Clinical characteristics and management of Listeria monocytogenes meningitis in children beyond the neonatal stage: a 10 years retrospective study

Abstract

Introduction

The data of Listeria monocytogenes (LM) meningitis in children beyond the neonatal stage has been limited. We aimed to summarize the clinical characteristics, management, and risk factors of neurological complications in LM meningitis children beyond the neonatal stage.

Methods

We retrospectively reviewed LM meningitis cases from January 2013 to December 2022 at Beijing Children’s Hospital. Clinical characteristics, pathogen detection results and management were analyzed.

Results

There were 41 LM meningitis patients at our center, with a median age of 2.3 years (ranging from 6 months to 9 years). Most patients (97.6%) were immunocompetent. Fourteen patients (34.1%) had a history of suspected food contamination. The most common symptom was fever (100%), and 29.2% of patients presented with diarrhea in the early stages of the disease. About 61% of patients showed monocyte predominance in their cerebrospinal fluid (CSF). Thirteen patients (31.7%) experienced neurological complications. Multivariate analysis indicated that a diagnosis delay of more than one week and a CRP level of 50 mg/L or higher were significant risk factors for these neurological complications (p < 0.05). CSF culture rates were much higher before hospital admission (85.7%) compared to after (31.7%, p < 0.05). Metagenomic next-generation sequencing (mNGS) identified pathogens in 3 culture-negative cases. In total, 97.5% of patients received meropenem, either alone or with other antibiotics, and all children recorded a Glasgow Outcome Scale (GOS) score of 5.

Conclusion

LM meningitis can affect immunocompetent children. Strengthening food hygiene and safety education is crucial to prevent LM infection. Penicillin or ampicillin are the preferred treatments, while meropenem may be considered as an alternative treatment.

Introduction

Listeria monocytogenes (LM) is a common gram-positive bacterium that causes listeriosis, a foodborne illness transmitted through contaminated foods such as fish, meat, fruits, vegetables, and cheeses, amongst others [1]. Following the ingestion of LM-contaminated foodstuffs, the LM traverses the intestinal epithelium and gains access to target organs through the lymphatic system and bloodstream [2]. Furthermore, it is important to note that the LM can breach the blood-brain barrier, resulting in the development of meningitis.

The newborns and immunocompromised children are at heightened risk for LM meningitis [3]. A nationwide surveillance study in Denmark demonstrated that the peak annual incidence of LM meningitis in neonates was 0.61/100,000 live births [4]. The incidence of LM meningitis is lower in non-neonatal children than in neonates. In Danish children (aged from 1 month to 17 years), the incidence rate of LM meningitis for the period 2000–2017 was 0.024 per 100,000 children [4]. Since 2000, China has invested significant efforts in the management and control of listeriosis, with a view to enhancing the surveillance system for this disease. In China, the incidence of LM infection was found to be low, with only sporadic cases reported. From 2008 to 2017, only 759 cases of LM infection were reported, of which approximately 50% were newborns, and the remaining non-neonatal children were less than 8% [5]. Listeriosis is an uncommon condition in children beyond the neonatal age, and the clinical presentation is comparable to that of other bacterial meningitis. It has been demonstrated that misdiagnoses are prevalent in non-neonatal children in the early stages of the disease.

In the present study, a retrospective analysis was performed on clinical data and outcomes of 41 children beyond the neonatal stage diagnosed with LM meningitis. The objective of this study is to provide empirical support for the diagnosis and management of LM meningitis in children.

Materials and methods

Patient recruitment

The subjects of this study were children with LM meningitis, aged between 29 days and 18 years, who were enrolled from January 2013 to December 2022 at Beijing Children’s Hospital. LM meningitis was defined by the following criterion [6]: (1) One or more of the following clinical symptoms: fever, headache, positive meningeal irritation sign, and consciousness change; (2) At least one of the following findings in the cerebrospinal fluid (CSF): elevated CSF leukocytosis (10 × 10⁶/L), decreased CSF glucose ( < 2.2 mmol/L) or elevated CSF protein ( > 450 mg/L); (3) Blood or CSF culture containing LM, or the LM-positive result in mNGS of the CSF. The final study population comprised forty-one patients with complete data.

