(259 days)
-
Enzyme-linked immunosorbent assay (ELISA) for the qualitative or semi-quantitative detection of β2-GPI IgA antibodies in human serum to aid in diagnosis of autoimmune thrombotic disorders associated with antiphospholipid syndrome (APS) and APS with systemic lupus erythematosus (SLE) in conjunction with other laboratory tests and clinical findings.
-
Enzyme-linked immunosorbent assay (ELISA) for the qualitative or semi-quantitative detection of β2-GPI IgG antibodies in human serum to aid in diagnosis of autoimmune thrombotic disorders associated with antiphospholipid syndrome (APS) and APS with systemic lupus erythematosus (SLE) in conjunction with other laboratory tests and clinical findings.
-
Enzyme-linked immunosorbent assay (ELISA) for the qualitative or semi-quantitative detection of β2-GPI IgM antibodies in human serum to aid in diagnosis of autoimmune thrombotic disorders associated with antiphospholipid syndrome (APS) and APS with systemic lupus erythematosus (SLE) in conjunction with other laboratory tests and clinical findings.
-
Enzyme-linked immunosorbent assay (ELISA) for the qualitative detection of β2-GPI IgA, IgG and IgM antibodies in human serum to aid in diagnosis of autoimmune thrombotic disorders associated with antiphospholipid syndrome (APS) and APS with systemic lupus erythematosus (SLE) in conjunction with other laboratory tests and clinical findings.
The test is performed as a solid phase immunoassay (ELISA) in B2GP1 coated microwells. Controls, Calibrators and patient serum samples are incubated in the antigen coated microwells to allow antibodies present in the serum to bind. Unbound antibody and other serum proteins are removed by washing the microwells are detected by adding an enzyme labeled anti-human lgA, lgG, lgM or lgA/IgG/IgM conjugate to the microwells. These enzyme conjugated antibodies bind specifically to the human immunoglobulin of the apropriate class. Unbound enzyme-labeled conjugate is removed by washing. Specific enzyme substrate (TMB) is then added to the presence of antibodies is detected by a color change produced by the conversion of the TMB substrate. The reaction is stopped and the intensity of color change, which is proportional to the concentration of antibody, is read by a spectrophotometer at 450 nm. Results are expressed in ELISA units per milliliter (EU/ml).
Here's a breakdown of the acceptance criteria and study information for the Immulisa Enhanced™ B2GP1 Antibody ELISA devices, based on the provided text:
Acceptance Criteria and Reported Device Performance
The acceptance criteria are generally implied by the reported performance in the method comparison study, where the device's agreement with a predicate device is evaluated. The FDA often evaluates substantial equivalence based on such comparisons. Specific pre-defined thresholds for agreement (e.g., >90% overall agreement) are likely part of the internal acceptance criteria, though not explicitly stated as "acceptance criteria" here.
| Performance Metric | Acceptance Criteria (Implied) | ImmuLisa Enhanced™ B2GP1 IgA Antibody ELISA (Reported) | ImmuLisa Enhanced™ B2GP1 IgG Antibody ELISA (Reported) | ImmuLisa Enhanced™ B2GP1 IgM Antibody ELISA (Reported) | ImmuLisa Enhanced™ B2GP1 IgA/IgG/IgM Antibody ELISA (Reported) |
|---|---|---|---|---|---|
| Method Comparison | |||||
| Positive % Agreement | High agreement with predicate device (likely >70-80%) | 74.8% (95% CI 65.7% - 82.2%) | 83.1% (95% CI 72.5% - 90.4%) | 91.7% (95% CI 84.8% - 95.7%) | 80.2% (95% CI 73.4% - 85.7%) |
| Negative % Agreement | High agreement with predicate device (likely >90%) | 99.1% (95% CI 96.3% - 99.8%) | 92.0% (95% CI 87.8% - 94.8%) | 99.1% (95% CI 96.7% - 99.9%) | 96.8% (95% CI 92.2% - 98.8%) |
| Overall % Agreement | High overall agreement with predicate device (likely >85-90%) | 90.5% (95% CI 86.7% - 93.4%) | 89.9% (95% CI 86.2% - 92.8%) | 96.6% (95% CI 94.3% - 98.4%) | 87.9% (95% CI 83.8% - 91.1%) |
| Cross-Reactivity | Low % Positive in various autoimmune/infectious conditions | - (Detailed in tables for each condition) | - (Detailed in tables for each condition) | - (Detailed in tables for each condition) | - (Detailed in tables for each condition) |
| Interference | No significant interference from common substances | No significant interference demonstrated | No significant interference demonstrated | No significant interference demonstrated | No significant interference demonstrated |
| Precision | Low imprecision (various CV% for within-run, between-day, operator) | Range of CV% from 3.5% to 7.0% | (Same study, assumed similar findings per assay) | (Same study, assumed similar findings per assay) | (Same study, assumed similar findings per assay) |
| Reproducibility | High qualitative agreement, especially near cutoff and for clear pos/neg samples | For cutoff samples: 76.7% Pos, 61.1% Neg for Ize, JgG, IgM; 96.7% Neg for 10% specimen; 100% for other specimens | (Same study, assumed similar findings per assay) | (Same study, assumed similar findings per assay) | For 120% below cutoff: 97.8% Neg; for cutoff: 50% Neg; 100% for other specimens |
| Limit of Detection | Low LoD values to detect low levels of antibodies | N/A | 3.6 EU/ml | 2.3 EU/ml | 2.7 EU/ml |
| Linearity and Recovery | Representative dilution results with acceptable slope, intercept, R2, and % recovery | Acceptable ranges reported for all | Acceptable ranges reported for all | Acceptable ranges reported for all | N/A (Only individual assays tested for linearity) |
| Hook Effect | No hook effect up to high concentrations | No hook effect up to 605.5 EU/ml (IgM) / 228.6 EU/ml (IgG) | No hook effect up to 605.5 EU/ml (IgM) / 228.6 EU/ml (IgG) | No hook effect up to 605.5 EU/ml (IgM) / 228.6 EU/ml (IgG) | N/A |
