Laboratory management of Crimean-Congo haemorrhagic fever virus infections: perspectives from two European networks.

Background Crimean-Congo haemorrhagic fever virus (CCHFV) is considered an emerging infectious disease threat in the European Union. Since 2000, the incidence and geographic range of confirmed CCHF cases have markedly increased, following changes in the distribution of its main vector, Hyalomma ticks. Aims To review scientific literature and collect experts’ opinion to analyse relevant aspects of the laboratory management of human CCHF cases and any exposed contacts, as well as identify areas for advancement of international collaborative preparedness and laboratory response plans. Methods We conducted a literature review on CCHF molecular diagnostics through an online search. Further, we obtained expert opinions on the key laboratory aspects of CCHF diagnosis. Consulted experts were members of two European projects, EMERGE (Efficient response to highly dangerous and emerging pathogens at EU level) and EVD-LabNet (Emerging Viral Diseases-Expert Laboratory Network). Results Consensus was reached on relevant and controversial aspects of CCHF disease with implications for laboratory management of human CCHF cases, including biosafety, diagnostic algorithm and advice to improve lab capabilities. Knowledge on the diffusion of CCHF can be obtained by promoting syndromic approach to infectious diseases diagnosis and by including CCHFV infection in the diagnostic algorithm of severe fevers of unknown origin. Conclusion No effective vaccine and/or therapeutics are available at present so outbreak response relies on rapid identification and appropriate infection control measures. Frontline hospitals and reference laboratories have a crucial role in the response to a CCHF outbreak, which should integrate laboratory, clinical and public health responses.


Introduction
Crimean-Congo haemorrhagic fever virus (CCHFV) is a tick-borne pathogen that causes a frequently lethal disease in humans and is considered to be a major emerging infectious disease threat spreading to and within Europe [1][2][3].
The severity of the disease, the presence of domestic and wild animal reservoirs and/or vectors, a large population of susceptible humans, limited diagnostic capacities and resources for epidemiological/   Crimean-Congo haemorrhagic fever virus strains with complete S segment available as at 5 December 2017, were collected from GenBank database and clustered at 100% with CD-HIT v4.6. Sequences were aligned with MAFFT v7.123b. Phylogenetic analysis were performed using RAxML v8.2.10 with GTRGAMMA model and 1,000 bootstrap inferences. For graphical exemplification, only representative sequences in each clade for each country were selected and reported in the phylogenetic tree.
Branches owing to different clades are presented in the following colours: Africa: clade I (red), clade II (light green) and clade III (brown); Asia: clade IV (dark green); Europe: clade V (purple), clade VI (light blue) and clade VII (blue). For each strain we reported GenBank ID, isolate ID, host, country and collection date, if present. Bootstrap values are shown for each clade.
ecological investigation, as well as the absence of effective prophylaxis and treatment render CCHFV a pathogen with outbreak potential [4].
Since 2000, the incidence and geographic range of CCHF cases have markedly increased [5,6] following an expanding distribution of its main vector, ticks of the genus Hyalomma, specifically the Hyalomma marginatumspecies [1,7,8]. In Turkey, nearly 900 new CCHF cases occur annually, with a total of 9,787 cases reported from 2002-15 [9]. CCHF is endemic in the Balkan region, in Kosovo, 228 cases were reported from 1995-2013 [10], In Bulgaria, over 1,500 cases have been reported from 1952 [11]. In the European region, cases of human infection have also been reported from Albania, Russian Federation, Georgia, Greece, and Ukraine [12]. Imported cases have been reported in France [13], United Kingdom [14], Greece [15] and Germany [16]. A detailed review of other outbreaks has been recently published by Papa et al. [11].
Public health systems (including diagnostic laboratories) should be prepared to respond to the increased circulation of the virus in endemic EU countries, the potential for importation of human CCHF cases or the emergence of virus circulation in new areas e.g. Spain [17].
The objectives of this study were to amalgamate the expertise of two EU expert networks (i) EMERGE (Efficient response to highly dangerous and emerging pathogens at EU level) [18] and (ii) EVD-LabNet (Emerging Viral Diseases Laboratory Network) [19], in order to select and analyse the relevant and some of controversial aspects of CCHF disease diagnostics with implications for laboratory management of human CCHF cases and any exposed contacts.

