Prospective Analysis of Central Nervous System Infection Risks in Varicella-Zoster Virus Reactivation Cases: A Single-Center Prospective Study of 1030 Cases

Introduction

The primary infection of varicella-zoster virus (VZV) usually occurs in children, leading to chickenpox. In the later stage, VZV lurks in the dorsal root ganglia of spinal cord or cranial ganglia, remaining dormant until reactivation, and this latent state is lifelong. When the host’s immune system is damaged, latent VZV can be reactivated, leading to local rash with or without severe pain, known as herpes zoster (HZ). The most common complication of HZ is postherpetic neuralgia, which can cause persistent chronic pain. HZ can also merge central nervous system (CNS) infections, which are closely related to the high risk of mortality and sequelae. The literature reports that the mortality of VZV-associated CNS infection can reach 5% [1], which brings huge physical, psychological, and economic burdens to patients and their families. Therefore, early identification of risk factors for CNS infection caused by VZV reactivation and active intervention for high-risk patients will help to reduce the incidence of intracranial infections and alleviate the burden on families and society.

However, the risk factors for VZV-associated CNS infection are currently unclear. HZ is closely related to the host’s immune function, and the probability of developing HZ and postherpetic neuralgia in immunocompromised individuals is significantly higher than that in immunocompetent individuals. According to our conventional thinking, immunocompromised patients with HZ may also have an increased chance of developing CNS infections. However, in clinical practice, there is still much difference of opinion on whether immune dysfunction is a risk factor for VZV-associated CNS infection. Some scholars believe that a compromised immune system in patients is associated with an increased risk for CNS complications of VZV reactivation [2–4]. However, some studies did not find differences in the frequency of CNS complications among HZ patients with or without immunosuppression [5]. In addition, Carey et al [6] found that patients with type 2 diabetes who develop HZ may be at high risk for CNS complications. Kim et al [7] evaluated clinical data from 578 patients with HZ and found that zoster with cardiovascular distribution and male sex were risk factors for HZ-associated meningitis. Although Kim et al retrospectively analyzed a large number of patients, the main focus was on analyzing their clinical data, and there was a lack of detection and analysis of serum markers. Herlin et al [8] also analyzed clinical data of 92 patients with VZV encephalitis nationwide in Denmark and found that, in addition to immunosuppression, elderly age is also a high-risk factor for VZV encephalitis. In summary, there are many inconsistent conclusions regarding the risk factors of VZV-associated CNS infection, which may be related to previous studies being mostly retrospective, with a small number of patients included and insufficient analysis indicators. Therefore, in this study, we prospectively included a large sample of patients with acute HZ, to explore the related factors (including clinical data and serum markers) of VZV-associated CNS infection, in order to provide clinical evidence for early assessment and identification of high-risk populations, early prevention, and control of VZV-associated CNS infection.

Material and Methods

CLINICAL CHARACTERISTICS:

We prospectively included 1030 patients admitted in our hospital due to acute HZ between January 2021 and June 2023, including 477 men and 553 women, with an average age of 60.5±15.3 years. he baseline clinical data of patients were analyzed, including sex, age, location of herpes, concomitant diseases, interval from onset to treatment, clinical symptoms, and immune status.

The inclusion criteria were as follows: (1) age ≥18 years; (2) within 2 weeks of onset; and (3) patients who had not undergone standardized treatment. The diagnosis of HZ was determined by dermatologists, based on diagnostic criteria.

The exclusion criteria were as follows: (1) zoster sine herpete; (2) HIV infection; (3) pregnancy or lactation; (4) patients with mental illness or severe dementia who could not accurately express clinical symptoms; and (5) patients or family members who do not agree to participate in this study.

The diagnosis of VZV-associated CNS infection was confirmed according to previous criteria in the literature [9,10]. It also followed the criteria of the European consensus-based guideline on the management of the herpes zoster, to include patients with VZV meningitis and meningoencephalitis [11].

This research protocol was approved by the Ethics Committee of Hangzhou Third People’s Hospital (2021KA013).

SERUM MARKERS/IMAGING DETECTION:

On the second day of admission, fasting venous blood was drawn from the included patients for routine blood tests (white blood cells, red blood cells, hemoglobin, blood progenitor cells, neutrophils, lymphocytes), C-reactive protein, and blood lipids (triglycerides, total cholesterol, low-density lipoproteins, high-density lipoproteins, liver function (albumin, globulin, total protein, bilirubin, total bile acid), renal function (urea nitrogen, creatinine), blood glucose, blood uric acid, homocysteine, thyroid function (T3, T4, TSH), T-lymphocyte subset level, and immune function (C3, C4, IgA, IgG, IgM). Head magnetic resonance imaging, electroencephalography, and other examinations were performed on patients with HZ suspected of CNS infection.

