Cache Valley virus: an emerging arbovirus of public and veterinary health importance

Abstract

Cache Valley virus (CVV) is a mosquito-borne virus in the genus Orthobunyavirus (Bunyavirales: Peribunyaviridae) that has been identified as a teratogen in ruminants causing fetal death and severe malformations during epizootics in the U.S. CVV has recently emerged as a viral pathogen causing severe disease in humans. Despite its emergence as a public health and agricultural concern, CVV has yet to be significantly studied by the scientific community. Limited information exists on CVV’s geographic distribution, ecological cycle, seroprevalence in humans and animals, and spectrum of disease, including its potential as a human teratogen. Here, we present what is known of CVV’s virology, ecology, and clinical disease in ruminants and humans. We discuss the current diagnostic techniques available and highlight gaps in our current knowledge and considerations for future research.

Introduction

Cache Valley virus (CVV) was first isolated in 1956 from Culiseta inornata mosquitoes collected in Utah’s Cache Valley (Holden and Hess 1959). Since the original description of CVV, the virus is known to be widely distributed across North America, including 22 US states, 4 Canadian provinces, and Mexico (Calisher et al. 1986, Blitvich et al. 2012a) (Fig. 1). Although CVV was demonstrated to cause encephalitis in a mouse model in 1956 (Holden and Hess 1959), it was not recognized as a pathogen until an outbreak of severe congenital central nervous system and musculoskeletal system dysfunction was described in sheep in 1987 (Edwards et al. 1989). Nearly an additional decade would pass before CVV would be recognized as a human pathogen (Sexton et al. 1997). Despite the long history of CVV and severe disease manifestations in humans and animals, CVV remains an understudied arbovirus.

Fig. 1.

Detection of Cache Valley virus in North America. The locations of samples collected with Cache Valley virus detection are shown by a circle corresponding to the state or province. The locations are colored by the source of CVV: mosquito (yellow), vertebrate (blue), or human (red), as documented in the literature (Calisher et al. 1986, Blitvich et al. 2012a, Waddell et al. 2019, Dieme et al. 2022b). Locations with multiple sources display overlapping circles off centered.

CVV is the exemplar isolate of the species Cache Valley orthobunyavirus, classified in the family Peribunyaviridae, genus Orthobunyavirus (Hughes et al. 2020). Tlacotalpan, Cholul, and some strains of Playas viruses are also included in the Cache Valley orthobunyavirus species. Serological tests such as complement fixation, hemagglutination-inhibition, and neutralization assays can antigenically separate orthobunyavirids into 1 of at least 18 serogroups (Casals and Whitman 1960, Calisher 1996). CVV was the first virus in the Bunyamwera serogroup to be isolated in the United States. While several additional viruses in the Bunyamwera serogroup are distributed in North America, including Tensaw, Maguari, Northway, Potosi, Main Drain, Fort Sherman, and Lokern viruses, Maguari virus is the only other known human pathogen (Groseth et al. 2017).

CVV infection is increasing as an important public health and one health concern. To date, there have been 6 documented outbreaks of CVV in sheep herds in the United States. and Canada, as recently as 2013 (Waddell et al. 2019), resulting in significant agricultural losses. Outbreaks on farms have been characterized by unusually high rates of stillbirths and fetal abnormalities. Human cases of CVV infection are rare but often present as severe central nervous system disease. Given the teratogenicity of CVV infection in sheep, infection in humans has been theorized to cause in-utero malformities. However, few studies have been conducted, and knowledge of the effects of CVV infection during human pregnancy is unknown and underwhelmingly studied.

This review will focus on the current state of knowledge for CVV. Descriptions of virology, transmission, clinical disease, and diagnosis will be presented. Lastly, areas for further research into this orphan pathogen will be highlighted.

Virology

Although CVV is an understudied virus, orthobunyavirids share many defining virological features. Virions are enveloped, spherical or pleomorphic, and 80–120 nm in diameter (Martin et al. 1985). The surface of the virion is a lattice of glycoprotein spikes in a tripod-like arrangement (Bowden et al. 2013), which are responsible for cellular attachment (Plassmeyer et al. 2005) and are the target of neutralizing antibodies (Hellert et al. 2019). Within the virion, the CVV genome consists of 3 segments of negative-sense, single-stranded RNA that exist in a coiled, circular structure (Obijeski et al. 1976). The terminal ends of each segment possess the canonical, conserved, and complementary sequence (in coding sense) 5ʹ-AGTAGTGT…ACACTACT-3ʹ (Obijeski et al. 1980). These conserved sequences form a pan-handle structure important in genome replication (Barr et al. 2003).

The 3 RNA segments are named relative to their size, large (L) segment, medium (M) segment, and small (S) segment (Fig. 2). The L segment (6,870 nt) encodes the L protein that possesses a central RNA-dependent RNA polymerase domain (Endres et al. 1989) and an amino-terminal endonuclease domain responsible for “cap snatching” that primes mRNA synthesis (Reguera et al. 2010). The M segment (4,463 nt) encodes for the structural glycoproteins Gn and Gc, and a nonstructural protein designated NSm between the Gn and Gc coding region (Fazakerley et al. 1988). The NSm protein is targeted to the Golgi complex and is thought to be involved in virion assembly (Fontana et al. 2008); however, it is dispensable for the generation of infectious virus (Tilston-Lunel et al. 2015). The S segment (950 nt) encodes for the nucleoprotein (N) that encapsidates the RNA and a nonstructural NSs protein in an overlapping reading frame (Fuller et al. 1983, Dunn et al. 1994). The NSs protein has been demonstrated to act as a virulence factor by disrupting the interferon response in host cells (Dunlop et al. 2018).