Methods
We carried out an on line research of published paper related to CCHFV molecular detection methods. References were obtained by an online search in PubMed using an intentionally wide search-query to Based on all available complete S segment CCHFV genomes (163 sequences as at 5 December 2017) retrieved from GenBank.
For all strains analysed, the collection countries are presented in the following colours: Africa: clade I (red), clade II (light green) and clade III (brown); Asia: clade IV (dark green); Europe: clade V (purple), clade VI (light blue) and clade VII (blue). ensure that a large number of papers was retrieved also for a rare disease such as CCHF.
The query produced a large number of papers, 20% of them were discarded after a narrative review, as they did not contain a detailed description of the detection methods employed including the nucleotide sequences of primers and/or probes. The search was done by one author and the results discussed among the authors. Papers related on non-previously retrieved molecular detection methods or to others relevant aspects discussed in this report have been directly provided by experts. For phylogenetic analysis all available CCHF virus genomes by 5 December 2017 were retrieved from GenBank (https://www.ncbi.nlm.nih.gov/nucleotide), using 'txid1980519(Organism)' as term of query. All analyses have been focused only on CCHFV S-segment, because it resulted as the most conserved gene across CCHFVs [8,20] and also because mostly all retrieved molecular methods has S segment as target. CCHF virus strains with complete S segment were selected and clustered at 100% with CD-HIT v4.6. A total of 163 sequences available at 5 December 2017 were obtained and aligned with MAFFT v7.123b in local pair mode. Phylogenetic analysis were performed with RAxML v8.2.10 using GTRGAMMA model and 1000 bootstrap inferences.
A preliminary text was drafted and discussed among the experts by email and during EMERGE and EVD-LabNet networks' 2017 and 2018 annual meetings. Most of the relevant and some of controversial aspects of CCHF disease with implications for laboratory management have been selected and analysed in the following sections. In the present paper, all the expressed opinions take into account both published data and personal experience of the experts.

Crimean-Congo haemorrhagic fever virus clades distribution
CCHFV (family Nairoviridae, genus Orthonairovirus) is tick-borne and is maintained in a tick-vertebrate-tick cycle with Hyalomma marginatum, the main vector species in Europe. Given the wide distribution of its vector, CCHFV has been detected over a wide geographic range: Africa, Europe, Asia and the Middle East [5,21].
CCHFV is an enveloped, tri-partite, negative-sense, RNA virus. The large genome segment (L) encodes the RNAdependent RNA polymerase (L protein), the medium segment (M) encodes the glycoproteins GN and GC, while the small segment (S) encodes the nucleocapsid protein (N).
Phylogenetic tree (Figure 1) was built, including only 65 of 163 representative strains with reported location of provenance either in GenBank records or in the associated papers. Taking into account similarity and geographic locations of the different viral lineages, seven genetic clades were identified: three prevalently diffused in Africa (clades I-III), three in Europe (clades V, VI and VII) and one in the south of Asia (clade IV).
Most of the isolates causing outbreaks in eastern Europe belong to clade V, whereas clade VI and VII include largely divergent strains isolated from ticks in Greece (including the prototype strain AP92) [21,22] and Russia (GenBank accession number KR814833 and KR814835).
Moreover, isolates belonging to the African clade III were collected from infected ticks in 2010 and 2014, and recently [23] a virus aligning to this clade was the cause of an outbreak in Spain [17].
For all strains analysed, the collection country was recorded and represented on the world map in Figure  2.

Transmission mode
Human infections are usually observed as single, sporadic cases when people in rural areas are bitten by ticks that have become infected by feeding on viraemic wild and domestic animals like hares, hedgehogs, horses, livestock and possibly birds [21]. The infection in animals is generally asymptomatic; at most, a mild fever may be noted.
In addition to tick exposure, CCHF infection can result from direct contact, especially through mucous membranes or skin wounds, with crushed infected ticks or the blood of infected animals (principally among shepherds, farmers, abattoir workers and veterinarians). Person-to-person transmission can also occur through contact with virus-containing bodily fluids of patients during the first 7-10 days of illness [21]. Unprotected contact with other bodily fluids like saliva or urine, may also represent a risk for humans [24]. Nosocomial transmission to healthcare workers, transmission among patients sharing the same room [25] and possible sexual transmission [26,27] have also been reported.

