Abstract
Background Real-Time PCR (RT-PCR) is widely used as the gold standard for clinical detection of SARS-CoV-2. However, due to the low viral load in patient throat and the limitation of RT-PCR, significant numbers of false negative reports are inevitable, which should not be ignored.
Methods We explored the feasibility of droplet digital PCR (ddPCR) to detect SARS-CoV-2 from 57 clinical pharyngeal swab samples and compared with RT-PCR in terms of the sensitivity and accuracy. Among 57 samples, all of which were reported as negative nucleic acid by officially approved clinical RT-PCR detection, 43 samples were collected from suspected patients with fever in clinic, and 14 were from supposed convalescents who were about to discharge after treatment. The experiment was double-blind.
Results The lower limit of detection of the optimized ddPCR is at least 500 times lower than that of RT-PCR. The overall accuracy of ddPCR for clinical detection is 94.3 %. 33 out of 35 negative pharyngeal swab samples checked by RT-PCR were correctly judged by ddPCR based on the follow-up investigation. In addition, 9 out of 14 (64.2 %) supposed convalescents with negative nucleic acid test twice by RT-PCR were positive by ddPCR detection.
Conclusions ddPCR shows superiority for clinical detection of SARS-CoV-2 to reduce the false negatives, which could be a powerful complement to the current standard RT-PCR. Before the ddPCR to be approved for diagnosis, the current clinical practice that the convalescent continues to be quarantined for 2 weeks is reasonable and necessary.
Introduction
The recent outbreak of coronavirus disease 2019 (COVID-19) caused by the infection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) poses a great threat to public health all over the world.1,2 On February 28, 2020, the world health organization (WHO) has upgraded the global risk level of this viral pneumonia from “high” to “very high”. According to WHO and Chinese Center for Disease Control and Prevention (CDC), the current gold standard for the diagnosis of SARS-CoV-2 infection is based on the real-time fluorescent quantitative PCR (RT-PCR), which means that the nucleic acid of SARS-CoV-2 could be detected in patient specimens using RT-PCR.3,4 However, the disadvantages of insufficient detection of RT-PCR are more and more prominent, especially the problem of detection dynamic range in the clinical application. At present, it has been found in clinical practice that some patients had fever, and chest CT showed symptoms of suspected viral pneumonia such as lower lobe lesions of the lungs, but the nucleic acid test of pharyngeal swab did not show positive results until 5-6 days after the onset of viral pneumonia. It was estimated that only 30 %-60 % positive results can be obtained among COVID-19 patients that further confirmed by chest CT.5 This might be explained by the relatively low viral load in the throat of patients and the sensitivity limitation of RT-PCR technology, which inevitably produced the false negatives during the clinical diagnosis, leading to a potential risk of viral transmission. Besides, supposed convalescent, who is about to discharge, also need multiple tests with negative results for confirmation. Therefore, it is a pressing needs for a more sensitive and accurate detection method for the pathogenic detection.
Digital PCR is based on the principles of limited dilution, end-point PCR, and Poisson statistics, with absolute quantification as its heart.6 It has broader dynamic range without external interference and robustness to variations in PCR efficiency. 7–9 In 2011, Hindson developed the droplet digital PCR (ddPCR) technology based on traditional digital PCR.10 The reaction mixture can be divided into tens of thousands of nanodroplets during the process. These vast and highly consistent oil droplets substantially improve the detection dynamic range and accuracy of digital PCR in a low-cost and practical format.11 In recent years, this technology has been widely used, such as analysis of absolute viral load from clinical samples, analysis of gene copy number variation, rare allele detection, gene expression, microRNA analysis and genome edit detection et al.12,13,14,15 Here, taking the advantages of ddPCR, we optimized the preparation of pharyngeal swab samples, and develop a workflow of ddPCR to detect SARS-CoV-2 using Chinese CDC approved primer and probe sets. Based on the results of this optimized ddPCR system, we showed that the overall accuracy of the ddPCR for clinical pathogen detection is 94.3 %, and 64.2 % of supposed convalescents with two consecutive negative nucleic acid tests by RT-PCR still carry SARS-CoV-2.
