Aetiology and Potential Animal Exposure in Central Nervous System Infections in Vietnam

Zoonoses are “diseases that can be transmitted between animals and humans via direct or indirect contacts” (European Centre for Disease Prevention and Control 2022). It is estimated that nearly three-quarters of emerging human pathogens are zoonotic in origin (Woolhouse and Gowtage-Sequeria 2005). Emerging and re-emerging pathogens, which are thought to be driven by factors ranging from change in land use and agriculture to international travel, are more likely to originate from an animal source, particularly in low- and middle-income countries (Woolhouse and Gowtage-Sequeria 2005; Rabaa et al. 2015). Southeast and East Asia are defined as global ‘hotspots’ for the emergence of zoonotic infectious diseases, due to the rapidly growing economies and populations and consequent encroachment on wildlife habitats (Grace et al. 2011). Vietnam is considered to be at risk of spill-over of zoonotic pathogens into humans and has seen a number of emerging pathogens in recent years including avian influenza virus (A/H5N1) (Dinh et al. 2006), and Streptococcus suis (Wertheim et al. 2009). Close exposure to livestock and wild animals plays a role in this risk particularly via occupations which involve the handling of raw meat at slaughterhouses or wet markets which often operate with limited biosecurity measures (Carrique-Mas and Bryant 2013). Additionally, food consumption practices, in particular eating raw or undercooked meats and fish and wild animals, are also risk factors for infection with a zoonotic pathogen (Carrique-Mas and Bryant 2013). Following the emergence of SARS-CoV-2, there has been a reinforcement of the regulation of wildlife trade in Vietnam (Borzée et al. 2020).

Central nervous system (CNS) infections in Vietnam are caused by a number of pathogens, many of which are zoonotic (Tan et al. 2014, 2010; Taylor et al. 2012; Trung et al. 2012). Despite the introduction of a vaccination programme in 1997, a common cause of CNS infections in children is Japanese encephalitis virus (JEV) (Tan et al. 2010; Yen et al. 2010). JEV, a flavivirus, is transmitted in an epizootic cycle between infected Culex mosquitoes, animal hosts, including pigs and wading birds, and humans (Solomon et al. 2000). The seroprevalence of JEV in pigs in Vietnam ranges from 60% in pigs younger than 6 months of age to 100% in adults (Lindahl et al. 2013; Ruget et al. 2018). The most common cause of bacterial meningitis in adults in Vietnam is Streptococcus suis, a Gram-positive bacterium. This infection is associated with eating undercooked pig blood and intestines, or being exposed to bacteria via handling raw pig products (Grace et al. 2011; Wertheim et al. 2009; Nghia et al. 2011). However, in most studies, despite intensive investigations, the aetiology of CNS infections remains unknown in about half of patients—both children and adults (Tan et al. 2014, 2010; Taylor et al. 2012; Trung et al. 2012).

The Vietnam Initiative on Zoonotic Infections (VIZIONS) was developed with the aim of characterizing endemic infections, novel infections and infections of unknown origin through a prospective hospital-based surveillance programme. Patients with either a respiratory tract infection, enteric infection, central nervous system (CNS) infection or jaundice were enrolled as part of the hospital surveillance component (Rabaa et al. 2015). Data from those with CNS infections were used for this study with the aims of (1) determining the aetiology of patients admitted with CNS infections; (2) describing the demographics and geographical distribution of the patients; and (3) understanding whether there is an association between animal contact and CNS infections of known and unknown aetiology.

