‘Very much a judgment call’: David Naylor on AstraZeneca and buyer’s remorse

TVO.org speaks with the COVID-19 Immunity Task Force co-chair about risks, studies — and the factors you should consider when booking your second shot
By Daniel Kitts - Published on Jun 07, 2021
David Naylor is a former dean of medicine and past president of the University of Toronto. (CAUT)



The changing advice on vaccines, particularly when it comes to AstraZeneca, has caused confusion and frustration for many people. 

At TVO.org, we’re trying to help answer some of the questions you may have about what vaccine you should take for your second dose. On Thursday, we published an interview with Zain Chagla, a consultant physician at St. Joseph’s Hospital and Hamilton Health Sciences and an assistant professor of medicine at McMaster University.

TVO.org also got a chance to speak with one David Naylor, co-chair of the federal government’s COVID-19 Immunity Task Force. A former dean of medicine and past president of the University of Toronto, he breaks down what the latest studies say about the various vaccines and addresses the public confusion around vaccines — and whether it could have been avoided.  

TVO.org: While you’ve made it clear you have a lot of respect for the people who work at the National Advisory Committee on Immunization, you did express disappointment last month when they suggested Pfizer and Moderna should be the “preferred” vaccines. I’m wondering what you think of NACI’s latest guidance that different vaccines can be mixed and matched?

A man filming in The Agenda studio

Our journalism depends on you.

You can count on TVO to cover the stories others don’t—to fill the gaps in the ever-changing media landscape. But we can’t do this without you.

David Naylor: I think it’s important to understand that the challenge in that context [last month] was one of potential vaccine chagrin, or buyer’s remorse. The other context was a direction that you would preferentially allocate a vaccine regarded as less preferred or less effective to those in higher-risk areas, which, as I said at the time, seemed perverse.

The evidence keeps evolving as to real-world effectiveness. And from the standpoint of what really is driving the change [in NACI guidance], one is supply. The other obviously is the sense that the Vaccine-Induced Thrombotic Thrombocytopenia (VITT) risk is higher than was initially estimated, according to most of the reports we’ve had. 

So with that in play, I think that we have to accept that AstraZeneca is not going to come back into use for first doses. And it’s reasonable, I think, to offer people some choice for a second dose, particularly now that second doses really are the priority in many ways, given the risk of the B.1.617.2 variant, which seems to have some vaccine resistance.

TVO.org: I have gotten a single shot of AstraZeneca. Why should I get a second dose of AstraZeneca when it is now safe for me to get an mRNA vaccine, which doesn’t have known blood-clot risks and may be even more effective at preventing COVID-19? [Editor’s note: Between the interview date and publication time, new data became available. Naylor submitted updated answers to this question via email.]

Naylor: Let's go back to the decision-making factors for taking another dose of AZ versus, say, Pfizer, 

1. Rare but severe side-effects

A key decision-making factor is how individuals perceive the risk of rare but very serious occurrences of blood clots affecting the brain or other vital organs after administration of Astra Zeneca [AZ]. This clinical syndrome, as you know, is called vaccine-induced immune thrombotic thrombocytopenia or VITT for short. The previous estimate, used by NACI, pegged the risk of VITT after a 2nd dose of AZ at 1/600,000 but cautioned that it might rise as reporting improved over time.

An update issued by the UK medicine regulator on June 3rd indicates that 13.4M 2nd doses of AZ had been given in the UK by May 26th. All the 2nd dose VITT reports came from persons aged 50+. That’s also the age bracket where the UK had concentrated the speed-up of its 2nd doses. Of note, even restricting the denominator to that age bracket, the VITT incidence after a 2nd dose improved slightly to ~1/700,000. One of the concerns about the British system, however, has been lags in adjudication of reports of adverse events after vaccination, so I still expect this rate to move towards 1/600,000 or higher in the weeks ahead.

That brings me back to the issue we discussed: concerns about underestimation of VITT in the UK reporting system. One approach as I explained is to generate a worst-case scenario by comparing risks of VITT after a 1st dose of AZ in the UK to jurisdictions that picked up the problem faster and reported the highest rates. Based on earlier data, I indicated that such an adjustment could raise the risk of VITT to as high as 1/187,500 — still very low, but a long way from 1/600,000. It's notable that, potential 2nd dose lags aside, the UK system now seems to be picking up more VITT after 1st doses. With allowance for that improved detection, the adjusted worst-case VITT moves to about 1/200,000.