Fig. 2.

Cache Valley virus genome organization. The general coding strategy for CVV is displayed for each segment. The viral cRNA is displayed in the 3ʹ–5ʹ direction, while the open reading frames are 5ʹ–3ʹ. The open reading frames show the relative sizes of the L protein; glycoproteins Gn and Gc, NSm; N, nucleocapsid, and overlapping NSs. Post-translational cleavage sites are indicated with triangles on the M open reading frame. Created with BioRender.

Cache valley orthobunyavirus strains circulate regionally as demonstrated genetically and antigenically. CVV isolates from Mexico were shown to have one-way cross-neutralization and group genetically with Tlacotalpan virus in a monophyletic clade distinct from US and Canadian CVV strains (Pabbaraju et al. 2009, Blitvich et al. 2012a). These distinct phylogenetic groups separated CVV into 2 lineages with spatial structures. Recently, however, replacement of lineage I in the United States and Canada with lineage II has been documented (Armstrong et al. 2015, Dieme et al. 2022b). Lineage II CVV strains, previously circulating in southern Mexico in 2008, were detected in mosquito pools in Connecticut in 2010 and became the dominant lineage in the state by 2014 (Armstrong et al. 2015). Similar surveillance in New York and Canada revealed lineage I displacement between 2010 and 2016 (Dieme et al. 2022b). Of particular interest, this study also documented increased vector competence of lineage II CVV strains by Anopheles species mosquitos (Dieme et al. 2022b), suggesting this as a mechanism for the displacement of lineage I CVV strains. Reassortment events whereby segments from 2 different lineages produce a progeny virus with a mixture of lineage I and II segments have also been documented. This phenomenon was first described for CVV during a 2015 case investigation of atypical CVV disease in a human (Baker et al. 2021) where the infecting virus had a lineage I L segment, but the S and M segments were lineage II. Similar reassortant viruses were described in 2015 in mosquitoes, in 2016 in a human (Dieme et al. 2022b), and in a 2020 case of transfusion-transmitted CVV (Al-Heeti et al. 2022). While these reassortant viruses have been shown to have higher transmissibility than pure lineage I viruses, lineage II viruses have a higher degree of vector competence (Dieme et al. 2022b). Continued surveillance and complete genome sequencing will be vital to understanding the role of these reassortant viruses in transmission and disease.

Virus Transmission and Prevalence

Like other viruses in the genus Orthobunyavirus, CVV is maintained in a dual-host life cycle between arthropods and vertebrate amplifying hosts and is similar to California serogroup Jamestown Canyon virus (JCV) and Bunyamwera serogroup Potosi virus (POTV) in that deer have been primarily implicated as the reservoir host (Fig. 3) (Neitzel and Grimstad 1991, McLean et al. 1996, Blackmore and Grimstad 1998). CVV isolations have been made from at least 44 species of mosquitoes (Kokernot et al. 1969b, Buescher et al. 1970, Iversen et al. 1979, Calisher et al. 1986, Main and Crans 1986, Mitchell et al. 1998, Ngo et al. 2006, Armstrong et al. 2013, Andreadis et al. 2014, Yang et al. 2018b, Waddell et al. 2019, Dieme et al. 2022a, 2022b). Though culicoides are implicated as vectors for other Bunyamwera serogroup viruses (Mellor et al. 2000), biting midges have not yielded isolations of CVV, nor have they demonstrated competence as a vector (Reeves and Miller 2013, Johnson et al. 2014). Hence, they are unlikely to contribute to circulation of the virus. Mosquito isolations have been most frequently acquired from Anopheles punctipennis (Say), Anopheles quadrimaculatus (Say), Aedes sollicitans (Walker), Coquillettidia perturbans (Walker), Cs. inornata (Williston), Aedes trivittatus (Coquillett), and Aedes canadensis (Theobald) among many others. However, the frequency of CVV isolations from mosquito species varies widely by geography, including nearly all regions of the United States, Alberta, Manitoba, and Ontario in Canada, Mexico, and Jamaica (Table 1) (Calisher et al. 1988, Farfan-Ale et al. 2010, Andreadis et al. 2014, Waddell et al. 2019, Dieme et al. 2022a, 2022b) (Fig. 1).

Table 1.