Of the 969 patients with CNS infections who were enrolled, 36 had more than one pathogen detected and were excluded from the analysis as it was not possible to ascribe cause to a single aetiology. Of the 933 patients included in the analysis, the highest number were recruited from Dak Lak General Hospital (n = 296, 31.7%). An aetiology was established in 291/933 (31.2%). Among those in whom a pathogen was detected, 247/291 (87.9%) had CSF white cell count of ≥ 5 cells/mm³ and 44/291 (15.1%) had a CSF white cell count of < 5 cells/mm³. Among those in whom no pathogen was detected, 435/642 (67.8%) had a CSF white cell count of ≥ 5 cells/mm³ and 207/642 (32.2%) had a CSF white cell count of < 5 cells/mm³. A total of 12 different bacterial pathogens were found with the most common being S. suis (n = 91 (9.8%)) and S. pneumoniae (n = 49 (5.3%)). Of the five viruses tested, the most detected were JEV (n = 72 (7.7%)) and enterovirus (n = 25 (2.7%)) (Table 1). The median age of the patients was 21 years [inter-quartile range (IQR) = 7–44]. The majority of patients were male (n = 606, 65.0%). Infections with S. suis were predominantly seen in adult males, with median age of 49 years [IQR = 39–58]. Infections with S. pneumoniae, JEV and enterovirus were predominantly seen in paediatric patients, with median age of 12 years [IQR = 2–33 years], 11 years [IQR = 6.8–15] and 13 years [IQR = 8–22], respectively. The median age of those in whom the aetiology was unknown was 23 years [IQR = 10–45] in those with a CSF white cell count of ≥ 5 cells/mm³ and 11 years [IQR = 2–36.5] in those with a CSF white cell count of CSF white cell count of < 5 cells/mm³. Half of the cases of JEV were recruited from Dak Lak General Hospital (n = 36 (50%)) and nearly half of the cases of S. suis were recruited from Hue Central Hospital (n = 43 (47.3%)). However, although there was statistical evidence of a difference in the aetiology of CNS infection by age of the patient and also, the site of recruitment (p = 0.001) there was not by gender of the patient (p = 0.062) (Fig. 1 and Table S1, supplementary data). There was evidence of a statistical difference between age groups admitted to the different hospital sites (p = 0.001) with a median age ranging from 16 years [IQR = 7–37] in Khanh Hoa to 29.5 years [IQR = 17–42] in the NHTD. No difference was seen between hospital site and gender (p = 0.102) (Fig. 2 and Table S2, supplementary data).

Three hundred and sixty-four patients (39%) raised an animal, 78 (8.4%) had slaughtered an animal and 371 (39.8%) handled, cooked or consumed raw meat, blood or viscera. In univariate analyses, patients with S. suis were more likely to keep, raise or handle an animal compared to those with another pathogen or an unknown aetiology (OR 1.6 (95% CI 1.04–2.48, p = 0.033)), and those with S. pneumoniae or an unknown aetiology with a CSF white cell count of < 5mm³ were less likely to keep, raise or handle an animal (OR 0.44 (0.21–0.83, p = 0.017) and OR 0.67 (0.48–0.93, p = 0.018), respectively). Those with an unknown aetiology with a CSF white cell count of < 5mm³ were less likely to have contact with an animal compared to those with a known pathogen or unknown aetiology with a CSF white cell count of > 5mm³ (OR 1.47 (1.08–2), p = 0.015)) (Fig. 3 and Table S3, supplementary data). Patients admitted to the hospital in Dak Lak were more likely to raise, keep or handle an animal (OR 1.91 (1.44–2.53), p < 0.001), have slaughtered an animal (OR 5.34 (3.28–8.91), p < 0.001), and handled, cooked or consumed raw meat, blood or viscera compared to other sites (OR 2.73 (2.06–3.63), p < 0.001). Patients admitted to the hospital in Dong Thap were also more likely to have handled, cooked or consumed raw meat, blood or viscera (OR 1.51 (1.04–2.18), p = 0.028) whereas those admitted to the hospital in Hue or Khanh Hoa were more likely to have had no contact with an animal (OR 1.82 (1.34–2.47), p < 0.001 and OR 2.44 (1.72–3.48), p < 0.001, respectively) (Figure S1 and Table S3, supplementary data). Adults aged 50 years and over were more likely to keep, raise or handle animals compared to other age groups (50–69 years: OR 2.43 (1.66–3.59, p < 0.001) and 70 years and over: OR 1.92 (1.08–3.42, p = 0.026)) with those aged 18–69 years more likely to have slaughtered an animal (18–49 years: OR 3.48 (2.17–5.69, p < 0.001) and 50–69 years: 2.5 (1.42–4.27, p = 0.001)) and handled, cooked or consumed raw meat, blood or viscera (18–49 years: OR 2.29 (1.74–3.02, p < 0.001) and 50–69 years: 1.7 (1.16–2.49, p = 0.007). Children aged under 5 years were more likely to have had no contact with animals (OR 2.93, 2.11–4.09, p < 0.001) (Figure S2 and Table S3, supplementary data). Females were more likely to have no contact with animals (OR 1.41, 1.08–1.85, p = 0.013) with males more likely to keep, raise or handle animals (OR 1.51, 1.14–2.01, p = 0.004) but with no difference between the genders in terms of slaughtering animals or handling, cooking or consuming raw meat, blood or viscera (Figure S3 and Table S3, supplementary data).

