Abstract
Although encephalomyocarditis virus (EMCV) infection has been commonly documented among domestic animals, less is known about EMCV transmission among humans. Recently, we described the isolation of EMCV from two febrile patients in Peru. To further investigate EMCV transmission in Peru, we screened febrile patients reporting to health clinics in Peru for serological evidence of recent EMCV infection. We also conducted a serological survey for EMCV-neutralizing antibodies in the city of Iquitos, located in the Amazon basin department of Loreto, Peru. Additionally, we screened serum from rodents collected from 10 departments in Peru for evidence of EMCV exposure. EMCV infection was found to be only rarely associated with acute febrile disease in Peru, accounting for <1% of febrile episodes analyzed. Despite the low acute disease burden associated with the virus, human exposure was quite common, as prevalence of EMCV-neutralizing antibodies ranged between 6.0% in the coastal city of Tumbes and >17% in cities in the tropical rainforest of northeastern Peru (Iquitos and Yurimaguas). On the basis of the serological survey conducted in Iquitos, risk factors for past infection include increased age, socioeconomic indicators such as residence construction materials and neighborhood, and swine ownership. Evidence from the rodent survey indicates that EMCV exposure is common among Murinae subfamily rodents in Peru (9.4% EMCV IgG positive), but less common among Sigmodontinae rodents (1.0% positive). Further studies are necessary to more precisely delineate the mode of EMCV transmission to humans, other potential disease manifestations, and the economic impact of EMCV transmission among swine in Peru.
Introduction
The impact of EMCV as a human pathogen is not well understood. Seroprevalence studies suggest a significant rate of exposure (3.2% to 50.6%) (Pope and Scott 1960, Jonkers 1961, Tesh 1978), but to date, only a handful of cases of human illness (most commonly encephalitis), based on seroconversion and virus isolation in laboratory mice, have been reported (Verlinde and Van Tongeren 1953, Gajdusek 1955, Craighead et al. 1963b, Kirkland et al. 1989). These isolation results are complicated by the possibility that the virus originated from the mouse during virus propagation instead of the inoculated human sera. Recent evidence has emerged convincingly linking EMCV to human febrile illness (Oberste et al. 2009), as EMCV closely related to previously described swine strains was isolated from two febrile patients in Peru, using cell culture in place of laboratory mice. Both patients were identified through an ongoing febrile surveillance system established between the Peruvian Ministry of the Health and the U.S. Naval Medical Research Center Detachment in Peru (U.S. NMRCD).
To further explore the role of EMCV in human febrile illness in Peru, we screened patient sera collected through the joint Peruvian Ministry of Health–U.S. NMRCD surveillance program for evidence of recent EMCV infection. In addition, we conducted a population-based survey to determine the prevalence of EMCV-neutralizing antibodies and to identify factors correlated with antibody positivity. Finally, we screened serum samples from wild rodents collected in Peru for EMCV-specific antibodies to better understand infection patterns in this potential disease reservoir.
Materials and Methods
Study design
Febrile surveillance study
Since 2000, the Peruvian Ministry of Health and U.S. NMRCD have been conducting a study called “Surveillance and Etiology of Acute Febrile Illness in Peru.” The study protocol was reviewed and approved by the Peruvian Ministry of Health and the U.S. Naval Medical Center Institutional Review Board (NMRCD.2000.0006). Detailed methods of this investigation have been described previously (Morrison et al. 2008). Briefly, patients presenting to clinics or hospitals in seven different cities in Peru (Cusco, Iquitos, Junin, Puerto Maldonado, Piura, Tumbes, and Yurimaguas) with fever ≥38°C for less than a week and at least one generalized symptom (e.g., myalgia, fatigue, and vomiting) but no clear focus of infection (e.g., cellulitis and urinary tract infection) were eligible for enrollment. Written informed consent was obtained from all adult participants (18 years and older). Parental consent was obtained for participants younger than 18. In addition, written assent was obtained from participants younger than 18 but older than 7. Serum samples were collected during the acute phase of the illness as well as 10 days to 4 weeks later during the convalescent period. Demographic data, address of residence, and clinical features of the illness were obtained with a standard questionnaire. Vital signs were also recorded.
