Assessing the Science of Ebola Transmission

As leading Ebola virologist Heinz Feldmann recently said, the discussion of Ebola prevention “should be fact-based; it should be data-based.” Interpretation of those facts and data is often clouded by politics and opinion. But they are not as ambiguous as debates in media have made them seem.

Earlier this week, The Atlantic ran an interview on the current Ebola epidemic with Steven Hatfill, a former employee of the U.S. Army Medical Research Institute of Infectious Disease (USAMRIID) who is famous for being wrongly accused of the 2001 anthrax attacks. Hatfill, presented as a world-leading expert on Ebola virus, gave his perspective on topics ranging from transmission of the virus to preparedness for an outbreak in the United States. Some of Hatfill’s concerns about our preparedness to stamp out outbreaks quickly are sound, but the bulk of the interview, about the virus and the disease it causes, goes well beyond the data.

The misinterpretations and misrepresentations of data that Hatfill espouses can be found in many places online, and especially on TV. In light of this, we can’t be surprised so many Americans really don’t know what to think and who to believe on Ebola. Going point-by-point through an interview is boring, unnecessary, and would make Hatfill the core issue. He isn’t. Misrepresentations of the scientific literature on Ebola aren’t unique to him. But I picked some of the most common and dangerous misrepresentations he upheld and dug into the primary scientific literature.

The most controversial part of the ongoing Ebola discussion has been the question of transmission, namely, whether it spreads by an airborne route. Other people have painfully detailed the differences between airborne, aerosol, and droplet transmission. That’s not my interest, or area of expertise. The question that people have, I think, is whether Ebola spreads like the flu; whether they can get it by breathing the same air as someone who has it. The biggest obstacle to clearly communicating the answer to this question, which is no, is the misinterpretation of a 2012 study in which Ebola virus was transmitted from pigs to monkeys without direct contact.

Hatfill references this study more than once as an example of “animal data [that] shows this can happen,” including in a series of answers culminating with the comment that “to say it's not aerosol transmitted is irresponsible.” Not only is that untrue, but the data he uses to get there isn’t properly interpreted, as has been the case for months in various forms of media.

That 2012 study was conducted by Gary Kobinger’s research group at the Public Health Agency of Canada, one of the best Ebola-research groups in the world today. Some have attacked this study in such a way as to make it sound useless or irrelevant, but that’s not the case. Its findings have simply been applied in the wrong way. The study was done because Reston virus, a filovirus closely related to Ebola virus, has been isolated from pigs in Asia, with evidence that they have spread it to farm workers. Reston virus was made famous by Richard Preston’s sensationalized account of its emergence in The Hot Zone, and does not cause disease in people. Nevertheless, Kobinger’s group wanted to know if pigs could spread Ebola virus through the air, the way they seem to transmit Reston virus in Asia.

Kobinger’s group infected several pigs with Ebola virus and set them loose in a room containing four cynomolgus macaques, a species of Asian monkey. The monkeys got sick, died, and analysis of their respiratory tracts strongly suggested that this was the route of inoculation. We knew monkeys could be infected by aerosol though, and have known it since 1995.

Two critical points are invariably missed when this study is cited to support the notion that airborne transmission of Ebola virus has been demonstrated. The first is that pigs develop a very different course of disease and pathology than primates do following Ebola virus infection. Pigs develop severe lung pathology and significant respiratory distress, which results in them expelling very high concentrations of virus from the nose and mouth. In contrast, primates develop a systemic, hemorrhagic disease in which the lungs do get infected but respiratory symptoms are rare. The second key point is that airborne infection isn’t equivalent to transmission. For airborne infection the virus needs to not only be able to get into the airway, it has to get out when a person or monkey breathes or coughs.

Fortunately, that’s been tested. Kobinger’s group followed up their pig study by specifically looking at whether Ebola virus could be transmitted between primates by an airborne route. That, after all, is what we’re interested in. Most Americans aren’t concerned about coming across an infected pig, but they may be worried about sharing breathing air with a person infected with Ebola virus. They set up this study so monkeys would share the same air but be unable to throw feces or other debris between cages, which could easily confound the results. When they did this, none of the control monkeys were infected, despite the experimentally infected monkeys getting sick and dying.

What it boils down to is not only is there no evidence that Ebola virus spreads between primates by an airborne route, there’s actually evidence it does not. If Ebola virus-infected pigs are found in an outbreak zone, which has never happened, they can easily be culled. Hatfill knows about the subsequent study, and that we have never seen airborne transmission in an outbreak; he mentions both. But he still circles back, relying on the pig study, to say that it’s irresponsible to say Ebola virus isn’t airborne. It’s a careful balance of not ignoring the inconvenient data, but emphasizing that which frightens but is far less relevant. It’s misleading without lying, and it leaves the reader, at best, unsure of what to believe.

In reality, there’s nothing equivocal about the data. No biologist would stand before you and absolutely discount the possibility of nearly anything, but based on the excellent experimental data we actually have, we can conclude that Ebola virus simply isn’t transmitted through the air between primates.

