Multi-laboratory evaluation of ReaScan TBE IgM rapid test, 2016 to 2017

Background Tick-borne encephalitis (TBE) is a potentially severe neurological disease caused by TBE virus (TBEV). In Europe and Asia, TBEV infection has become a growing public health concern and requires fast and specific detection. Aim In this observational study, we evaluated a rapid TBE IgM test, ReaScan TBE, for usage in a clinical laboratory setting. Methods Patient sera found negative or positive for TBEV by serological and/or molecular methods in diagnostic laboratories of five European countries endemic for TBEV (Estonia, Finland, Slovenia, the Netherlands and Sweden) were used to assess the sensitivity and specificity of the test. The patients’ diagnoses were based on other commercial or quality assured in-house assays, i.e. each laboratory’s conventional routine methods. For specificity analysis, serum samples from patients with infections known to cause problems in serology were employed. These samples tested positive for e.g. Epstein–Barr virus, cytomegalovirus and Anaplasma phagocytophilum, or for flaviviruses other than TBEV, i.e. dengue, Japanese encephalitis, West Nile and Zika viruses. Samples from individuals vaccinated against flaviviruses other than TBEV were also included. Altogether, 172 serum samples from patients with acute TBE and 306 TBE IgM negative samples were analysed. Results Compared with each laboratory’s conventional methods, the tested assay had similar sensitivity and specificity (99.4% and 97.7%, respectively). Samples containing potentially interfering antibodies did not cause specificity problems. Conclusion Regarding diagnosis of acute TBEV infections, ReaScan TBE offers rapid and convenient complementary IgM detection. If used as a stand-alone, it can provide preliminary results in a laboratory or point of care setting.


Introduction
Tick-borne encephalitis virus (TBEV) is the most important tick-transmitted virus causing human disease in Europe and Asia [1][2][3][4]. TBEV belongs to the genus Flavivirus, within the Flaviviridae family, and can be divided into three distinct subtypes: the European (TBEV-Eur, formerly known as Central European encephalitis virus), the Siberian (TBEV-Sib, formerly known as Siberian encephalitis virus), and the Far Eastern (TBEV-FE, formerly known as Russian Spring Summer encephalitis virus) subtypes [3]. Recently, two new subtypes of TBEV (Himalayan and Baikalian) have been characterised [5,6]. Several studies suggest that the case fatality rate for TBE caused by TBEV-Eur is 0-4% [1,7] by TBEV-Sib 2-3% [1,7] and by TBEV-FE 6-40% [1,3,7,8]. However, according to Ruzek et al., 2019 [9] the overall TBE mortality rate in Russia is approximately 2% (i.e. TBEV-Sib and TBEV-FE infections). Thus, data on fatality rates of TBEV are not comprehensive since, aside from the infecting subtype, other factors (such as healthcare system efficiency, population genetics or living conditions) may come into play. TBEV is maintained in ticks and in their wild vertebrate hosts in forested natural foci [10]. The main reservoir hosts are found among small mammals (e.g. rodents, insectivores), while larger animals (e.g. deer), despite being important feeding hosts for ticks, do not seem to play any considerable role in the maintenance of the virus within its foci [2,11]. The epidemiology of tick-borne encephalitis (TBE) is closely associated with the geographical distribution and ecology of tick vector species, and the periods of their feeding activity. Human infections usually occur as results of tick bites, but can also in rare occasions be acquired via consumption of unpasteurised milk and milk products from infected animals [2,12,13].
After a bite by an infected tick, a proportion of individuals will remain asymptomatic but 2-30% will develop an initial non-specific febrile illness lasting a few days, followed by an asymptomatic interval of 1-2 weeks [3,14]. Approximately 30% of the patients who initially showed clinical symptoms will further develop neurological symptoms (mild meningitis to severe encephalitis) during a second phase of the disease [14,15]. According to a 10-year follow-up survey in Germany, 80% of patients with primary myelitic manifestations suffered long-term sequelae [16]. A recent study on post-encephalitic syndrome (PES) after TBE in Slovenia revealed that the frequency and severity of PES diminished over time following acute illness [17]. After 12 months, PES frequency stabilised while severity continued to decline. Unfavourable outcomes at 12 months and at the final hospital visit (at 7 years post-acute illness) were strongly associated with the presence of PES at previous time points [17].
The clinical symptoms of TBE are unspecific and the diagnosis has to be verified in the laboratory. The criteria for confirmed TBE include central nervous system (CNS) inflammation symptoms (e.g. meningitis, meningo-encephalitis, encephalomyelitis, encephaloradiculitis) and at least one of the laboratory criteria, which are either detection of the virus or its nucleic acid in a clinical specimen or detection of specific IgM and IgG antibodies in serum, seroconversion, or specific IgM in cerebrospinal fluid (CSF) [18].
The laboratory diagnosis of TBE is usually straightforward; in almost all cases, TBEV-specific IgM and usually also TBEV-IgG antibodies are present in the first serum samples drawn when the CNS symptoms have manifested (i.e. during the second phase of the disease). Intrathecal IgM and IgG responses can be detectable in CSF several days after their appearance in serum and previous studies found these in all cases at Day 10 after onset of CNS symptoms in [19,20].
TBEV can be isolated, or detected by real-time (RT)-PCR, from blood during the first phase of the illness. The period of possible isolation and detection can be prolonged in patients with progressive disease and in immunocompromised patients [1,[21][22][23].
Enzyme immunoassays (EIA, ELISA), based either on purified virions, recombinant proteins, or recombinant virus-like particles obtained by expression of prM and E proteins, are usually used for specific serodiagnosis [1,24]. Haemagglutination inhibition (HI) has also been widely used but it measures all different antibody classes and thereby requires a significant rise in antibody titre for a clear-cut diagnosis.
High cross-reactivity of the antigenic sites among the various pathogenic flaviviruses may cause TBE diagnostic problems when other flavivirus(es) co-circulate (e.g. West Nile virus (WNV) in southern and central

