Molecular diagnostics

Common targets for both ebolaviruses and MARV NAAT testing include the NP, L and GP genes.33 The gene sequence of GP is strain specific and therefore could be used for differentiation between infecting species; the sequences of VP40 and NP are more conserved. As recently published by Clark et al, there are a number of laboratory designed tests (LDTs) using reverse transcriptase (RT)-PCR, quantitative RT-PCR (qRT-PCR) and RT-loop mediated isothermal amplification to detect ebolaviruses and MARV,34 as diagnostic testing has historically been carried out in international reference laboratories with varying performance when evaluated by independent external quality control proficiency testing.32 33 35

The 2014–2016 West Africa ebolavirus outbreak resulted in the rapid development of new commercially available benchtop and point-of-care (POC, ie, fully automated platforms ideal for use in low-resource sites) NAATs to allow for rapid detection of ebolavirus infection in lower resource laboratories.36–41 Many of these tests are available for purchase and are approved by a stringent regulatory authority (SRA), either through WHO Emergency Use Authorization and listing (WHO EUAL), US Food and Drug Administration (FDA) Emergency Use Authorization (FDA EUA) or European Commission (CE) marking.36 At this time, there are only two MARV diagnostics approved by SRAs, mainly as part of a panel of assays with the ability to detect and differentiate between multiple pathogens.36 41

Quantification of the viral load is useful for patient management as it can potentially predict the probability of recovery since fatal cases often exhibit rapid increases in viraemia with viral load in blood for ebolaviruses reaching approximately 10⁹ copies/mL, while survivors’ viral titre ranges around 10⁷ copies/mL.42 Quantitative detection is achievable with several real-time RT-PCR-based methodologies (online supplementary tables S1 and S2).

A major limitation of the majority of the available NAATs is their limited strain coverage. Only a few NAATs are able to test for multiple ebolavirus or MARV subtypes, which can potentially result in false negatives during outbreak responses or surveillance activities. Further development is needed for easily deployable, POC NAATs that can distinguish multiple ebolavirus and MARV subtypes for future outbreaks and improved surveillance.

Serological assays

Serology is useful in epidemiological studies of infected hosts by detection of ebolavirus-specific or MARV-specific antigen or IgM and IgG antibodies. Detection of anti-ebolavirus-IgM or anti-MARV-IgM indicates recent infection and can be detected as early as 2–4 days after symptom onset, while anti- ebolavirus-IgG or anti-MARV-IgG can be detected around 8–10 days after symptom onset and persist for up to 2 years after infection (figure 3). IgG detection is primarily used to identify individuals who have recovered from EVD or MVD or to assess seroprevalence in the community.8 32 43–46

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Diagnostic methodology and phase of illness of filovirus disease. Reprinted from J Pathology, 235(2), Martines RB et al., Tissue and cellular tropism, pathology and pathogenesis of Ebola and Marburg viruses, 153–74. Copyright (2014), with permission from John Wiley and Sons. RT-PCR, reverse transcriptase PCR.

There are currently no SRA-approved benchtop ELISAs for ebolavirus or MARV detection. Two commercial ELISAs are available for research use only to detect EBOV VP40 (online supplementary table S1) and IgG/IgM, and two LDTs are available to detect EBOV IgG/IgM.31 While ELISAs were developed for EBOV-specific antigen detection, RT-PCR is recommended for ebolavirus detection. For MARV, one ELISA LDT is available for viral detection.47

Rapid diagnostics

Rapid diagnostic tests (RDTs, also called immunochromatographic tests or lateral flow immunoassays) can leverage the same antibody/antigen capture agents as an ELISA but in a lateral flow strip format, allowing faster results with minimal specimen processing (although proper infection control practices are still required, and more definitive identification may be necessary depending on the assay’s limit of detection and clinical sensitivity).48 RDTs are ideal as screening tests, suitable for field testing and low infrastructure settings such as the clinic.32 During the 2014–2016 ebolavirus outbreak, rapid screening of cadavers to support local burial practice was critical in controlling the outbreak and preventing the spread of virus.

The 2014–2016 ebolavirus outbreak resulted in the rapid development of new RDTs. As of May 2018, five EBOV RDTs were approved for the detection of EBOV by SRAs (either through WHO EUAL, FDA EUA or CE marking),36 49 50 with new tests in development.51 At this time, there are no MARV RDTs approved by SRAs.

Implementation requirements of the different filovirus diagnostic types are shown in table 1. Molecular diagnostics typically require the highest laboratory infrastructure, including biosafety hoods and a clean room, while most ELISAs can be run on the benchtop in more modest laboratory environments (providing suitable PPE is used when handling samples). POC or near-POC NAAT tests are generally automated and robust, and may be performed without a biosafety hood depending on sample preparation requirements. RDTs are typically designed for field or clinic use.

