The mNGS group exhibited a lower mortality rate than the non-mNGS group (35.0% vs 57.1%, P = 0.034). Simultaneous detection in dual-sample resulted in a higher proportion of microorganisms identified as definite causes of sepsis alert compared to detection in either plasma or BALF alone. Acinetobacter baumannii, Stenotrophomonas maltophilia, Candida albicans, and human mastadenovirus B were the primary strains responsible for infections in PDS patients. Patients with lower white blood cells and neutrophil indices had a greater consistency in dual-sample mNGS. Patients in the mNGS group had more antibiotic adjustments compared to the non-mNGS group. The percentage of neutrophils was a risk factor for mortality in PDS patients.

Conclusion

Dual sample mNGS has the advantage of detecting and determining the pathogenicity of more pathogens and has the potential to improve the prognosis of patients with PDS.

Metagenomic next-generation sequencing

Pneumonia-derived sepsis

Dual sample mNGS

Clinical value

Sepsis is a life-threatening syndrome characterized by dysregulated host response to infection, resulting in organ dysfunction. Pneumonia is the predominant cause of sepsis, accounting for approximately half of all cases [1, 2]. Pneumonia-derived sepsis (PDS) exhibits the highest mortality rate in ICUs worldwide, with a mortality rate as high as 50% in patients with septic shock [3, 4]. Timely and accurate pathogen diagnoses are of great significance for better antibiotic treatment and improved clinical outcomes [5]. However, the low detection rate and time-consuming process (3–5 days) of conventional microbiological tests (CMT), including culture, serological tests, and PCR, are the problems faced by clinicians [6]. No aetiologic pathogen is identified in over 30% of patients undergoing culture-based diagnosis process [7]. For patients with PDS, the complexities of illness and prior antibiotic exposure further complicate the effective detection of pathogens.

Metagenomic next-generation sequencing (mNGS) is a promising diagnostic tool for PDS due to its broad range of detection and a shorter turnaround time compared to culture methods. Invasive specimens, such as bronchoalveolar lavage fluid (BALF) are commonly used to perform mNGS in pneumonia patients. In the case of most BALF samples, mNGS reports as many as 5–10 microorganisms, making it challenging to distinguish between pathogens associated with infection, colonization, or contamination [8]. Recent advancements in plasma cell-free DNA (cfDNA) sequencing have greatly expanded the capabilities of mNGS, enabling the non-invasive detection of pathogen nucleic acid circulating from various anatomical sites of infection. However, the extent to which microbial DNA circulates in the bloodstream of pneumonia patients remains unclear, and this knowledge could enhance our understanding of disease pathophysiology and clinical diagnosis.

Several studies have explored the efficacy of simultaneously conducting plasma and BALF samples mNGS [9–12]. In patients who test positive for both sample types, the concordance rate can reach 90% [9]. However, some studies have indicated that plasma is a less reliable indicator compared to respiratory sample. Among suspected pneumonia patients, only 26% had at least one plasma mNGS result that matched the BALF mNGS findings [11]. Importantly, the administration of antibiotics prior to sampling poses an inevitable and intricate bias for the mNGS assessment. Additionally, the disease severity can influence the detection results of pathogens. Therefore, the objective of this study is to analyze the concordance of pathogen detection between plasma and BALF samples using mNGS in PDS patients and explore the clinical utility of this concordance.

Study design and participants

This study involved 150 sepsis patients who were admitted to the ICU of the National Hospital of Guangxi Zhuang Autonomous Region between October 2021 and January 2023. The inclusion criteria were as follows: (1) age ≥ 18 years; (2) ICU stay > 24 hours; (3) presence of typical clinical signs of pulmonary infection, such as fever, cough, expectoration, and respiratory failure; (4) the diagnosis of pulmonary infection was supported by radiological evidence, including chest X-ray or computed tomography scan results; (5) adherence to the sepsis 3.0 diagnostic criteria issued by the Society of Critical Care Medicine (SCCM) [13]; (6) simultaneous collecting blood and BALF samples from each patient. Based on the inclusion criteria, we excluded 55 cases of sepsis unrelated to pneumonia, resulting in a final selection of 95 sepsis patients from the ICU for retrospective investigation (Fig. 1).

Some patients, due to personal or family circumstances, declined to undergo mNGS testing, opting solely for conventional microbiological tests (CMT). The type of testing conducted allowed for the following grouping criteria: (1) mNGS group: simultaneous mNGS testing of plasma and BALF, in addition to CMT; (2) non-mNGS group: CMT performed on both blood and BALF.

