Systemic reactogenicity following homologues and heterologous prime-boost AZD1222 and BNT162b2 COVID-19 vaccination of 2862 healthcare workers compared with an unvaccinated population

During spring 2021, AZD1222 and BNT162b2 were used as prime and BNT162b2 as booster COVID-19 vaccines in Denmark. We obtained self-reported information on systemic reactogenicity day-by-day during two weeks for 2862 healthcare workers vaccinated with heterologous AZD1222 + BNT162b2 or homologous BNT162b2 + BNT162b2 regimens and compared prevalences of symptoms with unvaccinated healthcare workers. We found comparable systemic reactogenicity during the first week in the two vaccine regimens and no reactogenicity during the second week. Most of the symptoms returned to a level equal to the control population four days after booster vaccination.


Abstract
During spring 2021, AZD1222 and BNT162b2 were used as prime and BNT162b2 as booster COVID-19 vaccines in Denmark. We obtained self-reported information on systemic reactogenicity day-by-day during two weeks for 2862 healthcare workers vaccinated with heterologous AZD1222+BNT162b2 or homologous BNT162b2+BNT162b2 regimens and compared prevalences of symptoms with unvaccinated healthcare workers. We found comparable systemic reactogenicity during the first week in the two vaccine regimens and no reactogenicity during the second week. Most of the symptoms returned to a level equal to the control population four days after booster vaccination.

Background
Heterologous AZD1222 and BNT162b2 COVID-19 vaccination regimens may activate strong immunogenicity and reactogenicity. [1][2][3][4][5] Mixing of vaccines can be a strategy to contain new severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants due to a higher efficacy. [5,6] However, reports differ on whether heterologous vaccine regimens induce greater reactogenicity [2] compared to homologues regimens, [1,[3][4][5] and more research is warranted. [5,6] In Denmark, BNT162b2 was used continuously as a prime dose from November 17, 2020 and as a booster dose from January 18, 2021. AZD1222 was used as a prime dose from February 15 until March 11, 2021, when it was put on hold due to reports on thrombotic response. [7] Since April 18, BNT162b2 was in the current population used as the booster following a prime dose with AZD1222. Symptoms of systemic reactogenicity are similar to symptoms resulting from common cold and other upper respiratory tract infections,[8] and we therefore compared the symptoms following AZD1222+BNT162b2 and BNT162b2+BNT162b2 regimens with individually matched unvaccinated controls.

Methods
We followed healthcare workers, technical, administrative, and other staff (hereafter, called healthcare workers) who received either heterologous AZD1222+BNT162b2 or homologous BNT162b2+BNT162b2 prime-boost vaccination and their daily matched controls, who were unvaccinated healthcare workers dayby-day for 14 days. Neither were previously tested positive for SARS-CoV-2 with reverse transcription polymerase-chain-reaction (RT-PCR). The study population was recruited into a COVID-19 risk factor study by e-mail sent to all healthcare workers of the Central Denmark Region November 17, 2020.[9, 10] Enrolment continued until March 17, 2021.
Participants were prompted daily via text message on their smartphone to report (yes/no) headache, muscle pain, fever, fatigue, and diarrhoea within the last 24 hours. Height and weight were reported at baseline.
Information on vaccination and RT-PCR testing for SARS-CoV-2 was obtained from regional registers with complete coverage.
Participants receiving the BNT162b2+BNT162b2 regimen were followed from January 18, 2021 and participants receiving the AZD1222+BNT162b2 regimen from April 14, 2021 when the first booster doses were administered. All were followed until testing positive for SARS-CoV-2, 13 days after receiving the booster, or end of study June 30, 2021, whichever came first.
The calendar time for the two vaccination regimens were on average displaced by 85 days (figure 1), and because the symptoms are similar to symptoms resulting from prevalent upper respiratory tract infections that show significant seasonal variation,[8] we randomly selected four unvaccinated controls with replacement for each day a vaccinated participant reported symptoms (i.e. date of follow-up). Controls were matched on date of follow-up +/-two days, sex and age (<40, 40-49, 50-59, ≥60 years). We plotted symptom prevalences by days since booster vaccination. We compared each vaccine regimen with the controls by odds ratios (ORs) obtained by conditional logistic regression analyses of all daily reports adjusted for days of follow-up, occupation (five groups), and BMI (<20, 20-24, 25-29, ≥30). ORs are reported for two one-week intervals in order to compare our findings with other vaccine studies. In additional sensitivity analyses we report ORs in shorter time intervals (days 0-3, 4-7, 8-13). We obtained 95% confidence intervals (CI) by 50 bootstrap samples to account for the repeated measurements within participants. A p value <0.5 was considered statistically significant. Analyses were conducted with STATA version 17 (StataCorp LLC, College Station, TX).
All participants gave written informed consent. The Central Denmark Regional Scientific Ethical Committee approved that ethical approval was not required (1-10-72-1-20).

Results
A total of 6861 healthcare workers participated in the dynamic risk factor study, 947 received the week interval among participants with a median age of 57 years. [2] In the two other studies longer intervals of 8-12 [3] and 10-12 [1] weeks were used in the heterologous regimens, while a similar interval of three weeks were used in the homologues regimen. [1] Also, their populations were younger (median age 37 years [1] and mean age 44 [3]). Our study population had a mean age of 50.3 years and 11-13 weeks between the vaccines in the heterologous regimen, and 51.0 years and 3-4 weeks in the homologous regimen.
Strengths of our study are the prospective design, day-by-day symptom reports beyond the standard seven days of vaccine trials, a high participation rate, and strict control for date because symptoms of reactogenicity are similar to symptoms resulting from common seasonal upper respiratory infections. [8] Furthermore, the inclusion of a non-vaccinated control group allowed us to document that after one week the vaccinated participants had no more symptoms than the control population.
Limitations include participants being aware of the administered vaccines, which may have affected symptom reporting, the lack of information on symptom severity, and the indirect comparison of the two regimens due to seasonal variation.
In conclusion, compared with the control population, heterologous AZD1222+BNT162b2 and homologous

Declaration of interests
☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
☐ The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: