Diseases carried by mosquitoes kill hundreds of thousands of people annually. (Philippe Huguen/AFP/Getty Images)
Henry T. Greely is director of the center for law and the biosciences at Stanford University. He is the author of "The End of Sex and the Future of Human Reproduction."

In 2018, about 440,000 people worldwide, mainly children younger than 5, died of malaria — nearly one every minute. Another 210 million people survived but suffered from one or two weeks of high fever, shaking chills and other symptoms. The toll in human suffering is immense but relatively hidden from American eyes because the vast majority of cases occur in sub-Saharan Africa.

But biotech holds out the potential of progress — possibly soon. People don’t give people malaria: The 460 species of mosquitoes in the genus Anopheles do, and researchers are taking aim at them. One approach uses genetic engineering to reduce, or even eliminate, populations of malaria-carrying mosquitoes — for example, by modifying males to ensure that all their offspring are infertile. Another more elegant tactic uses genetic modifications to make mosquitoes immune to the malaria plasmodium, the parasite that causes the disease. Either kind of change can be amplified by doing the genetic engineering along with a trick called “gene drive,” which speeds the spread of such changes through a population.

The stakes of deploying or not deploying such research are high: From the moment we develop such technologies, “every day we wait kills between 1,200 and 2,000 people,” notes Vox’s Dylan Matthews.

Can we in good conscience hold off from using these technologies as soon as possible? Yes, we can. In spite of the stakes — in some ways because of the stakes — it is important that we get this right before we try it. We have a moral imperative to fight malaria, but we also need a better understanding of the environmental, medical and political risks before we rush in. And we need to regulate the relevant experiments and field tests, not just to avoid unanticipated ecological side effects but also to forestall political resistance.

We’re at an extraordinarily sensitive moment in this debate. In November, the Conference of the Parties to the Convention on Biological Diversity met in Egypt; in the run-up to the meeting, a movement for a full ban on gene drives gained momentum. In the end, a bloc of African governments, concerned about their citizens’ health, derailed it, but opposition may grow again if a gene-drive experiment goes wrong or is perceived to have gone wrong. (The 196 parties to the convention do not include the United States, which signed on to it in 1993 but, due to conservative opposition, has never ratified it.)

At a very basic level, it is not clear whether or how well these new methods will work. Only one approach, pursued by a company called Oxitec, has had field tests (in Brazil, Panama and the Cayman Islands). Oxitec modified males of a non-malaria-carrying but still dangerous mosquito species, Aedes aegypti (a carrier of dengue, zika and yellow fever, among other diseases), so that their offspring would be infertile. The firm claims that the species’s populations in the test areas dropped by 80 percent, but more work is needed. One strong and enduring lesson of the past four decades of biotechnology is how often great ideas fail even after good initial results. Oxitec recently announced an agreement with the Gates Foundation to create an Anopheles mosquito using its genetic modification.

But once we know genetic modification of mosquitoes will work against malaria, should the techniques be rolled out quickly? Lives would be saved almost instantly, and unlike drugs given to humans, the side effects of modified mosquitoes won’t include inadvertently making people sicker. But other risks demand attention — especially ecological hazards.

In the 1950s and 1960s, an extremely effective pesticide called DDT wiped out malaria in some parts of the world, including the United States. But DDT wreaked havoc on some non-mosquito species, including eagles and pelicans, and is thought to have harmed human reproductive health and to have probably caused cancer. (High levels were clearly bad, but the effects of chronic low-level exposure are far more uncertain.) As a result, its use was greatly curtailed, a decision that remains controversial because it allowed malaria to persist in some places.

Anopheles mosquitoes pollinate some plants and provide food for some animals — how would plunging numbers of some of those species affect ecosystems? Even a drop in the number of malaria-carrying mosquitoes might have consequences; perhaps those mosquitoes help control the population size of some other animals (for example, rats) by infecting them with the human pathogen or a related pathogen also affected by the genetic engineering. Anopheles mosquitoes are not a classic keystone species — a species, like the wolves in Yellowstone, whose removal changes an ecosystem quickly and radically — but we simply don’t know what the ecological effects of reducing their population would be.