Mosquito species yielding Cache Valley virus isolations by region

Northeast United States	Midwest United States	Upper Midwest	Southeast/Atlantic Coast United States	Gulf Coast United States	Pacific Northwest/Western United States	Canadaa	Mexicob	Jamaica
Ae. albopictus	Ae. albopictus	Ae. canadensis	Ae. sollicitans	Ae. sollicitans	Cs. inornata	Ae. vexans	Ps. columbiae	Ae. taeniorhynchus
Ae. canadensis	Ae. canadensis	Ae. cinereus	Ae. taeniorhynchus	Ae. taeniorhynchus		Cs. inornata	Ae. taeniorhynchus
Ae. cantator	Ae. trivittatus	Ae. trivitattus	An. punctipennis	Ps. columbiae		Cx. tarsalis
Ae. cinereus	Ae. vexans	Ae. vexans	An. quadrimaculatus			Ps. ferox
Ae. japonicus	An. punctipennis	An. punctipennis	Ps. ferox
Ae. sollicitans	An. quadrimaculatus	An. quadrimaculatus
Ae. taeniorhynchus	Cq. perturbans	Cq. perturbans
Ae. triseriatus	Cx. tarsalis	Cs. inornata
Ae. trivittatus	Ps. ferox	Cx. tarsalis
Ae. vexans
An. punctipennis
An. quadrimaculatus
An. walkeri
Cq. perturbans
Cs. melanura
Cx. salinarius
Ps. ferox

^aIncludes data from Saskatchewan, Manitoba, Ontario.

^bIncludes data from Tamaulipas and Yucatan.

Fig. 3.

Transmission cycle of Cache Valley virus. Cache Valley virus circulates in an enzootic cycle with several mosquito species as potential vectors. CVV is amplified in an unknown vertebrate host and can be transmitted to humans and dead-end hosts. Specific vectors important in human transmission have not been described. Created with BioRender.

The large number of species and geographic range makes it difficult to incriminate primary vectors for CVV transmission. However, vector competence has been demonstrated experimentally for 11 common species: Aedes aegypti (Linnaeus), Aedes albopictus (Skuse), Aedes japonicus (Theobald), Aedes sollicitans, Aedes taeniorhynchus (Wiedemann), Aedes vexans (Meigen), An. punctipennis, Culex tarsalis Coquillett, An. quadrimaculatus, Cq. perturbans, and Cs. inornata (Yuill and Thompson 1970, Saliba et al. 1973, Corner et al. 1980, Blackmore et al. 1998, Ayers et al. 2018, 2019, Yang et al. 2018a, Dieme et al. 2022a, 2022b). Though the bionomics of these vectors vary greatly, they each readily feed on nonhuman mammals (Burkot and DeFoliart 1982, Savage et al. 1993, Molaei et al. 2008, Kent et al. 2009, Pruszynski et al. 2020, Little et al. 2022, Walter Reed Biosystematics Unit 2023.

Andreadis et al. 2014 noted that in addition to having no predictable yearly pattern, most CVV isolations from mosquitoes occur late in the collection season in August and September (Buescher et al. 1970, Calisher et al. 1986, Ngo et al. 2006, Andreadis et al. 2014). Interestingly, the longitudinal data for JCV, which also utilizes deer as a reservoir host, indicates the greatest frequency of isolates occur between mid-June to mid-July, though JCV-infected mosquitoes are detected from June through September (Andreadis et al. 2008). When examining the Connecticut collection data, there are several early spring mosquitoes in which only JCV and not CVV are detected between June and July, including Culex restuans (Theobald), Aedes abserratus (Felt and Young), Aedes aurifer (Coquillett), Aedes communis (De Geer), Aedes excrucians (Walker), Aedes provocans (Walker), Aedes sticticus (Meigen), and Aedes stimulans (Walker) (Andreadis et al. 2008, 2014). This distinct seasonality between the 2 viruses is interesting, considering they both are reasoned to utilize white-tail deer as reservoir hosts. Studies comparing the seropositivity of JCV and CVV in white-tailed deer in Minnesota indicated that the geometric mean antibody titers for JCV were generally lower than CVV end point titrations (Neitzel and Grimstad 1991). It has been proposed that these higher titers, in concert with the high rates of CVV seropositivity in deer, could result in a diminished availability of CVV naïve hosts early in the mosquito season and delay the transmission period (Andreadis et al. 2014, Clarke et al. 2022).

To date, the evidence for CVV transovarial transmission is lacking, with the only analysis in Cs. inornata demonstrating 2.9–3.3% of adult females transmitted to offspring with an average filial infection rate of 0.2% (Corner et al. 1980). In contrast, filial infection rates of California serogroup viruses have been shown to range from 2.7% to 71%, depending on the virus and vector (Tesh and Gubler 1975, Miller et al. 1977, Christensen et al. 1978, Turell et al. 1982, Woodring et al. 1998, Hughes et al. 2006). Comparatively lower rates of transovarial transmission and filial infection rates could also contribute to a later seasonal peak for CVV.

