key: cord-0916988-cqze6607
authors: Graham, Maryza; Williams, Eloise; Isles, Nicole; Buadromo, Eka; Toatu, Tebuka; Druce, Julian; Catton, Mike; Lin, Chantel; Howden, Benjamin P.; Williamson, Deborah A.
title: Sample pooling on the Cepheid Xpert® Xpress SARS-CoV-2 assay
date: 2020-10-15
journal: Diagn Microbiol Infect Dis
DOI: 10.1016/j.diagmicrobio.2020.115238
sha: c36a540ca5bb034513e832253bf73e61b5788bc4
doc_id: 916988
cord_uid: cqze6607

The COVID-19 pandemic has placed unprecedented global demand on laboratory supplies required for testing. Sample pooling has been investigated by laboratories as a strategy to preserve testing capacity. We evaluate the performance of Cepheid Xpert® Xpress SARS-CoV-2 RT-PCR assay for testing samples in pools of four and six. Clinical samples containing SARS-CoV-2, and confirmed negative clinical samples were used to create sample pools. Clinical samples had ‘neat’ Xpert® E gene cycle threshold (Ct) values ranging between 20 and 28 and all were detected qualitatively when contained in pools of four or six samples. For these samples, pooling had a median change in Ct value of 2.0 in pools of four, and of 2.9 in pools of six. With the use of Cepheid Xpert® Xpress SARS-CoV-2 RT-PCR assay, pooling of four or six samples may be an effective strategy to increase testing capacity.

Rapid and accurate diagnostic testing for SARS-CoV-2 is central to controlling the global COVID-19 pandemic. The Cepheid Xpert® Xpress SARS-CoV-2 assay (Cepheid, Sunnyvale, CA, USA) is a rapid, near-care, reverse-transcriptase PCR assay (RT-PCR), producing results within 45 minutes.

Manufacturer instructions for use claim a limit of detection (LoD) of 250 copies/mL, however recent work has demonstrated high analytical sensitivity of this assay, with a LoD approaching 100 viral copies/mL (1) (2) (3) (4) (5) . The sheer scale of global demand for laboratory reagents, including the Xpert® SARS-CoV-2 assay, has led many laboratories to investigate alternative strategies for optimizing the use of testing supplies, including sample pooling (6) (7) (8) (9) (10) (11) . The diagnostic performance of pooling depends on several factors, including assay sensitivity, prevalence of infection in the population being tested and sample types used for pooling (8) (9) (10) (11) . With increased recognition that asymptomatic SARS-CoV-2 infection may contribute to transmission, the populations tested have been expanded to include asymptomatic patients: in this context when prevalence rates are <5% sample pooling can substantially increase testing capacity (12) (13) (14) . On July 18th, 2020, the FDA issued its first Emergency Use Authorization (EUA) for sample pooling in diagnostic testing for SARS-CoV-2 by RT-PCR which applies for the Quest Diagnostics test for use with pooled samples containing up to four individual swab specimens (15) . Although pooling has previously been used successfully with the Xpert® MTB/RIF assay, there are limited data on the performance of pooling using the Xpert® SARS-CoV-2 assay (16) . Here, we investigated the performance of the Xpert® SARS-CoV-2 assay for detecting SARS-CoV-2 in pooled clinical samples. We chose to study pools of four and six samples based on the FDA EUA authorization, available literature for other assays (8) (9) (10) (11) (12) (13) (14) and our own experience with pooling using an in-house RT-PCR for SARS-CoV-2 (18) .

Nasopharyngeal swab samples were collected in viral transport media (Kang Jian, catalogue no In addition, serial dilutions of gamma-irradiated SARS-CoV-2 virus were prepared in viral transport media to allow correlation of viral concentration with Xpert® cycle threshold (Ct) values (Table 1) .

These serial dilutions of inactivated virus were used to create standards for a standard curve (Figure 1) from which the neat clinical sample extrapolated viral load was calculated.

All Xpert® testing was performed by one investigator (EW), blinded to previous testing results and pooling details. 

Clinical samples with 'neat' Xpert® E gene Ct values ranging between 20 and 28 were detected qualitatively when contained in pools of four or six samples (Table 2) . For these samples, pooling had a median change in E gene Ct (Ct) value of 2.0 in pools of four, and a Ct 2.9 in pools of six. All twenty-four negative samples were negative for SARS-CoV-2 by Xpert® when tested in four pools of six samples to confirm a specificity of pool testing of 100%.

