Thread by @ZdenekVrozina

New interesting mouse study out of Barcelona follows K18-hACE2 mice for 60 days after SARS2 infection. The trick - a deliberately low dose, so most animals survive the acute phase and can actually be followed this long. The goal - catch what’s left once acute COVID clears🧵

They recovered from mild COVID. Two (mouse) months later - nothing in the blood, but persistent immune dysregulation in tissue and a shrunken vagus nerve. A standard blood draw would’ve sent them home healthy.

At the group level, infected and control animals didn’t differ on behavior. At all. The signal only shows up in individual z-scores - a subset of animals carries the damage while the rest look normal. It’s not all the mice got worse, it’s some got notably worse.

So what were those animals doing differently? Not moving less - locomotion was normal. They hugged the walls, avoided open spaces, explored less. More anxious, more cautious - with motor function intact. In females mostly in the open field, in males in the maze.

But most summaries leave out - the worst hit animals never finished the study. Around 40% had massive acute neuroinvasion - brains full of virus - and didn’t survive the acute phase. So study no virus in the brain holds for the survivors, not for all of them. What follows is what’s left after you subtract the worst cases.

Why the clean blood part matters so much. The immune dysregulation was in tissue - in the lungs and the brain - while blood cytokines never moved the whole time. Blood and tissue don’t match here. Measure only blood and you miss the process entirely.

It’s not smooth, smoldering inflammation. Brain cytokines were actually down during the acute phase, the rise (MIP-1α, IL-2, RANTES) only came at 60 dpi, driven by a subset. A biphasic picture. Acute wave - quiet - late return. The inflammation comes back .

The vagus nerve. Reduced cross-sectional area in infected animals, significant at the latest timepoint (60 dpi). Small sample, no histology, so we don’t know if it’s fiber loss, demyelination, something else.

But the direction matches what’s reported in PCC patients (Papadopoulou, Lladós) - dysautonomia, weakened anti-inflammatory signaling.

The sex difference tracks the clinic. Males worse acutely, females worse long-term - more behavioral deficits in the post-acute phase. Exactly the pattern epidemiology describes in human PCC.

It fits a line of work. Mild respiratory COVID - neural dysregulation via lung chemokines (Fernández-Castaneda).

That second picture - CNS inflammation running independently of active replication, driven by myeloid and T cells - isn’t new. It’s how HIV/HAND works. The virus doesn’t need to be actively replicating in the brain for inflammation to do damage there. Persistent antigen, a broken barrier, immune momentum. SARS2 is walking a well-worn path.

It’s a mouse, the K18 promoter drives the receptor into neurons, so not a picture of human infection. Samples small, the behavioral signal is subset-level, and 60 days ≈ human years is a rough conversion. None of this refutes the headline .

The core holds, and it points out of the lab. Mild-course animals, healthy at the acute stage, carry measurable tissue and structural changes long afterward - with completely normal blood. Give them just a routine draw and they pass as fine.

This is exactly what happens to people. Routine workup finds nothing - there’s nothing wrong with you - patient written off.

Recovery from the acute phase and resolution aren’t the same thing - and this study shows an objective, blood-invisible gap between them - then a system betting that mild COVID means resolved COVID is doing the math wrong.

Ruiz-Casas at al., Persistent neurological and behavioral alterations after SARS-CoV-2 infection in an optimized K18-hACE2 mouse model. frontiersin.org/journals/micro…