Fighting back against flu

Why do we have a different flu vaccine every year?
25 September 2018

Interview with 

Derek Smith, Cambridge University


Vaccine injection


Every year, human strains of the flu circulate around the world. And as they go they subtly change their appearance. This means we need to update the vaccines we make so the immune system can still recognise them. So how do scientists know when and what to put into these vaccines? Chris Smith spoke with Derek Smith who advises the World Health Organisation on this topic; he’s based in the Zoology Department at the University of Cambridge...

Derek - It’s really remarkable global effort orchestrated by the World Health Organisation to do this. There are thousands of people around the world, in GP offices and in hospitals who see people who come in with respiratory diseases that look like they could be flu, take a throat swab and that swab is sent to the National Influenza Centre in that country.

The strains then are analysed to see if they really are flu. And if they are, they’re sent to one of five international WHO collaborating centres; Tokyo, one in Melbourne Australia, Beijing China, Atlanta in the US, and in the UK in London at the Crick Institute. There the strains are analysed in great detail in terms of how they differ from other viruses and how our immune systems will see them different.

Chris - When you say you’re analysing them, what are you looking for? How do you tell one strain of flu apart from another?

Derek - This is the key question. First of all it is laboratory work to test whether or not strains of flu that are already in the vaccine will be protective against the new strains of flu that may be emerging across the world.

Chris - I see. So I send you let’s say a sample of flu that I got from Joe Bloggs and you’re asking does the vaccine produce antibodies in someone I give it to that if they met that virus tomorrow it would stop it?

Derek - That’s exactly right. And as well as this happening locally in over 130 countries worldwide, these five international centres coordinate this and do all the laboratory work. And then collaborate with our laboratory here at the University of Cambridge where we also look at them with mathematical and computational methods to track the evolution of these viruses for the purpose of keeping that vaccine strain updated.

Chris - On a practical level, how do you actually do those experiments to know if the vaccine is going to defend me against that particular strain I got from Joe Bloggs last week? Is that somebody physically, with a test tube and some virus and some cells, growing these things and then proving yes, the immune response will stop Chris Smith from catching this virus?

Derek - That’s absolutely right. It’s 20 thousand viruses a year tested in exactly that way to see whether or not those strains can protect Chris Smith this year.

Chris - So it’s quite a lot of guesswork then because you’re getting samples that are doing the rounds now, but this is going to inform the vaccines you’re going to make into the future?

Derek - Yeah. This is absolutely the key thing. And in some ways it’s absolutely not guesswork because this is a very well oiled machine and comprehensive surveillance and very good analyses. But, the point that you raise, the virus has six to eight months to go do its own thing in that intervening time and may evolve such that the strain that’s in the vaccine is no longer a perfect match.

The four major types of flu that circulate there’s a strain of flu for each of those in the vaccine. And when there is one of these mismatches it’s typically just for one of those types of flu. That one component of the vaccine, that one fourth of the vaccine protecting against one fourth of the strains that are circulating doesn’t do as well as we’d all like it to do, and people are not as well protected against getting infected.

Even in those cases they are protected against the other three strains of flu - main strains that circulate. There are also typically protected against sever disease from that other strain. They may still get a cold, or something that feels like a cold; they’re less likely to die or end up in hospital.

Chris - Can you use say maths or other techniques to try and anticipate what the next move might be on the part of the virus, perhaps informed by what’s happened in the past, to make your guesswork odds a bit better?

Derek - Absolutely. And this is a research programme that we have been doing for something like 15 years now to see if we can understand what are the deep evolutionary processes that are going on. What are the constraints or are there constraints on how the virus might evolve?

And it turns out that there are constraints to the extent that we think that we can predict this. And there is a new generation of flu vaccines that are being produced where the strain that’s in the vaccine is not the best representative of the strain that’s circulating in February, but is actually an educated guess of what’s going to circulate the following year. These are vaccines that will enter clinical trials in about two years from now.

Chris - It’s a bit like when you’re driving down the motorway, you should always look at the car not directly in front but one in front of the car in front, because you see the brake lights go on on that one before the car in front of you is going to brake and so it give you advance warning. You’re sort of saying well if I look at what the virus is doing now and then I second guess where it’s going to be later I’ll get a much more accurate picture?

Derek - This is exactly what’s happening. And for me it’s a really beautiful integration of basic science, evolutionary biology, and fantastic surveillance because the other thing that we have when we drive down the motorway is we have the experience of doing this before. And because there is this great surveillance over so many years one can go back and do retrospective studies to imagine that it is 1989, and then see if we can predict what happens in 1990 in 1991/92 and know whether or not the methods are working or not.


Add a comment