The subsequent data were obtained retrospectively from each patient’s electronic medical records: (1) fundamental characteristics (e.g., age, gender); (2) clinical symptoms (e.g., fever, vomiting, diarrhea, convulsions); (3) history of consuming contaminated foodstuffs (the food is easily contaminated with LM, e.g. refrigerated dairy products, cheese, refrigerated cooked food, raw meat or seafood, and so on [1]); (4) acute period (within 1 week of symptom onset) laboratory tests (e.g., peripheral blood C-reactive protein [CRP], blood WBC count); (5) the first CSF test (e.g. CSF WBC count, CSF protein, CSF glucose); (6) pathogen information (e.g., pathogen identification method, pathogen identification time); (7) treatments (e.g., antibiotics and/or glucocorticoids); (8) neurological complications (include short-term complications such as focal neurological deficits, subdural effusions, ependymitis, brain abscess, cerebral infarction, etc.; long-term complications such as hearing loss, cognitive impairment, hydrocephalus, learning disability, epilepsy, etc.)

Prognostic evaluation

The long-term outcome was evaluated by telephone interviews and follow-up records. The telephone interviews were conducted from January 1, 2023 to January 31, 2023. The follow-up time was at least 12 months post-discharge. The outcomes were assessed with the Glasgow Outcome Score (GOS) [7]. The GOS was graded as follows: 5: good recovery, might have slight physical or mental impairments; 4: moderate disability, unable to resume previous activities; 3: severe disability, conscious but unable to independently carry out daily activities; 2: vegetative state; 1: death. A GOS of 5 was defined as a good prognosis, while a GOS of 1–4 was described as a poor prognosis.

Detection of metagenomic next-generation sequencing (mNGS)

Five patients underwent mNGS of CSF at our institution. During routine lumbar puncture, 1 mL of CSF was collected for mNGS analysis. The protocol comprised the following steps: (1) Total DNA was extracted directly from CSF samples using a commercial DNA extraction kit. DNA libraries were constructed via PCR amplification, with quality control performed using an Agilent 2100 Bioanalyzer (Agilent Technologies, USA) combined with PCR. Sequencing was carried out on the BGISEQ-100 platform (BGI-Tianjin, China). (2) Raw sequencing data were processed by removing short reads (35 bp), low-quality reads, and low-complexity reads. Clean reads were aligned to the human reference genome (hg19 and YH sequences) using Burrows-Wheeler Aligner (BWA, v.22). The remaining non-human reads were compared against microbial reference genomes from the National Center for Biotechnology Information (NCBI; https://www.ncbi.nlm.nih.gov/genomes). Sequencing depth and coverage for each species were calculated using SOAP coverage (http://soap.genomics.org.cn). The data of mNGS performed prior to admission was obtained through electronic medical records.

Statistical analysis

The subsequent processing and analysis of the data were conducted utilising SPSS Statistics 23.0. Categorical variables were presented as frequencies and proportions, while continuous variables were shown as the mean and standard deviation or median (interquartile range [IQR]). Categorical variables were analyzed using the Chi-square or Fisher’s exact tests as appropriate, while the continuous variables were compared using the Mann–Whitney U test. The risk factors associated with the extended antibiotic duration were evaluated using univariable logistic regression. When considering factors with a p-value < 0.05, multivariable logistic regression was performed. The odds ratio (OR) and confidence interval at 95% (CI 95%) were presented. P-value < 0.05 was considered significant.

Results

Clinical characteristics and examination

The study comprised 41 LM meningitis children at Beijing Children’s Hospital from January 2013 to December 2022 (Table 1). The age at onset of the patients varied from 6 months to 9 years, with a median age of 2.3 years. The analysis revealed that a single patient exhibited primary humoral immunodeficiency, while the remaining patients were immunocompetent. The data demonstrated that 70% (29/41) of cases occur during the spring and summer (spring: 15 cases, summer: 14 cases, autumn: 9 cases, winter: 3 cases, respectively). Of the 41 patients, 14 cases (34.1%) had a documented history of consuming contaminated foodstuffs (ice cream or refrigerated dairy products, 8 cases; refrigerated cooked food, 5 cases; raw seafood, 1 case). The median duration for diagnosis was 10 days (IQR: 7 days to 18 days), with four cases exceeding four weeks.

