THREAD: Big news today. For the first time, we have a confirmed report of SARS-CoV-2 reinfection by a patient in Hong Kong—his/her #COVID19 diagnoses were 4.5 months apart. This blows a big hole in the "let's stay locked down until there's a vaccine" position. Let me explain why.
We've been wondering how long immunity lasts after you've had #COVID19. Based on experience with common colds and other milder coronaviruses, we've expected somewhere around 6 months. This Hong Kong patient falls within that range. So why is it that immunity is so short?
A big reason for limited immunity is that coronaviruses mutate all the time, such that immunity to one version of SARS-CoV-2 doesn't necessarily confer immunity to another strain. We discussed this back in April in @FREOPP's paper on reopening the economy. freopp.org/a-new-strategy…
As we explain, "most SARS-CoV-2 vaccine developers are targeting the spike glycoproteins that dot the surface of the spherical coronavirus particle. But glycoproteins—a combination of sugar and protein molecules—are especially good at evading the human immune response." 
Sure enough, the @hkumed report states that the "viral genomes from 1st & 2nd episodes belong to different [genetic] lineages...amino acid differences can be found in 9 proteins, including a 58-amino acid truncation of the ORF8 protein." This is of great significance, because...
There's evidence that ORF8 (i.e., open reading frame 8 of the SARS-CoV-2 genome) is a key way that the coronavirus evades the human immune response. biorxiv.org/content/10.110… For the @hkumed patient, the truncated ORF8 may be why immunity from the first infection wasn't protective.
The good news is there's evidence that SARS-CoV-2 strains lacking a complete ORF8 are less able to evade the immune system, and therefore less virulent. virology.ws/2020/04/10/a-3… The @hkumed release gives us no info on the severity of illness (if any) in the reinfected patient.
Bottom line (and apologies for all the jargon). Bad news: vaccines may be very limited in their effectiveness because #COVID19 can be caused by many genetically different strains of the novel coronavirus. Good news: those mutated strains may be less lethal/virulent/dangerous.
The @hkumed report is a data point that illustrates the giant weakness in the "let's lock down until there's a vaccine" approach. We simply don't know when we'll have a vaccine that's fully protective against *every* dangerous strain of the coronavirus.
In other words: relying on vaccines to save us from dumb lockdown policies is itself dumb lockdown policy, because there is enough of a chance that vaccines aren't fully protective. Let's hope for the best with vaccine R&D, but also plan for the worst. freopp.org/covid19/home
Some people in the replies are asking: "What if the reinfection is a false positive?" That's highly unlikely in this case, because they isolated the coronavirus in the reinfected patient and closely examined its genetic code.
Some other people are arguing: "Well, doesn't this also blow up the strategy of herd immunity?" No, because herd immunity strategies don't lock down the economy or close schools. By contrast, keeping society closed until a vaccine comes along is extremely costly.
Muge Cevik says she has additional details on the @hkumed patient, who apparently was hospitalized with the first infection, but asymptomatic upon detection of the reinfection (was screened upon returning to HK from Spain).
If this patient remains asymptomatic, that's good news, as at least in this case it would mean that the second virus strain was not as dangerous in this specific patient.
More details from the upcoming publication submission, courtesy of @cwylilian:
In a @FREOPP article looking at international #COVID19 performance, I note that we've seen different viral strains in different parts of the world at different times. Probably the biggest reason why the Pacific Rim & the West have had different results. freopp.org/measuring-covi…
Some people in my timeline argue that we shouldn't pay much attention to the @hkumed story because it's 1 patient. That's not the point. The point is that the study shows how specific genetic changes in SARS-CoV-2 can evade prior immunity. That is significant new information.
Where the n=1 factor matters: as new strains of SARS-CoV-2 arise, will any of them be as virulent/lethal as the main ones we're dealing with now? The @hkumed case and other studies of ORF8 truncations/deletions give us reason to hope they won't be, but 1 case can't answer that.
But if immunity lasts ±6 months, we should see more of these cases over time. We can hope that newer strains are less virulent/lethal, but that will never be guaranteed. And if guarantees are what you need to stay locked down, you'll be locked down for a long time.
On the other hand, if we approach #COVID19 in the way we always should have—with a prudent and reasoned approach to risks, costs, & rewards—we'll do things like reopening schools, even as the pandemic endures. freopp.org/reopening-amer…
Some people are saying: "Well of course we know that a vaccine won't address all strains of SARS-CoV-2! After all, this is exactly the way influenza vaccines work. None of this is surprising." Correct, which is why we said the same thing in April: freopp.org/a-new-strategy…
The difference between influenza pandemics and #COVID19 is the *policy choices* we're making in response.
If we were making reasoned policy decisions, rather than panicked ones, we'd (1) focus our attention on protecting nursing home residents; (2) reopen schools; (3) pursue the Texas/Florida model of economic reopening. None of those things require a vaccine. freopp.org/the-covid-19-n…
What are those measures? "Better isolating [nursing home] residents"..."reopening schools"..."social distancing [taking] into account widely varying risks by age"..."masks"...but keeping bars closed. In other words, exactly what Texas & Florida are doing. wsj.com/articles/covid…
