Secret signals in sewage detect coronavirus outbreaks earlier

A person walking away from a toilet as a virus is seen traveling through the pipes
(Marisa Sloan/MEDILL)

By Marisa Sloan
Medill Reports

Infectious disease expert Rita Colwell began her career studying fruit fly genetics. Over the next several decades, she added marine and medical microbiology to her repertoire, assisted NASA scientists with oceanographic work and became the first woman to direct the National Science Foundation.

Now, like many researchers, she is tackling the challenge of COVID-19.

“That’s 21st century science,” she said. “By virtue of being shuttled from place to place, I landed right where the cutting-edge research is going on right now.”

Colwell’s work draws heavily from years spent tracking the spread of cholera, a bacterial disease caused by contaminated water, in the U.S. and in developing countries. Her success developing inexpensive tracking methods — along with infiltration methods that curtailed the amount of cholera in the drinking water of rural Bangladeshi villages to nearly half — has made her confident in her ability to accomplish the same with the novel coronavirus.

“COVID-19 is [caused by] a virus and cholera is caused by a bacterium, but nevertheless they are microorganisms and their behaviors seem to be driven quite similarly,” Colwell said. “So it was very easy to hypothesize and then test these hypotheses based on years of work.”

The coronavirus is surging in the U.S., surpassing 14 million cases and a quarter of a million deaths this November. Colwell and other researchers believe a technique that identifies remnants of the virus in wastewater could complement existing individual surveillance systems, such as nasal swab and antibody tests, when it comes to informing a community’s public policy decisions.

“The virus has a propensity for attaching to a specific receptor site, ACE2, and that receptor seems to be distributed on many of the organs in the human body,” Colwell said. “[That] accounts for the observation that not just the lungs are affected, but also the heart, the liver, the intestines and even the brain.”

In addition to a cough and difficulty breathing, many who contract the coronavirus also experience symptoms such as vomiting or diarrhea, a loss of taste or smell, and what has been referred to as “brain fog.”

Colwell explained that because the coronavirus is shed directly from the gut, it can be detected in the stool of about half of all people who test positive — even if they are otherwise asymptomatic. That makes wastewater surveillance a powerful tool for monitoring the spread of the virus within a community, particularly younger and under-resourced communities where people with less severe symptoms may not get tested as often as they should.

And the method does more than just track existing outbreaks. The proverbial canary in a coal mine can also predict when new ones will happen.

The coronavirus has been detected in wastewater approximately 7 to 10 days before cases show up at hospitals, likely because shedding of the virus in stool begins before more recognizable symptoms. If a sewershed notices a spike in the virus, leaders of that community are able to predict a future, similar spike in hospitalizations and reinstate restrictions like lockdowns to prevent it from worsening.

That, in turn, could be enough to prevent hospitals from becoming overwhelmed.

“[Similarly], if the number of virus in the sewage were to drop precipitously, it would mean the techniques of social distancing, masking and hand washing… were working,” Colwell said.

It isn’t a new idea. Variations of wastewater analysis have been used for over 50 years, beginning in the 1960s when researchers at Yale University conducted experiments to assess the effectiveness of vaccination campaigns for polio, another virus that can be found in stool. More recently, the growing field of wastewater-based epidemiology has used sewage to gather community-level data about everything from illicit drug use to nicotine levels.

Over the summer, Colwell’s bioinformatics company CosmosID began collaborating with the Maryland Department of the Environment to sample sewage from five major wastewater treatment plants across the state. Her method uses a technique called reverse transcription polymerase chain reaction (RT-PCR) to target two different parts of the virus, the N gene and the ORF1ab region, present in a sample and exponentially amplify its DNA until there is enough to detect.

“We have it really down to a science, which is what it should be,” Colwell said. “We can actually do collection, analysis and results in a single day if necessary. We’re the quickest, and I think the most reliable.”

Hers was the first COVID-19 wastewater test with guaranteed rapid turnaround time, meaning results came back within a single day of sampling. That’s a valuable head start on individual methods like nasal swabs and saliva tests, which have turnaround times ranging from a day to weeks.

A graph showing Colwell’s COVID-19 data from the Ballenger-McKinney wastewater treatment plant and Frederick County, Maryland's documented coronavirus cases
Colwell’s COVID-19 data from the Ballenger-McKinney wastewater treatment plant, the largest facility in Frederick County, Maryland, compared with the county’s documented coronavirus cases. (Marisa Sloan/MEDILL)

On Nov. 12, Maryland Governor Larry Hogan announced $1 million in funding for the Sewer Sentinel Initiative, a wastewater surveillance system grounded in the pilot program led by Colwell over the summer.

“Our administration is setting an example for the nation on how states can use cutting-edge science, innovative technology, and local partnerships to protect public health and help combat a second wave of COVID-19 infections,” Hogan said in a press release.

The Maryland Department of the Environment, in close coordination with the Maryland Department of Health, will focus its efforts on monitoring outbreaks in the state’s most vulnerable populations first: correctional facilities, nursing homes and low-income housing.

