key: cord-0696969-3du9tdej
authors: Nicolete, Vanessa C; Rodrigues, Priscila T; Johansen, Igor C; Corder, Rodrigo M; Tonini, Juliana; Cardoso, Marly A; de Jesus, Jaqueline G; Claro, Ingra M; Faria, Nuno R; Sabino, Ester C; Castro, Marcia C; Ferreira, Marcelo U
title: Interacting Epidemics in Amazonian Brazil: Prior Dengue Infection Associated with Increased COVID-19 Risk in a Population-Based Cohort Study
date: 2021-05-06
journal: Clin Infect Dis
DOI: 10.1093/cid/ciab410
sha: eb2bb9b5cbc160ef69b66957af5c6b3fd1074de9
doc_id: 696969
cord_uid: 3du9tdej

BACKGROUND: Immunity after dengue virus (DENV) infection has been suggested to cross-protect from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and mortality. METHODS: We tested whether serologically proven prior DENV infection diagnosed in September-October 2019, before the coronavirus 2019 (COVID-19) pandemic, reduced the risk of SARS-CoV-2 infection and clinically apparent COVID-19 over the next 13 months in a population-based cohort in Amazonian Brazil. Mixed-effects multiple logistic regression analysis was used to identify predictors of infection and disease, adjusting for potential individual and household-level confounders. Virus genomes from 14 local SARS-CoV-2 isolates were obtained using whole-genome sequencing. RESULTS: Anti-DENV IgG was found in 37.0% of 1,285 cohort participants (95% confidence interval [CI], 34.3% to 39.7%) in 2019, with 10.4 (95% CI, 6.7 to 15.5) seroconversion events per 100 person-years during the follow-up. In 2020, 35.2% of the participants (95% CI, 32.6% to 37.8%) had anti-SARS-CoV-2 IgG and 57.1% of the 448 SARS-CoV-2 seropositives (95% CI, 52.4% to 61.8%) reported clinical manifestations at the time of infection. Participants aged >60 y were twice more likely to have symptomatic COVID-19 than under-five children. Locally circulating SARS-CoV-2 isolates were assigned to the B.1.1.33 lineage. Contrary to the cross-protection hypothesis, prior DENV infection was associated with twice the risk of clinically apparent COVID-19 upon SARS-CoV-2 infection, with P values between 0.025 and 0.039 after adjustment for identified confounders. CONCLUSION: Higher risk of clinically apparent COVID-19 among individuals with prior dengue has important public health implications for communities sequentially exposed to DENV and SARS-CoV-2 epidemics.

A c c e p t e d M a n u s c r i p t 5

DENV was reintroduced in Brazil in 1981 and has since spread countrywide [11] , with 661 cases per 100,000 population in 2020 [3] . The study site, Mâncio Lima (urban population, 9,000), is situated in the upper Juruá Valley region of Acre State, next to the border with Peru ( Supplementary Fig. 1) , where the first dengue outbreak was recorded in 2014 [12] . Mâncio Lima experiences most DENV transmission during the rainy season, between November and April ( Supplementary Fig. 2 ). The first local COVID-19 case was notified on April 29, 2020, with a total of 1,766 laboratory-confirmed cases recorded as of 28 February 2021 ( Supplementary Fig. 2 ).

Cohort participants are members of randomly drawn (approximately 20%) urban households in 6 .0000.5467 and 30481820.3.0000.5467) approved the study protocols. Written informed consent was obtained from study participants or their parents/guardians. We analyzed sociodemographic and morbidity data and plasma samples from 1,285 individuals aged <1 to 92 y (mean, 29.9 y). Plasmas obtained in September-October 2019 ("2019 survey") were tested for anti-DENV IgG and those obtained in October-November 2020 ("2020 survey") were tested for anti-SARS-CoV-2 IgG (Figure 1 ). Information on housing quality and assets was used to derive a A c c e p t e d M a n u s c r i p t 6 household-level wealth index [13] as a proxy of socioeconomic status. To characterize SARS-CoV-2 lineages circulating at the peak of the current epidemic, we obtained two nasopharyngeal swab samples from 49 consecutive symptomatic patients (age range, 3-77 y) seeking COVID-19 testing in Mâncio Lima in August 2020. One swab was used for point-of-care antigen-based diagnosis (ECO F COVID-19 Ag test FA0054; Ecodiagnostica, Corinto, Brazil) and the other was preserved in RNA/DNA Shield (Zymo Research, Irvine, CA) for RNA extraction.