The study was approved by the Scientific Research and Clinical Trials Ethics Committee of the National Hospital of Guangxi Zhuang Autonomous Region (2022-55). Conducted in accordance with the principles of the Declaration of Helsinki.

Data Collection

The collected data included demographic information and the results of laboratory examinations, such as white blood cell count (WBC), hemoglobin (Hb), blood platelet count (PLT), neutrophils percentage (NE%), lymphocytes percentage (LY%), neutrophil count (NE), hematocrit (HCT%), C-reactive protein (CRP), and procalcitonin (PCT). Clinical diagnosis, symptoms, sepsis-related organ failure assessment (SOFA) scores, treatment details, and clinical outcomes were extracted from the electronic medical records of the Ethnic Hospital affiliated with Guangxi Medical University. A standardized data collection form was utilized for this purpose. Data on pre-admission antibiotic treatment, the initial antibiotic administered upon admission, and any subsequent adjustments made based on mNGS results were also recorded. CMT encompassed bacterial/fungal culture of peripheral blood and BALF, concurrent with mNGS testing, as well as sputum culture, detection of 1, 3-beta-D- glucan in BALF and RT-PCR testing.

Metagenomic next-generation sequencing and analysis

Sample collection and nucleic acid extraction

Whole blood samples (≥ 4 mL) and BALF sample (≥ 3 mL) were collected in anticoagulation tubes and transported at 4℃ after collection. cfDNA was extracted from plasma using PathoXtract Plasma Nucleic Acid kit (WYXM03001S, WillingMed Corp, Beijing, China) following the manufacturer’s protocol. Both DNA and RNA were extracted from BALF samples to cover all the potential pathogens. DNA was extracted using PathoXtract® Basic Pathogen Nucleic Acid Kit (WYXM03211S, WillingMed Corp, Beijing, China) and RNA was extracted using PathoXtract® Virus DNA/RNA Isolation Kit (WYXM03009S, WillingMed Corp, Beijing, China) following the manufacturer’s protocol[14]. DNA and RNA was eluted with 50 µL of nuclease-free water. The DNA and RNA were mixed and then reverse transcription of the RNA into complementary DNA (cDNA) was performed by using SuperScript® Double-Stranded cDNA Synthesis Kit (11917020, Invitrogen).

Library preparation and sequencing

For cfDNA libraries construction, the KAPA DNA HyperPrep Kit (KK8504, KAPA, Kapa Biosystems, Wilmington, MA, United States) was used according to the manufacturer’s protocol. Genomic DNA libraries was constructed using the Illumina® DNA Prep, (M) Tagmentation (20018705, Illumina) according to the manufacturer’s protocol. Pooled libraries were sequenced on NextSeq™ 550Dx system using a 75 bp, single-end sequencing kit (Illumina) and at least 20 million sequencing reads were acquired for each sample.

Bioinformatic analysis

Quality control was carried out after high-quality sequencing data were obtained. High-quality data consistent with the human reference genome sequence were screened and removed using Bowtie2 v2.4.3. By comparing the remaining sequences with the existing nucleic acid sequences of microorganisms in the database, the type of microorganism can be identified[14].

For pathogen identification, a threshold value called RPTM (reads per ten million) was used. RPTM represents the number of pathogen specific reads per ten million. For viral pathogens detection, reads count ≧ 3 was used as an empirical threshold. In the case of bacteria and fungi, positive pathogens in blood samples were required to meet an RPTM threshold ≧ 8. In BALF samples, positive bacteria and fungi were required to achieve an RPTM threshold ≧ 20. Special pathogens (including Cryptococcus and Mycobacterium) with RPTM ≧ 1 was considered positive.

Clinical adjudication of mNGS results

To confirm the composite clinical diagnosis, two experienced infectious disease specialists, who were well-versed in the ICU, independently reviewed the medical records and mNGS results of each patient. The panel of clinicians categorized the possibility of mNGS as the causative pathogen of PDS into four categories: definite, probable, possible, and unlikely. This categorization was based on previous reports by Blauwkamp et al. [15]. (1) Definite: the mNGS pathogen result was concordant with a microbiological test performed within seven days of sample collection for mNGS identification. (2) Probable: the mNGS pathogen result was considered the probable cause of the sepsis alert based on clinical, radiological or laboratory findings. (3) Possible: the mNGS pathogen result had potential for pathogenicity consistent with clinical presentation but an alternative explanation for the symptoms was more likely. (4) Unlikely: the cfDNA sequencing pathogen result had potential for pathogenicity but findings were inconsistent with clinical presentation. The final classification of the cases was at the discretion of the committee chair.