Relevance of viraemia
The typical duration of viraemia ranges from 1-9 days [28,29], and there is so far no evidence of detectable viraemia during the incubation period [30]. However, the positivity of CCHFV RNA in serum has been exceptionally reported up to 36 days from the onset of symptoms [31]. Studies investigating the presence and persistence of CCHFV in other body fluids are limited. Viral RNA was detectable up to 10 days and as late as 25 days after onset of symptoms in saliva [24] and urine [31] respectively, but no data on virus viability are available. Viral load is the most important prognostic factor: a value of viraemia higher than 10 8 copies/mL is associated with fatal outcomes [32]. Viraemia decreases significantly over time in surviving patients, but remains persistently high in non-survivors [32,33].

Table 1a
Published molecular assays for Crimean-Congo haemorrhagic fever virus detection as at 31  For each assay, the position of amplicon is reported to respect IbAr10200 (NCBI reference sequence NC_005302). Names of each primer and probe correspond to those reported in the reference.  For each assay, the position of amplicon is reported to respect IbAr10200 (NCBI reference sequence NC_005302). Names of each primer and probe correspond to those reported in the reference. For each assay, the position of amplicon is reported to respect IbAr10200 (NCBI reference sequence NC_005302). Names of each primer and probe correspond to those reported in the reference.

Table 1c
Published molecular assays for Crimean-Congo haemorrhagic fever virus detection as at 31 November 2017

Antibody kinetics
All CCHFV genotypes belong to one serogroup [21]. Cross-reactivity between CCHFV and other nairoviruses infecting humans ( Erve virus [34]), has not been described, although monoclonal and polyclonal antibodies to the N protein of CCHFV were found to crossreact weakly with Dugbe virus N protein [35].
Nucleocapside-directed IgM antibodies have been identified as the initial serological marker during infection, becoming detectable in a median of 2-3 days after disease onset, followed by glycoprotein precursor (GPC) directed IgM (4-6 days) and IgG antibodies (5-6 days) [28]. In another report, CCHFV IgM was detectable from 4 days after the onset of disease for up to 4 months. The maximum level of antibody titres was usually reached in the second to third week of illness [1]. IgM titre typically declines to undetectable levels 4 months after the onset of symptoms [29].
Antibody production against CCHFV is an important prognostic indicator for survival [37]. Patients with fatal outcome rarely develop measurable antibody responses (reviewed in [37]) and Saksida et al. observed a reverse correlation between viral load and antibody levels in fatal CCHF cases [33], indicating that an impaired immune response leads to uncontrolled replication of the virus. High levels of interleukin-10 (IL-10), an anti-inflammatory cytokine, were detected in patients with fatal outcomes and were lowest in patients with a moderate disease course [33]. It was hypothesised that CCHF could be the result of a delayed and downregulated immune response caused by IL-10, which leads to an increased replication and spread of CCHFV throughout the body [33].
There is a general agreement among the consulted experts that a critical aspect for laboratory biosafety and operation is the proper and reliable inactivation of specimens before they can be removed from the highlevel biocontainment environment for further diagnostic testing. Some of these inactivation methods include (i) chemical treatment (i.e. Buffer AVL or Buffer RLT, Qiagen, Hilden, Germany) + 100% ethanol, SDS, 0.5% Tween-20 (Thermofisher, Waltham, Massachusetts, USA) [46,47] or (ii) heat treatment plus riboflavin (vitamin B 2 ) [48].
It is opinion of the experts that further evaluation of inactivation procedures are needed, especially for their impact on other laboratory tests necessary for clinical evaluation and increased survival rates of patients. Although there is no direct evidence of its effects on CCHFV, it has been shown that Triton X-100 (Sigma-Aldrich, Saint Louis, Missouri, USA can decrease the biohazard risk of performing laboratory tests on samples from patients infected with other haemorrhagic fever viruses (i.e. Ebola), without affecting the results of biochemical tests [49][50][51][52].

Transport of diagnostic samples
General guidelines for suspected viral haemorrhagic fever infections apply for the transport of diagnostic samples from CCHF-suspected cases; these are listed under guidelines as Category A, Infectious Substances Affecting Humans UN 2814 and must be transported in packaging that meets the United Nations class 6.2 specifications and complies with Packing Instruction P620 [53].
Additionally, as for other biological resources, the exchange of CCHF samples or virus strains needs to comply with the Nagoya protocol on access to genetic resources and the fair and equitable sharing of benefits arising from their utilisation, which regulates transnational exchange between countries [54].

