Eastern Equine Encephalitis Virus Seroprevalence in Maine Cervids, 2012–2017

Eastern equine encephalitis virus (EEEV), an Alphavirus in the family Togaviridae, was first recognized as a cause of human disease in Massachusetts in 1938. The virus, originally isolated from a horse brain, is highly virulent in equids and has a high case-fatality rate ranging from 35% to 75% in humans.¹ Despite EEEV causing mortality and morbidity in humans and equids, these vertebrates do not contribute to the virus maintenance and transmission, and therefore are considered dead-end hosts. Enzootic cycling of EEEV is primarily maintained in nature between Culiseta melanura and songbirds.² Other mosquito species are thought to contribute to epizootic emergence as bridge vectors, primarily Coquillettidia perturbans and Aedes sollicitans, among others.¹

Although EEEV has long been known to circulate in most U.S. states east of the Mississippi River as well as Minnesota, South Dakota, and Texas,¹ virus activity has only recently been identified in the northernmost states of Vermont, New Hampshire, and Maine.³ In 2001, EEEV was initially detected in a dead songbird in the southernmost York County, Maine.⁴ Eastern equine encephalitis virus repeatedly, albeit sporadically, emerged in Maine before 2009 as evidenced by detection in two dead horses and 12 dead birds in 2005, a small batch of dead birds in 2006, and another equine case and two positive mosquito pools in 2008.^4–6 An epizootic in 2009 involving horses, llamas, and pheasants expanded the previously identified number of Maine counties, with EEEV activity from one to five.⁶ Seroprevalence studies of birds in Falmouth, ME (southern Maine), in 2013–2014 demonstrated an EEEV seroprevalence in mid-season (June–August) young birds of 12.9%, whereas seroprevalence of young birds tested late in the EEEV transmission season (September and October) had increased to 45.6%, indicating consistent enzootic activity at this locality.⁷ Despite initiation of active surveillance for human cases during the 2009 epizootic, no human cases were identified.⁵ To evaluate the distribution of EEEV activity and augment surveillance and control efforts, the CDC, Maine Centers for Disease Control, and Maine Medical Center Research Institute (MMCRI) initiated systematic serosurveys of Odocoileus virginianus (white-tailed deer) and Alces americanus (moose), as previously reported for 2009 and 2010.^4,8,9 This report describes additional serosurvey results acquired from 2012 to 2017 establishing the persistence of enzootic EEEV in Maine.

In brief, whole blood was collected from carcasses of O. virginianus and A. americanus at tagging stations during the 2011–2015 rifle hunting season in Maine.^4,8 Cavity blood was collected using disposable pipettes from field-dressed (all organs removed) carcasses presented at tagging stations. Blood samples were transferred to 7.5 mL Vacutainer tubes and stored on ice for 24–72 hours before shipping. The animal tag number and town of harvest were recorded on a log sheet, and the harvest location marked on an atlas by the hunter. Blood samples were shipped to the MMCRI Vector-borne Disease Laboratory for serum separation by centrifugation and subsequent storage at −20°C before shipping. Frozen serum samples were shipped on dry ice to the CDC, Division of Vector-borne Diseases, Arboviral Disease Branch in Fort Collins, CO, for neutralizing antibody screening and titration. Serum samples were inactivated at 56°C for 30 minutes before initial 1:10 screening for EEEV-neutralizing antibodies by plaque-reduction neutralization tests (PRNTs) at an 80% cutoff value as previously described.⁴ A Sindbis virus–EEEV chimera with structural regions derived from EEEV was used as the reference virus.¹⁰ Samples deemed positive by screening were fully titrated by PRNTs. Although Highlands J virus, an alphavirus in the western equine encephalitis antigenic complex, often co-circulates with EEEV, there is no serological cross-reactivity observed between these two viruses by PRNTs.¹¹

