August 7, 2013
The possibility that H7N9 avian influenza may evolve sufficiently to cause a pandemic has scientists turning again to controversial research —they must be careful how they justify the risks taken.
Researchers now want to make genetically engineered versions of H7N9 that are more transmissible and pathogenic in mammals. In a Correspondence published jointly this week in Nature and Science (see page 150), 22 scientists, including Ron Fouchier of the Erasmus Medical Center in Rotterdam, the Netherlands, and Yoshihiro Kawaoka of the University of Wisconsin-Madison, argue that such research can help to assess the ‘pandemic potential’ of H7N9. The dilemma is that should such engineered strains be accidentally or deliberately released from a lab, they could spark a flu pandemic.
The announcement is likely to prompt some replay of last year’s debate over the creation by Fouchier and Kawaoka of lab strains of H5N1 that could transmit between ferrets. And it offers the first test of some of the review and oversight structures put in place for this ‘gain-of-function’ flu research. As this journal has said before, scientists who push for such research should be wary of over-selling the benefits to public health, at least in the short term, as a way to justify the risks taken.
A sense of perspective is crucial here. The long-term benefits of such work are clear — as long as it is done to the highest biosafety standards. It will shed light on, for example, the mechanisms of virus transmissibility and pathogenicity. But the immediate benefits to public health and our short-term ability to counter the threat of H7N9 are less clear-cut. Scientists cannot predict pandemics, so to assess the pandemic potential of viruses — and to decide which strains warrant the manufacture of trial vaccines — comes down to judgements of relative risk.
Tests of how flu viruses behave in animal models such as ferrets can certainly provide information on the risk of transmissibility and pathogenicity, although it can be difficult to extrapolate those results to humans. A rash of papers this year has shown that H7N9 does have limited airborne transmissibility in ferrets, although the virus is not transmitting between people in the current outbreak in China.
Another way to assess pandemic potential is to monitor wild-type viruses for mutations that allow the virus more ready access to human cells. H7N9 has already acquired some of these mutations, which is why it infects humans more easily than does H5N1. But as researchers pointed out in June, there is no scientific evidence that such mutations predict the risk of a pandemic (D. M. Morens et al. N. Engl. J. Med. 368, 2345–2348; 2013). Transmissibility is more complex than that.
In creating mammalian- transmissible versions of H7N9, scientists would go a step further and hope to identify combinations of mutations that could increase virus transmissibility in ferrets or other models. Such work could yield information on the biological principles affecting transmission. But nature could well come up with combinations for transmission that are different from those obtained in experiments.
Following the H5N1 controversy, the US Department of Health and Human Services has introduced an extra layer of review that will apply to anyone seeking funding for work to make mammalian-transmissible strains of H7N9 (see page 151). The risks and benefits of the work will be assessed by a panel of experts in public health, security, risk assessment, law and ethics, and, importantly, any extra steps needed to mitigate biosafety risks will be considered. The way the review handles H7N9 will be an important test of the effectiveness and transparency of this new approach.