December 16, 2016

expert reaction to modelling study on universal and conventional flu vaccines

Publishing in PLOS Computational Biology, researchers use mathematical modelling to predict the possible effects of a universal flu vaccine.

Prof. Peter Openshaw, British Society for Immunology President and Professor of Experimental Medicine at Imperial College London, said:

“This is an interesting theoretical paper using mathematical modelling to predict the possible effects of influenza vaccines that have not yet been developed on the circulation and evolution of flu viruses. Immunologists are working on vaccines that don’t need to be reformulated each year: ‘universal vaccines’ that induce broad immunity, protecting against current and future strains of flu by mechanisms that are not just dependent on antibody.

“The authors reach the interesting conclusion that it might be possible for vaccines with certain profiles to affect the transmission and evolution of influenza viruses and that, theoretically, universal vaccines could control influenza epidemics while slowing the rate of viral evolution. By contrast, current vaccines that need to be changed each year could actually promote the speed of viral evolution.

“However, it’s important to say that the authors make a lot of assumptions. One is that there is no interaction between antibody-mediated immunity and T cell immunity (which we know is an oversimplification), and that the future universal vaccines do not actually protect against infection (this does seem a rather radical assumption). They also assume that current strain-matched vaccines illicit no cross protective immunity.

“So, this is an interesting but theoretical analysis based on speculation on what future influenza vaccines might do. We will need to see what does happen in practice when universal flu vaccines do become available.”

Prof. John Edmunds, Professor of Infectious Disease Modelling, London School of Hygiene & Tropical Medicine, said:

“This is a helpful study that shows how a universal flu vaccine could help control both seasonal and pandemic influenza.  However, ‘could’ is a key word here, as this is a purely theoretical exercise and we are still some way off having a universal flu vaccine. Meanwhile, the seasonal vaccine is by far the best way to protect yourself against flu.”

Dr Philip Minor, Head of Virology, National Institute for Biological Standards and Control (NIBSC), said:

“The people involved in this paper are very good.  Currently universal vaccines do not exist in spite of their attraction, so this study is purely theoretical.  It’s also worth considering what such a potential universal vaccine might be – the paper suggests they might offer the potential to act across all influenza A strains and serotypes.  However possible vaccines based on the HA stalk, which are mentioned, might not be quite so comprehensive.  The HA (hemagglutinin) is a lollipop-like structures in flu viruses – the stalk is the stick and does not vary much because it is a bit buried and the blob on the top is the bit that varies and that gets most antibody attention because it is most exposed.  So antibodies against the stalk (if you could make them) would work on different strains of virus.  However, there are two different classes of stalk structure (call them phylogroup 1 and phylogroup 2) and it would be unreasonable to expect to find a group 1 vaccine that works as well against group 2.  It might therefore theoretically be expected that a group 1 vaccine (if it existed) would mean a group 1 flu would not be able to infect as much, which would leave the field open to group 2 flu and could possibly drive the emergence of a group 2 pandemic.  A universal vaccine will only do the right thing if it really is a truly universal vaccine, and that is a tall order.

“The model compares the effect of long-lived specific immunity to hypothetical short-lived universal immunity, and the result that the universal version would be expected to be best by apparently all criteria is really interesting and not particularly obvious to me.  However flu is complicated and the devil is in the detail.

“This is an interesting paper.  The authors quote a number of possible ways to make a universal vaccine but no one is there yet.  Not even close, in my opinion.”

Prof. Keith Neal, Emeritus Professor in the Epidemiology of Infectious Diseases, University of Nottingham, said:

“This is a modelling paper that adds very little to what we know.  A flu vaccine that lasted many years to lifelong immunity would be a big advance; it’s just that we don’t have one at the moment.  This paper is merely calculating the benefit of such a potential vaccine over and above the current vaccine, so it is purely theoretical.”

Prof. Sarah Gilbert, Professor of Vaccinology, University of Oxford, said:

“This theoretical paper by Subramanian and others uses a mathematical model to predict the effect of future influenza vaccination, using either a strain-specific or cross-reactive (or ‘universal’) vaccine, which currently does not exist.  As vaccine recipients, we usually think only of the ability of the vaccine to protect us against a specific disease, but here the effects on the whole population and on the influenza virus itself are being considered.

“The first suggestion the authors make is that greatly increased population coverage with the strain-specific vaccines that we now use may serve to drive antigenic evolution of the influenza virus, making it even more difficult than it is now to produce a protective vaccine as the viruses would evolve even more quickly.  In contrast, the widespread use of a universal vaccine not only would not require the vaccine to ‘keep up’, it would also make it less likely that seasonal influenza viruses would change.  This means that looking at the population as a whole, there would be more benefits from using a universal vaccine than a strain-specific vaccine.  A high level of cross-reactive immunity in the population could even reduce the likelihood of a new pandemic occurring.  The problem is that a universal vaccine does not yet exist!

“One of the parameters that is required to set up the mathematical model is the duration of cross-reactive immunity, which is not well understood.  We know the cross-reactive immunity that is produced when an individual recovers from an influenza virus infection is not life-long, and probably only lasts for a small number of years.  However we should not assume that the duration of cross-reactive immunity induced by universal vaccines will be the same – it could last for considerably longer.  The authors accept that the situation they are modelling is extremely complex, but perhaps the main conclusion should be that when assessing new influenza vaccines that are now in development, we should not only consider the efficacy of the vaccine in protecting the person receiving the vaccine, but need to take into account the effects on the whole population as well.  Breadth of immunity – the ability to protect against many different versions of the influenza virus – may turn out to be what matters most.  Of course, no universal influenza vaccines are yet available, but this mathematical model strongly supports the continued attempts to develop such vaccines.

* ‘Universal or specific? A modeling-based comparison of broad-spectrum influenza vaccines against conventional, strain-matched vaccines’ by Rahul Subramanian et al. published in PLOS Computational Biology on Thursday 15 December 2016.

Declared interests

Prof. Peter Openshaw: “Prof Peter Openshaw’s research is funded by the Wellcome Trust, the MRC, BBSRC and the European Union. He has received honoraria or consultancy fees from GSK, Janssen, and Mucosis BV.”

Prof. John Edmunds: “I have no conflicts of interest to declare.”

Dr Philip Minor: “I have no conflicts.”

Prof Keith Neal: “No conflicts.”

Prof Sarah Gilbert: “I am scientific co-founder of a company that is developing a broadly cross-reactive influenza vaccine.”