Fig 1.
Near point-of-care and high-throughput sample-to-answer workflows.
(A-F) Near point-of-care workflow. (A) Equipment and consumables needed to run the RT-LAMP assay, including saliva collection and lysis tubes, a heat block, pipettes, mineral oil, and a benchtop fluorimeter. A nasal swab and lysis tube (inset) can be substituted for saliva collection materials. To perform the test, (B) either a nasal swab or saliva sample is collected and combined with lysis buffer, then (C) the sample is heated at 95° C for 5 (nasal) or 6 (saliva) minutes. (D) The lysed sample is then added to RT-LAMP reagents and (E) reactions are sealed and placed in a benchtop fluorimeter and heated at 65° C for 45 minutes with real-time fluorescence monitoring. (F) Results are displayed after the full incubation period. (G-M) High-throughput saliva workflow. (G) Consumables and pipettes needed to run the RT-LAMP assay, excluding Hamilton robotics consumables. To perform the test, (H) lysis buffer and (I) saliva are combined and (J) heated at 95° C for 6 minutes. (K) RT-LAMP master mix is prepared and added to a 96-well plate. (L) Inactivated samples are added to RT-LAMP master mix by a Hamilton MicroLab Prep. The plate is sealed and (M) placed onto the Bio-Rad CFX96 for amplification and detection. Results are analyzed within the Bio-Rad Maestro software.
Fig 2.
Lysis buffer evaluation and optimization.
Four lysis buffer concentrations (1X, 2X, 5X, and 10X) with and without 200 mM GuHCl were evaluated. Candidate buffers were combined with negative nasal swab eluate, spiked with Zeptometrix pseudovirus and heated at 95° C for 5 minutes. (A) Average Ct values resulting from lysed samples amplified in the CDC N1 RT-qPCR assay following column purification (± standard deviation; n = 6 for each condition). Significance determined using a two-way ANOVA with post-hoc Tukey’s test. (B) Average relative fluorescence values from lysed samples at the 1-hour timepoint of an RNase Alert assay (± standard deviation; n = 3 for each condition). Significance determined using two-way ANOVA. **** indicates p<0.0001, ** indicates p<0.01, * indicates p<0.05.
Fig 3.
Limit of detection in negative pooled matrix.
The limit of detection was assessed by spiking Zeptometrix pseudovirus into pooled nasopharyngeal matrix collected in lysis buffer that had previously tested negative for SARS-CoV-2 in RT-qPCR using the CDC assay, or negative pooled saliva combined with lysis buffer. Samples were heat-inactivated at 95° C for 5 minutes (A,B) or 6 minutes (C,D), and tested on either the Bio-Rad CFX96 or Axxin T8. The limit of detection at 95% probability was determined using probit analysis to be: (A) 20 copies per reaction on the Bio-Rad CFX96 and (B) 23 copies per reaction on the Axxin T8 for nasopharyngeal samples. The limit of detection in saliva was found to be: (C) 93 copies per reaction on the Bio-Rad CFX96 and (D) 116 copies per reaction on the Axxin T8.
Fig 4.
Testing and analysis flowchart by sample type.
Flowchart shows the number of participants who were enrolled and confirmed to have SARS-CoV-2 infections, and the corresponding number of samples by type that were collected and analyzed by RT-qPCR and by RT-LAMP.
Fig 5.
Ct distribution by sample type.
The frequency distribution of Ct values obtained by RT-qPCR for (A) nasopharyngeal swabs, (B) nasal swabs, and (C) saliva samples from patients who provided all three sample types (n = 41). The Ct values of corresponding nasopharyngeal, nasal, and saliva samples from the same patients are plotted and stratified by (D) nasal swab Ct<30, (E) nasal swab 30<Ct<35 (F) nasal swab Ct>35 and (G) negative nasal swab.
Fig 6.
RT-LAMP performance on clinical samples.
(A,B) Summary of RT-LAMP results for nasopharyngeal samples, stratified by the Ct value obtained in RT-qPCR. Nasopharyngeal samples were tested on (A) the Bio-Rad CFX96 and (B) the Axxin T8. (C) Positive agreement and 95% confidence intervals with RT-qPCR at varying Ct thresholds. (D,E) Summary of RT-LAMP results for nasal samples compared to the nasal sample Ct value obtained in RT-qPCR, when tested on (D) the Bio-Rad CFX96 and (E) the Axxin T8. (F) Positive and negative agreement with RT-qPCR and 95% confidence intervals for nasal samples. (G,H) Summary of RT-LAMP results for saliva samples compared to the saliva sample Ct value obtained in RT-qPCR, when tested on (G) the Bio-Rad CFX96 and (H) the Axxin T8. (I) Positive and negative agreement with RT-qPCR and 95% confidence intervals for saliva samples.
Fig 7.
Nasal Ct distributions in surveillance population and hospitalized patient population.
(A) Mean, standard deviation, and range of Ct values and (B) frequency distributions of Ct values for confirmatory nasal RT-qPCR tests after an individual tested positive with LAMP (n = 8), for individuals who tested positive as part of Rice University’s surveillance testing program with the testing provider who conducted confirmatory RT-qPCR tests (n = 136), and for all hospitalized patients who tested positive by nasal RT-qPCR (n = 35).