Pooling Upper Respiratory Specimens for Rapid Mass Screening of COVID-19 by Real-Time RT-PCR

To validate the specimen-pooling strategy for real-time reverse transcription PCR detection of severe acute respiratory syndrome coronavirus 2, we generated different pools including positive specimens, reflecting the distribution of cycle threshold values at initial diagnosis. Cumulative sensitivities of tested pool sizes suggest pooling of <6 specimens for surveillance by this method.


Materials and Methods for Protocol Development a. Specimen selection and preparation of pooled specimens
negative pooled specimens, we randomly combined 16 specimens from 300 negative specimens and made 60 negative pooled specimens.

c. Real-time RT-PCR (rRT-PCR)
We performed rRT-PCR (PowerCheck 2019-nCoV) with all pooled specimens. The interpretation guideline by manufacturer for SARS-CoV-2 positivity was Ct cutoff <35 for single specimen. However, we checked every amplified curve until the end of the 40 PCR cycles in this study. When either the E or RdRp gene was observed to have any amplified curve within the end of the amplification cycles, 40, we interpreted the result as positive for the pooled specimens.
When no amplification curves were observed until the Ct value reached 40 for both genes, the result was interpreted as negative.
The PCR protocol that we used was based on the protocol by Corman et al. (3). The Korea Centers for Disease Control and Prevention (KCDC) also used laboratory-developed tests (LDT) based on the same protocol. All of these protocols use the same primer and probe sequences. The KCDC also evaluated pooled specimens with various pool sizes with their laboratory-developed tests, and similar results were obtained (data not shown). Although other protocols based on the same protocol might show similar sensitivities with pooled specimens, we strongly recommend that the optimal pool sizes should be evaluated for each protocol.

Protocol
Based on the results obtained, the COVID-19 task force of KSLM and the KCDC Center for Laboratory Control of Infectious Diseases has recommended the following protocol for molecular diagnostic testing of pooled specimens of upper respiratory specimens. This protocol is not recommended for use in other situations, such as the initial diagnosis of symptomatic patients. This protocol is aimed to maximize testing capacity with minimal loss of sensitivity in the following situations:

1)
When rapid testing is critical for screening positive cases in large populations, e.g., screening the asymptomatic patients in the long-term care hospitals, or screening in the healthcare works without symptom.

2)
When testing resources are limited. Because differences in specimen collection, nucleic acid extraction, and real-time RT-PCR reagent can affect sensitivities of pooled specimens, laboratories must validate the possible effect of them before using this protocol.

a. Specimen types
-Upper respiratory specimens including oropharyngeal and nasopharyngeal swab specimens are recommended to be used for this protocol.
-When lower respiratory specimens are required from patients in accordance with clinical or epidemiological findings, testing using individual specimens is recommended rather than pooled specimens.

b. Specimen container
-Virus transport media; for example, UTM, or -Special media for virus molecular diagnostic testing; for example, guanidinium-based buffer, such as Copan eNAT, can be used.
Special media for molecular diagnostic testing contain substances to inactivate the virus, which is beneficial for specimen handling. Pooling must be performed with the specimens of the same media type.

c. Specimen collection, transport, and general biological safety of the specimen
Upper respiratory tract specimens should be collected and handled in accordance with the KCDC (4,5) or KSLM (1) guidelines.

d. Specimen pooling 1) Determination of the number of specimens pooled and volumes for nucleic acid extraction
The total number of specimens to be pooled needs to be determined. The volume for nucleic acid extraction should be checked in accordance with the manufacturer's instructions.
The probability of obtaining positive results from the pooled samples is higher as more specimen volume is used for nucleic acid extraction.
A predetermined dilution factor should be selected after careful consideration of the characteristics of the targeted population and the expected disease prevalence. We recommend the pooling of <6 specimens. If the 96% sensitivity is tolerable, the pooling of <10 specimens might be considered.

