Spatiotemporal transmission of SARS-CoV-2 lineages during 2020-2021 in Pernambuco - Brazil (preprint)

In recent years, the SARS-CoV-2 viruses emerged and spread around the world, leaving a large death toll and long-lasting impact on survivors. As of January 2023, Brazil is still among the countries with the highest number of registered deaths. This continental-size and pluralistic country experienced a heterogenous implementation of non-pharmacological and pharmacological interventions which, associated with large socioeconomic differences between the country regions, has led to distinct virus spread dynamics across the country. Here we investigate the spatiotemporal dispersion of emerging SARS-CoV-2 lineages and its dynamics in distinct epidemiological scenarios in the first two years of the pandemics in the Pernambuco state (Northeast Brazil). We generated a total of 1389 new SARS-CoV-2 genomes from June 2020 to August 2021 covering all major regions of the state. This sampling captured the arrival, communitary transmission and the circulation of the B1.1, B.1.1.28 and B.1.1.33 lineages in the first eight months of the pandemics, the emergence of the former variant of interest P.2 and the emergence and fast replacement of all previous variants by the more transmissible variant of concern P.1 (Gamma) lineage. Based on the incidence and lineage spread pattern we observed that there was an East-to-West to inner state pattern of transmission which is in agreement with the transmission of more populous metropolitan areas to medium and small size country-side cities in the state. Such transmission patterns may be partially explained by the main routes of traffic across municipalities in the state. Nevertheless, inter-state traffic was also another important source of lineage introduction and spread into the state. Our results highlight that the fine grained intrastate analysis of lineages and incidence spread can provide actionable insights for planning future non-pharmacological intervention for air-borne transmissible human pathogens.

Antigen concentration, viral load, and test performance for SARS-CoV-2 in multiple specimen types (preprint)

The relationship between N-antigen concentration and viral load within a specimen and across different specimens is essential for interpretation of rapid diagnostic tests (RDT) clinical performance in different use cases. A prospective study was conducted in Porto Velho, Brazil, to investigate RDT performance in different specimen types as a function of the correlation between antigen concentration and viral load. The study included 214 close contacts with recent exposures to confirmed cases, aged 12 years and older and with various levels of vaccination. Antigen concentration was measured in nasopharyngeal swab (NPS), anterior nares swab (ANS), and saliva specimens. Reverse transcriptase (RT)PCR was conducted on the NPS and saliva specimens, and two RDTs were conducted on ANS and one on saliva. Antigen concentration correlated with viral load when measured in the same specimen type but not across specimen types. Antigen levels were higher in symptomatic cases compared to asymptomatic/oligosymptomatic cases and lower in saliva compared to NPS and ANS samples. Discordant results between the RDTs conducted on ANS and the RT-PCR on NPS were resolved by antigen concentration values. The analytical limit-of-detection of RDTs can be used to predict the performance of the tests in populations for which the antigen concentration is known. The antigen dynamics across different sample types observed in SARS-CoV-2 disease progression support use of RDTs in nasal samples. Given lower antigen concentrations in saliva, tests using saliva is expected to require improved analytical sensitivity to achieve clinical sensitivity similar to testing of nasal samples.

The SARS-CoV-2 Omicron Variant of Concern and its Rapid Spread throughout the Western Brazilian Amazon (preprint)

Genomic surveillance represents an important strategy for understanding evolutionary mechanisms, transmission profile, and infectivity of different SARS-CoV-2 variants. We assessed the epidemiological profile of 366 individuals who tested positive for SARS-CoV-2 from 29 municipalities in Rondônia between December 2021 to March 2022. Samples were collected, RNA was ex-tracted and screened using RT-qPCR for Alpha, Beta, Gamma, Delta and Omicron VOCs and viral quantification was performed. Sequences were analyzed for phylogeny, mutations and lineages. Of the samples analyzed, 93.71% were positive for the Omicron variant and 6.28% were positive for the Delta variant. The symptoms observed were cough, sore throat, and fever, with a mean duration of 5 days; no hospitalizations or deaths were reported. We noted that among the positive individuals, 51% had been immunized with two doses, 22% received three doses, 13% received one dose, and 13% were not immunized. Just 242 samples were amenable to analysis for alignment and phylogenetic characterization; corresponding to variants BA.1 and BA.1.1; a total of 120 mutations were identified, 36% of which were found in the S gene. In conclusion, there was a high frequency of mutations in the SARS-CoV-2 genome, but no record of clinical severity, demonstrating the positive effect of vaccination.

