Utility of metagenomics next-generation sequencing in the diagnosis and treatment of severe infectious diseases in the intensive care unit

Abstract

BACKGROUND:

Metagenomic next-generation sequencing (mNGS) is a new method that combines high-throughput sequencing and bioinformatics analysis. However, it has not become as popular due to the limited testing equipment and high costs and lack of family awareness with not much relevant intensive care unit (ICU) research data.

OBJECTIVE:

To explore the clinical use and value of metagenomics next-generation sequencing (mNGS) in patients with sepsis in the ICU.

METHODS:

We conducted a retrospective analysis of 102 patients with sepsis admitted to the ICU of Peking University International Hospital from January 2018 to January 2022. Based on whether mNGS was performed, the identified patients were divided into the observation group ( $n=$ 51) and the control group ( $n=$ 51), respectively. Routine laboratory tests, including routine blood test, C-reactive protein, procalcitonin, and culture of suspicious lesion specimens were performed in both groups within 2 hours after admission to the ICU, while mNGS tests were performed in the observation group. Patients in both groups were routinely given initial anti-infective, anti-shock, and organ support treatment. Antibiotic regimens were optimized in a timely manner according to the etiological results. Relevant clinical data were collected.

RESULTS:

The testing cycle of mNGS was shorter than that of the conventional culture (30.79 $\pm$ 4.01 h vs. 85.38 $\pm$ 9.94 h, $P<$ 0.001), while the positive rate of mNGS was higher than that of the conventional culture (82.35% vs. 45.1%, $P<$ 0.05), with obvious superiority in the detection of viruses and fungus. There were significant differences in the optimal time of antibiotics (48 h vs.100 h) and length of ICU stay (11 d vs. 16 d) between the observation group and control group ( $P<$ 0.01) respectively, with no difference in 28-day mortality (33.3% vs. 41.2%, $P>$ 0.05).

CONCLUSION:

mNGS is useful in the detection of sepsis-causing pathogens in the ICU with the advantages of short testing time and high positive rate. There was no difference in the 28-day outcome between the two groups, which may be related to other confounding factors such as small sample size. Additional studies with extended sample size are needed.

Keywords

Intensive care unit microorganism culture mNGS sepsis health management

1. Introduction

With a growing aging population and the increasing incidence of chronic diseases and immunodeficiency diseases, sepsis is a major public health problem worldwide. The morbidity rate of sepsis is over 18 million per year, with a persistent rising tendency annually. The mortality rate of sepsis is as high as 30%, accounting for 20% of deaths worldwide [1]. Sepsis has become the major cause of hospitalization and death in the intensive care unit (ICU) setting [2]. Sepsis refers to a systemic inflammatory response syndrome caused by the invasion of pathogenic microorganisms including bacteria, fungus, viruses, parasites, etc. It manifests as abnormal pathological, physiological, and biochemical functions [3]. Common sites of infection include the lungs, blood, abdominal cavity, central nervous system, urinary tract, skin, and soft tissue. The condition of ICU patients with severe infection is complex, and multiple pathogens or multi-site infections can occur simultaneously leading to a strong inflammatory response and rapid disease progression. It is essential to determine the site of infection at the earliest, identify the characteristics of pathogenic microorganisms and provide appropriate initial anti-infective intervention to achieve successful treatment [4]. One of the most concerning and intractable problems for ICU clinicians is accessing early evidence of pathogenic microorganisms and providing precise anti-infective treatment. Conventional methods of detection mainly include routine blood tests, C-reactive protein (CRP), staining microscopy, conventional culture, and polymerase chain reaction (PCR). However, defects in sensitivity, specificity and periodicity exist in these methods to varying degrees. Moreover, these methods are susceptible to antibiotics, due to which they cannot effectively assist clinicians in the early stage, resulting in failure or delay of initial treatment. It has been reported that unreasonable initial anti-infective treatment is an important risk factor for 30-day all-cause mortality [5], which also leads to antibiotics abuse and microbial resistance, increases the risk of adverse reactions of combined drug use, and foists a heavy economic burden on patients. Therefore, it is crucial to detect pathogens with high specificity, high throughput, and rapid speed. Metagenomic next-generation sequencing (mNGS) is a new method that combines high-throughput sequencing and bioinformatics analysis, which has a high value in the diagnosis and treatment of ICU patients with severe infection [6, 7, 8]. It can not only improve the diagnostic accuracy and guide individualized treatment, but also play an important role in monitoring drug resistance and studying the dynamics of microbial communities. However, it has not become as popular due to the limited testing equipment and high costs and lack of family awareness with not much relevant ICU research data. In this study, we sought to investigate the utility of mNGS in the diagnosis and treatment of sepsis in the ICU.

2. Materials and methods

2.1 General information

We identified a total of 120 patients with sepsis admitted to the ICU of Peking University International Hospital from January 2018 to January 2022. Based on the exclusion criteria, 5 patients with incomplete clinical data, 8 patients with malignant tumors, and 5 patients lost to follow-up were excluded. There were 102 patients with sepsis who were finally included in the study. Among the 102 patients, there were 57 males and 55 females, with age ranging from 22 to 79 years.