A second risk is that the genetic modifications may drift from one mosquito species to others. Beyond the 460 species of Anopheles mosquito, at least another 2,500 or so mosquito species do not carry malaria but might be affected. How many of them mate and have offspring with Anopheles species or otherwise trade genes in the wild? And, if they did, what would be the consequences? We might end up eliminating species of mosquitoes, including those that never bite humans, with unknown consequences, compounding the ecological risk.

A risk of a different sort is that early benefits will shrink. The mosquitoes may develop resistance by mutating around the population-reducing changes. The gene drives that push the changes through populations may lose their power through mutations over generations.

All these risks can be studied. Small-scale deployments can offer hints about the effects of large interventions. We can consider ways to monitor, control and even reverse the introduced genetic changes, possibly by means of “kill switches” that allow the destruction of modified mosquitoes. One technique under study involves giving the engineered mosquitoes either a critical dependence on, or an acute vulnerability to, some otherwise benign substance. Withholding the substance, in the first case, or providing it, in the second, would kill the mosquitoes while harming nothing else. We might also use gene drives to reverse the original changes.

A very different set of complicated risks arise from human reactions to these experiments. Advocates of these technologies are terrified about a backlash like the one that has hit genetically modified foods. In 2002-2003, in the midst of a famine, several African countries at least initially refused to allow the distribution of genetically modified foods.

In November, the Conference of the Parties to the Convention on Biological Diversity ultimately approved a cautious approach to moving forward with deployment of this technology. Researchers would have to show that engineered insects posed no threat, and any distribution of modified mosquitoes would have to be preceded by the “free, prior and informed consent” of people who live in affected areas.

One good start to building confidence would be to open up research by making broadly available results and data from research into genetic modification of mosquitoes. Individual firms might lose some control and some advantages, but the field could gain enormously from decreased mistrust. Firms, researchers and government funders (including, for example, the Defense Advanced Research Projects Agency, which is interested in the extraordinary new DNA editing technique called CRISPR, with or without gene drives) should be required to describe publicly what they want to do, how and why. (With CRISPR, researchers can cheaply and easily edit specific “letters” of the DNA of the initial target mosquito. And supplemented with gene drive, CRISPR can be used to change all copies of that DNA stretch in each of that mosquito’s descendants — including “normal” versions of the gene it inherited from its other parent.)

Securing meaningful consent could be a tricky business. Oxitec’s releases of genetically modified Aedes aegypti mosquitoes in Brazil, Panama and the Caymans were made with government permission, the company says. But when it wanted to do a test release in the Florida Keys, it was stymied. The Florida Keys Mosquito Control District held a nonbinding referendum on the project in November 2016; the proposal was approved in the overall district but rejected by voters in the region where the release was planned, Key Haven. Oxitec is now trying to get approval for a different Florida test site.

Mosquitoes, of course, do not respect local or national borders, further complicating the issue of consent. Changed mosquitoes in countries that approve their release may lead to similar populations in places that rejected them. And not just immediate neighbors, either: Mosquitoes spread through international trade and human travel — the Asian tiger mosquito, Aedes albopictus, successfully invaded the United States by catching rides in imported used tires. Yet requiring worldwide consensus before proceeding in any one place is a recipe for stalemate.

There is lot to be done to determine whether and how to use genetically engineered mosquitoes to fight malaria. It may sound like an agenda for decades of work. But again, 440,000 people are dying every year. We don’t have decades for cautious debate.

History knows the name of Henry Gunther, the last soldier killed in World War I, one minute before the Armistice. How many died of AIDS who would have survived if the anti-retrovirals had been approved just a month, or a week, or a day earlier? And who will be the African toddlers whose deaths by malaria could have been prevented if only we had moved faster with the genetic modification of mosquitoes?

It’s never possible to know the perfect time to introduce a new intervention, but we need to investigate these new approaches and their likely consequences both carefully — and urgently.