Opinion | We Could Easily Make Risky Virological Research Safer

Lipsitch was among the loudest voices raising concerns about so-called gain of function research (defined in different ways by different groups, it generally refers to work manipulating pathogens to make them either more transmissible or more virulent) before the federal government established a moratorium in 2014. That year he and Tom Inglesby published estimates that “a laboratory-year of experimentation on virulent, transmissible influenza virus” was introducing risk equivalent to 2,000 to 1.6 million deaths during that period.

Lipsitch had been relatively quiet during the pandemic as heated lab-leak debate crowded out more measured conversations about biosafety and research oversight. And in Geneva his perspective was less alarmist than measured and technocratic. He said that even a vast majority of experiments described as gain of function are fairly safe, provide obvious benefits and need no additional review. Regulatory focus, he said, should be on a much narrower category that he called “gain of function research of concern.” The way he sees it, “over 90 percent of work in virology is not even in this category.” And while many virologists have recently argued that further oversight might hamper vaccine development, it would be very easy to avoid that problem, he said, for one simple reason: “The concerning experiments are not the ones that make vaccines.”

As for what experiments are concerning, the science advisory board’s recommendations, Lipsitch went on, offered a pretty good standard — broadening the category of pathogens requiring special scrutiny from those judged to be both likely highly virulent and highly transmissible, as was the standard set in 2017, to include those considered likely moderately virulent or moderately transmissible and likely to pose a severe threat to public health or national security. Research involving manipulation of the novel coronavirus, SARS-CoV-2, for instance, would not require additional review under the old framework but would under the new definitions. He said he’d prefer an even more expansive definition, including pathogens exhibiting low virulence and transmissibility, but on paper, the science advisory board recommendations offered what he called a pretty good definitional framework.

Implementation matters, too, and may be cause for concern. Since the 2017 standards were instituted, Lipsitch said, only three projects have come under the expanded review. Over the same period, he’s learned of at least five experiments he believed should have been reviewed but weren’t — meaning that just the experiments he happened to know about that slipped through without review outnumber the experiments processed by the relevant ethical authorities.

Those were very few experiments, over half a decade. Almost surely, there are more. But the small number suggests that regulation need not be all that difficult, bureaucratically cumbersome or even expensive and that a lot of additional safety could be achieved with relatively little additional review.