Screening for SARS-CoV-2 in close contacts of individuals with confirmed infection: performance and operational considerations (preprint)

Background Point-of-care and decentralized testing for SARS-CoV-2 is critical to inform public health responses. Performance evaluations in priority use cases such as contact tracing can highlight trade-offs in test selection and testing strategies. Methods A prospective diagnostic accuracy study was conducted among close contacts of COVID-19 cases in Brazil. Two anterior nares swabs (ANS), a nasopharyngeal swab (NPS), and saliva were collected at all visits. Vaccination history and symptoms were assessed. Household contacts were followed longitudinally. Three rapid antigen tests and one molecular method were evaluated for usability and performance against reference RT-PCR on NPS. Results Fifty index cases and 214 contacts (64 household) were enrolled. Sixty-five contacts were RT-PCR positive during at least one visit. Vaccination did not influence viral load. Gamma variants were most prevalent; Delta emerged increasingly during implementation. Overall sensitivity of evaluated tests ranged from 33%–76%. Performance was higher among symptomatic cases and cases with Ct<34 and lower among oligo/asymptomatic cases. Assuming a 24-hour time-to-result for RT-PCR, the cumulative sensitivity of an ANS rapid antigen test was >70% and almost 90% after four days. Conclusions The near immediate time-to-result for antigen tests significantly offsets lower analytical sensitivity in settings where RT-PCR results are delayed or unavailable.

A case series of SARS-CoV-2 reinfections caused by the variant of concern Gamma in Brazil (preprint)

The rapid spread of the SARS-CoV-2 Variant of Concern (VOC) Gamma during late 2020 and early 2021 in Brazilian settings with high seroprevalence raised some concern about the potential role of reinfections in driving the epidemic. Very few cases of reinfection associated with the VOC Gamma, however, have been reported. Here we describe 25 cases of SARS-CoV-2 reinfection confirmed by real-time RT-PCR twice within months apart in Brazil. SARS-CoV-2 genomic analysis confirmed that individuals were primo-infected between March and December 2020 with distinct viral lineages, including B.1.1, B.1.1.28, B.1.1.33, B.1.195 and P.2, and then reinfected with the VOC Gamma between 3 to 12 months after primo-infection. The overall mean cycle threshold (Ct) value of the first (25.7) and second (24.5) episodes were roughly similar for the whole group and 14 individuals displayed mean Ct values < 25.0 at reinfection. Sera of 14 patients tested by plaque reduction neutralization test after reinfection displayed detectable neutralizing antibodies against Gamma and other SARS-CoV-2 variants (B.1.33, B.1.1.28 and Delta). All individuals have milder or no symptoms after reinfection and none required hospitalization. The present study demonstrates that the VOC Gamma was associated with reinfections during the second Brazilian epidemic wave in 2021 and raised concern about the potential infectiousness of reinfected subjects. Although individuals here analyzed failed to mount a long-term sterilizing immunity, they developed a high anti-Gamma neutralizing antibody response after reinfection that may provide some protection against severe disease.

Emergence and Spread of the SARS-CoV-2 Variant of Concern Delta Across Different Brazilian Regions (preprint)

The SARS-CoV-2 Variant of Concern (VOC) Delta was first detected in India in October 2020. The first imported cases of the Delta variant in Brazil were identified in April 2021 in the Southern region, followed by more cases in different country regions during the following months. By early September 2021, Delta was already the dominant variant in the Southeastern (87%), Southern (73%), and Northeastern (52%) Brazilian regions. This work aimed to understand the spatiotemporal dissemination dynamics of Delta in Brazil. To this end, we employed a combination of Maximum Likelihood (ML) and Bayesian methods to reconstruct the evolutionary relationship of 2,264 of VOC Delta complete genomes (482 from this study) recovered across 21 out of 27 Brazilian federal units. Our phylogeographic analyses identified three major transmission clusters of Delta in Brazil. The clade BR-I (n = 1,560) arose in Rio de Janeiro in late April 2021 and was the major cluster behind the dissemination of the VOC Delta in the Southeastern, Northeastern, Northern, and Central-Western regions. The clade BR-II (n = 207) arose in the Parana state in late April 2021 and aggregated the largest fraction of sampled genomes from the Southern region. Lastly, the clade BR-III emerged in the Sao Paulo state in early June 2021 and remained mostly restricted to this state. In the rapid turnover of viral variants characteristic of the SARS-CoV-2 pandemic, Brazilian regions seem to occupy different stages of an increasing prevalence of the VOC Delta in their epidemic profiles. This process demands continuous genomic and epidemiological surveillance toward identifying and mitigating new introductions, limiting their dissemination, and preventing the establishment of more significant outbreaks in a population already heavily affected by the COVID-19 pandemic.