Samples
All samples were stored at −20 °C before analysis by Reascan TBE IgM. All TBE IgM positive samples used correspond to TBE-confirmed cases in accordance with the European Union (EU) case definition criteria [18].

HUSLAB, Helsinki, Finland
Serum samples from Finnish patients having prior tested positive (n = 55) and negative (n = 100) for TBE IgM by HUSLAB EIA as described below were analysed at HUSLAB in Helsinki, Finland.

National Institute for Health Development, Tallinn, Estonia
Estonian patients' serum samples (n = 47), which were previously found positive for TBE IgM by IMMUNOZYM FSME (TBE) IgM (Progen, Heidelberg, Germany) as described below were analysed at the Laboratory of Clinical Microbiology in Uppsala, Sweden.

Institute of Microbiology and Immunology, Ljubljana, Slovenia
Serum samples from Slovenian patients, which previously tested TBE IgM positive (n = 8) and TBE IgM negative (n = 31) by Enzygnost Anti-TBE virus ELISA IgM test (Siemens GmbH, Marburg, Germany) were analysed     TBEV antigen produced in insect cells infected with recombinant baculovirus expressing TBEV prM and E proteins [24]. Briefly, the diluted serum (or CSF) samples were incubated on goat anti-human IgM (Cappel/ MP Biomedicals, Santa Ana, California, United States (US)) coated EIA strips for 30 min at 37 °C. Unbound excess antibody was washed by phosphate-buffered saline (PBS) with 0.05% TWEEN-20. The recombinant antigen (diluted recombinant baculovirus-infected insect cell supernatant) was incubated with the strip for 45 min 37 °C and washed as above. A mouse monoclonal anti-TBEV-E protein antibody 1786 [28] was incubated and washed as above, after which a peroxidase-conjugated donkey anti-mouse IgG antibody (Jackson Immunoresearch, West Grove, Pennsylvania, US) was added, incubated and washed. The enzyme reaction was detected by TMB substrate (Sigma, Rockford, Illinois, US) and the reaction stopped by 0.5 M H 2 SO 4 . The absorbance was measured at 450 nm and the absorbance values adjusted as described in Jääskeläinen et al. 2003 [24].

