key: cord-0260271-ocu0ja6u
authors: Abu-Raddad, L. J.; Chemaitelly, H.; Ayoub, H. H.; Tang, P.; Coyle, P.; Hasan, M. R.; YASSINE, H. M.; Benslimane, F.; Al Khatib, H. A.; Al Kanaani, Z.; Al Kuwari, E.; Jeremijenko, A.; Kaleeckal, A. H.; Latif, A. N.; Shaik, R. M.; Abdul Rahim, H. F.; Nasrallah, G.; Al Kuwari, M. G.; Butt, A. A.; Al Romaihi, H. E.; Al Khal, A.; Al-Thani, M. H.; Bertollini, R.
title: Effect of vaccination and of prior infection on infectiousness of vaccine breakthrough infections and reinfections
date: 2021-07-30
journal: nan
DOI: 10.1101/2021.07.28.21261086
sha: 9929b6dd442f6f0868585324005f3e1eede114d2
doc_id: 260271
cord_uid: ocu0ja6u

SARS-CoV-2 breakthrough infections in vaccinated individuals and in those who had a prior infection have been observed globally, but the transmission potential of these infections is unknown. The RT-qPCR cycle threshold (Ct) value is inversely correlated with viral load and culturable virus. Here, we investigated differences in RT-qPCR Ct values across Qatar national cohorts of primary infections, reinfections, BNT162b2 (Pfizer-BioNTech) breakthrough infections, and mRNA-1273 (Moderna) breakthrough infections. Through matched-cohort analyses of the randomly diagnosed infections, the mean Ct value was higher in all cohorts of breakthrough infections compared to the cohort of primary infections in unvaccinated individuals. The Ct value was 1.3 (95% CI: 0.9-1.8) cycles higher for BNT162b2 breakthrough infections, 3.2 (95% CI: 1.8-4.5) cycles higher for mRNA-1273 breakthrough infections, and 4.0 (95% CI: 3.4-4.6) cycles higher for reinfections in unvaccinated individuals. Assuming a linear relationship between viral load and infectiousness, these differences imply that breakthrough infections are at least 50% less infectious than primary infections in unvaccinated individuals. Public health benefits of vaccination may have been underestimated, as COVID-19 vaccines not only protect against acquisition of infection, but also appear to protect against transmission of infection.

Coronavirus Disease 2019 vaccines have demonstrated protection against the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 1-3 . However, the efficacy against acquisition of infection is imperfect in that vaccines have an efficacy 100% S VE  , particularly against variants of concern 4,5 . Breakthrough infections in vaccinated individuals have been documented in various countries [4] [5] [6] , but the transmission potential of these infections is poorly understood. It is conceivable that vaccinated persons who acquire the infection may be less infectious than unvaccinated persons who acquire the infection, as the vaccine-primed immune response may attenuate the natural history of infection and reduce viral replication, leading to lower viral load and faster infection clearance 7 . There is evidence that seem to support this hypothesis for SARS-CoV-2 infection 8, 9 . Therefore, COVID-19 vaccines may not only be efficacious against acquisition of infection ( Figure 1D ).

The BNT162b2 and mRNA-1273 vaccines have been the vaccines of choice in the national immunization campaign in Qatar 4, 5, 11, 12 . In total, there has been 4,777 breakthrough infections in those fully vaccinated with BNT162b2 (0.61% of those fully vaccinated; Figure 1C ) and 306 mRNA-1273 breakthrough infections in those fully vaccinated with mRNA-1273 (0.09% of those fully vaccinated; Figure 1D ). Of note that mass immunization started with the BNT162b2 vaccine, and the mRNA-1273 vaccine was introduced only several weeks later.

Primary infection was defined as the first RT-qPCR-positive test for a given individual.