Diagnosis
The choice of which CCHF detection assays should be used for diagnostics with maximum sensitivity and specificity depends on the stage of disease and the specimens available.
Laboratory diagnosis of a patient with a clinical history compatible with CCHF is generally performed during the acute phase of the disease by viral RNA (RT-PCR) detection in blood [21]. In addition to blood (serum, plasma or whole blood), other possible specimens for molecular detection are saliva, urine [24,32] or post mortem biopsy of the liver and bodily fluids (including semen, for infection control purposes) [26,27].
The available data, limited to very few patients, do not allow detailed comparisons of the sensitivity of RT-PCR detection methods performed on different sample types and, in particular, of urine and saliva vs serum samples. In two of six infected patients reported in a review of CCHF cases in Kosovo* [31], viraemia was detectable up to 30 days after the onset of symptoms.
In the same investigation, one patient's urine was PCR-positive before the serum, and in another patient viruria continued longer than viraemia; however, more detailed studies on viruria are required. Further, both viruria and viraemia are detectable several days after the appearance of IgG response [31]. No chronological data are provided about detection of CCHFV RNA in saliva [24].
During the small outbreak in Spain involving two patients in 2016, positive RT-PCR results were obtained via saliva and vaginal swab, but they became negative when viraemia was still detectable [17]. Virus isolation was attempted from these samples, but was not successful (Maria Paz Sanchez-Seco, personal communication, EVD-LabNet 2 nd annual meeting, Rotterdam October 2017).

Molecular detection
There is high genetic diversity within the different CCHFV strains (Figure 1), which consequently hampers the performance of molecular tests. As a result, a range of different methods employing varied primer/probe combinations have been developed and a truly universal assay has been difficult to devise. Table 1 lists published molecular assays retrieved by our PubMed search: two single round PCR, six nested PCR, 10 realtime PCR, one loop-mediated isothermal amplification (LAMP) and one recombinase polymerase amplification (RPA). Indication on reference testing materials and sensitivity/specificity of the tests are also reported, when declared.
Therefore there is an agreement of experts that is advisable to perform more than one test to avoid exclusive reliance on a single assay and a single target, taking into account the travel history and the geographic distribution of the different strains.

Serological assays
In published investigations, the methods employed for the detection of antibodies are indirect immunofluorescence assays (IFAs) and enzyme-linked immunosorbent assays (ELISAs) [55][56][57]. Several commercial kits are available ( A CCHFV seroneutralisation test is not normally performed for diagnostic purposes; it requires work with an infectious virus, necessitating a BSL4 laboratory, and is difficult to perform. However, reverse genetic approaches employing non-infectious reporter viruses have been described recently [59], enabling neutralisation to be performed at low containment.

Virus isolation
Viral isolation, i.e. from blood or organ for further characterisation or infectivity studies, is performed under BSL4 conditions on either LLC-MK2, Vero, BHK-21 or SW-13.4 cell lines and can be achieved in 2-10 days [60]. CCHFV generally produces no or little cytopathogenic effect and viral growth can be detected by IFA with specific monoclonal antibodies [41] or by molecular tests. When viral isolation on cell cultures fails, it can be attempted in new-born or immunodeficient mice.
It is preferable that viral isolation is performed on samples collected during the first 5 days of infection, when the viraemia levels are high [58,61].

Laboratory diagnosis of CCHFV infection
There is no official, agreed-upon case definition for CCHF in the EU, though several case definitions adopted by EU countries were reviewed by the European Centre for Disease Prevention and Control (ECDC) [62,63]. According to a report published by ECDC in 2014 [60], most countries used, at least for surveillance purposes, the EU case definition established for Viral Haemorrhagic Fever [64].
Taking into account that laboratory screening is usually performed using molecular methods, we propose as expert opinion the following molecular diagnostic

Box
Criteria proposed for laboratory confirmation of a clinically suspected Crimean Congo haemorrhagic fever case For laboratory confirmation of a clinical CCHF diagnosis, the expert group opinion is that a CCHFV infection is laboratory confirmed when at least one criteria in the Box is fulfilled.
• Detection by molecular tests of CCHFV RNA, in blood (whole blood, serum or plasma) or in other bodily fluids or tissues; • Detection of CCHFV IgM or relevant (fourfold) increase in CCHFV IgG titres between two serologic samples (acute and convalescence phases); • CCHFV isolation and/or detection of CCHF viral antigens in blood (whole blood, serum or plasma).

