Amie and I walk the dog almost every day. It is a highlight of our calendar (well maybe the scotch at happy hour is more highly awaited), and a chance to get away from the news. Our conversation today centered on the production of vaccines and on how we will be able to use the virus tests and the antibody tests that are being developed.
I thought some discussion of what we can expect from a vaccine, and how we will be able to use the new tests as we attempt to open up society again would be valuable.
As I always try to do, I will be honestly blunt. Politically we ALL want to solve this problem, get it behind us and move on. Scientifically, we need to understand what we will actually expect to get and what we actually won’t.
Vaccines are quite effective in preventing disease, but NONE of them are 100% effective. What that means is that for people immunized against a specific viral agent, less than 100% of those people will actually generate a protective immune response to that virus. For the childhood diseases that we are immunized against, only about 85%-90% of the immunized are actually protected from the disease. Some of the vaccines, as you know, require “boosters”, second or even third challenges with the vaccine, in order to generate protection above 90%. As examples, two doses of the MMR vaccine (measles, mumps and rubella) will convey over 99% immunity, three doses of polio vaccine will generate 99% protection, but chicken pox is only around 85% effective and Influenza vaccines are more like 75% protective.
If vaccines are not totally protective, why do we vaccinate populations? This is where the term “herd immunity” arises. If you are at a party with 100 people and there is no immunity in that population, and if the R0 (the rate at which one infected person infects others) is say 4, and if there is an infected individual, there will be a 100% chance that that person will infect other people in close contact. If 50 of those people have immunity, the chance that the infected person will infect another drops because the chance that they would pass the infection on to another person is reduced by a chance of encountering a vulnerable individual. As the amount of immunity in the room increases, there becomes a tipping point in which it becomes statistically unlikely that the infected individual will pass on the infection. That is why it is ok if the measles vaccine isn’t 100% effective or if less than 100% of the population actually gets it. There will still be herd immunity to the general population unless the amount of immunized people drops below the tipping point described above, in which case the virus breaks out.
For our discussion, it is ok if the vaccine for COVID0-19 is not fully effective, if we are able to see limited breakout infections that can be quickly identified, isolated and contact-traced. That is what we would consider a “manageable” infection.
But, will we be able to produce a vaccine?
I remain cautiously optimistic, but as I have written on numerous occasions, it is not an absolute. Many viral infections have resisted successful vaccine development. There are lots of reasons for this.
- A vaccine may generate an immune response, but that response may not be “neutralizing”, meaning it doesn’t convey protection.
- Sometimes a vaccine that has proven effective in an animal model has the exact opposite effect than what you want, actually increasing the infection. This is termed “vaccine enhancement” and although rare, it does occur, and it becomes really important to make sure that your new vaccine doesn’t cause this.
- The vaccine itself is a protein, pieces of protein or some form of RNA plasmid that causes existing cells to express the protein. In all cases there is the presentation of a new protein to the body, and it is important to remember that the protein was something that the virus uses to attach to human cells. Sometimes, that protein itself may cause problems in human organs, because, after all, you are injecting people with a foreign chemical. It is critical to make sure that there are no harmful side effects to the injection of a vaccine into people.
- How many days will need to pass after a vaccination before the patient is protected from viral challenge? We know that it takes between 5 and 10 days for an immune response to appear; we can assume that it will take at least a week, and probably two, before a vaccinated person is actually safe from infection.
- It is always difficult to predetermine how long any protection that is generated will last. Months? Years? Decades? Life? These statistics will only be available retrospectively.
Nevertheless, we can be optimistic that a vaccine will be created, tested, produced and distributed. We can then be optimistic that an immunized population will provide sufficient herd immunity to allow the healthcare system to identify breakouts and deal with them efficiently.
Before a vaccine we will have testing. How will that help?
Let’s go back to that party with 100 people. Let us assume that 40 of those people were tested for the presence of virus 2 weeks ago, 25 were tested last week, 25 were tested today and 10 have not been tested. What do we know ourselves about the people we may be bumping into at the party?
- Of those not tested, we know nothing about whether they are infected.
- Of those tested 2 weeks ago, we know nothing about whether they were infected since then.
- Of those tested last week, we know that at least a week ago, although they may have been infected, they were not testing positive yet because positive demonstration of virus may not occur for up to 14 days before symptoms show.
- Of those tested yesterday, there is a better chance that they are virus free, although they may be infected, but not yet shedding detectable virus.
Ok, but what if we check for antibodies instead?
We currently believe that about 1% of the population has been infected and that at least 80% of those were either asymptomatic or had mild disease. Those people as well as an additional 12% of those infected who never presented to hospitals will, presumably have antibodies in their blood. The other 99% of the population who has avoided infection either because of fortune or due to their observation of self-isolation will not have antibodies.
So, we would expect that testing for antibodies would only show up in 1 person out of the 100 at the party. Not a lot of help there.
The other problem with antibody tests is drawing conclusions about infectivity along with the presence of antibodies. The best example here is HIV. The early tests for the presence of the disease were, in fact, antibody tests. The blood tests looked for antibodies to the AIDS virus, demonstrating that the individual was infected. However, that diagnostic test DID NOT mean that the person was now personally protected from infection. The presence of determinate antibodies neither indicates that the HIV patient is now virus-free, nor that he/she is non-contagious.
I know that this explanation might lead us to presume that testing is useless. Fortunately, that is not the case. There are several very important uses of testing. Remember that current testing for virus is done at hospitals to determine whether you need to be treated for COVID-19 or not, and in order to protect the healthcare workers.
First, if you were tested for the presence of virus a week ago and you were negative, and you are tested again today, we can be pretty sure that you are not infected
Second, if you have been tested for antibodies and were found positive, and you have been asymptomatic for 14 days, we can again be pretty sure that you are not only non-infective, but that you are, most likely, protected against future infection.
Third, if the antibody test is used to evaluate a well-described population, say an area with little current disease, or a group of factory workers, or a school, then you can determine how many people have actually had the disease and recovered although their symptoms were mild or non-existent. This is incredibly important data for analyzing the epidemiology of the disease.
Fourth, if a population is tested and there is no positive test for virus, and there is a good statistically valid random test for antibodies, finding no more than one would have expected in any random sample, then it is a strong indication that opening that group to social interactions is safe. Nothing is absolute, but if the infection rate is very low, then identification, isolation, contact tracing and treatment will manage the disease.
Finally, and this is probably the most important piece, once that statistical data is established, follow-up random testing will be able identify CHANGES in the presence of virus or the detection of antibodies in those populations. As long as those changes are either negligible or decreasing, restrictions can continue to be removed. However, if the new tests show a statistical increase, we will know that restrictions need to be reestablished in order to control any new expected outbreaks.