\documentclass[10pt]{article}
\usepackage[hmargin=1.5cm,top=2cm,bottom=2cm]{geometry}
\usepackage{mathtext}
\usepackage{multicol}
\setlength\columnsep{15pt}
\usepackage{amsmath}
\usepackage{amssymb}
\usepackage{array}
\usepackage{booktabs}
\usepackage{tabularx}
\usepackage[auth-sc]{authblk}
\usepackage{longtable}
\usepackage{multirow}
\usepackage{hyperref}
\usepackage{enumerate}
\usepackage[labelfont=bf]{caption}
\usepackage[usenames,dvipsnames]{xcolor}
\usepackage{mdframed}
\usepackage{graphics}
\usepackage{multirow}
\usepackage{rotating}
\usepackage{array}
\usepackage{capt-of}
\usepackage{lscape}
\usepackage{caption}
\usepackage{breakurl}
\usepackage{todonotes}
\usepackage{hanging}
\usepackage{pagecolor}
\usepackage[final]{pdfpages}
\usepackage[leftFloats,CaptionAfterwards]{fltpage}
\usepackage[numbers,super,sort&compress]{natbib}
\setlength{\bibsep}{0pt plus 0.3ex}
\usepackage{abstract}
\usepackage{enumitem}
\usepackage{soul}
\usepackage{titlesec}
\titleformat{\section}[block]{\large\bfseries\filcenter}{\thesection.}{0.4em}{}
\titleformat{\subsection}[block]{\normalsize\sc\bfseries\filcenter}{\thesubsection.}{0.4em}{}
\titleformat{\subsubsection}[block]{\normalsize\sc\itshape\filright}{\thesubsection.}{0.4em}{}
\setcounter{secnumdepth}{5}

\usepackage[hang]{footmisc}
\setlength\footnotemargin{0em}

\setlength{\skip\footins}{0.75cm}

\makeatletter
\renewcommand\footnoterule{%
  \kern-3\p@
  \hrule\@width \textwidth height 1.5pt
  \kern2.6\p@}
\makeatother

\makeatletter
\def\@biblabel#1{\@ifnotempty{#1}{#1.}}
\makeatother

\newcommand{\filllastline}[1]{
\setlength\leftskip{0pt}
\setlength\rightskip{0pt}
\setlength\parfillskip{0pt}
#1}

\newenvironment{Figure}
{\par\medskip\noindent\minipage{\linewidth}}
{\endminipage\par\medskip}


\title{\bf SARS-CoV-2 reinfections during the first three COVID-19 waves in Bulgaria}
\renewcommand\Authfont{\scshape\normalsize}
\author[1]{Georgi K. Marinov}
\author[2]{Mladen Mladenov}
\author[3]{Ivailo Alexiev}
\author[4]{Antoni Rangachev}
% \author[5]{LIST ADDITIONAL AUTHORS}
\renewcommand\Affilfont{\itshape\normalsize}
% \affil[1]{Department of Mathematics, University of Chicago, Chicago, IL 60637, USA}
\affil[1]{Department of Genetics, Stanford University, Stanford, CA 94305, USA}
\affil[2]{Premier Research, Morrisville, NC 27560, USA}
\affil[3]{National Center of Infectious and Parasitic Diseases, Sofia, Bulgaria}
\affil[4]{Institute of Mathematics and Informatics, Bulgarian Academy of Sciences, Sofia 1113, Bulgaria}
% \affil[$\#$]{Corresponding author}
% \affil[*]{These authors contributed equally}
\date{}

\begin{document}
\maketitle

% \centerline{}
% \centerline{}
\begin{abstract}

\noindent {\normalsize \textbf{Background: The COVID-19 pandemic has had a devastating impact on the world over the past two years. One of the key questions about its future trajectory is the protection from subsequent infections and disease conferred by a previous infection, as the SARS-CoV-2 virus belongs to the coronavirus, a group of a viruses the members of which are known for their ability to reinfect convalescent individuals. Bulgaria is presents a somewhat unique context in which to study this question, with high rates of previous infection combined with low vaccination rates and an elderly population.}  \\