Statistical analysis

All data in this study were statistically analyzed by SPSS 25.0 (USA) and plotted using GraphPad prism8 (USA). The t-test was used to determine the normal distribution and uniformity of variance. The Wilcoxon rank test was used to calculate the variance of measured data that were not normally distributed or had variance homogeneity. Chi-square test was used for discrete variables where appropriate. P-values ≦ 0.05 were considered significant, and all tests were two-tailed. Cox proportional hazards model was applied for discovering the independent risk factors for PDS. The model was built on the “survival” package in R version (v4.2.2).

Patient characteristics

The study included a total of 95 PDS patients, with 60 patients in the mNGS group, and 35 patients in the non-mNGS group. Table 1 documented the patient characteristics and baselines information. Statistical analyses were conducted to assess factors like age, sex, SOFA score, comorbidities, and patient outcomes. All factors, except for the SOFA score, showed no significant differences, suggesting a relatively similar baseline between the two groups. Despite the longer ICU length of stay in the mNGS group compared to the non-mNGS group (20 vs 8, P = 0.115), the mNGS group exhibited a significantly lower mortality rate (35.0% vs 57.1%, P = 0.034) (Table 1).

Table 1

**Clinical characteristics of participants.**
Characteristic	mNGS (n = 60)	Non-mNGS (n = 35)	P-value
Age, median (IQR)	68 (60–77)	73 (62–85)	0.177
Sex, n (%)
Male	47 (78.3%)	26 (74.3%)	0.652
Female	13 (21.7%)	9 (25.7%)
SOFA score, median (IQR)	11 (7–13)	8 (5–11)	0.039*
Comorbidities, n (%)
Respiratory failure	42 (70.0%)	27 (77.1%)	0.451
Hypertension	39 (65.0%)	19 (54.3%)	0.302
Diabetes	23 (38.3%)	8 (22.9%)	0.121
Hypoproteinemia	14 (23.3%)	13 (37.1%)	0.150
CAD	6 (10.0%)	8 (22.9%)	0.088
MODS	30 (50.0%)	15 (42.9%)	0.501
ARDS	20 (33.3%)	17 (48.6%)	0.142
CKD	14 (23.3%)	7 (20%)	0.706
Outcomes, median (IQR)
Hospital length of stay(days)	31 (22–53)	22 (9–46)	0.111
ICU length of stay(days)	20 (11–28)	8 (5–14)	0.115
Mortality(%)	21 (35.0%)	20 (57.1%)	0.034^*
IQR, interquartile range; SOFA, sepsis-related organ failure assessment; CAD, Coronary atherosclerotic heart disease; MODS, Multiple organ dysfunction syndrome; ARDS, Acute respiratory distress syndrome; CKD, Chronic kidney disease.,P-values ≦ 0.05; ,P-values ≦ 0.01; **,P-values ≦ 0.001

Detection performance of plasma and BALF mNGS

The performance of mNGS was compared to CMT in all patients (Fig. 2A). The pathogen positivity rate was higher in plasma and BALF mNGS compared to blood and BALF culture, respectively (61.7% vs 6.7%, P < 0.001; 100% vs 40%, P < 0.001). Specifically, BALF mNGS exhibited a 100% positivity rate. No significant difference was observed in the positivity rates of different CMT methods between the mNGS and non-mNGS groups.

Subsequently, we analyzed the quantity of microorganisms identified by various methods (Fig. 2B). Within the mNGS group, 8.3% of the patients exhibited a singular microorganism through plasma mNGS, whereas 55% identified more than 4 microorganisms through BALF mNGS. The CMT method identified 1 or 2 microorganisms in 78.3% of patients in the mNGS group and 65.7% of patients in the non-mNGS group (Additional file 1: Figure S1).

Then we assessed the consistency between mNGS and CMT methods. Among the mNGS group, a total of 33 patients were concurrently detected with microorganisms in plasma mNGS, BALF mNGS, and CMT (Fig. 2C). Out of the 33 patients who tested positively for both plasma mNGS and CMT, only 33% had consistent microbial detection (Fig. 2D). Of the 52 patients who tested positive for both BALF mNGS and CMT, 75% displayed at least one consistent microbial detection (Fig. 2E). When combing plasma and BALF sample mNGS results, out of the 52 patients who tested positively for both mNGS and CMT, 79% had consistent microbial detection (Fig. 2F).

