Do people with prior covid infection require vaccination? A summary of the data

1. Introduction
2. Evidence of immunity post natural Covid infection
3. Analysing vaccine clinical trial data to look for efficacy in individuals with prior infection — Pfizer, Moderna, Janssen, Astrazeneca, Sinopharm, Novavax SA, Astrazeneca SA
4. Real world studies looking at effect vaccinating convalescent patients

Introduction

To start off by saying, I think the development of COVID-19 vaccines are a brilliant scientific achievement, they are highly effective and I strongly recommend them to patients as risks outweigh benefits to a large section of the population. They have had a great contribution in reducing the number of deaths due to the covid-19 infection. I also believe that the right to bodily autonomy is enough reason for people who do not want to be vaccinated, to refuse a vaccine.

What about the significant proportion of the population who have already contracted covid (confirmed by either a positive pcr test, positive serology or both) — are there any benefits of vaccination in these people, and if there are, are they the same as the benefits obtained from vaccinating those without prior covid infection? So is there efficacy in vaccinating seropositive patients? Even if there is, does it apply to all age groups? If there is already protection from natural immunity, by definition absolute risk reduction in vaccinating seropositive patients would be lower than in seronegative patients. Do those possible benefits outweigh the risks of vaccination?

Most public health bodies have recommended the public to get vaccinated even if they have had covid infection before. Some people have stated “one jab is enough”. Is even one jab needed? Many previously infected individuals who recovered may still want to be fully vaccinated for peace of mind, and that’s reasonable as serious side effects from vaccines are rare. However there will be individuals who recovered who may be hesitant about vaccination wondering if the benefits outweigh the risks. Similar review will be needed in regards to the need for future boosters in the currently “fully vaccinated” cohort.

I hear a lot of people saying we just don’t enough know about natural infection induced immunity — the strength and duration of it — and thus it is safer to get vaccinated. A lot of countries have had major outbreaks since March 2020, the earliest public vaccination outside of clinical trials were in December 2020 — so if anything we should know more about natural immunity. The median followup period in the published data of the phase 3 trials for the EUA vaccines was about 2 months, yet there was no talk of “we dont know how long vaccine induced immunity will last” — of course it is also in the pharma companies interest that the immunity should only last ~1 year therefore ensuring a huge chunk of lifetime annual vaccine subscribers.

Evidence of immunity post natural Covid infection

There are a lot of studies and I shall post some of the good ones.