Figure 3
Algorithm for molecular diagnosis of Crimean-Congo haemorrhagic fever acute infection based on expert opinion CCHFV: Crimean-Congo haemorrhagic fever virus.
a The preferred biological specimen is blood (serum, plasma or whole blood). Other possible biological specimens are saliva, urine [24,31] and post-mortem biopsy or bodily fluids (including semen, for infection control purposes [26,27] algorithm for patients with suspected CCHF infection ( Figure 3). However, it is advisable, especially when the molecular tests are negative, to perform also serological tests on all suspected patients.
Other relevant aspects related to management of CCHF patients as clinical manifestation and discharging criteria are reported in the Supplementary material.

Discussion
CCHF is an important global health threat, as underlined by its inclusion in the list of priority diseases in the WHO document 'An R & D blueprint for action to prevent epidemics' [65].
In Europe, two autochthonous cases in Spain in 2016 [17] and the observed risk of importation of travelassociated cases [66] reinforce the notion that public health systems must be ready to respond to a potential emergence of CCHF.
Prompt and accurate laboratory diagnosis during the first days of the disease is critical to improve patient management, guide infection control measures and reduce case fatality. Early detection of viral RNA in blood is considered the gold standard diagnostic approach in the acute phase of the disease [21]. While the CCHFV RNA RT-PCR diagnostic test is most commonly performed in specialised laboratories, where non-commercial diagnostic assays and related reference biological material are available, this capability is more limited in rural areas and small cities where the majority of reported cases have occurred [9]. The availability of simple to use, commercial diagnostic tests will increase the number of laboratories performing RT-PCR or a similar NA detection strategy. However, laboratory capacity does not automatically mean capability, and laboratories implementing such tests would benefit from external quality assessments (EQA) of their capability to detect CCHFV in clinical samples, including monitoring of the effects of any corrective actions taken. The difficulty of clinical laboratories in establishing the diagnosis a CCHFV infection underlines the need to perform confirmatory tests in reference laboratories for both positive and negative samples collected from patients suspected of CCHF infection. In addition, appropriate biosafety measures must be in place when performing molecular testing.

Proposed measures to improve CCHFV laboratory preparedness and response
Measures to improve CCHFV laboratory preparedness and response should include: improving molecular tests to overcome the lack of sensitivity due to the high variability of the CCHFV genome; providing a positive control panel for molecular diagnostics, including the different CCHFV genotypes (including for serology testing) in order to support the improvement of diagnostic capability of clinical laboratories; enrolling diagnostic laboratories appointed for CCHFV diagnostics regularly in EQA programmes; improvement of diagnostic algorithms building on clinical experiences; validation of procedures to inactivate infectivity of clinical samples; and establishing of an international biorepository for the collection and storing of clinical samples, with the aim of validating new diagnostic tests and supporting pathogenicity studies. Some of these activities, including the evaluation of laboratory capability [67] and EQA [68]

Conclusions
There are knowledge gaps concerning the putative persistence of the CCHFV in various body compartments of survivors and the related consequences for infection transmission. Basic knowledge is needed to provide evidence to better inform hospital discharge guidelines and these issues need further research. Furthermore, a One Health approach is required for adequate public health preparedness for CCHF, and relevant measures should include vector and animal surveillance, focusing particularly on migratory birds [5,9,69]. Greater awareness of the circulation of CCHFV in vectors/animals in specific geographic areas is fundamental in order to alert public health systems. Information on the circulation of CCHFV can be obtained by a syndromic approach and by including CCHFV testing in the diagnostic algorithm of severe febrile infectious diseases of unknown origin.
Until an effective vaccine and/or therapeutics have been developed, the CCHFV outbreak response will continue to rely on rapid identification and appropriate infection-control measures. Front-line hospitals, as well as reference laboratories, have a crucial role in the outbreak response, which should integrate laboratory, clinical and public health responses. Union's Horizon 2020 research and innovation program', grant agreement no. 732732.

Conflict of interest
None declared.

Authors' contributions
BB data analysis, coordinating the activities and writing manuscript; GCEM data analysis and writing manuscript; CBR and ADC study coordinator, data analysis and writing manuscript. BB, CEMG, MK, TA, SB, IC, RG, RH, GK, CML, AM, AP, MPSS, AVS, HZ, CN, MRC, GI, CBR, and ADC contributed to the conception and design of the work, the interpretation of data, the revision of the manuscript and the approval of the final version.