Clinical adjudication of plasma and BALF mNGS results

In the mNGS group, 37 patients had both plasma and BALF mNGS test results positive, and among them, 28 patients were identified to have at least one consistent microorganism (Fig. 3A). A total of 284 microorganisms were identified through mNGS. Out of these, 60 microorganisms were identified in the plasma samples, while 260 microorganisms were identified in BALF samples, and 36 of them were consistently identified in both sample types (Fig. 3B). Out of the 36 detections, 55.6% (20/36) was classified as definite causes of the sepsis alert (Fig. 3C). Among the microorganisms identified exclusively in either plasma or BALF mNGS, only 20.8% (5/24 in plasma) and 18.8% (42/224 in BALF) were classified as definite causes of the sepsis alert (Fig. 3C). The proportion of definite causes increased by about 35% when microorganisms were identified in both samples compared to those identified in only one sample, the proportion of microorganisms classified as "Definite + Probable" in cases where microorganisms were simultaneously detected in both samples (88.9%) increased by 9 to 26% compared to microorganisms detected in a single sample alone (62.5% in plasma and 79.9% in BALF). This means that detecting consistent microorganisms can better clarify the clinical significance of pathogens.

Microorganisms identified in patients with PDS

A total of 26 kinds of pathogens detected by plasma and BALF mNGS were classified as definite causes of the sepsis alert, including 15 bacteria, 5 fungi, 4 viruses, and 2 special pathogens (Fig. 4). Acinetobacter baumannii, Stenotrophomonas maltophilia, Candida albicans, and Human mastadenovirus B were the most common pathogens in PDS patients. Noting the particular sensitivity of plasma in identifying special pathogens such as Orientia tsutsugamushi and Mycobacterium tuberculosis complex. A total of 43 pathogens were classified as probable causes of sepsis alert. The most frequently detected bacteria were Corynebacterium striatum, Staphylococcus haemolyticus, Enterococcus faecium, and Burkholderia cepacia, as well as Epstein-Barr virus and cytomegalovirus. Similar to the “definite” pathogens, viruses were frequently detected in both plasma and BALF, whereas fungi were exclusively detected in BALF (Additional file 2:Figure S2). The CMT method identified a total of 19 pathogens in both the mNGS and non-mNGS groups. The most common pathogens in both the mNGS groups and non-mNGS groups were Candida app (Additional file 3:Figure S3).

Then we analyzed the pathogen burden based on RPTM among microorganisms identified by plasma and BALF mNGS. There were no significant differences in pathogen burden among microorganisms classified as definite, probable, possible, and unlikely causes of the sepsis alert in plasma (Fig. 5A). The pathogen burden of microorganisms classified as definite causes of the sepsis alert was significantly higher than that of microorganisms classified as probable, possible, and unlikely causes of the sepsis alert in BALF (Fig. 5B). Moreover, the pathogen burden of microorganisms identified in both plasma and BALF samples was significantly higher than that of microorganisms identified only in plasma or BALF samples (Fig. 5C and D).

The associations between inflammatory markers and mNGS results

To investigate potential factors associated with the matched results between plasma and BALF mNGS findings, we compared differences in inflammatory markers, including WBC, Hb, PLT, NE%, LY%, NE, HCT%, CRP, and PCT. The inflammatory markers between the mNGS and non-mNGS groups were generally similar, except for the Hb index, which was lower in the mNGS group (Additional file 4: Table S1). Subsequently, we compared the differences between the matched or partially matched cases and unmatched cases. There were no significant differences in Hb, PLT, NE%, LY%, HCT%, CRP, and PCT between these two groups. However, the unmatched cases exhibited significantly higher WBC indices (11.03 ± 5.42 versus 14.30 ± 6.50 10^9/L, P = 0.041) and NE indices (8.75 ± 5.27 vs 12.17 ± 5.78 10^9/L, P = 0.021) compared to the matched or partially matched cases (Table 2).

Table 2

**Comparison of inflammatory markers between matched or partially matched cases and unmatched cases of plasma and BALF mNGS detection.**
Laboratory tests	Plasma and BALF mNGS matched or partially matched cases (n = 28)	Plasma and BALF mNGS unmatched cases (n = 32)	P-value
WBC (10^9/L)	11.03 ± 5.42	14.30 ± 6.50	0.041*
Hb (g/L)	90.64 ± 22.60	88.42 ± 22.95	0.707
PLT (10^9/L)	187.54 ± 120.41	242.75 ± 138.20	0.107
NE%	82.14 ± 15.64	83.44 ± 8.87	0.689
LY%	11.22 ± 12.10	11.89 ± 17.83	0.866
NE (10^9/L)	8.75 ± 5.27	12.17 ± 5.78	0.021*
HCT%	30.22 ± 22.46	27.41 ± 7.60	0.507
CRP (mg/L)	108.03 ± 88.30	116.61 ± 94.32	0.736
PCT (ng/ml)	1.32 ± 2.32	8.84 ± 22.37	0.153
WBC, white blood cell; Hb, hemoglobin; PLT, blood platelet; NE%, Percentage of neutrophils in blood; LY%, Percentage of lymphocytes and total white blood cell; NE, neutrophil; HCT%, hematocrit; CRP, C-reactive protein; PCT, procalcitonin.

