Emerging Trends in the Epidemiology of West Nile and Usutu Virus Infections in Southern Europe

The epidemiology of West Nile (WNV) and Usutu virus (USUV) has changed dramatically over the past two decades. Since 1999, there have been regular reports of WNV outbreaks and the virus has expanded its area of circulation in many Southern European countries. After emerging in Italy in 1996, USUV has spread to other countries causing mortality in several bird species. In 2009, USUV seroconversion in horses was reported in Italy. Co-circulation of both viruses was detected in humans, horses and birds. The main vector of WNV and USUV in Europe is Culex pipiens, however, both viruses were found in native Culex mosquito species (Cx. modestus, Cx. perexiguus). Experimental competence to transmit the WNV was also proven for native and invasive mosquitoes of Aedes and Culex genera (Ae. albopictus, Ae. detritus, Cx. torrentium). Recently, Ae. albopictus and Ae. japonicus naturally-infected with USUV were reported. While neuroinvasive human WNV infections are well-documented, USUV infections are sporadically detected. However, there is increasing evidence of a role of USUV in human disease. Seroepidemiological studies showed that USUV circulation is more common than WNV in some endemic regions. Recent data showed that WNV strains detected in humans, horses, birds, and mosquitoes mainly belong to lineage 2. In addition to European USUV lineages, some reports indicate the presence of African USUV lineages as well. The trends in WNV/USUV range and vector expansion are likely to continue in future years. This mini-review provides an update on the epidemiology of WNV and USUV infections in Southern Europe within a multidisciplinary “One Health” context.


INTRODUCTION
West Nile virus (WNV) and Usutu virus (USUV) are mosquitoborne flaviviruses characterized by similar clinical manifestations and overlapping geographic distribution, host and vector species. The epidemiology of WNV and USUV has changed dramatically over the past two decades. Since 1996, an increasing number of WNV outbreaks in humans and horses were detected, and the area of its circulation expanded in many Southern European countries (1). After USUV emergence in Austria in 2001, it subsequently spread to neighboring countries causing mortality in several wild bird species, mainly Eurasian blackbirds (2). However, a retrospective analysis of archived bird tissue samples from Italy (Tuscany region) in 1996 identified USUV, indicating a much earlier introduction of this virus into Europe (3).
Since the early 2000s, WNV circulation has been continuously monitored in some European countries with varying number of human and horse cases. USUV infections are reported in birds, while human clinical cases are rarely detected (4)(5)(6). However, there is increasing evidence of a role of USUV in human disease. Seroepidemiological studies showed that USUV circulation in humans is more common than WNV in some endemic regions where both viruses circulate (7).
Until the introduction of WNV lineage 2 in 2004 (8), WNV lineage 1 was identified as the cause of human outbreaks in Europe. In the following years, lineage 2 dispersed to the eastern part of Austria and to Southern European countries (9)(10)(11). Recent data showed that strains detected in humans, horses, birds, and mosquitoes mainly belong to WNV lineage 2 (6,12,13). Although the majority of USUV strains belong to European USUV lineages, some reports indicate the presence of African USUV lineages as well (5).
In 2018, WNV infections in Europe increased dramatically compared to previous transmission seasons. A total of 2,083 human cases and 285 outbreaks among equids were reported with the largest number detected in Italy, Serbia, and Greece (14). USUV was detected in asymptomatic blood donors, birds, and mosquitoes in Italy (15,16) and in three patients with neuroinvasive disease in Croatia (6).

ALBANIA
Data on WNV infections in Albania are scarce. To date, only two human cases were reported in 2011 (14). A serological study conducted among horses from 12 districts in Albania showed WNV seropositivity of 22.2%, with significant regional differences (7.7-66.7%). The highest seroprevalence rates were reported in districts near the Mediterranean Sea, whereas low or negative seroprevalence was detected in districts located further inland (41). Data on USUV infection are not available.

BOSNIA AND HERZEGOVINA
There is only one report, in 2013, of two cases of WNV neuroinvasive disease (WNND) in Bosnia and Herzegovina (detected in the Tuzla and Kladanj region). Three out of nine patients screened for WNV met clinical criteria for WNND, of which two had high serum IgM titres. RT-PCR for WNV RNA was negative. Since the neutralization test was not performed, cases were classified according to the European Union case definition as probable WNV (42). So far, there are no data on USUV infections in Bosnia and Herzegovina.

BULGARIA
In 2015, the first confirmed human case with fatal outcome reported in Bulgaria was caused by a Central/Southern-European lineage 2 WNV (43). Human cases were continuously detected in the following seasons (12). A nationwide study conducted among residents of all 28 districts in Bulgaria revealed a 1.5% WNV seroprevalence with the highest seropositivity (up to 10%) detected in districts near the Danube River (44). USUV infections were not documented in Bulgaria.

