Streptococcus suis Meningitis: A Systematic Review and Meta-analysis

Background Streptococcus suis is the most common cause of meningitis in pork consuming and pig rearing countries in South-East Asia. We performed a systematic review of studies on S. suis meningitis to define the clinical characteristics, predisposing factors and outcome. Methodology Studies published between January 1, 1980 and August 1, 2015 were identified from main literature databases and reference lists. Studies were included if they were written in West-European languages and described at least 5 adult patients with S. suis meningitis in whom at least one clinical characteristic was described. Findings We identified 913 patients with S. suis meningitis included in 24 studies between 1980 and 2015. The mean age was 49 years and 581 of 711 patients were male (82%). Exposure to pigs or pork was present in 395 of 648 patients (61%) while other predisposing factors were less common. 514 of 528 patients presented with fever (97%), 429 of 451 with headache (95%), 462 of 496 with neck stiffness (93%) and 78 of 384 patients (20%) had a skin injury in the presence of pig/pork contact. The case fatality rate was 2.9% and hearing loss was a common sequel occurring in 259 of 489 patients (53%). Treatment included dexamethasone in 157 of 300 (52%) of patients and was associated with reduced hearing loss in S. suis meningitis patients included in a randomized controlled trial. Conclusion S. suis meningitis has a clear association with pig and pork contact. Mortality is low, but hearing loss occurs frequently. Dexamethasone was shown to reduce hearing loss.


Introduction
Bacterial meningitis is a severe infectious disease with a high mortality and morbidity. The estimated incidence is 2.6-6 per 100,000 adults per year in developed countries and several times higher in low-income settings [1]. Most pathogens causing bacterial meningitis are transmitted between humans (e.g., Streptococcus pneumoniae and Neisseria meningitidis), while others can be acquired through food ingestion (e.g., Listeria monocytogenes) [1,2]. Transmission of pathogens causing bacterial meningitis can also occur directly from animals to humans, a condition referred to as zoonotic bacterial meningitis.
One of the most common zoonotic pathogens causing bacterial meningitis is Streptococcus suis. This pathogen has its natural reservoir in pigs and may cause meningitis, endocarditis and sepsis in humans after contact with pig or pork [3,4]. Due to high pork consumption and frequent small scale pig rearing, S. suis infection is endemic in South-East Asia, where several outbreaks and cohort studies of S. suis meningitis have been reported [5][6][7][8]. Nevertheless, cases of S. suis meningitis occur all over the world [9], particularly in patients having occupational contact with pigs or pork, such as abattoir workers and butchers [10]. The clinical manifestations, epidemiology and outcome of S. suis infection in humans were described in a recent systematic review and meta-analysis [9]. This review included studies through 2012 and did not review characteristics of S. suis meningitis separately (the condition comprised 68% of cases). We performed a systematic review on studies on S. suis meningitis to define the clinical characteristics, risk factors and outcome of S. suis meningitis.

Methods
We searched the main databases (PubMed, ScienceDirect, Google scholar) for published articles describing cases of S. suis meningitis, published from January 1980 to August 2015. We used the search terms "Streptococcus suis AND meningitis", and searched the literature for cohort studies using the term "Streptococcus suis". We also searched the reference lists of the articles identified by this search strategy and selected those that we judged to be relevant. Articles written in English, Dutch, French, German, Spanish, Italian and Portuguese were included. Articles describing at least 5 patients with S. suis meningitis were included if at least one clinical characteristics or ancillary investigation was described, unless no sub-analysis for S. suis meningitis was performed (e.g. S. suis infection or streptococcal meningitis).
All articles meeting the inclusion criteria were read and systematically processed into a database of clinical data. The variables were as follows: patient characteristics, predisposing factors, clinical presentation, ancillary investigations, and outcome. Predisposing factors were defined as 1) Contact with pigs or pork, defined as preparing pork, consumption of raw pork or other swine materials (e.g. raw pig blood), occupations related to pigs or pork (e.g. abattoir workers, butchers), or breeding pigs at home [4], and/or 2) An immunocompromised status for bacterial meningitis caused by infection with Human Immunodeficiency Virus (HIV), a history of immunosuppressive medication, cancer, splenectomy, or alcoholism [2]. When patients were reported to be 'not immunocompromised', we assumed no immunosuppressive medication, splenectomy or HIV-infection in these patients. Skin injury was defined as cuts or scrapes, since skin rash could be misidentified as bruises (as seen in meningococcal sepsis).
As data description was heterogeneous between studies, all data are presented as number for which a characteristic was present out of the total number for which the characteristic was evaluated. We described the relevant characteristics using proportions with 95% confidence intervals (CIs) for categorical factors (sex, predisposing factors), and mean with standard deviation (SD) for continuous factors (age, laboratory parameters). For the latter, medians were converted to means by using proposed formulas [11].

