Thread by @DiseaseEcology

Long term antibody dynamics & persistent RNA in #COVID19 patients & from Shenzhen, China.
tl;dr 93% developed long-term neut Ab (up to 17 weeks); RNA fragments, not infectious virus, not re-infection
Thread
medrxiv.org/content/10.110…

Background
The standard way we test for SARS-COV-2 infection is by nasal swab (or other tissues) & then look for viral RNA. Importantly, 99% of all testing does NOT test for live virus. That requires a specialized lab. This will be important in a couple tweets.

>95% of infected people test + in the first few days after they show symptoms (& possibly 1-3 days before, but poor data, so a negative test when asymptomatic doesn't rule out infection

https://twitter.com/DiseaseEcology/status/1270093309126438912

).
Viral loads & detection drop quickly after.
medrxiv.org/content/10.110…

In *most* people viral RNA becomes undetectable by ~28d after symptom onset & 2 - tests are sometimes used as evidence of clearance. However, we've known since April that in a small subset of patients RNA is detectable after recovery, sometimes after a period of testing negative.

This raised fears that people might either be chronically infected, or be able to be re-infected soon after their first infection. Recent reports of reinfection from NJ doctor fall into this camp.

This new paper looks at 479 patients that were discharged from hospital in China after clearing symptoms & testing - twice.
19% (N=93) of patients tested + at least once again for viral RNA (more likely in younger patients) & RNA detectable up to 90-112d after symptom onset.

Suggests that residual viral shedding for weeks after recovery is actually not uncommon (~10% of patients for wks 4-8 PSO; >5% for wks 3-12).
HOWEVER, there was no evidence that these patients were infectious or were shedding live virus.

They attempted to culture 9 samples; 0 had live virus. They also tested 96 contacts of 23 RNA persistent shedders by both swab and antibodies & 0/96 were infected. This indicates that patients w/ persistent viral RNA are likely NOT infectious. 2nd study to show this.

Data on antibody dynamics over longest time I've seen (17 weeks!). All normal caveats about how protective any immune marker is apply (

https://twitter.com/DiseaseEcology/status/1283282941729103872

). Data suggest that neutralizing Ab slightly rise over time & persist through 17 weeks (no diff for shedders/non-shed).

Paper proposes 1:32 dilution for neut Ab as cutoff for vaccine trials. 60% of patients exceeded 1:32 (but only 28% had higher 1:64 level).
Obv. need to know functional protection (infection, disease, infectiousness) of this & other immune markers for COVID19 to fully interpret.

Only 7% of patients didn't mount a neutralizing Ab response (T-cell responses not investigated)
6% of patients developed neut Ab response after initial discharge suggesting it can take time to detect this response as @florian_krammer has also shown. medrxiv.org/content/10.110…

Note that this study was done in Shenzhen where there was little transmission after initial epidemic so patients were likely not being re-exposed to virus (so nAb persistence w/out boosting from exposure).

Conclusions
-chronic RNA shedding for up to 3 mo is not uncommon (5-10%; more in younger patients) but is NOT infectious virus & is not due to re-infection
-Neutralizing antibody develop in >90% of patients & persist for at least 4mo