Table 1.

Clinical and demographic characteristics of children with lm meningitis

Characteristic	Case n (%) or Median (IQR)
Sex
Male	21 (51.2%)
Female	20 (48.8%)
Age at onset (years)	2.3 (1.6~3.4)
Season at onset
Spring (March-May)	15 (36.6%)
Summer (June-August)	14 (34.1%)
Autumn (September-November)	8 (19.5%)
Winter (December-February)	4 (9.8%)
History of contaminated food and drink	14 (34.1%)
Duration from symptom onset to diagnosis (days)	10 (7~18)
Clinical manifestations and physical examination
Fever	41 (100.0%)
Chill	3 (7.3%)
Vomiting	24 (58.5%)
Diarrhoea	12 (29.3%)
Altered state of consciousness	10 (24.4%)
Headache	10 (24.4%)
Convulsive seizures	12 (29.3%)
Stiff neck	18 (43.9%)
Positive Babinski sign	10 (24.4%)
Concomitant diseases	7 (17.1%)
Spesis	2 (4.9%)
Immunodeficiency disease	1 (2.4%)
Respiratory failure/respiratory distress	2 (4.9%)
Digestive tract disease	2 (4.9%)
Initial blood WBC count (109/L)	11 (8.3~11.5)
neutrophil predominance	36 (87.8%)
Disease-initial C-reactive protein (mg/L)
< 50	24 (58.5%)
≥50， < 100	9 (22.0%)
≥100	8 (19.5%)
Disease-initial procalcitonin (ng/dL)	0 (0.0%)
< 1	33 (80.5%)
≥1， < 2	4 (9.8%)
≥2	4 (9.8%)
Disease-initial CSF white blood cell count (106/L)	847 (405~1586)
Monocyte predominance	29 (70.7%)
Disease-initial CSF glucose concentration (mmol/L)	2.1 (1.1~3.0)
Disease-initial CSF protein concentration (mg/L)	810 (648~1200)
Glucocorticoids	20 (48.8%)
Central nervous system complications	13 (31.7%)
Subdural effusion	5 (12.2%)
Hydrocephalus	6 (14.6%)
Ependymitis	7 (17.1%)
Brain abscess	1 (2.4%)
Cerebral infarction	1 (2.4%)
Visual impairment	3 (7.3%)
Dyskinesia	2 (4.8%)
Epilepsy	1 (2.4%)

Abbreviations: IQR interquartile range, WBC white blood cell, CSF cerebrospinal fluid

The study population displayed three major clinical manifestations: fever (41 cases, 100%), vomiting (24 cases, 58%), and neck stiffness (18 cases, 43.9%). Headaches were reported in 10 cases (24.3%), primarily in children aged 3 and older, while convulsions were more common in those under 3 years (12 cases, 29.3%). Diarrhea affected 12 cases (29.2%), often preceding or coinciding with central nervous system symptoms (Table 1).

A laboratory analysis revealed that 48.7% of patients displayed elevated blood WBC counts, with all cases exhibiting increased neutrophil ratios. The majority of patients (82.9%, 34/41) exhibited elevated C-reactive protein (CRP) levels, while only 19.6% (8 out of 41) demonstrated procalcitonin (PCT) levels exceeding 1 ng/dL. Lumbar punctures were performed between one and 47 days following the onset of symptoms. The analysis of CSF revealed a median cell count of 847 × 10⁶/L (with a range from 57 to 4070 × 10⁶/L) and a median glucose level of 2.1 mmol/L (ranging from 0.24 to 4.9 0 mmol/L). Monocyte predominance was identified in 61% of patients (25/41).