\noindent \textbf{Methods: We use detailed governmental data on registered COVID cases to evaluate the incidence and outcomes of COVID-19 reinfections in Bulgaria in the period between March 2020 and December 2021.}  \\

\noindent \textbf{Results: For the period analyzed, a total of 4,106 cases of individuals infected more than once were observed, including 31 cases of three infections and one of four infections. The number of reinfections increased dramatically during the Delta variant-drive wave of the pandemic towards the end of 2021. We observe a comparable rate of severe outcomes  (hospitalization and death) in primary infections and reinfections, and a reduction in breakthrough infections in vaccinated individuals. }

\noindent \textbf{Conclusions: In the available datasets from Bulgaria, prior infection appears provide no to limited protection from severe outcomes.} 
}
\centerline{}
\centerline{}
\end{abstract}

\begin{multicols}{2}

\section*{Introduction}

The COVID-19\cite{Wang2020,Zhou2020,Huang2020} pandemic has become the most significant public health crisis in more than a century, and is still rapidly developing. An important question for its future trajectory, especially given the large and steadily growing number of infected individuals in most countries, is the degree of protection from subsequent infection and serious disease that prior SARS-CoV-2 infection and recovery confers.

SARS-CoV-2 is belongs to the coronavirus family, of which four different endemic human viruses were known prior to the pandemic -- HCoV-OC43\cite{McIntosh1967OC43,McIntosh1967PNAS}, HCoV-229E\cite{Hamre1966}, HCoV-NL63\cite{Fouchier2004,Pyrc2004} and HCoV-HKU1\cite{Woo2006,Lau2006,Vabret2006}. These usually cause common colds (around 10-15\% of colds, depending on the source\cite{Wat2004,Makela1998,Larson1980,Nicholson1997}, are considered to be caused by them), and, as is common with respiratory viruses\cite{Yewdell2021}, they cause repeated reinfections through people's lifetimes\cite{Callow1990}. Large epidemics are thought to occur at two- to three-year intervals\cite{Monto1974,KahnMcIntosh2005}, though these are generally not noticed by society due to the overall mild nature of common-cold coronaviruses.

Given that SARS-CoV-2 belongs to the same family of viruses, the concern that a similar host-pathogen dynamics involving frequent reinfections will be osberved with it too was natural.

The first reports of repeated infections appeared very early in the pandemic\cite{Mahase2020}, but at the time it was difficult to exclude the possibility of simple persistence of viral RNA as opposed to true reinfections. Viral genomic sequencing (showing that distinct viral lineages infected the same individual more than once) eventually proved beyond reaosnable doubt that reinfection occurs but it was still initially seen as an exotic and surprising phenomenon\cite{deVrieze2020,Arafkas2021,Ledford2020,JabbariRezaei2020,Duggan2021,YorkA2020,Law2020}. Since then, however, reinfection has been proven to be far from rare phenomenon as a large body of case reports has accumulated from around the world\cite{Tang2021,LeeJS2020,Mulder2020,ToKK2020,LarsonD2020,GoldmanJD2020,Tillett2021,Prado-Vivar2021,Sevillano2021,GuptaV2020,Shastri2021,Resende2021,Nonaka2021,Mahajan2021,Garvey2021,Harrington2021,Zucman2021,Ramirez2021,VanElslande2021,Tomassini2021,Hoang2020,Bongiovanni2021,Salzer2021,Sicsic2021,Salcin2021,Selhorst2020,Scarpati2021,Garg2021,Goes2021,Ahmadian2021,Amorim2021,Bongiovanni2021a,Salehi-Vaziri2021,Bonifacio2020,SilvaMSD2020,YuALF2021,Romano2021,West2021,Sharma2020,Ozaras2020,LeeJS2020,Colson2021,Selvaraj2020,AlFehaidi2021,deBrito2020,Tillett2021,MulderM2020,Hanif2020,Ferrante2021,Arteaga-Livias20201,Novoa2021,Alzedam2021,Camargo2021,DiazY2021,Aguilar-Shea2021,Rodriguez2021,Sanyang2021,Loconsole2021,Garduno-Orbe2021,Vora2021,Sanchez2021,Fageeh2021,Novazzi2021,Letizia2021,Staub2021,Brehm2021,Konstantinou2021,Shoar2021,Rani2021,Vancsa2021,Das2021,Fintelman-Rodrigues2021,Teka2021,RoyS2021,Fernandes2021,Ul-Haq2020,Krishna2021,LeungHossain2021,Salehi-Vaziri2021b,Romera2021,Yadav2021,Cavanaugh2021,Vetter2021,LeeJT2021,Kulkarni2021,AdrielleDosSantos2021,Inada2021,Fonseca2021,ZhouX2021,Massachi2021,Massanella2021,Romera2021,Alshukairi2021,Bader2021,Awada2021,Zanferrari2021,ZhangN2021,Naveca2021}, including most recently even cases of third infections\cite{Hasanzadeh2021,Shastri2021,Pulliam2021}.