“This, I think, is not going to go away after COVID-19,” Colwell said. “It is a public health tool.”

Maryland is joined by various other local governments in its endeavors. In Chicago, the Metropolitan Water Reclamation District supports a similar project led by researchers from Argonne National Laboratory, Northwestern University and the University of Illinois, Chicago. The district’s main role is to provide samples and information about the city’s water reclamation plants and collection systems.

“In Phase I, which is already underway, a plan is to be developed for sampling and testing procedures and logistics among collaborators,” said Allison Fore, a public affairs officer at the MWRD, in an email. “Phase II entails implementation of the plan.”

Phase I has proven to be no walk in the park, thanks to the mammoth 4,400-mile sewage system that lies below Chicago’s streets. In addition to communicating data to local public health departments, Fore said the team of researchers uses computer modeling to estimate where neighborhood-level sampling could best identify future outbreaks of the virus.

Many municipalities already have the infrastructure in place to implement such communitywide viral testing. According to the U.S. Environmental Protection Agency, about 80% of U.S. households are connected to a central sewer system and thousands of wastewater treatment plants already collect samples to monitor the amount of pollutants in sewage and storm runoff.

There are challenges, however, particularly with older systems. Sewage flow is never just sewage ― think of how much water rushes into storm drains during a downpour. Pipes that have cracks or holes can allow groundwater to percolate in as well and exacerbate the problem.

“We had a pretty rainy summer, for example, a lot of big rainstorms,” said Larry Cahoon, a marine biologist at the University of North Carolina at Wilmington. “They dilute down your signal, so your signal may be going down in the raw sewage, when in fact it’s steady or going up in the population.”

Like Colwell and many other researchers, Cahoon has transitioned from his usual work — in aquatic ecology — to monitoring the spread of the coronavirus via his campus’ sewer system. Right now, he is doing the math to correct these varying volumes of water and ensure the calculated concentrations of the virus in his wastewater samples are as accurate as possible.

According to The New York Times, coronavirus cases in college towns dramatically increased and cases in non-college towns decreased when academic institutions reopened their doors for the fall semester. In the face of this, a growing number of campuses have turned toward wastewater surveillance, a method precise enough to identify outbreaks even in individual dorms, to remain safely open and operational.

A graph showing coronavirus cases in college towns versus non-college towns
When schools across the country reopened at the end of August and beginning of September, the seven-day rolling average of coronavirus infections increased more in counties where students comprised at least 10% of the population. (Marisa Sloan/MEDILL)

Cahoon said UNCW’s campus is a perfect candidate.

“Our sewer system collects only sewage from the university,” he said. “And our sewer system is relatively new, because the campus has been expanding rapidly over time, so we have a lot of access points.”

Over the summer, the university’s College of Arts & Sciences spent over $100,000 on a new PCR instrument and began sampling five manholes across campus, including a dorm used exclusively for quarantining students and an empty dorm designated for quarantine overflow. Using the former as a positive control and the latter as a negative, Cahoon and his coworkers developed a method that was sensitive enough to detect the virus shed from just two quarantined students and fast enough to give results in just 5 hours.

Sure enough, they began getting hits as students came back to campus in fall. Cahoon said it became obvious that the virus was spreading through campus, and he soon found himself sending emails to the university’s administration after every test he ran — almost every other day.

He was met with radio silence, he said.

“The university is not using our data,” Cahoon said. “If your bottom line is just paying for dorms, well, it doesn’t matter how many sick people you’ve got in them.”

In less than three months, UNCW faced at least 11 clusters on campus, defined as five or more positive cases in close proximity, four of which involved athletic teams. Cahoon said that although his wastewater monitoring data of the athletics buildings were ahead of the university’s detection of those clusters, the administration chose to act retroactively rather than proactively. Throughout September, that meant moving more residential students to single dorm rooms and rewarding mask-wearers with Chick-fil-A coupons and other prizes.

Neither Chancellor Jose Sartarelli nor the Office of the Provost responded to a request for comment.

“Part of the problem with the university’s approach is that until mid-September, testing was only voluntary,” Cahoon said. “It was obvious that a very small portion of the student body was getting tested.”

For the first few weeks of the semester, students who did seek out a test could expect to be charged $55 if they were not enrolled in UNCW’s health insurance program. The university’s health center averaged only 13 tests per day during this period.

The coronavirus pandemic is a multidimensional problem and it has become clear that wastewater surveillance looks at only one dimension of that problem: community spread. Although it is a powerful tool for identifying and predicting future outbreaks, only lockdowns, mask mandates and social distancing can then prevent such outbreaks from occurring.

As the semester comes to a close, Cahoon is optimistic that the university’s administration is finally convinced of the efficacy of wastewater surveillance and will more effectively utilize it when students return in January.

“I think we’ve convinced them that we can help,” he said. “Our fundamental responsibility is to keep the students safe, and if we fail to do that, we’ve screwed up.”

Marisa Sloan is a health, environment and science reporter at Medill. You can follow her on Twitter at @sloan_marisa.