Phylogenetic-based inference reveals distinct transmission dynamics of SARS-CoV-2 variant of concern Gamma and lineage P.2 in Brazil (preprint)

The COVID-19 epidemic in Brazil experienced two major country-wide lineage replacements, the first driven by the lineage P.2, formerly classified as variant of interest (VOI) Zeta in late 2020 and the second by the variant of concern (VOC) Gamma in early 2021. To better understand how these SARS-CoV-2 lineage turnovers occurred in Brazil, we analyzed 11,724 high-quality SARS-CoV-2 whole genomes of samples collected in different country regions between September 2020 and April 2021. Our findings indicate that the spatial dispersion of both variants in Brazil was driven by short and long-distance viral transmission. The lineage P.2 harboring Spike mutation E484K probably emerged around late July 2020 in the Rio de Janeiro (RJ) state, which contributed with most (~50%) inter-state viral disseminations, and only became locally established in most Brazilian states by October 2020. The VOC Gamma probably arose in November 2020 in the Amazonas (AM) state, which was responsible for 60-70% of the inter-state viral dissemination, and the earliest timing of community transmission of this VOC in many Brazilian states was already traced to December 2020. We estimate that variant Gamma was 1.56-3.06 more transmissible than variant P.2 co-circulating in RJ and that the median effective reproductive number (Re) of Gamma in RJ and SP states (Re = 1.59-1.91) was lower than in AM (Re = 3.55). In summary, although the epicenter of the lineage P.2 dissemination in Brazil was the heavily interconnected Southeastern region, it displayed a slower rate of spatial spread than the VOC Gamma originated in the more isolated Northern Brazilian region. Our findings also support that the VOC Gamma was more transmissible than lineage P.2, although the viral Re of the VOC varied according to the geographic context.

Unusual SARS-CoV-2 intra-host diversity reveals lineages superinfection (preprint)

The SARS-CoV-2 has infected almost 200 million people worldwide by July 2021 and the pandemic has been characterized by infection waves of viral lineages showing distinct fitness profiles. The simultaneous infection of a single individual by two distinct SARS-CoV-2 lineages provides a window of opportunity for viral recombination and the emergence of new lineages with differential phenotype. Several hundred SARS-CoV-2 lineages are currently well characterized but two main factors have precluded major coinfection/codetection analysis thus far: i) the low diversity of SARS-CoV-2 lineages during the first year of the pandemic which limited the identification of lineage defining mutations necessary to distinguish coinfecting viral lineages; and the ii) limited availability of raw sequencing data where abundance and distribution of intrasample/intrahost variability can be accessed. Here, we have put together a large sequencing dataset from Brazilian samples covering a period of 18 May 2020 to 30 April 2021 and probed it for unexpected patterns of high intrasample/intrahost variability. It enabled us to detect nine cases of SARS-CoV-2 coinfection with well characterized lineage-defining mutations. In addition, we matched these SARS-CoV-2 coinfections with spatio-temporal epidemiological data confirming their plausibility with the co-circulating lineages at the timeframe investigated. These coinfections represent around 0.61% of all samples investigated. Although our data suggests that coinfection with distinct SARS-CoV-2 lineages is a rare phenomenon, it is likely an underestimation and coinfection rates warrants further investigation. DATA SUMMARYThe raw fastq data of codetection cases are deposited on gisaid.org and correlated to gisaid codes: EPI_ISL_1068258, EPI_ISL_2491769, EPI_ISL_2491781, EPI_ISL_2645599, EPI_ISL_2661789, EPI_ISL_2661931, EPI_ISL_2677092, EPI_ISL_2777552, EPI_ISL_3869215. Supplementary data are available on https://doi.org/10.6084/m9.figshare.16570602.v1. The workflow code used in this study is publicly available on: https://github.com/dezordi/IAM_SARSCOV2.