Commercially available tick-borne encephalitis IgM enzyme immunoassays
The Enzygnost Anti-TBE virus ELISA IgM, IgG test was used according to the manufacturer's instruction in the Slovenian and Swedish laboratories. The IMMUNOZYM FSME (TBE) IgM was used according to the manufacturer's instructions in the Estonian laboratory. The SERION ELISA classic FSME Virus/TBE Virus IgG and IgM was used according to manufacturer's instructions in the Dutch laboratory.

Other serological and molecular assays
Assays used by the different laboratories for diagnosis of other infections than TBEV are listed in Table 1 as well as any serological or molecular assays for TBE diagnosis.

Ethical statements
Specific ethical approvals were at the time of the study considered not needed for Sweden, Estonia or Finland for anonymous/within hospital laboratory pseudonyms. For Slovenia, this study was performed as a part of the ongoing study on tick-transmitted infections at the Institute of Microbiology and Immunology, Ljubljana, and ethical clearance was obtained according to national legislation (NMEC number 131/06/13). For the Netherlands: ethical approval was obtained from the Erasmus MC Medical Ethical Committee (MEC-2015-306) to anonymously analyse the used samples of all patients.

Results
The results from the evaluation by the five diagnostic reference laboratories are shown in Figure A-D.

Assay sensitivity
Based on a locally available total of 172 serum samples from earlier diagnosed TBE patients, the sensitivity was calculated to be 99.4%, i.e. all except one serum sample were found positive by the assay ( Table  2 and Table 3).

Assay specificity
Based on 141 locally-available samples prior testing TBEV IgM negative using the routine method for each laboratory (i.e. other commercial TBE assays than ReaScan TBE or quality assured in-house methods) in Finland, Slovenia and Sweden, a specificity of 97.9% (138/141) was found ( Table 2 and Table 3).
To further analyse the specificity, 73 patient serum samples from CMV, EBV, HSV and VZV infected patients and 10 serum samples from A. phagocytophilum infected individuals were tested. Based on those 83 potentially interfering samples, a specificity of 96.4% (80/83) was found (Table 4 and Table 5).

Other flavivirus/togavirus infections
The flaviviruses are known to share a number of common antigenic sites causing serological cross-reactions and thereby problems in serological diagnostics. To evaluate the level of potential cross-reactivity of the assay to other flavivirus infections, a total of 57 serum samples from DENV, JEV, WNV, and ZIKV infected individuals were analysed. In addition, 15 sera from JEV-and YFV-vaccinated individuals and 10 sera from individuals infected with CHIKV (a togavirus) were analysed. Of these 82 sera 81 showed a negative result and one sample (from an YFV-vaccinated individual) was positive (Table 4), giving a specificity of 98.8% ( Table 5).

ReaScan TBE compared to Enzygnost IgM ELISA
Thirty-one TBEV IgM negative patient serum samples from Slovenia, 18 of which were previously tested as false TBEV IgM positive by Enzygnost Anti-TBE virus ELISA, were analysed by the ReaScan TBE. These 18 false TBEV IgM positive samples, initially tested by the Siemens Enzygnost IgM assay were samples from patients whose consecutive second (one week later) and third serum (two weeks later) samples were available. In all consecutive serum samples they did not present any specific IgG against TBEV, i.e. no IgG seroconversion within a three-week time period. Besides, their CSF samples were IgM and IgG negative. Based on these data, these TBEV IgM positive samples were considered as false positive. One of the samples showed equivocal result by the ReaScan IgM, all other 17 samples showed a negative result (Table 6).
Of 18 samples from seven Slovenian PCR-confirmed TBE patients, 11 and 10 tested positive by the ReaScan TBE and Enzygnost, respectively (Table 6). A total of 26 serum samples from vaccine failure TBE patients showed similar results with both methods (Table 6).

