Wastewater testing gains traction as a Covid-19 early warning system

What only a month ago had been merely an intriguing laboratory finding about analyzing wastewater to detect the virus that causes Covid-19 has quickly leapt to the threshold of real-world use.

With swab tests still plagued by capacity issues, inaccuracy, and slow turnaround, testing wastewater for the novel coronavirus’ genetic signature could give communities a faster way to spot a rebound in cases — as soon as this fall.

“There is real hope that this can be a sensitive, early warning” if, as officials ease social distancing measures, Covid-19 begins to spread again, said Peter Grevatt, CEO of the nonprofit Water Research Foundation. “Several labs have achieved a proof-of-concept in terms of demonstrating the ability to detect the RNA [genetic material] of the virus in wastewater.” Studies in the U.S. and the Netherlands, among others, have shown you can pick up a signal about a week before the first clinical case.

Grevatt and his colleagues briefed congressional staffers last week on the potential for wastewater analysis to be the canary in the Covid-19 coal mine, and on Wednesday the National Academies’ Water Science and Technology Board hosted a panel discussion on how to build a surveillance network and what additional research is needed to make it work. Water utilities from southeastern Virginia to Portland, Ore., are already conducting the analysis on their own. And by next week, Grevatt’s group will have identified the labs that will participate in a quality control test deemed crucial for rolling out a nationwide effort to analyze wastewater for coronavirus: The foundation will send wastewater samples gathered by several utilities to all participating labs and have them run the analysis, compare results, and agree on best practices.

“We hope to have results of this lab-to-lab comparison by the end of the summer,” Grevatt said.

Countries are not waiting for every scientific question to be answered. Finland, Germany, and the Netherlands have already launched national wastewater surveillance programs to catch any resurgence of Covid-19. “People are already starting to scale this up,” environmental engineer David Sedlak of the University of California, Berkeley, told the Academies’ panel.

The stakes are enormous. As states ease social distancing, they need to know if they are reopening too quickly. If they can’t detect a surge of cases until patients show up in emergency rooms, it could be too late to prevent a repeat of this spring, when hospitals in New York City and elsewhere were overwhelmed. If, as many hope, the country makes it through the summer with a “flattened curve” — keeping cases below the number that hospitals can handle — then it will be crucial to be on high alert for any second wave in the fall.

Otherwise, the body count will rise and the nation could face another economically crippling shutdown.

“It’s tremendously important to get this right,” Grevatt said. “These are life-and-death questions.” And although more research is needed to figure out if, for instance, the amount of virus in stool and therefore wastewater varies with the severity of disease, “this has the potential to be implemented in the current pandemic.”

“Wastewater epidemiology” has been used for decades to detect polio in countries where the disease remains endemic and, more recently, to estimate the prevalence of opioid abuse in U.S. communities.

“We know that SARS-CoV-2 [the virus that causes Covid-19] is shed in stool, which means it can be collected in sewage systems,” said Megan Murray of the Harvard T.H. Chan School of Public Health. In small studies so far, the detection of the new coronavirus in sewage samples “has correlated very nicely with the arrival of Covid-19 into different communities,” she added, including picking up “significant amounts of viral material” in Boston sewage weeks before cases arrived in March.

In addition to detecting cases sooner, wastewater analysis can also get around the well-known limitations of the swab tests used to detect Covid-19. If the test (which looks for the virus’ genes) is given to 100 infected people, it will miss about 67 if they are four days after infection and 38 if they are on the first day of experiencing symptoms, incorrectly pronouncing them virus-free, Johns Hopkins researchers reported this month in Annals of Internal Medicine.

And although Covid-19 testing in the U.S. has become more available and faster, testing capacity is still nowhere near enough to test everyone every few days, which would be necessary to, for instance, safely re-open schools and workplaces.

“Wastewater offers the opportunity to provide near real-time trend data to evaluate the impact of policymaking,” such as re-opening the economy, said Mariana Matus, CEO and co-founder of Boston-based Biobot Analytics. It can provide “early warning for second waves, and the opportunity to mass-test up to 75% of the U.S. population on a regular basis at a fraction of the cost of clinical testing.”

Getting even a weeklong head start on detecting cases can make a huge difference. If the U.S. had begun imposing social distancing measures one week earlier than it did in March, the death toll by early May would have been 36,000 lower, Columbia University researchers calculated this month. Two weeks earlier would have averted 54,000 deaths.

Biobot is working with about 400 water facilities in 42 states to see if wastewater epidemiology for Covid-19 in the real world can live up to the hopes inspired by early lab tests. (Many of the utilities are also running their own, independent tests.) When it analyzed wastewater from a Massachusetts plant from March 18 to 25, it found high levels of several telltale SARS-CoV-2 genes.

Although that indicated that some number of people had Covid-19, it wasn’t clear how many. Biobot’s analysis “can distinguish between 10 and 100 [cases],” Matus said, “but we are running experiments to get more precise estimates.”

Doing that should be at the top of the research agenda, Grevatt said: “Work is moving in a direction where you may be able to count the cases in a community, but we aren’t there yet.” To get there, scientists must figure out when during the course of infection people shed virus in stool, whether the amount of viral shedding is the same in severe and mild or even asymptomatic cases, and whether the quality of an area’s water — mineral content, industrial effluent, and other components — affect the detection of viral genes.

They’d also like to know where, in the water treatment facility, is the best place to sample for the coronavirus. On Tuesday, scientists in Spain reported a preliminary answer: remnants of virus, including its genes, tend to collect at the “sludge line” in the large settling tanks that separate solids and liquid.

Even if wastewater analysis can’t yet quantify cases, it can detect when their prevalence changes. Scientists in France reported that the rise and fall of confirmed infections in Paris correlated with the amount of virus detected in sewage from March 5 to April 23 (France’s lockdown began on March 17). “Viral genomes could be detected before the beginning of the exponential growth of the epidemic,” the scientists found, and “a marked decrease in [viral RNA] was observed” as case numbers fell.

That and other studies show that “you can track changes like those we’re seeing right now in many communities” as they ease social distancing, Grevatt said. “There are many areas of the country, especially rural ones, that may not even have an infection yet. Wastewater analysis can tell them when they do,” days before someone shows up sick.

Utilities aren’t waiting for every research question to be answered. Oregon’s Clean Water Services, a utility that serves 613,000 people west of Portland, is part of Biobot’s network, sending weekly samples to the company. It is also doing its own research, trying to answer key questions such as how rain affects viral concentration, said Ken Williamson, director of regulatory affairs.

Engineers at the Southern Nevada Water Authority found high levels of five SARS-CoV-2 genes in samples taken around March 9, which lines up with when the state has its first confirmed case, said civil engineer Dan Gerrity of the authority and the University of Nevada Las Vegas. They found a peak in late March and early April, with ups and downs through late May — also mimicking reported case counts.

Hampton Roads Sanitation District in southeastern Virginia started sampling for SARS-CoV-2 on March 11 “to look at trends and estimate what prevalence might look like,” said Jim Pletl, director of water quality for the utility. Cost: $70,000 for an automated RNA extractor (to get the genetic material out of cells), $70,000 for a digital PCR machine (to identify the genes), and $100 per sample.

For its inter-lab study, the Water Research Foundation will see if the analyses agree, hoping to come up with best practices that can guide nationwide implementation as soon as this fall. “That,” Grevatt said, “is going to be important for the scientific credibility of this approach.”