Type of infection: There were 15 patients with conventional or perioperative intracranial infection after craniocerebral surgery, with a score of 3–9 points on the Glasgow Coma Scale (GCS), 20 patients with bloodstream infections from various sources, 34 patients with community-acquired or hospital-acquired severe pneumonia, while 21 patients had complicated abdominal infection, 8 patients had urinary tract infections, and lastly, 4 patients had skin and soft tissue infections.

2.2 Methods and observation indicators

2.2.1 Inclusion and exclusion criteria

Inclusion criteria: 1) Patients who met the Sepsis-3 diagnostic criteria jointly published by the Society of Critical Care Medicine (SCCM) and the European Society of Critical Care Medicine (ESICM); 2) The source of infection was not clear when admitted to the ICU; 3) Blood culture and mNGS were done simultaneously.

Exclusion criteria: 1) Blood culture and mNGS were not done simultaneously; 2) Sepsis following cardiopulmonary resuscitation; 3) Sepsis in end-stage chronic diseases; 4) Sepsis in malignant tumor; 5) Incomplete clinical data; 6) Prognosis within 28 days was not known.

At present, due to equipment and cost problems in the ICU, mNGS cannot be accepted by family members as a first-line detection method. Most patients refuse to be tested, or accept mNGS detection when the early treatment effect is not ideal. Therefore, the number of sepsis patients who can be tested by mNGS at the same time with other conventional detection methods in the early stage of ICU admission is small, resulting in a low sample size of the experimental group. The sample size of the control group was randomly selected from the patients who met the inclusion criteria according to the data of the experimental group.

This study was approved by the Research and Clinical Trials Ethics Committee of the Peking University International Hospital (No. 2022-KY-0004-01).

2.2.2 Sample collection

mNGS and conventional culture samples were collected within 2 hours after sepsis was diagnosed in the ICU. mNGS blood samples were submitted for testing immediately while conforming to the requirements of genetic testing labs (such as using an exclusive specimen collection tube, following aseptic procedures, and avoiding hemolysis). If samples could not be sent immediately, the samples were stored in a $-$ 4 ${{}^{\circ}}$ C refrigerator for not more than 24 hours. The microbial nucleic acid sequences of the samples were analyzed using high-throughput sequencing and identified by comparing the sample with the nucleic acid sequences of existing microorganisms in the database.

2.2.3 Observation indicators

Based on whether mNGS was performed or not, the included patients were divided into the observation group ( $n=$ 51) and the control group ( $n=$ 51) respectively. General clinical data in both groups were collected after admission to ICU, including age, gender, Acute Physiology and Chronic Health Evaluation (APACHE II) score, white blood cells, CRP, and procalcitonin (PCT). All patients were given empirical initial anti-infective treatment. The results of general culture and mNGS sequencing were recorded. The anti-infection regimens were adjusted in a timely manner and optimized. Statistical differences in timeliness and sensitivity between mNGS and conventional culture were analyzed. Meanwhile, differences in optimization time of the antibiotic regimen, length of ICU stay, and 28-day mortality were also observed between the two groups.

2.3 Statistical analysis

SPSS version 23.0 software was used for data analysis. Continuous variables conforming to normal distribution were represented as mean $\pm$ standard deviation (Mean $\pm$ SD) and analyzed by independent sample $t$ -test. Data conforming to non-normal distribution were represented by median [interquartile distance] (median [IQR]) and analyzed using Mann-Whitney U test. Quantitative data were analyzed by Pearson Chi-square test or Fisher exact test. X ${}^{2}$ test was used for comparison between groups, and $P<$ 0.05 was considered statistically significant.

3. Results

Observation group: In terms of gender composition, there were 28 males (54.9%), age was 60 years, APACHE II score was 21. Control group: In terms of gender composition, there were 29 males (56.9%), age was 58 years, APACHE II score was 21. There were no significant differences in general data and type of infection between the two groups ( $P>$ 0.05) (Table 1).

Table 1
Clinical characteristics and distribution of infection sites

Index	Observer group ( $n=$ 51)		Control group ( $n=$ 51)		Z/ $\chi^{2}$	$P$
Age	60	(24)	58	(23)	0.331	0.740
Sexuality (males $n$ %)	28	(54.9)	29	(56.9)	0.04	0.842
APACHE 2	21	(7)	21	(7)	0.419	0.675
WBC	15.5	(7.9)	15.09	(6.6)	0.917	0.359
CRP	166	(124.99)	201.35	(145.33)	$-$ 1.017	0.309
PCT	2.5	(11.44)	4.6	(5.18)	$-$ 1.784	0.074
Infection sites ( $n$ %)					7.423	0.191
Respiratory system	19	(37.2)	15	(29.4)
Bloodstream infection	9	(17.6)	11	(21.5)
Nervous system infection	10	(19.6)	5	(9.8)
Urinary system infection	2	(3.9)	6	(11.7)
Abdominal infection	8	(15.6)	13	(25.4)
Skin and soft tissue infection	3	(5.8)	1	(2.0)

The reporting time of mNGS was 30.79 $\pm$ 4.01 h, with a positive rate of 82.35% while the reporting time of conventional culture was 85.38 $\pm$ 9.94 h, with a positive rate of 44.11%. The differences in detection time and positive rate between the two groups were statistically significant ( $P<$ 0.05) (Table 2). In the observation group, 5 cases of mNGS results were judged to be background bacteria and miscellaneous bacteria, which were inconsistent with the clinical results, and 3 cases were negative, but the infection could not be excluded. The two groups had good consistency in the detection results of pathogenic microorganisms (Kappa $=$ 0.702, $p<$ 0.001) (Table 3).