Materials and methods
Ethics statement
This study was approved by the Ethics Committee of the Renmin Hospital and Zhongnan Hospital of Wuhan University. The analysis was performed on existing samples collected during standard diagnostic tests, posing no extra burden to patients, as described previously.2
Specimen collection and RNA extraction
Pharyngeal swab samples were obtained from clinical suspected patients with fever or rehabilitation quasi-discharged patients of COVID-19 at Renmin Hospital and Zhongnan Hospital of Wuhan University according to the interim guidance of WHO. Pharyngeal swabs were soaked in 500 μl PBS and vortexed with diameter of 3 mm beads (Novastar, China) for 15 seconds immediately. Total RNA was extracted from the supernatant using QIAamp viral RNA mini kit (Qiagen) following manufacturer’s instruction. First strand cDNA was synthesized using PrimeScript RT Master Mix (TakaRa) with random primer and oligo dT primer.
Primers and probes
The primers and probes targeted the ORF1ab and N of SARS-CoV-2 according to Chinese CDC. Target 1 (ORF1ab), forward: 5’-CCCTGTGGGTTTTACACTTAA-3’,
reverse: 5’-ACGATTGTGCATCAGCTGA-3’, probe: 5’-FAM-CCGTCTGCGGTATGTGGAAAGGTTATGG-BHQ1-3’;
Target 2 (N), forward: 5’-GGGGAACTTCTCCTGCTAGAAT-3’,
reverse: 5’-CAGACATTTTGCTCTCAAGCTG-3’,
probe: 5’-FAM-TTGCTGCTGCTTGACAGATT-TAMRA-3’.16
Droplet Digital PCR workflow
All the procedure follow the manufacture instructions of the QX200 Droplet Digital PCR System using supermix for probe (no dUTP) (Bio-Rad). Briefly, the TaqMan PCR reaction mixture was assembled from a 2× supermix for probe (no dUTP) (Bio-Rad), 20× primer and probes (final concentrations of 900 and 250 nM, respectively) and template (variable volume) in a final volume of 20 μl. Twenty microliters of each reaction mix was converted to droplets with the QX200 droplet generator (Bio-Rad). Droplet-partitioned samples were then transferred to a 96-well plate, sealed and cycled in a T100 Thermal Cycler (Bio-Rad) under the following cycling protocol: 95 °C for 10 min (DNA polymerase activation), followed by 40 cycles of 94 °C for 30 s (denaturation) and 60 °C for 1 min (annealing) followed by an infinite 4-degree hold. The cycled plate was then transferred and read in the FAM channels using the QX200 reader (Bio-Rad).
RT-PCR
The primers and probes used in ddPCR are also used in RT-PCR. A 30-μl reaction was set up containing 10 μl of RNA, 18.5 μl of reaction buffer provided with the one step RT-PCR system and 1.5 µl enzyme mix (BGI BIOTECHNOLOGY). Thermal cycling was performed at 50 °C for 20 min for reverse transcription, followed by 95°C for 10 min and then 40 cycles of 95 °C for 15 s, 60 °C for 30 s in BIO-RAD CFX96 Touch RT-PCR system.
Data statistical analysis
Analysis of the ddPCR data was performed with Quanta Soft analysis software v.1.7.4.0917 (Bio-Rad) that accompanied the droplet reader calculate the concentration of the target DNA sequences, along with their Poisson-based 95 % confidence intervals. The positive populations for each primer/probe are identified using positive and negative controls with single (i.e., not multiplexed) primer–probe sets. The concentration reported by QuantaSoft equals copies of template per microliter of the final 1× ddPCR reaction, which was also used in all the results. In addition, plots of linear regression were conducted with GraphPad Prism 7.00, and probit analysis for lower limit of detection (LLoD) was conducted with StatsDirect software v3.2.9. Lower limit of quantitation (LLoQ) and LLoD were defined as the lowest concentration at which 95 % and 50 % of positive samples were detected, respectively.
