For group 1, serum and CSF samples from 103 patients with encephalitis (at that time of unknown etiology) were received and tested during the study. Samples were from 60 male and 43 female patients; age range was 1–89 years (median age 48 years). For group 2, bornavirus serologic analysis was conducted for serum samples from 121 patients who had no clinical history of encephalitis but for whom bornavirus serologic analysis was requested. Samples were from 55 male and 66 female patients; age range was 4–84 years (median age 45 years).

Figure 2

Figure 2. Germany showing locations of residences of human case-patients who had encephalitis and other conditions and were tested for bornavirus etiology, 2018–2020. Among 103 encephalitis cases with unknown etiology, 4 bornavirus...

Figure 3

Figure 3. Reactivity of samples from BoDV-1- and VSBV-1-infected patients with homologous and heterologous bornavirus P antigens, Germany. Bornavirus indirect immunofluorescence antibody test–positive serum and cerebrospinal fluid samples were tested by using...

For group 1, a total of 4 (3.9%) confirmed bornavirus encephalitis case-patients were detected: 1 VSBV-1 case-patient in northern Germany (Schleswig-Holstein) (case-patient 1) and 3 BoDV-1 case-patients in Bavaria in southern Germany (case-patients 2–4; Figure 2). None of these case-patients had a travel history outside Germany in the 5 months before symptom onset. Initial testing by IFAT resulted in detection of bornavirus-reactive antibodies in serum and CSF samples for all 4 case-patients (Table 2). Samples of case-patients 1–3 also showed positive results for a bornavirus P line blot assay. Serum and CSF samples from case-patient 1 showed higher signals for VSBV-1 P than BoDV-1 P, and serum samples from case-patients 2 and 3 showed higher signals for BoDV-1 P (Table 2). The discriminatory potential of this test was confirmed by analysis of reference serum samples from several laboratory-confirmed BoDV-1 and VSBV-1 encephalitis cases (Figure 3). qRT-PCR and full-genome sequencing confirmed the bornavirus infection for all 4 case-patients.

Case-Patient 2: Acute BoDV-1 Encephalitis

Figure 4

Figure 4. Phylogenetic analysis of BoDV-1 nucleotide sequences from virus-endemic areas, Germany. Shown are partial bornavirus sequences (nucleoprotein gene to phosphoprotein gene, 1,824 nt, representing genome positions 54–1877 of BoDV-1 reference genome...

Case-patient 2 was a 55-year-old woman, a part-time cleaner from Bavaria who was given a diagnosis of BoDV-1 infection in February 2019. The patient had encephalitis, fever, headache, and coma develop in mid-January 2019. IFAT titer was 640 for serum and 80 for CSF. Line blot results for antibodies against BoDV-1 P were 30 units for serum and 4 units for CSF (Table 2). A qRT-PCR result for BoDV-1 in CSF was weakly positive (Cq 35.3). Case-patient 2 died 1 day after diagnosis and 3 weeks after onset of disease. Postmortem virus isolation and sequencing from brain tissue were successful (GenBank accession no. LR722643; Figure 4) (12). This case-patient was included in a recent case series from Bavaria as case-patient P8 by Niller et al. (12), and histopathologic results were described in detail as case-patient 6 by Liesche et al. (10).

The epidemiology of human bornavirus encephalitis is still largely unknown. In our prospective study, we identified 4 bornavirus case-patients (3.9%) in a group of 103 case-patients who had cryptic encephalitis. The VSBV-1 case-patient had a disease course during 2007–2019, whereas the BoDV-1 case-patients died from acute infection during 2019 and 2020. Regarding incidence of infection, bornavirus encephalitis cases caused by VSBV-1 and BoDV-1 appear to be rare. Cases were only found among the group of encephalitis patients and not among patients with unspecified symptoms for whom a bornavirus serologic analysis was nonetheless requested. This finding supports previous studies that scientific evidence for a possible bornavirus etiology for other clinical entities apart from encephalitis continues to be lacking (15,17,18).