Central nervous system complication and risk factor analysis

All patients underwent cranial imaging examination, 13 (31.7%) patients developed neurological complications during the clinical course of their disease. The most prevalent complications included ependymitis (7 cases, 17%), hydrocephalus (6 cases, 14.6%), subdural effusion (5 cases, 12.2%), and brain abscesses (3 cases, 7.3%) (Table 1). Univariate analysis showed that changes in consciousness, duration from symptom onset to diagnosis > 1 week and onset CRP ≥ 50 mg/L were associated with neurological complication (all p < 0.05; Table 2). Multivariate analysis showed that delay diagnosis time > 1 week and increased initial CRP (≥50 mg/L) were independent risk factors for neurological complication (all p < 0.05; Table 3).

Table 2.

Univariate analysis of risk factors associated with neurologic complications in children with LM meningitis

Characteristics		Neurologic complication (n = 13) [n(%) OR median (IQR)	No-neurologic complication (n = 28) [n(%) OR median (IQR)	P-value
Sex				0.326
Male		5 (38.5%)	16 (57.1%)
Female		8 (61.5%)	12 (42.9%)
Age at onset（years)		2.4 (1~3.5)	2.2 (1.6~3.25)	0.726
Season at onset
Spring(March-May)		8 (61.5%)	7 (25.0%)	0.038
Summer(June-August)		4 (30.8%)	10 (35.7%)	1
Autumn(September-November)		1 (7.7%)	7 (25.0%)	0.398
Winter(December-February)		0	4 (14.3%)	0.288
History of contaminated food and drink		4 (30.8%)	10 (35.7%)	1
Duration from symptom onset to diagnosis
≤1 week		1 (7.7%)	13 (46.6%)	0.031
> 1 week, < 2 weeks		6 (46.2%)	5 (17.9%)	0.073
≥2 weeks, < 4 weeks		5 (27.8%)	7 (25.0%)	1
> 4 weeks		1 (7.7%)	3 (10.7%)	1
Clinical manifestations and physical examination
Fever		13 (100%)	28 (100%)
Chill		1 (7.7%)	2 (7.1%)	1
Vomiting		5 (27.8%)	19 (67.9%)	0.098
Diarrhoea		6 (46.2%)	6 (21.4%)	0.146
Changes in consciousness		6 (46.2%)	4 (14.3%)	0.049
Headache		4 (30.8%)	6 (21.4%)	0.698
Convulsive seizures		5 (27.8%)	7 (25.0%)	0.469
Stiff neck		7 (53.8%)	11 (39.3%)	0.503
Positive Babinski sign		3 (23.1%)	7 (25.0%)	1
Concomitant diseases		3 (23.1%)	4 (14.3%)	0.659
Sepsis		1 (7.7%)	1 (3.6%)
Immunodeficiency disease		0	1 (3.6%)
Respiratory failure/respiratory distress		1 (7.7%)	1 (3.6%)
Digestive tract disease		1 (7.7%)	1 (3.6%)
Initial blood WBC count (109/L)		17.5 (3.1~24.2)	10.8 (8.0–19.0)	0.078
neutrophil predominance		11(84.6%)	25 (89.3%)	0.645
Disease-initial C-reactive protein (mg/L)
< 50		4 (30.8%)	20 (71.4%)	0.020
≥50， < 100		6 (46.2%)	3 (10.7%)	0.181
≥100		3 (23.1%)	5 (17.9%)	0.692
Disease-initial procalcitonin (ng/dL)
< 1		9 (69.2%)	24 (85.7%)	0.237
≥1， < 2		1 (7.7%)	3 (10.7%)	1
≥2		3 (23.1%)	1 (3.6%)	0.086
Disease-initial CSF white blood cell count (106/L)
< 100		1 (7.7%)	1 (3.6%)	0.539
≥100， < 500		2 (15.4%)	8 (28.6%)	0.458
≥500， < 1000		5 (27.8%)	6 (21.4%)	0.280
≥1000		5 (27.8%)	13 (46.6%)	0.742
Monocyte predominance		9 (69.2%)	20 (71.4%)	1
Disease-initial CSF glucose concentration (mmol/L)
≤1		4 (30.8%)	9 (32.1%)	1
> 1， < 2		3 (23.1%)	4 (14.3%)	0.659
≥2		6 (46.2%)	15 (53.6%)	0.744
Disease-initial CSF protein concentration (mg/L)
< 450		0	0
≥450， < 1000		7 (53.8%)	19 (67.9%)	0.492
≥1000， < 2000		4 (30.8%)	6 (21.4%)	0.698
≥2000		2 (15.4%)	2 (7.1%)	0.618
Glucocorticoids		9 (69.2%)	11 (39.3%)	0.100

Abbreviations: IQR interquartile range, WBC white blood cell, CSF cerebrospinal fluid

Table 3.

