Portable molecular diagnostic platform for rapid point-of-care detection of mpox and other diseases

Abstract

The World Health Organization's designation of mpox as a public health emergency of international concern in August 2024 underscores the urgent need for effective diagnostic solutions to combat this escalating threat. The rapid global spread of clade II mpox, coupled with the sustained human-to-human transmission of the more virulent clade I mpox in the Democratic Republic of Congo, highlights a critical gap in point-of-care diagnostics for this emergent disease. In response, we developed Dragonfly, a portable molecular diagnostic platform for point-of-care use that integrates power-free nucleic acid extraction (<5 minutes) with lyophilised colourimetric LAMP chemistry. The platform demonstrated an analytical limit-of-detection of 100 genome copies per reaction for monkeypox virus, effectively distinguishing it from other orthopoxviruses, herpes simplex virus, and varicella-zoster virus. Clinical validation on 164 samples, including 51 mpox-positive cases, yielded 96.1% sensitivity and 100% specificity for orthopoxviruses, and 94.1% sensitivity and 100% specificity for monkeypox virus. Here, we present a rapid, accessible, and robust point-of-care diagnostic solution for mpox, suitable for both low- and high-resource settings, addressing the global resurgence of orthopoxviruses in the context of declining smallpox immunity.

Fig. 1. The Dragonfly platform.

a Overview of the platform, including consumable components, optional tablet and companion app, as well as deployed disposable mobile workstation, and low-cost isothermal heat block with reusable fixed-volume pipette. b High-level overview of rapid nucleic acid extraction process and test panel loading for one sample. c High-level overview of test panel incubation and colourimetric result interpretation. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/d45n457.

Fig. 2. The SmartLid nucleic acid extraction technology.

a Graphical illustration of SmartLid usage to transfer magnetic beads from one tube to another with a removable magnet. b Close-up images showing the rapid magnetic bead collection process, with beads visible on the underside of the SmartLid after collection. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/d45n457.

Fig. 3. The Dragonfly sample-to-result workflow.

a Sample input using a disposable exact-volume pipette. b Insertion of the SmartLid to initiate the rapid power-free extraction process. c Close-up of extracted nucleic acids being loaded into the open test panel using a 20 µL fixed-volume pipette. d Simultaneous incubation of two test panels, with a capacity of up to four panels per low-cost, portable isothermal heat block. e Fully incubated colourimetric test panel being removed from the heat block. The attached panel tag allows for easy removal, reduces the risk of accidental tube opening, and verifies correct panel selection through a comparison with the scanned tag at the beginning of the process. f Example of a test panel result indicating a positive (yellow) reaction for OPXV, along with valid test controls. Screenshots from the companion application, showing the test panel loading screen (g), incubated test panel result capture screen (h), result confirmation screen (i), and result interpretation and recording screen (j). Created in BioRender. Cavuto, M. (2025) https://BioRender.com/d45n457.

Fig. 4. Validation of the Dragonfly Skin Infection Viral Test Panel.

a Evaluation of virucidal activity of eNAT® against vaccinia virus (VACV): Plaque assays were performed using confluent monolayers of BSC40 cells. The bar plot shows the virucidal activity after a 2-minute exposure, showing 8-log reduction, alongside images of crystal violet-stained plaque assays (1% crystal violet, 70% ethanol) after 30 minutes. b Analytical sensitivity of LAMP assays: range of detected concentrations of synthetic DNA (log10 of copies per reaction) and corresponding TTP values in minutes. c Analytical specificity of LAMP assays: Using extracted nucleic acids from various commercially available viral particles (MXPV, HSV-1/HSV-2, and VZV), and a control sample consisting of synthetic DNA at 5 × 10³ copies per reaction. The bar plot displays mean TTP values and data points (in minutes) with error bars representing the standard deviation (SD). d qPCR C_t values distribution: A histogram illustrating the distribution of qPCR C_t values obtained from OPXV and MPXV-positive clinical samples (purple dots for OPXV and yellow dashed lines for MPXV). Confusion matrices of the diagnostic performance are included below. e Sample-to-result demonstration: Images showing the results for vaccinia (VACV), cowpox (CPXV), MPXV, and combined HSV-1/HSV-2/VZV viral particles, with spiking concentrations indicated below each set of reactions. VACV and CPXV viral particles, both spiked at 5 × 10⁵ PFU/mL, demonstrate the specificity of the OPXV assay, showing no cross-reactivity with other target assays. Similarly, spiking MPXV (2.5 × 10³ copies/mL) and combined HSV-1/HSV-2/VZV viral particles (2.5 × 10³ copies/mL for VZV and HSV-1, 3.5 × 10³ copies/mL for HSV-2) showed no cross-reactivity. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/d45n457.

Fig. 5. Dragonfly OPXV/MPXV Multi-patient.

a Overview of the complete mobile testing kit, reconfigured for the simultaneous extraction and detection of OPXV and MPXV from two patient samples. b Dual-sample vortex tool enabling the simultaneous vortex mixing of two samples. c, d Magnetic bead collection performed concurrently for both samples. e Sequential loading of the test panel; the second half of the panel is loaded only after the tubes in the first half are sealed to minimise the risk of cross-contamination. f Example of a valid negative result for both patients, along with the corresponding dual-patient result card. Detailed instructions for use can be found in the Supplementary Methods. Created in BioRender. Cavuto, M. (2025) https://BioRender.com/d45n457.