I suspect you came here all the way from Amanz. Anyway, happy new year. I hope it has been going on great for you.
Those who have been following me closely as of late, it is not a secret that I have been writing article on COVID-19 for publication in Bahasa. If you thought this was it, the answer is “no”. In fact, this vaccine fact sheet was something I thought on Thursday (7 January 2021) after going through some random online discussions about vaccines for COVID-19. This also means there is potentially another one coming up if I am keeping up the pace well.
As of Friday (8 January 2021), after 5 pages and 1.8k words written, and with the help from my dear friends, Miji & Yan plus few others, I thought it reached feature completeness for publication. Then, I started writing this editorial to go along with it. I was happy knowing that I spent about a day to get this one up. That did not happen quite often in my long-reads writing career.
Just a quick reminder that viral immunology and vaccinology are the fields that I consider dear to myself and it is my turf, so if you dare to fight me with nonsense I would be elated to stuff you inside a flow cytometer and shoot you with 4 colorful lasers. And that is not the worst part yet.
Carl Zimmer tweets current status on vaccine clinical trials daily, linking his tweet to NY Times Coronavirus Vaccine Tracker page. This page is very helpful. I could find information on efficacy results and emergency use authorizations (EUAs) here with relative ease.
I decided to only include 3 manufacturers: Pfizer/BioNTech, Moderna/NIAID, and AstraZeneca/Oxford. Exclusion of Russian’s Gamaleya & Vector Institute and China’s CanSino, Sinopharm, Sinovac, & Sinopharm is because I could not find more information on their products. Ideally, I would like to see their phase I/II and their phase III data. For example, recently China’s Sinovac (vaccine name: CoronaVac) reported over 50% efficacy from their trial done in Brazil, but withheld full results (as of writing). As covered by Nature blog article, CoronaVac (killed/inactivated SARS-CoV-2 virus vaccine) so far showing inconsistent results, probably because the study protocol is not available on the public domain (as of 23 Jan 2021).
In contrast, for the main 3 vaccines:
- For Pfizer/BioNTech, here is phase I/II trial data (Mulligan et al. 2020), then here is their phase III trial data (Polack et al. 2020).
- For Moderna/NIAID, we can look at their phase I trial data here (Jackson et al. 2020), and then their FDA briefing document here.
- For AstraZeneca/Oxford, phase III data is available here (Voysey et al. 2020).
All the information are available for public to look at. The whole thing is like reading a beautifully written novel, except with a lot of alphabet soup.
Anyway, if you did see that I have 2 bars for AstraZeneca/Oxford, one at 62% and another at 90%? Well, it was fun reading about that. It gives me sort of assurance that I will have (stable) job after getting a PhD, hopefully.
Going through the responses I have received on both the main website (on Amanz) and tweets directed to me, I noticed a mistrust issue of the general public against the pharmaceutical companies. I understand this mistrust. I am aware of some misdeeds conducted by the pharmaceutical companies. As far as the scandals go, most of the seem to be in the domain of non-communicable diseases, i.e. disease that will not get transmitted between human-to-human like infection does. In other words, most of them kind of related to metabolic-related diseases.
Here is the problem comparing between communicable (infectious) vs. non-communicable diseases: you can’t. For non-communicable diseases, take high blood pressure for example, you really do not have a single cause of disease. You also do not die in the same year or month you are first diagnosed as you tend to live with it generally long enough, and once diagnosed and depending on the severity, medical intervention are usually to slow/halt the progression (usually coupled with life style changes).
So, what does this mean?
It is surprisingly challenging to measure the efficacy of intervention because the nature of these metabolic-related diseases being so multifactorial. To further speculate, drug companies are sort of (for the lack of better term) pressured to get drugs out to the market as soon as possible (something that has to do with patent expiration, I think), because the longer it has been patented but not marketed, the shorter the window to make return from millions (?) of investment, so on and so forth.
Communicable diseases like infections (for example, COVID-19), is entirely different. You usually have a single cause (the virus), we know how to make the vaccines, we know exactly the mechanism to induce protective antibody, and we know exactly what to measure, and we know what bad things to expect (e.g. allergic reactions). On top of that, getting infected has worst outcomes than getting vaccinated.
