(444 days)
The ARCHITECT Toxo IgG assay is a chemiluminescent microparticle immunoassay (CMIA) used for the quantitative determination of IgG antibodies to Toxoplasma gondii in human serum and plasma on the ARCHITECT i System.
The ARCHITECT Toxo IgG assay is to be used as an aid in the detection of immune status to Toxoplasma gondii in individuals, including women of child-bearing age, and as an aid in the diagnosis of Toxoplasma gondii infection.
Not intended for use in screening blood, plasma, or tissue donors.
The ARCHITECT Toxo IgG reagent kit contains:
- Microparticles: Recombinant Toxoplasma gondii antigen coated microparticles in MES buffer with protein (bovine). Minimum concentration: 0.03% solids. Preservative: ProClin 300.
- Conjugate: Murine acridinium-labeled anti-human IgG in MES buffer with protein (bovine) stabilizer. Minimum concentration: 0.05 µg/mL. Preservatives: antimicrobial agents.
- Assay Diluent: TRIS buffer with protein (murine) and protein (bovine). Preservative: ProClin 300.
The ARCHITECT Toxo IgG Calibrators:
- Calibrator A - an aqueous solution with protein (bovine) stabilizer. Preservative: ProClin 300.
- Calibrators B through F - contain anti-Toxo IgG in an aqueous solution with protein (bovine) stabilizer. Preservative: ProClin 300.
The ARCHITECT Toxo IgG Controls:
- Negative Control - contains recalcified human plasma with protein (ovine) stabilizer. Preservatives: ProClin 950 and sodium azide.
- Positive Control 1 - contains anti-Toxo IgG in aqueous solution with protein (bovine) stabilizer. Preservative: ProClin 300.
This assay is a two-step immunoassay for the quantitative determination of IgG antibodies to Toxoplasma gondii in human serum and plasma using chemiluminescent microparticle immunoassay (CMIA) technology.
The provided text describes the Abbott ARCHITECT Toxo IgG assay, a chemiluminescent microparticle immunoassay for the quantitative determination of IgG antibodies to Toxoplasma gondii. This document is a 510(k) premarket notification summary, which means it aims to demonstrate substantial equivalence to a legally marketed predicate device, not necessarily to prove its own absolute effectiveness against a clinical ground truth like disease outcome or pathology.
Therefore, the acceptance criteria are primarily demonstration of analytical performance that is substantially equivalent to the predicate device and supports the stated indications for use. The "study that proves the device meets the acceptance criteria" refers to the nonclinical and clinical laboratory studies presented in the 510(k) summary that support this substantial equivalence.
Here's an analysis of the requested information based on the provided text:
1. A table of acceptance criteria and the reported device performance
The document doesn't explicitly list "acceptance criteria" as a separate table. Instead, it presents performance data from various studies. The implicit acceptance criteria are the successful demonstration of performance characteristics typically required for an in vitro diagnostic device to be found substantially equivalent to a predicate.
Below is a table summarizing key performance metrics and the reported results. The "Acceptance Criteria" column reflects the generally expected performance for such assays or thresholds implicitly met by the reported data, rather than specific numerical targets stated in the document.