In 2012, 2013, 2014, 2016, and 2017, 26/393 (6.6%), 34/377 (9.0%), 35/305 (11.5%), 4/35 (11.4%), and 7/33 (21.2%) O. virginianus tested positive for neutralizing EEEV antibodies, respectively (Table 1). The following PRNT₈₀ titer frequencies were observed: 1:10 (n = 33), 1:20 (n = 25), 1:40 (n = 12), 1:80 (n = 10), 1:160 (n = 15), 1:320 (n = 7), and 1 ≥ 640 (n = 4). Because of collections being limited to the northern portion of the state in 2015, no O. virginianus samples were tested for 2015. Throughout the testing period, three of the 16 counties from which O. virginianus were tested, Franklin, Oxford, and Sagadahoc, showed no EEEV seropositivity. However, previous serosurveys detected seropositive O. virginianus from the three aforementioned counties.⁹ Seroprevalence rates for A. americanus were 15/226 (6.6%), 36/403 (8.9%), 18/203 (8.9%), 3/28 (10.7%), and 11/136 (8.1%) for 2012, 2013, 2014, 2015, and 2017, respectively (Table 1). Neutralizing antibody titer frequencies observed were 1:10 (n = 28), 1:20 (n = 15), 1:40 (n = 17), 1:80 (n = 7), 1:160 (n = 7), 1:320 (n = 5), and 1 ≥ 640 (n = 4). No A. americanus samples were tested in 2016. Of the 11 counties from which A. americanus samples were collected, Aroostook, Franklin, Piscataquis, and Somerset were found to have EEEV seropositivity. The propensity for seropositive A. americanus to be primarily found in the northernmost counties as compared with O. virginianus, which demonstrated seropositivity in all counties, is largely explained by geographic density differences between the two species. Whereas densities of O. virginianus have been described to range from approximately 8/km² (20/mi²) in the south to 2/km² (4/mi²) in the north, densities of A. americanus range from 0.1/km² (0.2/mi²) in the south to 0.7/km² (1.7/mi²) in the north.^12–14 Overall, densities of A. americanus were lower than those of O. virginianus throughout Maine and are more likely found in the northern portion of the state. Collection locations of seropositive and seronegative samples for both A. americanus and O. virginianus can be seen in Figure 1.

View Full Size

Map of Maine illustrating collection origin of cervid serum and seropositivity status of samples collected from 2012 to 2017.

Citation: The American Journal of Tropical Medicine and Hygiene 103, 6; 10.4269/ajtmh.19-0874

• Download Figure
• Download figure as PowerPoint slide

Logistic regression analysis examining year as a predictor of seropositivity showed an association with a slight, yet positive, slope (β1 = 0.026) in O. virginianus by log likelihood (P < 0.005), but not in A. americanus. However, area under the curve and goodness-of-fit evaluations of the model indicated poor predictive value (area under the curve = 0.58; Tjur’s R² = 0.0085). Similarly, it must be noted that this model cannot account for any cumulative antibody persistence across years. Despite the lack of data regarding how long neutralizing antibody endures in cervids, it is likely that the population attrition rate removing seropositive individuals as a result of general mortality as well as seasonal hunting negates any seropositivity carryover effect. However, the balance of this trade-off has yet to be established empirically.

Since the emergence of EEEV in horses, birds, and mosquitoes in the mid-2000s, it is clear the virus has demonstrated pervasive activity throughout the state of Maine. Although there have been a few examples of apparent EEEV-induced encephalitis and mortality in O. virginianus outside of Maine,^15,16 this has not yet been documented in Maine. It is intriguing to postulate whether O. virginianus or A. americanus are truly dead-end hosts or whether they may become viremic and contribute to the perpetuation of the EEEV enzootic activity. Whereas the primary vector, Cs. melanura, has been shown repeatedly to be principally ornithophilic, other implicated bridge vectors such as Cq. perturbans, Aedes canadensis, and Anopheles quadrimaculatus, among others, have been shown to have a predilection for feeding on O. virginianus and potentially A. americanus.^17,18 The predilection of A. americanus to inhabit wetlands, swamps, and marshes which act as primary habitat for multiple vectors of EEEV including Cs. melanura, Cq. perturbans, Ae. canadensis, and Ae. sollicitans increases the likelihood of exposure. In addition, cervids, specifically O. virginianus, have been established as viremic hosts for Potosi, Cache Valley, and Jamestown canyon viruses.^19,20 To date, there is no documentation of A. americanus serving as a reservoir for any arbovirus; however, the paucity of publications on the topic suggests it has yet to be fully explored.

Examinations of the closely related North American EEEV and Madariaga virus (formerly South American EEEV) in reservoir models demonstrate that juvenile rodents develop viremias greater than 300 times that of mature rodents for representative strains of both viruses, suggesting that host age may be a factor in magnitude and duration of viremia.²¹ Similarly, adult animals are more likely to have been exposed and have neutralizing antibodies removing them from the susceptible population. Albeit tenuous, the potential for O. virginianus and A. americanus to act as reservoir hosts in EEEV enzootic cycle cannot be ruled out, and further studies are warranted to explore such potentials. The seroprevalence of EEEV in Maine cervids was originally assessed in 2009–2011 following an epizootic that expanded the previous range of the virus. Continued serosurveys from 2012 to 2017 described herein demonstrate that EEEV has established and maintained enzootic cycling for 8 years following the expanded epizootic in 2009, suggesting that the virus has become endemic in Maine.