Table 2

Comparison of positive rate and testing time between mNGS and bacterial culture

Index	mNGS ( $n=$ 51)	Bacterial culture ( $n=$ 102)	t/ $\chi^{2}$	$P$
Positive rate (%)	42 (82.35)	45 (44.11)	4.923	$P=$ 0.027
Testing time (h)	30.79 $\pm$ 4.01	85.38 $\pm$ 9.94	30.10	$P=$ 0.000

Table 3

Consistency evaluation of two test results

	Result	Progressive standard error	Approximate value T	Approximate value sig.
Consistency measure Kappa	0.702	0.092	7.050	0.000
Valid case N	102

Significance of the two detection methods with respect to the type of etiology: mNGS covered all microorganisms such as bacteria, viruses, and fungi, with obvious superiority in viral infections. Among the 51 patients in the observation group, there were 42 patients whose mNGS test results were positive. Among these, bacterial infection accounted for 73.8%, mainly including Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Staphylococcus. At the same time, mNGS also show a good advantage in the detection of bacterial resistance genes. A total of 16 cases of pathogenic bacteria containing drug resistance genes (including 6 strains of Acinetobacter baumannii, 5 strains of Enterobacteriaceae, 3 strains of Staphylococcus aureus and 2 strains of Enterococcus faecium) were reported. The detected drug resistance genotypes included, amongst others, blaOXA-23, blaOXA-51, blaCTX-M, blaKPC, blaZ, mecA, mecR1. Viral infection accounted for 19%, and included human herpesvirus IV (EBV), cytomegalovirus (CMV), and porcine herpesvirus type I. Fungal infections accounted for 7.2%, including Candida albicans (Candida albicans, Candida glabrata, and Candida tropicalis) and Rhizopus microspora. It must be noted that conventional culture cannot effectively identify viruses (Fig. 1).

Figure 1.

The positive detection rate of mNGS and culture for different pathogens.

Figure 2.

The positive detection rate of mNGS and culture in different specimen types.

mNGS showed a good positive rate in all types of specimens, especially in urine and pus samples with a positive rate of up to 100%. The sensitivity of mNGS in sterile body fluids such as blood and cerebrospinal fluid was significantly higher than that of conventional culture ( $P<$ 0.05). mNGS was superior in the diagnosis of severe bloodstream infections and central nervous system infections (Fig. 2).

The time of optimal antibiotics regimens, the length of ICU stay, and 28-day outcome of the two groups were compared. The time of optimal antibiotics regimens was 48 h in the observation group and 100 h in the control group, showing a statistically significant difference ( $P<$ 0.01). The length of ICU stay was 11 d in the observation group and 16 d in the control group, also showing a statistically significant difference ( $P<$ 0.01). There was no significant difference in 28-day outcomes between the two groups ( $P>$ 0.05) (Table 4, Fig. 3).

Table 4

Comparison of antibiotic optimization time, ICU hospitalization time and 28-day mortality between observation group and control group

Index	Observation group ( $n=$ 51)		Control group ( $n=$ 51)		Z/ $\chi^{2}$	$P$
Antibiotic optimization time (h)	48	(26)	100	(30)	$-$ 6.076	0.000
ICU hospitalization time (d)	11	(7)	16	(6)	$-$ 4.725	0.000
28-day mortality (%)	17	(33.3%)	21	(41.2%)	0.671	0.413

Figure 3.

Observation and control group of 28 days survival curve (2).

Figure 4.

Comparison of duration stay in ICU and 28-day mortality between negative group and positive group.

Among the 102 patients with sepsis, 28 cases had negative etiological results, including 8 cases in the observation group (5 cases of mixed bacteria and 3 cases of negative mNGS results), 20 cases in the control group had negative culture results, and the remaining 74 cases had positive etiological results. The length of ICU stay and 28-day mortality were compared in patients with negative and positive results. We found that the length of ICU stay was 16.93 $\pm$ 3.839 d in the negative group and 10.36 $\pm$ 2.475 d in the positive group, respectively, with a significant difference ( $\chi^{2}=$ 10.18, $P<$ 0.001). Likewise, the difference in 28-day mortality between the negative group and positive group was also statistically significant ( $\chi^{2}=$ 5.204, $P<$ 0.05) (Fig. 4).

