key: cord-0840942-c9dt9oaf
authors: Anahtar, Melis N; McGrath, Graham E G; Rabe, Brian A; Tanner, Nathan A; White, Benjamin A; Lennerz, Jochen K M; Branda, John A; Cepko, Constance L; Rosenberg, Eric S
title: Clinical assessment and validation of a rapid and sensitive SARS-CoV-2 test using reverse-transcription loop-mediated isothermal amplification without the need for RNA extraction
date: 2020-12-21
journal: Open Forum Infect Dis
DOI: 10.1093/ofid/ofaa631
sha: da1e4ecdeb2a3ad38cc1c57874cdb8b4d6530452
doc_id: 840942
cord_uid: c9dt9oaf

BACKGROUND: Amid the enduring pandemic, there is an urgent need for expanded access to rapid, sensitive, and inexpensive COVID-19 testing worldwide without specialized equipment. We developed a simple test that uses colorimetric reverse-transcription loop-mediated isothermal amplification (RT-LAMP) to detect SARS-CoV-2 in forty minutes from sample collection to result. METHODS: We tested 135 nasopharyngeal specimens from patients evaluated for COVID-19 infection at Massachusetts General Hospital. Specimens were either added directly to RT-LAMP reactions, inactivated by a combined chemical and heat treatment step, or inactivated then purified with a silica-particle based concentration method. Amplification was performed with two SARS-CoV-2-specific primer sets and an internal specimen control; the resulting color change was visually interpreted. RESULTS: Direct RT-LAMP testing of unprocessed specimens could only reliably detect samples with abundant SARS-CoV-2 (>3,000,000 copies/mL), with sensitivities of 50% (95% CI, 28 to 72) and 59% (95% CI, 43 to 73) in samples collected in universal transport medium and saline, respectively, compared to qPCR. Adding an up-front RNase inactivation step markedly improved the limit of detection to at least 25,000 copies/mL, with 87.5% (95% CI, 72 to 95) sensitivity and 100% specificity (95% CI, 87 to 100). Using both inactivation and purification increased the assay sensitivity by ten-fold, achieving a limit of detection comparable to commercial real-time PCR-based diagnostics. CONCLUSION: By incorporating a fast and inexpensive sample preparation step, RT-LAMP accurately detects SARS-CoV-2 with limited equipment for about US$6 per sample, making this a potentially ideal assay to increase testing capacity especially in resource-limited settings.

The worldwide spread of COVID-19 has led to an unprecedented need for rapid, accurate, affordable, and readily available SARS-CoV-2 tests. Hundreds of molecular assays for the detection of SARS-CoV-2 RNA have received regulatory approval in the US, Europe, and Asia to date, but they have not met the need for widespread testing demand due to several critical factors including: a high cost per reportable result (in the 15-40 US dollar range), and costly up-front capital equipment such as proprietary testing platforms, real-time amplification and detection platforms, or automated RNA extraction equipment and consumables that are in limited supply 1, 2 . In general, these tests must be performed by highly trained molecular laboratory professionals, in well-resourced laboratories. The development of more simple, rapid, and low-cost diagnostics that do not rely on the same supply chains, reagents or consumables as other COVID-19 tests could help rapidly and substantially expand testing capabilities, especially in resource-limited settings.

Alternative rapid tests to detect SARS-CoV-2 rely on detection of viral antigen using lateralflow immunoassays (LFA). While extremely convenient, respiratory viral LFAs tend to be less sensitive than nucleic-acid amplification methods, with an average sensitivity of 61-75% 3, 4 . As an alternative to antigen detection methods and resource-intensive real-time PCR tests, isothermal amplification methods such as loop-mediated isothermal amplification (LAMP) 5 and recombinase polymerase amplification (RPA) 6, 7 enable sensitive detection of nucleic acids with just the use of a stable heat source in as little at 15 minutes. Colorimetric RT-LAMP expands on the basic LAMP technology with a one-pot reaction that contains both reverse transcriptase and DNA polymerase with visual detection of nucleic acid amplification due to a pH indicator dye within the master mix, obviating the need for additional detection equipment 8 . LAMP has been used to detect many pathogens including SARS 9 , Zika virus 10 , Mycobacterium tuberculosis 11 , malaria 12 , and human leishmaniasis 13 . RT-LAMP has also been performed for SARS-CoV-2 on extracted RNA with a colorimetric read-out [14] [15] [16] [17] [18] or for subsequent CRISPR-Cas12-based detection 19 .