Serological surveys and questionnaire
An outbreak of Venezuelan equine encephalitis virus (VEEV) was detected in Iquitos in 2006 by the U.S. NMRCD febrile surveillance program (Morrison et al. 2008). In response to the outbreak, serological surveys were initiated in four regions of urban Iquitos (Belen, Bella Vista Nanay, San Juan, and 22 blocks located near the center of the city). The central neighborhoods are known to be of higher socioeconomic class than the outbreak neighborhoods (ODEI-Loreto 2006). Consent was obtained from 1195 of 1314 participants to test samples for antibodies against other arthropod-borne and zoonotic pathogens, including EMCV. The protocol was reviewed and approved by the NMRCD Institutional Review Board (PJT.NMRCD.014), and the data were collected from early November through mid-December of 2006.
Phlebotomists working in teams were assigned maps and proceeded door to door explaining the study and recruiting participants. Participation was offered to all individuals 5 years of age or older. If the residents agreed to participate, the consent and assent forms were signed before blood samples were obtained. A standardized questionnaire to collect demographic and risk factor information was administered to all participants. All blood samples were initially screened for EMCV-reactive antibodies using an IgG enzyme-linked immunosorbant assay (ELISA). All samples that tested positive by IgG ELISA were further evaluated for anti-EMCV antibodies by a microneutralization (MN) assay.
Seroprevalence of EMCV in rodents
A total of 497 blood samples were collected between 2004 and 2005 from wild rodents, trapped in 10 departments of Peru (Arequipa, Ancash, Cajamarca, La Libertad, Lambayeque, Madre de Dios, Moquegua, Piura, Tacna, and Tumbes; Fig. 1). The taxonomic identification for each rodent trapped was made by the Museum of Natural History of The San Marcos University in Lima, Peru. Sera were inactivated at 56°C for 30 min, diluted to 1/100, and screened for EMCV-reactive IgG by ELISA. Approval from the NMRCD Institutional Animal Care and Use Committee was obtained before initiation of the project.
Location of Peru (inset) and study sites within Peru. Cities where human serosurveys were performed (Tumbes City, Iquitos, Yurimaguas, and Cusco) are indicated. Departments in Peru where rodent sera were collected are indicated in darker gray. ( indicates a lake.)
Laboratory assays
IgG and IgM ELISA
IgM levels were determined by using an adapted IgM-capture ELISA (Kuno et al. 1987). Briefly, 96-well plates were coated with anti-human IgM antibody, followed by the addition of human serum samples (1:100 dilution) and incubation for 1 h at 37°C. Wells were subsequently incubated with EMCV antigen for 1 hour at 37°C. Viral antigens were recognized by the addition of anti-EMCV hyperimmune mouse ascitic fluid (HMAF) for 1 h at 37°C, followed by incubation with horseradish-peroxidase-labeled anti-mouse IgG for an additional hour. Finally, (2,2′-Azinobis [3-ethylbenzothiazoline-6-sulfonic acid]-diammonium salt) (ABTS) substrate (KPL, Inc., Gaithersburg, MD) was added, and the optical density (OD) was measured using a spectrophotometer at a wavelength of 405 nm. All acute- and convalescent-phase samples were initially screened at 1:100. Positive samples were titrated using fourfold serial dilutions (1:100, 1:400, 1:1600, and 1:6400).
IgG levels for both human and rodent sera were determined by standard ELISA. Ninety-six-well plates were coated with antigen overnight and then incubated with human or rodent serum (1:100 dilution), which was then incubated with horseradish-peroxidase-linked anti-human or anti-mouse IgG.