2) Protocol for specimen pooling
a) It is recommended for the personnel performing the specimen pooling to wear appropriate personal protective equipment (N95, KF94, or equivalent or superior respiratory protectors, full-body gown, gloves, etc.) and to process specimens, using a ≥Class II biosafety cabinet (BSC) in a biosafety level 2 facilities. Despite processing specimens in a molecular diagnostic testing tube with the inactivation substance, the same procedure should be followed. When specimens are processed outside of the BSC, personal protective equipment, including an N95 respiratory mask or other superiorquality respiratory protectors, must be worn. After processing, the safety cabinet should be sterilized with 70% isopropyl alcohol or equivalent. Therefore, calculations should be performed with more than 10% overfill to prepare pooled specimens (Table 1). Numbers in Table 1 can be modified when preparing other volumes of pooled specimens.

e. Nucleic acid extraction
-Nucleic acid extraction from the pooled specimens should be performed in accordance with the protocol of the manufacturers of the nucleic acid extraction reagent/kit or the test reagents.
-When using a device with high extraction efficiency and when using more specimen volume for nucleic acid extraction, chances of viral detection are higher in pooled specimens. Therefore, the number of specimens used for nucleic acid extraction and the eluted amount and concentration of the extracted nucleic acid should be confirmed before using the device.
-An automated extraction device is advantageous for efficient extraction of numerous specimens.

f. Test implementation
-Tests will be performed in accordance with the manufacturer's guidelines, using RNA extracted from pooled specimens.

-This protocol is based on the results by PowerCheck 2019 nCoV Real-Time Detection
Kit, emergency use authorization version (Kogene Biotech). Any modification of the protocol should be verified by each laboratory before implementation. The extent of its application can be modified in accordance with future scientific evidence.

g. Interpretation of the test results using pooled specimens
Based on COVID-19 laboratory guidelines, the test result from an individual patient should be considered positive only when all genes targeted in the diagnosis kit are detected (1,6). However, this protocol is aimed at screening positive specimens from multiple patients and not at making a confirmed diagnosis (Table 2).  Below cutoff Positive results on pooled samples screening (SARS-CoV-2 genes detected in the pooled specimens) Individual specimens should be tested Amplification within the maximum cycles predetermined per the manufacturer's protocol for >1 genes but not satisfying the standard for positive results for pooled specimens Undetermined results on pooled samples screening (Possibility of SARS-CoV-2 genes in the pooled specimens) ® Individual specimens should be tested Amplification beyond the maximum cycles predetermined per the manufacturer's protocol is NOT observed for either gene Negative results on pooled samples screening (Absence of SARS-CoV-2 genes in the pooled specimens)

Recommendations for Adaptation of Authors' Protocol
This article is based on the opinion of authors. We discussed the factors related to design, the experiment, and guideline development. The opinions expressed in the supplemental data by authors do not necessarily reflect the opinions of the Korea Centers for Disease Control and Prevention or the institutions with which the authors are affiliated.

Analytical Performance and Characteristics of RT-PCR Kits
A. Analytical sensitivity (limit of detection) is a critical factor for the performance of pooling specimens. Poor sensitivity can cause false negatives, causing disastrous situations. Therefore, for the adaptation of the RT-PCR kit to the pooling strategy, the analytical sensitivity of the intended kit should be evaluated.
B. Analytical specificity should also be evaluated for the possibility of false-positive results, as it will influence the efficiency, including cost-effectiveness.
C. Most countries have their own systems for the Emergency Use Authorization (EUA) or regular approval of in vitro diagnostic (IVD) products. We highly recommend using IVD kits approved by the domestic EUA or Regular System. D. When using multiplex PCR kits, the criteria for interpretation of pooled specimens should be determined before using. The kits use variable target genes and show variable amplification efficiency, and their sensitivity among genes may differ.
E. The threshold for fluorescence, the cutoff for Ct values, and the maximum amplification cycles should be considered together to determine the maximum pool size.

RNA Extraction Method
A. The purity of RNA and efficiency of the RNA extraction method may influence the results. Therefore, the RNA extraction method should be validated in a laboratory with the selected RT-PCR kit.
B. One of the main factors determining the final volume for pooling is the sample input volume used for RNA extraction. The most frequently recommended volume for RNA extraction by commercial kits is approximately 200 μL. Some systems can be used with larger volumes up to 1 mL. Users can validate various protocols with different sample volumes to determine the efficiency, cost-effectiveness, and performance of the test.

Type and Quality of Specimens
A. When making the specimen pool, the volume from each sample should be the same. To obtain the same amount of the specimen, the specimen should be homogenous and liquid. Therefore, upper respiratory specimens in media are good for pooling, in contrast to sputum.
B. Regarding the specimen buffer, to avoid the matrix effect, we recommend using the specimens with same specimen buffer for pooling.
C. After pooling, to make the pooled sample homogenous, rigorous vortexing is required.

Pooling Method
There might be three types of pooling methods. We recommend using media pooling for efficiency and ease of handling.