| Performance Metric (Implicit Acceptance Criteria) | Reported Device Performance (ARCHITECT Toxo IgG) |
|---|---|
| Within-Laboratory Precision (Low %CV, repeatability) | - Negative Control: Mean 0.0 IU/mL, SD 0.01 IU/mL (N/A %CV) |
- Positive Control 1: Mean 6.4 IU/mL, SD 0.16 IU/mL (2.5 %CV)
- Other Panels: %CVs for Panels 3, 4, 5 were 3.8%, 2.8%, 2.5% respectively. |
| Lower Limits of Measurement (LoB, LoD, LLoQ within acceptable ranges) | - LoB: 0.1 IU/mL - LoD: 0.2 IU/mL
- LLoQ: 0.2 IU/mL (at 20% CV) |
| Linearity (Assay measures concentrations accurately across a range) | Linear across 0.2 to 75.0 IU/mL. |
| Analytical Specificity / Interference (No significant interference from common substances or cross-reacting conditions) | - No significant interference observed from listed endogenous substances (e.g., Bilirubin, Hemoglobin, Triglycerides) and drugs (e.g., Ascorbic Acid, Atovaquone). - Zero reactive results (0/164) from specimens with various potentially interfering medical conditions (e.g., other viral infections, autoimmune markers, rheumatoid factor). |
| CDC Panel Agreement (High agreement with a well-characterized reference panel) | - Sensitivity (Positive Agreement): 97% (68/70 positive specimens detected as reactive) - Specificity (Negative Agreement): 100% (30/30 negative specimens detected as nonreactive) |
| System Reproducibility (Multi-site) (Consistent performance across different sites/operators) | - Negative Control: Mean 0.0 IU/mL, SD 0.02 IU/mL (N/A %CV) - Positive Control 1: Mean 6.4 IU/mL, SD 0.22 IU/mL (3.5 %CV)
- Other Panels: Reproducibility %CVs for Panels 3, 4, 5 were 5.6%, 5.9%, 6.7% respectively. |
| Percent Agreement with Predicate Device (High agreement with a legally marketed comparator) | - Routine Order Samples:- Negative % Agreement: 98.54% (1082/1098)
- Positive % Agreement: 94.87% (148/156)
- Preselected Positive Samples:
- Negative % Agreement: 100.00% (1/1)
- Positive % Agreement: 96.73% (148/153)
- Pregnant Females Samples:
- Negative % Agreement: 100.00% (186/186)
- Positive % Agreement: 93.33% (14/15) |
2. Sample size used for the test set and the data provenance (e.g., country of origin of the data, retrospective or prospective)
- Precision and Reproducibility Studies:
- Within-Laboratory Precision: 120 replicates per panel/control (e.g., Positive Control 1 N=120, Panel 3 N=120). Data provenance implicitly laboratory-based from Abbott.
- System Reproducibility: 360 replicates per control/panel (e.g., Positive Control 1 N=360). Conducted at 3 clinical sites (US based: New York City and Farmingdale, New York, and Palo Alto, California).
- Lower Limits of Measurement: N ≥ 60 replicates per analyte level. Data provenance implicitly laboratory-based from Abbott.
- Analytical Specificity:
- Interfering Endogenous Substances, Drugs, and Other Substances: Not specified exact number of replicates per interferent, but tested at 2 analyte levels.
- Potentially Interfering Other Conditions: 164 specimens.
- CDC Panel Agreement: 100 specimens (70 true positive, 30 true negative) from the CDC Toxoplasma 1998 Human Serum Panel (implies retrospective, pre-characterized samples).
- Percent Agreement (Method Comparison Clinical Study):
- Total N = 1414 specimens.
- Routine Order: 777 specimens (US) and 482 specimens (outside US). Total 1259.
- Preselected Positive: 84 specimens (US) and 71 specimens (outside US). Total 155.
- Pregnant Females: 200 specimens (US).
- Data Provenance: The document explicitly states the clinical study was performed in the US (at 3 clinical testing sites located in New York City and Farmingdale, New York and Palo Alto, California) and with samples collected in the US and outside of the US. The study is described as a "clinical study," implying prospective collection and testing, but the use of "preselected positive" specimens indicates some retrospective sample selection.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g., radiologist with 10 years of experience)
The document does not specify the number or qualifications of experts used to establish the ground truth for the test sets.
- For the CDC Panel, the "gold standard" was the Dye Test. This test itself is highly specialized and would have been performed by experts at the CDC, but their specific qualifications are not detailed here. The panel itself is a "masked, characterized serum panel," implying its ground truth was established prior to this study.
- For the Percent Agreement (Method Comparison) study, the "ground truth" was established by a "current FDA cleared commercially available anti-Toxo IgG assay" (the comparator device). For discordant samples from this study, the Dye Test was used as a reference. Again, the experts performing these comparator or Dye Tests are not detailed.