Reinfection was defined as the first RT-qPCR-positive test that occurred ≥90 days after the primary infection [13] [14] [15] [16] . Breakthrough infection in a vaccinated individual was defined as an RT-qPCR-positive test 14 or more days after the individual received the second vaccine dose, conditional on this RT-qPCR-positive test being the first ever positive for this individual.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.28.21261086 doi: medRxiv preprint Study populations Figure 1 shows the process for identifying eligible primary infections, reinfections, BNT162b2 breakthrough infections, and mRNA-1273 breakthrough infections. Figure 2 schematizes the six pairwise comparisons conducted between these cohorts of infections, after matching in a 1:1 ratio by sex, 10-year age group, reason for RT-qPCR testing, and calendar week of the RT-qPCR test, to control for differences in biology by sex and age, as well as exposure risk 17, 18 

In the comparisons including all RT-qPCR-confirmed infections, regardless of the reason for the RT-qPCR testing, the mean RT-qPCR Ct value was higher in all cohorts of breakthrough infections compared to the cohort of primary infections in unvaccinated individuals (Table 1 

In the comparisons including only the randomly diagnosed (asymptomatic) infections (Methods), which are perhaps most representative for differences between these cohorts of infections, given the random diagnosis, the mean RT-qPCR Ct value was also higher in all cohorts of breakthrough infections compared to the cohort of primary infections in unvaccinated individuals ( 

In the comparisons including only the symptomatic infections (Methods), the mean RT-qPCR Ct value was also higher in all cohorts of breakthrough infections compared to the cohort of primary infections in unvaccinated individuals, but the difference was smaller for BNT162b2 breakthrough infections (Table 3 and Supplementary Figure 2 ). The Ct value was 0.2 (95% CI: -All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (Table 3) .

These differences in Ct values can be translated into approximate relative differences in viral load (Methods). Since the mathematical formula linking viral load to Ct value has an exponential form (viral load is proportional to 2 n − where n is the RT-qPCR Ct value) 21 , small differences in Ct value represent large differences in viral load.

Empirically, SARS-CoV-2 viral load strongly correlates with culturable virus 10 , just as viral load for other viruses, such as HIV, is a measure of its infectiousness 22-24 . Therefore, relative differences in viral load can be considered as a proxy measure of relative differences in infectiousness (Methods), that is I VE 7 . For instance, assuming that SARS-CoV-2 infectiousness is linearly proportional to viral load, the simplest default assumption, a Ct value higher by 1.0 cycle corresponds to a lower viral load (and infectiousness) by a factor of 2 ( 1.0

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.28.21261086 doi: medRxiv preprint Table 5 shows approximate estimates for the relative infectiousness of SARS-CoV-2 infections across the cohorts of primary infections, reinfections, BNT162b2 breakthrough infections, and mRNA-1273 breakthrough infections, assuming that SARS-CoV-2 infectiousness is linearly proportional to viral load. Meanwhile, Table 6 shows the same estimates in a sensitivity analysis assuming that infectiousness is (non-linearly) proportional to some power of viral load ( vl  ), just as for HIV infection which has perhaps the best characterized relationship between viral load and infectiousness in the literature 22-24 . Here, the exponent  has been estimated at around 0.4 22-24 -that is infectiousness is approximately proportional to the square rate of viral load rather than the viral load itself.

In both sets of estimates, breakthrough infections in those vaccinated or who had a prior infection had substantially lower infectiousness than primary infections in unvaccinated persons. 

Breakthrough infections in those vaccinated or who had a prior infection have higher RT-qPCR Ct values than primary infections in unvaccinated persons. While these breakthrough infections are not uncommon, the results indicate that they have lower viral load and are less likely to be All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.28.21261086 doi: medRxiv preprint infectious than primary infections. While some of these breakthrough infections could lead to secondary transmissions, and indeed some of them did have high viral loads ( Figure 3 and Supplementary Figure 1 and 2), the risk of onward transmission appears to be substantially reduced, compared to primary infections. Thus, they are of less public health concern.