The cases detected in our study confirm previously identified risk factors: acquiring VSBV-1 encephalitis through contact with exotic squirrel species and association of BoDV-1 encephalitis with residence in mainly rural environments in areas endemic for animal BD and thus BoDV-1. The incidence of human VSBV-1 continues to be restricted to the small group of zoo animal caretakers and squirrel owners exposed to infected exotic squirrels. The denominator of this group is unknown, but small. Known confirmed human case-patients include 3 private breeders (1) and 2 zoo animal caretakers (case-patient 1 reported here; case reported in 11). All of these patients died, although the patient identified in this study survived for 12 years after the acute phase. In another study (6), a squirrel breeder had a serologically positive result after recovery from a transient neurologic disease. This case was classified as a probable case because diagnostic material suitable for confirmation of the infection by direct virus detection (Table 1) was not available. Two additional breeders identified in the same study had died of unclear encephalitis, but no archived diagnostic materials were available. These cases were categorized as possible cases (Table 1).

The incidence of diagnosed BoDV-1 encephalitis cases is currently ≈2 cases/year in Germany; there is a strong restriction to known areas to which BoDV-1 is endemic (12). In these regions, BoDV-1 might be a major cause of previously cryptic encephalitis: In a recent study of archived brain tissue material from patients who had fatal encephalitis without a known cause from 1 center in Bavaria in an area to which animal BD is endemic, 7 (78%) of 9 had a BoDV-1 infection (12). At the same time, a seroprevalence study performed in BoDV-1–endemic areas found only 1 seropositive person among a presumed risk group of 736 veterinarians (0.14%; clinical and diagnostic follow-up was not possible) and none among 373 healthy blood donors (0%; 15). Thus, these data suggest that BoDV-1 infection and BoDV-1–induced encephalitis are rare, even in virus-endemic areas, but have a high case-fatality rate.

Bornavirus encephalitis cases caused by VSBV-1 and BoDV-1 are clinically similar, severe, and in nearly all reported cases fatal. The course of BoDV-1 encephalitis appears to be more rapid (1–3,10–13). All but 1 case-patient who had confirmed bornavirus encephalitis died; only in a transplant-associated BoDV-1 encephalitis cluster did 1 patient survive but had neurologic sequelae (2). In our study, all 4 patients died of the disease; however, a patient who had severe VSBV-1 encephalitis/encephalopathy had a 12-year chronic course of disease. More thus far undetected chronic bornavirus infections in persons who have neurologic deficits after encephalitis or chronic encephalopathy might be present and should undergo diagnostic testing for a bornavirus etiology. Clinical awareness, particularly for severe encephalitis cases in areas to which BoDV-1 is endemic or after contact with exotic squirrel species, should lead to early testing for a possible bornavirus etiology while the person is still alive. Bornavirus infection needs to become a routine target for differential diagnostics. An early diagnosis would be a prerequisite for antiviral chemotherapy.

Following graded case definitions for VSBV-1, BoDV-1, and unspecified bornavirus encephalitis/encephalopathy (Table 1), we used a serologic testing scheme (Figure 1) for rapid initial intra vitam diagnosis of bornavirus encephalitis. Positive serologic results were subsequently confirmed by direct pathogen detection for all 4 bornavirus encephalitis cases. In this study, as well as in previous studies, higher and earlier detectable antibody titers were usually observed in serum samples than in CSF samples (2,12,13). The time of seroconversion during human bornavirus encephalitis is variable; some patients are already seropositive at the time of hospitalization, and others show development of detectable antibodies only shortly before death (2,12). Thus, serologic follow-up testing (likely within days) should be performed in encephalitis case-patients who have epidemiologic risk factors but initially negative results because such patients might be infected with zoonotic bornaviruses but might not yet have seroconverted.

