In this blog we’ll mostly take a 35,000-foot view of a number of issues (with some exceptions), and take a long, hard look at how the pandemic will progress, and finally end.
Basic epidemiology
With the delta variant (strain b1617.2), the virus upped its game. It developed in India as a result of uncontrolled spread in the spring of 2021 in a virus-naïve dense population that was minimally vaccinated. Its R0 is on a par with measles—something like 7 or 8, so it is incredibly infectious. My own calculations based on estimated deaths of 4.5 million and a change in viral seropositivity from about 28-68% show that the IFR (infection mortality rate) was around 0.6%. That’s a tad higher than the median IFR of 0.4% a year ago. There are some studies that do indicate a slightly higher IFR, notably the Scottish study, but in the USA we do not know for sure whether it is higher than the alpha variant (b117), which is a good comparator.
Mutation
Most of the viral evolution has been focused on improving the virus’s R0 from around 1.5 in late 2019. Since the introduction of vaccines, there has also been a notable increase in variants that elude vaccine targets, primarily the spike protein; the lambda variant, common in South America is a good example. This is mainly the result of genetic evolution pressure due to vaccines, but it’s not the only evolutionary pressure. One of the concerns I have had for some time is where the virus goes from here. Let’s take a look at its genome: it’s large for a virus coming in at 29.9 kb. While the press talks about the spike protein mainly, we forget that are 3 other structural proteins and 16 non-structural proteins (NSPs) (Figure 1).
Fig 1. the SARS-Cov2 genome. Schematic presentation of the SARS-CoV-2 genome Structure. SARS-CoV-2 has a spherical structure. The virus has an outer lipid envelope, covered with spike glycoprotein. The SARS-CoV-2 represents a typical Betacoronavirus genome organisation. The full-length RNA genome comprises of approximately 29,903 nucleotides and has a replicase complex (comprised of ORF1a and ORF1b) at the 5′UTR. The ORF1a encodes for nsp1–nsp10, while ORF1b encodes for nsp1–nsp16. Four genes that encode for the Structural proteins: Spike gene, Envelope gene, Membrane gene, Nucleocapsid gene and a poly (A) tail at the 3′UTR. The accessory genes are distributed in between the structural genes. Courtesy Rastogi et al, 2020 (https://respiratory-research.biomedcentral.com/articles/10.1186/s12931-020-01581-z).
In my opinion, one of the most important NSPs is 14, which represents a proofreader that is part of a huge replication complex. The vast majority of positive-sense single-strand RNA viruses don’t have a proofreader, which makes coronaviruses almost unique. It’s far from a perfect proofreader, so mutations can still occur during replication but at a much lower rate. If there was no proofreading, not much of the viral progeny would be viable, since the genome is so large. But, nevertheless, it is a “sloppy” proofreader, which has produced much speculation. A recent study has looked at it and the authors noted: “We find that proofreading nearly optimizes the rate of functional virus production…. We speculate that cost optimization [translation: higher proofreading rates versus energy costs to do that] could help avoid cell defense responses.” I find it creepy that the virus can optimize a process in which it essentially controls the mutation rate of its progeny and also defeat the cell’s immune responses.
Another NSP of interest is 12, which is an RNA-dependent RMA polymerase, and is involved in replication and methylation activities. A key mutation of NSP 12—323PL; swap of proline for leucine)—is perhaps one of the most lethal mutations associated with any NSP, likely because it increases viral replication efficiency. It is also commonly associated with the infamous D614G spike protein mutation. It’s also creepy because in the limited assays to date of viral isolates from deceased patients it’s nearly always present.
The point is, the more cases we have, the more the opportunity for viable NSP mutations that could increase viral lethality. If a new strain develops that is more lethal and evades current vaccines, the pandemic may be prolonged by many more years with many more deaths. Because history shows us that viral strains tend to become less lethal over time I think this situation is unlikely, but it cannot be ruled out.
