The SensiTox B. anthracis Toxin Test for use with the MultiPath Analyzer, is a qualitative immunofluorescence assay to aid in the diagnosis of inhalation anthrax. The test is intended for the rapid, qualitative detection of lethal factor, a biomarker associated with Bacillus anthracis). The test can be used with whole blood collected with dipotassium EDTA anticoagulant by venipuncture. This testing samples from individuals who have signs and symptoms consistent with inhalation anthrax and a likelihood of exposure to B. anthracis. A positive SensiTox B. anthracis Toxin Test result is presumptively diagnostic for B. anthracis infection. Diagnosis of B. anthracis infection must be made in conjunction with medical history, likelihood of exposure, signs, and symptoms of disease, as well as other laboratory evidence. The definitive identification of B. anthracis from blood samples additional testing and confirmation procedures in consultation with public health or other authorities for whom reports are required. Testing should be performed and reported in accordance with current guidelines provided by the appropriate public health authorities. The level of lethal factor present in blood from individuals with early systemic infection is unknown. Negative results do not preclude infection with the biothreat microbial agents targeted by the device and should not be used as the sole basis for diagnosis, treatment, or other patient management decisions.

Laboratories performing the SensiTox B. anthracis Toxin Test must have the appropriate biosafety equipment, personal protective equipment (PPE), containment facilities, and personnel trained in the safe handling of clinical specimens potentially containing B. anthracis.

The SensiTox B. anthracis Toxin Test is for prescription use only.

This assay is not FDA-cleared or approved for testing blood or plasma donors.

The SensiTox B. anthracis Toxin Test, run on the MultiPath Analyzer, detects lethal factor in venous whole blood samples using an immunofluorescence assay and the proprietary MultiPath detection technology.

A whole blood specimen, collected in dipotassium EDTA, from individuals with signs and symptoms consistent with inhalation anthrax and a likelihood of exposure, is used for the test. The blood sample is added directly to the SensiTox B. anthracis Cartridge, a single use consumable that contains all the reagents required to run the test. The Cartridge is loaded onto the MultiPath Analyzer for processing through the steps of the assay.

Once loaded onto the Analyzer, the barcodes on the cartridge that identify the test type and associated test specific information (manufacturer installed barcode) and sample (laboratory affixed barcode) are read. The cartridge is moved to the fluidics station where it is first heated to 35°C. The sample is then split into 3 equal aliquots in 3 distribution wells within the cartridge. The sample aliquots flow from the distribution wells to the reagent wells containing target-specific antibody conjugated fluorescent and magnetic particles in the form of lyophilized beads. Upon contact with the sample, the lyophilized beads rehydrate and the reaction mixtures flow into the imaging wells, the bottoms of which are coated with a dye cushion reagent. Upon contact with the reagents, the dye-cushion dissolves forming a dense opaque aqueous layer that separates the sample and reagents from the bottom optical surface of the Imaging Well. In the upper assay layer, the toxins, if present, bind to the magnetic and fluorescent particles tethering them together. The cartridge is incubated for 12 minutes to allow the reaction to take place and then is moved to the magnetics station. At the magnetics station, the imaging well is placed over permanent magnets that draw the magnetic particles and any fluorescent particles that are tethered to them via the target molecules through the dye-cushion layer, depositing them on the bottom imaging surface. The captured fluorescent particles are imaged and quantified using nonmagnified digital imaging.

The Analyzer can be run in batch mode or by random access. Up to 20 cartridges can be loaded onto the Analyzer in parallel. The first result is reported in approximately 21 minutes of loading the cartridge onto the Analyzer with subsequent results being reported in 2.5-minute increments. The results are interpreted using the MultiPath applications software as valid or invalid, and if valid, the results are reported as Lethal Factor detected. Results are displayed on the instrument touch screen and can be printed.

Here's a summary of the SensiTox B. anthracis Toxin Test's acceptance criteria and the study that proves the device meets them, based on the provided FDA 510(k) summary:

Acceptance Criteria and Reported Device Performance

The acceptance criteria are not explicitly stated as distinct pass/fail thresholds in the provided document, but rather implied by the design and results of the performed studies. The reported performance metrics from the non-clinical and clinical tests are:

Study Details

1. Sample sizes used for the test set and the data provenance:

   • Limit of Detection (LoD):
     • Test set: Minimum of 5 serial dilutions, tested in replicates of 20 each.
     • Data provenance: In-house generated data using recombinant Lethal Factor or native Lethal Factor from B. anthracis culture supernatants spiked into whole blood from healthy donors.
   • Reproducibility:
     • Test set: 270 samples (90 per site, 3 replicates per sample for each operator). Samples were whole blood spiked with low positive (1.5X LoD), moderate positive (3X LoD) recombinant LF, and negative (unspiked).
     • Data provenance: In-house generated data using spiked whole blood from healthy donors.
   • Analytical Reactivity (Inclusivity):
     • Test set: 29 B. anthracis strains. Supernatants spiked into whole blood at 3X LoD, tested in triplicate.
     • Data provenance: In-house generated data using B. anthracis strains.
   • Analytical Specificity (Cross-Reactivity):
     • Test set: Over 100 bacterial, viral, protozoan, and fungal pathogens (see Table 5.5 in the document). Each tested in triplicate by addition to pre-screened K₂EDTA human venous whole blood.
     • Data provenance: In-house generated data using specified pathogens.
   • Analytical Specificity (Interference):
     • Test set: Over 100 bacterial, viral, fungal, and protozoan pathogens (see Table 5.6 in the document). Each tested in triplicate by addition to pre-screened K₂EDTA human venous whole blood containing 3X LoD recombinant LF.
     • Data provenance: In-house generated data using specified pathogens spiked into blood.
   • Interfering Substances:
     • Test set: 8 endogenous and 33 exogenous substances (see Table 5.7 in the document). Each tested in triplicate in unspiked and 3X LoD spiked K₂EDTA whole blood.
     • Data provenance: In-house generated data using specified interfering substances.
   • Hook Effect:
     • Test set: 10 dilutions of recombinant LF from 50 pg/mL to 50 µg/mL. Three replicates per dilution.
     • Data provenance: In-house generated data using recombinant LF spiked into whole blood.
   • Clinical Performance Evaluation:
     • Negative Percent Agreement (NPA) - Residual specimens: 318 residual whole blood specimens (from a single clinical site).
     • Positive Percent Agreement (PPA) and NPA - Prospective specimens: 105 prospectively collected whole blood specimens from febrile patients (contingent on being eligible after initial 21 withdrawals). These were used for both unspiked and spiked testing. Contrived specimens (prepared from these prospective samples, spiked with native or recombinant LF at 1.5X LoD, 5X LoD, or 5 µg/mL) were also tested blinded at a second site.
     • Data provenance: Clinical sites in the US (residual and prospective).

2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

   • For the non-clinical studies (LoD, Reproducibility, Analytical Reactivity, Specificity, Interference, Hook Effect), the ground truth was based on the precisely controlled spiking of known concentrations of Lethal Factor, specific B. anthracis strains, various pathogens, or interfering substances. No human expert consensus was used for these.
   • For the Clinical Performance Evaluation, the ground truth for NPA in residual specimens was that they were presumed negative for B. anthracis Lethal Factor as they were collected for routine hematology analysis. For the prospective study, the ground truth was established by spiking the collected blood with known concentrations of Lethal Factor for PPA calculation. The intrinsic "negative" status of the febrile patients for naturally occurring anthrax Lethal Factor served as the ground truth for NPA. The document does not specify the qualifications of personnel involved in sample collection or initial characterization, but it implies standard clinical laboratory practices.

3. Adjudication method for the test set: Not applicable for this device as it's an in vitro diagnostic testing for a biomarker, not image-based or human-interpreted data that would require adjudication. The device provides a qualitative "Lethal Factor detected" or "Lethal Factor not detected" result.

4. If a multi-reader, multi-case (MRMC) comparative effectiveness study was done, if so, what was the effect size of how much human readers improve with AI vs without AI assistance: Not applicable. This is an automated immunofluorescence assay for a biomarker, not an AI-assisted diagnostic device requiring human interpretation of results.

5. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done: Yes, the SensiTox B. anthracis Toxin Test, run on the MultiPath Analyzer, is a standalone device. The results are "interpreted using the MultiPath applications software as valid or invalid, and if valid, the results are reported as Lethal Factor detected." There is no explicit human-in-the-loop performance described beyond the loading of the cartridge and reading of the digital display.

6. The type of ground truth used:

   • Non-clinical studies: Laboratory-controlled ground truth established by known concentrations of recombinant or native Lethal Factor, quantified culture supernatants, and characterized pathogen strains.
   • Clinical Performance Evaluation:
     • Residual specimens (NPA): Presumed negative as they were "residual whole blood specimens collected in K₂EDTA for routine hematology analysis."
     • Prospective specimens (PPA): Artificially created by spiking prospectively collected patient blood with known concentrations of Lethal Factor.
     • Prospective specimens (NPA): Based on the absence of spiked Lethal Factor and the enrollment criteria (febrile patients without confirmed anthrax, ensuring presumed negativity for the target biomarker).

7. The sample size for the training set: Not explicitly stated as a separate "training set" in the context of machine learning. For this in vitro diagnostic device, the developmental and validation process involves extensive analytical studies (LoD, cross-reactivity, interference, etc.) using numerous spiked samples and characterized strains. The "training" in this context refers to the development and optimization of the assay and software algorithms during product development, which would likely have involved an internal set of samples, but no specific training set size is provided in this summary.

8. How the ground truth for the training set was established: As above, the ground truth for the development and optimization of the assay and software would have been established by precisely controlling the input (e.g., specific concentrations of LF, specific pathogen strains) in a laboratory setting. This is a common approach for in vitro diagnostic device development.

Ask a specific question about this device