Potential new treatments

At the time of writing, the treatment of Ebola is based on supportive care and symptom relief. Experimental therapies are in trials and the position is changing fast. In the video Professor Peter Smith discusses the rationale and plans for testing convalescent plasma as a therapy. In this step we consider the issues and constraints in evaluating this and other new treatments. Early work in new therapies and vaccines for Ebola predated the 2014 outbreak and was done in response to a perceived threat of Ebola as a biological weapon. No new drug or vaccine candidates were taken as far as human trials. With the 2014 outbreak the need to accelerate production and testing of these candidates was realised. In a statement, an ethics panel at the World Health Organization (WHO) agreed that “In the particular circumstances of this outbreak, and provided certain conditions are met, the panel reached consensus that it is ethical to offer unproven interventions with as yet unknown efficacy and adverse effects, as potential treatment or prevention.’¹ The need for improvements in supportive therapy was discussed in the ‘Research priorities’ step. This step concentrates on new therapies.

Possible therapies

There are broadly three categories of therapy: 1) Convalescent blood or plasma 2) Antiviral drugs that are already available 3) New products

Testing new treatments

The usual process in testing potential new treatments is as follows:²

1. Animal testing. Consideration of the results and formal approval is required before moving to human testing.
2. Testing in (typically 20-80) healthy human volunteers (Phase 1). This is to check for safety, measure side effects and refine the dosage.
3. Preliminary testing in individuals with the disease (Phase 2). This is usually done as a randomised controlled trial and gives early, preliminary data on effectiveness and information on safety and side-effects in those with the disease. The numbers of individuals in these trials are usually too small to be sure about effectiveness, although ‘Phase 2B’ trials try to achieve this.
4. If Phase 2 studies suggest the new treatment may be effective, Phase 3 trials are conducted. These should include large enough numbers of individuals to get a clear answer on the drug’s effectiveness compared to previous treatments (or a placebo if there are no previous treatments). Safety and side-effects are closely monitored. Results from Phase 3 trials are usually needed for applications for formal approval of the treatment.
5. After licensing it is important to continue to monitor the effectiveness and long-term side-effects on larger numbers of patients (Phase 4).

The full process from animal testing to licensing and wide-spread use usually takes many years. There are fast track processes for speeding up the time taken for formal approval of the different stages. Also, for life-threatening conditions there are precedents for granting provisional approval based on less stringent evidence. The process will be faster if starting with treatments already used for other conditions, as safety data will be available. In these situations it will be possible to bypass Phase 1. For Ebola, expedited evaluation and licensing is planned. For Ebola there has been very small-scale unlicensed use of experimental therapies in patients treated in Europe and the US. Trials are needed to know if these therapies are actually effective.

Practical issues when planning trials for the Ebola outbreak include:

• There are few Ebola treatment sites with adequate facilities to support a clinical trial.
• Trials require data collection but data collection is difficult in the high risk areas because of the limited time and difficulties of recording information while wearing personal protective equipment, and because it is difficult to transfer data out of these areas. This further limits the number of suitable sites.
• The number of potential therapies for Ebola is larger than the number of suitable trial sites, so priorities for testing need to be agreed centrally. It is important to keep a database of negative results to counter false claims of efficacy.
• The baseline standard of care and the case fatality rates vary between and within treatment units. For non-randomised trials, sites with uniform and stable baseline standards of care would be needed.
• The case numbers vary drastically over time and place, so it is difficult to predict which units will have enough patients for a trial but not be completely overburdened and unable to conduct a trial safely.
• It is essential that trials do not interfere with routine care. However, in other settings it is often found that the protocols required by trials actually improve routine care.

Trial designs are discussed in the next step. Key issues include:

• ‘compassionate’ use vs product licensure.
• rapid access vs data to demonstrate safety and efficacy.
• randomised controlled trials vs alternative trial designs.
• maintaining public trust.

First trials of therapies for Ebola

The first trials are with convalescent plasma, and two existing antiviral drugs, favipivavir and brincidofovir, all in collaboration with Médecins Sans Frontières (MSF)-run Ebola Treatment Units.³ As discussed in the video, there are good theoretical reasons to believe that convalescent plasma could be helpful, and very limited evidence from earlier outbreaks. Given the scale of the current outbreak there are many survivors who can donate plasma, and they have proven willing. Plasma donation may also help to reduce the stigma that surrounds survivors. The first non-randomised trial started in Conakry, Guinea, led by the Antwerp Institute of Tropical Medicine.³ Favipiravir has been approved in Japan against influenza (in specific circumstances), has been used in 1000 people and has a good safety profile (although it should not be used in pregnant women because of a risk of birth defects). It was effective in mouse studies of Ebola but not in non-human primates. Large numbers of doses are available. It is being tested in Guéckédou, Guinea, led by the French biomedical research agency, INSERM.³ Preliminary results were presented at the CROI conference in February 2015. These were inconclusive: patients with high viral loads continued to have high mortality. Those with lower viral loads had lower mortality than measured previously, but this was not a randomised comparison and it is not clear if the difference is due to the drug. Brincidofovir was developed as a broad spectrum anti-viral drug. It has been used to treat over 1000 patients with viral diseases, including studies for the treatment of cytomegalovirus and adenovirus. It was being tested in Liberia, in a study led by the University of Oxford.³ But the trial was stopped in February 2015 after recruiting only a few patients, when the incidence of Ebola had declined in Liberia: following discussions with the US Food and Drug Administration (FDA), the pharmaceutical company Chimerix, which manufactures brincidofovir, unexpectedly announced that it would cease to participate in any current or future trials of brincidofovir for Ebola.

Other existing therapies:

Interferon is licensed for treatment of hepatitis and multiple sclerosis and is available. It caused a delayed time to death in non-human primates with Ebola, but had no impact on the case fatality rate. Side effects occur and the benefit to risk ratio is unknown.

New therapies

Monoclonal antibodies (including ZMapp, made by MappBio, USA,which was used in 7 Ebola patients early in the epidemic) and other potential new therapies are all only available in very small quantities. ZMapp production is being scaled up, and NIH announced plans for a trial of ZMapp in February 2015. TKM-Ebola, made by Tekmira Pharmaceuticals, Canada consists of small interferring RNA which targets genes in Ebola RNA once inside the cell, blocking replication. It is sequence-specific to the current strain of Ebola. A phase II trial started in Sierra Leone in March. It is not placebo controlled, but in practice it is unlikely that all the patients in the treatment centre will receive it, as it is complex to give: it requires slow intravenous infusion each day for a week, and close monitoring of the patients afterwards. There are some candidate therapeutics that have been through Phase 1 trials, while others are at an earlier stage. The latest information is available on the WHO website. As WHO states, ‘[..] [I]t should be noted that the potential compassionate use and further investigation of these compounds should not detract attention from the implementation of effective clinical care, rigorous standards of practice in infection prevention and control… careful contact tracing and follow-up, effective risk communication, and social mobilization, which will be crucial to terminate the epidemic.’⁴