COVID severity tied to reactivation of dormant viruses

Clinical takeaway: Reactivation of dormant herpesviruses tracked with COVID severity in immunocompetent patients, which widens the group of hospitalized patients in whom it may be worth considering.

When a hospitalized COVID patient deteriorates in the second week, the differential is well-worn: bacterial superinfection, thromboembolism, progressive lung injury. Chronic viral reactivation rarely joins that list unless the chart indicates that the patient is immunocompromised.

Reactivation has been documented in COVID before, but based on thin evidence: small cohorts, single time points, and antibody titers that show exposure rather than active replication. How often these viruses actually switch back on, and when, stayed open. This study set out to measure active viral replication itself, repeatedly, in enough patients to answer both questions.

Just under half of the cohort, 47.9%, reactivated at least one chronic virus within 40 days, with detection climbing across all five severity groups for Epstein-Barr virus, cytomegalovirus (CMV), herpes simplex virus 1 (HSV1), and Anelloviridae. Among the critically ill, CMV in a respiratory sample was tied to higher one-year mortality, as was nasal Epstein-Barr virus. Neither immunosuppressive medication nor transplant history explained the herpesvirus reactivations.

Epstein-Barr virus was already detectable in about one in four patients during the first week after admission, then tapered. Anelloviridae held steady through day 20 before declining slowly. CMV and HSV1 surfaced later, clustering around days 19 to 23.

In convalescent samples collected months out, Anelloviridae was more common in patients reporting the persistent physical deficits of long COVID, including fatigue and disability. That's association that did not appear when the same viruses were measured during the acute period. CMV also tracked with renal complications alongside sustained rises in urea, raising the question of whether reactivation deserves attention in COVID patients with acute kidney injury.

Investigators sequenced viral RNA from nasal swabs, blood cells, and endotracheal aspirates across up to 10 visits over a year, in 1,154 adults hospitalized for COVID at 20 US hospitals from May 2020 through March 2021. Sequencing detects transcripts from replicating virus, so a positive result reflects active reactivation rather than past exposure. Participants were unvaccinated and predominantly infected with ancestral strains, which leaves open whether reactivation behaves the same way in patients with hybrid immunity or milder illness.

Cause or consequence remains an unsettled question, and only a trial can answer it. Much of the necessary apparatus already exists: antivirals for the herpesviruses, validated PCR for herpesviruses and anelloviruses alike, and metagenomic sequencing that keeps getting cheaper. Anelloviridae is a hard case, detectable but with no established pathogenic role and no directed therapy. The authors call for work dissecting what the family does in patients left with lingering physical symptoms.

"This association with long COVID is an interesting finding as millions around the world suffer from this chronic condition," said Ofer Levy, MD, PhD, director of the Precision Vaccines Program at Boston Children's Hospital and a site principal investigator for the study. "Having new insight as to the molecular and viral associations with long COVID could point the way to better understanding and ultimately better diagnostics and treatments."

Source: Maguire C, et al. (2026 Aug 5) Nature. Virus reactivation in acute and long COVID-19