4. Discussion

Sepsis has become one of the independent risk factors for ICU death, and one of the challenges in treatment in the ICU. Along with adequate fluid resuscitation and organ support, early identification of the pathogens and active and appropriate antibiotics (correct initial treatment $+$ precise sequential treatment) are still the cornerstone of successful treatment [9], which is crucial for prognosis. Guidelines also point out that delayed or inappropriate initial treatment are important risk factors that increase mortality [10]. It is estimated that about 35% of deaths are related to inappropriate antibiotic treatment, and about 50% of antibiotics used are considered unnecessary or have an ultra-broad spectrum coverage, thus inducing drug resistance. At the time of administering initial antibiotic treatment, namely empirical anti-infective treatment, there must be a comprehensive analysis of the possible source, nature, and site of infection, and local epidemiological characteristics, to identify potential pathogens to the maximum extent possible. At the same time, drug-drug interactions and damage to organ function should also be considered. While determining the initial treatment plan, etiological examination should be completed at the earliest to identify the pathogens quickly, and adjust and optimize the antibiotic regimens for precise treatment. However, conventional laboratory methods of detection have their limitations. For example, the detection cycle in a conventional culture is relatively long, especially for special bacteria such as Mycobacterium tuberculosis, which requires an extended culture time. The positive rate of Cryptococcus and Brucella culture is low, while viral cultures are mostly negative [11]. Although it has improved the pathogen diagnosis rate, polymerase chain reaction (PCR) still has a narrow detection target, delayed positive results, and a high false positive rate. Therefore, the conventional detection methods that are currently used, often are unable to identify the nature of infection in the early stage and cannot effectively guide the direction of subsequent treatment, eventually leading to the delay of treatment or antibiotics abuse.

As a novel method for pathogen diagnosis, mNGS has unique advantages in the identification, classification, and detection of drug-resistant mutations and identification of novel pathogens [12]. By extracting nucleic acids from samples obtained from the site of infection, it combines high-throughput sequencing and bioinformatics analysis to realize rapid whole genome (DNA or RNA) sequencing. mNGS can accurately identify bacteria, fungus, viruses, parasites and other atypical pathogens by comparing the sample with the pathogen genome database and analysis [13]. The advantages of mNGS include a short detection cycle, high positive rate, and high specificity. The use of mNGS in infectious diseases has been gradually increasing, especially in rare diseases and infectious diseases with difficulties in definite diagnosis, such as viruses, mycobacterium tuberculosis, anaerobic bacteria, and fungi [14], providing the etiological basis for more targeted drug treatment in patients with severe infections. Because of the unbiased characteristics, mNGS not only covers the genome of microorganisms, but also covers the plasmids carrying drug resistance/virulence genes. While identifying drug-resistant bacteria, the analysis of drug resistance genes can guide the diagnosis and treatment of some patients [15]. However, due to the limited testing equipment, relatively high cost and payment methods, mNGS has not yet become a first-line clinical testing method although the consensus clearly recommends it [16], and is rarely studied in ICU-related research. Therefore, in this study we analyzed and discussed the positive rate of mNGS in different samples and pathogenic microorganisms, and its influence on treatment and prognosis in ICU patients with sepsis.

A total of 102 patients with sepsis were included in this study. Among them, both mNGS and conventional culture were used in 51 patients in the observation group, and only conventional culture was used in 51 patients in the control group. The detection cycle of mNGS was significantly shorter than that of conventional culture ( $\chi^{2}=$ 30.1, $P<$ 0.001), which provided the etiological basis for early precise anti-infective treatment. The positive rate of mNGS was significantly higher than that of conventional culture (82.35% vs. 44.11%), with a statistically significant difference ( $\chi^{2}=$ 4.923, $P<$ 0.05). Among the positive mNGS results, in terms of strain distribution, common bacteria accounted for 73%. The top five ranking pathogens were as follows: Klebsiella pneumoniae, Enterococcus faecium, Acinetobacter baumannii, Pseudomonas aeruginosa, and Escherichia coli/Staphylococcus aureus. The results of conventional culture were similar to that of mNGS apart from a slightly lower proportion of Enterococcus faecium, and this finding was consistent with the current epidemiology of hospital-associated infections [17]. In addition, mNGS also performed well in the case of pathogens that are difficult to be detected or have a long detection cycle in conventional methods, such as viruses (which accounted for about 19%, including various types of human herpes virus and swine herpes virus) and Candida (which accounted for about 8%), to name a few. Among different specimen types, the positive rate of mNGS in pus and urine samples was up to 100%, which was significantly superior to that of conventional culture, but this may be related to the small numbers of such cases in this study. The positive rates of mNGS in alveolar lavage fluid, blood, peritoneal drainage fluid, and cerebrospinal fluid were 86.7%, 73.3%, 71.4%, and 54.4%, respectively. The pathogen coverage rate in cerebrospinal fluid was the lowest in mNGS, which was consistent with that in conventional culture, reaching only 33%. These findings, consistent with previous reports [18, 19], demonstrate the challenges in the diagnosis of infections of the nervous system. However, mNGS still has obvious advantages over conventional culture. In the observation group, a patient with Eisenmenger syndrome complicated with brain abscess (Staphylococcus), a patient with multi-drug-resistant Acinetobacter baumannii following pituitary cyst surgery, and a rare case of herpes porcine encephalitis were diagnosed using mNGS cerebrospinal fluid genome sequencing, respectively (Fig. 5). It is worthwhile to note that the conventional culture of cerebrospinal fluid of all these three patients were negative. This highlights the clinical advantages of mNGS in identifying specific infection sites and rare pathogens [20]. The above results indicate that mNGS has a shorter cycle, a higher positive rate, and covers a wide range of disease pathogens when compared to conventional culture, showing obvious advantages in detection of virus infections and rare pathogens, and functioning as a better clinical guide.

Figure 5.

Eisenmenger syndrome with brain abscess, MDR-AB infection after pituitary cyst surgery and porcine herpesvirus type iencephalitis.