1. Antibody Status and Incidence of SARS-CoV-2 Infection in Health Care Workers — Lumley et al. NEJM Feb 2021 — Adjusted risk ratio of covid infection for antibody positive healthcare workers was 0.11 (89% less likely) — median follow up 139 days. All reinfections were asymptomatic.
2. SARS-CoV-2 infection rates of antibody-positive compared with antibody-negative health-care workers in England: a large, multicentre, prospective cohort study (SIREN) — Hall et al. Lancet Apr 2021 — Adjusted incidence rate ratio for covid infection — 0.159 (84% protection) and symptomatic covid — 0.074 — (93% protection) — median follow up 7 months, by the end of the study, the alpha variant was already dominant in UK and previous infection still provided protection from the alpha variant.
3. SARS-CoV-2 antibody-positivity protects against reinfection for at least seven months with 95% efficacy — Abu-Raddad et al — EClinicalMedicine April 2021 — Efficacy of natural infection against reinfection was 95.2%, reinfections were less severe than primary infections. Median follow up 114 days.
4. Assessment of SARS-CoV-2 Reinfection 1 Year After Primary Infection in a Population in Lombardy, Italy — Vitale et al — JAMA Internal Med — May 2021 — Patients who had 1 PCR positive tests had Hazard ratio 0.06 of further positive PCR — 94% protection — mean follow up 280 days.
5. The risk of symptomatic reinfection during the second COVID-19 wave in individuals previously exposed to SARS-CoV-2 — Manica et al — PREPRINT — April 2021 — Antibody positive patients had odds ratio of 0.054 of symptomatic covid infection (95% protection) — 7 month follow up.
6. Reinfection Rates among Patients who Previously Tested Positive for COVID-19: a Retrospective Cohort Study — Sheehan et al Clinical Infectious Diseases — Mar 2021 — After 1 positive PCR, Protection against any COVID — 81.8%, symptomatic COVID — 84.5% — minimum of 6 month follow up — protection increased over time.
7. SARS-CoV-2 infection and reinfection in a seroepidemiological workplace cohort in the United States — Finch et al — PREPRINT — May 2021 — Patients with positive antibodies — 91% protection against reinfection over atleast 6 month period
8. Risk of reinfection after seroconversion to SARS-CoV-2: A population-based propensity-score matched cohort study — Leidi et al — Clinical Infectious Diseases May 2021 — seropositive patients had 94% reduction in hazard of testing positive — mean follow up 8 months
9. A 1 to 1000 SARS-CoV-2 reinfection proportion in members of a large healthcare provider in Israel: a preliminary report — Perez et al. PREPRINT March 2021 — Those who had positive PCR test had 1 in 1000 chance of having a subsequent true reinfection. Study period over 10 months
10. Assessment of protection against reinfection with SARS-CoV-2 among 4 million PCR-tested individuals in Denmark in 2020: a population-level observational study — Hansen et al — Lancet — Mar 2021 — Those who were PCR positive in first wave in Denmark had 80.5% protection against reinfection, however when they looked at it by age, those above 65 only had 47.1% protection — something that is worth noting. They did not look at symptoms at all either, and also worth noting that Denmark has by far the highest tests per population ratio in the world. 9 month study period.
11. SARS-CoV-2 seropositivity and subsequent infection risk in healthy young adults: a prospective cohort study — Letizia et al — Lancet — April 2021 — US marine recruits; Seropositive recruits had incidence ratio of testing positive of 0.18–82% protection. 10% of seropositive recruits tested positive in this study. This was used by the press to state young people are not immune from reinfection and must be vaccinated even if previously infected. They forgot to report that reinfections were more likely to be asymptomatic (84% vs 68%) and that 48% of recruits without antibodies tested positive — this is one of the highest rates of infection in any study surely, way way higher than in the vaccine trials. Study period just 6 weeks.
12. Incidence of SARS-CoV-2 infection according to baseline antibody status in staff and residents of 100 long-term care facilities (VIVALDI): a prospective cohort study — Krutikov et al — Lancet — June 2021 — Nursing home residents and staff — Antibody positive residents had 85% protection from reinfection, staff had 61% protection. None of the reinfections were hospitalised (pretty amazing for nursing home residents!) Follow up 144 days
13. Association of SARS-CoV-2 Seropositive Antibody Test With Risk of Future Infection — Harvey et al — JAMA Internal Med — Feb 2021 — During the follow-up periods, the ratio of positive PCR among those with antibodies vs those with a negative antibody test at index was 2.85 0 to 30 days, 0.67 at 31 to 60 days, 0.29 at 61 to 90 days, and 0.10 at more than 90 days. 90% protection after 90 days. Early PCR positives (0–30 days) are likely prolonged RNA shedding — important to note this for later.
14. Protection of previous SARS-CoV-2 infection is similar to that of BNT162b2 vaccine protection: A three-month nationwide experience from Israel- Goldberg et al PREPRINT — April 2021 — Those with previous PCR positive tests had 94.8% protection against reinfection, 94.1% protection against hospitalization and 96.4% protection against severe illness. 3 month study period
15. Impact of vaccination on new SARS-CoV-2 infections in the United Kingdom — Pritchard et al Nature Medicine June 2021 — Comparing those who were fully vaccinated and those with previous infection, no statistically significant difference in infection rates, less symptoms in the natural infection rates and higher cycle thresholds (but not statistical significant difference)
16. Necessity of COVID-19 vaccination in previously infected individuals — Shrestha et al — PREPRINT — June 2021 — in 1359 employees with previous infection who remained unvaccinated — not a single reinfection. Median follow up for previous infected individuals was 143 days
17. New national surveillance of possible COVID-19 reinfection, published by PHE — updated July 2021 Public Health England — out of over 4.3 million infections, only ~23105 possible reinfections, 839 probable and 66 confirmed. There is some evidence of slightly increased risk of reinfection with delta compared to alpha variant. However this does not necessarily mean a reduction in protection, as primary infection is more likely with delta compared to alpha (established greater transmissibility of delta compared to alpha).

Ok that’s great, there is clear evidence of post infection immunity, but can we not boost it more by vaccinating them?

Let’s go back to the vaccine trial data.