DETECTION OF CEREBROSPINAL FLUID:

Lumbar puncture examination was performed on patients with acute HZ suspected of having CNS infection. All patients collected cerebrospinal fluid (CSF) samples after signing the consent form. Sterile test tubes were used to collect 5 mL of cerebrospinal fluid, and routine cerebrospinal fluid, biochemical (glucose, chlorine, total protein, lactate dehydrogenase, adenosine deaminase), and VZV DNA measurements were performed.

ANTIVIRAL THERAPY:

After admission, patients with herpes zoster received systematic antiviral treatment (valciclovir hydrochloride tablets 1000 mg 3 times daily, or bromofloxacin tablets 125 mg 4 times daily; alternatively, acyclovir 10 mg/kg can be administered intravenously every 8 h). The course of treatment was 7 to 10 days. Patients with diagnosis of VZV-associated CNS infection received acyclovir 10 mg/kg intravenously every 8 h. The treatment course lasted for 10 to 21 days (1 patient with renal insufficiency had an adjusted dosage of acyclovir according to creatinine clearance rate). There were 13 patients with VZV-associated CNS infection (13/40, 32.5%) who received intravenous methylprednisolone 40 to 60 mg/day for 3 to 7 days. Patients with herpes zoster-associated pain in both groups were treated with oral gabapentin or pregabalin capsules.

STATISTICAL ANALYSIS:

Statistical analysis was conducted using SPSS 25.0 software. Counting data is expressed in terms of number of examples (percentage, n), and intergroup comparisons were conducted using the chi-square test or Fisher’s method. Quantitative data that followed a normal distribution are expressed as mean±standard deviation (x±SD), and intergroup comparisons were conducted using 2 independent-sample t tests. Data that were not normally distributed are represented as median (M [Q1, Q3]), and inter-group comparisons were conducted using the Mann-Whitney U test. First, univariate analysis was applied to screen for risk factors for concurrent CNS infections in HZ patients, and then multivariate logistic regression analysis was applied to screen for independent risk factors for concurrent CNS infections in patients. The difference was considered statistically significant with P<0.05.

Results

GENERAL INFORMATION OF PATIENTS IN THE HZ GROUP AND VZV-ASSOCIATED CNS INFECTION GROUP:

Among the 1030 included HZ patients, 990 were in the HZ group (without concurrent CNS infection), with 454 men (45.9%) and 536 women (54.1%). The average age was 60.7±15.3 years, and the interval from onset to treatment was 5.3±2.6 days. There were 40 patients in the VZV-associated CNS infection group (including 33 cases of meningitis, 6 cases of meningoencephalitis, and 1 case of myelitis), with 23 men (57.5%) and 17 women (42.5%). The average age was 55.7±15.0 years, and the interval from onset to treatment was 6.5±3.1 days. Compared with patients in the HZ group, the patients in the VZV-associated CNS infection group were younger (P<0.05) and had a significantly longer interval from onset to treatment (P<0.05). However, there was no significant difference in sex (P>0.05), as shown in Table 1.

In the HZ group, the main concomitant diseases were hypertension (32.4%), type 2 diabetes (9.5%), and tumors (6.4%). Similarly, the most common comorbidities in the VZV-associated CNS infection group were hypertension (37.5%), type 2 diabetes (10.0%), and tumors (5.0%). There was no significant difference in comorbidities between the 2 groups (P>0.05), and there was no difference in the immune diseases and use of immunosuppressive agents between the 2 groups (P>0.05).

Herpes were mainly distributed in the head and neck of 30 patients (75.0%) in the VZV-associated CNS infection group, followed by 9 cases (22.5%) in the chest and back, and 1 case (2.5%) in the limbs. Herpes in the HZ group were distributed in 515 cases (52.0%) of the head and neck, 173 cases (17.5%) of the chest and back, 177 cases (17.9%) of the waist and abdomen, and 125 cases (12.6%) of the limbs. The incidence of head and neck herpes in patients in the VZV-associated CNS infection group was significantly higher than that in the HZ group (P<0.01; Table 1).

COMPARISON OF SERUM MARKER LEVELS BETWEEN HZ GROUP AND VZV-ASSOCIATED CNS INFECTION GROUP:

Univariate analysis showed that compared with the HZ group, the VZV-associated CNS infection group had lower levels of blood chloride and albumin, and higher levels of C-reactive protein and glutamyltranspeptidase, with significant differences (P<0.05, P<0.01). There were no significant differences in levels of lymphocyte subsets, C3, C4, IgA, IgG, and IgM (P>0.05; Tables 2–4).