Results showed that the major cause of sepsis in ICU patients in our hospital was severe pneumonia (accounting for about 34%), followed by abdominal infection (accounting for about 21%), and bloodstream infection ranked third (accounting for 18%). On receipt of the positive results, antibiotics were adjusted in time and optimized in both groups. The optimal treatment time was 48 h in the observation group and 100 h in the control group, respectively, with a statistically significant difference ( $P<$ 0.001), indicating that the observation group was able to receive a precise treatment plan guided by the etiological results obtained faster through mNGS, which avoided the adverse outcome that could have resulted from the initial treatment failure. Moreover, the time of empirical broad-spectrum anti-infective treatment was shortened, the drug resistance induced by antibiotic abuse was reduced, and the risk of adverse reactions caused by combined drug use was lowered, thereby improving the rational application of antibiotics. These findings are consistent with the report of Zhou et al. [21, 22]. Our results also showed an apparently shorter length of ICU stay in the observation group than that in the control group (11 d vs. 16 d), with a statistically significant difference ( $P<$ 0.01). However, there was no significant difference in the 28-day outcome between the two groups ( $P>$ 0.05), which may be related to other confounding factors leading to clinical death (such as patient-specific factors, severe comorbidities, and small sample size). In this study, we also observed the occurrence of complications in the two groups. All patients had related complications after admission. Different degrees of liver injury (10 cases in the observation group and 15 cases in the control group), acute kidney injury required CRRT (8 cases in the observation group and 12 cases in the control group); acute coronary syndrome (8 cases in the observation group and 7 cases in the control group). The damage of liver and kidney function in the control group was more common than that in the observation group. The reason may be related to the infection itself and the high intensity of antibiotic use in the observation group. but the condition is controllable and has no effect on the mortality rate. Among the 102 patients with sepsis in the two groups, there were still 28 patients with unknown sources of infection. There were significant differences in terms of duration of ICU stay and 28-day mortality between patients with positive and negative results. Being the basis of clinical treatment, the detection of pathogenic microorganisms can improve the final outcome by influencing clinical treatment strategies [23]. mNGS, in particular, offers more possibilities for clinicians to obtain pathogenic evidence, especially in severe cases [24, 25].

At present, the rapid development of medical technology and equipment is of great help to clinicians [26], especially in the ICU of major hospitals, with the most advanced and comprehensive diagnosis and treatment equipment, to provide the most accurate diagnosis and treatment monitoring for critically ill patients, as far as possible to help save the lives of patients. although there is a great demand for mNGS monitoring technology in clinical practice, due to reasons such as capital, staffing and technology, there are few mNGS detection equipment in domestic general hospitals. most of them are completed in cooperation with third-party testing institutions with mature and reliable technology and qualification audit, so as to ensure the scientificity of the testing process and the accuracy of the testing results, and eliminate the treatment risk and damage caused by measurement error and uncertainty to patients [27, 28, 29]. The premise of effective application of mNGS detection methods is to select appropriate patients, use appropriate specimens and make reasonable interpretation [30]. It is suggested that it can be used as a first-line detection method in patients with immunodeficiency, severe patients and suspected explosive infectious diseases. At the same time, it is recommended as a supplementary means when the clinical treatment effect is poor and the pathogen infection is unknown. It is recommended to use specimens to select suspected infection sites with high pathogenic microbial load and relatively small impact on non-pathogenic microorganisms, such as alveolar lavage fluid. In addition, the interpretation of mNGS results is very important for us [31]. First, we can refer to the number of sequences of the pathogenic bacteria, but at the same time, we should consider the coverage, specificity and conservation of the sequence on the genome. For pathogenic bacteria that are difficult to detect, even if the number of sequences is low, they can be judged as positive, such as viruses and Brucella. For environmental bacteria and symbiotic bacteria with high sequence numbers, since mNGS cannot distinguish colonization or infection, it is necessary to first exclude conditional pathogenic bacteria by combining pathogens, samples and clinical characteristics.which requires comprehensive judgment or multidisciplinary discussion (ICU, laboratory, pharmacy, infection department). It should be noted that mNGS cannot be used as the only evidence for clinical decision-making. It is also necessary to avoid the failure of clinical treatment due to the wrong interpretation of the results.

As an increasingly sophisticated molecular diagnostic method, mNGS has been widely used in the diagnosis, treatment, and monitoring of infectious diseases in a big way. It has been reported that mNGS can establish a sepsis diagnosis model by integrating host and microbial characteristics, predict the incidence of sepsis, and greatly advance the diagnosis time [32, 33, 34]. At present, mNGS has not been widely accepted by patients and their families due to the limitation of testing equipment and relatively high cost, resulting in the lack of large-sample studies on the use of mNGS in ICU patients and its influence on clinical prognosis [35]. However, it is believed that with mNGS detection technology continuously evolving, in particular, the addition of artificial intelligence technology in the future will help to further improve the efficiency and accuracy of gene detection, while reducing the cost of detection from the side, greater awareness among patients about this technique, it is highly probable that mNGS may become a routine detection method [36]. We recommend that patients with sepsis in the ICU undergo a combination of investigations using mNGS, conventional cultures, other molecular biology, and serological tests to arrive at accurate diagnosis of infectious diseases and obtain better clinical benefits.