Analysing vaccine clinical trial data to look for efficacy in individuals with prior infection

Most of the vaccine trials recruitment policy was to exclude patients with known history of covid-19 infection. Despite that they do find a small proportion of patients recruited are seropositive. Pretty much every vaccine trial aims to report vaccine efficacy in seronegative patients. Why? because they suspected that there will be a likely protective effect from prior infection and this would blunt reported vaccine efficacy. However they included seropositive patients in their safety analysis looking at adverse effects. The proportion of seropositive patients was small and thus admittedly it would be harder to prove statistically significant efficacy in this group anyway — but even trends of efficacy would be useful. Annoyingly the pharma companies did not make the data easily available.

Pfizer phase 3 trial, interim analysis published in NEJM — 43448 patients randomized to placebo or vaccine, median followup 2 months, max follow up — 112 days. 95% efficacy in patients without evidence of prior infection.

Of their initially recruited patients, 2.9% had evidence of prior infection.

So what was the effect in patients with evidence of prior infection. Ofcourse they dont report it in the NEJM journal article. But the data is available on the FDA website, presentation named “FDA Review of Efficacy and Safety of ​​Pfizer-BioNTech COVID-19 Vaccine ​Emergency Use Authorization Request” as well as on the EMA page for their vaccine underpublic assessment report.

Post dose 1, efficacy in antibody negative group is 86.2%, but in antibody positive group it is negative efficacy, it is -17.9% (doesnt reach statistical significance). However there is something interesting here… that is also seen in the Novavax South African and Astrazeneca South African trial.. Will comment later in the article — also more detail in a followup article I wrote.

Post dose 2, In those with positive baseline antibodies, again efficacy is actually negative, it is -7% (of course does not reach statistical significance either).

If you were paying attention to the numbers, you will see post dose 1, there are actually higher numbers in the seropositive group including those vaccinated when compared to seronegative. This effect seems to have disappeared post dose 2 — so vast majority cases in the seropositive group are between dose 1 and 2.

They are quite deceptive with their language too in the journal article, they use the term “with or without evidence of previous infection” and show additional cases in this group when compared to “without evidence of previous infection”, but these are mostly in patients with missing baseline serology as opposed to positive baseline serology.

UPDATE

Pfizer released new data on 28/7/2021 — preprint — 6 month follow up data.

Post dose 1, 13 cases in vaccine group and 17 cases in placebo group, 19.2% efficacy — non statistically significant. But when they look at the subset who were only positive for antinucleocapsid antibody and not by PCR positive, they had 2 in vaccine and 7 in the placebo group — 70.5% efficacy, not statistically significant though.

The above data led Pfizer to state “These data support the current practice of immunizing without screening for evidence of prior infection.” I dont think the matter is settled yet. They have not given a breakdown post dose 2. From the interim analysis, we noted there were 19 cases in the seropositive post dose 1, but only 2 cases post dose 2 — suggesting most of the cases were between dose 1 and 2 — possibly not true reinfections and more residual RNA shedding.

Verdict —Updated: With the longer follow up data, in a select group, post dose 1, maybe a trend of efficacy, but not statistically significant and requires greater clarification.

Moderna phase 3 trial interim analysis published in NEJM, 30420 patients randomized to placebo or vaccine, median followup — 63 days, maximum follow up — 118 days. 94.1% efficacy in patients without evidence of prior infection. 2.2% of their recruited patients did have evidence of prior infection. Effect in patients with prior infection — again not in the main NEJM article but is present in the supplementary appendix.

So in those with positive antibodies, the placebo group had 1 case and the vaccine group had 0 cases. So they did reach 100% efficacy, ofcourse with such a small number of cases, they did not reach statistical significance as confidence intervals ranged from negative efficacy to 100%

Verdict — No statitistically significant benefit demonstrated, maybe with a much larger sample size, this could change. Absolute risk reduction is very small even if are willing to take into account the difference between placebo and vaccine group.

Janssen (Johnson&Johnson) phase 3 interim analysis published in NEJM. 43783 patients randomized to placebo or vaccine, median followup was 58 days –66% efficacy in those without prior infection. 10% of their recruited patients had positive antibodies, so what was the efficacy in these patients?

Again this data is not in the NEJM article but on the FDA website as a presentation named “FDA Review of Efficacy and Safety of the Janssen COVID-19 Vaccine Emergency Use Authorization Request”.

In those with positive antibodies, 3 cases in vaccine group and 4 cases in placebo group, efficacy is 28.5% and in no way reaches statistical significance again demonstrating no benefit in vaccinating those with prior infection.