Spread of Gamma (P.1) sub-lineages carrying Spike mutations close to the furin cleavage site and deletions in the N-terminal domain drives ongoing transmission of SARS-CoV-2 in Amazonas, Brazil (preprint)

Summary The Amazonas was one of the most heavily affected Brazilian states by the COVID-19 epidemic. Despite a large number of infected people, particularly during the second wave associated with the spread of the Variant of Concern (VOC) Gamma (lineage P.1), SARS-CoV-2 continues to circulate in the Amazonas. To understand how SARS-CoV-2 persisted in a human population with a high immunity barrier, we generated 1,188 SARS-CoV-2 whole-genome sequences from individuals diagnosed in the Amazonas state from 1st January to 6th July 2021, of which 38 were vaccine breakthrough infections. Our study reveals a sharp increase in the relative prevalence of Gamma plus (P.1+) variants, designated as Pango Lineages P.1.3 to P.1.6, harboring two types of additional Spike changes: deletions in the N-terminal (NTD) domain (particularly Δ 144 or Δ 141-144) associated with resistance to anti-NTD neutralizing antibodies or mutations at the S1/S2 junction (N679K or P681H) that probably enhance the binding affinity to the furin cleavage site, as suggested by our molecular dynamics simulations. As lineages P.1.4 (S:N679K) and P.1.6 (S:P681H) expanded (Re > 1) from March to July 2021, the lineage P.1 declined (Re < 1) and the median Ct value of SARS-CoV-2 positive cases in Amazonas significantly decreases. Still, we found no overrepresentation of P.1+ variants among breakthrough cases of fully vaccinated patients (71%) in comparison to unvaccinated individuals (93%). This evidence supports that the ongoing endemic transmission of SARS-CoV-2 in the Amazonas is driven by the spread of new local Gamma/P.1 sub-lineages that are more transmissible, although not more efficient to evade vaccine-elicited immunity than the parental VOC. Finally, as SARS-CoV-2 continues to spread in human populations with a declining density of susceptible hosts, the risk of selecting new variants with higher infectivity are expected to increase.

Attack rate of COVID-19 and the onset of herd immunity in Manaus, Brazil (preprint)

Buss et al. (2021) and Faria et al. (2021) reported that ~76% of the residents of the capital city of Manaus, had been infected by SARS-CoV-2 by October 2020 suggesting that herd immunity had been achieved by the end of the first wave. But the announcement of herd immunity, which would imply reasonable protection from future outbreaks, only provided the Manaus population with a false sense of security. Within two months later, a second wave of COVID-19 was initiated with death rates much larger than the first attributed to the appearance of the new P.1 Variant of Concern. Faria et al. (2021) suggest that large scale reinfections played an important role in enabling the huge second epidemic wave. In this Technical Comment we challenge such interpretations, and provide quantitative arguments that suggest the attack rate of the first wave was well below 76%. We then present alternative interpretations of the data.

Identification of SARS-CoV-2 P.1-related lineages in Brazil provides new insights about the mechanisms of emergence of Variants of Concern (preprint)

One of the most remarkable features of the SARS-CoV-2 Variants of Concern (VOC) is the unusually large number of mutations they carry. However, the specific factors that drove the emergence of such variants since the second half of 2020 are not fully resolved. In this study, we described a new SARS-CoV-2 lineage provisionally designated as P.1-like-II that, as well as the previously described lineage P.1-like-I, shares several lineage-defining mutations with the VOC P.1 circulating in Brazil. Reconstructions of P.1 ancestor sequences demonstrate that the entire constellation of mutations that define the VOC P.1 did not accumulate within a single long-term infected individual, but was acquired by sequential addition during interhost transmissions. Our evolutionary analyses further estimate that P.1-ancestors strains carrying half of the P.1-lineage-defining mutations, including those at the receptor-binding domain of the Spike protein, circulated cryptically in the Amazonas state since August 2020. This evolutionary pattern is consistent with the hypothesis that partial human population immunity acquired from natural SARS-CoV-2 infections during the first half of 2020 might have been the major driving force behind natural selection that allowed VOCs' emergence and worldwide spread. These findings also support a long lag-time between the emergence of variants with key mutations of concern and expansion of the VOC P.1 in Brazil.