Multivariate analysis of risk factors associated with neurologic complications in children with lm meningitis

	β	Wald	P-value	OR	95% CI
Changes in consciousness	0.675	0.904	0.342	1.965	0.488	7.907
Duration from symptom onset to diagnosis > 1 week	3.215	7.276	0.007	24.912	2.408	257.68
Disease-initial C-reactive protein ≥ 50 mg/L)	2.433	12.082	0.001	11.397	2.89	44.947

Abbreviations: OR odds ratio, CI confidence intervals

Microbiological detection and data of mNGS

The positivity rates for blood and CSF cultures were 26.8% (11/41) and 82.9% (34/41), respectively. The combined culture yielded a positivity rate of 90.2% (37/41). CSF culture rates were significantly higher before hospital admission compared to after admission (85.7% vs 31.7%, p < 0.05) (Table 4). Notably, two cases had negative CSF cultures before but positive cultures after admission. mNGS was conducted on CSF samples from seven pediatric patients, all producing positive results. Four cases showed negative results in CSF cultures, while three exhibited negative cultures in blood specimens. The duration between symptom onset and sample collection ranged from 5 to 37 days. (Table 5).

Table 4.

Blood and CSF culture results of 41 patients

Method	Before admission positive rate, (n)	After admission positive rate, (n)	Total positive rate, (n)	χ2	P
Blood culture	25%（8/29)a	14.6%(6/41)	26.8%（11/41）	0.16	0.68
CSF culture	85.7%(24/29)	31.7%(13/41)	82.9%（34/41）	18.80	0.00
Total	86.2%（25/29）	36.5%（15/41）	90.2%（37/41）	18.90	0.00

Abbreviations: a: Positive cases/Total cases cultured, CSF, cerebrospinal fluid

Table 5.

Summary of the microbiological investigations in the patients performing mNGS

Patient number	mNGS	Symptom onset to sampling date	Specific sequences of mNGS	Blood culture	CSF culture
1	positive	18	9529	positive	negative
2	positive	5	unknown	positive	positive
3	positive	12	1011	negative	positive
4	positive	8	unknown	positive	positive
5	positive	9	34	negative	negative
6	positive	21	2561	negative	negative
7	positive	37	25	negative	negative

Abbreviations: Metagenomic next-generation sequencing,mNGS; CSF, cerebrospinal fluid

Treatment

Antibiotic susceptibility testing and LM serotyping were not performed for any patient, and they were given empirical antibiotic therapy. Before the pathogen was diagnosed, 60.9% (25/41) of the children received cephalosporin antibiotics empirically, with no improvement in symptoms. Following the acquisition of the pathogen, the predominant antibiotic regimen observed was monotherapy with meropenem (60.9%, 25/41) (Table 6). In 17 cases (17/25, 68%) where there was a poor response to antibiotic treatment, combination therapy was implemented with linezolid, vancomycin, amoxicillin clavulanate potassium, or trimethoprim-sulfamethoxazole (TMP-SMZ), depending on the individual patient’s clinical status. Six patients who were initially treated with Penicillin exhibited a poor response and were subsequently transitioned to meropenem therapy. (Table 6). Ten patients were initially treated with meropenem in combination with linezolid, vancomycin, or amoxicillin clavulanate potassium. Among these, only two patients responded poorly and were combined with TMP-SMZ.

Table 6.