Verdict — No benefit demonstrated, large number of seropositive patients in this trial which is useful, even with over 4000 seropositive patients of whom over 2100 got vaccines, they could not achieve efficacy.

Astrazeneca interim phase 3 (with UK and Brazil data only) –62% efficacy against symptomatic covid, 55% against positive swabs in those with no evidence of prior infection. So pretty average efficacy even in those without prior infection. How about those with prior infection?

Looking at their product assessment on the EMA website. They state

Efficacy in seropositive subjects There were few subjects seropositive at baseline (373 subjects in total, DCO 4 November 2020). The number of seropositive participants in Any Dose Efficacy was too small for a meaningful analysis of the incidence of COVID-19 (0/185 cases in the AZD1222 group and 1/188 cases in the control group). No reliable estimates of VE by serostatus at baseline can be presented.

Verdict: Really small numbers, possible efficacy? I will come back to AZ vaccine very shortly…

Sinopharm (BBIBP-CorV and WIBP-CorV)

80% and 78% efficacy in seronegative patients.

Looking at the supplementary content from the JAMA article.

1 case in placebo and 0 in both vaccine groups, 100% efficacy, no confidence intervals declared. Absolute risk reduction is much smaller in seropositive individuals in any case. Weirdly in the missing baseline serology group, efficacy is much lower despite a large number of individuals in this group

Verdict: Efficacy with no confidence intervals, ?statistical significance seems unlikely. Absolute risk reduction is much smaller in seropositive participants

Ok so looking at the clinical trial data there’s minimal evidence for benefit for vaccinating seropositive patients.

WAIT… but variants…

Novavax South African trial phase 2b trial data

nearly 30% antibody positive patients

From the NEJM article

“Notably, during the initial 60 days of follow-up in the placebo group, the preliminary incidence of Covid-19 that was observed among participants who were seronegative at baseline (5.3%; 95% CI, 4.3 to 6.6), which included 33 mild and 47 moderate cases among 1516 participants, was similar to the incidence among seropositive participants (5.2%; 95% CI, 3.6 to 7.2), which included 14 mild and 21 moderate cases among 674 participants (Figure 2C). “ — No protection from previous infection against the South African variant?? This is quite alarming!

Also positive efficacy against South African variant in vaccinating seropositive patients? Even though not reaching statistical significance, it isnt far off.

Something does not look right about the Kaplan-Meier curve, a lof early cases in the seropositive group and then a fall off after 49 days. Are they simply picking up post infectious RNA shedding?

The duration of followup is very short too, just 60 days.

Not surprisingly they released a statement later dated Mar 11 — “A previously reported initial analysis from the study through 60 days indicated that prior infection with the original COVID-19 strain might not completely protect against subsequent infection by the variant predominantly circulating in South Africa. However, the complete analysis of the South Africa trial indicates that there may be a late protective effect of prior exposure with the original COVID-19 strain. In placebo recipients, at 90 days the illness rate was 7.9% in baseline seronegative individuals, with a rate of 4.4% in baseline seropositive participants.” So clearly with just 30 more days follow up, there is a significant protective effect from previous infection against the South African variant and surely this should increase further with time. No more claims of efficacy in patients with previous infection either.

Verdict: Interesting stuff, clearly we need to wait for data after the longer follow up. Remember from one of our earlier trials that people were more likely to test in the first 30 days after a positive antibody test, but with time this quickly turns the other way.

Still the south african variant seems tricky right?

Finally AZ South African variant trial

Some nice data in this one!

Over 15% were seropositive

21.9% efficacy against any covid variant, 10.4% efficacy against South African variant even in patients without previous infection, and even that wasnt statistically significant. Fair to say no efficacy, but seems all cases were mild-moderate, so experts said they expected efficacy agaisnt severe infection — which seems reasonable to me but no data in this trial supportive of it.

You can infer subtracting the the “any numbers” from “seronegative numbers”, 4 cases were in seropositive or missing serology group, (1 was in the placebo and 3 in the vaccine) — from the later data, you can further infer out of 4, 3 were in seropositive and 1 in missing serology group. From this you can both infer a protective effect of previous infection and no efficacy in vaccinating seropositive, but look down for further analysis. Unlike the pfizer data, this trial did not have many participants with missing serology data, they were mostly negative or positive.

But wait, is there any evidence from the trial data, that previous infection protects against the South African variant? That would be interesting right

Have a look in the supplementary index data on the article page

Data until 31st October 2020 (before the emergence of the South African variant and actually even before the patient recruitment window ended — which was june 24 to November 9), strangely there are way more cases in the antibody positive group, but this would be at short duration of follow up and may be prolonged viral rna shedding — interesting though and has to be noted.