In addition to white-tail deer (Odocoileus virginianus) and mule deer (Odocoileus hemionus hemonius) (Buescher et al. 1970, Main and Crans 1986, Neitzel and Grimstad 1991, Aguirre et al. 1992, McLean et al. 1996, Blackmore and Grimstad 1998), CVV seropositivity has been detected in multiple mammals, including jackrabbits (Lepus californicus) and Eastern cottontail rabbits (Sylvilagus floridanus) (Aguirre et al. 1992, Blackmore 1996), and horses (Equus ferus) (Kokernot et al. 1969b, Buescher et al. 1970, McLean et al. 1987, Blackmore 1996, Blitvich et al. 2012c), wild boar (Sus scrofa) and domestic pigs (Sus scrofa domesticus) (Buescher et al. 1970, Blackmore 1996), cattle (Bos taurus) (Kokernot et al. 1969b, Yuill et al. 1970, McConnell et al. 1987, Blackmore 1996, Walters et al. 1999, Sahu et al. 2002, Uehlinger et al. 2018), swift foxes (Vulpes velox) and kit foxes (Vulpes macrotis) (Buescher et al. 1970, Miller 1997), domestic dogs (Canis lupus familiaris) (Kokernot et al. 1969b), raccoons (Procyon lotor) (Kokernot et al. 1969b, Buescher et al. 1970), and groundhogs (Marmota monax) (Buescher et al. 1970). Isolated studies have also detected antibodies in brown rats (Rattus norvegicus) and meadow voles (Microtus pennsylvanicus) in Maryland and Virginia (Buescher et al. 1970). The seroprevalence history highlights the geographical range of CVV circulation, which include detection of the virus in Mexico, United States, and Canada (Fig. 1). It should be noted that additional studies originally identifying CVV seropositivity in Jamaica, Trinidad and Tobago, Guyana, Panama, and Argentina, were all using a virus strain that has since been reclassified as Maguari virus (Downs et al. 1961, Sabattini et al. 1965, Belle et al. 1966, Groseth et al. 2017).

In Mexico, cattle (46.7%) (Scherer et al. 1967) and horses (26.4%) (Blitvich et al. 2012c) produced the highest seroprevalence for CVV in studies with sample sizes greater than 10. Testing throughout all regions of the United States showed horses (66–94%) (Buescher et al. 1970, McLean et al. 1987), cattle (51–90%) (Buescher et al. 1970, Yuill et al. 1970, McConnell et al. 1987), white-tailed deer (16–79%) (Buescher et al. 1970, Neitzel and Grimstad 1991, McLean et al. 1996, Blackmore and Grimstad 1998), and Roosevelt elk (Cervus elaphus roosevelti) (40%) (Eldridge et al. 1987) yield the most CVV seropositivity. Sampling in Canada followed suit with horses (69%) and mule deer (50.8%), exhibiting the highest CVV exposure rates. Goats (Capra aegarus) and various sheep breeds (Ovis aries musimom, O. aries aries, and O. ammon) tested in these regions typically ranged from 3% to 35% seropositivity (McConnell et al. 1987, Chung et al. 1991, Blitvich et al. 2012c, Meyers et al. 2015); however, flocks with previous incidence of CVV-associated congenital malformations in Texas and Saskatchewan demonstrated seroprevalence ranging as high as 71% and 64%, respectively.

Few experimental studies have pinpointed a natural amplifying host for CVV. Following the original isolation, Holden and Hess challenged 0.5-day-old chicks and observed no viremia or development of neutralizing antibodies (Holden and Hess 1959). Lack of viremia in chicks and a dearth of seropositivity in birds (Buescher et al. 1970, Blitvich et al. 2012c) suggest birds are not a likely amplifying reservoir. In experiments with needle-inoculated eastern cottontails (S. floridanus), animals did acquire a low viremia for 1–3 days, but adult Cq. perturbans were unable to become infected following oral feeding on needle-inoculated rabbits. Though needle inoculation has limitations in comparison to natural infection by mosquitoes, orally infected Cq. perturbans were unable to transmit to naïve rabbits leading researchers to conclude they are not a likely enzootic amplifying host for CVV (Blackmore and Grimstad 2008). A study from Kokernot et al. 1969a found opossums, raccoons, and a 4-month-old pig did not generate a viremia following subcutaneous or intravenous inoculation (Kokernot et al. 1969a). However, an adult female pig did demonstrate a trace viremia (supported by mortality in inoculated mice in the absence of measurable viremia) and subsequently demonstrated neutralizing antibody. Goats and calves developed a trace viremia, apart from 1 calf with apparent low viremia on days 4 and 5 postinoculation. Both goats and calves seroconverted 1–2 months postinoculation. The authors concluded that these data, in conjunction with previous field seropositivity surveys, suggest that cattle may contribute to the enzootic transmission of CVV (Kokernot et al. 1969b, 1969a), though this has not been examined further. White-tailed deer developed a viremia of 3 log₁₀ PFU/ml following needle inoculation, which supports the theory that deer may serve as propagative hosts (Blackmore and Grimstad 1998). To date, no viremia data exist for horses, which have also demonstrated high rates of widespread seroprevalence.

Clinical Disease in Animals

While CVV has been known to circulate in most of the United States (Calisher et al. 1986), it was not until 1987 that it was linked to severe congenital malformations in sheep (Crandell et al. 1989). Between January and February 1987, CVV was identified as the etiological agent of an outbreak of congenital abnormalities in sheep in San Angelo, Texas, in an experimental station flock at Texas A&M. Abnormalities included abortions, weak lambs, stillbirths, mummified fetuses, and defects in the central nervous system (CNS) and musculoskeletal system (MSS) in 92 of the 360 lambs (or about 25.6%) born during this outbreak. Arthrogryposis hydranencephaly (AGH) occurred in roughly 69 of the 360 lambs (or about 19.2%) (Crandell et al. 1989, Chung et al. 1990). The total loss of lambs in the season was 25.6%, a significant increase from the previous lambing season, which saw only a 0.6% lamb loss from MSS and 4.8% total lamb loss, including stillborn births (Chung et al. 1990). A link to CVV was determined based on the detection of anti-CVV antibodies in ewes’ serum and pre-colostral serum from affected lambs. Anti-CVV antibodies were detected in all ewes with AGH lambs, while only 62% of ewes with normal lambs had anti-CVV antibodies (Chung et al. 1990). Isolation of CVV from a sentinel sheep in the same pasture later confirmed the presence of circulating CVV (Chung et al. 1990). At this time in Texas, an increase in CVV seroprevalence occurred from 5% in the spring of 1986 to 62% in the winter of 1987. It is thought that the low seroprevalence combined with increased rainfall in the region during this time led to the epizootic (Chung et al. 1990). Similar epizootics of CVV-causing fetal abnormalities in lambs were identified in Michigan, Nebraska, and Illinois during the same period (Crandell et al. 1989).

CVV causes a similar disease in ruminants as orthobunyavirids, Akabane virus (AKAV), and Schmallenberg virus (SBV). AKAV causes epizootics of AGH, abortions, and stillbirths in cattle, goats, and sheep in Japan, Australia, and Israel, while SBV causes similar epizootics in Belgium, Germany, France, Italy, Luxemburg, the Netherlands, Spain, and the United Kingdom (Kurogi et al. 1975, Della-Porta et al. 1976, Brenner et al. 2004, Gibbens 2012). Experimental models have been developed to study the teratogenicity of CVV in lambs. Previous studies have shown that ovine fetal CVV infections with CNS and MSS complications occur most frequently between gestational days 28 and 48, with no complications arising in animals inoculated after gestational day 48 (Edwards 1994). For AKAV, teratogenesis is most frequent when infections occur between gestational days 30 and 36 (Parsonson et al. 1977, Hashiguchi et al. 1979). To study CVV vaccine efficacy, 6-month-old male Rambouillet lambs were used to determine immunogenicity of a candidate live-attenuated CVV vaccine and were shown to produce a neutralizing antibody response after vaccination. Challenge experiments were not conducted as CVV does not cause clinical symptoms in adult ruminants, demonstrating that an ovine model to study vaccine efficacy in pregnant ewes would be useful (Ayers et al. 2023).

CVV targets CNS and MSS tissue in the infected fetus. Viral RNA and infectious virus have been detected in the brain, spinal cord, and skeletal muscle as early as 7 days postinfection of ovine fetuses from ewes infected 35 days postbreeding. Hydrocephalus, micromyelia, and muscular loss were observed in CVV-infected fetuses (Rodrigues Hoffmann et al. 2012). CVV viremia is short lived, and virus or viral antigen is not detected in fetal tissues after 14 days postinfection (Rodrigues Hoffmann et al. 2012). Therefore, detection of neutralizing antibody is needed to diagnose CVV in aborted fetuses or stillborn lambs (Chung et al. 1990, Edwards 1994). While the adaptive immune system begins to develop in the ovine fetus starting at 25 days postgestation when lymphocyte production begins in the thymus, the ovine fetus produces an antiviral innate immune response to CVV infection by upregulating production of Mx protein, an interferon GTPase known to inhibit growth of orthobunyavirids (Rodrigues Hoffmann et al. 2013). Between gestational days 45 and 50, lymphocytes begin to circulate in the bloodstream, and IgG- and IgM-positive cells can be in infected fetal tissues by gestational day 49, which is much earlier than in AKAV infections in lambs when IgM and IgG-positive cells are detected in tissues between 59 and 100 days postgestation (McClure et al. 1988, Rodrigues Hoffmann et al. 2013).

Few mouse models to study pathogenesis of CVV or vaccine efficacy in vivo have been evaluated. Outbred ICR mice were resistant to CVV infection and viral-induced teratogenicity in pregnant mice (Edwards et al. 1998). However, interferon receptor-deficient mice are susceptible to CVV infection via peripheral routes of inoculation and vertical transmission of the virus resulting in spontaneous abortions and congenital malformations (Lopez et al. 2021, Skinner et al. 2022). The lack of an effective interferon response in these mice leading to productive CVV infections is not surprising given that interferon has been shown to play an essential role in restricting growth of orthobunyavirids, including CVV and La Crosse virus (Hefti et al. 1999, Kochs et al. 2002, Rodrigues Hoffmann et al. 2013). This animal model may prove to be very useful in future CVV vaccination studies.

Clinical Disease in Humans

To date, only 7 human clinical cases of CVV infections have been documented, 3 of which were fatal (Table 2) (Sexton et al. 1997, Campbell et al. 2006, Nguyen et al. 2013, Wilson et al. 2017, Yang et al. 2018b, Baker et al. 2021, Al-Heeti et al. 2022). In most cases, CVV causes severe neurological symptoms, including meningitis or encephalitis, and in some cases, organ failure (Sexton et al. 1997, Wilson et al. 2017, Baker et al. 2021). Four of the documented CVV patients were immunocompromised either because of a preexisting condition or because of treatment with immunosuppressive drugs (Wilson et al. 2017, Yang et al. 2018b, Baker et al. 2021, Al-Heeti et al. 2022). Immunosuppressive drugs, like rituximab that causes B-cell depletion in the receiving patient, have led to several severe arboviral infections, including infections with West Nile virus, tick-borne encephalitis virus, eastern equine encephalitis virus, Jamestown Canyon virus, and Powassan virus (Kapadia et al. 2022). In these cases, the arboviral infection diagnosis was delayed due to atypical clinical presentation and no detectable antibody response secondary to the administration of immunosuppressive drugs in the patient. Serological diagnosis is heavily relied upon for detecting most arboviral infections as viremia may be transient and usually undetectable once the patient has presented with symptoms. In these cases, a positive diagnosis was made by virus isolation or nucleic acid detection.