Application of mNGS for antibiotic adjustment

All hospitalized patients received antibiotic treatment. Among the patients in the mNGS group, 56 individuals received antibiotic treatment prior to undergoing mNGS testing (Table 3). Amon the 56 patients, 85.71% (48/56) had adjustments made to their antibiotic treatment, a higher percentage compared to the non-mNGS group (33.33%, 11/33) (Table 3). Based on the mNGS results, 19.64% (11/56) of the patients had a reduction in the number or spectrum of initial empiric antimicrobial drugs, while no patients in the non-mNGS group experienced de-escalation. Antibiotic escalation was observed in 42.86% (24/56) of patients in the mNGS group, primarily due to the addition of targeted antimicrobial drugs. Additionally, one patient in the mNGS group showed symptom improvement and was discharged from the hospital before the mNGS results were obtained (Table 3).

Table 3

The adjustment of antibiotics in the mNGS group and the non-mNGS group.
Antibiotic adjustments	mNGS (n = 56)	Non-mNGS (n = 33)	P value
Adjustments	48 (85.71%)	11 (33.33%)	< 0.001***
De-escalation	11 (19.64%)	0 (0.00%)	0.007**
Reduce spectrum of agents	3 (5.36%)	0 (0.00%)	0.176
Reduce number of agents	8 (14.29%)	0 (0.00%)	0.023*
Escalation	24 (42.86%)	4 (12.12%)	0.003**
Increase number of agents	20 (35.71%)	4 (12.12%)	0.015*
Increase spectrum of agents	4 (7.14%)	0 (0.00%)	0.116
Same level replacement	13 (23.21%)	7 (21.21%)	0.827
Unadjustments	7 (12.5%)	22 (66.67%)	< 0.001***
No change	7 (12.5%)	22 (66.67%)	< 0.001***
Discharged before mNGS	1(1.79%)

Next, we compared the antibiotic adjustment between the matched or partially matched cases and unmatched cases from dual sample mNGS tests. Among the 27 patients with matched or partially matched results, 81.48% (22/27) had adjustments in antibiotic treatment based on the mNGS results, while 89.66% (26/29) had adjustments in antibiotic treatment in the 29 patients with unmatched results (Additional file 5: Table S2). Compared to patients with unmatched results, more patients in the group with matched or partially matched results increased spectrum of agents (14.81% vs 0.00%, P = 0.032) (Table S2).

Factors influencing prognosis in patients with PDS

Compared to the non-mNGS group, the mNGS group exhibited a prolonged length of stay in the ICU and a reduced mortality rate (Table 1). Further analysis was conducted to compare the clinical outcomes among cases with matched or partially matched plasma and BALF mNGS results and those with unmatched results. The mortality rate was lower in the plasma and BALF consistent mNGS group compared to the inconsistent group, but the difference was not statistically significant (35.71% vs 40.61%, P = 0.696) (Table 4).

Table 4

Clinical outcomes between matched or partially matched cases and unmatched cases of plasma and BALF mNGS detection.
Outcomes	Blood and BALF mNGS matched or partially matched cases (n = 28)	Blood and BALF mNGS unmatched cases (n = 32)	P-value
Hospital length of stay (days)	31 (22–52)	30 (22–51)	0.510
ICU length of stay (days)	20 (11–28)	19 (11–28)	0.625
Mortality (%)	10 (35.71%)	13 (40.61%)	0.696

The initially univariate regression analysis identified potential factors influencing sepsis prognosis. The univariate Cox regression analysis indicated a significant association between age (adjusted HR = 0.98) and CMT positive (adjusted HR = 2.36) with adverse outcomes (P ≦ 0.05) (Table 5). To conduct a comprehensive assessment of independent risk factors and determine the impact of matched dual mNGS results on patient prognosis, we incorporated ‘dual mNGS consistency’ and variables with P < 0.2 in the univariate analysis into the multivariate analysis. The multivariate analysis revealed that NE% emerged as an independent risk factor for poor prognosis among sepsis patients (adjusted HR = 0.94, P = 0.002) (Table 5). The consistency of dual-mNGS exhibited a trend as a risk factor for mortality in sepsis patients (adjusted HR = 1.35, P = 0.451).

