(309 days)
The AmplideX® Fragile X Dx & Carrier Screen Kit is an in vitro diagnostic device that uses polymerase chain reaction (PCR) and capillary electrophoresis to detect and identify the number of cytosine-guanine (CGG) repeats in the fragile X mental retardation-1 (FMR1) gene using genomic DNA isolated from peripheral whole blood specimens. It is solely intended as an aid in the post-natal diagnosis of fragile X syndrome, and fragile Xassociated disorders [i.e., fragile X-associated tremor/ataxia syndrome (FXTAS) or fragile Xassociated primary ovarian insufficiency (FXPOI)], and for carrier testing in adults of reproductive age. Assay results are solely intended to be interpreted by healthcare professionals who are board certified in molecular genetics and to be used in conjunction. with other clinical and diagnostic findings, consistent with professional standards of practice. Reflex testing, clinical genetic evaluation, and genetic counseling should be offered as appropriate. The test is for use on the 3500Dx Series Genetic Analyzer.
This test is not indicated for use for fetal diagnostic testing, newborn screening or for standalone diagnostic purposes.
The AmplideX Fragile X Dx & Carrier Screen Kit (hereafter referred to as the AmplideX Kit) includes reagents sufficient for 100 reactions and are for use on the Applied Biosystems® 3500 Dx Series Genetic Analyzer (8 and 24 capillary) with AmplideX reporter software.
The AmplideX® Fragile X Dx & Carrier Screen Kit is an in vitro diagnostic device used to detect and identify the number of cytosine-guanine (CGG) repeats in the FMR1 gene, which is associated with fragile X syndrome and related disorders. The device uses PCR and capillary electrophoresis.
Here's a breakdown of the acceptance criteria and the studies proving the device's performance:
1. Acceptance Criteria and Reported Device Performance
The device's performance was evaluated against specific criteria for genotype categorization and allele size precision.
| Criterion | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Precision (CGG Repeat Length) | - 1-70 CGG repeats: ± 1 |
- 71-120 CGG repeats: ± 3
- 121-199 CGG repeats: ± 5%
- ≥ 200 CGG repeats: N/A (reported as >200) | Three-site Reproducibility:
- For 10 out of 11 samples, >97% genotype agreement was observed for categorical calls. Sample ASGN-018 (Full Mutation mosaic) showed 54.7% agreement due to the mosaic allele not being found in 81/179 replicates, prompting a limitation statement in IFU.
- Reproducibility of expected allele lengths (excluding mosaic alleles) showed >98% within target precision range.
Lot-to-Lot Reproducibility:
-
99% categorical genotype agreement was observed for all samples (excluding invalids). Samples ASGN-112 and ASGN-016 were not 100% due to third alleles in a few replicates.
- Reproducibility of all alleles (excluding mosaic alleles) showed 100% within target precision specifications.
Mosaicism Reproducibility/LoD:
- Greater than 95% hit rate demonstrated for claimed LoD for mosaic alleles (Table 9 shows various MAFs from 2.0% to 10.0% achieved 94.4% to 100% detection). |
| Diagnostic Performance (Full Mutation) | PPA, NPA, and OPA ≥ 95% | Full Mutation vs. Southern Blot: - PPA: 95.71% (95% CI: 88.1-98.5)
- NPA: 99.30% (95% CI: 96.0-99.9)
- OPA: 97.6% (95% CI: 94.5-99.0)
- One discordant sample (194-repeat AmplideX, >200 Southern Blot) was borderline. |
| Diagnostic Performance (Premutation) | PPA, NPA, and OPA ≥ 95% | Premutation vs. Normal or Intermediate (Southern Blot): - PPA: 100% (95% CI: 94.7-100.0)
- NPA: 97.10% (95% CI: 90.0-99.2)
- OPA: 98.6% (95% CI: 94.9-99.6) |
| Carrier Screening Performance | Overall agreement ≥ 95% for Full Mutation/Premutation, Intermediate, or Normal genotypes. | AmplideX vs. FMR1 Dual-PCR Reference Method: - Premutation/Full Mutation: 100% (95% CI: 94.7-100.0)
- Intermediate: 85.7% (95% CI: 75.7-92.1)
- Normal: 98.6% (95% CI: 92.2-99.7)
- Note: When considering assay precision, 8 borderline samples (54/55 CGG) were reclassified as concordant, increasing Intermediate agreement to 97.1%. |
| LoD for mosaic alleles | > 95% hit rate for claimed LoD | - Intermediate on Normal: 2.0% MAF, 100.0% detected - Premutation on Normal: 2.0% MAF, 94.4% detected
- Full Mutation on Normal: 6.1% MAF, 100.0% detected
- Normal on Premutation: 2.0% MAF, 100.0% detected
- Premutation on Premutation: 5.0% MAF, 97.1% (low), 100.0% (high) detected
- Full Mutation on Premutation: 7.0% MAF, 100.0% detected
- Normal on Full Mutation: 2.0% MAF, 100.0% detected
- Premutation on Full Mutation: 10.0% MAF, 100.0% detected |
