The Evidence On 4th Doses of Covid Vaccines & the Thorny Question of Age Thresholds

When I last posted on boosters in January, there were at least 14 countries offering fourth vaccine doses. Almost all were only for people with compromised immune systems, for whom a primary course is widely regarded as 3 doses. So a fourth dose would be their first booster, and there’s good reason for that. I discuss that and other strategies for immunocompromised people below.

Back then, the US was recommending a fourth dose only for certain people with compromised immune systems. Now the US Food and Drug Administration has approved a fourth dose for people from 50 years of age, and a fifth for the immunocompromised. Has the evidence for that age group changed that much in these last few weeks?

Short answer: not really. The FDA extended their approval for the Moderna and Pfizer vaccines to people aged 50+ arguing that the “known and potential benefits…outweigh the known and potential risks in the authorized populations when given at least 4 months following the first booster dose”.

They don’t explain the basis for the potential in this specific age group. What about what’s known? They cited a single study from Israel for both vaccines, and for the Pfizer vaccine, they also included this statement on the question of risks: “Safety surveillance data from the Ministry of Health of Israel on the administration of approximately 700,000 fourth doses of the Pfizer-BioNTech COVID-19 Vaccine given at least 4 months after the third dose in adults 18 years of age and older (approximately 600,000 of whom were 60 years of age and older) revealed no new safety concerns”.

STAT reports that Pfizer also provided data on waning for their vaccine, but applications from Moderna and Pfizer weren’t published.

There are 4 studies now in the public domain for these vaccines, all from Israel. The one cited by the FDA is the only one published in a journal on March 16 (Regev-Yochay 2022). The others are preprints, one posted on February 1 (Bar-On 2022) [subsequently published in a journal], and the others posted on March 24 (Arbel 2022 and Gazit 2022).

Regev-Yochay 2022 – the study cited by the FDA:

This is a study from Sheba Medical Center, of a fourth dose of Moderna or Pfizer vaccine 4 months after the third. The participants are 274 healthcare workers matched with another 548 who had 3 doses. They were tested for Covid every week, so the study could assess risk of infection, as well as data on signs of immune response and illness. The numbers were small, so there’s a lot of uncertainty. Here’s what they found, comparing the advantage gained by that fourth dose:

• Immune response after the fourth dose was stronger than after the third.
• Effectiveness against infection: for Moderna, 11% (CI -43 to 44); for Pfizer, 30% (CI -9 to 55).
• Effectiveness against symptomatic disease: for Moderna, 31% (CI -18 to 60); for Pfizer 43% (CI 7-65).

In both vaccine and control groups, people who got symptomatic disease generally had mild symptoms, and no one was hospitalized for Covid-19.

For age, they reported a calculation for 2 age groups: 40-59 and 60+ (compared to 18-39 year-olds). (That calculation was log incidence rate ratio, with -0.2996 for the 40-59 year-olds, and -0.9183 for 60+.) There were only 208 people aged 46 or older who had fourth doses in this study.

Arbel 2022:

This study assessed risk of death. It’s from the Clalit Health Services, with 563,465 people aged 60+ who were eligible for the fourth dose – 58% had it. The researchers adjusted for some confounding variables for their calculations, but it wasn’t a matched or case control study. Those variables were quite extensive, including both socioeconomic data, Covid risk factors, and Covid antiviral treatment. They compared people with 3 doses to people at least 7 days after a fourth dose.

There were 40 days of follow-up, and in that time, 92 people who had had a fourth dose died of Covid versus 232 who had had three doses – that’s about a 78% reduction (adjusted hazard ratio 0.22 [CI 0.17 to 0.28]). That varied enormously according to age, though. Compared to the 60-69 year-olds, Covid-related death was twice as likely in the 70-79 age group (2.24 [CI 1.51-3.34]) and 10 times as likely for 80+ (9.95 [CI 6.93-14.28]).

Bar-On 2022:

The data here come from the Israeli Ministry of Health database, for 1,138,681 over 60, and therefore eligible for a fourth dose. They chose only people who had the Pfizer vaccine, and had no confirmed Covid-19 in their record before January 1: the fourth dose vaccination campaign began on January 3. They collected infections from January 15 to 27, and severe illness from January 21 to 27.

The researchers adjusted for some confounding variables, but it wasn’t a matched or case control study, and it wasn’t a group of people all using the same testing service (or being regularly tested). The factors they included were date of infection/illness, age group (60-69, 70-79, and 80+ years), sex, and demographic group (general Jewish, Arab, ultra-Orthodox Jewish).

Calculated differences in rates between people with 3 doses and those with 4 doses (12 or more days after the fourth injection): people without the fourth dose were twice as likely to have a positive test recorded (rate ratio 2.0 [CI 2.0-2.1]) and 4 times as likely to have severe Covid (rate ratio 4.3 [CI 2.4-7.6]).

Gazit 2022:

This one is from the Maccabi Healthcare Services, a health fund. It’s a test-negative case-control study with 97,499 people who got a Covid test, were aged 60+ and had 3 doses (so were eligible for 4th): 27,876 of them had a 4th dose.