- For the Precision, Linearity, and Analytical Specificity studies, the samples are either manufactured/spiked or are pre-screened based on other assays or clinical categories. The ground truth here is analytical rather than clinical, and is established by the methods of sample preparation and characterization typically done by laboratory scientists.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
The document does not describe a formal reader (e.g., radiologist) adjudication method (like 2+1 or 3+1) because this is an in vitro diagnostic (IVD) assay, not an image-based AI device. The "reading" or determination of results is done by the automated ARCHITECT i System based on chemiluminescent reactions, not human interpretation of complex images.
For discordant results in the method comparison study, the Dye Test was used as an arbitration method for clarification. For example, in the routine order comparisons, 24 discordant samples were tested using the Dye Test.
5. 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
No MRMC comparative effectiveness study was done. This type of study (MRMC) is relevant for diagnostic imaging systems, particularly those that incorporate Artificial Intelligence (AI) to assist human readers (e.g., radiologists interpreting medical images).
The ARCHITECT Toxo IgG assay is an in vitro diagnostic immunoassay. Its "reader" is the ARCHITECT i System, an automated instrument. There is no human "reader" in the loop interpreting results in the same way a radiologist interprets an image. Therefore, questions about human reader improvement with AI assistance are not applicable to this device.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
Yes, the performance of the ARCHITECT Toxo IgG assay is presented as standalone (algorithm/instrument only) performance. The ARCHITECT i System automatically processes samples and yields quantitative results (IU/mL) and qualitative interpretations (reactive, grayzone/equivocal, nonreactive). There is no "human-in-the-loop" affecting the primary measurement or interpretation by the device itself after it has been loaded.
The various precision, linearity, analytical specificity, and agreement studies (including the CDC panel agreement) demonstrate the standalone performance of the device.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
The ground truth for evaluating this IVD device employed a combination of methods:
- Well-characterized Reference Panels: Specifically, the CDC Toxoplasma 1998 Human Serum Panel, which was characterized using the Dye Test (a highly regarded serological reference method for Toxoplasma IgG).
- Comparison to a Legally Marketed Predicate/Comparator Device: For the method comparison studies, the results of the ARCHITECT Toxo IgG assay were compared against a "current FDA cleared commercially available anti-Toxo IgG assay." This serves as a practical ground truth for demonstrating substantial equivalence within regulatory frameworks for IVDs.
- Arbitration with Reference Method: For discordant results in the method comparison study, the Dye Test was used as an arbitration method to resolve discrepancies and provide a more definitive "ground truth" for those specific samples.
- Analytical Ground Truth: For studies like linearity, precision, LoB/LoD/LLoQ, and analytical specificity, the "ground truth" is established by precisely prepared samples (e.g., known concentrations, spiked interferents, samples verified to be negative).
8. The sample size for the training set
The document describes premarket studies for a commercial IVD kit, not an AI/ML algorithm development. Therefore, there is no explicit "training set" in the context of machine learning model development.
The studies described are for validation and verification of the device's performance characteristics. For an IVD assay like this, the "training" aspect would relate to the assay's biochemical development, reagent formulation, and calibration curve generation, which are usually proprietary development processes and not "data sets" in the AI/ML sense.
9. How the ground truth for the training set was established
As there is no explicit "training set" for an AI/ML model described for this IVD assay, this question is not directly applicable.
The development and internal optimization of an immunoassay typically involve extensive laboratory work where known standards and characterized samples are used to optimize reagent concentrations, reaction conditions, and calibration algorithms. The "ground truth" during such development would be based on reference methods, gravimetric/volumetric preparations, and established scientific principles of immunology and chemistry.
§ 866.3780
Toxoplasma gondii serological reagents.(a)
Identification. Toxoplasma gondii serological reagents are devices that consist of antigens and antisera used in serological tests to identify antibodies toToxoplasma gondii in serum. Additionally, some of these reagents consist of antisera conjugated with a fluorescent dye (immunofluorescent reagents) used to identifyToxoplasma gondii from clinical specimens. The identification aids in the diagnosis of toxoplasmosis caused by the parasitic protozoanToxoplasma gondii and provides epidemiological information on this disease. Congenital toxoplasmosis is characterized by lesions of the central nervous system, which if undetected and untreated may lead to brain defects, blindness, and death of an unborn fetus. The disease is characterized in children by inflammation of the brain and spinal cord.(b)
Classification. Class II (performance standards).