CROATIA
The first outbreak of human WNND in Croatia was detected in 2012 in eastern counties, thereafter outbreaks (2013, 2017) and sporadic cases (2014-2016) were continuously notified in continental Croatia (4,45). The largest outbreak was recorded in 2018 with 54 cases of WNND and 7 cases of WNV fever in 10 of the 21 counties (6). During the 2013 WNV outbreak, the first three cases of neuroinvasive USUV disease were detected in Zagreb and surrounding areas (4,46). Three additional USUV cases were confirmed in the 2018 outbreak, of which one was fatal (6). A serological study conducted from 2010 to 2011 showed WNV antibodies in 3.43% horses and 0.11% cattle with the highest seropositivity in eastern counties bordering Hungary and Serbia (47). Seropositivity and acute asymptomatic WNV infections in horses were detected continuously in subsequent years. Seroprevalence varied greatly by year and region (0-26%) with the highest seropositivity in counties with documented human cases (4,48). Although passive monitoring of WNV in birds was established in Croatia in 2012, WNV infections were not detected until the summer of 2018, when WNV infection was confirmed serologically in one buzzard presenting with neurological symptoms, and WNV RNA was detected in two  dead goshawks from Northwest Croatia. Two USUV seropositive horses were documented in 2011 in Northwest Croatia (49). In 2018, USUV RNA was detected in one dead blackbird in Zagreb County (6). To date, none of the mosquito pools tested were WNV RNA positive. However, USUV-positive mosquito pools were found in 2016 (Aedes albopictus), 2017 (Culex pipiens), and 2018 (Cx. pipiens) in Northwestern counties (6,50). Sequenced strains from humans, wild birds, and mosquitoes (2018) confirmed circulation of WNV lineage 2 and USUV Europe 2 lineage (6).

GREECE
No WNV clinical cases in humans or horses had been reported in Greece prior to the large WNV outbreak in 2010 when 262 human cases with extremely high fatality rates (17%) were reported mostly in Central Macedonia (51). The outbreaks continued in 2011 and 2012 in areas that had not been affected before (52,53). Thereafter, outbreaks occurred in humans every year except 2015 and 2016 (14,54 (86,87). In 2014, seroconversion was detected in 2.57% horses from the whole country. In addition, in 2015, 2017, and 2018, acute WNV infections were detected in 0.53, 0.41, and 1.47% of horses tested, respectively (88)(89)(90). WNV antibodies were detected in 7.6% wild birds sampled from 2011 to 2012 in Vojvodina, of which 133 birds were found dead. Virus presence was confirmed in 9.87% of tissue samples and in the blood sample of one bird (91). From 2014 onwards, WNV was detected in wild birds, with the highest percentage of positive samples being observed in 2018 when WNV was detected in 11.61% of tissues and 6.56% of pharyngeal swabs (88)(89)(90). All WNV isolates belonged to WNV lineage 2 (88,91). Additionally, WNV antibodies were identified in 15.4% farm pigs, 17.6% wild boars, and 18.7% roe deer sampled from 2011 to 2012 (Vojvodina)

CONCLUSIONS AND FUTURE PERSPECTIVES
An integrated "One Health" surveillance of mosquitoes and birds in several countries has proven to be useful for early detection of WNV/USUV circulation and identification of enzootic areas (48,61,116). Virus detection in mosquitoes and birds preceded human and horse cases (48,62). In addition, a high seroprevalence in sentinel animals was detected in areas with documented human cases (4,22,48,61). There are still many challenges in the epidemiology of WNV/USUV. In addition to birds, which are well-known reservoirs, WNV and USUV antibodies were found in different animal species. WNV and USUV neutralizing antibodies were documented in red deer in Spain (117). WNV seropositive dogs were detected in Italy, Spain, and Corsica Island (118), while USUV antibodies were found in hunting dogs in Italy (119). Moreover, WNV and USUV antibodies were found in gray squirrels in Italy, broadening the host range for these viruses (120). WNV neutralizing antibodies were also found in two wild rodents (Apodemus flavicollis) captured in forested areas of Italy (121). However, the short-term and low-level viremia makes it unlikely that these animal species play a role in the WNV transmission cycle. Culex pipiens mosquitoes appear to be a major vector for WNV and USUV transmission in Europe, but Cx. modestus and Cx. perexiguus play an important role in marshlands of some southern countries (115). However, detection of WNV and USUV in different fieldcollected native (Ae. vexans) and invasive (Ae. albopictus, Ae. japonicus) mosquito species indicates their possible role in promoting the overwintering of these viruses (75,83,(122)(123)(124)(125). Additionally, a recently published study identified Cx. torrentium as a highly competent vector for WNV in Central and Northern Europe (126).
While several licensed veterinary WNV vaccines are currently available, there is no WNV or USUV vaccine for humans (127,128). Since data from Europe indicate that both WNV and USUV appear to be expanding their geographical ranges and the trends indicate that this spread is likely to continue in future years, the development of an effective vaccine is urgently needed to protect at-risk populations from neurological complications.

AUTHOR CONTRIBUTIONS
TV-C and TP made contributions to conception and design of the study, involved in data collection, and drafting the manuscript. VSa, LB, ITa, IH-C, MB, AK, AM, VSt, PD-K, LR, FM, and EL were involved in data collection and drafting the manuscript. ITo, DP, and GS revised the manuscript critically. All authors read and approved the final manuscript.