Results
Comparison of the lower limit between ddPCR and the standard RT-PCR
Using a manual threshold to define positivity, 9 % of negative controls (3/32) were scored as positive due to one single positive droplet (data not shown). The presence of two positive droplets or more was not observed for negative controls. Serial dilutions of a positive control DNA fragment of SARS-CoV-2 were tested with primers/probe sets targeting ORF1ab and N of SARS-CoV-2, respectively for ddPCR. It shows good linearity (R2: 0.9932 and 0.9824, respectively) (Fig. 1A and 1B). Reportable range of ddPCR is from 10 copies/μl to 2500 copies/μl for both ORF1ab and N primes/probe sets. In contrast, the dynamic range of RT-PCR is from 50 copies/μl to 105 copies/μl for both ORF1ab and N primes/probe sets (Fig. 1C and 1D). To define the limit of quantification of ddPCR, five low concentrations of plasmid control were analyzed with 8 replicates. The lower limit of quantitation (LLoQ) of the optimized ddPCR is 1.003 copies/μl and 0.415 copies/μl for ORF1ab and N primers/probe sets, respectively. The lower limit of detection (LLoD) of the optimized ddPCR is 0.109 copies/μl and 0.021 copies/μl for ORF1ab and N primers/probe sets, respectively (Fig. 2), which is at least 500 times lower than the RT-PCR detection kit used in current clinical test. Therefore, the ddPCR is more sensitive for samples with low level analyte.
Plot of results from a linearity experiment to determine reportable range of ddPCR and RT-PCR targeting for ORF1ab and N of SARS-CoV-2. (A and B) Expected values (converted to log10) were plotted on the X axis versus measured values (converted to log10) on the Y axis using Graph Pad Prism for ddPCRa targeting ORF1ab and N. (C and D) Expected values (converted to log10) were plotted on the X axis versus measured Ct values on the Y axis using Graph Pad Prism for RT-PCR targeting ORF1ab and N. Data are representative of three independent experiments with 3 replicates for each concentration.
Probit analysis sigmoid curve reporting the lower limit of quantitation (LLoQ) and the lower limit of detection (LLoD) of ddPCR. Replicate reactions of SARS-CoV-2 (A) ORF1ab and (B) N were done at concentrations around the detection end point determined in preliminary dilution experiments. The X axis shows expected concentration (copies/μl). The Y axis shows fraction of positive results in all parallel reactions performed. The inner line is a probit curve (dose-response rule). The outer lines are 95 % CI. Data are representative of three independent experiments with 8 replicates for each concentration.
Detection of SARS-CoV-2 from patient specimens with ddPCR
57 clinical pharyngeal swab samples (Fig. 3), which were judged to be negative by both officially approved clinical RT-PCR detection and the commercial RT-PCR detection kit for double check (generally referred to as RT-PCR), were tested with ddPCR in double-blind. We did not know any information, results of clinical diagnosis and status of enrolled patients during the tests. The follow-up investigation revealed those information after ddPCR tests. Compared with the information and clinical diagnosis, our results show that the overall accuracy of the optimized ddPCR is 94.3 % and 64.2 % of supposed convalescents are still carrying SARS-CoV-2. Details are as follows (Fig. 3) (Table 1 and 2):
Detection results of ddPCR for febrile and suspected patients of COVID-19.
Detection results of ddPCR for supposed convalescent patients who is about to be discharged after treatment.
Information diagram of detection results with ddPCR and subsequent clinical diagnosis for both convalescent and febrile suspected patients.
Firstly, among 27 febrile suspected patients whose SARS-CoV-2 nucleic acid were negative initially tested by RT-PCR, 25 out of 27 were detected with ddPCR as positive and 2 out of 27 were negative. However, all 27 patients were diagnosed with SARS-CoV-2 infection by chest CT as well as RT-PCR in subsequent follow-up investigations, and all of them were hospitalized. As a result, 92.6 % of patients with false negative nucleic acid test could be identified as positive by the optimized ddPCR (Table 1).
Secondly, pharyngeal swabs of 8 febrile patients with negative results tested by RT-PCR were also tested negative by ddPCR. In the follow-up investigation COVID-19 was excluded based on the normal results of chest CT and RT-PCR (Table 1).
Thirdly, pharyngeal swabs collected from 8 febrile suspected patients in the clinic recently with negative nucleic acid tests by RT-PCR, were detected positive by ddPCR. However, chest CT of these 8 patients did not show any abnormalities upon their first visit the clinic. According to official clinical guidelines, these 8 patients were home quarantined and no further followed-up by us (Table 1).
Finally, pharyngeal swabs of 14 supposed convalescent were tested negative in two consecutive tests by RT-PCR (Table 2). However, using ddPCR, 9 out of 14 were positive with a positive rate of 64.2 %. Therefore, the current clinical practice that the convalescent continues to be quarantined for 2 weeks is reasonable and necessary. In conclusion, compared with RT-PCR, ddPCR show superiority for clinical detection of SARS-CoV-2 to reduce the false negatives, which could be a powerful complement to the current standard RT-PCR.