A number of cohort studies have also been published\cite{VitaleJ2021,Babiker200,Boyton2021,Hansen2021,Gousseff2020,Lawandi2021,IwasakiA2020,LeidiA2021,HallVJ2021,Abu-Raddad2020,Ghorbani2021,Santiago-Espinosa2021,Peghin2021,Wilkins2021,AliAM2021,Lawandi2021,Ringlander2021,Dobano2021,SanchesMontava2021,Lutrick2021,Crellen2021,Breathnach2021,OMurchu2021,Leidi2021,Iruretagoyena2021,Turnerjs2021,Qureshi2021,ZareF2021,Fabianova2021,Davido2021,Brouqui2021,Dimeglio2021,Breathnach2021b,Rennert2021,RaddadLJ2021,Hanrath2021,Lumley2021,Hamed2021,Addetia2020,Mack2021,Pulliam2021,Abu-Raddad2021,Bean2021,Mensah2021,Levin-Rector2021,Malhotra2022,Mao2021,Salehi2021,Mensah2022,McKeigue2021,Maier2021,Goldberg2021,Chivese2021} but most of these suffer from various drawbacks, such as the inclusion of a very narrow time window after initial infection, focus on healthcare workers (meaning that the age distribution is not representative of the overall population), and the fact that most such studies were carried out prior to the appearance of the more highly derived SARS-CoV-2 variants that have come to dominate the pandemic in 2021. The importance of comprehensive population sampling was shown by a recent reinfection study from Denmark\cite{Hansen2021}, which found protection from reinfection of only 47.1\% among those 65 and older during the late-2020 surge as opposed to 80.5\% for the general population. The importance of variants was stressed by the placebo trial of the Novavax vaccine in South Africa\cite{Shinde2021}, which showed no protection of prior infection against infection with the dominant at the time there B.1.351 variant\cite{Tegally2021}. Later, in the end of 2021, the Omicron variant emerged, with a very high degree of immune escape\cite{Muik2022,Dejnirattisai2022,Hui2022,Meng2022} and the ability to reinfect convalescent individuals at a high rate\cite{Pulliam2021,Altarawneh2022,Nunes2022,Lyngse2021,Chivese221,Lacy2022}.

In this work, we analyze available reinfection data in Bulgaria prior to the emergence of the Omicron variant, when largely antigenically homologous variants were circulating. Bulgaria has been one of the most seriously affected by the pandemic countries\cite{Rangachev2021}, having experienced three major COVID waves in 2020-2021 and exhibiting excess mortality approaching 1\% of its population\cite{Karlinsky2021}. In the same time, only a small portion of the population has been vaccinated $(\leq$30\% by the end of 2020), meaning that it provides a unique context in which the clinical impact of reinfections can be observed in a previously severely impacted population with an age structure skewed towards the elderly, and without the confounding factor of high vaccination coverage. We identify 4,106 reinfected individuals out of $leq$700,000 cases in the country prior to December 2021. The frequency of reinfection increased substantially during the third wave driven by the Delta variant, at which point reinfection represented $\sim$2.2\% of cases, with protection conferred by previous infection $\sim$77.5\%. The severity of reinfection was comparable to that of primary infection, while severity was reduced in breakthrough infections in vaccinated uninfected subjects.