MULTIVARIATE LOGISTIC REGRESSION ANALYSIS THE RISK FACTORS OF CNS INFECTION IN HZ PATIENTS:

Using the occurrence of CNS infection as the dependent variable, a multivariate logistic regression analysis was conducted using the interval from onset to treatment, herpes distribution, age, and levels of serum chlorine, C-reactive protein, glutamyltranspeptidase, and albumin as independent variables (model 1), with P<0.05 between the HZ group and VZV-associated CNS infection group in univariate analysis. The results showed that the interval from onset to treatment, herpes distribution (head and neck), C-reactive protein, and albumin content were independent risk factors for CNS infection in patients with HZ (P<0.05), while age, chlorine, and glutamyltranspeptidase content were not independent risk factors (P>0.05). After adjusting for potential confounding factors, such as sex, concomitant tumor, type 2 diabetes, immunological diseases, immunosuppressant use (model 2), the interval from onset to treatment, herpes distribution (head and neck), C-reactive protein, and albumin content remained independent risk factors for predicting CNS infection in patients with HZ (Table 5).

The AUC of CNS infection in patients with acute HZ predicted by the interval from onset to treatment, head and neck herpes, C-reactive protein, and albumin content were 0.587, 0.634, 0.718, and 0.682, respectively. The AUC of 4 indicators combined to predict CNS infection in patients with acute HZ was 0.787, with a sensitivity of 63.5% and specificity of 81.9% (Figure 1).

Discussion

Cerebrospinal fluid testing is the criterion standard for diagnosing VZV-associated CNS infections. However, in clinical practice, the compliance with lumbar puncture for patients with acute HZ suspected of CNS infection is not high, especially for patients with normal immune function and relatively mild clinical symptoms. Therefore, it is extremely important to preliminarily assess the risk of CNS infection and intervene early through clinical manifestations and blood marker detection. However, the risk factors for CNS infections related to VZV are currently unclear, and the number of patients included in previous studies, and mostly retrospective studies or case reports, is not large. Therefore, a prospective analysis of the relevant risk factors for CNS infection in a larger number of patients with acute HZ will provide reference for clinicians to identify high-risk HZ patients in the early stage and prevent the occurrence of VZV-associated CNS infections. Our study included 1030 patients with acute HZ. The advantage of this study is that we recruited a relatively large number of patients and analyzed more data, including clinical data and peripheral blood markers. The results showed that delayed onset to treatment, head and neck herpes, elevated C-reactive protein, and low albumin content were independent risk factors for CNS infection in patients with acute HZ. Clinicians should pay attention to the risk factors of CNS infection and actively intervene in advance.

Most patients with HZ can recover after early standardized treatment, but some can experience complications, among which postherpetic neuralgia is the most common complication. In addition, postherpetic neuralgia can lead to CNS infections, mainly encephalitis, meningitis, cerebellitis, myelitis, and stroke [13], related to the high risk of patient mortality and sequelae – this places a huge burden on families and society. With the advancement of PCR technology, there are increasing reports of CNS infections related to VZV, among which VZV meningitis/encephalitis is the most common CNS complication. Research has found that VZV is the second most common virus that causes encephalitis or meningitis [3,8,13]. Shukla et al found that VZV is the second most common pathogen causing aseptic meningitis, next to enteroviruses [14]. A clinical study from Spain showed that among 100 patients with complications of the nervous system, there were 88 cases of postherpetic neuralgia, 2 cases of meningitis, and 1 case of encephalitis [15]. We included 1030 HZ patients, of which 40 (40/1030, 3.9%) had complication with CNS infections, including 33 cases of meningitis, 6 cases of meningoencephalitis, and 1 case of myelitis.

The site of herpes is one of the potential risk factors for the progression of CNS infection in patients with acute HZ. Skripuletz et al [16] retrospectively evaluated 282 HZ patients admitted to the neurology department from 2005 to 2013, and among them, 34 patients developed CNS infections and 19 patients (56%) developed trigeminal ganglia inflammation (18 cases of V1 infection and 1 case of V3 infection). In another study of HZ complicated with aseptic meningitis, it was also found that HZ patients with craniocervical distribution accounted for 87.5% in the aseptic meningitis group and 54.3% in the uncomplicated group (P=0.043). Among the craniocervical ganglia, the V1 trigeminal ganglia were most commonly involved in the meningitis group (P=0.003), with a 4-fold increased risk, compared with other ganglia [7]. Consistent with previous studies, our study also found that among 40 patients with VZV-associated CNS infection, 30 cases of herpes distribution were in the head and neck, accounting for 75%. Therefore, early and sufficient antiviral treatment for herpes in the head and neck has important clinical significance.