A c c e p t e d M a n u s c r i p t 5 To develop a truly accessible sample-to-answer nucleic acid-based diagnostic test, one must couple a simple detection method with an equally simple sample preparation method. The simplest sample preparation method is to directly add sample to the amplification reaction 18 , but this can be problematic for several reasons. Endogenous RNases present in body fluids can degrade target RNA and infectious virus contained in the sample may increase the risk of laboratory acquired infection among technologists who handle the specimens. While heat inactivation alone can partially reduce RNase activity and inactivate virions 20 , we and others have shown that RNAses can be fully inactivated by combining heat inactivation with chemical inactivation using the shelf-stable reducing agent, Tris(2-carboxyethyl)phosphine (TCEP), and the divalent cation chelator EDTA 21, 22 . An additional challenge specific to colorimetric RT-LAMP is that the buffer and phenol red in viral transport media (VTM) may interfere with the pH-mediated color change. To circumvent these issues, most SARS-CoV-2 molecular tests use extracted RNA as input, but RNA extraction is expensive, time consuming, laborious, and extraction kits are in short supply.

As part of an ongoing quality improvement initiative, we tested 135 clinical nasopharyngeal samples collected from Massachusetts General Hospital patients who were admitted or evaluated in the Emergency Department during the COVID-19 pandemic to determine the testing characteristics of three diagnostic strategies using colorimetric RT-LAMP (Fig. 1 ). The first was the direct-fromsample approach, including samples collected in either universal transport media or sterile physiologic saline. The second incorporated an upfront five-minute chemical and heat inactivation step to inhibit RNases and lyse virions. The third strategy uses both the aforementioned inactivation step with an additional nucleic acid purification step using a solution of silica particles ("glass milk") to increase the effective sample input volume into the RT-LAMP reaction 21 .

A c c e p t e d M a n u s c r i p t 6

Nasopharyngeal samples were collected in 1 mL of sterile physiologic saline from the inpatient units and the Emergency Department (ED) of Massachusetts General Hospital (MGH) between March and April 2020. The inpatient samples were a prospectively collected convenience set obtained from patients whose COVID-19 status was known (20 qPCR positive, 17 qPCR negative).

The ED samples were collected prospectively as part of a laboratory quality improvement initiative from patients who presented within a 24-hour period and required clinical COVID-19 testing (22 qPCR positive, 45 qPCR negative). In addition, the nasopharyngeal samples collected in 3 mL of universal viral transport media were obtained from excess material collected for routine clinical care had Ct values obtained using the cobas instrument, 11 with the LDT, and 1 with the Xpert. The cobas ® 6800 system's cycle threshold tends to be within 2 cycles of the LDT's. If the Ct of a saline specimen was not available, the Ct from the paired VTM specimen that was collected simultaneously was used as a proxy. Though these qPCR assays are not truly quantitative due to variability in sample input, approximate conversions between cycle thresholds and viral copies/µL were calculated with a standard curve generated on the LDT by spiking 0, 10 1 , 10 2 , 10 3 , and 10 4 copies/µL of SARS-CoV-2 N gene RNA into COVID-negative nasopharyngeal specimens. Samples were aliquoted and quickly frozen at -20 C for additional testing to avoid RNA degradation. RT-LAMP assay performance was not affected after samples underwent a freeze-thaw cycle. 

The SARS-CoV-2 ORF1a gene (HMS Assay 1e) 21 , SARS-CoV-2 N gene (NEB N-A) 23 , and human actin B gene (generously provided by New England Biolabs) primer sequences are listed in Table S1 .

The ORF1a primers were combined into a 10X primer stock using 16 μM of Forward Inner Primer 

RT-LAMP testing was performed in biosafety level 2, Clinical Laboratory Improvement Amendments (CLIA)-certified clinical laboratory space. Regardless of the sample preparation method, each sample was amplified with two SARS-CoV-2-specific primer sets for the ORF1a and N genes. An additional primer set for the human actin B gene also served as an internal specimen control to detect the presence of inhibitory substances. A negative and positive control were tested with every set of clinical samples. Each 25µL RT-LAMP reaction was performed as described by the manufacturer's protocols with WarmStart® Colorimetric RT-LAMP 2X Master Mix (New England Biolabs, Ipswich, MA, USA) using a 1 µL sample input for samples collected in VTM and 5 µL input for samples collected in normal saline. After the thirty-minute heating step, the results were visually interpreted ( Fig. 1 and Table 1 .) The interpretative criteria are listed in Table 1 . After interpretation, A c c e p t e d M a n u s c r i p t 8 tubes were discarded or stored in sealed bags without re-opening to prevent post-amplification contamination of workspaces.