Table 2.

Demographics, clinical manifestations and outcomes of human CVV infections

Reference	Year	Gender	Age (years)	Location	Potential exposure	Clinical symptoms	Underlying condition	Outcome
Sexton et al. 1997 PMID: 9023091	November 1995	Male	28	North Carolina, USA	Reported mosquito bites 2 wk before illness onset	Myalgia, fever, chills, headache, vomiting, confusion, maculopapular rash	none reported	Died in June 1996 of pulmonary complications with severe neurologic sequelae
Campbell et al. 2006 PMID: 16704854	October 2003	Male	41	Wisconsin, USA	none reported	headache, nausea, vomiting, fatigue	none reported	recovered; reported frequent headaches thereafter
Nguyen et al. 2013 PMID: 23515536	September 2011	Female	63	New York, USA	Reported outdoor activity mid-to-late August in Wyoming and New York	Fever, headache, neck stiffness, photophobia macular, nonpruritic lesion, petechial rash	Hypertension, hypothyroidism, meningioma, migraine headaches	Recovered; reported frequent headaches and memory loss thereafter
Wilson et al. 2017 PMID: 28628941	Spring, Summer 2013	Male	34	Traveled from Australia to North and South Carolina, Michigan, USA	Reported mosquito bites while in North and South Carolina	Fatigue, anorexia, lethargy, fever, confusion, drowsiness	Hypochondroplasia, X-linked agammaglobulinemia	Initial recovery; severe neurologic decline; died in 2017
Yang et al. 2018b PMID: 29282755	July 2016	Male	58	Ney York, USA	None reported	Confusion, irritability, memory loss, gait instability, fatigue, weight loss	TP53 wild-type chronic lymphocytic leukemia (CLL), lymphadenopathy and undergoing treatment with bendamustine and rituximab	Neurologic decline; died in August 2016
Baker et al. 2021 PMID: 33630998	September 2015	Male	60s	Missouri, USA	Reported mosquito bites 2 wk before illness onset	Fever, fatigue, fever, chills, rigors, cough, abdominal pain, anorexia, nausea, diarrhea, vomiting	Diabetes mellitus, coronary artery disease, thymectomy	Recovered
Al-Heeti et al. 2022 PMID: 35883256	Fall 2020	Female	60	Illinois, USA	Blood transfusion	Weakness, confusion, fatigue, weight loss, diarrhea, back pain, increased urinary frequency, and pain	Kidney transplant in fall 2020; undergoing treatment with alemtuzumab, rituximab, tacrolimus, and mycophenolate	Severe neurologic sequelae

Evidence of widespread infection in humans has been documented in multiple serosurveys conducted in North and Central America, and seroprevalence rates vary widely among the different geographic areas tested. One of the first studies of its kind was conducted between 1961 and 1963 with healthy residents living in areas of Maryland and Virginia on or near the Del Mar Va Peninsula. Overall, a 12% seroprevalence rate was determined using a qualitative neutralization test in which protective efficacy of the serum samples was measured in weanling mice challenged with a CVV mosquito isolate from the study site In this study, most residents positive for CVV-neutralizing antibody resided in Virginia (Buescher et al. 1970). In an investigation to determine the disease burden caused by various bunyaviruses on the Yucatán Peninsula, Mexico, researchers found an 18% orthobunyavirus-seropositivity rate among febrile patients measured by comparative plaque reduction neutralization test (PRNT) with CVV, Cholul, Kairi, Maguari, and Wyemomyia viruses. Of these patients positive for a previous orthobunyaviral infection, 4% were specifically seropositive for CVV, 3.4% were seropositive for Cholul virus, a reassortant virus of CVV and Potosi virus, and 26% were seropositive for an indistinguishable or undetermined orthobunyavirus (Blitvich et al. 2012b). Likewise, serologic studies of sera from patients with a recent suspected WNV infection from Manitoba and Saskatchewan, Canada in 2009 found seropositive rates from 5 to 16% measured by CVV-specific PRNT. As such, the public health agency of Canada named CVV a major emerging vector-borne zoonotic disease of public health importance (Kulkarni et al. 2015). A serosurvey of national parks and forest employees in the eastern and western United States found a 3% seroprevalence rate of anti-CVV neutralizing antibodies measured by CVV-specific PRNT (Kosoy et al. 2016). These 2 studies only included CVV-specific PRNT data, and the possibility a related orthobunyavirus circulating in these regions may be responsible for the detection of neutralizing antibodies in these samples cannot be discounted. Nevertheless, these data suggest that most CVV infections go undetected either because of a lack of severe disease manifestations (most cases may be asymptomatic or cause mild disease in humans) or because of a lack of knowledge by physicians or available diagnostic testing. The lack of understanding of the geographic distribution and the true burden of CVV disease in humans requires further study.