The most common animals kept, raised or handled were dogs (264/364 patients (72.5%)) and chickens (204/364 patients (56.0%)) both in those with and without a known aetiology. Chickens were the most common animals slaughtered (61/78 patients (78.2%)), and the raw meat, blood or viscera from pigs were the most common product handled, cooked or consumed (n = 344/371 (92.7%)) again both in those with and without a known aetiology (Figs. 4 and 5 and Table S4, supplementary data). Wild animals were only slaughtered by those with an unknown aetiology and CSF white cell count of ≥ 5 cells/mm³ (other wild bird (0.2%, n = 1) and rat (0.2%, n = 1)). However, the raw meat, blood or viscera of wild/farmed wild animals were handled, cooked or consumed by those with a known aetiology (bamboo rat (0.7%, n = 2) and pigeon (0.7%, n = 2)); those with an unknown aetiology and CSF white cell count of ≥ 5 cells/mm³ (bamboo rat (0.2%, n = 1), other wild bird (0.2%, n = 1), pigeon (0.2%, n = 1) and squirrel (0.2%, n = 1); and those unknown aetiology and CSF white cell count of < 5 cells/mm³ (deer (0.5%, n = 1) and pigeon (1%, n = 2). In the univariate binomial regression, those with S. suis were more likely to have slaughtered a pig (OR 4.88, 95% CI 1.66–12.91, p = 0.002). However, there was no difference between those with and without S. suis who kept or raised pigs; or ate, cooked or handled the raw meat, blood or viscera from pigs (OR 1.19 (95% CI 0.77–1.83, p = 0.438) and OR 1.02 (95% CI 0.65–1.59, p = 0.918), respectively). No difference was between those with and without JEV and keeping or raising pigs (OR 1.24, 95% CI 0.76–2.01, p = 0.383), slaughtering pigs (OR 1.41, 95% CI 0.86–2.28, p = 0.167) or eating, cooking or handling the raw meat blood or viscera from pigs (OR 0.70, 95% CI 0.04–3.48, p = 0.730).

In the multivariate binomial regression, there was no evidence of a difference between exposure to animals and an aetiology of S. suis, S. pneumoniae, JEV or enterovirus. Similarly, there was no difference between animal exposure and an unknown aetiology, either with or without a CSF white cell count of ≥ 5 cells/mm³. Men were more likely than women to have exposure to animals (OR 1.35, 95% CI 1.00–1.83, p = 0.047); however, there was no difference by age (Table 2).

In this multicentre study of CNS infections in Vietnam, an aetiology was not identified in 69% of patients. In a study of adults with a CNS infection of presumed viral aetiology admitted to the Hospital for Tropical Diseases, HCMC, conducted by Tan et al. (2014), the same percentage of patients had an unknown aetiology despite this study also testing for pathogens including cytomegalovirus (CMV), Epstein–Barr Virus (EBV), Nipah virus, influenza A and B virus, Mumps virus, rubella virus, rabies virus and generic flaviviruses which with the exception of JEV and DENV were not included in VIZIONS (Tan et al. 2014). However, the percentage of those with an unknown aetiology in our study was lower than reported by Tan et al. (2010) in the NHTD, Hanoi (73%) (Taylor et al. 2012). Tan et al. (2010) did not, however, test for JEV or DENV but given that they only recruited adults, it is unlikely that the absence of diagnostics for JEV resulted in a much higher proportion of unknown aetiologies. Additionally, although CSF was cultured for mycobacteria if tuberculous meningitis (TBM) was suspected, only 4/352 cases were positive (Taylor et al. 2012).