Discussion
More and more nucleic acid detection kits have been developed for SARS-CoV-2 recently based on RT-PCR to meet the requirement of large-scale clinical molecular diagnosis. It has been reported that 6 kinds of RT-PCR detection kits were compared and analyzed for their detection performance. Results showed that there are differences in the detection ability of these kits for weakly positive samples, and the accuracy, sensitivity and reproducibility of some reagents are not ideal 17 In the meantime, many efforts have been focusing on developing better and complementary technology for clinical diagnosis of SARS-CoV-2, due to the limited sensitivity and precision of RT-PCR for viral quantitation. Different from RT-PCR that the data are measured from a single amplification curve and a Cq value, which is highly dependent on reaction efficiency, primer dimers and sample contaminants, ddPCR is measured at reaction end point which virtually eliminates these potential pitfalls. Results in this work proved that ddPCR is more sensitive (Fig. 1) and accurate for low viral load diagnosis (Fig. 2), which can greatly reduce the false negatives detection (Fig 3).
Based on two primers/probe sets targeting ORF1ab and N of SARS-CoV-2, results showed that N primers/probe set was more sensitive compared to that of ORF1ab. Among 42 samples that were judged as positive with ddPCR, 40 in 42 were detected as positive by N primers/probe set, and 12 in 42 were detected as positive by ORF1ab primers/probe set. This could be explained by the subgenomic RNA discontinuous replication and transcription model of coronavirus. The genome RNA of SARS-CoV-2 encodes single copy of ORF1ab and N, respectively. In contrast, a nested set of around 10 subgenomic RNAs (sgRNAs), each of which encodes one copy of N, are synthesized by viral replication and transcription complex in a manner of discontinuous transcription.18,19,20 Therefore, the copy numbers of N gene is significantly higher than that of ORF1ab gene in SARS-CoV-2 infected cells.
Although 2 patients, who were clinically confirmed by chest CT and RT-PCR subsequently, were reported as negative nucleic acid in pharyngeal swabs by our ddPCR, leading to 2 false negative reports by ddPCR in 35 cases (5.7 % missing rate), the overall accuracy of SARS-CoV-2 detection is significantly improved, which will benefit to the early diagnosis, intervention and treatment.
Notably, 64.2 % supposed convalescent patients, who are negative for pharyngeal swab nucleic acid tests twice by RT-PCR, are still carrying SARS-CoV-2 based on our work. Although there is no evidence that such COVID-19 convalescent carrying SARS-CoV-2 will be infectious to other healthy person, the risk still exists. Therefore, the current clinical practice that the convalescent continues to be quarantined for 2 weeks is reasonable and necessary. And we recommend that ddPCR could be a complement to the current standard RT-PCR to re-confirm the convalescent, which will benefit to reduce the risk of the SARS-CoV-2 epidemic and social panic.
Author Contributions
YC, KL conceptualized the study design. TS, WH, LD, TC, YX, and GC recruited the patients, collected specimens, collected demographic, clinical data; XL, MG, QZ, XW, YY, MS, DG and ZH did the laboratory tests. JF, YL and QZ plotted the figures; XL, MG, JF and YC analyzed the data; ZH, XK, YL, YlL and YC interpreted the results; JF wrote the initial drafts of the manuscript; YC and KL revised the manuscript and FL and KX commented on it. All authors read and approved the final report.
Funding
This study was supported by National Science and Technology Major Project (#2018ZX10733403 and #2018YFA0900801), China NSFC grants (#81672008) and Hubei Natural Science Foundation (#2018CFA035), Basic Scientific Research Foundation of Central Universities (#2042019gf0026), Ministry of Science and Technology of China, the National Mega Project on Major Infectious Disease Prevention (#2017ZX10103005) and National Key Research and Development Program of China (#2018YFE0204500). None of the funders had any role in the study design and the collection, analysis, and interpretation of data or in the writing of the article and the decision to submit it for publication. The researchers confirm their independence from funders and sponsors.
Data Availability
The authors confirm that the data supporting the findings of this study are available within the article.
Declaration of interests
No authors have received research funding from the company whose commercial products were used in this work. All authors report no competing interests. All authors have completed the Unified Competing Interest form.
Acknowledgement
We are grateful to Taikang Insurance Group Co., Ltd and Beijing Taikang Yicai Foundation for their great support to this work.
Reference
