(275 days)
The DiaSorin LIAISON® FT4 assay is an in vitro chemiluminescent immunoassay intended for the quantitative determination of free thyroxine (FT4) in human serum using the LIAISON® Analyzer. It is intended for use as an aid in the clinical assessment of thyroid status.
The DiaSorin LIAISON® Control Thyroid 1, LIAISON® Control Thyroid 2 and LIAISON® Control Thyroid 3 are intended for use as assayed quality control samples to monitor the accuracy and precision of the DiaSorin LIAISON® FT4 assay.
The DiaSorin LIAISON® FT4 assay's method for quantitative determination of FT4 is based on the Solid Phase Antigen Linked Technique (SPALT) principle. A T4-protein-conjugate is coated on the magnetic particles (solid phase); a monoclonal antibody is linked to an isoluminol derivative (isoluminol-antibody conjugate). All assay steps and incubations are performed by the LIAISON® Analyzer.
Results are determined by a 2 point calibration conversion of the master curve to a working curve. The light signal, and hence the amount of isoluminol-antibody conjugate, is measured by a photomultiplier as relative light units (RLU) and is inversely related to FT4 concentration present in calibrators, patient samples or controls.
The DiaSorin LIAISON® FT4 assay is an in vitro chemiluminescent immunoassay intended for the quantitative determination of free thyroxine (FT4) in human serum using the LIAISON® Analyzer. It is intended for use as an aid in the clinical assessment of thyroid status.
The DiaSorin LIAISON® Control Thyroid 1, LIAISON® Control Thyroid 2 and LIAISON® Control Thyroid 3 are intended for use as assayed quality control samples to monitor the accuracy and precision of the DiaSorin LIAISON® FT4 assay.
Here's an analysis of the provided text regarding acceptance criteria and the study proving the device meets them:
1. Table of Acceptance Criteria and Reported Device Performance
The documents do not explicitly list formal "acceptance criteria". However, performance data is presented, which implicitly demonstrates the device's acceptable performance. For clarity, I've inferred common performance metrics for this type of device and presented the reported results.
| Metric (Implicit Acceptance Criteria) | Reported Device Performance (LIAISON® FT4) |
|---|---|
| Method Comparison (Against Predicate) | |
| Linearity (Slope) | 1.0707 (95% CI: 1.03 to 1.09) |
| Linearity (Intercept) | -0.1211 ng/dL (95% CI: -0.17 to -0.08 ng/dL) |
| Correlation Coefficient (R²) | 0.9633 |
| Reference Range | |
| 95% Normal Range (Serum) | 0.69 - 1.59 ng/dL (8.93 - 20.4 pmol/L) |
| Reproducibility/Precision (20-Day Study) | |
| Kit Control 1 (Mean 1.02 ng/dL) - Total %CV | 7.9% |
| Kit Control 2 (Mean 3.18 ng/dL) - Total %CV | 6.9% |
| Kit Control 3 (Mean 6.59 ng/dL) - Total %CV | 7.0% |
| Sample 1 (Mean 1.19 ng/dL) - Total %CV | 8.0% |
| Sample 2 (Mean 1.90 ng/dL) - Total %CV | 7.5% |
| Sample 3 (Mean 5.56 ng/dL) - Total %CV | 7.1% |
| Sample 4 (Mean 5.72 ng/dL) - Total %CV | 7.1% |
| Reproducibility/Precision (5-Day Study - Near Medical Decision Points) | |
| Sample 1 (Hypo, Mean 0.851 ng/dL) - Total %CV | 9.5% |
| Sample 2 (Normal, Mean 1.57 ng/dL) - Total %CV | 8.6% |
| Sample 3 (Hyper, Mean 3.25 ng/dL) - Total %CV | 7.5% |
| Dilution Linearity | Linear across the measuring range (0.29 - 7.7 ng/dL) |
| Limit of Blank (LoB) | ≤ 0.06 ng/dL |
| Limit of Detection (LoD) | 0.13 ng/dL |
| Limit of Quantitation (LoQ) | 0.29 ng/dL |
| Interfering Substances | No interference observed at tested concentrations |
| Specificity (Cross-Reactivity) | D-T4: 100%, 3-iodo-L-Tyrosine: 0%, 3,5-diiodo-L-Tyrosine: 0%, 3,3',5,5'-tetra-iodothyroacetic acid: 0% |
2. Sample Sizes Used for the Test Set and Data Provenance
- Method Comparison: 201 samples.
- Reference Range Determination: 130 apparently healthy adult human serum samples.
- Reproducibility/Precision (20-Day Study): A coded panel of 4 frozen serum samples prepared by DiaSorin, plus 3 levels of LIAISON® Control Thyroid. Each sample/control tested 80 times (2 replicates/run, 2 runs/day, 20 days).
- Reproducibility/Precision (5-Day Study): A panel of three frozen serum samples prepared by DiaSorin with FT4 concentrations near medical decision points. Each sample tested 60 times (12 replicates/run, 1 run/day, 5 days).
- Dilution Linearity: One spiked sample, diluted into ten evenly spaced intervals, tested in four replicates for each dilution.
- LoB/LoD/LoQ: Not explicitly stated, but typically involves multiple replicates of blank and low-concentration samples.
- Interfering Substances: Not explicitly detailed, but involved testing various potential interferents.