For both IgG and IgM studies, viral antigen stock was prepared from the lysates and supernatant, respectively, of Vero-E6 (African green monkey) cells infected with EMCV isolated from a febrile patient in Iquitos (IQD6726) (Oberste et al. 2009). The supernatant was inactivated with 3 mM binary ethylenimine. HMAF for EMCV was used as positive control, and normal ascitic fluid as negative control. After the addition of colorimetric substrate, absorbance was read at 405 nm and OD reported. The positive cutoff OD value was calculated from the mean adjusted OD of antibody-negative control sera plus three standard deviations.
Both acute and convalescent blood samples were tested for a range of viruses, including EMCV (Oberste et al. 2009). For this study, samples were considered positive for recent EMCV infection if elevated titers (≥1:100) were observed in either acute or convalescent serum. All samples collected through the febrile surveillance system were also screened for all four dengue virus serotypes, VEEV, Mayaro virus, Oropouche virus, Group C viruses, and hantaviruses through both cell culture and serological methods.
MN test
For MN assays, ∼80–100 median CCID50 of isolate IQD6726 and twofold serial dilutions of serum (starting at 1:20 and ending at 1:1280) were incubated at 37°C for 1 h before Vero-E6 cells were added to the wells. After incubation for 5 days at 37°C, each plate was stained with 0.1% napthol blue black and 1.36% sodium acetate in 6% acetic acid. Convalescent samples from patients from whom EMCV had been isolated were used as positive controls. Serum samples were considered positive in the case of 50% or greater virus neutralization at a 1:40 dilution, based on the absence of observed cytopathic effects after staining. Negative controls included antibodies against two other Picornaviridae family members: human sera with Hepatitis A virus neutralizing antibodies and HMAF for Coxsackie Virus.
Statistical analysis
Proportions were compared using a chi-square test in SPSS (SPSS Version 16, 2007; SPSS Inc., Chicago, IL.). Risk factors for infection with EMCV were evaluated by bivariate and multivariate logistic regression using SPSS. Models were constructed with the dichotomous dependent variable (MN positive for EMCV antibody) and the following independent variables: age (5–10 years, 11–20 years, >20 years), place of residence in Iquitos (Belen, Bella Vista Nanay, San Juan, and central neighborhoods), housing material (concrete/brick or wood), travel history (report of multiple day trips outside Iquitos or an overnight trip), and animal ownership (cats and swine).
Results
Evidence for human disease associated with EMCV infection
After the first isolations of EMCV from two febrile patients in Peru in 2004 (Oberste et al. 2009), we screened samples from febrile participants reporting to a clinic or hospital in seven cities in Peru between January 2004 and December 2008 for EMCV-reactive IgM. From 10,419 febrile participants we identified 14 cases, in addition to the 2 previously reported (Oberste et al. 2009), with evidence for recent EMCV infection based on IgM ELISA (Table 1). Four cases with IgM seroconversions were also tested for virus neutralization activity. All four lacked EMCV-neutralizing antibodies in the acute sample but had robust EMCV-specific neutralizing response in the convalescent sample. Disease most commonly presented as undifferentiated febrile episodes, although cardiac complications were detected in one patient (Table 1; Patient 13, 16-year-old girl). In addition to fever, the most common symptoms described were malaise (13/14), arthralgia and myalgia (11/14), headache (11/14), chills (11/14), nausea/vomiting (8/14), retro-orbital pain (7/14), rash/petechiae (5/14), and cough (5/14). Three participants had concurrent serologic evidence of recent dengue virus infection, including the participant with cardiac complications. Overall, 73.2% (7622/10,419) of participants tested for EMCV-reactive IgM were from cities in the Loreto Department (Iquitos or Yurimaguas). Nearly all participants positive for EMCV IgM (13/14; 92.9%) were from Iquitos or Yurimaguas (Table 1), whereas no EMCV IgM-positive samples were detected in sites along the desert coast (Piura and Tumbes; 0/885) or the highlands (Cusco; 0/386).
“IgM” indicates results by IgM ELISA and “MN” indicates results from microneutralization assay.