| DNA Input Range (20-80 ng) | 100% categorical genotype agreement for valid samples, all expected allele peaks within target precision range, and acceptable QC failure rates within the recommended range. | - 100% categorical genotype agreement and all expected allele peaks within target precision for all valid samples and all input levels. - QC failure rates: 1 ng (6.25%), 10 ng (1.88%), 20 ng (0.6%), 40 ng (1.3%), 80 ng (0%), 160 ng (0.63%). These results support the recommended 20-80 ng range. |
| DNA Extraction Methods | 100% categorical genotype agreement and 100% of all measured alleles within target precision range for common commercial DNA extraction methods. | - 100% categorical genotype agreement and 100% of all measured alleles within target precision range for the three common commercial DNA extraction methods tested (solution precipitation, manual silica spin column, and manual magnetic bead). |
| Thermal Cycler Equivalence | 100% genotype category calls agreement and 100% of all identified independent alleles within their respective target precision ranges across all thermal cycler types. | - 100% genotype category calls agreement and 100% of all identified independent alleles within their respective target precision ranges across three separate thermal cyclers. |
| Interfering Substances (Hemoglobin, EDTA) | For processable samples, >95% genotype category agreement and 100% of alleles within expected precision range. (Limitation statement regarding impact on assay failure rate for excess hemoglobin/EDTA is included in IFU). | - Effects on DNA concentration yield observed (hemoglobin reduced, EDTA enhanced). - For samples that could be processed, genotype category agreement was >95% and 100% of alleles within expected precision range. (Limitation statement for IFU). |
| Specimen Handling Stability | 100% genotype category agreement and all expected allele peaks within target precision for DNA extracted from whole human blood stored at 2-8°C for up to 2 weeks. | - 100% genotype category agreement and all expected allele peaks within target precision for all timepoints across all samples (up to 2 weeks storage at 2-8°C). |
| DNA Freeze/Thaw Stability | 100% genotype category agreement and all expected allele peaks within target precision for extracted DNA across (b)(4) freeze/thaw cycles. | - 100% genotype category agreement and all expected allele peaks within target precision for up to (b)(4) freeze/thaw cycles. |
| PCR & CE Reagent Product Stability | PCR Product: 100% categorical genotype agreement and >95% of measured alleles within target precision for PCR products stored up to 3 days at 2-8°C.
CE Product: 100% categorical genotype agreement and 100% of measured alleles within target precision for samples prepared for CE analysis and stable at ambient temperature for up to 24 hours. | - PCR Product: 100% categorical genotype agreement for all timepoints across all samples and >95% of all measured alleles within target precision range (up to 3 days at 2-8°C). - CE Product: 100% categorical genotype agreement for all timepoints across all samples and 100% of all measured alleles within target precision range (up to 24 hours at ambient temperature). |
| Real Time Kit Stability | 100% categorical genotype agreement and 100% of measured alleles within target precision. (Supports a minimum of 1 year, with ongoing study for 24 months). | - 100% categorical genotype agreement for all timepoints across all samples and 100% of all measured alleles within target precision range. (Supports at least 1-year storage stability at present). |
| Freeze/Thaw Kit Stability | 100% categorical genotype agreement and 100% of measured alleles within target precision for up to eight freeze/thaw cycles. | - 100% categorical genotype agreement for all timepoints across all samples and 100% of all measured alleles within target precision range (up to eight freeze/thaw cycles). |
| Kit Shipping Stability | 100% of measured alleles within target precision range and 100% genotype category agreement across different shipping configurations. | - 100% of their measured alleles within the target precision range and 100% genotype category agreement across all different shipping configurations (based on ISTA 7D, 24-hour summer profile). |
| WHO Standards Testing Accuracy | Correctly classifies and sizes international Fragile X reference standards. | - The test correctly classified and sized the 5 WHO International Standard Fragile X Syndrome Reference Panel samples (NIBSC code: 08/158) with 100% of measured values falling within the expected CGG range, even when reported as ">200" for full mutations. |
2. Sample Sizes and Data Provenance (Test Set)
- Precision (Reproducibility) Studies:
- Three-site reproducibility: 11 samples (7 clinical whole blood specimens, 4 contrived from cell lines spiked into leukocyte-depleted whole human blood). This yielded 1980 total sample measurements (excluding controls).