The percentage of patients with an unknown aetiology in VIZIONS is higher than that found in a multicentre study of CNS infections conducted in provincial hospitals in southern and central Vietnam by Trung et al. (2012) where the aetiology was unknown in 48% of adults and 49% of children (Trung et al. 2012) and a study of children with encephalitis admitted to the Children’s Hospital 1 in HCMC (45%) by Tan et al. (2010). In the study by Trung et al. (2012), more intensive testing for MTB was performed compared to VIZIONS. This included PCR and mycobacterial growth indicator tube (MGIT) culture of the CSF in addition to a smear for acid-fast bacilli (AFB) with 5% (n = 34) of adults and 2% (n = 11) of children diagnosed with MTB (Nghia et al. 2011). In the study by Tan et al. (2010), PCR was used to detect a wider range of pathogens including CMV, influenza A virus, Me Tri virus, human parechoviruses and generic flaviviruses (Tan et al. 2010). In summary, compared to the four previous studies conducted in Vietnam, we identified fewer patients with a known aetiology compared to two studies, the same number with a known aetiology compared to one study, and more patients with a known aetiology compared to one study.

Pigs were the second most common animals slaughtered and raw pork or pig blood/viscera the most common produced eaten, cooked or handled. The consumption of raw pig blood is common in Vietnam with 35% of respondents in surveys conducted in Hanoi reporting consumption of the dish Tiet canh (raw blood pudding) in the previous year (Huong et al. 2014). The risk of meningitis from S. suis from infected pigs is common, particularly in Asia (Mai et al. 2008) and another analysis of the VIZIONS data found that over 26% of patients admitted with an enteric, respiratory or CNS infection had had contact with pigs, while eating/handling of raw meat, blood or viscera was the most common form of contact across all hospital sites (Robertson et al. 2020). However, in our study, we only found an association between S. suis and slaughtering pigs. In addition to slaughter and consumption, 15% of the patients kept pigs determined as a possible risk factor for JEV in some studies (Liu et al. 2010; Rayamajhi et al. 2007) and Nipah virus (Goh et al. 2000). However, no association was seen in our study.

In addition to contact with livestock and poultry, exposure to wild animals may be associated with novel pathogens, which have the potential to cause CNS infections. Although numbers were very small, 2/4 people with recent history of slaughtering or handling, cooking or consuming of rodents including bamboo rats (n = 3), rats (n = 1) and squirrels (n = 1; one person reported eating both a squirrel and a bamboo rat) had a CNS infection with unknown aetiology and a CSF white cell count of ≥ 5 cells/mm³. Rodents are known hosts for a number of zoonotic pathogens with a study in the Mekong delta, Vietnam, detecting antibodies to Tick-borne encephalitis virus (TBEV) and hantavirus in both humans and rodents (Cuong et al. 2015). Rodents are also known hosts of Leptospira spp. (Cosson et al. 2014), Orientia tsutsugamushi (Wei et al. 2017) and Rickettsia typhi (Vallée et al. 2010), which have been determined as causes of CNS infections in Vietnam and neighbouring Lao People’s Democratic Republic (Dittrich et al. 2015; Nadjm et al. 2014).

Although most common pathogens were tested for in our study, further testing for zoonotic pathogens such as O. tsutsugamushi and Leptospira spp. and increased diagnostics for MTB could have reduced the proportion of patients in whom no aetiology was identified. This could include quantitative PCR (qPCR) and indirect immunofluorescence assays for O. tsutsugamushi and qPCR and microscopic agglutination tests for Leptospira spp. as described by Dittrich et al. (2015); and the inclusion of GeneXpert to aid diagnosis of TBM (Bahr et al. 2016).

This study shows that while there was no evidence of a difference between those with and without a CNS infection of unknown aetiology and animal exposure after adjusting for age, gender and site of hospital admission, contact with animals including high-risk contact such as consuming raw meat, blood or viscera is not uncommon and awareness of the risk of infection following these practices should continue to be emphasised. The expansion of testing for pathogens causing CNS infections in the region such as MTB, O. tsutsugamushi and Leptospira spp. may increase the likelihood of obtaining an aetiology while also enquiring about exposure to rats.