With COVID-19 and most vaccine designs, the genetic code for making these vaccines are publicly available. Moderna/NIAID and Pfizer/BioNTech publicly share their formulation (there are in their paper, FDA briefing documents, etc., I am too lazy to link them). In other words, you could use their formulation. However, the question now is do you have the right equipments with proper biosafety level? Or can you be sure that your formulation isn’t contaminated? This simple answer has 1 thing in common: a lot of financial investment up front. Maybe I should write about how expensive science could get.
To say we cannot trust big pharma in the face of people dropping dead and millions are infected, is a moot point.
Another contentious issue is how vaccination is viewed through Islamic rulings. I do not like explaining this part, because really it is not my job. It is the job of religious clerics (imams). However, I feel like some of them have abdicated this very important role in favor of spreading misinformation, and then go off and say “wallahu a’lam” (truly, God knows better). It is just plain lazy.
As reported by Quartz (paywalled), UAE and Bahrain have vaccinated 10.99% and 5.25% of their population as of 11 January 2021. This is because of a few contributing factors, chiefly the population sizes are smaller and the issues with logistics are less apparent. But not just because of these two. The imams communicated this early on. For example, UAE Fatwa Council chaired by Shaykh Abdallah bin Bayyah issued that COVID-19 vaccination program is allowed. Furthermore, in the report,
“Coronavirus vaccination is classified under preventive medicines for individuals, as recommended by the Islamic faith, particularly in times of pandemic diseases when the healthy happen to be prone to infections due to the high risk of contracting the disease, therefore posing risk to the entire society,” the Council explained.
The Fatwa Council added that even though the vaccine in question contains non-halal ingredients banned by Islam, it’s permissible to use it in implementation of the Islamic rule that permits the use of such products in case there are no alternatives.
Furthermore, the Syakh Abdullah himself received the jab publicly. That is how you, imams, build public confidence.
When we talk about vaccine profiles, we usually mean its safety, efficacy, immunogenicity, durability, and reactogenicity. I thought about teasing apart each of this profile but I was like “hm, I am not writing a textbook here”. Instead, I opted to mostly focus on its safety profile together with efficacy, and saved the reactogenecity later to explain allergic reaction and adverse events. In bullet points, when we talk any of the profiles:
- Safety: It does not make people to become so sick
- Efficacy: Does it work the way it is supposed to work?
- Immunogenecity: Does the body mount appropriate immune response after being administered (e.g. Th1- or Th2-skewed, neutralizing Abs, etc.)
- Durability: Does it generate and maintain immune memory (e.g. memory B, memory T, LLPCs, etc.)
- Reactogenecity: Any adverse events and how bad are those?
What I did not mention in the text regarding the decline in popularity of inactivated and attenuated vaccines is because the biosafety level requirement for level 3 pathogens. Currently, SARS-CoV-2 is a level 3 pathogens. To generate inactivated or attenuated vaccine for SARS-CoV-2, you need access to biosafety level 3, which is expensive to build and expensive to maintain. Now, you need industrial-grade level 3 for large scale vaccine production on the scale of millions of doses. That sounds like a lot of money now. Arguably, probably vectored vaccine is cheaper due to lower biosafety requirement at level 2. Lower biosafety level means lower restriction, lower cost, and generally more accessible.
Another reason to not use attenuated or inactivated SARS-CoV-2 virus for vaccine production is because we are (kind of) lacking great tools to manipulate the genome of the virus through reverse genetics. With influenza A viruses, we have great techniques to modify them with relative ease and speed. On top of that, SARS-CoV-2 has a very large genome at 29-31 kilobases, and that alone poses challenge to modify them. On the other hand, vectored virus (e.g. adenovirus) has genome that is small but good enough for it to be a good vaccine platform (with few caveats, of course).
At first, I thought about comparing it with the development of dengue vaccine (Dengvaxia), which I covered here in 2017. But then I thought “wow, what a bad comparison to make”, because the concern with Dengvaxia is something completely unique and something that is not very likely to happen with COVID-19 vaccines. That is the problem of “original antigenic sin” (OAS), and the reason why second infection with heterosubtypic dengue can be fatal.
