(582 days)
The Vysis CLL FISH Probe Kit is intended to detect deletion of the LS1 TP53, LSI ATM, and LSI D13S319 probe targets and gain of the D12Z3 sequence in peripheral blood specimens from untreated patients with B-cell chronic lymphocytic leukemia (CLL). The assay may be used to dichotomize CLL (the 13q-, +12, or normal genotype group versus the 11q- or 17p- group) and may be used as an aid in determining disease prognosis in combination with additional biomarkers, morphology, and other clinical information. The Vysis CLL FISH Probe Kit is not intended for use in selection of therapy or in monitoring of residual disease.
The Vysis CLL FISH Probe Kit uses fluorescence in situ hybridization (FISH) DNA probe technology to determine deletions of the locus-specific identifier (LSI) TP53, LSI ATM, and LSI D13S319 probe targets and gain of the D12Z3 sequence.
The Vysis CLL FISH Probe Kit (List No. 4N02-020) consists of two DNA FISH probe sets and three general purpose reagents sufficient to process 20 assays.
- . LSI TP53 SpectrumOrange/ATM SpectrumGreen Probe
- LSI D13S319 SpectrumOrange/13q34 SpectrumAqua/CEP 12 SpectrumGreen Probe .
- DAPI II Counterstain .
- NP-40 .
- 20X SSC Salt .
Here's a summary of the acceptance criteria and the study details for the Vysis CLL FISH Probe Kit, based on the provided 510(k) summary:
Acceptance Criteria and Device Performance
Note: The 510(k) summary for the Vysis CLL FISH Probe Kit primarily focuses on analytical performance (specificity, sensitivity, normal cut-off values, precision, reproducibility) and concordance with an existing clinical assay rather than traditional "acceptance criteria" related to a new AI/CADe device. The provided tables outline the device's technical performance.
| Acceptance Criterion / Performance Metric | Reported Device Performance (Mean/Overall) |
|---|---|
| Analytical Specificity | 100% for all probes (LSI TP53, LSI ATM, LSI D13S319, LSI 13q34, CEP 12) |
| Analytical Sensitivity (Expected normal signal pattern: 2 signals/nucleus) | LSI TP53: 97.98% |
| LSI ATM: 98.68% | |
| LSI D13S319: 98.60% | |
| CEP 12: 98.94% | |
| Normal Cut-off Values (Percentage of abnormal nuclear FISH patterns) | LSI TP53 (1 signal): 7.0% (14/200 nuclei) |
| LSI ATM (1 signal): 6.0% (12/200 nuclei) | |
| LSI D13S319 (1 signal): 5.5% (11/200 nuclei) | |
| LSI D13S319 (0 signal): 1.5% (3/200 nuclei) | |
| CEP 12 (3 signals): 2.5% (5/200 nuclei) | |
| Precision (Mean % abnormal cells for various probes/abnormalities) | Ranges from 0.0% to 73.2% across different samples and probes (SDs provided in Tables 4-8) |
| Reproducibility (Overall Agreement, Site to Site by Probe) | TP53 (17p-): 100% |
| ATM (11q-): 90% | |
| CEP 12 (+12): 100% | |
| D13S319 1x (13q-): 90% | |
| D13S319 2x (13q-): 90% | |
| Reproducibility (Prognostic Category, Generalized Kappa) | Kappa = 0.86 (Strength: "Almost Perfect") |
| Method Concordance (AMT vs. RFT for Prognostic Category) | Overall Agreement: 97% (62/64) |
| Lower bound one-sided 95% CI: 90% |
Study Details
-
Sample sizes used for the test set and the data provenance:
- Analytical Specificity: 5 karyotypically normal male peripheral blood cultures (retrospective, assumed domestic, not explicitly stated).
- Analytical Sensitivity: 25 karyotypically normal patient peripheral blood cultures (retrospective, assumed domestic, not explicitly stated).
- Analytical Characterization of Normal Cut-off Values: 25 karyotypically normal patient peripheral blood cultures (retrospective, assumed domestic, not explicitly stated).
- Precision:
- Precision Study 1: 2 negative peripheral blood specimens (lacking abnormalities) and 8 additional specimens with at least one abnormality. Likely retrospective, but not explicitly stated.