The 100X inactivation reagent and purification reagents were prepared as described elsewhere 21 When purification was performed, 250-500 µL of the inactivated sample was mixed with 5 µL of glass milk in a 1.5 mL tube, thoroughly resuspended, and mixed with half the initial sample volume of binding reagent. The binding reagent was comprised of 6 M NaI (MilliporeSigma), 10 mM HCl (Millipore Sigma), and 2% Triton X-100 (MilliporeSigma). The sample was then incubated at room temperature for 10 minutes with manual inversions approximately every two minutes to resuspend the silica. The samples were briefly spun in a mini benchtop centrifuge for several seconds and the supernatant was poured off. The pellet was washed with 700 µL of 80% ethanol and briefly spun.

The supernatant was poured off again and briefly re-spun. Any visibly remaining ethanol was removed with a P20 pipette and the pellet was air-dried on a heat block at 65 C for five minutes or until the pellet was visibly dry. Twenty-five µL of colorimetric RT-LAMP reaction mix was added to the pellet, resuspended, and transferred to a 0.2 mL tube for incubation at 65 C for 30 minutes, briefly placed in ice to enhance the color change, and visually inspected. 

With the goal of directly adding unprocessed sample into the RT-LAMP reaction ( Fig. 2A) , we first optimized the transport media input volume. We added increasing amounts of transport media to a standardized reaction containing 1,000 copies of SARS-CoV-2 control RNA. VTM interfered with the colorimetric readout, with complete inhibition of the pH-mediated color change with 3 µL of input, while saline had little effect (Fig. 2B) . Subsequent experiments were conservatively performed with 1 µL of VTM and 5 µL of saline sample input to facilitate robust assay performance in the setting of intrinsic clinical sample variability.

We next asked whether SARS-CoV-2 could be consistently detected from unprocessed clinical VTM and saline samples. We tested 16 qPCR-positive and 15 qPCR-negative NP specimens collected in VTM by adding 1 µL of VTM directly to the RT-LAMP reactions. When compared to qPCR on an FDA EUA approved platform, the sensitivity of RT-LAMP performed with the SARS-CoV-2 N gene and human actin gene primer sets and direct addition of a VTM specimen was only 50% (95% CI, 28 to 72) (Fig. 2C) . RT-LAMP could only detect VTM samples with a cycle threshold less than 23, corresponding to approximately 3,000,000 copies/mL in internal validation studies. There were two A c c e p t e d M a n u s c r i p t 11 false positives results, possibly related to interpretation difficulties due to a limited dynamic color range and higher background of the N gene primer set.

We next tested NP specimens directly inoculated into saline transport media using both the N and ORF1a primer sets and the interpretation criteria listed (Table 1) . Of 40 qPCR-positive samples tested with direct saline addition, RT-LAMP consistently detected samples with cycle thresholds less than 25 and as high as 32, yet the assay sensitivity was only 59% (95% CI, 43 to 73) ( Figure 2D ).

Among 45 qPCR-negative saline samples tested, the color changes were crisper and easier to interpret compared to VTM and there were no false positives. This was consistent with in silico and in vitro analyses that did not demonstrate cross reactivity between the ORF1a and N gene primer sets and other coronaviruses or respiratory viruses (Table S2) . Overall, normal saline appeared to be a more amenable sample collection media compared to VTM but direct sample addition to the RT-LAMP reaction remained too insensitive for routine clinical use.

We next tested whether the assay sensitivity would improve with a simple inactivation step consisting of TCEP/ EDTA addition to neutralize endogenous RNase activity and heat to release the viral RNA contained within virions and human cells (Fig. 3A) 20, 22 . The inactivation step appeared highly effective in nasopharyngeal specimens spiked with serially diluted SARS-CoV-2 control RNA ( Fig. S1 ) and enabled performance of limit of detection (LoD) studies. As little as 25 copies/µL (25,000 copies/mL) of control SARS-CoV-2 RNA could be detected in all 20 replicates using the ORF1a primers, and the N gene primers appeared slightly less sensitive (Fig. 3B) .

We then incorporated the inactivation step into the testing of clinical samples and observed a substantial improvement in the assay sensitivity. COVID-positive samples that were previously falsely negative with unprocessed sample addition were subsequently positive with both SARS-CoV-2 A c c e p t e d M a n u s c r i p t 12 primer sets after inactivation (Fig. 3C) . Importantly, COVID-negative samples remained negative after inactivation (Fig. 3C, Sample 3) . To systematically test the efficacy of inactivation, we repeated the assay using inactivation with the available 32 qPCR-positive and 30 qPCR-negative samples that had originally been tested by direct sample addition. We found 100% specificity (95% CI, 87 to 100) and 87.5% sensitivity (95% CI, 72 to 95) with this sample set (Fig. 3D ). In addition, we found that inactivation enabled the detection of over 95% of samples with a cycle threshold below 30, corresponding to about 40 viral copies/µL (40,000 copies/mL), and could detect SARS-CoV-2 in samples with cycle thresholds as high as 33.5 (Fig. 3E) . Thus, the combination of a very simple inactivation step followed by RT-LAMP provided a robust rule-in test for SARS-CoV-2 with a sample to result time of approximately forty minutes and minimal labor.