- Specificity (Cross-Reactivity): Not explicitly detailed regarding sample count, but involves specific cross-reactants.
Data Provenance:
The studies appear to be prospective, performed internally by DiaSorin Inc. (e.g., "A twenty day reproducibility/precision study was performed at DiaSorin Inc.", "An additional five day precision study was performed at DiaSorin"). The "country of origin" is implied to be the US, given the submission to the FDA by DiaSorin Inc. in Stillwater, MN.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
For an in vitro diagnostic (IVD) device like a Free Thyroxine Assay, the "ground truth" is typically established by reference methods or established laboratory practices, not by expert interpretation in the same way an imaging device might use radiologists.
- Method Comparison: The comparison was made against a "commercially available immunoassay," which serves as the reference or "ground truth" for comparative performance. The specific details of how its "ground truth" was established are not provided, but it is implied to be a legally marketed, accepted method (the Roche Elecsys® FT4, K961489, is named as the predicate device).
- Reference Range: Established by testing samples from "130 apparently healthy adults." These samples themselves represent the "ground truth" of a healthy population.
- Reproducibility/Precision, Dilution Linearity, LoB/LoD/LoQ, Interfering Substances, Specificity: These studies rely on scientifically controlled preparations (e.g., spiked samples, controls, known concentrations) and established assay principles, rather than expert consensus on patient data.
Therefore, the concept of "number of experts" and their "qualifications" for ground truth establishment, as it might apply to image-based diagnostics, is not directly applicable here. The "ground truth" for the test set is either a comparative, legally marketed device or samples with analytically defined characteristics.
4. Adjudication Method for the Test Set
Not applicable. As described above, the "ground truth" for this type of IVD device is based on analytical measurements against a comparative method or defined standards, not on subjective interpretations requiring adjudication among experts.
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
Not applicable. This is an in vitro diagnostic (IVD) assay, not an AI-powered diagnostic imaging device that involves human readers. Therefore, an MRMC study to assess human reader improvement with AI assistance is not relevant to this submission.
6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done
This device is a standalone algorithm, in the sense that it is a fully automated immunoassay system that quantitatively determines FT4 concentration without human interpretative input on the result itself (beyond clinical interpretation of the quantitative value). The performance data presented (Method Comparison, Precision, Linearity, etc.) represents the standalone performance of the LIAISON® FT4 assay. There is no human "in the loop" for the measurement process; the LIAISON® Analyzer performs all assay steps and incubates.
7. The Type of Ground Truth Used
The ground truth used for various performance evaluations includes:
- Comparative Device: For method comparison, the Roche Elecsys® FT4 Test (K961489) served as the reference.
- Clinically Defined Samples: For reference range determination, samples from "apparently healthy adults" were used to establish expected values.
- Analytically Defined Standards/Samples: For precision, linearity, LoB/LoD/LoQ, interfering substances, and specificity, the ground truth was established by:
- Commercial Controls: LIAISON® Control Thyroid (1, 2, 3), whose target values are established through rigorous internal procedures involving multiple analyzers and kit lots.
- Internal Panels/Spiked Samples: Frozen serum samples prepared by DiaSorin, or buffer samples spiked with known concentrations of T4 and other substances.
8. The Sample Size for the Training Set
The document describes performance studies, not a machine learning model. IVD assays like the LIAISON® FT4 are based on well-established chemical and immunological principles, not on "training sets" in the context of AI. The "training" in this context refers to the development and optimization of the assay itself and its reagents.
However, if we interpret "training set" more broadly as data used to develop and optimize the assay, rather than for final validation:
- Calibrator Value Assignment: Master standards are prepared and concentrations are assigned by testing in "10 different assays against the Primary Reference Standard." Kit calibrator values are then assigned using "multiple LIAISON® analyzers with several kits calibrators over several runs." This is part of the internal development and calibration process.
- Control Value Assignment: A minimum of "60 valid test results for each control" are used to establish the target value and range, using "a minimum of 3 LIAISON® Analyzers, using 2 different approved LIAISON® FT4 assay kit lots, at a minimal time period of 3 days." This data essentially "trains" the acceptable control ranges.
9. How the Ground Truth for the Training Set was Established
Again, this is not an AI model, so "training set ground truth" is interpreted in the context of IVD assay development.
- Calibrator "Ground Truth": The calibrators are referenced to an "in-house Primary Reference Standard preparation." This primary standard itself would be meticulously characterized and traceable to internationally recognized standards if available, or rigorously defined internally. Concentrations are assigned via testing against this standard.
- Control "Ground Truth": The target values for the LIAISON® Controls (Thyroid 1, 2, 3) are established by calculating the mean from a minimum of 60 valid test results, spanning multiple analyzers, kit lots, and days. The "ground truth" for control values is therefore derived empirically from the performance of the LIAISON® FT4 assay itself, establishing an expected range for quality control.
§ 862.1695 Free thyroxine test system.
(a)
Identification. A free thyroxine test system is a device intended to measure free (not protein bound) thyroxine (thyroid hormone) in serum or plasma. Levels of free thyroxine in plasma are thought to reflect the amount of thyroxine hormone available to the cells and may therefore determine the clinical metabolic status of thyroxine. Measurements obtained by this device are used in the diagnosis and treatment of thyroid diseases.(b)
Classification. Class II.