The antibiotic treatment regimens after positive cultures

Initial treatment	Modified treatment	Replacement ratio
Penicillin（6）		83.30%
Ampicillin （2）	Meropenem+Vancomycin（1）/Meropenem+Amoxicillin-Clavulanate potassium（1）
Penicillin G（2）	Meropenem+Linezolid+TMP-SMZ（1）; Meropenem+Penicillin G（1）
Amoxicillin-Sulbactam（1）	Meropenem+Vancomycin+TMP-SMZ（1）
Piperacillin-Tazobactam（1）	Piperacillin-Tazobactam（1）
Meropenem（25）		68%
	Meropenem（8）
	Meropenem+Amoxicillin-sulbactam（3）
	Meropenem+TMP-SMZ（3）
	Meropenem+Linezolid（1）
	Meropenem+Vancomycin（5）
	Meropenem+Linezolid/Vancomycin+TMP-SMZ（5）
Combined treatment（10）		20%
Meropenem+Amoxicillin-Clavulanate potassium（1）	Meropenem+Amoxicillin-Clavulanate potassium（1）
Meropenem+Vancomycin（7）	Meropenem+Vancomycin（5）
	Meropenem+Vancomycin+TMP-SMZ（2）
Meropenem+Linezolid（2）	Meropenem+Linezolid（2）

Abbreviations: Trimethoprim-sulfamethoxazole TMP-SMZ

Prognosis

All patients were monitored for at least 12 months without fatalities. Two patients with V-P shunts exhibited mild dyskinesia abnormalities, while the remaining patients showed satisfactory recovery with no discernible neurological complications. The GOS score for all the children was recorded as 5.

Discussion

LM infection is recognized as one of the most severe foodborne illnesses because of its high pathogenicity and fatality rates. In healthy individuals, the immune system usually eliminates LM upon invasion. However, those with compromised or deficient immune systems may experience uncontrolled growth of LM in their bodies, which can lead to central nervous system infection. LM infection has been identified as the causative agent in approximately 4% of community-acquired bacterial meningitis cases among patients over the age of 16 [8]. Reports indicated that between 50% and 70% of adult patients diagnosed with LM meningitis also have immune-related diseases [9, 10]. The prevailing consensus in international research suggests that LM meningitis is rarely seen in immunocompetent children beyond the neonatal period. However, recent studies have shown that LM meningitis can occur in immunocompetent children as well [9, 11, 12]. In this study, only one child exhibited humoral immunodeficiency, while the others were previously healthy. This finding aligns with observations reported in both domestic and international literature.

Consistent with previous research, the clinical presentations of LM infections often exhibit nonspecific symptoms of bacterial meningitis [13, 14]. Notably, over 90% of human infections caused by LM are foodborne, primarily resulting from the consumption of contaminated foods such as raw meat, milk, and ready-to-eat meals [15, 16]. Seasonal variation is also observed, with LM infections in the gastrointestinal tract being more prevalent during the summer months [17]. In this study, 34.1% of patients reported a history of consuming refrigerated dairy products or cooked food, and about 70% experienced symptoms in the spring or summer, which aligns with a similar study conducted in Chongqing province [12]. When assessing patients with purulent meningitis who present gastrointestinal symptoms during the prodrome, it is essential to consider the possibility of a LM infection. Compared to bacterial meningitis caused by more common pathogens like Streptococcus pneumoniae, meningitis due to LM typically has a protracted disease course, a lower incidence of septic shock, and tends to affect older children more frequently [18].

The typical predominance of monocytes in LM meningitis differs from what is observed in other types of bacterial meningitis. This similarity was noted in the present study [13, 19]. This phenomenon may be due to the proliferation of monocytes in response to LM, a facultative intracellular pathogen that can invade non-phagocytic cells or be engulfed by phagocytic cells. This interaction leads to the activation of monocytes and their transformation into macrophages [2, 20]. The predominance of monocytes can complicate the early diagnosis of LM meningitis, sometimes resulting in misdiagnosis as tuberculous meningitis, which is relatively common in northern China [21].

Previous research has shown that the main imaging characteristics of LM meningitis include meningitis, obstructive hydrocephalus, ependymitis, and brainstem encephalitis, with brainstem encephalitis being a significant risk factor for mortality [22]. In this study, none of the children exhibited brainstem encephalitis, which may explain the relatively favorable prognosis observed in this cohort.