Then I decided to compare it with mumps and measles, which I think would be more comparable.
Regarding trial duration, the concern about speeding it up way too fast has its own merit. The scientific community feels this way too. But, with great danger comes great effort to make sure we could put the flame out. I think in our lifetime, there is no greater feat than seeing these pharmaceutical companies managed to enroll tens of thousands of people into their trial. Also, the trial design is interesting: a multi-site system where each site enroll about hundreds to thousands of trial participants. It is a horizontal scaling, the way nowadays we build our cloud computing architecture. I might cover this one later in the other article!
I also thought about including the phase III interim result showing the symptomatic infection data, but I was like “this would require 1 or 2 paragraphs to explain” so I scrapped that plan and might write about that later in a separate article exclusively covering on study trials.
Astute readers might be wondering “hol' uppa minut, how did they get that many trial participants?”, well I do have an answer for that. Dr. Eric Topol tweeted that phase 3 vaccine trials were done at COVID-19 hot spots, which would make sense for 2 main reasons: you get the study done very quickly and you could potentially save people while doing that as well. And you know what makes this really difficult? First, you can’t model short-term human behavior that accurately, but you could statistically reason yourself in places with higher concentration of people with a sizeable proportion of people that will not follow physical distancing rules, that would be one of your best bets. So, it is not surprising if large cities are being picked because you have large population and most of them are healthy young adults (which are also very likely to not follow rules for many reasons). Second reason why this is so difficult is that… you have a lot of competition (refer: WSJ’s article on competing to enroll subjects).
Vaccine interim endpoint analysis
This part that scared me the most because language/wording matters, and I have seen a few instances where the interim endpoint analysis (as mentioned in the formal trial protocol) being weaponized to dissuade people from being vaccinated.
As stated in the study protocols (for Pfizer/BioNTech, Moderna/NIAID, and for AstraZeneca/Oxford), the interim endpoint analysis mentions about measuring the efficacy of the vaccine to prevent hospitalization or severe disease. First of all, let’s break these down:
- Interim: Because there are other parts of the study that are still not fully investigated, and being investigated separately after the publication of interim efficacy data, for example (AstraZeneca/Oxford’s single-dose trial by Ewer et al. 2020).
- To prevent hospitalization: Because among the main goals is to reduce strain on the healthcare system. If healthcare resources are stretched too far out, a lot of life will be lost. Not just the severely infected ones, but those who needed access to ICU but could not due to limited capacity.
- Study protocol: When submitting documentation for trial, a study protocol has to be put in place on exactly every facet of the study. I cannot even begin to describe how exhaustive they are, but you can read it here for Pfizer/BioNTech (available on ClinicalTrials).
As mentioned, once a trial has gathered enough symptomatic infection events to determine efficacy, an interim report will be published. This does not mean the trial is fully completed. It means “okay we have enough evidence that this is working”. In other words, manufacturers are still hard at work to piece together all the data pertaining to other aspects of their vaccine. For example, Moderna/NIAID submitted a document to support efficacy against asymptomatic infection (available on FDA website, see page 6, also here is a slide on other analyses they have planned).
Note: Thanks to Dr. Jesse Bloom for asking this question on Twitter. The responses were helpful, especially the link to FDA by Dr. Florian Krammer. Also, I met with these 2 experts before, and I had a virtual lunch with Dr. Jesse Bloom when he virtually came over to Rochester for a seminar.
Vaccine dosage and scheduling
I covered the bare minimum here that you need all 2 doses for most vaccines in advanced phase or the ones that are approved already, while excluding Johnson & Johnson’s single-dose Ad26.COV2.S vectored-vaccine (Sadoff et al. 2020, medRxiv).
So, why 2 doses? While I went over spewing name of difficult immunology concepts to the length of mentioning germinal center (GC) reaction and generation of antibody & plasma cells, I did not include somatic hypermutation (SHM) and affinity maturation in my explanation. Arguably, SHM and affinity maturation are the important ones when describing the need for 2-doses regiment.