We tested plasmas from the 2019 survey for anti-DENV IgG, which persists for life following an infection [14] , using an ELISA with DENV serotype 2 viral particles (EI 266b-9601 uro mmu bec erm y) w t 98 5 se s t v ty d 95 7 s ec f c ty ccord to t e m uf cturer

During the 2020 survey, we tested plasmas from the same individuals for anti-SARS-CoV-2 IgG with an ELISA that uses the recombinant subdomain S1 of the Spike protein as antibody-capture antigen (EI 2606-9601 G; Euroimmun) [15] , with a sensitivity of 82.5% to 93.3% and specificity of 98.0% to 98.5% [16, 17] . To identify DENV seroconversion events between surveys, we tested 186 randomly chosen samples collected in 2020 from donors who were DENV seronegative in 2019. To determine whether antibodies cross-reactive to SARS-CoV-2 existed before the COVID-19 epidemic, we tested for anti-SARS-CoV-2 IgG in 105 randomly chosen plasmas obtained in 2019 from donors found to be seropositive in 2020.

A c c e p t e d M a n u s c r i p t 7

Template RNA was prepared using QIAamp Viral RNA mini kits (Qiagen, Hilden, Germany). We detected SARS-CoV-2 RNA in all 15 antigen-positive samples using a 2-plex TaqMan assay that targets the RdRP and E genes (GenScript, Piscataway, NJ) and selected 14 samples with cycle threshold <30 for whole-genome sequencing as part of a countrywide SARS-CoV-2 genomic surveillance project [18] . Nanopore sequencing on the MinION platform (Oxford Nanopore, Oxford, UK) was carried out using the ARTIC V3 multiplexed amplicon protocol [19] (Supplementary Methods online). Assembled sequences yielded at least 75% coverage of the SARS-CoV-2 genome, with at least 20-fold depth. Lineages were classified using the Pangolin software [20] and maximum likelihood phylogenetic analysis with complete reference genomes. New genome sequences were deposited in the GISAID platform (https://www.gisaid.org/; accession numbers, EPI_ISL_1251212 to EPI_ISL_1251225).

Four primary outcomes were considered ( confidence intervals (CIs) to quantify the influence of each predictor on the outcome, while M a n u s c r i p t 9 controlling for all other covariates. Models 1 and 3 do not provide OR estimates for household-level covariates. We also computed the intraclass correlation coefficient (ICC), the proportion of the variability in the outcome attributable to differences between versus within households.

Altogether, 37.0% (95% CI, 34.3% to 39.7%) of the study participants had serological evidence of DENV infection prior to the current SARS-CoV-2 pandemic. Seropositivity rates increased with age ( Figure 2A ), which is not unexpected, as the number of infections accumulates over time [23] .

However, all participants who were born in Mâncio Lima and neighboring municipalities (95.5% of the study population), except under-five children, had exactly 5 y of exposure (assuming that they did not leave the region) to DENV since dengue outbreaks started in the region [12] . However, only 24.7% of the 6-15-y old individuals had anti-DENV IgG detected in 2019, compared with 58.3% in the 60-y old individuals, suggesting that the cumulative exposure to DENV (e.g., due to human mobility) has varied significantly across age strata since 2014.

Mixed-effects multiple logistic regression analysis identified older age to be significantly associated with DENV seropositivity (Table 1) < Table 1 >

Serological evidence of SARS-Cov-2 infection was found for 452 individuals tested in 2020 (35.2%; 95% CI, 32.6 to 37.8%), with similar antibody positivity rates across age groups ( Figure 2B ).

Seropositivity was not predicted in multiple logistic regression analysis by age, sex, overnight stays out of the town, or socioeconomic status; prior DENV infection did not emerge as a risk factor for SARS-Cov-2 infection either ( groups were uniformly exposed to SARS-CoV-2 ( Figure 2B ), individuals >15 y were twice more likely to develop COVID-19 than their younger counterparts ( Figure 2C ). Of note, 6 Figure 2D ). Indeed, older age, prior DENV infection, and (marginally) wealth were positively associated with the risk of disease following SARS-CoV-2 infection (Table 4 ).