Table 5

Multivariate regression analysis of predictors of clinical outcomes.
Variables	Univariate analyses		Multivariate analysis
Variables	HR (95% CI)	P-value	HR (95% CI)	P-value
Gender	0.77 (0.38–1.54)	0.457	——	——
Age	0.98 (0.97-1.00)	0.048*	0.99 (0.96–1.01)	0.256
SOFA	1.00 (0.94–1.08)	0.914	——	——
RF	0.70 (0.36–1.38)	0.307	——	——
Dual mNGS consistent	0.94 (0.49–1.82)	0.860	1.35 (0.62–2.93)	0.451
CMT positive	2.36 (1.02–5.46)	0.044*	0.54 (0.15-2.00)	0.358
WBC	1.02 (0.98–1.08)	0.332	——	——
Hb	1.01 (1.00-1.02)	0.156	1.01 (0.99–1.03)	0.194
PLT	1.00 (1.00–1.00)	0.356	——	——
NE%	0.98 (0.96-1.00)	0.075	0.94 (0.91–0.98)	0.002**
HCT	0.99 (0.97–1.01)	0.270	——	——
CRP	1.00 (1.00–1.00)	0.747	——	——
PCT	1.01 (1.00-1.02)	0.072	1.01 (1.00-1.02)	0.064
HR, hazard ratio; CI, confidence interval; SOFA, sequential organ failure assessment; CMT, conventional microbiological tests.

Our study involved conducting a dual sample mNGS test on 60 PDS patients and comparing the antibiotic adjustment and clinical outcomes of these patients to those who did not undergo mNGS testing. Overall, mNGS detected a wider range of pathogens compared to CMT methods, even when most patients have already undergone antibiotic treatment before specimen collection. The results of metagenomic NGS can be valuable even when they are concordant with results of conventional testing or confirm empirical treatment plans. In those cases, it provides reassurance of the diagnosis and potentially rules out non-infection cases [12].

The presence of consistent pathogens in both plasma and BALF samples indicates a substantial dissemination of pathogens from the lungs to the bloodstream. In our study, 46.67% (28/60) of patients demonstrated the presence of at least one matched microorganism in both plasma and BALF through mNGS. The microorganisms that were consistent between plasma and BALF samples in mNGS analysis tends are more likely to be associated with sepsis. Compared to the microorganisms exclusively identified in either plasma or BALF samples, the percentage of microorganisms that were classified as definite causes of sepsis alert increased by 35%. Interpreting mNGS results of respiratory specimens is challenging. While subjective judgment by doctors is necessary, distinguishing between colonization and infection remains a significant challenge when formulating follow-up treatment plans using mNGS results from respiratory samples only [8]. Our findings suggest that dual samples mNGS can conveniently and promptly confirm the pathogen in patients, enabling clinicians to initiate targeted treatment at an earlier stage.

Among the pathogens identified, we found that viruses were the most easily detectable in both plasma and BALF simultaneously. This finding suggests that viruses have a higher likelihood of spreading into the bloodstream compared to bacteria or fungi. Several studies have indicated that plasma mNGS can effectively monitor viral pneumonia and viral sepsis in clinical practice [10, 16]. In bacteria, Klebsiella pneumoniae, a common pathogen of pneumonia, was found to be more easily detected in both plasma and BALF simultaneously. Plasma mNGS exhibited a lower detection rate for fungi compared to BALF, consistent with previous studies’ findings.

We conducted an analysis of clinical indicators in the mNGS group. The results indicated that individuals with lower levels of WBC and NE index were more likely to exhibit consistent findings between plasma and BALF sample mNGS results. Previous research have also suggested a potential correlation between reduced NE counts and an increased likelihood of common human microbiome entering the bloodstream [17]. However, further investigation is necessary to validate the feasibility of these findings and elucidate the underlying mechanisms. These research findings can provide valuable insights for clinicians in selecting appropriate specimens for mNGS testing.

In this study, a higher proportion of patients in the mNGS group experienced adjustments in antibiotic treatment compared to non-mNGS patients (85.71% vs 33.33%, P < 0.001). Those whose treatment was modified based on mNGS results received not only antibiotic escalation but also antiviral drugs (e.g., ganciclovir), antifungal drugs (e.g., caspofungin, fluconazole), and antibiotics targeting specific bacteria (e.g., vancomycin). Patients with consistent dual sample mNGS results were more likely to receive an upgrade in the spectrum of antibiotics and the addition of specific antimicrobial agents compared to those with inconsistent results.

Several patient-related factors, such as age, gender, SOFA, and APACHE II scores, have been recognized as independent risk factors for sepsis-related mortality, as mentioned in previous studies [18, 19]. In our study involving PDS patients, similar results were observed. In the univariate analysis, age and positive CMT were identified as independent risk factors for mortality. Furthermore, the multivariate analysis showed that NE% was a significant risk factor for mortality in PDS patients. While there was a tendency for consistent results from dual sample mNGS to be a risk factor, it did not reach statistical significance.