Finally, we asked whether increasing the effective sample input volume using a concentration and purification step could enable detection of samples with very low levels of virus that would otherwise not be detected (falsely negative) with inactivation alone (Fig. 4A) . We used the glass milk protocol to concentrate up to 500 µL of heat and TCEP-inactivated sample into a single RT-LAMP reaction 21 . Using serial dilutions of qPCR-positive samples, we demonstrated that glass milk purification improved the LoD by ten-fold (Fig. 4B) . We then re-tested 20 qPCR-positive and 20 qPCR-negative samples using glass milk inactivation. Unsurprisingly, qPCR-positive samples that were previously positive using inactivation alone were also positive after purification. Two of the four clinical samples that were falsely negative by inactivation alone were positive after the addition of a purification step, demonstrating the improved limit of detection using glass milk purification.

The two remaining specimens that tested negative after the purification step may have been truly qPCR-negative due to undergoing approximately three freeze-thaw cycles, but insufficient material remained for repeat testing. Thus, we may be underestimating the sensitivity of inactivation with A c c e p t e d M a n u s c r i p t 13 purification. All twenty qPCR-negative samples tested negative after purification (100% specificity ;   Fig. 4C ). Overall, purification improved the assay sensitivity by increasing the effective sample input volume.

Here we have demonstrated a simple and inexpensive loop-mediated isothermal amplification assay for the detection SARS-CoV-2 that achieves 87.5% overall sensitivity and 100% specificity compared to qPCR when performed directly from a clinical sample with the inclusion of an upfront, five-minute sample inactivation step. Performing an additional glass milk purification step resulted in increased assay sensitivity that was comparable to qPCR with an additional assay cost of only US$0.07 per sample. Laboratorians can determine whether the increased sensitivity afforded by glass milk purification justifies the additional labor and twenty-minute longer assay turn-aroundtime based on considerations such as their regional disease prevalence (Table S3 ). This assay can be performed in any clinical laboratory or even ad hoc settings, like a mobile laboratory, as it does not require any specialized equipment or highly trained laboratory personnel. Since the required reagents are easily manufactured by multiple manufacturers, access to this test does not rely on traditional commercial diagnostic supply chains that have hindered the broad distribution of SARS-CoV-2 testing. The manufacturer of colorimetric RT-LAMP master mix has large-scale production in place with millions of reactions worth of product available. We estimate an overall per-sample cost approximately US$6, though personnel and overhead costs will also contribute and vary greatly depending upon the setting. We believe this forty-minute sample-to-answer assay addresses a pressing need for COVID-19 diagnostics worldwide.

Sample preparation is often a time consuming and expensive step of the testing process. We have demonstrated that RT-LAMP can be performed directly from a nasopharyngeal sample but that the assay sensitivity increases by 30% with chemical RNase inactivation using TCEP/EDTA and heatmediated lysis. In addition to improving assay performance, the inactivation step described here A c c e p t e d M a n u s c r i p t 14 likely reduces the infectivity of the sample as well 24 , reducing the risk of exposure for laboratory personnel. The assay's sensitivity is further improved by glass milk purification, which is both extremely inexpensive compared to commercial RNA extraction kits (approximately US$5 per sample) and can be performed without a microcentrifuge, enabling its use in low-resource laboratory environments.

We foresee this assay being used in three ways. The first is a 40-minute rule-in test that uses inactivation followed by RT-LAMP. If the sample is positive and controls are valid, the test is reported as positive for SARS-CoV-2 to enable effective infection control practices and clinical management. If a rule-out test is desired, a sample that tests negative using inactivation alone can then be reflexed to a glass milk purification or a qPCR-based test. The third method tests pooled specimens using the inactivation and purification protocols for the rapid screening of large groups, i.e. in a school or employment setting, assuming adequate assay performance in this setting is demonstrated with additional clinical validation.

While this clinical validation is focused on nasopharyngeal swabs, which are recommended by the WHO and U.S. CDC as the most sensitive specimen type for SARS-CoV-2 detection 25, 26 , the same methods can be applied to other sample types as well, possibly including saliva 21 .