A consequence of these findings is that the public health benefits of vaccination may be underestimated. In addition to the conventional vaccine efficacy against acquisition of infection ( S VE ) that is assessed in randomized clinical trials 1-3 , there is an additional "breakthrough" efficacy against transmission ( I VE ) that augments the benefits of S VE (Tables 5 and 6 ), at least for the BNT162b2 and mRNA-1273 vaccines investigated in this study. The existence of this additional I VE efficacy implies that the reproduction number ( 0 R ) after vaccination is lower than current estimates 7 . SARS-CoV-2 incidence may decline faster with vaccine scale up 7 , and herd immunity may be easier to achieve, at somewhat lower vaccine coverage than previously predicted 7 .

One finding of this study is that there appears to be a hierarchy in infectiousness of SARS-CoV-2 infections, where primary infections in unvaccinated persons are most infectious, followed by that dominated incidence since the onset of vaccination 19,20 ), while protection against acquisition of infection was considerably higher among those vaccinated with mRNA-1273 5 and those with All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.28.21261086 doi: medRxiv preprint a prior infection [13] [14] [15] 25 . This may indicate that both This study has limitations. The number of documented mRNA-1273 breakthrough infections was small, thereby limiting the statistical precision of the comparisons involving these infections, leading to estimates with wider 95% confidence intervals, and perhaps making them more prone to bias. The study was implemented on documented infections, but other infections may have occurred and gone undocumented. It is possible that breakthrough infections in those vaccinated or who had a prior infection are less likely to be documented, perhaps because of minimal/mild or no symptoms. However, with the high rate of PCR testing in Qatar, the majority of infections are identified not because of testing symptomatic cases, but because of testing for other reasons, such as random testing campaigns, contact tracing, individual requests, routine healthcare testing, pre-travel, and at ports of entry. The results of our study were also consistent when we included only the randomly diagnosed infections.

Imperfect assay sensitivity and specificity of RT-qPCR testing may have affected infection ascertainment, but RT-qPCR testing was performed using a validated commercial platform that has been used globally and has essentially 100% sensitivity and specificity 26 (Methods). While it is established that SARS-CoV-2 viral load strongly correlates with culturable virus 10 and infectiousness, the exact mathematical relationship linking viral load to infectiousness is not known. Therefore, we generated estimates for the relative infectiousness assuming a linear dependence on viral load, and alternatively assuming a non-linear (power-law) dependence on All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. viral load, similar to that rigorously delineated for HIV infection 22-24 . Estimates for the relative infectiousness presented in Tables 4 and 5 are thus approximate and qualitative in nature, and should not be seen as precise quantitative measures.

Unlike blinded, randomized clinical trials, the investigated observational cohorts were neither blinded nor randomized. Our cohorts predominantly included working-age adults; therefore, results may not necessarily be generalizable to other population groups, such as children or the elderly. Matching was done for sex, age, reason for the RT-qPCR testing, and calendar week of the RT-qPCR test, but could not be done for other factors, such as comorbidities, as these were not available to study investigators. But inclusion of additional factors in the matching would have considerably reduced the sample sizes-breakthrough infections in those vaccinated or who had a prior infection are relatively uncommon. It is noteworthy that matching by age and sex may have served as a proxy for matching by co-morbidity, as co-morbidities are associated with older age and may differ between women and men.

In conclusion, prior immunity, whether due to vaccination or prior infection, is associated with lower SARS-CoV-2 viral load upon infection. While breakthrough infections have been observed globally, they appear less infectious than primary infections; thus, they constitute a lesser public health concern. The public health benefits of vaccination are underestimated, as both the BNT162b2 and mRNA-1273 vaccines seem not only to protect against acquisition of infection, but also against transmission of infection. These findings justify optimism and stress the urgency to scale up vaccination globally in order to robustly control infection transmission and the extent of the pandemic.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. 

LJA conceived and co-designed the study, led the statistical analyses, and co-wrote the first draft of the article. HC co-designed the study, performed the statistical analyses, and co-wrote the first draft of the article. All authors contributed to data collection and acquisition, database development, discussion and interpretation of the results, and to the writing of the manuscript.

All authors have read and approved the final manuscript.

Dr. Butt has received institutional grant funding from Gilead Sciences unrelated to the work presented in this paper. Otherwise, we declare no competing interests.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.28.21261086 doi: medRxiv preprint 

Estimates in the sensitivity analysis assuming that infectiousness is (non-linearly) proportional to some power of viral load, similar to that for HIV 22-24 , are found in Table 5 . Table 4 . (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.28.21261086 doi: medRxiv preprint health system at the port of entry upon return to Qatar, given the national requirements and to benefit from privileges associated with vaccination, such as quarantine exemption 12 In total, six types of pairwise comparisons were conducted on the cohorts of this study, after matching in a 1:1 ratio by sex, 10-year age group, reason for RT-qPCR testing, and calendar

week of the RT-qPCR test, to control for differences in biology by sex and age, as well as exposure risk 17, 18 and variant exposure 4, 19, 20 . It is noteworthy that the first SARS-CoV-2 epidemic wave in Qatar occurred before introduction of any variant of concern and peaked in late May, 2020 17, 18 (Delta 30 ) variant has been introduced only recently in Qatar, and as of July 11, 2021, it remains at low incidence 19,20 . There is no evidence that any other variant of concern is or has been responsible for appreciable community transmission in Qatar 19,20 .

Comparisons across the cohorts of infection were implemented for all RT-qPCR-confirmed infections, for only the symptomatic infections defined as RT-qPCR-positive tests conducted because of clinical suspicion due to symptoms compatible with a respiratory tract infection, and for only the randomly diagnosed (asymptomatic) infections defined as RT-qPCR-positive tests conducted with no prior reason to suspect infection and no reported symptoms compatible with a All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.28.21261086 doi: medRxiv preprint respiratory tract infection. The latter was strictly defined as an RT-qPCR-positive test conducted as part of a survey (random testing campaigns), for routine health care testing, as a pre-travel requirement, or at a port of entry upon arrival in the country 4, 5, 12, 17 .

All records of RT-qPCR testing in Qatar were examined in this study, but only samples of Reporting of the study followed the STROBE guidelines (Supplementary Table 3 ).

Nasopharyngeal and/or oropharyngeal swabs (Huachenyang Technology, China) were collected for PCR testing and placed in Universal Transport Medium (UTM). Aliquots of UTM were: extracted on a QIAsymphony platform (QIAGEN, USA) and tested with real-time reversetranscription PCR (RT-qPCR) using TaqPath™ COVID-19 Combo Kits (100% sensitivity and specificity 26 ; Thermo Fisher Scientific, USA) on an ABI 7500 FAST (ThermoFisher, USA); extracted using a custom protocol 41 on a Hamilton Microlab STAR (Hamilton, USA) and tested All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.28.21261086 doi: medRxiv preprint using AccuPower SARS-CoV-2 Real-Time RT-PCR Kits (100% sensitivity and specificity 42 ; Bioneer, Korea) on an ABI 7500 FAST; or loaded directly into a Roche cobas® 6800 system and assayed with a cobas® SARS-CoV-2 Test (95% sensitivity, 100% specificity 43 ; Roche, Switzerland). The first assay targets the viral S, N, and ORF1ab regions. The second targets the viral RdRp and E-gene regions, and the third targets the ORF1ab and E-gene regions.

All tests were conducted at the HMC Central Laboratory or Sidra Medicine Laboratory, following standardized protocols.

Only the RT-qPCR-confirmed infections diagnosed using the TaqPath 

Socio-demographic characteristics of study samples were described using frequency distributions and measures of central tendency. Differences in proportions across categorical variables between study groups were evaluated using Chi-square tests. The distributions of the RT-qPCR Ct values were illustrated using scatter plots and boxplots, and summarized using measures of central tendency and dispersion. Mean differences in the RT-qPCR Ct values between study groups and associated 95% confidence intervals (CIs) were calculated using independent t-tests. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. where n  is the difference in the RT-qPCR Ct values between these two cohorts.