The teratogenicity of CVV in humans is another understudied area of public health interest. Only 2 studies to date have investigated the link between neural tube defects of infants born to mothers with anti-CVV neutralizing antibodies. A retrospective study using serum samples from mothers of infants with micro- or macrocephaly collected in the 1960s discovered a significant correlation between the presence of anti-CVV neutralizing antibodies in the mothers and infants with macrocephaly. In contrast, no correlation was found between the presence of anti-CVV antibodies in the mothers and infants with microcephaly (Calisher and Sever 1995). Interestingly, in 8 of the patients with paired samples from early and late pregnancy, 2 patients that had given birth to infants with macrocephaly had a 4-fold rise in neutralizing antibody, indicating a recent CVV infection (Calisher and Sever 1995). In 1990 and 1991, following an increase in neural tube defects in south Texas, a serologic study of 74 patients who had given birth to infants with anencephaly found none of the patients had anti-CVV neutralizing antibodies suggesting CVV was not the causative agent in these cases (Edwards and Hendricks 1997). This study followed the initial outbreak of congenital malformations and fetal death in sheep in San Angelo, TX in 1987 (Edwards 1994). While this study found no link between CVV and neural tube defects, CVV was not known to be circulating in the area and therefore does not prove that CVV is not teratogenic in humans. Relying on neutralization tests that can only identify a past CVV infection may not be the optimal test to use for studying potential teratogenicity of CVV in humans, and serologic assays that detect anti-CVV IgM which is produced early in infection may prove to be more useful in examining the teratogenicity of the virus in future studies.

Molecular Diagnosis

Presently, there are no FDA-approved or commercially available tests for the molecular diagnosis of CVV in humans. Given the rarity of human CVV infection, clinical or surveillance testing is not widely available in the public health systems and is often restricted to high-capacity state public health labs or the US Centers for Disease Control and Prevention. Diagnosis of human CVV infections has relied on classical reference testing techniques, laboratory-developed tests (LDT), and more recently, metagenomic or next-generation sequencing.

The first 2 human diagnoses of CVV in 1995 (Sexton et al. 1997) and 2003 (Campbell et al. 2006) were accomplished through classical reference testing techniques. Virus isolated in cell culture from patient samples identified bunyavirus virions under electron microscopy. Using confirmatory reverse-transcription polymerase chain reaction (RT-PCR) primers targeting the complementary terminal ends of the S segment (Dunn et al. 1994), the full-length S segment sequence of CVV was sequenced from cDNA clones and identified as the pathogen.

RT-PCR techniques using virus-specific primers were among the first LDTs described for molecular detection of California and Bunyamwera serogroup viruses, including CVV in mosquito pool homogenates (Kuno et al. 1996). This technique relies on visualization of the amplicon products on a gel to determine positive amplification. This size discrimination ability of amplicons was further applied for broad detection of several arboviruses in a given sample (Huang et al. 2001). Similar methods of targeting areas of broad genomic consensus have resulted in LDTs capable of detecting many orthobunyaviruses, including CVV, and represent a reference testing approach for detecting emerging and orphan viruses (Lambert and Lanciotti 2009). Limitations of RT-PCR include the need for amplicon sequencing to determine the specific virus adding additional time and testing required for definitive results.

Real-time RT-PCR techniques using fluorogenic probes represent an improved method over traditional RT-PCR for the specific detection of CVV. Like traditional RT-PCR, real-time RT-PCR uses gene-targeted primers but includes an additional virus-specific probe. During cycling amplification, the fluorophore is dissociated from the probe, and a logarithmic increase in fluorescence intensity is measured and compared to baseline fluorescence. A multiplexed real-time RT-PCR LDT was developed to detect encephalitic bunyaviruses (CVV and California serogroup viruses) in clinical samples and mosquito pools (Wang et al. 2009). This LDT has been used in clinical settings for CVV diagnosis (Nguyen et al. 2013) as well as in vector competence and surveillance studies (Ayers et al. 2018, 2019). Following the recognition of lineage II CVV in the United States, a second LDT was described to better detect viruses of both lineages in mosquito surveillance and clinical diagnostics (Dieme et al. 2022a, 2022b). Limitations of all targeted RT-PCR approaches include the emergence of new or replacing lineages or genetic drift of the genome from error-prone RNA replication may reduce the sensitivity or specificity of the primers over time.

Metagenomic next-generation sequencing (mNGS) represents recent advances in molecular detection. mNGS has advantages over primer-specific approaches as a non-targeted, unbiased method to detect minor virus populations in a sample. Recently, mNGS has detected CVV RNA in clinical samples and determined CVV as the causative agent in 4 additional infections, including chronic infection, atypical disease presentation, and blood transfusion-associated infection (Wilson et al. 2017, Yang et al. 2018b, Baker et al. 2021, Al-Heeti et al. 2022). The recognition of CVV as the pathogen in these case investigations demonstrates the utility of mNGS as a non-targeted approach to detect emerging and rare infections, especially in atypical cases. Limitations of mNGS include cost, sequencing bias of host genome, inconsistent method standards, and the need for data analysis pipelines. Although mNGS is a revolutionary tool, presently, this method cannot replace more classical RT-PCR techniques for routine testing.

