Large-Scale Screening of Asymptomatic Persons for SARS-CoV-2 Variants of Concern and Gamma Takeover, Brazil

We performed a large-scale severe acute respiratory syndrome coronavirus 2 screening campaign using 2 PCR-based approaches, coupled with variant genotyping, aiming to provide a safer environment for employees of Federal University in Curitiba, Brazil. We observed the rapid spread of the Gamma variant of concern, which replaced other variants in <3 months.


The Study
During October 10, 2020-May 24, 2021, asymptomatic and mildly symptomatic persons within the community of the Federal University of Paraná were called for voluntary participation through social media and email. Eligible participants were members of the academic community (students, technicians, professors, or outsourced employees) or their relatives (grandparents, parents, siblings, or children) or household members. Saliva samples were selfcollected by using an individually wrapped plastic drinking straw, transferred to a prelabeled 2.0 mL microtube, and stored at 4°C. Samples were transported to the laboratory in <1 h; total turnaround time to results was <48 h.
Samples were homogenized and allowed to settle for 30 min or centrifuged for 2 min (2,000 × g). A quantity of 200 µL from each specimen was

Large-Scale Screening of Asymptomatic Persons for SARS-CoV-2 Variants of Concern and Gamma Takeover, Brazil
We performed a large-scale severe acute respiratory syndrome coronavirus 2 screening campaign using 2 PCR-based approaches, coupled with variant genotyping, aiming to provide a safer environment for employees of Federal University in Curitiba, Brazil. We observed the rapid spread of the Gamma variant of concern, which replaced other variants in <3 months.
then pooled (8) (2). If the pool rendered a positive result, we reprocessed samples individually. We further evaluated positive samples by using 2 probe-based genotyping systems to detect VOCs. The first one was the Vogels et al. (7) multiplex approach to detect Spike Δ69-70 and Orf1a Δ3675-3677 deletions as an outcome for distinguishing Alpha, Beta, or Gamma and wild-type or other lineages (7). For this approach, we also included the Centers for Disease Control and Preventon N1 target and defined a cycle threshold (C t ) of <28 on this particular target to evaluate the gene dropouts.
A total of 16 collection dates were recorded; 12,558 examinations were processed (Table) from the 7,249 persons who attended because some participants engaged in >1 day of collection. The number of attendees per collection date ranged from 162 to 1,737. The overall prevalence rate was 1.28% (161/12,558). Comparing these numbers to cases in the state of Paraná by the epidemiologic week of diagnosis (Figure 1, panel A), we found prevalence similar to the prevalence rate at the beginning of the state's second wave of SARS-CoV-2 infections (Figure 1, panel A).
We also evaluated all SARS-CoV-2-positive cases by using multiplex and singleplex genotyping approaches (Figure 2). From all 161 positive cases evaluated, the Vogels et al. (7) multiplex assay was invalidated in 46 (28.6%) against 50 (31.1%) in a Thermo Fisher 3-assay allelic detection approach because of the high C t values. Comparing the original C t value of detection, performance depreciated in samples with C t >30 ( Figure 2, panel B), as stated in the Thermo Fisher manual. Nevertheless, all genotyped cases were concordant between the 2 assays, considering that the Vogels et al. (7) assay alone does not discriminate between wild-type and B1.1.28/P.2 VOI.
Detection of the Gamma variant occurred on January 21, 2021, <2 weeks after the collapse of the healthcare system in Manaus. Prevalence of the Gamma variant was 9.1% on this date and increased to 42.9% 2 weeks later (Figure 1, panel B). A possible explanation for this scenario is the increased transmission of this variant in >20% compared with the wild-type transmission rate (10), which is supported by the observation of a reproduction number of 1.5 in Paraná (http://shiny.leg.ufpr.br/elias/co-vid19time) in the weeks before the death peak, the highest reproduction number observed during the pandemic. This increase in cases could be correlated with the subsequent collapse of the Curitiba city healthcare system and a surge of coronavirus disease deaths in the Paraná state, reaching values >5% in the subsequent weeks. A similar scenario was also observed in Manaus, the origin of the Gamma variant; both SARS-CoV-2 cases and excess of burials in the city reached their highest levels during the pandemic to that point (4). Two factors could explain those observed surges: increased lethality of the Gamma variant-which is not yet defined (10)-and the actual collapse of healthcare systems, leading to poorer patient support. When testing activities resumed after the healthcare collapse, all cases became Gamma variant, completely displacing B1.1.28/P.2 VOI and wild-type cases in 3 months.

Conclusions
Analysis of saliva in pools as described in this study offers an inexpensive and easy-to-implement asymptomatic screening strategy. Thus, given the high rates of SARS-CoV-2 transmission, the risk for asymptomatic coronavirus disease spread and the importance of social distancing should continue to be stressed to the public until the vaccine is viable for large-scale application. Our mass testing pro-gram was intended to be accessible (every test was free-of-charge for the participant), reliable (all participants received their results and positive persons had a follow-up opportunity), and aimed to reach all social strata within the academic community (from professors to outsourced employees, which consisted mainly of socially, economically, and ethnically vulnerable groups), which are key characteristics of a strong mass testing system (11). We found that both multiplex PCR and singleplex PCR approaches were valuable tools to evaluate the proportion of variants within genomic surveillance and were faster and less expensive than whole-genome sequencing approaches. Although those methods do not serve as substitutes for whole-genome sequencing, they could be an essential method to screen and select samples for further variant classification. Nevertheless, those approaches could demonstrate rapid spread of new variants and predict surges of SARS-CoV-2 infections, acting as a lighthouse for farreaching public health decisions.