| Clinical Sensitivity (APS) | Adequate sensitivity for diagnosis of APS | 38.3% (30.0-47.3%) | 58.8% (48.3-68.5%) | 64.1% (55.7-71.8%) | 84.8% (76.1-90.8%) |
| Clinical Specificity (APS) | High specificity for diagnosis of APS | 92.0% (89.3-94.1%) | 93.8% (91.3-95.7%) | 96.1% (93.9-97.5%) | 91.9% (89.1-94.0%) |
Study Details
-
Sample size used for the test set and the data provenance:
- Method Comparison (Test Set for Equivalence):
- IgA: 328 samples (115 Pos, 213 Neg by predicate)
- IgG: 338 samples (77 Pos, 261 Neg by predicate)
- IgM: 355 samples (120 Pos, 235 Neg by predicate)
- IgA/IgG/IgM: 331 samples (177 Any Pos, 154 All Neg by predicate)
- Cross-Reactivity (Test Set):
- Various sample sizes depending on the condition, ranging from 7 to 60 for individual Ig classes, and 7 to 45 for the combined IgA/IgG/IgM assay.
- Clinical Performance (Test Set):
- The exact total number of clinical samples for APS, APS with SLE, and SLE is not explicitly stated, but the sensitivity and specificity are provided, implying a sufficient number were tested. The percentages are accompanied by 95% Confidence Intervals.
- Data Provenance: Not explicitly stated (e.g., country of origin). The studies appear to be retrospective as they involve testing existing clinical samples and disease controls.
- Method Comparison (Test Set for Equivalence):
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- The document implies that the ground truth for the method comparison study was established using a predicate device (INOVA QUANTA Lite™ ß2 GPI IgA, IgG, IgM and Screen ELISA).
- For the clinical performance studies, the ground truth for "APS," "APS with SLE," and "SLE" diagnoses would typically be established by clinical diagnostic criteria, likely involving an expert consensus based on patient history, other laboratory tests, and clinical findings, but the number and qualifications of experts are not specified in this document.
-
Adjudication method (e.g., 2+1, 3+1, none) for the test set:
- For the method comparison, the predicate device served as the reference standard, so no human adjudication method (like 2+1) is indicated for the test set results.
- For clinical performance, the ground truth regarding the disease status (APS, SLE, etc.) would be based on established diagnostic criteria, implying a form of clinical consensus and data, but a specific "adjudication method" involving multiple readers for interpreting the device's results is not mentioned.
-
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, an MRMC comparative effectiveness study was not done. This device is an in-vitro diagnostic (ELISA) kit, not an AI-assisted diagnostic tool that requires human interpretation of images or complex data where AI might "assist" a human reader. Therefore, this question is not applicable to this type of device.
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- Yes, standalone performance was done. The device itself is the standalone "algorithm" (the ELISA assay) that produces a numerical result (EU/ml) or a qualitative (positive/negative) call. The performance metrics (method comparison, sensitivity, specificity, precision, etc.) all represent the standalone performance of the device.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- Method Comparison: The ground truth was based on the results of a legally marketed predicate device (INOVA QUANTA Lite™ ß2 GPI IgA, IgG, IgM and Screen ELISA).
- Clinical Performance: The ground truth for disease classifications (APS, APS with SLE, SLE) is implied to be based on clinical diagnostic criteria, which typically involves a combination of clinical findings, patient history, and other laboratory tests (similar to expert consensus/outcomes data). The document states "Sensitivity/specificity exclude healthy human blood on the reference laboratory testing with a possible diagnosis of APS or SLE," indicating samples were classified based on their clinical status.
-
The sample size for the training set:
- The document does not explicitly state a separate "training set" size for algorithm development. For in-vitro diagnostic assays like ELISAs, assay development and analytical validation typically involve iterative testing with various samples, but these are generally referred to as optimization or development samples, not a distinct "training set" in the context of machine learning. The studies presented (method comparison, clinical performance, etc.) are essentially validation studies, testing the final, developed assay.
-
How the ground truth for the training set was established:
- As no explicit "training set" is mentioned in the context of algorithm development, similar to machine learning models, this question is not directly applicable. The assay's parameters (e.g., cutoffs, analytical ranges) would have been established during product development using samples whose characteristics (e.g., positive, negative, various concentrations) were known, likely through reference methods or clinical diagnosis, but this is a standard part of ELISA development rather than "ground truth establishment for a training set" as understood in AI/ML.
§ 866.5660 Multiple autoantibodies immunological test system.
(a)
Identification. A multiple autoantibodies immunological test system is a device that consists of the reagents used to measure by immunochemical techniques the autoantibodies (antibodies produced against the body's own tissues) in serum and other body fluids. Measurement of multiple autoantibodies aids in the diagnosis of autoimmune disorders (disease produced when the body's own tissues are injured by autoantibodies).(b)
Classification. Class II (performance standards).