Our study has several limitations. Firstly, the small sample size necessitates larger-scale clinical trials to validate our findings. Secondly, although consensus was attained by two experienced senior physicians in this study, incorporating clinical manifestations and other laboratory results, subjectivity may have introduced certain bias. Finally, the performance of mNGS at different stages of the disease for each patient could introduce bias in the detection results.

Simultaneously conducting plasma and BALF mNGS facilitates the identification of the causative agent of sepsis, especially when the NE% of patients is lower. mNGS has the potential to improve the prognosis of patients with PDS.

mNGS	Metagenomic next-generation sequencing
BALF	Bronchoalveolar lavage fluid
PDS	Pneumonia-derived sepsis
ICU	Intensive care unit
CMT	Conventional microbiological tests
WBC	White blood cell
Hb	Hemoglobin
PLT	Platelet
NE	Neutrophils
LY	Lymphocytes
NE%	Neutrophils percentage
LY%	Lymphocytes percentage
HCT%	Hematocrit percentage
CRP	C-reactive protein
PCT	Procalcitonin
SOFA	Sepsis-related organ failure assessment
cfDNA	Cell-free DNA
cDNA	Complementary DNA
RPTM	Reads per ten million

Ethics approval and consent to participate

The study was approved by the Research Ethics Committee of National Hospital of Guangxi Zhuang Autonomous Region (2022-55), which waived the written informed consent requirement. Conducted in accordance with the principles of the Declaration of Helsinki.

Clinical trial number: not applicable.

Consent for publication

Not applicable.

Availability of data and material

The data presented in the study are deposited in the SRA (https://www.ncbi.nlm.nih.gov/sra/) repository, accession number PRJNA1063543.

Competing interests

Yuxin Guo, Yafeng Zheng and Wei Gai are employed by WillingMed Technology (Beijing) Co., Ltd. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Authors' contributions

GQ, JYL and WG had the idea for and designed the study and had full access to all data in the study and took responsibility for the integrity of the data and the accuracy of the data analysis. JYL, DXP and YXG wrote and revised the manuscript. XLL, CD, FFX, ZNL, LHS and QHC participated in data collection. BZ, YFZ, and SHN performed the data analysis. RFQ and FFJ contributed to critical revision of the manuscript for important intellectual content.All authors read and approved the final manuscript.

Acknowledgements

We thank the participants in the study.