Effectiveness against infection was high at first, but dropped by 9 weeks’ follow-up. Effectiveness against infection peaked at 64% (CI 62-66) & dropped to 29.2% (CI 18-39). They classified Covid-related hospitalization and Covid-related deaths as severe Covid. Effectiveness against severe Covid stayed over 73% throughout though, and in the last few weeks of the study it was 86.1% (CI 73-93%).

So that’s quite a lot of data for fourth doses for those who are 60+. But what about 50 as the age threshold? The evidence the FDA cites doesn’t relate to that age.

Is the age 50 because of international consensus? No, it’s definitely not that. Countries with age thresholds for fourth doses are setting it anywhere from 50 to 80:

• Australia: 65 (and 50 for Aboriginal and Torres Strait Islander people)
• Chile: 55
• Dominican Republic: 60
• France: 80
• Germany: 70
• Israel: 60
• Sweden: 80
• UK: 75
• USA: 50

The Australia vaccination authority was explicit about the reason for the 65-year-old threshold: “Older age is the strongest risk factor for severe COVID-19 outcomes…The age cut-off of 65 years aligns with eligibility for receiving influenza vaccine under the National Immunisation Program, which may facilitate implementation and uptake of the additional booster dose”. That’s good to know.

The FDA provided no rationale for choosing 50 years in their documentation. But STAT reports their spokesperson said “the FDA landed on 50 as the age cutoff because one in three people 50 to 65 have a health condition that leaves them vulnerable to more severe Covid-19”. (They also didn’t go through their more usual process of having the subject discussed by their vaccines committee – those discussions, including data presentations, have been live streamed.)

I’m not sure about the other countries, though at least in some cases, it looks to correspond with the threshold they use for reporting increased risk in the elderly. In their detailed 62-page report, the German authority included a chart of hospitalization rates in age groups, including each decade from 60. That showed very elevated levels for the age range they chose.

I don’t envy the people having to make these decisions in a pandemic, with so much uncertainty and so much potential for disaster from a wave of Covid if they get it wrong. Unless authorities explain the rationale for their choices, though, it can look pretty arbitrary.

What isn’t arbitrary, though, is the insanity of vaccine distribution globally. While a healthy 50-year-old working at home in the US could have 4 doses of mRNA vaccine and 74% of people in high-income countries have had a full primary course, only 15% of people of in low-income countries have had at least 1 shot. After more than a year of vaccinating people, the goal of 70% of the world vaccinated – the minimum needed to keep the planet safe – isn’t even in sight.

Fourth and fifth doses in people with compromised immune systems

There isn’t data from randomized trials to test the results of having 4 doses. The level of risk exposure to Covid-19 and vaccine effectiveness is very different for people with compromised immune systems, though. The following section in this post covers randomized trials that have been registered on interventions to improve vaccination outcomes for immunocompromised people – but only 1 of those includes fourth doses, and that part is only a pilot study. So it isn’t designed to answer questions about fourth dose effectiveness.

In my January post, I included some reports of outcomes have been published for people with compromised immune systems – all from France – with fourth doses of Moderna or Pfizer vaccine:

• 92 people with kidney transplants after a weak response to their third (Caillard 2021)
• 67 people with kidney transplants after a weak response to their third (Benotmane 2021)
• 37 people with kidney transplants in France given a fourth dose of Moderna or Pfizer (Kamar 2021)

Since then, there have been some studies of 5 doses as well. In both, immune response rates were higher after the 5th dose:

• 18 people with solid organ transplants in the US (Abedon 2022)
• 603 people with kidney transplants and 3 doses, 250 with 4 doses, and 40 with 5 doses, sometimes with treatment pause, in Germany (Osmanodja 2022)

I’ve identified 2 additional trials since I last summarized the evidence on vaccination and people with compromised immune systems – and no additional results. You can read that summary of issues and evidence here.

Trials comparing interventions for people with compromised immune systems

(Trials in chronological order of announcement – added since last booster post marked *)