Serial dilutions of inactivated SARS-CoV-2 virus between 4.17 x 10 2 copies/mL and 1 x 10 6 copies/mL were reliably detected by the Xpert® assay. Virus was detected when testing the lowest dilution of 4.17 x 10 2 copies/mL in a pool of six but not when tested in a pool of four negative samples. Figure 1 ).

We have found that testing samples in pools of four or six using Xpert® assay can increase testing capacity and that virus from samples with neat Ct values of between 20 and 28 can be reliably detected during pooling. Consistent with other studies, we have found that samples with viral load around the LoD for the assay used may be missed when testing in pools (14) . Since the required input into the Xpert® cartridge is fixed at 300µL, when testing a pool size of four, 75µL of each original sample is tested. In comparison, for assays where RNA extraction of the pool can be performed as a separate step greater possible original sample input volumes can be accommodated to increase sensitivity. However use of more sensitive assays such as the Xpert® to test sample pools is less likely to miss samples with low viral load when using lower sample input volumes. Furthermore, samples with low viral load, particularly in asymptomatic persons, may indicate the presence of noninfectious virus since studies have found that samples with higher Ct values are less likely to yield culturable virus (17) .

Strategies to optimize the performance of pooling include limiting its use to low prevalence situations such as testing of asymptomatic populations. By implementing pooling in low prevalence settings, the work of 'de-coupling' of pools for individual testing in the event of a positive pool is minimized. For a test with assumed sensitivity and specificity of 99% and a SARS-CoV-2 prevalence of under 5%, the expected number of PCR reactions required for testing of 1,000 samples in pools of four is under 500, inclusive of initial pool testing and deconstruction of positive pools (18) . The Xpert® assay is authorised to be used in patient care settings outside of the clinical laboratory environment and it has been widely used in low-resource settings particularly for the diagnosis of tuberculosis (16) . Its ease of use, including for pool testing, makes it an ideal assay for use in settings where expertise for SARS-CoV-2 testing may not be readily available. However, in view of high global demand for test kits the supply of cartridges does not always meet demand in some settings, which further supports the use of pooling with this assay.

During our study we developed laboratory strategies to mitigate against errors during the pooling process. These included: (i) oversight of the pool assembly process by a second staff member; (ii) checking of manual transcriptions by a second staff member; (iii) holding back reporting of negative pooled samples until individual testing of positive pools is complete; (iv) use of standardised worksheets for recording two identifiers (specimen number and patient name) of each specimen in the pool, and (v) requirement for operator signatures at each step of the testing process.

Here we demonstrate that with the use of a sensitive and specific molecular assay that is easy to use, pooling of four or six samples is an effective strategy to increase testing capacity. With limited testing resources, sample pooling may preserve testing capacity; findings of our study may be particularly valuable for low-resource settings. 

Multicenter evaluation of the Cepheid Xpert Xpress SARS-CoV-2 test

Detection of SARS-CoV-2 by use of the Cepheid Xpert Xpress SARS-CoV-2 and Roche cobas SARS-CoV-2 assays

Comparison of commercially available and laboratory-developed assays for in vitro detection of SARS-CoV-2 in clinical laboratories

Clinical evaluation of three sample-toanswer platforms for detection of SARS-CoV-2

Multi-center evaluation of cepheid xpert® xpress SARS-CoV-2 point-of-care test during the SARS-CoV-2 pandemic

Pooling of nasopharyngeal swab specimens for SARS-CoV-2 detection by RT-PCR

Saliva as a Noninvasive Specimen for Detection of SARS-CoV-2

Evaluating the efficiency of specimen pooling for PCR-based detection of COVID-19

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

Pooling of SARS-CoV-2 samples to increase molecular testing throughput

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

Assessment of specimen pooling to conserve SARS CoV-2 testing resources

Simulation of pool testing to identify patients with coronavirus disease 2019 under conditions of limited test availability

Large-scale implementation of pooled RNA extraction and RT-PCR for SARS-CoV-2 detection

COVID-19) Update: FDA Issues First Emergency Authorization for Sample Pooling in Diagnostic Testing FDA

Testing pooled sputum with Xpert MTB/RIF for diagnosis of pulmonary tuberculosis to increase affordability in lowincome countries

Viral RNA load as determined by cell culture as a management tool for discharge of SARS-CoV-2 patients from infectious disease wards

Sample Pooling is a Viable Strategy for SARS-CoV-2 Detection in Low-Prevalence Settings

We thank Professor Patrick Reading (The Peter Doherty Institute) for discussions during the planning stages and assistance with proof-reading of the draft manuscript.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.