\begin{figure*}
\begin{center}
\includegraphics[width=15cm]{Fig1.png}
\captionsetup{singlelinecheck=off,justification=justified}
\caption{
{\bf Suspected SARS-CoV-2 reinfections Bulgaria over time}. 
(A) Primary infections in Bulgaria over time. Bulgaria has so far experienced three distinct epidemiological waves of COVID, with peaks in November 2020, March 2021, and October 2021.
(B) Number people eligible to be considered for reinfection, i.e. people who have tested positive and 90 days have elapsed since that positive test.
(C) Dominant variants in Bulgaria over time. The first wave Bulgaria was driven by initial B.1/B.1.x derivative variants. The second wave was associated with the Alpha/B.1.1.7 variant. The third wave was dominated by the Delta/B.1.617.2 variant and its AY.x derivatives
(D) Number of probable reinfections over time in Bulgaria (per week).
}
\label{Fig1}
\end{center}
\end{figure*}

\begin{figure*}[!ht]
\begin{center}
\includegraphics[width=18.5cm]{Fig2.png}
\captionsetup{singlelinecheck=off,justification=justified}
\caption{
{\bf Time between primary infections and reinfections and impact of variants}. 
(A) Distribution of the length of the interval between primary infection and reinfection.
(B) Primary infections and reinfections by wave and dominant variant. Waves were defined as follows: ``initial infections'' refers to the period prior to September 2020; ``1st wave'' refers to the period between September 2020 and the middle of January 2020, during which D614G variants without other major mutations were dominant; the ``2nd wave'', between mid-January 2021 and June 2021 was dominated by the B.1.1.7/Alpha variant; the ``third wave'', dominated by the B.1.617.2/Delta variant, began in July 2021.
}
\label{Fig2}
\end{center}
\end{figure*}

\section*{Methodology}

\subsection*{Datasets}

\subsubsection*{Primary infection and reinfection data}

\hl{XXX WRITE DESCRIPTION OF DATA}

Reinfections were defined as cases of two positive PCR tests spaced $\geq$90 days apart. 

Breakthrough reinfections were defined as cases of a second such positive tests at least one day after the second dose of vaccine received.

\subsubsection*{SARS-CoV-2 sequencing data}

Information about sequenced SARS-CoV-2 genomes was obtained from the GISAID database\cite{GISAID}.

% \subsection*{Statistical analysis}

% \hl{XXX}

% We calculated the rate of infection as the number of individuals with positive PCR tests during the second surge divided by the cumulative number of person-days at risk. We calculated the number of days at risk for each individual in the sample as the number of days from Sept 1, 2020, until the first positive test, or Dec 31, 2020, whichever came first. We censored follow-up time in the event of death. This non-informative censoring mechanism essentially assumed a similar infection rate would have been observed among those who died if they had survived, as was observed among the survivors with the same exposure status (whether previously infected or uninfected). We calculated the adjusted rate ratio (RR) and accompanying 95% CI using Poisson regression, adjusted for sex, age group (0–5, 6–14, 15–24, 25–34, 35–44, 45–54, 55–64, 65–74, 75–84, and ≥85 years), and test frequency (number of PCR tests done on each person in 2020 categorised as 1–2, 3–5, 6–10, and ≥11 tests) to control for potential confounding. Protection against repeat infections was calculated as 1 – adjusted RR, analogous to the method of estimating vaccine effectiveness from observational data.

% https://www.medrxiv.org/content/10.1101/2021.10.17.21265101v1

\subsection*{Data Availability}

All datasets and associated code can be found at \hl{XXXX}.

