Thread by @jbloom_lab

Excited to contribute to this extremely important #SARSCoV2 study by @GreningerLab that provides first direct evidence that neutralizing antibodies are a correlate of protection against #COVID19 in humans! (1/9)
medrxiv.org/content/10.110…

Perhaps *the* most important question in the #SARSCoV2 field right now is what immune responses protect against re-infection in humans. Knowing the answer to this question is critical for vaccine design and epidemiology. (2/9)

Lab studies show neutralizing antibodies are protective in animals, but there isn't comparable data in humans. To address this, large & expensive studies are currently being set up.

But @GreningerLab came up with a creative way to start answering this question right now! (3/9)

In May, a fishing boat with 122 passengers on its manifest left port in Seattle. To avoid a #SARSCoV2 outbreak on board, nearly all crew (120 / 122) were tested both for virus and for anti-N antibodies (Abbott assay) in tests run by UW Lab Medicine. (4/9)

Unfortunately, there was still a #SARSCoV2 outbreak on board, and 103 of the 120 passengers with pre-boarding serology got infected. @GreningerLab realized it was possible to look back at pre-boarding serology to see if pre-existing antibodies affected who got infected. (5/9)

6 of 120 passengers were positive for anti-N antibodies pre-boarding, & @khdcrawford @AdamDingens in our lab determined just 3 of 6 had anti-Spike & neutralizing antibodies. (Of other 3, some may have been false positives in anti-N screening as low signal in Abbott assay) (6/9)

So assuming that only individuals positive in anti-N Abbott assay have neutralizing antibodies (a good assumption, see pre-print for rationale), that means that 3 of 120 had neutralizing antibodies, and 117 of 120 did not (7/9).

During the outbreak, 0 of 3 with the neutralizing antibodies were infected, and 103 of 117 without neut antibodies were infected. This is highly significant by Fisher's exact test.

Therefore, neutralizing antibodies are correlate of protection in humans in this outbreak (8/9).

Limitations:

- neutralizing antibodies are *correlate* of protection, can't be sure they are causal vs some co-occurring immune response.

- Only 3 individuals in neut antibody group and we can't be certain they were exposed, but result is statistically significant. (9/9)