Background
Viral meningitis is characterized by the acute onset of fever, headache, and neck stiffness, often accompanied by nausea and vomiting. Non-polio enteroviruses (EV), which are a genus in the family Picornaviridrae, are known as the most common pathogen of aseptic meningitis in children [
1] and account for approximately 60–80% of aseptic meningitis cases [
1‐
3]. The outcomes of EV infections of the central nervous system are usually favorable in otherwise healthy patients, but EV meningitis causes considerable morbidity, with moderate or high fever despite antipyretics and several days of severe headache warranting opiate analgesia [
1,
4]. Thus, careful attention is needed to confirm the diagnosis.
Several studies have reported the lack of pleocytosis in children with EV meningitis, RT-PCR is presently considered to be the gold standard for diagnosis of EV infections [
2]. CSF analysis takes several hours, but RT-PCR requires several days depending on the facilities; therefore, it is difficult to confirm EV via RT-PCR during a patient’s stay in the emergency department (ED). Although disposition should be based on suspicion of meningitis and symptom severity rather than the presence of CSF pleocytosis, in reality, due to lack of medical resources many patients are discharged from the ED if CSF cell count is normal with well controlled symptoms. Some of these patients revisited the ED with aggravated symptoms, upon which EV meningitis was confirmed via positive CSF RT-PCR results using records of the previous ED visit. Thus, we aimed to analyze the prevalence of enterovirus (EV) meningitis without the presence of cerebrospinal fluid (CSF) pleocytosis and identify patient factors and clinical features.
Discussion
We studied 305 cases of RT-PCR-confirmed EV meningitis—the largest number of cases from a single-center study—and compared clinical features of patients and the frequency of pleocytosis during periods of EV prevalence.
This study shows that patients with apleocytosis were younger, had lower WBC and ANC count, compared with those characteristics in patients with pleocytosis. Several studies have reported the absence of CSF pleocytosis in children with EV meningitis [
7,
11‐
14]. For example, Yun et al. reported that about 30% of patients aged < 3 months with EV meningitis did not have CSF pleocytosis [
7]. Other studies report a similar proportion of cases lacking pleocytosis in EV meningitis among patients aged < 3 months: 34% of patients aged < 90 days in Singapore [
12], 31% in Pennsylvania, US [
13], and 30% of patients aged < 2 months in Missouri, US [
15]. Furthermore, a study from the Netherlands also showed that 40% of pediatric patients aged < 16 years with EV meningitis did not demonstrate CSF pleocytosis and these patients were significantly younger (
p = 0.001) [
11]. Compared with previous studies, we showed higher prevalence of EV meningitis with apleocytosis; overall 55.4, and 82.8% in patients younger than 3 months.
The mechanism underlying the lack of pleocytosis in younger patients has been suggested to be connected with the immaturity of the immune system of young patients [
13]. Mulford et al. speculated that the immunologic maturity of these patients is insufficient to recruit leukocytes to the site of infection [
15]. Moreover, consistent with the current results, lower peripheral WBC count in patients without pleocytosis has also been observed in other studies [
7,
11,
13]. The association between lower peripheral WBC count and a lack of CSF pleocytosis in young patients may also be explained by immunologic immaturity [
11,
13].
Another hypothesis to explain the frequent absence of pleocytosis at younger ages is the short time from the onset of EV meningitis to lumbar puncture [
7,
12]. Yun et al. hypothesized that the immunologic response that recruits WBCs to CSF may not be complete due to the short duration of symptoms [
7]. In the present study, 82.8% of infants younger than 3 months had a higher frequency of the absence of pleocytosis than that in previous studies. The reason for this higher rate might be because 94.5% of patients aged ≤3 months underwent lumbar puncture within 24 h. In Korea, domestic caregivers are taught that fever in infants is a dangerous symptom of infection, and they can easily access an emergency room at any time [
16]. Especially if the child is younger, caregivers visit the outpatient clinic or emergency room as soon as possible, so the examination and diagnosis tends to be done quickly [
17].
Heo et al. reported a lower CSF protein level in patients without pleocytosis (apleocytosis group 18.9 ± 4.3 mg/dl vs. pleocytosis group 34.9 ± 18.3 mg/dl,
p = 0.002) and explained that at the initial stage of illness, blood-brain barrier permeability is maintained, thus WBCs have not yet infiltrated the cerebrospinal cavities and CSF protein level dose not increase [
18]. On the other hand, the CSF protein level in patients without pleocytosis was higher in our study, and this is in discord with another study [
14]. To explain our result, we must consider the normal protein level in CSF according to age. The normal range of CSF protein level according to the actual age is higher in neonates and infants than it is in children (preterm infant 115 mg/dl (65–150), term neonate 90 mg/dl (20–170), child 5-40 mg/dl) [
19,
20]. Our result suggests that the higher protein level in patients without pleocytosis is because the apleocytosis group contained more patients aged ≤3 months.
Additionally, our results reveal that patients with pleocytosis had complain of more meningitis symptoms such as headache and vomiting than patients without pleocytosis. This might be because those with pleocytosis were older, had more advanced verbal skills, and thus could verbally explain their symptoms such as headache more easily.
Another important finding from our analysis is that there was a significantly different frequency of pleocytosis according to season of the year. EV meningitis shows a pronounced seasonality, being more frequent in summer and fall [
9,
10]. In our study, while the frequency of EV meningitis without pleocytosis was higher compared with pleocytosis meningitis during prevalent EV seasons, the frequency was similar during non-predominant months. During the traditional EV season, we recommend that patients with symptoms suspicious for meningitis undergo a lumbar puncture, with RT-PCR testing.
The diagnosis of meningitis affected the dispositions; discharge or admission. Among 305 patients, 43 patients (14.1%) were discharged from the ED. In patients with pleocytosis, all were hospitalized except for one; the remaining discharged 42 patients did not demonstrate pleocytosis. 6 patients of these 42 patients (14.3%) revisited the ED with aggravated symptoms and poor condition. When they returned to the ED, the results of RT-PCR during the previous visit were confirmed, and they finally received a diagnosis of EV meningitis without pleocytosis. This suggests that although the first result of the CSF profile was negative, patients with clinical features suspicious for meningitis should have their diagnosis confirmed by RT-RCR.
This study has several limitations. First, because of the retrospective design, clinical features such as vomiting, headache, irritability and meningeal irritation sign may be inadequate based solely on medical records. Also, as younger patients may not have been able to communicate their symptoms, this data in the younger age group may not be accurate. As retrospective medical reviews, we had small numbers in 3 months - 3 years. Though we analyzed non-parametric methods, this is weak point of our study. Second, this study was a single-center study, thus the results may not be widely applicable. In addition, our RT-PCR data contained only positive or negative results of EV infection and did not indicate a subtype; accordingly, we were unable to determine whether differences in the present study and those of previous reports are due to EV subgroup differences. For example, compared with previous studies, we found a higher frequency of apleocytosis in older children, but this can be influenced by EV subtype. Taken together, we consider that a prospective multicenter study is needed to address these points.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.