Beutz MA, Abraham E. Community-acquired pneumonia and sepsis. Clin Chest Med. 2005;26(1):19–28.
Montull B, Menéndez R, Torres A, Reyes S, Méndez R, Zalacaín R, Capelastegui A, Rajas O, Borderías L, Martin-Villasclaras J, et al. Predictors of Severe Sepsis among Patients Hospitalized for Community-Acquired Pneumonia. PLoS ONE. 2016;11(1):e0145929.
Li LL, Dai B, Sun YH, Zhang TT. Monocytes Undergo Functional Reprogramming to Generate Immunosuppression through HIF-1α Signaling Pathway in the Late Phase of Sepsis. Mediators of inflammation 2020, 2020:4235909.
Pereira JM, Paiva JA, Rello J. Severe sepsis in community-acquired pneumonia–early recognition and treatment. Eur J Intern Med. 2012;23(5):412–9.
Lindberg O, De Geer L, Chew M. Nonadherence to antibiotic guidelines in patients admitted to ICU with sepsis is associated with increased mortality: A registry-based, retrospective cohort study. Eur J Anaesthesiol. 2020;37(2):113–20.
Ferreira J, Sarda C, Rello J. Burden of Community-Acquired Pneumonia and Unmet Clinical Needs. Adv therapy. 2020;37(4):1302–18.
Novosad SA, Sapiano MR, Grigg C, Lake J, Robyn M, Dumyati G, Felsen C, Blog D, Dufort E, Zansky S. Vital signs: epidemiology of sepsis: prevalence of health care factors and opportunities for prevention. Morb Mortal Wkly Rep. 2016;65(33):864–9.
Geng S, Mei Q, Zhu C, Fang X, Yang T, Zhang L, Fan X, Pan A. Metagenomic next-generation sequencing technology for detection of pathogens in blood of critically ill patients. Int J Infect Dis. 2021;103:81–7.
Chen J, Zhao Y, Shang Y, Lin Z, Xu G, Bai B, Zheng J, Li P, Mao Y, Deng Q et al. The clinical significance of simultaneous detection of pathogens from bronchoalveolar lavage fluid and blood samples by metagenomic next-generation sequencing in patients with severe pneumonia. J Med Microbiol 2021, 70(1).
Chen X, Ding S, Lei C, Qin J, Guo T, Yang D, Yang M, Qing J, He W, Song M et al. Blood and Bronchoalveolar Lavage Fluid Metagenomic Next-Generation Sequencing in Pneumonia. The Canadian journal of infectious diseases & medical microbiology = Journal canadien des maladies infectieuses et de la microbiologie medicale 2020, 2020:6839103.
Sun T, Liu Y, Cai Y, Zhai T, Zhou Y, Yang B, Wu X, Zhan Q. A Paired Comparison of Plasma and Bronchoalveolar Lavage Fluid for Metagenomic Next-Generation Sequencing in Critically Ill Patients with Suspected Severe Pneumonia. Infect Drug Resist. 2022;15:4369–79.
Xie F, Duan Z, Zeng W, Xie S, Xie M, Fu H, Ye Q, Xu T, Xie L. Clinical metagenomics assessments improve diagnosis and outcomes in community-acquired pneumonia. BMC Infect Dis. 2021;21:1–9.
Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, Bellomo R, Bernard GR, Chiche J-D, Coopersmith CM. The third international consensus definitions for sepsis and septic shock (Sepsis-3). JAMA. 2016;315(8):801–10.
Li D, Gai W, Zhang J, Cheng W, Cui N, Wang H. Metagenomic Next-Generation Sequencing for the Microbiological Diagnosis of Abdominal Sepsis Patients. Front Microbiol. 2022;13:816631.
Blauwkamp TA, Thair S, Rosen MJ, Blair L, Lindner MS, Vilfan ID, Kawli T, Christians FC, Venkatasubrahmanyam S, Wall GD, et al. Analytical and clinical validation of a microbial cell-free DNA sequencing test for infectious disease. Nat Microbiol. 2019;4(4):663–74.
Fang X, Xu M, Fang Q, Tan H, Zhou J, Li Z, Li F, Yang S. Real-time Utilization of Metagenomic Sequencing in the Diagnosis and Treatment Monitoring of an Invasive Adenovirus B55 Infection and Subsequent Herpes Simplex Virus Encephalitis in an Immunocompetent Young Adult. Open Forum Infect Dis. 2018;5(6):ofy114.
Horiba K, Torii Y, Okumura T, Takeuchi S, Suzuki T, Kawada J-i, Muramatsu H, Takahashi Y, Ogi T, Ito Y. Next-generation sequencing to detect pathogens in pediatric febrile neutropenia: a single-center retrospective study of 112 cases. Open Forum Infectious Diseases: 2021. Oxford University Press US; 2021. p. ofab223.
Sun L, Zhang S, Yang Z, Yang F, Wang Z, Li H, Li Y, Sun T. Clinical application and influencing factor analysis of metagenomic next-generation sequencing (mNGS) in ICU patients with sepsis. Front Cell Infect Microbiol 2022:972.
Tong Y, Zhang J, Fu Y, He X, Feng Q. Risk Factors and Outcome of Sepsis in Traumatic Patients and Pathogen Detection Using Metagenomic Next-Generation Sequencing. Canadian Journal of Infectious Diseases and Medical Microbiology 2022, 2022.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
25 Sep, 2024
Reviews received at journal
18 Sep, 2024
Reviewers agreed at journal
17 Sep, 2024
Reviewers agreed at journal
03 Sep, 2024
Reviews received at journal
31 Aug, 2024
Reviewers agreed at journal
26 Aug, 2024
Reviewers agreed at journal
26 Aug, 2024
Reviewers invited by journal
26 Aug, 2024
Editor invited by journal
23 Aug, 2024
Editor assigned by journal
22 Aug, 2024
Submission checks completed at journal
22 Aug, 2024
First submitted to journal
21 Aug, 2024

You are reading this latest preprint version

Clinical application value of simultaneous plasma and bronchoalveolar lavage fluid metagenomic next generation sequencing in patients with pneumonia-derived sepsis

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

Materials and methods

Study design and participants

Data Collection

Metagenomic next-generation sequencing and analysis

Sample collection and nucleic acid extraction

Library preparation and sequencing

Bioinformatic analysis

Clinical adjudication of mNGS results

Statistical analysis

Results

Patient characteristics

Detection performance of plasma and BALF mNGS

Clinical adjudication of plasma and BALF mNGS results

Microorganisms identified in patients with PDS

The associations between inflammatory markers and mNGS results

Application of mNGS for antibiotic adjustment

Factors influencing prognosis in patients with PDS

Discussion

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1