Discussion
TBE is an important and growing public health problem in Europe and Asia [1][2][3][4]. In 2016, France reported a marked increase of TBE cases, following which Sweden reported its highest numbers of TBE cases since 1956 when comparable records were first kept, with 391 and 385 cases during 2017 and 2018, respectively, while in Finland the number of TBE cases has more than doubled during the last decade, fatalities have occurred and the disease has spread to new areas [29][30][31][32]. The Netherlands, previously TBE-free, has reported its emergence from 2016 [33], and in Denmark, TBE was reported in a new geographical hot spot (Tisvilde Hegn, in Northern Zealand) [34]. Furthermore, TBEV was recently detected for the first time in ticks in the United Kingdom (2019) [35]. The cause of the increase in TBE case numbers despite an increase in the frequency of TBE vaccinations [36] in several European countries is not yet fully understood, but climate change (increased temperature and humidity) may be Table 4 ReaScan TBE a performance on serum samples containing potentially cross-reactive or problematic sera b (n = 209 samples)  one important factor. Other potential reasons can be increased awareness and testing.
Due to the very low amount, or in most cases complete absence, of detectable TBEV RNA at the onset of the CNS symptoms in immunocompetent patients, serology is generally required for TBEV diagnostics. Routine laboratory diagnosis of acute TBE cases is therefore usually performed by immunoassays such as EIAs or ELISAs, designed for the detection of TBEV-specific IgM. The major potential problems in serological TBEV diagnostics, as also seen for the diagnosis of all other human-pathogenic flavivirus infections (e.g. DENV, YFV, JEV, ZIKV), are due to high level of cross-reactive epitopes among the flaviviruses: (i) cross-reactions by antibodies induced by infection by another flavivirus, (ii) cross-reactive antibodies induced by vaccination against another flavivirus, and (iii) less specific viral antigens used in the assay. Several commercial TBE IgM assays are available and some comparisons between these assays are available at present [27,[37][38][39]. In this study, we compared the ReaScan TBE IgM rapid test with the conventional routine methods of several laboratories in a clinical laboratory setting. Our results revealed a high sensitivity (99.4%) when serum samples from 172 previously diagnosed TBE patients from five European countries were analysed. We also found a high specificity (97.9%) when 141 samples from previously diagnosed non-TBEV infections were analysed. Interestingly, there were no major differences in sensitivity or specificity in the five laboratories that participated in this study, despite the differences in the respective reference assays.
The specificity of the assay evaluated here was further analysed by the use of well-known 'serologically problematic' samples (i.e. from patients with different acute herpesvirus (CMV, EBV, HSV and VZV) or A.
The assay presented here is based on soluble recombinant TBE virus antigen expressed in insect cells. It is well known that recombinant antigens expressed by mammalian or insect cells usually result in higher specificities as compared with antigens expressed in bacterial systems [24]. Also, the assay format, where the measurement of the test line intensity is performed by a dedicated test reader (instead of visual interpretation), may improve both the sensitivity and the specificity of the assay and reduces the user-dependent errors. The fact that the IgM responses to the various flaviviruses are generally more virus-specific, as compared with the later IgG responses, is in line with the observed high specificity, also when analysing acute sera from e.g. DENV, JEV, YFV and ZIKV infected patients.
Our results thereby indicated that the recombinant antigen used in the assay is antigenically sufficiently different from the corresponding antigens among DENV, JEV, YFV and ZIKV. This is likely due to the more distant relationship of the tick-borne flaviviruses (e.g. TBEV) to the mosquito-borne flaviviruses such as DENV, JEV, YFV and ZIKV.
In conclusion, our study revealed a similarly high sensitivity and high specificity of the ReaScan TBE as compared with the assays used by the diagnostic laboratories that participated in this study. Other advantages of the assay are its rapid format (20 min), its suitability for analysing one sample at each time (i.e. no need to collect samples to have enough numbers to perform an EIA test), and the possibility to use it as a Point-of-Care (POC) test to allow rapid diagnosis in primary care settings/outside central laboratories by non-specialised personnel.
However, to fulfil the EU case definition criteria, detection of TBEV IgM alone is not sufficient. Our results suggest that the assay is valuable as a convenient complement for diagnosis of acute TBE infections, either as a complementary method for rapid IgM detection for laboratories that already have a TBE EIA/ELISA-method or as a stand-alone method for preliminary results in a laboratory or POC setting.

Table 6
Results of Enzygnost IgM ELISA a and ReaScan TBE b assays on various types of samples (n = 75 samples)