Towards a General Theory of Corona and Why No Interventions Ever Seem to Work As Intended
Our attempts to manage the pandemic have presented us with a variety of problems and paradoxes: Why have lockdowns and other containment measures proven so generally ineffective against SARS-2? What causes waves of infection to begin, and why do they collapse long before all the susceptible have been infected? What is the significance of asymptomatic infection? Why do the vaccines seem to boost infections for a few weeks after the first and third doses? Why did the SARS-2 spike protein begin to experience heavy selection pressure in the latter half of 2020, giving rise to variants like Alpha and Delta? How did these variants manage so completely to replace prior lineages?
Increasingly, I wonder whether all of these questions aren’t just facets of the same phenomenon – one that was first described by R. Edgar Hope Simpson in his fascinating book on The Transmission of Epidemic Influenza (New York, 1992). As I said on Twitter, Hope-Simpson is one of the only people to have attempted what might be called a zoology of viruses. That is to say, he tried to analyse and explain how influenza interacted with its human hosts more broadly, and not just how it infected cells and responded to antibodies. This is precisely where modern epidemiology and virology are most deficient, with extremely simplistic understandings of how viruses behave across the population.
Hope-Simpson wrote to explain a twofold problem: On the hand, influenza as studied by scientists seems to be hardly infectious at all, with very low observed household attack rates, and strange failures of challenge studies to actually cause infection in volunteers. On the other hand, seasonal and pandemic influenza viruses spread rapidly across the globe every year, often in highly synchronised waves, totally undeterred by the lack of observed infectiousness. Hope-Simpson therefore concluded that influenza transmission was a far more subtle matter, than sick people infecting the healthy.
He proposed instead a two-phase process:
1) Acute infection, lasting a week or two, during which influenza is only transmitted very rarely.
2) Latency, whereby a subset of recovered people fail to clear the virus completely, and instead become asymptomatic carriers for many months, until seasonal factors correlated with latitude cause them to become infectious once again, generally without any noticeable recurrence of illness.
For Hope-Simpson, spread via symptomless carriers who had been infected in a prior wave was the dominant mode of influenza transmission. If he’s right here, we’re permitted to wonder whether the same strategy might also be available to other viruses, like the one we’ve spent the past years losing our minds over. There is substantial evidence of transmission during the acute phase of Corona infection, but this doesn’t exclude the possibility that the virus also depends in some substantial way on longer-term phase 2) carriers. In fact, I’m increasingly tempted to think that it does, for this would explain almost all of the questions posed in the opening paragraph:
The failure of lockdowns: This is hardly a mystery, if we imagine that some significant portion of SARS-2 infections have already been seeded months ago in the prior wave, and these latent carriers are simply being re-activated by environmental factors. Indeed, lockdowns often had a curious property, of seeming to simply move infections around rather than stop them. Closing one sector of society would simply shift the case burden elsewhere.
The spontaneous rise and fall of infection waves: As with influenza, we’d simply posit that a significant part of infections are triggered by environmental stimuli that cause the carriers to transmit. Waves lose energy after all of the latent carriers have done their infecting.
Asymptomatic infections: Some subset of these people would be carriers in the latent phase who have recovered from a prior infection, perhaps with mild symptoms they hardly noticed.
Surging infections following vaccination: Perhaps the vaccines work much like the environmental stimuli that trigger seasonal waves, by temporarily reducing the immune response of latent carriers and causing them to become infectious.
The variant SARS-2 lineages: This is where the theory begins to offer very intriguing solutions indeed.
Hope-Simpson found traditional explanations for the antigenic drift of influenza unconvincing:
The current belief holds that antigenic drift occurs because of “herd immunity,” that is, the immune pressure on the virus of a partially immunized human community. On that view it is the recipient who is seen as exerting the proposed immune pressure on the virus. … In a partially immunized community the minor variants of the virus are supposed to find themselves at a selective advantage over the previous prevalent strain and so will tend to replace it.
This superficially attractive hypothesis cannot survive closer scrutiny. The “partially immune community” is not a community of partially immune persons, but a community in which some persons have become immune and the others are nonimmune. Second attacks are relatively uncommon, so the recipients who develop influenza and from whom the drifted variant is isolated are persons in the nonimmune portion of the community who could not have exerted the proposed immune selective pressure to cause the antigenic drift. (pp. 97–98)
Put a different way, and with different emphasis for our rather different virus: Only a very small minority of people are ever infected in a Corona wave; and before Omicron, reinfections were extremely rare. Alpha might have spread faster than wild-type SARS-2, but how did it actually succeed in wiping out these wild-type strains? Hope-Simpson calls this the “disappearing act,” as it is also observed with successive strains of influenza, and is very hard to account for.
Hope-Simpson finds the beginning of an answer in old influenza A antibody experiments:
During the great influenza A epidemic of the 1950–51 season, Isaacs in London, England and independently Archetti and Horsfall in the United States made a seminal observation. Two minor variants of the A (H1N1) subtype, “Scandinavian” and “Liverpool,” were co-circulating in many parts of the world. In both laboratories it was discovered that if the Scandinavian strain were made to infect a fertilized chicken egg in the presence of Scandinavian antibody, the strain harvested a few days later would turn out to be Liverpool. Vice versa, an appropriate dose of homologous antibody induced Liverpool virus-infected eggs to yield a harvest of Scandinavian virus.
The findings have been repeated with other strains of influenza virus in many laboratories throughout the world. (p. 99)
Some influenza strains, in other words, emerge spontaneously, via mutation, as a response or solution to the antibodies generated against other strains. Thus Hope-Simpson held that selection for new influenza strains happened primarily in the latent carriers of the virus, after acute infection.
Over months, the virus that lurked in carrier tissues was able to train itself against their immune response. Various mutant strains arose, and when the carriers became infectious again, the most contagious of these mutants went on to infect other people as the new influenza strain. Because the set of optimal solutions was in every case limited, all the latent carriers infected with the prior strain ended up incubating the same set of successor strains.
Note that this is exactly the same explanation modern science provides for the emergence of the variant SARS-2 lineages, with the proviso that these prolonged training sessions are occurring in immuno-compromised people. The distance from current orthodoxy to crazy Hope-Simpson theory is in this case very short. We just imagine that some subset of ordinary people end up carrying the virus between waves, and that subsequent strains emerge in their tissues during the latent period. This makes the emergence of the first variant strains, around the time of the vaccine trials, potentially highly intriguing. These strains emerged as one product of the spike protein – and the spike protein alone – experiencing heavy selection during these months. The vaccines, of course, induce immunity against the spike protein alone.
While I’m not sure Hope-Simpson’s theory is precisely right, I’m increasingly convinced that something roughly like this must be going on. The standard SIR model can’t account for the behaviour of SARS-2, and the vaccines have had an array of unexpected effects on transmission and infection dynamics that suggest they’re interfering in a multi-phase process.