Neg., negative at 1:100 dilution for IgM ELISA or 1:20 dilution for MN assay.
n/a, information or serum sample not available; nd, not determined.
ELISA, enzyme-linked immunosorbant assay.
To determine if EMCV infection was more prevalent in the Loreto Department than other regions of Peru, we conducted a cross-sectional antibody prevalence survey for prior EMCV exposure, utilizing the convalescent samples from febrile patients reporting to clinics in or around Cusco, Iquitos, Tumbes, and Yurimaguas (Table 2). Convalescent serum samples from febrile patients were screened by IgG, followed by confirmatory MN assay. The prevalence of EMCV-neutralizing antibodies was highest in cities in Loreto (19.1%; 322/1689 for Iquitos and Yurimaguas), with significant differences between Yurimaguas and Cusco and Tumbes (Table 2; p = 0.01 and p < 0.001, respectively).
Participant samples were initially screened by IgG and subsequently confirmed by neutralization assay.
EMCV antibody prevalence in Iquitos
Population
To further investigate the epidemiology of EMCV transmission in the Loreto Department, we utilized serum samples originally collected through a cross-sectional survey of healthy participants (n = 1195) to study VEEV transmission in Iquitos (Morrison et al. 2008). The mean age of the 1195 participants was 29.3 years (range 5 to 87); 11% were 10 or younger, 27% were between 11 and 20, and 62% were 21 or older. Sixty-seven percent of participants were female. The most commonly reported occupations were housewife (43%) and student (32%). Roughly one-third of participants reporting keeping cats (32.1%) and dogs (35.5%), whereas only 1.9% reported keeping swine. Approximately one-third of participants (36.4%) also reported having left Iquitos for a day trip or overnight excursion in the prior 6 months.
There were significant demographic differences among the neighborhoods, as previously reported (Morrison et al. 2008, Forshey et al. 2010). Houses made of brick or concrete (vs. wood) were more common in the central neighborhoods (66.2% vs. 8.2% in Belen, 8.8% in Bella Vista Nanay, and 19.7% in San Juan). Inhabitants of central neighborhoods were also more likely to keep domestic pets (50.3% vs. 27.3% in Belen, 25.7% in Bella Vista Nanay, and 29.6% in San Juan). Residents of Belen and Bella Vista Nanay were more likely to report rodents in or around the house than those in the central zones (95.2% in Belen, 85.8% in Bella Vista vs. 79% in the control neighborhoods).
EMCV antibody prevalence
Serum samples were screened for EMCV IgG by ELISA. ELISA-positive samples were subsequently confirmed by MN assay (Table 3). Of the 1195 participants, 444 (37.2%) were positive for EMCV antibodies by ELISA; of these 251 (56.5% of the IgG positive participants) were confirmed by neutralization assay. Antibody prevalence was highest in the neighborhoods of Belen (29.6%) and San Juan (23.0%) and was lowest among participants from the central neighborhoods (14.1%).
Participant samples were initially screened by IgG and subsequently confirmed by neutralization assay.
Risk factors
In univariate analysis the following variables were found to be associated with EMCV antibody positivity: age older than 21 years, residence in Belen or San Juan, living in a house made of wood, and keeping pigs (Table 4). Participants with jobs in agriculture, street vendors, and housewives had a significantly higher association with seropositivity than did students, likely due to at least in part to age differences. Factors associated with decreased proportions of seropositivity included a recent history of travel (day or overnight trips) and cat ownership. Factors not significantly associated with EMCV antibody prevalence included sex, rodent sightings in the home, and dog or chicken ownership (Table 4; data not shown).
OR, odds ratio; 95% CI, confidence interval.
Multivariate logistic regression was performed with antibody positivity as the outcome variable, incorporating variables significantly associated with antibody positivity in univariate analyses in stepwise backward fashion. The majority of the risk factors identified in univariate analysis remained in the model (age >21 years, residence in Belen or San Juan, dwelling made of wood, swine ownership, and no history of recent travel) with the exceptions of cat ownership and the occupation-associated risk factors (Table 5).