\section*{Results}

\subsection*{Suspected SARS-CoV-2 reinfection cases in Bulgaria}

In order to identify SARS-CoV-2 reinfection cases in Bulgaria, we obtained datasets on the incidence and clinical outcomes of primary and subsequent infections up to December 9 2021. We classified cases as reinfections if $\geq$90 days have passed between testing positive on at least two different occasions.

After largely escaping the first global wave of infections in the first half of 2020, Bulgaria experienced three major waves of COVID-19, in October-December 2020, in February-April 2021, and in the later months of 2021, of roughly equal magnitude (Figure \ref{Fig1}A). Under this criterion, the eligible population to be considered for reinfection was $\sim$200,000 individuals after the first wave, doubling to $\geq$400,000 after the second  (Figure \ref{Fig1}B). These waves were driven by different variants of the SARS-CoV-2 virus. The first wave was dominated by B.1 lineages antigenically identical to the ancestral strain. The second wave consisted almost entirely of the Alpha/B.1.1.7 variant\cite{Kraemer2021,Davies2021}, while in the third wave the globally dominant by then Delta variant\cite{Mlcochova2021} constituted almost all cases (Figure \ref{Fig1}C). We have defined for the purposes of our analyses the dividing lines between these waves as mid-January 2021 and beginning of June 2021.

In total, we identified 4,106 cases of individuals infected more than once, including 31 cases of people infected three times and one case of a quadruple infection.

The number of reinfections in the first wave in late 2020 was small, peaking at $\leq$100 such case weekly, reflecting the low incidence of COVID earlier that year (Figure \ref{Fig1}C). A larger though still small number of reinfections were observed during the Alpha wave in the first half of 2021. The bulk of reinfections came during the Delta wave in the second half of the year, peaking at 755 a week at the end of October 2021. During the Delta wave reinfections constituted $\sim$2.3\% of cases in Bulgaria. Taking into account the number of eligible for reinfection individuals, this corresponds to protection from reinfection of $\sim77.5$\% \hl{(XXX CI MISSING XXX)}. 

We then examined the time between primary and subsequent infections. We observe a peak at approximately a year from the initial infection, but the distribution is heterogeneous and a large number of reinfections are observed all throughout the interval from 90 to 360 days (Figure \ref{Fig2}A). These numbers correspond to a cohort of people who were infected in the first wave and then reinfected in the Delta wave ($n=1,674$), and another group of people infected during the Alpha wave and then reinfected during the Delta wave  ($n=1,435$). 

\begin{figure*}
\begin{center}
\includegraphics[width=18.5cm]{Fig3.png}
\captionsetup{singlelinecheck=off,justification=justified}
\caption{
{\bf Clinical severity of SARS-CoV-2 reinfections in previously infected individuals in Bulgaria }. 
(A) Percentage of hospitalizations among cases in primary infections, breakthrough infections (infections in vaccinated individuals), reinfections (divided into reinfections in the unvaccinated and breakthrough reinfections) 
(B) Percentage of deaths among cases in primary infections, breakthrough infections (infections in vaccinated individuals), reinfections (divided into reinfections in the unvaccinated and breakthrough reinfections). Binomial proportion confidence intervals were estimated using the Clopper-Pearson exact binomial interval method.
}
\label{Fig3}
\end{center}
\end{figure*}

\subsection*{Clinical severity of reinfections}

Next we analyzed the clinical outcomes of reinfections and compared it to outcomes from primary infections and from infections in vaccinated individuals (``breakthorugh infections''). 

Among the 4,106 reinfections, 413 were also ``breakthrough reinfections'', i.e. the reinfection occurred after a vaccination course was completed. We divided the reinfection cases into separated unvaccinated and breakthrough reinfection categories.

A total of 84 fatalities were recorded within the reinfected cases, one of them within the set of 31 third infections. This corresponds to an apparent lower case fatality rate (CFR) than the total CFR in Bulgaria for the studied period ($\sim$2\% compared to $\sim$4.2) but the populations in such a comparison are not age matched. 