Syndromic multiplex approach

Viral haemorrhagic fever (VHF) refers to a diverse group of animal and human illnesses in which fever and haemorrhage are caused by viral infection. In regions where VHF viruses may be endemic and maintained through natural reservoirs, a panel for distinguishing the viral family causing infection and, potentially, co-infection will be useful for surveillance as well as appropriate isolation and infection prevention and control measures. While the majority of ebolavirus outbreaks are caused by EBOV and SUDV, current tests do not have full strain coverage. A multiplexed, syndromic approach is a more useful and efficient strategy for surveillance, differential diagnosis and outbreak determination than relying on single, pathogen-specific assays. Currently, there are few commercially available molecular diagnostics that can detect multiple fever-causing agents.41 52

Lack of a regulatory approval pathway for filovirus diagnostics

During the 2014–2016 West Africa ebolavirus outbreak, WHO established an EUAL procedure for in vitro diagnostics (IVDs) in order to expedite the introduction of critically needed technologies during an outbreak setting.53 Similarly, the US FDA developed an EUA programme to expedite the approval process of American-made products.54

However, on 29 March 2016, WHO announced that the ebolavirus situation in West Africa no longer constituted a Public Health Emergency of International Concern. While previously submitted and approved products under WHO EUAL may still be purchased and used, both WHO and FDA recommend test manufacturers to submit for full regulatory approval.53 54 However, the high cost, limited availability of well-characterised clinical samples and time needed for resubmission may deter manufacturers from pursuing additional regulatory approvals of their existing product or new manufacturers to submit their product. This challenge, coupled with the evolving CE-IVD marking system in Europe,55 makes the diagnostic regulatory pathway for filoviruses and other emerging threats unclear.

In addition, as the ebolavirus outbreak waned, so has access to confirmed positive samples, which are required in order to fully validate an assay for regulatory approval. These circumstances make the regulatory approval process after an outbreak a significant challenge that must be addressed in order to ensure access to reliable and robust diagnostics in the future.

Need for improved sensitivity and specificity of screening and triage tools

The need for rapid, sensitive, safe and simple ebolavirus diagnostic tests was highlighted in a November 2014 Call for Diagnostics by WHO since efforts to contain the ebolavirus outbreak were hampered by cumbersome, slow, complex and costly diagnostic tests. In order to stimulate product development, WHO issued a detailed product profile of the ‘ideal’ rapid, sensitive, safe and simple diagnostic test considered most likely to accelerate interruption of virus transmission in severely resource-constrained settings.56 It was suggested that the ‘ideal’ test should be designed for use in peripheral health clinics with no laboratory infrastructure in place, no biosafety precautions beyond the wearing of PPE, with three or fewer steps, and a time-to-result less than 30 min. Shortly after the WHO Call for Diagnostics was published, a target product profile (TPP) for ebolavirus diagnostics was developed by FIND, MSF, WHO and partners in 2014.57

Of the commercially available tests with publicly available performance data, two ebolavirus tests listed meet the ‘acceptable’ TPP criteria (ie, sensitivity >95% and specificity >99%) and one meets the ‘desired’ TPP criteria (ie, sensitivity >98% and specificity >99%).40 41 When surveillance testing in centralised settings is considered, without the decentralisation criteria, 10 tests meet the ‘acceptable’ criteria as listed in the TPP, including two LDTs. No POC tests meet the ‘desired’ or ‘acceptable’ criteria listed in the TPP. The RDT did not meet the sensitivity or specificity criteria when used with whole blood and just missed the cut-off for sensitivity when used to test buccal swabs (94%).48 The other POC tests did not meet the sensitivity criteria. If rapid, simple tests are needed most during an outbreak, one near-POC test is available, but POC tests that meet the TPP performance criteria are still needed. A number of laboratory-designed RDTs are under development that may meet these TPPs criteria (online supplementary table S1).

In terms of MARV, although a TPP for diagnostic tests to detect MARV does not exist, it may be helpful to extend the ebolavirus TPP criteria to evaluate the MARV tests, considering the viruses are within the same family and have many overlapping characteristics. Unfortunately, none of the MARV tests would be classified as ‘acceptable’ according to the ebolavirus TPP (although performance data are limited) due to the lack of deployable tests in areas with limited infrastructure and resources.

Need for refined testing algorithms

Specifically for MARV, there are limited recommendations from WHO regarding the diagnosis of MVD. The Centers for Disease Control and Prevention recommends that ELISA, PCR or virus isolation be used to diagnose MVD within a few days of symptom onset. The IgG-capture ELISA is recommended when testing patients later in the course of the disease or post-recovery.58 Further refinement of MARV testing algorithms, especially as more sensitive diagnostics tests become available, should be considered.

Use of alternative sample types for diagnosis

The usual clinical samples for testing for EVD and MVD are WB, plasma and serum. However, venepuncture is an invasive procedure with high risk for both the patient and the health worker. Moreover, blood separation into plasma or serum often requires additional equipment (eg, centrifuges) or expensive collection tubes. With the difficulties of blood collection and POC analysis under field conditions, less invasive, simpler to obtain and more stable clinical specimens must be considered for new diagnostic development.

There is evidence that ebolaviruses can also be detected in saliva, breast milk and semen.59–62 In the case of an outbreak in hard-to-reach areas, buccal swab testing to confirm fatality due to filovirus infection is important to inform behaviour in the field (eg, burial method, contact tracing). However, there is only one POC test that can accept oral fluids.38 Increasing the performance of RDTs to be used for screening via oral fluid from living individuals could provide a rapid triage test that could ultimately decrease the risk of infection for healthcare workers and others within the community, clinic and/or healthcare facility.