- Precision Study 2: 8 different patient specimens (blinded panel). Likely retrospective, but not explicitly stated.
- Reproducibility: A blinded 20-member slide panel representing five Döhner classifications. Likely retrospective, but not explicitly stated.
- Method Concordance: 64 specimens with pre-established Döhner classifications. Retrospective, as classifications were "based on previous results using the RFT." Provenance not explicitly stated.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Analytical Specificity: One technologist. Qualifications not specified beyond being a "technologist."
- Analytical Sensitivity: Not explicitly stated, but each technologist evaluated 100 nuclei per specimen, implying multiple technologists were involved. Qualifications not specified.
- Analytical Characterization of Normal Cut-off Values: Not explicitly stated, but implies multiple technologists based on the sensitivity study. Qualifications not specified.
- Precision: Not specified, but involved counting signals, implying trained technologists.
- Reproducibility: Not specified, but involved three different laboratories, implying multiple trained personnel.
- Method Concordance: The "Reference FISH Test (RFT)" was used to establish initial Döhner classifications. The original Shanafelt study (referred to as the RFT) implies multiple pathologists/cytogeneticists. This study used "previous results" from the RFT as its ground truth.
-
Adjudication method for the test set:
- No formal adjudication method (like 2+1 or 3+1) is explicitly described for establishing the ground truth or validating results in the analytical studies. Results were typically enumerated by technologists and aggregated.
- For the Reproducibility study, "overall agreement" between three testing sites was assessed, but a specific adjudication process for discrepancies is not detailed beyond reporting disagreement counts.
- For Method Concordance, the ground truth was the "Reference FISH Test (RFT) used in the Shanafelt study," which effectively served as the gold standard, and the AMT results were compared against it.
-
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 involving human readers improving with AI vs. without AI assistance was reported. This device is a FISH probe kit, not an AI-powered diagnostic system. The studies focused on the analytical and clinical validity of the probe kit itself.
-
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
- This is not an algorithm-only device. It is a laboratory assay (FISH probe kit) where human technologists perform the analysis by counting fluorescent signals under a microscope. The studies performed were of the laboratory kit's performance, not an automated algorithm.
-
The type of ground truth used:
- Analytical Specificity, Sensitivity, Normal Cut-off, Precision: Based on visual identification and enumeration of FISH signals by trained technologists, comparing observed signal patterns to expected normal or abnormal patterns in karyotypically normal and patient samples. The "expected normal" pattern serves as a form of expert-derived ground truth.
- Reproducibility: Comparison of results across multiple labs/readers for the same samples. The "agreement" between sites implies a consensus or majority rule for comparison, but the ultimate ground truth for a given sample's true Döhner classification is not explicitly detailed but likely derived from expert pathological/cytogenetic review.
- Method Concordance: The ground truth was established by a "Reference FISH Test (RFT) used in the Shanafelt study," which is stated to have its clinical validity documented via peer-reviewed literature. This implies a ground truth based on established clinical and pathological diagnosis as determined by a validated method.
-
The sample size for the training set:
- This device is a diagnostic kit, not an AI/machine learning algorithm, so there is no "training set" in the conventional sense of AI development. The studies described are for analytical validation and clinical concordance.
-
How the ground truth for the training set was established:
- Since there's no AI "training set," this question is not applicable. The device's performance characteristics (specificity, sensitivity, cut-offs) are established through testing on defined sets of samples (e.g., karyotypically normal individuals, patients with confirmed CLL aberrations) where the expected outcome is known or determined by expert review using established cytogenetic methods.
§ 866.6040 Gene expression profiling test system for breast cancer prognosis.
(a)
Identification. A gene expression profiling test system for breast cancer prognosis is a device that measures the ribonucleic acid (RNA) expression level of multiple genes and combines this information to yield a signature (pattern or classifier or index) to aid in prognosis of previously diagnosed breast cancer.(b)
Classification. Class II (special controls). The special control is FDA's guidance document entitled “Class II Special Controls Guidance Document: Gene Expression Profiling Test System for Breast Cancer Prognosis.” See § 866.1(e) for the availability of this guidance document.