PDA

View Full Version : Possible Ebola Vaccine Breakthrough



Swift
2015-Jul-31, 07:12 PM
From the BBC (http://www.bbc.com/news/health-33733711#end)


A vaccine against the deadly Ebola virus has led to 100% protection and could transform the way Ebola is tackled, preliminary results suggest.

...

This trial centred on the VSV-EBOV vaccine, which was started by the Public Health Agency of Canada and then developed by the pharmaceutical company Merck.

It combined a fragment of the Ebola virus with another safer virus in order to train the immune system to beat Ebola.

A unique clinical trial took place in Guinea. When a patient was discovered, their friends, neighbours and family were vaccinated to create a "protective ring" of immunity.

...

There is caution as the results are still preliminary, with more data coming in.

But officials at the WHO believe the effectiveness of the vaccine will end up being between 75% and 100%.

...

One hundred patients were identified in the trial between April and July and then close contacts were either vaccinated immediately, or three weeks later.

In the 2,014 close contacts who were vaccinated immediately there were no subsequent cases of Ebola.

In those vaccinated later there were 16 cases, according to the results published in the Lancet medical journal.


Here is The Lancet article (http://www.thelancet.com/pb/assets/raw/Lancet/pdfs/S0140673615611175.pdf)

Ken G
2015-Aug-02, 03:06 AM
What will be interesting to see is if the phenomenon we are discussing on another thread will occur here. That phenomenon is, when there is a particularly "headline worthy" result coming from a small set of clinical trials, there is a tendency to select from a larger set than what is actually being recognized as relevant to the efficacy of the treatment, creating a false sense of the statistical significance of the outcome. The way that could work here is, there might be N different vaccines being tested, but only the one with the best results gets reported. If N=20, and the one being reported achieves a 95% confidence level, then that really doesn't mean anything-- but it might be reported that we have 95% confidence the vaccine works. It sounds like the WHO is aware of this problem and is hedging their bets a little, saying that it might turn out the vaccine is only 75% effective or something, but it might be 100% effective. What we decided on that other thread is that the best we can really say is that what matters now is followup confirmation that the vaccine works in additional trials. This particular vaccine is now in the limelight, so it can no longer be said to be culled from a larger more anonymous set, and it is only trials taken from that perspective that can demonstrate a reliable statistical effect in their own right.

JohnD
2015-Aug-02, 09:24 AM
KenG,
Did you read the Lancet article? Swift provided a link.

90 'clusters' of people, all family and contacts of a known Ebola victim, were included. "N" was not 20 but 7651 subjects, half of whom had immediate vaccination, half delayed.
There were 16 cases of Ebola in the delayed group and none in the immediate.

This is a real advance in Ebola prevention, not some media hype.
But it isn't the last word by any means. This is a live vaccine, a vesicular stomatitis virus that infects animals and humans, that has been genetically engineered to produce a glycoprotein that appears on the surface of the Ebola virus. The human immune response to infection with this virus thus includes Ebola. It is a real infection, not merely exposure to a dead viral component, and there were 43 "serious adverse events" among the recipients. No doubt work will continue to make this vaccine safer while maintaining its 100% success rate.

John

demeter
2015-Aug-02, 12:24 PM
KenG,
Did you read the Lancet article? Swift provided a link.

I think perhaps someone other than KenG needs to read something a little more carefully.


90 'clusters' of people, all family and contacts of a known Ebola victim, were included. "N" was not 20 but 7651 subjects, half of whom had immediate vaccination, half delayed.
There were 16 cases of Ebola in the delayed group and none in the immediate.

Here is an excerpt from KenG's post.


The way that could work here is, there might be N different vaccines being tested, but only the one with the best results gets reported. If N=20, and the one being reported achieves a 95% confidence level, then that really doesn't mean anything-- but it might be reported that we have 95% confidence the vaccine works.

As you may note, his "N" is measuring something different than your "N". And his point is valid. 95% confidence means there is only a one in twenty chance that we would see the observed results just due to luck, if the vaccine were actually useless. And if 20 (independent - perhaps this word should have been included in KenG's post) vaccines are being tested, then you would expect to get positive results from one of them, with 95% confidence, even if all 20 are completely useless.

Ken G
2015-Aug-02, 01:26 PM
Yes. I'll admit that 16 vs. 0 cases in cohorts numbering about 3600 sounds pretty good. The WHO expects at least 75% efficacy, but what confidence level is this? If it's 95%, then we need to know how many other vaccines were being tried by all the other nations and pharma companies, before we can believe that 75% minimum. Of course, if the efficacy is lower, they can always make further modifications to improve it, so I'm not saying this isn't a good thing-- merely that statistical interpretations can be slippery.

JohnD
2015-Aug-08, 04:25 PM
Apologies, Ken and thanks, demeter. Your N was the number of vaccines tested, not the number of subjects.
But I think my other comments, that this is a pretty good, in fact spectacularly good, result, to contain a continuing and devastating epidemic.
But not as good when the objective is prevention and the recipients could be a whole population.

John

Ken G
2015-Aug-08, 06:19 PM
I don't dispute this is very promising. I'm just saying that the conclusion from the other thread was, one can never actually assess the significance of a statistical outcome until one has a firm idea of the true size of the sample from which it was drawn. That can be very hard to do, given that null results never make it to the light of day! So the more reliable way to avoid disappointment is to reserve judgement of the efficacy until it is replicated by a study that focuses just on it, when it is no longer cherry-picked from a much larger body of potential outcomes. Until that happens, we can't assess its true significance, but there's no question that replication should be a top priority, and resources should be moved in the direction of reaching fruition of this potential game-changer.

BigDon
2015-Aug-11, 02:56 PM
Will the technique used here be able to produce a vaccine against Ebola's bigger, meaner brother Marburg?