We divided cases in all four categories into age groups and compared the rates of hospitalizations and fatalities in each (Figure \ref{Fig3}). This analyses reveals a comparable rate of hospitalizations between primary infections and reinfections across all age groups, and a slight reduction of risk of death (though within overlapping confidence intervals). In contrast, the severity of breakthrough infections was significantly reduced compared to primary infections in the unvaccinated, although that effect diminished in the higher age groups (consistent with previous findings of lower vaccine efficacy in the elderly\cite{Cohn2022,Bar-On2021,Collier2021}).

% \begin{figure*}
% \begin{center}
% \includegraphics[width=18.5cm]{Fig4.png}
% \captionsetup{singlelinecheck=off,justification=justified}
% \caption{
% {\bf Relative severity of primary infections and reinfections}. 
% (A) Distribution of severity in primary infections and reinfections \hl{no hospitalization, hospitalization, ICU, death; stacked bar plots, side by side, by age group} 
% (B) \hl{Odds ratios for hospitalization, ICU, death by age group -- bar plots with bars}
% (C) \hl{Odds ratios for hospitalization, ICU, death by vaccination status -- bar plots with bars}
% }
% \label{Fig4}
% \end{center}
% \end{figure*}

\section*{Conclusions}

In this study we evaluated the rate of incidence and the clinical outcomes of SARS-CoV-2 reinfections during the first three waves of the COVID-19 pandemic in Bugaria. The bulk of reinfections happened during the Delta variant-driven wave, with prior infection providing protection from reinfection in the $\sim$75-80\% range. However, within the recorded reinfections, clinical severity was not reduced relative to primary infections, in contrast to the observed reduction in severity in breakthrough infections in the vaccinated. Results regarding the relative severity of reinfections in the literature have ranged from finding no difference in the severity of reinfections and primary infection to finding considerable (though rarely very high) degree of reduction from severe outcomes\cite{Mensah2022}. With the possible caveat that case ascertainment rates in Bulgaria have likely been quite low throughout the pandemic, due to the lack of as comprehensive testing and recording of positives as in other countries, thus leading to a bias towards documenting symptomatic infections, the available data from Bulgaria points to reinfections being as severe as initial infections.

\section*{Notes}

% \subsection*{Reproducibility Statement}

\subsection*{Competing Interests} 

The authors declare no competing interests.

\section*{Author contributions}

A.R. and G.K.M. conceptualized the study. A.R. and M.M. collected datasets. G.K.M. and M.M. carried out data analysis. G.K.M. wrote the manuscript with input and supervision from all authors.

\section*{Acknowledgments and Funding}

\hl{XXXX}

\begin{thebibliography}{100}

% \section*{References}

\input{references}

\end{thebibliography}

\end{multicols}

\end{document}

\clearpage

\setcounter{table}{0}
\renewcommand{\tablename}{Supplementary Table}
\setcounter{figure}{0}
\renewcommand{\figurename}{Supplementary Figure}

\setcounter{page}{1}
\renewcommand\thepage{{SM }\arabic{page}}

\begin{center}
% {\LARGE \textbf{\begin{spacing}{1.1}XXXX. \\ Supplementary Materials\end{spacing} }}
{\LARGE \textbf{Supplementary Materials}}
\end{center}

% \section*{Supplementary Figures}\label{SM}

% \begin{figure*}[!ht]
% \begin{center}
% \includegraphics[width=18.5cm]{FigS1-age-groups-65-69.png}
% \end{center}
% \captionsetup{singlelinecheck=off,justification=justified}
% \caption{
% {\bf \hl{XXXXX} }. 
% (A) \hl{XXXXX}
% (B) \hl{XXXXX}
% (C) \hl{XXXXX}
% }
% \label{FigS1}
% \end{figure*}

% \clearpage

% \section*{Supplementary Tables}\label{ST}

% \input{TableS1-Europe-excess-mortality-total}

\end{document}