Changes to the SARS-CoV-2 genome—including some of those found in currently circulating variants—can negatively affect the detection of the virus by reverse transcription (RT) PCR, according to a study published on April 26 in the Journal of Clinical Microbiology. The researchers propose that mutations in the loci recognized by DNA primers may reduce the amplification of viral sequences and, as a result, potentially hinder the detection of the virus in samples from COVID-19–positive individuals.

This finding isn’t cause for full-blown panic, the authors say. “We thought maybe this could be more common than not. But it turns out, it’s actually fairly rare,” says coauthor David Wang, a virologist at Washington University. Wang and his colleagues recommend that diagnostic tests include more than one target to ensure proper SARS-CoV-2 detection. While a number of products already include multiple genetic targets, some COVID-19 RT-PCR assays authorized for emergency use check only one.

The primers used for RT-PCR assays were developed early in the pandemic, when the SARS-CoV-2 virus was first sequenced. Based on information about other coronaviruses, researchers designed PCR primers to amplify sequences in the viral genome thought to remain relatively stable. The approach has been used to great success, detecting SARS-CoV-2 in samples from nasopharyngeal swabs, saliva, and even sewage.

At Washington University, the molecular diagnostics lab of Barnes-Jewish Hospital has been using the Roche cobas SARS-CoV-2 test to process patient samples. It looks for the viral gene ORF1ab as well as the E gene, which encodes the envelope protein. For any given sample, these targets should take roughly the same number of PCR cycles to be detected, a value known as the cycle threshold.

“If you have a high value in one gene, you have a high value in the other gene—and vice versa,” says coauthor Bijal Parikh, a clinical pathologist at Washington University and the medical director of the molecular diagnostics laboratory. While most samples his team has processed have had similar cycle threshold values for the two targets, a handful strayed from the expected correlation: at times, the E gene wasn’t amplified to the same degree as was ORF1ab.

Despite this curious result, the tests still correctly identified SARS-CoV-2–positive samples based on the ORF1ab signal. To figure out what was going on with the E gene, the team sequenced a handful of viral samples. They found three samples had a common mutation in the E gene, one not present in any of the common variants now circulating in the population. The researchers propose that the mutation affects the binding of the PCR primer and interferes with amplification.

Unfortunately, they couldn’t confirm this hypothesis because the assay’s primer and probe sequences aren’t publicly available. The mutated site, however, falls within the sequence targeted by primers used in a widely used E gene assay developed by researchers at Charité – Universitätsmedizin Berlin in Germany.

Another study published in September in the Journal of Clinical Microbiology reported similar results, where mutations at a different site within the E gene sequence, targeted by the Charité primer-probe set, also interfered with RT-PCR in the Roche cobas SARS-CoV-2 assay. And in December 2020, Emily Crawford, a biochemist at the University of California, San Francisco, and colleagues reported in the Journal of Clinical Microbiology a related finding: mutations in the SARS-CoV-2 N gene, which encodes the virus’ nucleocapsid protein, reduced RT-PCR test sensitivity by interfering with primer binding.

“We are still learning about what are the most conserved regions of the virus and what are the least conserved,” says Crawford. “Until we really pin that down, we have to be on our toes, to be ready for a mutation to pop up anywhere.”

The scientific community needs more of these types of studies, says Chantal Vogels, a virologist at the Yale School of Public Health, who was not involved with the new work. She led a study published in July 2020 in Nature Microbiology that analyzed the sensitivity and efficiency of SARS-CoV-2 primer sets. The researchers found that a mismatch between one primer and its target sequence resulted in decreased test sensitivity.

PCR assays to capture an evolving virus

The US Food and Drug Administration (FDA) has acknowledged that SARS-CoV-2 mutations could interfere with COVID-19 tests and continues to monitor variants and evaluate potential effects on diagnostic assays. The agency has also provided recommendations to developers to design tests so that viral mutations have minimal evasion, such as by including multiple genetic targets. The FDA also advises developers to watch for mutations that could alter test performance and clearly convey test limitations.

“From the very beginning, we’ve been keeping a very close eye on all the different variants,” says Palani Kumaresan, the head of research and development for Roche Diagnostic Solutions. At this moment, the company isn’t planning to change its test. The E gene signal is pan-sarbecovirus, the virus subgenus that includes both SARS-CoV and SARS-CoV-2. A SARS-CoV-2–specific signal comes from Orf1ab. “When we combine those two, we do not see any impacts,” Kumaresan says. Still, Roche is monitoring coronavirus variants for any reduction in performance and has launched a separate SARS-CoV-2 variant test for research purposes.

The US Centers for Disease Control and Prevention (CDC) regularly monitors the primers and probes for its COVID-19 diagnostic panel and multiplex test for flu and COVID-19, a CDC media representative tells The Scientist by email. CDC also actively tracks and characterizes coronavirus variants through genomic surveillance efforts.

“As long as the virus keeps changing—which is a natural thing—you just need to keep monitoring,” Vogels says. “Having these studies out there that look at that only makes our tests better.”

S. Tahan et al., “SARS-CoV-2 E gene variant alters analytical sensitivity characteristics of viral detection using a commercial RT-PCR assay,” J Clin Microbiol, doi:10.1128/JCM.00075-21, 2021.