With the promotion of mNGS technology, its application value will not only be limited to the diagnosis and personalized precision medicine of patients with severe infection and tumor in hospital, but also be gradually applied to pre-hospital diagnosis and even early prevention and health care [37, 38]. For example, genetic testing is used to understand the genetic susceptibility of individuals to certain diseases, such as cancer and some chronic diseases, so as to carry out targeted health management and reduce the risk of illness. It can help people realize early knowledge, early prevention and early treatment of diseases, and has a certain predictive effect on the development trend of diseases.

5. Conclusion

In short, mNGS will run through the entire stage of human health management from prevention to treatment, and will also receive the attention of clinicians and more people.

Competing interests

The authors declare that they have no competing interests.

Funding

No external funding was received to conduct this study.

Ethics statement

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Ethics Committee of the Peking University International Hospital. Written informed consent was obtained from all participants (Approval no. 2022-KY-0004-01).

Footnotes

Acknowledgments

The authors are particularly grateful to everyone who helped them with the article.

References

Rudd

Johnson

Agesa

Shackelford

Tsoi

Kievlan

Colombara

Ikuta

Kissoon

Finfer

Fleischmann-Struzek

Machado

Reinhart

Rowan

Seymour

Watson

West

Marinho

Hay

Lozano

Lopez

Angus

Murray

CJL

Naghavi

. Global, regional, and national sepsis incidence and mortality, 1990–2017: Analysis for the Global Burden of Disease Study. Lancet. 2020 Jan 18; 395(10219): 200-211. doi: 10.1016/S0140-6736(19)32989-7. PMID: 31954465; PMCID: PMC6970225.

Abe

Ogura

Shiraishi

Kushimoto

Saitoh

Fujishima

Mayumi

Shiino

Nakada

Tarui

Hifumi

Otomo

Okamoto

Umemura

Kotani

Sakamoto

Sasaki

Shiraishi

Takuma

Tsuruta

Hagiwara

Yamakawa

Masuno

Takeyama

Yamashita

Ikeda

Ueyama

Fujimi

Gando

; JAAM FORECAST group. Characteristics, management, and in-hospital mortality among patients with severe sepsis in intensive care units in Japan: The FORECAST study. Crit Care. 2018 Nov 22; 22(1): 322. doi: 10.1186/s13054-018-2186-7. PMID: 30466493; PMCID: PMC6251147.

Singer

Deutschman

Seymour

Shankar-Hari

Annane

Bauer

Bellomo

Bernard

Chiche

Coopersmith

Hotchkiss

Levy

Marshall

Martin

Opal

Rubenfeld

van der Poll

Vincent

Angus

. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016 Feb 23; 315(8): 801-10. doi: 10.1001/jama.2016.0287. PMID: 26903338; PMCID: PMC4968574.

Niederman

Baron

Bouadma

Calandra

Daneman

DeWaele

Kollef

Lipman

Nair

. Initial antimicrobial management of sepsis. Crit Care. 2021 Aug 26; 25(1): 307. doi: 10.1186/s13054-021-03736-w. PMID: 34446092; PMCID: PMC8390082.

Holmbom

Andersson

Berg

Eklund

Sobczynski

Wilhelms

Moberg

Fredrikson

Balkhed

ÅÖ

Hanberger

. Prehospital delay is an important risk factor for mortality in community-acquired bloodstream infection (CA-BSI): A matched case-control study. BMJ Open. 2021 Nov 18; 11(11): e052582. doi: 10.1136/bmjopen-2021-052582. PMID: 34794994; PMCID: PMC8603295.

Wang

Miao

Pan

Jin

Zhang

Yao

Huang

Wang

Cai

Luo

Zhou

. The clinical value of metagenomic next-generation sequencing in the microbiological diagnosis of skin and soft tissue infections. Int J Infect Dis. 2020 Nov; 100: 414-420. doi: 10.1016/j.ijid.2020.09.007. Epub 2020 Sep 6. PMID: 32898669.

Thoendel

Jeraldo

Greenwood-Quaintance

Yao

Chia

Hanssen

Abdel

Patel

. Identification of prosthetic joint infection pathogens using a shotgun metagenomics approach. Clin Infect Dis. 2018 Oct 15; 67(9): 1333-1338. doi: 10.1093/cid/ciy303. PMID: 29648630; PMCID: PMC6186856.

Brown

Bharucha

Breuer

. Encephalitis diagnosis using metagenomics: Application of next generation sequencing for undiagnosed cases. J Infect. 2018 Mar; 76(3): 225-240. doi: 10.1016/j.jinf.2017.12.014. Epub 2018 Jan 2. PMID: 29305150; PMCID: PMC7112567.

Niederman

Baron

Bouadma

Calandra

Daneman

DeWaele

Kollef

Lipman

Nair

. Initial antimicrobial management of sepsis. Crit Care. 2021 Aug 26; 25(1): 307. doi: 10.1186/s13054-021-03736-w. PMID: 34446092; PMCID: PMC8390082.

10.

Ferrer

Martin-Loeches

Phillips

Osborn

Townsend

Dellinger

Artigas

Schorr

Levy

. Empiric antibiotic treatment reduces mortality in severe sepsis and septic shock from the first hour: Results from a guideline-based performance improvement program. Crit Care Med. 2014 Aug; 42(8): 1749-55. doi: 10.1097/CCM.0000000000000330. PMID: 24717459.