Additionally, previous investigations have identified factors such as delayed initiation of antibiotics, young age ( < 12 months), altered level of consciousness, initial CSF-protein > 2 g/L, CRP > 100 mg/L are associated with neurological complications [23–25]. Our study also showed delayed diagnosis and significant CRP elevation are associated with neurological complications. Corticosteroids have the potential to mitigate hearing loss and neurological sequelae associated with bacterial meningitis. However, there remains ongoing debate concerning the effects of corticosteroids therapy on the prognosis of LM meningitis. The ESCMID guidelines recommended to stop dexamethasone if the patient is discovered the bacterium causing the meningitis is a species other than Haemophilus influenzae or Streptococcus pneumoniae [26]. In the present study, twenty patients received corticosteroids as adjunctive treatment prior to the determination of the etiology, and all patients had prognoses this study.

Distinguishing LM meningitis from tuberculous meningitis, cryptococcal meningitis, and other types of bacterial meningitis can be challenging based on clinical symptoms and routine cerebrospinal fluid (CSF) examinations. Therefore, the diagnosis relies on CSF and blood cultures. The present study observed a considerable decrease in the positivity rates of both blood and cerebrospinal fluid (CSF) cultures post-admission. This finding indicates that completing blood and CSF cultures before antibiotic treatment may improve pathogen detection rates.

The mNGS has emerged as a novel method for identifying microorganisms, overcoming the limitations of traditional culture-based techniques. By sequencing total DNA or RNA, mNGS can detect all nucleic acids in specimens, including those from both the host and microbes. This technique increases pathogen coverage, allows for the detection of pathogens even at lower concentrations, improves sensitivity, is less affected by prior antibiotic exposure and treatment duration, and provides rapid results (in less than 24 hours) [27, 28]. These advantages are of particular significance in the diagnosis of LM meningitis. This study identified LM through mNGS in three patients whose CSF and blood cultures were negative. Despite its high sensitivity, clinicians should exercise caution when interpreting mNGS results due to certain inherent limitations.

The empirical treatment of meningitis in China usually involves third-generation cephalosporins and vancomycin to target prevalent pathogens [29]. However, it is essential to note that LM exhibits inherent resistance to third-generation cephalosporins. Currently, amoxicillin, Penicillin G and ampicillin are recommended as first-line treatments, with meropenem and TMP-SMZ as alternative options [30, 31]. Other viable treatment options for LM meningitis include quinolones, rifampicin and linezolid [31, 32]. In our institution, we did not routinely perform antimicrobial susceptibility testing for LM and could not obtain ampicillin or amoxicillin for treatment. Eighty-three percent of children who were initially treated with penicillin were eventually switched to meropenem. This data highlights the importance of being aware of antibiotic resistance. Additionally, a recent meta-analysis indicated that resistance to amoxicillin and penicillin G in LM has been reported [33]. In our study, meropenem, either as monotherapy or in combination with other agents, was administered to 40 pediatric patients, and favorable therapeutic outcomes were achieved. The above results are consistent with prior studies, indicating that meropenem may be viable for empirical therapy. Nevertheless, the clinical evidence remains inconclusive, with reports of treatment failures in individual cases. Only one patient with mild disease was treated with piperacillin-tazobactam alone. Piperacillin-tazobactam exhibits significant antibacterial efficacy against LM. Although it possesses the capability to cross the blood-brain barrier, its concentration within the CSF remains relatively low, thereby limiting its utility in treating central nervous system infections. In our syudy, the patient demonstrated marked clinical improvement following initial empirical therapy with piperacillin-tazobactam. Consequently, the therapeutic regimen was maintained post-pathogen identification, resulting in favorable clinical outcomes. Previous studies have indicated that during inflammatory episodes, increased permeability of the blood-brain barrier facilitates enhanced drug penetration into the CSF, thereby augmenting antimicrobial efficacy against CNS infections [34]. Additionally, research by Chie Fukasawa et al. [35] monitored piperacillin-tazobactam concentrations in the CSF of five pediatric patients with Haemophilus influenzae meningitis, revealing that the drug could achieve bactericidal concentrations within the CSF. These results offer a potential reference point for the application of piperacillin-tazobactam in the treatment of LM meningitis. Nevertheless, additional research is necessary to ascertain its therapeutic efficacy in this specific condition.