Study characteristics
In total, 382 articles were screened for eligibility (375 by searching the databases and 7 by cross-checking references) (Fig 1). 54 articles did not meet the inclusion criteria as they described S. suis infection in animals. 304 articles were excluded from the review as no cases were described (183 articles), reporting less than 5 cases (88 articles), no sub-analysis possible for S. suis meningitis (10 articles), no S. suis infection described (9 articles), no meningitis described (7 articles), foreign language (5 articles) and duplicate articles (3 articles). The 24 articles included in the review described 913 patients [7,8,10,. The number of included patients per study varied between 5 and 151 (median 21). The median described time-period of the studies was 6 years (ranging from 1 to 23 years). Studies were performed in Thailand (8 studies), Vietnam (6 studies), Hong Kong (5 studies), the Netherlands (3 studies), China (1 study) and Japan (1 study). Studies composed 10 single center studies, 4 multi-center studies and 10 nationwide studies. 11 studies included patients prospectively and 13 were retrospective studies.

Ancillary investigations
The mean blood leukocyte count was 17.4 x 10 9 /L (SD 0.9, reported in 322 cases). The mean blood thrombocyte count was 166.3 x 10 9 /L (SD 19.1, reported in 213 cases). The mean cerebrospinal fluid (CSF) leukocyte count was 1920/mm 3 (SD 757); it was reported in 395 patients and abnormal in all 913 patients. The mean CSF protein was 2.4 g/L (SD 0.8, reported in 380 patients) and the mean CSF glucose was 1.09 mmol/L (SD 0.60, reported in 177 patients).

Treatment
The majority of patients was treated with ceftriaxone (250 patients) or penicillin (102 patients) monotherapy; no antibiotic resistance for these antibiotics was found in the 182 cases where the resistance pattern was determined. Antibiotic resistance for tetracycline was reported in 2 studies [7,33]. In some studies, patients were treated with either penicillin or ceftriaxone (101 patients), but the exact number of patients receiving either treatment was not reported [23,24,29,31]. In 454 patients, the type of antibiotic treatment was unknown. 157 of 300 patients (52%, 95% CI 44-60%) received adjunctive dexamethasone. The majority of these patients were included in a randomized controlled trial in which 71 patients received adjunctive dexamethasone and 69 patients received placebo [7]. In the other studies, dexamethasone was given at the discretion of the treating physician.

Outcome
The case fatality rate was 2.9% (17 of 581 patients, 95% CI 1.9-3.9%) and 116 of 320 patients (36%, 95% CI 31-41%) recovered without sequelae. An association between dexamethasone and death could not be established because numbers of patients who died were small. Data from the RCT showed no patients died in dexamethasone group versus three in the placebo group [34]. Hearing loss was present in 259 of 489 patients (53%, 95% CI 49-57%). 68 of these patients were screened at admission for hearing loss and this was present in 60 patients (88%, 95% CI 80-96%). According to a study describing 41 patients with hearing loss in S. suis meningitis, 38 had hearing loss on admission and 3 developed hearing loss during admission [23]. Another study described 16 patients with S. suis meningitis and hearing loss, with hearing loss persisting in 7 patients (44%) [28]. Other neurological sequelae were present in 35 of 286 patients (12%, 95% CI 8-16%) and consisted of ataxia in 19 patients, cognitive impairment in 2, tinnitus in 2, and were not specified in 12.
A randomized controlled trial showed that dexamethasone was significantly associated with a reduction in hearing loss in at least one ear (38% to 12%, p = 0.003) and a reduction in severe (>80 dB) hearing loss (odds ratio 0.23 [95% CI, 0.06-0.78]), using a multivariate analysis including age >50 and CSF bacterial load [7]. A recent case series from the Netherlands showed that despite dexamethasone treatment 6 out of 7 patients with S. suis meningitis had hearing loss upon discharge [32].