Now after including data up to Jan 15 2021 (about 75 days later) the proxy for the window for infection by the beta variant, we see symptomatic cases in the antibody positive group have gone from 5 to just 7 (extra 2 cases), whereas they have gone from 9 to 53 (extra 44) in the antibody negative group — so for the beta variant, 75% less likely for the seropositive group after adjusting for the size of both groups. If just looking at all PCR positives, the seropositive group has increased from 12 to 14 (extra 2 cases), the seronegative from 24 to 85 (extra 61 cases) — so 82% less likely in the seropositive group. If this isnt evidence of protection against the South African variant, I dont know what is. Also note post dose 1, 12.4% efficacy in symptomatic disease (non significant) and negative efficacy in all NAAT confirmed infections and moderate covid disease.

More negative efficacy

Verdict: No efficacy, also great information regarding protection from natural immunity against the south african variant which is the variant which has showed the greatest reduction in neutralisation in in-vitro studies. Never mind that South african variant (beta) itself is being outcompeted by delta variant now in South Africa. Remember beta is the variant with the greatest reduction in neutralizing antibody titres (moreso than delta) when looking at vaccinated and convalescent sera. So the protection of natural infection induced immunity even from beta is a good sign. However do need to acknowledge like in the novavax trial, at the early interim analysis point there are an excess number of cases in the seropositive group. However since a very clear and large later protective effect emerges, it seems likely to me that the early cases in seropositive is picking up prolonged RNA shedding. I liked this trial data so much I wrote a follow up article about it.

Sputnik, Covaxin and Novavax UK phase 3 trials do not report any data of the number of cases in the seropositive patients, maybe when they apply for fda or ema authorisation they will provide this data and I will be on a lookout for it.

So to summarise, pharma companies have made the data hard to get and not very forthcoming with it in journal articles, but even when considering variants of concern, there really isnt much data to convince that there is efficacy when vaccinating people with evidence of prior infection. I dont think it is impossible that there is efficacy, what is highly likely is that the absolute risk reduction will be much smaller in people with prior infection.

Real world studies comparing vaccinated and convalescent individuals

1. COVID-19 vaccine coverage in health-care workers in England and effectiveness of BNT162b2 mRNA vaccine against infection (SIREN): a prospective, multicentre, cohort study — Hall et al. Lancet May 2021 — again some excellent data from the Siren study.

They were comparing vaccinated, prior infected and infection naive.

They stated 85% efficacy 7 days post second dose, 70% efficacy 21 days post first dose. They stated the previously infected had 90% protection from reinfection — which is even greater than the 85% efficacy post dose 2 — but the time periods of comparison are different — so let’s not focus too much on that.

The most interesting finding though is the relative and absolute risk reduction comparison between the negative and the postitive cohort. Both the relative and the absolute risk reductions are much smaller after vaccination in the positive cohort. Is the risk reduction in the positive cohort post vaccination even statistically significant? Even the study authors were unable to comment on hazard ratios after vaccination in the positive cohort due to “insufficient information”.

2. Reduced Risk of Reinfection with SARS-CoV-2 After COVID-19 Vaccination in
Kentucky, May–June 2021 — Cavanaugh et al. CDC August 2021.

The first real world data showing benefit in vaccinating seropositive patients? Patients in the reinfected cohort had a 2.34 odds ratio of being unvaccinated when compared with the non-reinfected cohort.

What’s really lacking is breakdown of vaccination rate by age group. It is a retrospective observational study with several limitations — I have written a separate article addressing it — please have a read. I welcome the study though, and what we need is similar studies but with controlling for testing frequency, severity outcomes and vaccination rates by age.

When you have no or minimal benefits to vaccination, then the risks even if small might not worth it anymore. What about the vaccination of seropositive children, or seropositive pregnant women? Can anyone really convince us that the risk-benefits are positive in the direction of vaccination?

I have not even started talking about in vitro lab studies and the data is often conflicting, but for me real world studies are much more useful than trying to extrapolate from invitro results.

Of course I have to state that all current vaccines aim to produce the spike protein of the “Wuhan wild type” variant, if booster vaccines to cover variants are introduced, the above analysis could change.

Please do suggest if you have further data in support or against my findings above.

Feel free to share my post elsewhere if needed.