IFAT appears to be more sensitive than the line blot assay used in this study and previous studies (12). In this study, line blot results were below the cutoff value for case-patient 4 in serum and CSF samples. This finding emphasizes the need for serologic follow-up testing for some case-patients, with the expectation that the screening test result will be confirmed by positive single-antigen assay results in follow-up samples that have increased antibody titers. Without the molecular confirmation, case-patient 4 would have been classified as having a possible case, stressing the need for molecular testing. In addition to sensitivity, specificity is a crucial issue of serologic testing for bornavirus infections, emphasizing the need for careful evaluation of IFAT results with a specific granular intranuclear pattern observed only in the bornavirus-infected cells (2,11,15,16).

Orthobornaviruses show considerable cross-reactivity among each other (16). Furthermore, antibodies against individual antigenic epitopes might be detected despite lack of any known previous bornavirus contact, leading to false-positive results (19). Establishing and optimizing assays such as Western blot, line blot, or ELISA, for detection of antibodies against individual bornavirus antigens are needed to enable subsequent confirmation of positive IFAT results and help to partly overcome the shortcomings of bornavirus serologic analysis. Furthermore, comparative testing by using antigens from different orthobornaviruses by IFAT (16) or line blot (as shown in this study) can enable discriminatory prediction of VSBV-1, BoDV-1, or other orthobornavirus infections. In the line blot used in this study, serum samples from VSBV-1–infected and BoDV-1–infected patients reliably showed higher arbitrary antibody units for the respective homologous P antigen than for heterologous antigen (Table 2). This provisional discrimination might be useful for prognosis because BoDV-1 infections appear to be more rapidly fatal than VSBV-1 infections. Despite these measures, incidental, singular, bornavirus-reactive antibodies might remain indistinguishable from truly positive results.

Because of these limitations of serologic testing, direct pathogen detection is mandatory to confirm of the initial serologic diagnosis and discrimination of VSBV-1 and BoDV-1 infections, as indicated in the graded case definitions (Table 1). However, intra vitam direct detection of the virus is severely hampered by the strong cell-associated and neurotropic nature of the virus and its almost exclusive restriction to the central nervous system in dead-end hosts, such as humans (2,11,12). The virus is not detectable in blood, and viral RNA detection by qRT-PCR in CSF samples often shows negative or only weakly positive results; the negative predictive value appears to be low (12). Bornavirus RNA detection in biopsy specimens from affected brain areas seems to be the most sensitive diagnostic method, but the difficulties with these procedures are high and unlikely to be met early in the disease course. Postmortem, unequivocal confirmation of infection can be made by using qRT-PCR, in situ hybridization, and immunohistochemical analysis of brain tissue (2,3,9–13). Sequencing the viral genome and subsequent phylogeographic analysis might provide information on the regional source of infection (2,12).

The first limitation of this study is that, although it was a large screening program of encephalitis cases for bornavirus infections, a complete or representative sampling was not performed. Thus, the prevalence of human bornavirus encephalitis remains to be investigated. Second, only a few bornavirus cases were found, which was partly caused by the overall low incidence of bornavirus encephalitis in humans. This limitation also emphasizes the need for increased awareness for this disease. Detection of future cases will provide further opportunity to approve the herein proposed case definition criteria. Third, only encephalitis cases from Germany were investigated in this study. However, the virus-endemic region for BoDV-1 also includes neighboring Austria, Liechtenstein, and Switzerland. Undetected human BoDV-1 cases are also expected in these countries.

In conclusion, human bornavirus encephalitis cases remain rare in the general population in Germany, even for BoDV-1 infections in areas to which animal BD is endemic. Human bornavirus encephalitis is often fatal. Chronic cases occur at least for VSBV-1 infections, but might be exceptional. There is no evidence that zoonotic bornaviruses might cause diseases other than encephalitis (e.g., encephalomyelitis, encephalomyeloradiculits) in humans. Antibody testing in serum samples is sensitive but requires confirmation by direct detection of virus. The proposed testing scheme and case definitions proved useful. All patients who have encephalitis (especially a severe course) from virus-endemic areas or after contact with exotic squirrels should be tested for a bornavirus infection, ideally early in the disease course by antibody testing in serum samples (21–23).