Evidence for EMCV infection in rodents in Peru
A survey of EMCV antibody prevalence was performed using rodents trapped in fields, farms, and peridomestic settings during 2004 and 2005. A total of 497 rodents from 23 species (298 Murinae and 199 Sigmodontinae) were trapped in 10 different departments of Peru (Table 6; Fig. 1). ELISA was used to detect the presence of EMCV-reactive IgG. Thirty individual rodents in 5 different species tested positive for EMCV IgG: 4 of 14 (29%) Rattus norvegicus, 19 (13%) of 151 R. rattus, 5 (4%) of 133 Mus musculus, 1 (4%) of 24 Phyllotis limatus, and 1 (4%) of 25 Akodon molli. Only 2 of the 199 New World Sigmodontinae rodents tested showed evidence of EMCV infection (Table 6), including 1/41 Akodon spp., 1/101 Phyllotis spp., 0/25 Oligoryzomys spp., 0/10 Calomys sorellus, 0/8 Oryzomys spp., 0/7 Bolomys lasiurus, 0/3 Abrothrix andinus, 0/3 Thomasomys taczanowskii, and 0/1 Holochilus sciureus.
Discussion
On the basis of the results presented here, EMCV transmission is common among humans in Peru, but acute febrile disease only rarely results from infection. On the basis of data from four different cities in Peru, >17% of the population had EMCV-neutralizing antibodies. In a study of banked serum samples collected conducted in 1978, Tesh found that worldwide EMCV antibody prevalence varied greatly among adults, ranging 3.2% in Omo Valley, Ethiopia, to 50.6% on the island nation of Vanuatu (Tesh 1978). Prevalence of EMCV antibodies varied widely within Peru as well. Of the four locations studied, seroprevalence was lowest in the coastal desert site (Tumbes) and the highlands site (Cusco), and highest in the two tropical rainforest sites (Iquitos and Tumbes). It is interesting to note that a serosurvey of Peruvian Indians conducted by Jonkers in 1961 (Jonkers 1961) found a nearly identical prevalence of EMCV-neutralizing antibodies among Cusco residents as found in this current study (10.5% vs. 10.9%), although possible age differences between the two sampled populations should be considered. The reasons for the region-specific differences in EMCV transmission are unclear. One possibility is that there may be sociological factors that promote increased exposure to infected reservoirs in the tropical rainforest regions. Alternatively, viral persistence or transmissibility may be greater in regions with elevated temperatures and rainfall.
On the basis of the serosurvey of healthy participants, we found that roughly one in five residents of Iquitos, Peru (21.0%), had serologic evidence for prior exposure to EMCV. This number was comparable to the prevalence calculated from the febrile surveillance study, although clearly differences exist even within Iquitos. In our study, participants were designated seropositive when a positive IgG ELISA was confirmed by MN. We observed an unexpectedly high false-positive rate with the ELISA: only 56.5% of the positive ELISAs were confirmed as positives by MN (false-positive rate: 43.5%). This could be due to cross-reactivity of the IgG ELISA with other Cardioviruses or other members of the Picornaviridae family. In a recent study Zoll and co-authors demonstrated that a newly identified Cardiovirus, Saffold virus, was nearly ubiquitous in populations in The Netherlands, Mali, Cameroon, and Indonesia (Zoll et al. 2009). On the basis of these results, it is likely that Saffold virus or related viruses circulate in Peru and could cross-react in the IgG ELISA.
Statistically significant predictors of seropositivity fell into three categories: swine exposure, indicators of socioeconomic status, and age. Swine ownership was the strongest independent predictor of seropositivity (odds ratio = 2.64, confidence interval 1.08–6.46). It is interesting to note that the EMCV strains previously isolated from patients in Peru were genetically similar to porcine strains identified elsewhere (Oberste et al. 2009). EMCV-related disease is well documented in pigs, but to our knowledge no direct evidence for swine–human transmission has been reported (Murnane et al. 1960, Craighead et al. 1963a, Gainer et al. 1968, Acland and Littlejohns 1975, Dea et al. 1991, Brewer et al. 2001, Maurice et al. 2005, 2007). Considering the high mortality and reproductive failure documented in pigs and the resultant economic losses, our data suggest that swine herds in Peru should be monitored for EMCV activity.