11.

Cai

Miao

Song

Fang

. High-throughput metagenomics for identification of pathogens in the clinical settings. Small Methods. 2021 Jan 4; 5(1): 2000792. doi: 10.1002/smtd.202000792. Epub 2020 Dec 13. PMID: 33614906; PMCID: PMC7883231.

12.

Blauwkamp

Thair

Rosen

Blair

Lindner

Vilfan

Kawli

Christians

Venkatasubrahmanyam

Wall

Cheung

Rogers

Meshulam-Simon

Huijse

Balakrishnan

Quinn

Hollemon

Hong

Vaughn

Kertesz

Bercovici

Wilber

Yang

. Analytical and clinical validation of a microbial cell-free DNA sequencing test for infectious disease. Nat Microbiol. 2019 Apr; 4(4): 663-674. doi: 10.1038/s41564-018-0349-6. Epub 2019 Feb 11. PMID: 30742071.

13.

Evans

Rhodes

Alhazzani

Antonelli

Coopersmith

French

Machado

Mcintyre

Ostermann

Prescott

Schorr

Simpson

Wiersinga

Alshamsi

Angus

Arabi

Azevedo

Beale

Beilman

Belley-Cote

Burry

Cecconi

Centofanti

Coz Yataco

De Waele

Dellinger

Doi

Estenssoro

Ferrer

Gomersall

Hodgson

Møller

Iwashyna

Jacob

Kleinpell

Klompas

Koh

Kumar

Kwizera

Lobo

Masur

McGloughlin

Mehta

Mer

Nunnally

Oczkowski

Osborn

Papathanassoglou

Perner

Puskarich

Roberts

Schweickert

Seckel

Sevransky

Sprung

Welte

Zimmerman

Levy

. Surviving sepsis campaign: International guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021 Nov; 47(11): 1181-1247. doi: 10.1007/s00134-021-06506-y. Epub 2021 Oct 2. PMID: 34599691; PMCID: PMC8486643.

14.

Miao

Wang

Pan

Zhang

Jin

Yao

Huang

Wang

Zhou

. Microbiological diagnostic performance of metagenomic next-generation sequencing when applied to clinical practice. Clin Infect Dis. 2018 Nov 13; 67(suppl_2): S231-S240. doi: 10.1093/cid/ciy693. PMID: 30423048.

15.

Serpa

Deng

Abdelghany

Crawford

Malcolm

Caldera

Fung

McGeever

Kalantar

Lyden

Ghale

Deiss

Neff

Miller

Doernberg

Chiu

DeRisi

Calfee

Langelier

. Metagenomic prediction of antimicrobial resistance in critically ill patients with lower respiratory tract infections. Genome Med. 2022 Jul 12; 14(1): 74. doi: 10.1186/s13073-022-01072-4. PMID: 35818068; PMCID: PMC9275031.

16.

Consensus Expert Group on Clinical application of metagenomics sequencing technology in moderate and severe infections, Sepsis and Shock Professional Committee of Chinese Research Hospital Society, Microtoxin Professional Committee of Chinese Society of Microbiology, et al. Expert consensus for the application of metagenomic next generation sequencing in the pathogen diagnosis in clinical moderate and severe infections (first edition). Chinese Critical Care Emergency Medicine. 2020; 32(5): 531-536.

17.

Geng

Mei

Zhu

Fang

Yang

Zhang

Fan

Pan

. Metagenomic next-generation sequencing technology for detection of pathogens in blood of critically ill patients. Int J Infect Dis. 2021 Feb; 103: 81-87. doi: 10.1016/j.ijid.2020.11.166. Epub 2020 Nov 20. PMID: 33227513.

18.

Brouwer

Thwaites

Tunkel

van de Beek

. Dilemmas in the diagnosis of acute community-acquired bacterial meningitis. Lancet. 2012 Nov 10; 380(9854): 1684-92. doi: 10.1016/S0140-6736(12)61185-4. PMID: 23141617.

19.

Polage

Cohen

. State-of-the-art microbiologic testing for community-acquired meningitis and encephalitis. J Clin Microbiol. 2016 May; 54(5): 1197-202. doi: 10.1128/JCM.00289-16. Epub 2016 Feb 17. PMID: 26888896; PMCID: PMC4844713.

20.

Zheng

Lin

Zhang

Fang

Chen

Pan

Lin

. Value of metagenomic next-generation sequencing in children with severe infectious diseases. Zhongguo Dang Dai Er Ke Za Zhi. 2022 Mar 15; 24(3): 273-278. English, Chinese. doi: 10.7499/j.issn.1008-8830.2110003. PMID: 35351257; PMCID: PMC8974643.

21.

Zhou

Larkin

PMK

Zhao

Yao

Wang

Zhou

Wang

Feng

Chen

Zhou

Hua

Zhou

Yang

. Clinical impact of metagenomic next-generation sequencing of bronchoalveolar lavage in the diagnosis and management of pneumonia: A multicenter prospective observational study. J Mol Diagn. 2021 Oct; 23(10): 1259-1268. doi: 10.1016/j.jmoldx.2021.06.007. Epub 2021 Jun 29. PMID: 34197923.

