Ebola: What’s in the toolbox?

For the almost 40 years that preceded the most recent epidemics, Ebola virus (EBOV) had only caused limited outbreaks of up to 400 cases. However, by the time the 2013 epidemic was over in 2016, the virus had caused 28,616 infections and over 11,300 deaths. A new outbreak currently on-going in the Democratic Republic of Congo reminds us that this virus is constantly and frequently re-emerging from the wild, most likely from the bat population, and despite some very encouraging breakthroughs, there are no approved vaccines or antiviral therapies for Ebola.

The difficulties of designing anti-EBOV strategies are due not only to its complex and extremely severe pathology but also to the difficulties of studying a highly lethal virus for which animal models are limited. Nevertheless, a lot can be learned from the scientific community’s attempt to design and implement anti-EBOV strategies. These efforts are very exhaustively summarised by Cross and colleagues, who comprehensively examine post-exposure treatments to the virus.  What emerges is a large variety of tools and the difficulty of proving their effectiveness in humans.

Ebola and Marburg

The Filoviridae family is composed of the Marburg virus (MARV) and Ebolavirus. The Ebolavirus genus, in turn, is composed of five species: Zaire, Sudan, Tai Forest, Bundibugyo, and Reston. Zaire and Sudan are the most pathogenic in humans, while Reston is not known to have infected humans. Besides infecting humans, EBOV is also pathogenic for great apes, further decimating this already endangered species.There is no immune cross-protection between Marburg and Ebola, and only limited cross-protection between Ebola species, further complicating the development of a vaccine.

1. Antibodies and blood products

While convalescent blood has been used since the first EBOV outbreaks, its effectiveness is very hard to demonstrate, and studies in both non-human primates (NHP) and humans did not show substantial evidence of protection. The risk of host rejection also complicates the applicability of this treatment.

Some monoclonal and purified antibodies have also been tested, with mixed results. KZ52, purified from an EBOV survivor and targeting the viral glycoprotein GP, although strongly neutralizing in vitro, had no impact on disease or viral load. The rise of escape mutants able to evade neutralization is likely to be one of the main reasons for this difference, compromising the validity of this strategy.

Polyclonal IgGs from the convalescent serum of NHPs vaccinated with experimental vaccines, although less neutralizing that KZ52 in vitro, showed better protection than KZ52 in NHPs, possibly as polyclonal antibodies are less likely to select for escape mutants.

The most famous monoclonals, brought into the spotlight during the 2013–2016 epidemics, are ZMab and ZMapp. ZMab is a pool of three mouse monoclonals (1H3, 2G4, and 4G7) which showed 50–100% protection in NHPs. ZMapp comes from the combination of humanized ZMab, where 1H3 was replaced by 13C6, which in turn came from the MB003 humanized cocktail. ZMapp was produced in tobacco plants at the time of the 2013 epidemics and was shown to be 100% protective in NHPs, even when administered at day 5. ZMapp was used in 2014 on compassionate bases on repatriated healthcare personnel; however, as all patients underwent a complex and aggressive supportive therapy, it is hard to confirm that survival was actually due to ZMapp administration. In 2015, a randomized trial of ZMapp + standard of care vs. standard of care reported 22% deaths in the first case and 37% in the second; this difference is not large enough to make conclusions about the effectiveness of the therapy, particularly as the treatment was administered towards the end of the epidemics, and to individuals that died before receiving full dosage.

2. siRNA and phosphorodiamidate morpholino oligos

Chemically modified siRNAs that do not activate an immune response have been designed against Ebola proteins L, VP24, and VP35; proteins important for viral replication, but also for the virus’ ability to evade the innate antiviral response. Administered in various formulations of novel lipid nanoparticles (NLN), repeated doses of siRNA showed protection in NHPs, but again, results in humans are unclear, mainly due to a variety of technical problems in the studies. Similar levels of protection are seen in NHPs for morpholino oligos, suggesting that these approaches have potential. However, multiple dosages seem to be necessary, generally via intravenous administration.

3. Small molecules

No specific drugs against EBOV have yet been approved, and most of the small molecules tested are broad-spectrum nucleoside analogs that work by blocking RNA synthesis by the viral polymerase.

Favipiravir is a pyrazine derivative that blocks RNA synthesis. No NHP studies are available, and the few cases of administrations to patients are inconclusive due to the simultaneous implementation of other treatments. Comparison with historical controls does not support a convincing antiviral effect, and the limited levels of protection observed seem to be inversely correlated to viral load. The ability of the compound to reach sufficient concentrations in the blood or the affected organs remains unclear.

GS-5734 is a monophosphoraminidate prodrug of an adenosine analog. Intravenous administration for 12 days in NHPs resulted in 100% protection when starting before day 3 post-infection. This drug is currently used in the PREVAIL trial in South Africa, in men with persistent viral RNA in their semen.