The breakdowns in infection control in Dallas have also been in the news, with a focus on the inadequacy of the CDC’s original guidelines on the use of Personal Protective Equipment when treating Ebola patients. Hatfill offered the opinion that the problem here is that “We've taken a BSL-4 disease, and we're treating it in BSL-3 conditions.” BSL-4 though, means positive pressure suits, like those seen in the movies Outbreak and Contagion. In reality, the gold standard for clinical Ebola PPE, recommended by Doctors Without Borders and now the CDC, is something less than that. This entails full skin coverage with an impermeable gown or suit, use of a respirator to protect the worker during procedures like intubation, double gloves, and show covers. This all sounds a lot like what we wear to work with SARS or MERS, two viruses that require BSL-3 containment and procedures.

The specialized facilities we have to treat diseases like Ebola do reflect this. None of the hospitals that have safely treated Ebola patients, Emory, University of Nebraska, the NIH, or Bellevue have BSL-4 medical suites. What they have are contained-isolation rooms and the rigorous and careful use of BSL-3-like PPE by highly trained staff. The BSL-4 medical suite at USAMRIID wasn’t closed for budgetary reasons, it was closed because the NIH isolation facility used to treat Nina Pham is fully sufficient for treating any USAMRIID researchers exposed to Ebola or other dangerous viruses during lab accidents. Ebola is a scary virus, but making it ten feet tall won’t help us save lives. To safely and successfully treat patients we need excellent training and effective PPE that enables nurses and doctors to do their jobs. Not only is medical treatment in a restrictive BSL-4 suit that much more difficult and cumbersome, the recent experience in non-BSL-4 medical facilities shows us that it’s unnecessary for the safe clinical management of infected people.

Another area of concern, mentioned by Steven Hatfill, Darrell Issa, and many riders of the New York City subway, is the possibility of contracting the virus by brushing up against the skin of an infected person. Hatfill talks about Ebola being shed from the skin, and while it’s true that Ebola virus RNA can be found on the skin of infected people, Hatfill’s explanation of how this occurs and the significance of it is scientifically unsound. He goes into a quasi-technical, but inaccurate, explanation of Langerhans cells and their ability to shed virus through the skin. Langerhans cells are skin-resident dendritic cells, a type of cell that ingests a virus or bacteria and goes rapidly to the lymph node, where it activates T-cells specific for that pathogen. Ebola virus infects Langerhans cells and other so-called antigen presenting cells like macrophages, and its trip to the lymph nodes inside them is one of the ways it facilitates its spread throughout the body.

Langerhans cells can’t shed virus onto the outside of your skin, though. The outer layer of skin consists of dead keratinocytes, skin epithelial cells; Langerhans cells are deeper in the tissue. That outer skin cells are dead is a great defense against viruses, because viruses require a living cell to replicate. A virus that lands on unbroken skin is inert; it can’t enter a dead cell and will quickly be destroyed by enzymes on the skin. Likewise, it can’t enter dead cells from inside the tissue and emerge from those cells onto the surface of the skin. Langerhans cells exist so that if a virus gets past that dead outer layer through a break in the skin, it will be quickly recognized and used to activate the immune system. Infected Langerhans cells can certainly produce virus that infects other cells and tissues. They can’t shed virus onto the surface of your skin. Sweat may be able to do that, but very late in infection. Even then, it’s not clear whether virus found on the skin is actually infectious, or dead particles still full of the viral RNA.

Hatfill is right in at least one area, that we need to fight these outbreaks at their source. Ideally we would stamp them out quickly, like the Democratic Republic of the Congo and Nigeria seem to have done with their Ebola outbreaks in the last few months. Sometimes, like in West Africa, that won’t work, or won’t happen. Should we then give up? Whether there’s one case of Ebola in Africa or 10,000, the only way to stop it here is to stop it there. Yet Hatfill undermines his own insistence that we do this by suggesting that the 4,000 troops we are sending to fight this epidemic are at risk of contracting Ebola from bats themselves.

Evidence that fruit bats harbor Ebola virus has mounted in recent years, with the most prominent study finding evidence of infection in three species of bat. The percentage of bats with evidence of infection varied considerably, making it unclear exactly how prevalent the virus is in the bats. It’s certainly not likely that a given bat carries the Ebola virus, and it’s unlikely many American soldiers will be coming into close contact with any bats at all. More importantly, Hatfill’s comment about bats infecting American soldiers through defecation is at odds with everything we know about how Ebola virus cross-species transmission occurs. It’s been hypothesized that bats might drop infected fruit to the forest floor, infecting humans or other animals that subsequently handle or eat it. Overwhelmingly, though, Ebola outbreaks have been linked to the handling of wild-game carcasses, such as bat, gorilla, duiker, or chimpanzee. The current epidemic began with a 2-year-old child, likely infected by touching an infected bat carcass.

American soldiers are unlikely to engage in any of these high-risk behaviors, nor will they be in areas of high bat density. Transmission through bat bodily fluids can’t be entirely ruled out, but why such a rare, theoretical event should threaten 4,000 soldiers to the point that we should reconsider the deployment is a question with an elusive answer.

Ultimately this Ebola outbreak, as even Hatfill acknowledges, will end—though with many more dead than we would have ever imagined. In the aftermath we will learn more about Ebola virus than we know now and use this to ready ourselves for its next appearance. We know that our current knowledge isn’t perfect, but where the data is strong, as it is in the areas I’ve discussed, we can’t ignore it or make it mean something other than what it does. It leaves the public rightfully unsure as to who to trust, and makes the job of those called to fight this epidemic and keep us safe at home that much more difficult.