Molecular diagnosis of CVV in agricultural settings is not well documented in the literature (Waddell et al. 2019). The LDTs and mNGS approaches described above often apply to CVV diagnosis in agricultural settings, though validation has yet to be described. Specialized testing labs such as the USDA or University veterinary hospitals may offer CVV molecular testing for diagnosis in livestock.

Serologic Diagnosis

Currently, no commercial assays are available to detect anti-CVV antibodies, and testing may only be performed in biosafety level 2 (BSL-2) laboratories. Serologic diagnostic tests are essential for diagnosing human arboviral infections, particularly those that cause neurological symptoms because viremia is transient and usually undetectable when patients present with disease symptoms. Until recently, the only serologic diagnostic test available for detection of CVV infections has been the plaque reduction neutralization test (PRNT) which requires the use of live virus. In this assay, recent and past infections are indistinguishable in specimens taken at a single timepoint because only neutralizing antibody, which could be IgM elicited during an acute infection or IgG from a previous infection, is detected. If paired specimens are available a rise in antibody titers between acute and convalescent samples can be used to indicate a more recent infection. PRNT, mNGS, and virus isolation were used in a human clinical investigation to determine the source of CVV infection in a human kidney transplant case (Al-Heeti et al. 2022). PRNT has also been useful in investigations of outbreaks of congenital abnormalities in sheep (Crandell et al. 1989, Edwards et al. 1989, Chung et al. 1991).

Anti-viral IgM is produced early in the course of infection, making it an ideal diagnostic analyte for detection of an acute and recent convalescent CVV infection in humans and livestock. The recent development of murine anti-CVV monoclonal antibodies (MAbs) has led to the development of an IgM antibody capture enzyme-linked immunosorbent assay (MAC-ELISA) that is currently undergoing validation at the US Centers for Disease Control’s Arboviral Diseases Branch (ADB) diagnostic laboratory (Skinner et al. 2022). ADB’s diagnostic laboratory uses the MAC-ELISA platform to detect IgM to several medically important arboviruses, including dengue viruses, West Nile virus, St. Louis encephalitis virus, La Crosse virus, Zika virus, and Eastern equine encephalitis virus from acute or convalescent human specimens (Martin et al. 2000, 2002).

One of the limiting factors in performing the CVV MAC-ELISA is the need for human positive control to include in the test because of a lack of human CVV infections. To overcome this, the development of a recombinant murine-human IgM antibody expressed in a stable HEK293 cell line using MAb CVV17 previously described has been undertaken (Skinner et al. 2022). This recombinant antibody expressing the variable regions of the murine MAb CVV17 and the human constant regions of IgM can replace human positive control sera making the distribution and implementation of this assay more accessible (unpublished data).

Need for Further Study

While CVV is potentially an important livestock pathogen that can cause severe disease in humans, more research needs to be undertaken for preventative measures or understanding the economic impact of the disease on agriculture. Due to the inconsistent yearly patterns, lack of longitudinal data, and limited diagnostic reagents, there are considerable gaps in knowledge regarding CVV ecology. As human cases are likely underrecognized, it will be important to understand the risk that CVV poses moving forward (Kapadia et al. 2022). More ecological longitudinal studies like those done in New York and Connecticut (Andreadis et al. 2014, Dieme et al. 2022a, 2022b) will help identify regionally important vectors and ideally pinpoint peak transmission periods for targeted mosquito surveillance and early detection. More information regarding the transovarial transmission rates of CVV in a broader range of implicated vectors is needed to identify vectors that have the potential to initiate seasonal transmission. Additionally, experimental studies in animals with historically high seropositivity rates could highlight alternative amplifying cycles and targets for control. For instance, if cattle and horses contribute to circulation, a veterinary vaccine may be a feasible way to interrupt some aspect of enzootic circulation.

Recent advancements in animal models and vaccines such as the immunocompromised mouse model capable of eliciting in-utero disease similar to natural disease in sheep are advantageous for studying pathogenesis or vaccines because of its accessibility (Lopez et al. 2021). Recent research on vaccine development has described the creation of a live-attenuated CVV vaccine candidate that lacks both nonstructural proteins (NSs and NSm) (Ayers et al. 2022). Presently, a lack of transmissibility in mosquitoes by the deletion mutant virus and immunogenicity was demonstrated (Ayers et al. 2023); however, protective efficacy has not been investigated. To understand the benefit of vaccination, the economic impact of CVV disease must be understood, but no such studies have been conducted (Waddell et al. 2019, Ayers et al. 2022). While CVV outbreaks at farms cause large agroeconomic losses, the true direct and indirect costs have not been evaluated. One study investigating the impact of animal health diseases on the economy highlighted the importance of both the direct cost of the disease to ranchers and the indirect cost to the overall economy, including rural business and potential tourism (Barratt et al. 2019).

Finally, more research on CVV is needed to determine the burden of disease in humans and animals. Serologic tests that do not rely on biocontainment will be useful tools to determine the seroprevalence of CVV in any given geographic area and population. Incorporating serologic assays that detect IgM will help to rapidly identify epizootic outbreaks, and less severe cases of human disease. Ultimately, understanding the prevalence of CVV will aid in our ability to determine whether the virus poses a teratogenic risk to humans.

Acknowledgment

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of CDC.