Other compounds (BCX4430, Brancidovir, and Amiodarone) have been tested/suggested based primarily on in vitro studies, but their activity in human remains dubious.

4. Post-exposure vaccines

The recombinant vesicular stomatitis virus expressing EBOV GP (rVSV-EBOV) is the only successful phase III vaccine to show statistically significant protection (= no EBOV cases amongst vaccinated individuals) in a ring-vaccination trial in Guinea. Unfortunately, having to rely on the onset of a vaccine-specific immune response, the same vaccine is only effective post-exposure in NHPs when administered no later than day 1–2 post-infection. Limited protection was also observed in a few cases of accidental laboratory exposure or repatriated medical personnel in the 2013 epidemics: all patients recovered, but whether they were exposed is not certain.An adenovirus-based vaccine (Ad5-EBOV-GP) has also shown some promising results in NHPs.

5. Modulation of the host response

Given that most of Ebola’s symptoms and pathology are due to abnormal immune responses and coagulation disorders (rather than to the virus itself), most interventions on the host reaction to the virus have focused on regulating these responses.

Attempts have been made at modulating immunity, either by activating an interferon (IFN) response (as some EBOV proteins are dedicated to evading IFN synthesis and signaling) or suppressing the cytokine storm that is associated with high viral titers and is responsible for hypovolemic shock and organ failure. Unfortunately, IFN administration has not shown significant benefit and might only work in combination with more directed therapies, while the effectiveness of the opposite approach (i.e., reducing the inflammatory response) also lacks evidence in NHPs and humans. However, administration of antibodies that block the IFN-I receptor in rhesus monkeys, which were then challenged with EBOV, resulted in faster death than controls, suggesting that suppressing immunity is likely to boost viral load. It is possible that more information on the correct balance between immunity and inflammation might be needed before we can fine-tune a host response, and we suggest that timing may be an important factor as well.

Anticoagulant proteins administered to NHPs challenged with MARV or EBOV showed very limited protection, while no significant data is available on the use of heparin in humans, although this has been tried in only a few individuals. These efforts suggest that this strategy might also be more suited to a combinatorial approach.

Outlook

While we might not be too close to a post-exposure treatment against Ebola, the number of approaches tested and the information available from studies in relevant primate models is impressive and extremely informative.

Monoclonal antibodies and cocktails of monoclonals are one of the most promising strategies, although small changes in the viral GP can be problematic and the antibodies themselves might induce the evolution of escape mutants.

siRNA and other silencing approaches also appear very encouraging in NHPs, and have the advantage that they can be easily modified to neutralize different strains and mutants. However, these require multiple administrations, and this would likely be difficult to implement in the field.

Small molecules should be the traditional post-exposure treatment. However, so far, their success is limited, possibly also due to the need to carefully define and improve their pharmacokinetics to ensure that sufficient levels of the active form reach the sites of viral replication.

Vaccines are showing great promise when used prophylactically, but their therapeutic usage still relies on the development of immune response, meaning that they can only be effective when administered very early after infection, which is something difficult to define.

Finally, how to correctly modulate the host response to infection requires further study, but is also likely to depend on several factors that might be difficult to control, like the viral dose, strain, and the characteristics of each patient. However, in combination with other therapies, these strategies might hold promise, as a synergistic approach may be required to tackle a virus that quickly replicates to high titers, mostly shielded from the host immune response.

Strikingly, many studies that show promise in NHP did not seem to be equally effective in humans, highlighting the fact that even highly relevant animal models challenged and treated in a laboratory setting remain very different from humans infected in the field, particularly for a virus like EBOV where timing is crucial. Also, for primate studies, it is difficult to achieve sufficiently large numbers to make conclusions robust, particularly for a BSL4 virus.

Studies in the field are also extremely complicated due to the lethality of the disease, which raises a meaningful ethical objection to, for instance, having untreated control groups. Equally, the small available doses of experimental treatments administered to repatriated medical personnel have been used in combination with extensive supportive care, and in some cases various drugs and treatments have been applied at once: the vast majority of repatriated nurses, doctors, and volunteers have survived the infection, but what was this due to?

Another interesting aspect of EBOV infection that has emerged since the 2013 epidemic is the persistence of the virus in some recovered patients, even years after the infection. It has become clear that in some people EBOV can replicate in immune privileged sites and also cause clinical complications, such as encephalitis, ocular problems, and muscoloskeletrical pain. The implications of this prolonged replication on the patient and on the virus itself are not entirely clear. In the meantime, it is important that the scientific community retains its focus on this devastating virus, as we know for certain that it will continue to re-emerge.

While we do not work with full EBOV, we have viral tools and systems that can be used for cell-based screenings. If your projects involve vaccines or antiviral development for BSL2 and BSL3 pathogens, from target identification to compound testing, contact us and find out how we can help!