Noisy origins

COVID-19 probably did emerge in a Wuhan market – but I can understand why people think it didn’t

In late 2019, I saw a talk by a research team who were sampling coronaviruses from animals in central China, with the aim of predicting – and ideally preventing – future pandemic threats.

Listening to the talk, I was sceptical. That’s all well and good, I thought, but how likely is it that the next pandemic threat emerges in central China with a coronavirus? We’d already had a SARS outbreak begin in China in 2002-3. Wasn’t this just fighting the last battle?

Within a couple of months, it was clear I’d been wrong about the likelihood of a SARS-like pandemic emerging from China. But I think I was right to be sceptical about the benefits of this research for pandemic response. Despite spending years looking in the right kind of place for the right kind of viruses, the research hadn’t produced much in the way of actionable knowledge that could help us predict – or get ahead of – COVID-19.

Jemma Geoghegan and Eddie Holmes previously pointed out the limitations of predicting virus emergence in their 2018 paper:

For example, it has long been known that influenza viruses bind to sialic acid-containing molecules as receptors. However, this information has not improved prediction of influenza virus emergence and re-emergence. More generally, machine learning requires very large amounts of data to predict common events, whereas studies of disease emergence necessarily use data on rare events to predict rare events.

Yet when it comes to studying potential pandemic viruses, there’s another challenge. Research into these viruses may increase the threat of a virus, either by collecting new viruses from isolated wild animals and bringing them into labs staffed by humans, or running so-called ‘gain of function’ studies that introduce new mutations into viruses that are not yet a threat.

I recently had a new Bloomberg article looking more at this trade off in knowledge and risk in the context of viruses like H5N1 and SARS-CoV-2.

Of course, when it comes to lab experiments and COVID-19, one of the most prominent stories in recent years has been the possibility that the pandemic began with a lab leak in Wuhan.

While researching the Bloomberg piece, I reminded myself of the timeline and emergence of evidence. And it struck me just how hard it must have been for the average non-scientist to make sense of what was happening, even if they were acting in good faith.

First, we had the scattergun conspiracy theories in early 2020. The virus had come from snakes, or was a part-HIV hybrid, or was a US bioweapon. These conspiracies were followed by rebuttals from scientists (including my own Guardian article).

We also had data showing that the earliest reported COVID cases didn’t have a link to the Huanan Seafood Market in Wuhan, even if later ones did. Plus the fact that Wuhan had a major lab working on coronaviruses. Even if we discarded the conspiracies, which way was the evidence pointing?

During February 2020, I found the intense media speculation around origins quite unhelpful, because I felt it crowded out discussion of the much bigger problem heading our way. As I put it in the Bloomberg piece:

It was a bit like standing on a beach debating which underwater volcano was responsible for the massive incoming tsunami.

As countries shut down in response to growing COVID epidemics, the investigations would continue. In May 2020, the Chinese Center for Disease Control and Prevention reported that animal samples collected from the market in early January had tested negative. Rather than being the origin of the outbreak, they said the market was ‘a victim of COVID-19’. This was followed by speculation in China that the virus originally came from frozen food imported from abroad.

Then there was the 2021 WHO-convened report on possible origins, which was designed to be the first phase of several. In particular, it mapped the path for follow-up scientific studies to narrow down the likely origin. ‘Crucially, the window is rapidly closing on the biological feasibility of conducting the critical trace-back of people and animals inside and outside China,’ the authors noted in August 2021. But, hampered by politics, these follow-up studies did not go ahead.

Next, there was a pre-print paper in early 2022, which reported that some environmental samples from the market that tested positive for SARS-CoV-2 also contained human DNA. The interpretation was that humans – not animals – first brought the virus into the market.

For the first three years of the pandemic, it was hard to point to much that convincingly identified the Wuhan market as the source of COVID-19 pandemic.

But something unexpected would happen. In early 2023, researcher Florence Débarre discovered previously unreported genetic sequences from the Wuhan market. These sequences indicated wildlife presence at market stalls that had been positive for SARS-CoV-2. Although the research team couldn’t confirm that these animals were infected or that the market was the outbreak’s origin, the new evidence made alternative explanations less plausible.

It’s not definitive evidence the market was the origin. But we’ll probably never have that. Instead, we must weigh up imperfect evidence pointing to the market and – in my view – even more imperfect evidence pointing to a lab or somewhere else. Given a pandemic happened, we must pick the somewhat unsatisfying explanation that is the least unsatisfying.

As I’ve written about previously, this is the kind of situation where your prior beliefs really matter. If you strongly think a lab leak is possible – perhaps because they’ve happened before, from smallpox to influenza – then the current evidence may not be enough to sway you. Indeed, for much of the pandemic, a lab leak of a novel coronavirus collected from wildlife seemed a plausible possibility to me, simply because of the patchy evidence available to suggest otherwise.

But suppose we strip away the three years of conspiracies and noise and denials and contradictory reports and unreported data. And, instead, we think about how we’d have weighed up this evidence if we’d had it all from the start. This is what I tried to do while writing the above Bloomberg piece. And it leads, I think, to a different place compared to the journey many went through in real-time.

Earlier this month, there were also reports of preliminary analysis suggesting that samples from animals at the Wuhan market had genomic markers indicative of an immune response to SARS-CoV-2, which would be consistent with the theory that there were infected animals there.

In addition, new preliminary data were announced recently about the coronaviruses collected between 2004 and 2021 and studied at Wuhan Institute of Virology. These data reportedly suggest that none of the collected viruses were a closer relation to SARS-CoV-2 than other viruses that are already known, which would suggest none of these newly collected viruses could be the source of the COVID pandemic.

Yet even if COVID wasn’t the result of a lab leak, it doesn’t mean collecting and creating risky viruses should go without scrutiny. For certain experiments, the risks are real and arguably too large – especially if researchers try to justify them with appeals to authority rather than solid evidence of benefit for pandemic preparedness.

Cover image: China News Service