The recommended treatment duration for LM meningitis ranges from 3 to 4 weeks, depending on the severity of the cases. For children with immune deficiencies and brain abscesses, the treatment course should be extended to 6 to 8 weeks [36]. In our study, the treatment duration for patients ranged from 3 to a maximum of 10 weeks, with successful infection control and a favorable long-term prognosis observed, including in three patients with ventricular-peritoneal shunts (V-P). Currently, there is no consensus on the optimal duration of therapy. In this study, 25 cases (60.9%) received antibiotic treatment lasting more than 4 weeks.

Limitation

This study has several limitations. Firstly, it was a single-center retrospective study that included only 41 patients. Therefore, the results need to be confirmed in future prospective multicenter studies. Secondly, some participants initially sought treatment at other hospitals, which meant that some laboratory and culture results were obtained from different institutions. This could have led to variations in methods and standards, potentially affecting our findings. Additionally, there is currently a lack of research on the long-term prognosis for children with LM meningitis. More prospective multicenter studies with larger sample sizes are necessary to better identify risk factors associated with poor prognosis in LM meningitis cases. Such studies could help develop more effective treatment strategies, ultimately improving the prognoses for children with LM meningitis and alleviating the burden of this disease on families and society.

Conclusions

In summary, the poor efficacy of empirical cephalosporin treatment in a child with bacterial meningitis should prompt consideration of LM infection, even in immunocompetent hosts, especially if supported by a history of exposure to contaminated food. Timely completion of microbiological examinations of both cerebrospinal fluid (CSF) and blood is essential. For patients with challenging pathogen diagnoses, mNGS is recommended as soon as possible. Penicillin or ampicillin should be administered promptly, ideally in combination with aminoglycoside antibiotics. When necessary, meropenem should be used for targeted treatment. A combination of meropenem with linezolid, vancomycin, and trimethoprim-sulfamethoxazole may be considered in severe cases. Strengthening food hygiene and safety education is crucial to prevent LM infection.

Abbreviations

LM

Listeria monocytogenes

CSF

Cerebrospinal fluid

mNGS

Metagenomic next-generation sequencing

GOS

Glasgow Outcome Score

WBC

White blood cell

CRP

C-reactive protein

PCT

Procalcitonin

V-P

Ventriculoperitoneal

TMP-SMZ

Trimethoprim-sulfamethoxazole

OR

Odds ratio

CI95%

Confidence interval at 95%

IQR

Interquartile range

Author contributions

All the authors had access to the full dataset (including the statistical reports and tables) and take responsibility for the integrity of the data and the accuracy of the data analysis. and L.QJ conceived and designed the study L.G and L. QJ conceived and designed the study. H. B, G. LY, D. ZZ, F. WY, N. X and X.HJ were involved in the case and data collection, designed the analysis, and interpreted the data. L. QJ wrote the first draft of the paper. C.TM made preliminary revision. L. QJ, and L. G reviewed and approved the final report. All authors have read and approved the final manuscript.

Funding

This work was supported by the Capital's Funds for Health Improvement and Research [grant numbers:2024-1-2092];Training Plan for High level Public Health Technical Talents Construction Project [grant numbers: Discipline Leader-02-02]; Beijing Municipal Administration of Hospitals Incubating Program [grant number: PX2021047]; Young Talent Training Project of Beijing Hospital Authority [grant number: QML20191203]; Beijing Excellent Talents Training Program [grant numbers:2018000021469G274].

Data availability

The datasets analyzed during the current study are available from the cor‑ responding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the principles outlined in the Declaration of Helsinki. Ethical approval for this research was obtained from the Ethics Committee of Beijing Children’s Hospital Affiliated to Capital Medical University (approval No. [2024]-E-076-R). Informed consent to participate was waived by the Ethics Committee of Beijing Children’s Hospital Affiliated to Capital Medical University because this was a retrospective study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.