In addition to swine exposure, a number of factors related to socioeconomic status (poorer neighborhoods, cheaper housing materials, and no history of recent travel) were also independent predictors of prior EMCV infection. Assuming poorer residents of Iquitos have decreased access to clean bathing and drinking water, these data support a risk of transmission related to lower levels of hygiene (fecal–oral or otherwise). In this study we also observed a significant association between prior exposure and increasing age. Previous studies have indicated that for EMCV and other Cardioviruses infection is most commonly due to childhood exposure (Jonkers 1961, Tesh 1978, Zoll et al. 2009). Here we also observed substantial transmission among younger participants (∼13% among the 5–10-year-old age group); however, EMCV exposure seems to continue to later in life, as the oldest age group (>21 years) had significantly higher antibody prevalence.
On the basis of the small sample size presented here, EMCV appears to be a common infection of Murinae rodents in Peru and significantly less common in Sigmodontinae rodents. As no rodent captures were conducted in the Loreto department, it is difficult to extrapolate the results of the rodent study to the prevalence of EMCV-specific antibodies in Iquitos. In Iquitos, while no specific correlation between rodent sightings and seropositivity was made, the percentage of participants reporting rodents in their homes was higher in the neighborhood with highest prevalence of EMCV-neutralizing antibodies (Belen) than in the central neighborhood for example (95.2% vs. 80%). Further, in univariate analysis cat ownership was associated with a protective effect against EMCV antibody positivity, suggesting the possibility that rodent control in the home is important for preventing infection. As with the role of swine in human EMCV infection, clearly delineating the role of rodents in EMCV transmission in Iquitos will require further investigation. Further studies should also be focused on the role of unhygienic practices in EMCV transmission. While our data suggest that acute febrile disease is not commonly associated with EMCV infection, as observed with heterologous Cardioviruses (Chiu et al. 2008, Drexler et al. 2008, Blinkova et al. 2009, Li et al. 2009) other potential acute disease manifestations such as gastroenteritis (Liang et al. 2008, Ren et al. 2009) as well as chronic sequelae should be investigated.
Footnotes
Acknowledgments
We thank Carolina Guevara for laboratory support; Stalin Vilcarromero, Miguel Villanueva, and Tatiana Saldarriaga for support at the field sites; and Rebeca Carrion for coordination of field personnel in Iquitos. We also thank the local health authorities, including DIRESA-Loreto, for their support of this and other ongoing studies.
Disclaimers
The study protocols (NMRCD.2000.0006 and PJT.NMRCD.014) were approved by the U.S. NMRC and NMRCD Institutional Review Boards in compliance with all U.S. Federal regulations governing the protection of human subjects. The experiments reported herein were conducted in compliance with the Animal Welfare Act and in accordance with the principles set forth in the Guide for the Care and Use of Laboratory Animals, Institute of Laboratory Animals Resources, National Research Council, National Academy Press, 1996. This study was funded by the U.S. Department of Defense Global Emerging Infections Systems Research Program, WORK UNIT NUMBER: 847705.82000.25GB.B0016. The sponsor had no role in this study other than providing funding. The views expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, or the U.S. Government. Some authors of this artile are military service members or employees of the U.S. Government. This work was prepared as part of their official duties. Title 17 U.S.C. § 105 provides that “Copyright protection under this title is not available for any work of the United States Government.” Title 17 U.S.C. § 101 defines a U.S. Government work as a work prepared by military service members or employees of the U.S. Government as part of those persons' official duties.
Disclosure Statement
No competing financial interests exist.