- Lot-to-lot reproducibility: 11 samples (same panel as above). This yielded 2376 total sample measurements (excluding controls).
- Mosaicism reproducibility/LOD: Panel of mixed cell line DNA with known genotypes and a panel of 30 samples involving mixing DNA from clinical specimens. This yielded 36 replicate measurements per specimen.
- DNA Input Study: 8 samples (6 clinical, 2 contrived). This resulted in 960 sample measurements (20 replicates per dilution level per sample).
- Clinical Performance Studies:
- Diagnostic Performance: 207 leftover clinical specimens (111 female, 96 male) obtained from patient samples submitted for routine FMR1 5'UTR mutation testing across multiple clinical sites.
- Carrier Screening Performance: (b)(4) specimens from females ((b)(4) years of age) enrolled across (b)(4) clinical sites, resulting in 207 evaluable subjects.
- Mosaic Allele Calling Accuracy: 49 specimens identified as mosaic by AmplideX across the two clinical studies (27 from diagnostic, 22 from carrier screening). An additional 40 randomly selected samples were further evaluated for mosaicism resolution after initial discordant cases were addressed.
- WHO Standards Testing: 5 samples from the WHO International Standard Fragile X Syndrome Reference Panel (NIBSC code: 08/158). This yielded 135 sample measurements (excluding control).
Data Provenance: The data primarily appears to be retrospective for the clinical performance studies, using "leftover specimens" and "patient samples submitted for routine FMR1 5'UTR mutation testing" or "enrolled across clinical sites." The analytical performance studies used a mix of clinical specimens and contrived samples (cell lines spiked into blood). The document does not explicitly state the country of origin for the clinical data.
3. Number of Experts and Qualifications (Ground Truth for Test Set)
The document does not explicitly state the number of experts or their specific qualifications for establishing the ground truth for the clinical test sets (Southern Blot and FMR1 Dual-PCR Reference Method). However, it mentions that the interpretation of assay results is "solely intended to be interpreted by healthcare professionals who are board certified in molecular genetics."
4. Adjudication Method (Test Set)
The document does not describe a formal adjudication method for the test set.
- In the diagnostic performance study, results were compared directly against the reference methods (Southern Blot or FMR1 Dual-PCR). Discordant samples were noted, and in one case for full mutation, a borderline sample was identified.
- For mosaicism, there was a manual evaluation of orthogonal PCR method results for cases discordant with AmplideX.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
There is no mention of a multi-reader multi-case (MRMC) comparative effectiveness study to assess the effect size of human readers improving with AI vs. without AI assistance. This device is a diagnostic kit for laboratory use, and its performance is evaluated against reference methods, not typically in a human-in-the-loop context.
6. Standalone Performance Study (Algorithm Only)
Yes, the studies described are for standalone performance (algorithm only, as applied to the output of the PCR and capillary electrophoresis). The AmplideX Fragile X Reporter Software performs automated analysis and categorization of FMR1 alleles based on size. The performance is assessed by comparing the device's output (categorical calls and CGG repeat lengths) directly against established reference methods (Southern Blot, FMR1 Dual-PCR) or known values (WHO standards, contrived samples).
7. Type of Ground Truth Used
- Clinical Performance (Diagnostic):
- Southern Blot: For Full Mutation classification and Premutation vs. Normal or Intermediate assessment. Southern Blot is a molecular biology method considered a gold standard for detecting large FMR1 expansions.
- Clinical Performance (Carrier Screening):
- FMR1 Dual-PCR Test: An alternate orthogonal independently validated PCR assay was used as the comparator for carrier screening, as Southern blot is noted to not have accurate sizing in the premutation range.
- Analytical Performance (Precision, LoD):
- Clinical Samples: Whole blood specimens from patients with known FMR1 genotypes.
