'Just stupid stuff': Logistics and lack of testing stymied finding the Omicron variant in the US

After South African scientists alerted the world last week to the new, concerning Omicron variant of SARS-CoV-2, researchers here in the U.S. warned that the highly mutated strain was likely already circulating on American soil. It was just a matter of time before someone found it.

On Thursday, health officials reported the country’s second Omicron infection, in a Minnesota man who attended an anime conference in New York days before his symptoms began. Notably, he had not traveled internationally, unlike the first case — a California resident recently returned from South Africa — indicating the occurrence of domestic transmission. But the timing of discovery has many wondering, “why did it take so long?”

According to details shared by the Minnesota Department of Health, the man, who is a resident of Hennepin County, home to the state’s most populous city, Minneapolis, first developed symptoms on Nov. 22. He took a PCR test on Nov. 24. News of Omicron broke on Nov. 25. Minnesota health officials didn’t confirm from sequencing data that his infection was caused by the Omicron variant until a week later, on the evening of Dec. 1.

By contrast, public health agencies in the U.K. found that country’s first cases of the new strain less than 48 hours after South Africa sounded the alarm.

“Even though we’ve increased sequencing in recent months, we still have a sampling issue,” said Eric Topol, the founder and director of Scripps Research Translational Institute in La Jolla, Calif., which is part of a SARS-CoV-2 surveillance consortium organized by the Centers for Disease Control and Prevention last year. The U.S. is only testing at a rate a quarter of what the U.K. is, and there’s not a systematic approach for testing travelers and people with breakthrough infections. So it’s less likely that something new and scary will ever get onto a swab, let alone a sequencer.

And that puts the U.S. at a big disadvantage, said Topol. “It all goes back to the same theme from the beginning of this pandemic,” he said. “We don’t have unity in this country, we have a balkanized, well, everything.”

The Thanksgiving holiday likely contributed to the slowdown. But that isn’t the whole story. While the U.S. has greatly ramped up its overall SARS-CoV-2 sequencing efforts over the past year, progress has been patchy. Urban centers close to large academic centers tend to be well covered, while rural areas are less so. That means public health departments in large parts of the country are still flying blind, even as they are figuring out ways to prioritize Omicron-suspicious samples.

“People are trying to make a concerted effort to triage as best they can, but we didn’t do enough as a nation to increase sequencing capacity since the start of the pandemic, and now here we are, and it’s a problem again,” said Lane Warmbrod, of the Johns Hopkins Center for Health and Security, who co-authored a report earlier this year outlining what an effective national surveillance program should look like.

Even when a patient lives in a major city like the Minnesota case, logistical challenges can slow the process to a crawl. In that state, it typically takes four to seven days for positive samples to get packaged and shipped to a facility that can perform sequencing, Doug Schultz, a spokesman for the Minnesota Department of Health, told STAT. Sequencing takes another four to five days. With all the testing turnarounds, that means that genomes coming off the sequencer are usually from samples collected 10 to 14 days prior.

“We’ve built in all the speed we can to the process,” said Schultz. “There’s just a certain amount of time it takes to collect samples, package, and transport them baked in.”

Minnesota is hardly alone. The fractured nature of U.S. health care means that when someone is swabbed for SARS-CoV-2 in a hospital or clinic, the majority of those samples go to one of the large private testing providers like LabCorp and Quest. Some subset of those samples will then be set aside to be sent to county or state public health labs, which either do their own sequencing, or send the samples off to academic labs or other industry partners to decode the genomes of the viruses inside.

“In a perfect world, we would have a much stronger health system that is more integrated with our public health agencies so we could have testing and sequencing being done within the same network,” Warmbrod said.

All that form-filling and ferrying about can take days to weeks. Sequencing itself can be as fast as a 24- to 48-hour turnaround. It’s the logistics of moving samples around that’s the real bottleneck.

“It’s really just stupid stuff,” said Amy Mathers, associate director of clinical microbiology at University of Virginia Health, whose lab team sequences positive coronavirus samples for the state of Virginia. Currently, they receive up to 300 samples a week, delivered in shipments once or twice weekly via courier from the state public health lab in Richmond. And if the public health department flags some samples that are suspicious because of travel history or contact tracing data, they’ll push them to the front of the line.

“We’re more integrated into the health system than we ever were before the pandemic,” said Mathers. But the data they pull off the sequencer isn’t reflective of real-time infections. Typically it shows what was circulating in the state 10 to 14 days earlier. Right now they’re running samples from mid-November.

“We tried to and are still lobbying for a system where people can send us positive samples directly,” said Mathers. “But the problem was it was hard to find a way to log them into the public health system. The forms were just too confusing for the hospitals.”

Her team has yet to find any cases of the Omicron variant, either in their sequence database or in any new samples, including ones taken from travelers and rushed to the front of the line. Mathers said her lab is sequencing every sample that tests positive in her area.

During the summer Delta surge, the backlog stretched back even further. And in the end, all that sequencing wasn’t that helpful, she said. In the month of September, her team sequenced over 1,000 samples. Every single one of them was the Delta variant. But now, it’s more important than ever to sequence every last vial of virus in the state. “As these new variants emerge we have to do heavy sampling to capture the one Omicron that’s lurking,” she said. “So I know that if someone comes to Charlottesville with Omicron, we’ll find it. But it’s not the same everywhere.”

In a press conference Tuesday, CDC director Rochelle Walensky said the U.S. is currently sequencing about 80,000 samples per week, between public health, academic, and private labs. But it’s not clear what percentage of them represent real-time windows into the virus’s spread.

That’s not information that’s readily available, Kelly Wroblewski, director of infectious diseases for the Association of Public Health Laboratories, told STAT via email. And there doesn’t seem to be an urgent, coordinated push by all players to speed things up, she said. “Having sequencing distributed to all of the states and many local jurisdictions is helpful in this way as it brings sequencing closer to patients,” she wrote.

Until broader geographic sequencing is available, there’s another step scientists can take now. Because of Omicron’s unique constellation of mutations, it leaves a tell-tale pattern with some types of PCR testing. These assays hunt for snippets of the virus’s genome, including on the S gene, which codes for its spike protein. On some assays, Omicron’s S gene snippet doesn’t show up — a phenomenon called “S gene dropout.” If scientists see that, it’s a signal that the virus likely isn’t the widespread Delta variant, and it could be Omicron. And in fact, that’s how officials in California and Minnesota found their infections.

In the case of the California individual, after he tested positive at a city testing location (using a PCR test that didn’t pick up the S gene dropout), his travel history spurred officials to send the sample to the lab of Charles Chiu, a researcher at the University of California, San Francisco, where his team ran a test looking for the S gene dropout. Within a few hours, they saw the signal so they expressed the sample to a handheld gene sequencer, which runs faster than the large ones typically used for mass sequencing. It gave them confirmation later that same night.

When they heard of the Omicron variant, Minnesota officials contacted one of their clinical lab partners in New Jersey, whose standard assay already looks for the S gene dropout. They found a handful of samples with the suspicious pattern, which they sent back to Minnesota for sequencing. The samples arrived Tuesday, and it took the state’s team less than 48 hours to complete the genomic analysis and discover that one of them was the new variant.

On Thursday night, officials in Colorado confirmed an additional case of the Omicron in a person recently returned from South Africa, and in New York, officials confirmed five cases, likely caused by local transmission. Scientists expect more to be identified in the coming days and weeks. “We were late to the Omicron party, but they’ll start cropping up fast,” said Topol. But without improvements, these discoveries will provide a picture of the variant’s spread in the past — and an incomplete one at that — not where it is today.