< Table 4 >

Age is likely to confound the association between prior DENV infection and COVID-19 risk, because it is strongly associated with COVID-19 outcomes and the exposure of interest. We tested whether adjusting for age as a continuous variable (model 3) would reduce residual confounding and attenuate the magnitude of association observed in models 1 and 2 (with age as a categorical variable) and model 0 (unadjusted analysis). We obtained similar OR estimates, regardless of the age variable used, for the association between dengue and COVID-19 analyzed with different multiple logistic regression models ( Figure 3 ). OR estimates were slightly larger in adjusted analyses, but the corresponding 95% CIs were wider than in model 0. There has been little temporal overlap between dengue and COVID-19 epidemics in South America [3] . Dengue cases peaked during the 2019-20 summer season in the Southern Hemisphere, rainy season in our study site, when COVID-19 transmission was at its nadir. However, the sequential epidemics have affected populations potentially exposed to nutritional deficiencies and other pathogens that can increase the susceptibility to both dengue and COVID-19 [7] Our study has many strengths. First, the longitudinal design enables us to discern temporal associations between DENV exposure and the outcomes of interest. Second, serology allows to objectively identify a prior DENV infection, regardless of symptoms. Third, we applied standardized clinical criteria [21] to define clinically apparent COVID-19 upon serologically proven SARS-CoV-2

infection.

The study also has some limitations. First, false-positive DENV serology may arise from yellow fever (YF) vaccination, which is near universal in the Amazon, but we have previously found similar DENV 

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 manuscript.

A c c e p t e d M a n u s c r i p t M a n u s c r i p t M a n u s c r i p t 

The global distribution and burden of dengue

World Health Organization. Fact Sheets. Dengue and Severe Dengue

Pan American Health Organization/World Health Organization

Genomics and epidemiology of a novel SARS-CoV-2 lineage in Manaus

Emerging SARS-CoV-2 variants

The SARS-CoV-2 pandemic: A syndemic perspective

How super-spreader cities, highways, hospital bed availability, and dengue fever influenced the COVID-19 epidemic in Brazil

Previous dengue infection and mortality in COVID-19

Computational modelling supports that dengue virus envelope antibodies can bind to SARS-CoV-2 receptor binding sites: Is pre-exposure to dengue virus protective against COVID-19 severity?

History, epidemiology and diagnostics of dengue in the American and Brazilian contexts: a review

The introduction of dengue follows transportation infrastructure changes in the state of Acre, Brazil: A network-based analysis

Estimating wealth effects without expenditure data-or tears: an application to educational enrollments in states of India

The molecular specificity of the human antibody response to dengue virus infections

Severe Acute Respiratory Syndrome Coronavirus 2-specific antibody responses in Coronavirus Disease patients

Evaluation of SARS-CoV-2 IgG antibody response in PCR positive patients: Comparison of nine tests in relation to clinical data

Detection of SARS-CoV-2 antibodies using commercial assays and seroconversion patterns in hospitalized patients

Evolution and epidemic spread of SARS-CoV-2 in Brazil

Improvements to the ARTIC multiplex PCR method for SARS-CoV-2 genome sequencing using nanopore

A dynamic nomenclature proposal for SARS-CoV-2 lineages to assist genomic epidemiology

Safety and efficacy of the BNT162b2 mRNA COVID-19 vaccine

Research electronic data capture (REDCap) -a metadata-driven methodology and workflow process for providing translational research informatics support

Dengue seroprevalence: data from the clinical development of a tetravalent dengue vaccine in 14 countries

The use of longitudinal cohorts for studies of dengue viral pathogenesis and protection

Outcome: anti-SARS-CoV-2 IgG in 2020 (n = 1285) Outcome: clinical COVID-19 in 2020 (n = 1281) Missing morbidity data

Outcome: clinical COVID-19 once infected in 2020 (n = 448) Negative for anti-SARS-CoV-2 IgG

A c c e p t e d M a n u s c r i p t