In 11 of 74 Autopsied Hearts, the Coronavirus Was Still Replicating
The abstract calls it evidence that the coronavirus can still be replicating in heart tissue months after someone catches COVID, and that it lines up with the symptoms doctors label cardiac long COVID. It's also just a conference abstract, and the antiviral trials built on exactly this theory have already come back negative.
Pathologists autopsied 74 people who had lived with COVID for at least 60 days before they died. In 11, the virus was still replicating inside the heart, and those 11 were far more likely to have died with the chest pain, palpitations, and breathlessness of cardiac long COVID: 82% versus 37%.
The finding is a conference abstract, not a published paper. A physician in the thread flagged the catch: the same theory was already tested with extended Paxlovid in real long COVID trials, and both came back negative.
Live virus, not debris, turned up in eleven of seventy-four hearts
At this year's USCAP pathology meeting, researchers presented an autopsy series of 74 people who had lived with COVID-19 for at least 60 days before they died. Left ventricle tissue from every case went through NanoString transcriptomic analysis, immunohistochemistry, and in-situ hybridization for the viral spike gene, testing specifically for the reverse strand: the molecular signature of a virus actively copying itself, distinct from inert genetic fragments left behind after an infection clears.
Eleven of the 74 hearts tested positive for that reverse strand. In those eleven, 82% had documented cardiac long COVID, palpitations, chest pain, dyspnea, fatigue, compared with 37% in the virus-negative group (p=0.0075). The virus-positive hearts were also more often enlarged, with left ventricle dilatation in 64% of cases versus 24%, plus more pericardial fluid. A differential gene expression analysis found 44 altered genes, with interferon regulatory factor 4 standing out among them.
The authors' own framing: persistent viral replication may be driving the structural and immune changes behind cardiac long COVID symptoms, with "implications for future diagnostic and therapeutic approaches." One caveat belongs up front. This is a conference abstract presented at a pathology meeting, not a peer-reviewed paper: the full methods aren't published, the statistics haven't been through journal review, and every one of the 74 cases is a death, so the study can't say what fraction of living long COVID patients carry the same signal.
Why a physician in the thread wasn't ready to celebrate
The HN thread stayed thin, 24 comments, but one of them did the most editorial work the abstract itself couldn't: a reader who lists himself as a practicing physician and AI SaaS founder walked through why the finding is promising but not proof of anything actionable yet.
"This abstract only reports a cut-off of >60 days from COVID infection, so it's hard to know how many of these deceased patients are truly outside the acute infection phase." — a practicing physician and AI SaaS founder
He then pointed to the trials that already tested the obvious next step: if the virus is still replicating, would killing it with antivirals fix long COVID symptoms?
"The major anti-viral trials for long COVID were negative. STOP-PASC trial and PAXLC were both 15-day courses of Paxlovid and showed no benefit across patient reported outcomes." — the same physician
Both trials, one published in JAMA Internal Medicine, tracked patient-reported outcomes after extended courses of Paxlovid and found no improvement. He added one more asterisk: the biomarker those trials measured, circulating spike protein, isn't the same thing this autopsy study detected. Spike presence in blood doesn't prove the virus is still actively copying itself the way the reverse-strand signal in heart tissue claims to show.
A founder's reversal: from "what doesn't kill you" to "avoid it entirely"
A separate voice in the thread, the founder of SignNow and tidy.com, used the autopsy finding to describe a change in his own thinking about infection generally, not just COVID.
"I used to think that it was kind of good to train your immune system (kind of whatever doesn't kill you makes you stronger). I no longer believe that. I believe that diseases often cause unknown effects and it's better to avoid disease entirely." — founder of SignNow and tidy.com
The physician, replying, tied the same finding to a wider pattern beyond COVID:
"Viral disease is being associated with more and more long term diseases. EBV is being linked to MS, cervical cancer to HPV, and so on." — the same physician
Those two links are established, not speculative: the Epstein-Barr virus is now the leading suspected trigger for multiple sclerosis, and HPV causes nearly all cervical cancer. What the thread's physician was doing was placing this abstract inside that pattern rather than treating it as a one-off. It's a reasonable place to file 11 hearts out of 74. Not a treatment, not yet a diagnosis, but one more data point in a list of chronic diseases that turn out to have had a virus at the start of them.
Get breakdowns like this every day
The top Hacker News stories, unpacked — by email, Telegram, or RSS.
A story blaming modern decor for headaches misread its own source study
StudyFinds ran a piece claiming striped floors and flickering LEDs are quietly overloading people's brains, citing a new review in the journal Vision. Commenters went and read that review themselves - and found the site's central science claim stated backwards.