This 2-doses regiment is also known as prime-boost regiment. The first injection is to prime the immune system. It would result in production of (polyclonal) antibodies and B & T memory cells. After a few weeks, the second injection would further boost the response by extending (I am waving my hand here while using that word) the duration of the GC in the local draining lymph node, or coaxed memory B cells generated from the prime to get activated for another separate round of SHM and affinity maturation. As a result, this would further lead to more SHM and producing BCR/Abs with higher affinity towards the antigen, resulting in more potently neutralizing immunity. Yeah it is difficult to simply explain it without turning the whole discussion into a textbook.
Now, note that the scheduling for the main 3 ones are different:
- Pfizer/BioNTech: 3 weeks apart
- Moderna/NIAID: 4 weeks apart
- AstraZeneca/Oxford: 4 weeks apart
Why the difference? Ideally, when designing a study you would include multiple groups where each would receive different dose. If you have 3 different doses already, and if you want to test every possible permutation, you would have 9 cohorts. Throw scheduling into the mix, you would see quickly how it would turn into a nightmare. For convenience but still adhering to the biology of immune response, 3-6 weeks scheduling is usually ideal. Sometimes, the decision is based on logistics, i.e. Moderna/NIAID’s vaccine is stable at 2-8 °C for 30 days, so spacing it 4 weeks apart is ideal. Another reason is that, really, 1-month is ideal because stretch it longer than that, you are likely going to forget to go back to the clinic/pharmacy for a second visit.
Note: Recent episode of This Week in Virology (TWiV) discusses vaccine dosage, see TWiV 703.
Distribution and logistics
Ugh. Another nightmare. And let’s not talk about vaccine hesitancy.
Ideally, vaccines need to be as widely distributed as possible. Ideally, everyone should get it. But in reality, and reality is always the force acting on an opposite direction, gets in the way. In the U.S. alone, I have seen some real troubles getting vaccines to the frontline healthcare workers. The night will only grow once countries start mass vaccination. I am not an expert, but I think Wendover can teach you something about it. Head over to Wendover’s YouTube video on distributing the COVID vaccine.
Protection against novel variants
Okay I am out of time to write more because I was hoping to hit the deadline, so here is quick overview about mutations that the scientific community is seeing. I talked about B117 lineage and that does not seem to be a problem with regard to vaccines that are designed by referencing it to the Wuhan-Hu-1 sequence.
What we do not know quite fully yet is the other variant, B1351. Not an expert on this lineage yet, but I have seen a Twitter thread talking about it.
Allergic reaction and adverse events
CDC released a report on 6 January 2021 on events relating to allergic reactions including anaphylaxis after receiving the first dose of Pfizer/BioNTech between 14 to 23 December 2020. The system, VAERS (vaccine adverse event reporting system), detected 21 cases of anaphylaxis after administering 1,893,360 of first doses of the vaccine. This is equivalent to 11.1 cases per million doses. As for non-anaphylaxis allergic reactions, about 4.3k of those were detected (0.2%). There is a nice table on anaphylaxis cases. However, it seems like they haven’t finished analyzing data on the non-anaphylaxis cases, which is understandable given that they probably wanted to get this report out as quickly as possible and it would take time to analyze a large dataset.
As of week of 18-24 January 2021, more evidence about adverse reactions just came out. Here is a commentary published in journal Vaccine, outlining steps that should be taken to understand the root cause of the reaction, which now seems to point the culprit to PEG (Kelso, 2021).
Recently on JAMA, a report came out describing 21 events of anaphylactic reaction to the first dose of Pfizer/BioNTech out of approximately 1.9 million vaccinees (Shimabukuro & Nair, 2021), most with documented prior allergies, and mostly had fast onset (fastest at 2 min, slowest at 150 min). Similarly with Moderna/NIAID vaccine, out of 4,041,396 vaccinees, about 1,266 (0.03%) of adverse events were reported, 108 of those were further reviewed, and 10 events were deemed as anaphylactic events (CDC COVID-19 response team).