Vaccines and Vaccination
Social media reports over the past few months suggest that the trickle of breakthrough SARS-Cov2 infections has become a flood in fully vaccinated individuals. But we need to take a step back and look and some recent studies to see what the numbers are and their implications. This week the CDC released a number of large real-world studies: study 1; study 2; and study 3. There are also the Israeli and UK studies. The bad news is that vaccine efficiency dropped some 10-30% for general populations but around 45% for nursing home residents. The good news is that efficiency at preventing hospitalization or death did not drop much at all and that means some 97-98% of vaccinated individuals won’t get hospitalized even though some may experience the equivalent of a flu bout if they get infected. We expected some drop in efficiency against infection over time and a steeper drop against variants such as delta, which can partially evade vaccines, but while the declines are nowhere as bad as many people have been parroting around the Internet, they are serious. We can also expect some further decline over the next year. So, our effective window to vaccinate one time is closing very rapidly.
Based on the CDC data the Biden administration has recommended booster shots. With the exception of the elderly and the immunocompromised, who might benefit from a booster shot, I am profoundly disappointed with this decision. First, in order to get this pandemic under control we need as many people as possible around the world to be vaccinated. Siphoning off vaccines for booster shots rather than sending them to countries that are still desperate for vaccines is morally reprehensible and scientifically wrong and sends the wrong message. Second, for many individuals, booster shots won’t do that much. To understand this, we need to look at a phenomenon called original antigenic sin (OAS).
We have known about OAS for over 65 years starting with influenza data. In a nutshell the antibody-mediated immunity achieved post-vaccination may not be fully specific to a distinct antigen variant contained in the vaccine since the antigenic determinants may be shared across various strains of the respective types or subtypes of the viral pathogen, as is the case with influenza. Most importantly, from flu vaccines we learned that while specific antibodies targeting antigens contained in the vaccine (i.e., particular kinds of neuraminidases and hemagglutinins) can be induced, a vaccination series targeting a new strain also raised the levels of antibodies specific to antigens produced in response to previous vaccination or influenza. The rate of production of the original antibodies can also be significantly faster. The current crop of vaccines were trained on pre-Greek letter SARS-Cov2 strains (i.e., prior to alpha), so booster shots will only get you more of the same, and since with delta we already know that they don’t work that well, you won’t get a different qualitative response. The best you will get is some quantitative boost to compensate for loss of efficiency over time from the first inoculation. I also worry that inappropriate use of same-same booster shots may increase the chance of ADE (antibody-dependent enhancement) occurring; we already are suspicious that this might have occurred to some extent in Brazil or possibly India based on people who had been infected with earlier strains and re-infected with third-generation strains. We certainly don’t need a vaccine-induced ADE problem.
Covid waves
In the USA, we’re seeing another wave, this time focused primarily on areas in the south of the country where vaccine uptake has been the lowest. The vast majority of hospitalized people continue to be relatively young and unvaccinated. This wave (see the end of Figure 2) will likely peak in 4-6 weeks and thereafter gradually decline until November when we could see a similar wave to 2020 albeit less lethal.
Fig 2. US SARS-Cov2 cases through August 18, 2021. Red line is 7-day moving average.
We are looking at stabilization—meaning a low daily case count of 5,000 or so—by the end of 2022 but it’s likely to be another year or more before the world achieves anything like that and a declaration that the pandemic will be declared over. In the end SARs-Cov2 will probably become seasonal with outbreaks much like flu. If, however we see development of more vaccine-evasive strains emerge, then all bets are off and stabilization gets further pushed ever further forward.
Personally, it’s taken a lot of grief to for me to find some balance to being risk-aversive and suffering from isolation and taking some risk so I can see friends and do some business travel. I don’t think endless lockdowns are the answer as no doubt countries like Australia and New Zealand will find out in their quest to achieve zero cases. But, I also take heart from the results in the UK where for now with over 60% of the country vaccinated they are open and finally experiencing declining cases.