key: cord-1048250-7cgzxtcq
authors: Wunsch, M.; Aschemeier, D.; Heger, E.; Ehrentraut, D.; Krueger, J.; Hufbauer, M.; Syed, A. S.; Horemheb-Rubio, G.; Dewald, F.; Fish, I.; Schlotz, M.; Gruell, H.; Augustin, M.; Lehmann, C.; Kaiser, R.; Knops, E.; Silling, S.; Klein, F.
title: Safe and effective pool testing for SARS-CoV-2 detection
date: 2021-04-08
journal: nan
DOI: 10.1101/2021.04.08.20205781
sha: ea729d96ad6898a7b0f2a6dd1cc58456126ba76c
doc_id: 1048250
cord_uid: 7cgzxtcq

Background / Objectives: The global spread of SARS-CoV-2 is a serious public health issue. Large-scale surveillance screenings are crucial but can exceed diagnostic test capacities. We set out to optimize test conditions and implemented high throughput pool testing of respiratory swabs into SARS-CoV-2 diagnostics. Study design: In preparation for pool testing, we determined the optimal pooling strategy and pool size. In addition, we measured the impact of vortexing prior to sample processing, compared pipette- and swab-pooling method as well as the sensitivity of three different PCR assays. Results: Using optimized strategies for pooling, we systematically pooled 55,690 samples in a period of 44 weeks resulting in a reduction of 47,369 PCR reactions. In a low prevalence setting, we defined a preferable pool size of ten in a two-stage hierarchical pool testing strategy. Vortexing of the swabs increased cellular yield by a factor of 2.34, and sampling at or shortly after symptom onset was associated with higher viral loads. By comparing different pooling strategies, pipette-pooling was more efficient compared to swab-pooling. Conclusions: For implementing pooling strategies into high throughput diagnostics, we recommend to apply a pipette-pooling method, using pool sizes of ten samples, performing sensitivity validation of the PCR assays used, and vortexing swabs prior to analyses. Our data shows, that pool testing for SARS-CoV-2 detection is feasible and highly effective in a low prevalence setting.

The SARS-CoV-2 pandemic is a serious public health problem of unprecedented 44 magnitude in recent times. In particular individuals at older ages or with comorbidities 45 are at a high risk to develop an acute respiratory distress syndrome (ARDS) requiring 46 hospitalization and intensive care [1] . Therefore, it is essential to control person-toperson transmission in order to protect vulnerable individuals and limit the number of 48 severe cases. Until herd immunity is achieved by vaccination, nonpharmaceutical 49 interventions need to be applied. Many countries could successfully contain the 50 spread of COVID-19 through social distancing or lock-down measures, contact 51 tracing, quarantine, and large-scale testing in the ongoing pandemic [2, 3] . In order to 52 control viral transmission when lifting lock-down strategies, large-scale testing and 53 surveillance are critical interventions. These approaches are based on frequent tests 54 of individuals e.g. by rapid antigen-based tests or reverse transcription-real-time PCR 55 (rRT-PCR) to detect SARS-CoV-2 in swab specimens. However, large-scale 56 surveillance screenings can exceed the test capacities of diagnostic laboratories. 57

Pooling swab specimens for PCR testing can increase test capacities and limit 58 the consumption of reagents [4] . Pool testing is highly efficient in a setting of low 59 disease prevalence and the availability of highly sensitive test methods [5] . It can be 60 applied to enable surveillance screenings of asymptomatic individuals in public 61 institutions e.g. hospitals, schools or retirement homes, which carry a high risk for 62 superspreading events and severe disease courses. When pool testing is 63 established, test conditions need to be optimized including (a) the type of pooling 64 strategy and pool sizes, (b) sample preparation and pooling method, (c) the quality of 65 SARS-CoV-2 detection by PCR. In this study, our aim was to determine and 66 . CC-BY-NC-ND 4.0 International license It is made available under a 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 April 8, 2021. ; https://doi.org/10.1101/2021.04.08.20205781 doi: medRxiv preprint the same specimens before and after vortexing, human β-globin-gene quantification 97 was performed as published [7] . 98