Discussion
Meningitis is the most frequently described presentation of S. suis infection, occurring in approximately 50-60% of reported S. suis infected patients [9]. Despite the geographical distribution, there were no significant differences for clinical presentation and outcome in S. suis meningitis between the different studies and low-/high-income countries. In our meta-analysis the main risk factor for S. suis meningitis was exposure to pigs or raw pork. This confirms the findings by a single center case-control study from Vietnam of 101 patients with S. suis infection which showed an odds ratio of 6.33 for occupations related to pigs [16]. Another previously reported potential risk factor was alcoholism, which we identified in 16% of patients. Alcoholism was not an independent risk factor for contracting S. suis meningitis when corrected for other predisposing factors in Vietnam [16]. However, alcoholism has been associated with an increased risk of infection in general and of an unfavorable outcome of bacterial meningitis [35].
Skin injury in the presence of pig/pork contact was described in 20% of the cases, which is similar to the previously observed 25% skin injuries in all types of S. suis infections [9]. S. suis may directly pass into the blood stream after exposure to pigs or pork in the presence of skin injuries, even without visible wound infection [10,16,36]. Patients with an increased risk of infection, e.g. because of splenectomy or use of immunosuppressive medication, should avoid direct pig or pork contact when skin lesions, particularly on the hands, are present. Skin protection has been suggested to reduce the incidence of S. suis infection [16].
Direct exposure to pigs or pork was described for 61% of meningitis cases. Direct pig exposure was documented in the majority of the European cases of S. suis infection, but was reported in less than half of the Asian cases, suggesting that other mechanisms may be involved in those patients [16]. A recent study showed that the gastro-intestinal tract is an entry site for S. suis [37], supporting the epidemiological evidence that ingestion of S. suis contaminated food is a risk factor for infection [9,16,38].
The sensitivity of the classic triad of bacterial meningitis consisting of fever, neck stiffness, and altered mental status was low (9%). This was mainly due to the low frequency of altered mental status, since other symptoms and signs of bacterial meningitis were present in a large proportion of patients. In patients with a history of regular pig exposure or pork consumption, hearing loss and these symptoms, meningitis due to S. suis should be suspected, and CSF examination should be performed to get diagnostic certainty [3].
We found that the mortality of S. suis meningitis was low (3%), especially when compared to pneumococcal meningitis (20%) and Listeria monocytogenes meningitis (36%) [39,40]. The mortality rate was also lower than reported for general invasive infection caused by S. suis (13%) [12]. The difference between mortality in S. suis meningitis and other types of S. suis infection (such as sepsis) has been noted before [6,8,9,19], but the mechanism causing this difference needs to be further elucidated [9]. Similar differences between meningitis and sepsis case fatality rates have been reported for invasive meningococcal disease [41].
The mortality rate was low but many surviving patients have sequelae. The most common sequel is hearing loss occurring in 53% of the patients; variable rates of hearing loss have been reported in other types of bacterial meningitis, with 8% in meningococcal meningitis and 22% in pneumococcal meningitis [2]. Hearing loss in S. suis meningitis may be a presenting symptom or develop during admission [23], and does not always persist [28]. Different hypotheses for hearing loss in S. suis meningitis are described in the literature such as direct infection of the auditory nerve and suppurative labyrinthitis [42]. For patients with meningitis in whom S. suis is identified, it is important to consult the otorhinolaryngologist early in the clinical course for audiometry and evaluate whether cochlear implantation is possible [43].
Dexamethasone has been shown to decrease mortality in pneumococcal meningitis and to decrease hearing loss and neurological sequelae in all bacterial meningitis cases [44,45]. For S. suis meningitis, an effect on mortality has not been established [34]. One randomized controlled trial on dexamethasone in bacterial meningitis, performed in Vietnam, included a substantial number of S. suis meningitis [34]. A subsequent analysis of all S. suis patients showed dexamethasone reduced hearing loss in a multivariate analysis [7]. As a recent case-series showed, hearing loss is still observed in patients treated with dexamethasone [32], additional randomized clinical trials on the effect of dexamethasone in S. suis meningitis would be desirable to further evaluate whether there is a benefit. However, it is unlikely such a trial is going to be performed for practical and financial reasons. Based on the available evidence, dexamethasone treatment in regions with high rates of S. suis as cause of meningitis appears reasonable to potentially reduce the very high rate of post-meningitic hearing impairment. This review has several limitations. First, most included studies show a selection bias due to a retrospective character. A recent study showed evidence of publication bias in S. suis meningitis [9]. S. suis meningitis is probably underreported, and often in numbers of less than 5 cases, which was an exclusion criterion for this study. Second, reporting of clinical characteristics, ancillary investigations and outcome was highly diverse between the included studies. We have presented the total number of patients in whom the specific characteristic was reported, but we could not perform a risk factor analysis due to heterogeneity in data. Third, cases of S. suis meningitis might have been missed due to a negative CSF culture caused by pre-treatment with antibiotics.
In conclusion, S. suis meningitis is predominantly seen in men after contact with pigs or pork and is endemic in pig rearing and pork consuming countries such as Vietnam, Thailand and China. The typical clinical presentation consists of hearing loss, fever, headache and neck stiffness, and skin injury in the presence of pig/pork contact is present in 20% of the cases.
Although the mortality of S. suis meningitis is low compared to S. suis infection in general and other causes of bacterial meningitis, 53% of patients end up with hearing loss.