Oropharyngeal specimens are likely to be compatible given their similar composition to nasopharyngeal specimens. Sputum is generally a challenging sample type due to high viscosity and heterogeneity, but we expect the TCEP/EDTA chemical inactivation step to mimic the current recommendation to pre-treat sputum with dithiothreitol, an alternative reducing agent to TCEP 27 .

Further clinical studies to assess the range of compatible sample types are underway.

There are several limitations of this assay. The assay is qualitative and does not provide a semi-quantitative cycle threshold number. Additionally, the visual interpretation affords substantial flexibility, but it can also be prone to user errors. Objective color measurements can be performed by measuring the absorbance at 432 and 560 nm, or potentially with a smartphone application 28 . It A c c e p t e d M a n u s c r i p t 15 is critical to use the positive and negative controls as interpretative aids to avoid misinterpreting an orange intermediate color change as positive. The N gene primer set is more likely to give subtle background color changes and additional primer sets will be tested in the future. Like any nucleic acid amplification test, systems must be in place to avoid environmental and sample contamination with post-amplification products. One such precaution is to refrain from opening the reaction vessel after amplification; reactions should be discarded or transferred to a sealed container for later reference, since the color change remains stable for days to weeks. While the assay generally requires very little infrastructure, the operator must abide by laboratory biosafety guidelines and the procedure is most safely performed within a class II biosafety cabinet, although we recognize in many setting this may not be possible 29 . Additionally, the RT-LAMP master mix currently requires storage at -20 C, which is not ideal for low-resource or remote settings, but this may be ameliorated by lyophilization.

In summary, we present the implementation of a simple RT-LAMP assay for the detection of SARS-CoV-2 that achieves a high sensitivity and specificity in a challenging clinical sample set obtained during the peak of the Spring 2020 COVID-19 pandemic. Future work includes validating additional sample types, validating a specimen pooling approach, eliminating the cold chain requirement through reagent lyophilization, and assessing the feasibility and performance of the assay in resource-limited settings. A c c e p t e d M a n u s c r i p t 28 Figure 4 

Answer Platforms for the Detection of SARS-CoV-2

Comparison of Four Molecular In Vitro Diagnostic Assays for the Detection of SARS-CoV-2 in Nasopharyngeal Specimens

Rapid Tests for Influenza, Respiratory Syncytial Virus, and Other Respiratory Viruses: A Systematic Review and Meta-analysis. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America

Evaluation of a rapid diagnostic assay for detection of SARS CoV-2 antigen in nasopharyngeal swab

Accelerated reaction by loop-mediated isothermal amplification using loop primers. Molecular and cellular probes

DNA detection using recombination proteins

Rapid detection of SARS-CoV-2 by low volume real-time single tube reverse transcription recombinase polymerase amplification using an exo probe with an internally linked quencher (exo-IQ)

Visual detection of isothermal nucleic acid amplification using pH-sensitive dyes

Development and evaluation of a novel loop-mediated isothermal amplification method for rapid detection of severe acute respiratory syndrome coronavirus

Rapid colorimetric detection of Zika virus from serum and urine specimens by reverse transcription loop-mediated isothermal amplification (RT-LAMP). PloS one

World Health Organization. The use of loop-mediated isothermal amplification

Loop-mediated isothermal amplification (LAMP): An advanced molecular point-of-care technique for the detection of Leishmania infection. PLoS neglected tropical diseases

Rapid and visual detection of

CoV-2) by a reverse transcription loop-mediated isothermal amplification assay. Clinical microbiology and infection : the official publication of the European Society of Clinical Microbiology and Infectious Diseases

Development of a reverse transcription-loop-mediated isothermal amplification as a rapid early-detection method for novel SARS-CoV-2. Emerg Microbes Infect

Rapid detection of COVID-19 coronavirus using a reverse transcriptional loop-mediated isothermal amplification (RT-LAMP) diagnostic platform. Clinical chemistry

A colorimetric RT-LAMP assay and LAMPsequencing for detecting SARS-CoV-2 RNA in clinical samples. Science translational medicine

Fomsgaard AS, Rosenstierne MW. An alternative workflow for molecular detection of SARS-CoV-2 -escape from the NA extraction kit-shortage

SARS-CoV-2 detection using isothermal amplification and a rapid, inexpensive protocol for sample inactivation and purification

Field-deployable viral diagnostics using CRISPR-Cas13. Science

25. World Health Organization. Laboratory testing for coronavirus diseas in suspected human cases: interim guidance

Centers for Disease Control and Prevention. Processing of Sputum Specimens for Nucleic Acid Extraction

A c c e p t e d M a n u s c r i p t 19 28 . Sun A c c e p t e d M a n u s c r i p t