Effectiveness of CoronaVac among healthcare workers in the setting of high SARS-CoV-2 Gamma variant transmission in Manaus, Brazil: A test-negative case-control study (preprint)

Background Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variant, Gamma, emerged in the city of Manaus in late 2020 during a large resurgence of coronavirus disease (COVID-19), and has spread throughout Brazil. The effectiveness of vaccines in settings with widespread Gamma variant transmission has not been reported. Methods We performed a matched test-negative case-control study to estimate the effectiveness of an inactivated vaccine, CoronaVac, in healthcare workers (HCWs) in Manaus, where the Gamma variant accounted for 86% of genotyped SARS-CoV-2 samples at the peak of its epidemic. We performed an early analysis of effectiveness following administration of at least one vaccine dose and an analysis of effectiveness of the two-dose schedule. The primary outcome was symptomatic SARS-CoV-2 infection. Findings For the early at-least-one-dose and two-dose analyses the study population was, respectively, 53,176 and 53,153 HCWs residing in Manaus and aged 18 years or older, with complete information on age, residence, and vaccination status. Among 53,153 HCWs eligible for the two-dose analysis, 47,170 (89%) received at least one dose of CoronaVac and 2,656 individuals (5%) underwent RT-PCR testing from 19 January, 2021 to 13 April, 2021. Of 3,195 RT-PCR tests, 885 (28%) were positive. 393 and 418 case- control pairs were selected for the early and two-dose analyses, respectively, matched on calendar time, age, and neighbourhood. Among those who had received both vaccine doses before the RT-PCR sample collection date, the average time from second dose to sample collection date was 14 days (IQR 7-24). In the early analysis, vaccination with at least one dose was associated with a 0.50-fold reduction (adjusted vaccine effectiveness (VE), 49.6%, 95% CI 11.3 to 71.4) in the odds of symptomatic SARS-CoV-2 infection during the period 14 days or more after receiving the first dose. However, we estimated low effectiveness (adjusted VE 36.8%, 95% CI -54.9 to 74.2) of the two-dose schedule against symptomatic SARS-CoV-2 infection during the period 14 days or more after receiving the second dose. A finding that vaccinated individuals were much more likely to be infected than unvaccinated individuals in the period 0-13 days after first dose (aOR 2.11, 95% CI 1.36-3.27) suggests that unmeasured confounding led to downward bias in the vaccine effectiveness estimate. Interpretation Evidence from this test-negative study of the effectiveness of CoronaVac was mixed, and likely affected by bias in this setting. Administration of at least one vaccine dose showed effectiveness against symptomatic SARS-CoV-2 infection in the setting of epidemic Gamma variant transmission. However, the low estimated effectiveness of the two-dose schedule underscores the need to maintain non-pharmaceutical interventions while vaccination campaigns with CoronaVac are being implemented. Funding Fundação Oswaldo Cruz (Fiocruz); Municipal Health Secretary of Manaus Research in Context Evidence before this study We searched PubMed for articles published from inception of the pandemic until April 3, 2021, with no language restrictions, using the search terms “P.1” AND “vaccine” AND “SARS-CoV-2”. Additionally, we searched for “CoronaVac” AND “SARS-CoV-2”. Early studies have found plasma from convalescent COVID-19 patients and sera from vaccinated individuals have reduced neutralisation of the SARS-CoV-2 variant, Gamma or P.1, compared with strains isolated earlier in the pandemic. Pfizer BNT162b2 mRNA, Oxford-AstraZeneca ChAdOx1, and CoronaVac are the only vaccines for which such data has been published to date. No studies reported effectiveness of any vaccine on reducing the risk of infection or disease among individuals exposed to P.1 or in settings of high P.1 transmission. Added value of this study This study finds that vaccination with CoronaVac was 49.4% (95% CI 13.2 to 71.9) effective at preventing COVID-19 in a setting with likely high prevalence of the Gamma Variant of Concern. However, an analysis of effectiveness by dose was underpowered and failed to find significant effectiveness of the two-dose schedule of CoronaVac (estimated VE 37.1%, 95% CI -53.3 to 74.2). Implications of all the available evidence These findings are suggestive for the effectiveness of CoronaVac in healthcare workers in the setting of widespread P.1 transmission but must be strengthened by observational studies in other settings and populations. Based on this evidence, there is a need to implement sustained non-pharmaceutical interventions even as vaccination campaigns continue.

The ongoing evolution of variants of concern and interest of SARS-CoV-2 in Brazil revealed by convergent indels in the amino (N)-terminal domain of the Spike protein (preprint)

Mutations at both the receptor-binding domain (RBD) and the amino (N)-terminal domain (NTD) of the SARS-CoV-2 Spike (S) glycoprotein can alter its antigenicity and promote immune escape. We identified that SARS-CoV-2 lineages circulating in Brazil with mutations of concern in the RBD independently acquired convergent deletions and insertions in the NTD of the S protein, which altered the NTD antigenic-supersite and other predicted epitopes at this region. These findings support that the ongoing widespread transmission of SARS-CoV-2 in Brazil is generating new viral lineages that might be more resistant to neutralization than parental variants of concern.

