Vaccine for hemorrhagic fevers
This article (probably only free for subscribers but NYTimes article here) reports the development of vaccines for both Marburg and Ebola viruses, which are closely related members of the filovirus family.
The researchers used another, minimally pathogenic, virus that generally doesn't infect humnas, called vesicular stomatitis virus. They then used genetic engineering so the vaccine viruses would express a protein from the outer coat of either Marburg or Ebola. They vaccinated 6 cynomolgus monkeys with each vaccine, and then challenged 4 of 6 with high doses of the virus they should be protected against (Ebola for the Ebola immunized and vice versa) and used the other 2 as a control group, challenging them with the virus they were not vaccinated against (Marburg for the Ebola group and vice versa).
All the monkeys tolerated immunization without obvious symptoms and no live vaccine virus was detected in swabs from various monkey surfaces and orifices. Live vaccine virus was detected in the blood of monkeys, which was expected since the idea was that the vaccine virus would cause an infection and thereby generate a vigorous immune response
All 8 monkeys challenged with the wild-type filovirus (Ebola or Marburg) they had been vaccinated with were protected from infection (didn't have virus in the blood, get sick or die), but the four that were challenged with the other virus died with high levels of virus in their blood. While numbers are small (monkeys are near prohibitively expensive to do research on) the results are impressive.
As a further experiment, the investigators challenged the four monkeys who had been vaccinated against Ebola and survived Ebola infection with another Ebola strain. Unfortunately only one of four monkeys survived, indicating the vaccine is pretty specific for the strain chosen. An ideal vaccine would probably express proteins from several Ebola strains and Marburg to give protectin against all of these with one vaccination.
One interesting point about this experimennt is how the vector virus (the one which is used to express the Ebola or Marburgh protein) is chosen. The immune system reacts much more strongly to a true infection that to just injecting proteins into the body. An ideal vector virus would establish infection but not cause significant symptoms or be easily spread to others, all properties possesed by VSV. Additionally, the exposure of the human population to the virus should be low, otherwise many people will have already be infected and vaccination will not lead to productive infection nor a good immune response. This is one problem with adenovirus which is commonly used in similar trials. While minimally pathogenic, VSV infection rates naturally in humans are quite low, at least according to this article.
This vaccine, while still some time from human use, might be particularly useful for vaccinating people in areas of an outbreak, especially those with known exposure and health care workers.