For the pipette-pooling method and to simulate various pool sizes, positive 99 specimens with various Ct-values were combined with increasing volumes of 100 negative samples combined as a stock, accordingly. For the swab-pooling method, 101 nine SARS-CoV-2 negative and one positive swab were used. After removal of the 102 transport medium, 1.2 ml PBS was added to the tube containing the swab and 103 vortexed 5 seconds. The PBS was then transferred to the next swab tube and 104 vortexed. Following this principle, ten swabs were merged. Preparation time was 105 measured. Detection of SARS-CoV-2 was performed on a Roche cobas® 6800 using the 120 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 8, 2021. ; https://doi.org/10.1101/2021.04.08.20205781 doi: medRxiv preprint

Pool testing can be performed using different strategies. In this study we conducted 139 two-stage hierarchical pooling procedures ( Figure 1A ). Swab samples were 140 combined and tested in a single PCR reaction. If the pool test was positive, the 141 remaining sampling material of the included specimens was retested separately to 142 detect the infected individual. If the pool test was negative, all individuals were 143 declared as not infected [5, 9, 10] . . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 8, 2021. ; https://doi.org/10.1101/2021.04.08.20205781 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 8, 2021. ; https://doi.org/10.1101/2021.04.08.20205781 doi: medRxiv preprint CC-BY-NC-ND 4.0 International license It is made available under a 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 April 8, 2021. applying both methods, respectively. The mean processing time was 3 minutes, 47 187 seconds for a swab-based pool (95% CI: (2 min,59sec.)-(4min,36sec.)) and 1 188 minute, 55 seconds for a pipette-based pool (95% CI: (1min,33sec.)-(2min,16sec.)). 189

In order to investigate the sensitivity, we generated 16 different pools with CC-BY-NC-ND 4.0 International license It is made available under a 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 April 8, 2021. Figure 2C) . 213

The lowest detectable copy number was 200 copies for instrument I, 2,000 for 214 instrument II, and 20 copies for instrument III as determined using two INSTAND 215 standards ( Figure 2F ). 216

To determine the detection-rate for different pool sizes, 25 positive samples 217

with Ct-values ranging from 18.96 to 34.99 were each diluted in a stock of negative 218 specimens and tested in duplicates on instrument III. All pools were SARS-CoV-2-219 positive, however, for two pools the 1:20 and 1:50 dilution resulted in one negative 220 and one positive replicate, respectively ( Figure 2G ). 221 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Mann-Whitney test was performed. Sample preparation time was measured for four 225 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (n=2,640 samples) had to be retested individually ( Figure 3B ). As Figure 3C (Figures 3 F and G) . 264 . CC-BY-NC-ND 4.0 International license It is made available under a 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 April 8, 2021. ; https://doi.org/10.1101/2021.04.08.20205781 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a 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 April 8, 2021. ; https://doi.org/10.1101/2021.04.08.20205781 doi: medRxiv preprint single round of testing. In our study, we used the hierarchical testing strategy due to 299 logistics of a high-throughput diagnostic setting. 300

Pre-analytical handling can substantially impact test sensitivity, however, 301 limited data on this topic are available. Test results are influenced by improper 302 . CC-BY-NC-ND 4.0 International license It is made available under a 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 April 8, 2021. We developed a feasible pooling procedure that can readily be implemented in 316 diagnostic routines. The preparation for pool testing contained besides extensive 317 technical investigations, also changes in the laboratory logistics and adaptions of the 318 laboratory software. The data communicated here will contribute to the process of 319 finding and implementing a consensus pool testing strategy enabling larger test 320 capacities to effectively combat the SARS-CoV-2 pandemic. 321 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

. CC-BY-NC-ND 4.0 International license It is made available under a 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 April 8, 2021. 

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 8, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 8, 2021. ; https://doi.org/10.1101/2021.04.08.20205781 doi: medRxiv preprint

Clinical course and risk factors for mortality

To Interpret the SARS-CoV-2 Test

Ct values and infectivity of SARS-CoV-2 on 376 surfaces. The Lancet Infectious Diseases

What reinfections mean for COVID-19

SARS-CoV-2 380 detection, viral load and infectivity over the course of an infection

Pooling of 382 samples for testing for SARS-CoV-2 in asymptomatic people

Sample pooling as a strategy to detect community 384 transmission of SARS-CoV-2