Last, any death caused by any vaccine is tragic and disturbing, and VITT can be lethal. That risk, fortunately, seems to be extremely low. Combining current UK and Canadian case fatality rates from VITT with a rescaling of UK data, the risk of death after a 2nd dose of AZ is about 1 in a million in a negative-case scenario.   

I think VITT risks, even at these low levels, will remain a factor in decisions for many who had a first dose of AZ. Risk perceptions and tolerances are highly individualized. Given the limited but very positive evidence on mixing doses to date, it makes perfect sense for anyone who is anxious about those risks to switch to an mRNA vaccine where the VITT risk is effectively zero.  

2. Common Side-Effects ("Reactogenicity") 

Two more studies have appeared as preprints since we spoke …

Both come from Germany. These are mid-sized non-randomized cohort studies. Hillus et al compared short-term side-effects of two regimens: AZ followed 10-12 weeks later by a dose of Pfizer, versus two doses of Pfizer spaced 3 weeks apart. Barros-Martins et al compared AZ followed by a second dose of AZ 10+ weeks later to another group that got Pfizer on roughly the same interval.  

What's important here is that both studies mixed a 1st dose of AZ with a 2nd dose of Pfizer on extended intervals pretty similar to what would occur in current Canadian conditions. There is nothing in the reports or comparisons to alternative regimens that suggests a worrisome excess of the usual side effects such as pain at the vaccination site, myalgias, low grade fever, fatigue and so on. It’s unlikely, therefore that this factor is going to have much impact on the choice of mixing versus matching after a 1st dose of AZ. 

3. Effectiveness

The two recently released German studies are very encouraging with respect to the potential effectiveness of mixing versus matching. They are also relevant to the Canadian situation in that they addressed how a 1st dose of AZ followed by a dose of Pfizer 10-12 weeks later compared to either 2 doses of AZ or 2 doses of Pfizer. Both studies analyzed not only key antibody responses but also measures of cell-mediated immunity that were not reported in the Spanish CombiVacS trial. And unlike CombiVacS, they directly compared 2-dose regimens. 

In brief, the mixed protocol was fully comparable in immune responses to those obtained with two doses of Pfizer given three weeks apart per the manufacturer’s specifications. There may even have been an edge in T-cell responses for the AZ-Pfizer series. As compared to the AZ-AZ series, the mixed protocol showed stronger immune responses in general and specifically against key variants of concern.

The limitations of this evidence are obvious. Neither report has been peer-reviewed. Neither was a randomized trial. There were no clinical outcomes reported. And neither addressed coverage of B.1.617.2, or ‘delta’ in the new WHO terminology – the variant that appears to combine higher infectivity with lower response to vaccines. While the Oxford Com-CoV trial is likely to yield more randomized data later this month, it too will lack clinical outcomes.

I learned early in my career as a clinical epidemiologist to be very skeptical about studies reporting proxy measures as opposed to clear-cut clinical outcomes. But the two German studies do offer exciting and mutually reinforcing evidence in support of mixing rather than matching.    

The bottom line here is simple. There’s no wrong choice. Some Canadians will be persuaded by these indirect but very impressive indicators of the effectiveness of following a first dose of AZ with a second dose of Pfizer. Others will be skeptical about the lack of direct clinical proof of equal or superior effectiveness, and get on with taking a second dose of AZ ...

I think most decisions will therefore come down to individual risk perceptions and how each person interprets the effectiveness data.  

The most important message to those making a decision about a second dose is this: Thank you for protecting yourself and others with an effective vaccine during the Third Wave. What matters more than anything now is not what product you choose but the fact that you’re getting a 2nd dose to improve your protection against new variants.    

TVO.org: I spoke to Zain Chagla of McMaster University, and he made a rough estimate that now, with this new guidance, only about 25 per cent of one-dose AstraZeneca recipients will come back for a second AstraZeneca dose. What does that mean for our efforts to vaccinate the population as quickly as possible — and for all those AstraZeneca doses we already have?

Naylor: I would be surprised if it was only 25 per cent. But it’s possible that’s where it will end up. What that’s going to mean, I think, is that we’ve had only a one-month extension for some of the current doses. Some of them could end up being wasted. And it will raise questions as to what should be done with any further doses that have been ordered. Some of them could end up being sent abroad. I certainly hope we don’t end up in a situation where we simply have them coming in and don’t know what to do with them. That would be unethical.

TVO.org: In March, when reports emerged of people insisting on getting an mRNA vaccine instead of AstraZeneca, I thought those people were being foolish. I didn’t question the message “just get the first vaccine you’re offered.” But knowing what we know now, were those people I thought were foolish right to be leery of AstraZeneca?

Naylor: I think that they were right to be concerned, depending upon their own risk assessment about a blood clot associated with the first dose. I think people who got AstraZeneca followed what was very reasonable public-health advice and had a very limited risk of a serious blood clot. Certainly, in my case, since I took AstraZeneca, I anticipated the possibility of going to a heterologous booster [a different vaccine for the second dose]. This was not something that was in any way unusual. We do heterologous boosting with many vaccines.

I don’t think anyone made an error here. I don’t think anyone was duped. Those people who got the doses were not only protecting themselves and their families, but protecting others. Let’s remember we were in the midst of the third wave. And it was the right strategy, I think, at the time. 

A week may be a long time in politics. A day is a long time in a pandemic. The evidence changes; the supplies change. And I think one has to simply adapt. As for myself, I’m still pondering what I will do when my 12 weeks [until the second dose] are up. But I’m perfectly comfortable and happy with the choice I made. I don’t think anyone should have any particular buyer’s remorse about having tried to do the right thing in the midst of a very vicious third wave.

TVO.org: I think at this stage, if getting an AstraZeneca shot meant I could get to two doses faster, I would do it. But if waiting a couple of weeks meant getting an mRNA shot, I think my calculus — and, I imagine, the calculus of a lot of people — will be, “Well, why take the risk with AstraZeneca if in a few weeks you can get something where don’t have to spend a month worrying about blood-clot symptoms?”

Naylor: As I said to others, these are going to be decisions taken more in your innards, your guts, than necessarily between your ears, because there is a lot of this that has to do with how one trades off risks and benefits. It would be nice if there were a simple formula you could just put up on a computer and fill in a few blanks and be told to go Direction A or Direction B. But it’s not quite that simple. It is very much a judgment call.

Above all, I have to say getting a second dose is the right thing to do. It’s nice to get it sooner rather than later. But if the delay is not particularly long, and there’s a concern, fine. The point is we need Canadians to get second doses.

TVO.org: This is an unprecedented situation where billions of people are lining up for vaccines that didn’t exist a year ago. The data we have around these vaccines is obviously changing constantly. As a result, it makes me wonder if this confusion over vaccines was unavoidable. How could the messaging around these vaccines have been any clearer given the circumstances?

Naylor: It is almost unavoidable. And it is almost impossible to convey subtleties of risk and to sort out what to do — particularly in this federation, where, for all intents and purposes, you have 13 jurisdictions that have the authority to make their own calls on vaccine rollouts. And remember that NACI was basically a peacetime body that was set up originally to deal with childhood-immunization issues. It has a great record. No history exists for a national guideline body in any discipline having to issue almost weekly updates in the context of a flood of evidence and in the face of product development that we’ve never seen in the history of medicine. To me, the communications challenges here have been incredible. 

And one lesson, I think, for the next pandemic, which I hope is a long time away, is that we are going to have to think about what configuration we use to deal with vaccines and therapeutics and what communication strategies we put in place to try to make this smoother. But, frankly, it’s almost unavoidable to have bumps. And we are where we are. I think this latest decision is a good one. And we’ll just keep tracking the evidence and try to sort our way through it. The good news here is the country’s zooming along with coverage. We’re making really good strides. As always, Canadians found a way to muddle through in good faith and with a reasonable amount of social solidarity and get the job done.

This interview has been condensed and edited for length and clarity.

Thinking of your experience with tvo.org, how likely are you to recommend tvo.org to a friend or colleague?
Not at all Likely
Extremely Likely

Most recent in Coronavirus