Vaccines	Schedule	Participants	Location(s)	Research phase
CoronaVac	Randomized to not take the 2 doses of methotrexate due after vax or continue treatment as usual	138 people with stable rheumatoid arthritis using methotrexate	Brazil	Results
Moderna, Pfizer	Head-to-head trial of 2-dose courses of Moderna vs Pfizer	700 people with HIV or transplants	Switzerland	Phase 3 (protocol)
AZ, Moderna, Pfizer	2 doses of Moderna or Pfizer vax, randomized to either a 3rd dose of the same vax or a dose of AZ	60 people on rituximab who did seroconvert after 2 doses of mRNA vax	Austria	Phase 2 Results
Moderna	2 doses of Moderna a month apart, randomized to either a 3rd dose 2 months later or placebo	120 people with solid organ transplants	Canada	Phase 4 Results
J&J, Moderna, Pfizer	2 doses of Moderna or Pfizer vax, randomized to a 3rd dose of J&J, Moderna, or Pfizer (with pilot study of 4th dose)	197 people with kidney transplants who had not developed spike antibodies after 2 doses of mRNA vax	Austria	Phase 4 Results
Pfizer	Randomized to reduction of mycophenolic acid plus 3rd dose of Pfizer vax or 3rd dose only	504 people with kidney transplants with inadequate humoral response after 2 doses of Pfizer vax	Israel	Phase 4 Called BECAME (protocol)
AZ, Moderna, Novavax, Pfizer	Randomized to a 3rd dose of Moderna, Novavax, or Pfizer	1,200 people with compromised immune systems who had a poor response to AZ or Pfizer vaxes	UK	Phase 3 Called OCTAVE DUO
City of Hope, J&J*, Moderna, Pfizer	Randomized to 2 doses 28 days apart of either of City of Hope vax or a US-authorized vax (* not clear if J&J included)	240 people with blood cancer and stem cell transplant or cellular therapy	US	Phase 2
J&J, Moderna, Pfizer	2-dose course of Moderna or Pfizer, or single-shot J&J, randomized to a dose of 1 of the 3 vaccines, with or without stopping some immunosuppressive treatments before and shortly after vaccination	600 people with any of 5 autoimmune diseases (including rheumatoid arthritis and MS), on immunosuppressive treatment, and with suboptimal response to any of the 3 vaccines	US	Phase 2
AZ, Moderna, Pfizer	Randomized to 2-week break from methotrexate or not while receiving booster dose	560 people with autoimmune disease (including rheumatoid arthritis) who have had methotrexate for at least 3 months	UK	Phase 3/4 Called VROOM
Moderna, Pfizer	2 doses of Moderna or Pfizer vax, randomized to a third dose of the same or the other vax	300 people with kidney transplants or on dialysis who had a poor response to 2 doses of mRNA vax	Canada	Phase 2/3 Called Boost Kidney
J&J, Moderna, Pfizer	2-dose course of Moderna or Pfizer, or single-shot J&J, followed by dose 3 of Moderna	171 people with hematological (blood) cancers who had a poor response to any of the 3 vaccines	US	Phase 2 (unrandomized)
J&J, Moderna	People on triple immunosuppressive therapy: randomized to 3rd dose of Moderna, with or without discontinuation of mycophenolate mofetil (MMF) or mycophenolic acid (MPA); others randomized to 3rd dose of Moderna, 3rd and 4th dose of Moderna, or J&J	460 people with kidney transplants who did not seroconvert after 2 doses of Moderna	Netherlands	Phase 4 Called RECOVAC
J&J, Moderna	Head-to-head trial of J&J vs Moderna for 3rd dose	386 people with solid organ transplants who have had 2 doses of Moderna vax	Spain	Phase 3 Called REIN-TX
J&J, Pfizer	Head-to-head trial of J&J vs Pfizer for 3rd dose	200 people with solid organ transplants who have had 2 doses of Pfizer vax	US	Phase 3
Moderna, Pfizer	Randomized to no change in immunosuppression for 3rd dose of mRNA vax or to reduction in dose of mycophenolate mofetil/mycophenolic acid (MMF) or azathioprine before and after mRNA vax	50 people with kidney transplants	US	Phase 4 Called ADIVKT
J&J, Moderna, Pfizer	Randomized to homologous booster (2nd dose for J&J, 3rd dose for Moderna and Pfizer) or heterologous booster (2nd dose of mRNA if originally J&J, 3rd dose of J&J if originally mRNA)	60 people with multiple sclerosis	US	Phase 4
CoronaVac	2 doses of CoronaVac alone or 3 doses of CoronaVac or 2 doses of CoronaVac plus a double-dosage 3rd shot	240 people with pulmonary tuberculosis	China	Phase 4
J&J after Moderna or Pfizer	People with low protein spike test after 2 or 3 doses of Moderna or Pfizer, randomized to a dose of J&J (with or without change in immunosuppression medication); if response still low, randomized to a further J&J dose, again with or without medication change	1,200 people with kidney transplants	US	Phase 3
* AZ, Moderna, or Novavax	4th dose, heterologous (different vaccine than original course)	287 people people with rheumatic disease & immunocompromise	Canada	Phase 2/3 (unrandomized)
* Moderna	4th dose in half, standard, or double-dose	60 people with lung transplants	US	Phase 1/2

~~~~

For details on how I track Covid vaccine progress, see my background post. Notes on my collection of studies are here. The collection is in a public Zotero library you can dig into here.

All my Absolutely Maybe Covid-19 vaccine posts

All previous Covid-19 posts at Absolutely Maybe

My posts at The Atlantic, at WIRED, and debunking posts at my personal website.

Disclosures: My interest in Covid-19 vaccine trials is as a person worried about the virus, as my son is immunocompromised: I have no financial or professional interest in the vaccines. I have worked for an institute of the NIH in the past, but not the one working on vaccines (NIAID). More about me. The age threshold issue: I fall into a 5-year age group that’s at elevated risk of severe Covid outcomes, and was over the age of eligibility for Pfizer and Moderna vaccines for most of last year and under the age eligibility for a fourth dose now.

The cartoons are my own (CC BY-NC-ND license). (More cartoons at Statistically Funny.)