22.

Liu

. A study on the mechanism of the protective effect of GuangeFang on sepsis-associated acute kidney injury. World J Tradit Chin Med. 2021; 7: 414-8.

23.

Ullah

Hussain

Ali

Samad

, Shah

Yousef

Khan

. A prospective observational study assessing the outcome of Sepsis in intensive care unit of a tertiary care hospital, Peshawar. Pak J Med Sci. 2016 May-Jun; 32(3): 688-93. doi: 10.12669/pjms.323.9978. PMID: 27375715; PMCID: PMC4928424.

24.

Liu

Yen

Liu

Bhan

Chi

Sheng

. Clinical application of metagenomic next-generation sequencing in patients with hematologic malignancies suffering from sepsis. Microorganisms. 2021 Nov 6; 9(11): 2309. doi: 10.3390/microorganisms9112309. PMID: 34835435; PMCID: PMC8624204.

25.

Gai

Zhang

Cheng

Cui

Wang

. Metagenomic next-generation sequencing for the microbiological diagnosis of abdominal sepsis patients. Front Microbiol. 2022 Feb 2; 13: 816631. doi: 10.3389/fmicb.2022.816631. PMID: 35185847; PMCID: PMC8847725.

26.

Badnjevic

. Evidence-based maintenance of medical devices: Current shortage and pathway towards solution. Technology and Health Care. 2023; 31(1): 293-305.

27.

Badnjevic

Almir

, et al. A Novel Method for Conformity Assessment Testing of Mechanical Ventilators for Post-market Surveillance Purposes. 2023; 31(1): 367-376.

28.

Liu

Wang

Weihan

Wang

Tian

, et al. Application of the data mining algorithm in the clinical guide medical records. World J Tradit Chin Med. 2022; 8: 548-55.

29.

Badnjević

Pokvić

Deumić

Bećirović

. Post-market surveillance of medical devices: A review. Technol Health Care. 2022; 30(6): 1315-1329. doi: 10.3233/THC-220284. PMID: 35964220.

30.

Chinese Thoracic Society. Consensus of clinical pathways of metagenomic next-generation sequencing test in diagnosis of lower respiratory tract infections in China. Zhonghua Jie He He Hu Xi Za Zhi. 2023 Apr 12; 46(4): 322-335. Chinese. doi: 10.3760/cma.j.cn112147-20220701-00553. PMID: 36787941.

31.

Laboratory Medicine Branch of Chinese Medical Association. Expert consensus on clinical standardized application of metagenomics next-generation sequencing for detection of pathogenic microorganisms. Chinese Journal of Laboratory Medicine. 2020; 43(Issue 12): 1181-1195.

32.

Kalantar

Neyton

Abdelghany

Mick

Jauregui

Caldera

Serpa

Ghale

Albright

Sarma

Tsitsiklis

Leligdowicz

Christenson

Liu

Kangelaris

Hendrickson

Sinha

Gomez

Neff

Pisco

Doernberg

Derisi

Matthay

Calfee

Langelier

. Integrated host-microbe plasma metagenomics for sepsis diagnosis in a prospective cohort of critically ill adults. Nat Microbiol. 2022 Nov; 7(11): 1805-1816.

33.

Sharean

Johncy

. Deep learning models on Heart Disease Estimation – A review. Journal of Artificial Intelligence. 2022; 4(2): 122-130.

34.

Sujithra

Rathika

Velvadivu

Marimuthu

. An intellectual decision system for classification of mental health illness on social media using computational intelligence approach. Journal of Ubiquitous Computing and Communication Technologies. 2023; 5(1): 23-35.

35.

Sun

Zhang

Yang

Wang

Sun

. Clinical Application and Influencing Factor Analysis of Metagenomic Next-Generation Sequencing (mNGS) in ICU Patients With Sepsis. Front Cell Infect Microbiol. 2022 Jul 13; 12: 905132. doi: 10.3389/fcimb.2022.905132. PMID: 35909965; PMCID: PMC9326263.

36.

Gams

, et al. Artificial Intelligence and Ambient Intelligence. 1 Jan. 2019: 71-86.

37.

Yuan

Wang

Zhao

Mao

Wang

Zhou

Jin

Zhang

Liang

Jiang

Zhang

Song

Yang

Guo

Zhang

Zhou

Wang

Qiu

Jiang

Zhang

. Early identification and severity prediction of acute respiratory infection (ESAR): A study protocol for a randomized controlled trial. BMC Infect Dis. 2022 Jul 20; 22(1): 632.

38.

Badnjevi

Avdihodži

Gurbeta Pokvi

. Artificial intelligence in medical devices: Past, present and future. Psychiatr Danub. 2021 May; 33(Suppl 3): S336-S341. PMID: 34010259.

Utility of metagenomics next-generation sequencing in the diagnosis and treatment of severe infectious diseases in the intensive care unit

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Materials and methods

2.1 General information

2.2 Methods and observation indicators

2.2.1 Inclusion and exclusion criteria

2.2.2 Sample collection

2.2.3 Observation indicators

2.3 Statistical analysis

3. Results

Table 1 Clinical characteristics and distribution of infection sites

Competing interests

Funding

Ethics statement

Footnotes

Acknowledgments

References

Table 1
Clinical characteristics and distribution of infection sites