SARS-CoV-2 viral load correlates strongly with cultivable virus 10 , and thus infectiousness, just as viral load for other viruses, such as HIV, is a measure of its infectiousness 22-24 . Therefore, the ratio of two viral loads in two cohorts can be considered to provide a measure of the relative infectiousness of these two cohorts. This implicitly assumes that SARS-CoV-2 infectiousness is linearly proportional to vl , the simplest default assumption for lack of data. This assumption was used in the baseline analysis for the relative infectiousness.

However, infectiousness could be related to viral load through a non-linear mathematical function. For HIV infection for instance, which has perhaps the best characterized relationship All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.28.21261086 doi: medRxiv preprint between viral load and infectiousness in the literature 22-24 , infectiousness is (non-linearly) proportional to vl  where the exponent  has been estimated at 0.3-0.5 22-24 -that is infectiousness is approximately proportional to the square rate of viral load rather than the viral load itself. In a sensitivity analysis, we estimated SARS-CoV-2 relative infectiousness assuming that infectiousness is proportional to vl  with an 0.4

, similar to that for HIV 22-24 , to provide an alternative estimate assuming a power-law mathematical function linking viral load to infectiousness.

The study was approved by the Hamad Medical Corporation and Weill Cornell Medicine-Qatar

Institutional Review Boards with waiver of informed consent.

The dataset of this study is a property of the Qatar Ministry of Public Health that was provided to the researchers through a restricted-access agreement that prevents sharing the dataset with a third party or publicly. Future access to this dataset can be considered through a direct application for data access to Her Excellency the Minister of Public Health (https://www.moph.gov.qa/english/Pages/default.aspx). Aggregate data are available within the manuscript and its Supplementary information.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. , and mRNA-1273-vaccine breakthrough infections ....................................... 2 Supplementary Table 2 ............................................................................................................................................... .......................................................................................... 6 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Tables  1-5 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 30, 2021. ; https://doi.org/10.1101/2021.07.28.21261086 doi: medRxiv preprint

Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine

Covid-19 Commission of Accademia Nazionale dei Lincei, R. COVID-19 vaccines: where we stand and challenges ahead

Effectiveness of the BNT162b2 Covid-19 Vaccine against the B.1.1.7 and B.1.351 Variants

mRNA-1273 COVID-19 vaccine effectiveness against the B.1.1.7 and B.1.351 variants and severe COVID-19 disease in Qatar

COVID-19 Vaccine Breakthrough Infections Reported to CDC -United States

Epidemiological Impact of SARS-CoV-2 Vaccination: Mathematical Modeling Analyses. Vaccines (Basel)

SARS-CoV-2 seropositivity and subsequent infection risk in healthy young adults: a prospective cohort study

Prevention and Attenuation of Covid-19 with the BNT162b2 and mRNA-1273 Vaccines

Duration of infectiousness and correlation with RT-PCR cycle threshold values in cases of COVID-19

Pfizer-BioNTech mRNA BNT162b2 Covid-19 vaccine protection against variants of concern after one versus two doses

Associations of Vaccination and of Prior Infection With Positive PCR Test Results for SARS-CoV-2 in Airline Passengers Arriving in Qatar

SARS-CoV-2 antibody-positivity protects against reinfection for at least seven months with 95% efficacy

Assessment of the risk of SARS-CoV-2 reinfection in an intense re-exposure setting

Reinfections with the SARS-CoV-2 B.1.351 variant and efficacy of natural immunity against reinfection under review at

Assessment of protection against reinfection with SARS-CoV-2 among 4 million PCR-tested individuals in Denmark in 2020: a population-level observational study

Characterizing the Qatar advanced-phase SARS-CoV-2 epidemic

Mathematical modeling of the SARS-CoV-2 epidemic in Qatar and its impact on the national response to COVID-19

Abbreviations: IQR, interquartile range. * Study groups were matched in a 1:1 ratio by sex, 10-year age group, reason for RT-qPCR testing, and RT-qPCR test calendar week. † Nationalities were chosen to represent the most populous groups in the population of Qatar.