We also found that the interval from onset to treatment in patients with acute HZ complicated CNS infection was significantly prolonged (P<0.05). The reactivation of VZV can manifest as various neurological disorders, even without rash, as the absence of typical skin lesions can lead to delayed treatment and poorer prognosis [17]. Arruti et al [18] studied the factors associated with poor outcomes of CNS infections associated with reactivated VZV and found that older age and delayed initiation of acyclovir treatment were closely related to patient mortality. This reminds us that starting treatment immediately when patients present acute HZ with clinical symptoms of CNS infection can prevent adverse outcomes. Research on herpes zoster ophthalmopathy further supports the benefits of early treatment, with a 5.2-day delay in treatment after initial consultation doubling the risk of severe visual loss. Antiviral therapy within 2 to 4 weeks typically stops the progression of retinitis [19–21]. Due to the good efficacy of early treatment in patients with CNS involvement, rapid diagnosis, adequate treatment, and early CSF analysis are crucial.

Albumin is mainly synthesized in the liver, and serum albumin levels can partially reflect the patient’s immune, inflammatory, and nutritional status. A decrease in serum albumin level indicates that the body is in a state of malnutrition, while the immune system is inhibited to varying degrees [22]. Its antioxidant capacity and regulation of intracellular signals in neurons or glial cells contribute to its neuroprotective effects [23,24]. Multivariate logistic regression shows that a decrease in albumin concentration is the only independent predictor of mental disorder in patients with a confirmed infectious pathogen. The decrease in serum albumin concentration has good calibration and discrimination ability for predicting early mental disorders in the process of infectious encephalitis [25]. A study on the correlation between albumin contents and intracranial infection found that serum albumin levels in patients with anti-NMDA receptor encephalitis and cryptococcal encephalitis were significantly lower than those in healthy participants (P<0.001). Patients with poor treatment response have significantly lower serum albumin levels than those with good outcomes [26]. This study found that serum albumin levels are an independent risk factor for VZV-associated CNS infections. The possible mechanism is that elevated serum albumin levels play an important role in reducing brain tissue edema and promoting hemodynamic stability. Meanwhile, hypoalbuminemia can indirectly reflect an increase in vascular permeability, and the blood-brain barrier plays a crucial role in the progression of CNS infection, thereby increasing the risk of intracranial infection.

C-reactive protein is one of the biomarkers reflecting systemic inflammation [27]. C-reactive protein levels can regulate the development of inflammation and improve chronic inflammation, and the relative stability of pro-inflammatory and anti-inflammatory cytokines helps to stabilize the inflammatory system, while their imbalance can lead to postherpetic neuralgia [28]. Some potential mechanisms, including immunosuppression, dysregulation of pro-inflammatory and anti-inflammatory factors, oxidative stress, and imbalance in antioxidant levels, are associated with the reactivation of VZV [29,30]. In the present study, we found that C-reactive protein levels were elevated in patients with VZV-associated CNS infection. The possible mechanism is speculated to be as follows. First, the disruption of patient cytokine levels stimulates an increase in C-reactive protein. Second, inflammatory-related oxidative stress increases. Third, the invasion of pathogens disrupts the body’s redox state, causing oxidative stress and promoting inflammation of the skin cells, blood, and CNS.

With the increase in an individual’s age, the incidence of HZ and its complications will gradually increase. This study identified risk factors that affect VZV-associated CNS infection, including the interval from onset to treatment, head and neck herpes, C-reactive protein, and albumin content, and the AUC of the 4 indicators combined to predict CNS infection in acute HZ patients was 0.787, with sensitivity and specificity of 64.5% and 81.9%, respectively. The indicators can early screen high-risk patients with CNS infection in acute HZ and achieve early intervention to reduce the incidence of CNS infection.

There are also some limitations in this study. First, this was a single center study, and all the patients included in this study were admitted in our hospital mainly with a large herpes area and relatively severe clinical symptoms. However, some patients with mild symptoms, such as small herpes area, mild pain symptoms, and no systemic symptoms, such as fever and headache, were unwilling to be hospitalized. Therefore, these patients did not complete the relevant examinations and could not be included in the study. There may be some bias in the patients included, but the patients in the study reflected the real situation when seeking treatment for herpes zoster, and selection bias was naturally formed and should be considered a systematic error. We conducted the multivariate logistic regression analysis to further reduce the probability of such errors. Second, when analyzing the risk factors of VZV-associated CNS infections in this study, we focused mainly on clinical laboratory testing indicators, and some clinical data, such as herpes area and presence of prodromal pain, were not included in the analysis. In the future, multicenter studies are needed to include more patients with HZ and further analyze clinical data to validate the results of this study.

Conclusions

In summary, clinicians should pay attention to early intervention in patients with acute HZ with delayed treatment, herpes distribution in the head and neck, elevated C-reactive protein levels, and decreased albumin levels. Early intervention can reduce the incidence of concurrent CNS infections.