Antibody evasion by the Brazilian P.1 strain of SARS-CoV-2 (preprint)

Terminating the SARS-CoV-2 pandemic relies upon pan-global vaccination. Current vaccines elicit neutralizing antibody responses to the virus spike derived from early isolates. However, new strains have emerged with multiple mutations: P.1 from Brazil, B.1.351 from South Africa and B.1.1.7 from the UK (12, 10 and 9 changes in the spike respectively). All have mutations in the ACE2 binding site with P.1 and B.1.351 having a virtually identical triplet: E484K, K417N/T and N501Y, which we show confer similar increased affinity for ACE2. We show that, surprisingly, P.1 is significantly less resistant to naturally acquired or vaccine induced antibody responses than B.1.351 suggesting that changes outside the RBD impact neutralisation. Monoclonal antibody 222 neutralises all three variants despite interacting with two of the ACE2 binding site mutations, we explain this through structural analysis and use the 222 light chain to largely restore neutralization potency to a major class of public antibodies.

SARS-COV-2 genomic surveillance in the Brazilian Western Amazon region: evolutionary history of epidemic dissemination and genetic signature in Rondônia, Brazil (preprint)

SARS-CoV-2 has spread rapidly around the world, with Brazil currently considered an epicenter of the pandemic. The northern region of the country has the highest incidence and mortality rates. This study aimed to investigate information about the evolutionary history of epidemic spread and genetic aspects of strains isolated on the Western Amazon, in the State of Rondônia, Brazil. It was possible to detect a total of 22 mutations. Some of these alterations may possibly be related to effects on transmissibility, the fidelity of RNA replication, the ability of cancer patients to respond to infection, beyond a mutation that emerged after the introduction of SARS-CoV-2 in Rondônia. At least two events of introduction were detected, corresponding to the B.1 and B.1.1 European lineages. An introduction was observed possibly through Argentina, where strains originated that circulated in the Minas Gerais and Ceará Brazilian states, prior to Rondônia (B.1.), as well as through the Minas Gerais state and the Federal District, which gave rise to strains that spread to Rondônia, from the capital to more rural parts of the state (B.1.1.). The findings show the need to monitor the genetic epidemiology of COVID-19, in order to surveil the virus’s evolution, dispersion and diversity.

Chloroquine diphosphate in two different dosages as adjunctive therapy of hospitalized patients with severe respiratory syndrome in the context of coronavirus (SARS-CoV-2) infection: Preliminary safety results of a randomized, double-blinded, phase IIb clinical trial (CloroCovid-19 Study) (preprint)

Background There is no specific antiviral therapy recommended for the disease caused by SARS-CoV-2 (COVID-19). Recent publications have drawn attention to the possible benefit of chloroquine (CQ). Our study aimed to comprehensively evaluate the safety and efficacy of two different CQ dosages in patients with established severe COVID-19. Methods We performed a parallel, double-blinded, randomized, phase IIb clinical trial, aiming to assess safety and efficacy of two different CQ dosages as adjunctive therapy of hospitalized patients with SARS in Manaus, Brazilian Amazon. Eligible participants were allocated to receive orally or via nasogastric tube high dose CQ (600mg CQ twice daily for 10 days or total dose 12g); or low dose CQ (450mg for 5 days, twice daily only on the first day, or total dose 2.7g). In addition, all patients received ceftriaxone and azithromycin. This study was registered with ClinicalTrials.gov, number NCT04323527. Findings Out of a pre-defined 440 patients sample size, 81 patients were enrolled. The high dose CQ arm presented more QTc>500ms (25%), and a trend toward higher lethality (17%) than the lower dosage. Fatality rate was 13.5% (95%CI=6.9-23.0%), overlapping with the CI of historical data from similar patients not using CQ (95%CI=14.5-19.2%). In 14 patients with paired samples, respiratory secretion at day 4 was negative in only one patient. Interpretation Preliminary findings suggest that the higher CQ dosage (10-day regimen) should not be recommended for COVID-19 treatment because of its potential safety hazards. Such results forced us to prematurely halt patient recruitment to this arm. Given the enormous global push for the use of CQ for COVID-19, results such as the ones found in this trial can provide robust evidence for updated COVID-19 patient management recommendations.