- Contrived Samples: Genomic DNA extracted from cell lines with known gender, genotype, and expected allele sizes, sometimes spiked into leukocyte-depleted whole human blood.
- Mixed Cell Line DNA: With known genotypes and mosaic allele frequencies for LoD studies.
- WHO Standards Testing:
- WHO International Standard Fragile X Syndrome Reference Panel (NIBSC code: 08/158): These are internationally established reference materials with characterized mean repeat lengths and ranges.
8. Sample Size for the Training Set
The document does not provide a specific sample size for a "training set." This device is a diagnostic assay with software for interpretation, but the context here suggests a traditional verification and validation study rather than a machine learning model where a distinct training set (for model parameters) and test set (for performance evaluation) would be explicitly separated and reported. The analytical and clinical studies serve as the validation of the device's performance.
9. How the Ground Truth for the Training Set Was Established
As mentioned above, there isn't an explicit "training set" in the context of a machine learning-based device. The ground truth for the reference materials and clinical samples used in the validation studies was established through:
- Reference Methods: Southern Blot analysis and an independently validated FMR1 Dual-PCR assay are considered the ground truth for clinical performance.
- Known Characteristics: For contrived samples and cell lines, the genotypes and allele sizes are "known" based on prior characterization.
- International Standards: The WHO/NIBSC standards have established and published characterizations from a study consortium.
§ 866.5970 Inherited nucleotide repeat disorder DNA test.
(a)
Identification. An inherited nucleotide repeat disorder DNA test is a prescription in vitro diagnostic device that is intended to detect and identify the number of nucleotide repeats in a gene using genomic DNA isolated from post-natal patient specimens. It is solely intended as an aid for carrier testing and as an aid for the diagnosis of inherited nucleotide repeat-associated disorders. Assay results are solely intended to be used in conjunction with other clinical and diagnostic findings. These tests do not include those indicated for use for fetal diagnostic testing or newborn screening.(b)
Classification. Class II (special controls). The special controls for this device are:(1) The intended use on the device's label required under § 809.10(a)(2) of this chapter and device's labeling required under § 809.10(b)(2) of this chapter must include a statement that assay results are solely intended to be used in conjunction with other clinical and diagnostic findings, consistent with professional standards of practice, and that reflex testing, clinical genetic evaluation, and genetic counseling should be offered as appropriate.
(2) The labeling required under § 809.10(b) of this chapter must include:
(i) A warning that mosaicism detected in one tissue may not reflect mosaicism in other tissues and that the significance of mosaicism should be interpreted with caution in conjunction with other laboratory and clinical information (
e.g., sex of patient, diagnostic testing or carrier screening, patient symptoms) and should include appropriate genetic counseling.(ii) A prominent statement that this test is not indicated for use for fetal diagnostic testing, newborn screening or for stand-alone diagnostic purposes.
(iii) Information that addresses how to interpret different result outputs specific to the technology, such as (peaks) in the electropherograms.
(3) Design verification and validation must include the following:
(i) Appropriate design features and control elements incorporated into the testing procedure that mitigate the risk of incorrect clinical results. These include controls as determined acceptable by FDA that:
(A) Enable the user to determine when the amplification may yield incorrect results,
(B) Enable the user to determine when cross contamination may have occurred;
(C) Software risk control measures that address device system hazards;
(D) Provide software traceability that ensures all hazards are adequately controlled and that all controls have been validated in the final device design; and
(E) Ensure the instructions for use and test reports appropriately inform the user about the limitations of the assay.
(ii) Validated and acceptable, as determined by FDA, criteria for test result interpretation and reporting, including result outputs.
(iii) Acceptable, as determined by FDA, evidence demonstrating the clinical validity of the device which supports each indicated diagnostic use, including for each genotype and associated phenotype used in providing a clinical determination for the target population.
(iv) Evidence demonstrating acceptable, as determined by FDA, analytical device performance. Patient specimens must represent the full spectrum of expected clinical results and be obtained through unbiased collection. Specimens must be representative of all categories of results and across the range of repeat sizes (
e.g., categories and repeat sizes for Fragile X syndrome are: normal 1-44 repeats; intermediate 45-54 repeats; premutation 55-200 repeats, full mutation greater than 200 repeats), across a range of allelic combinations, be near decision points, and be from both male and female subjects. The number of specimens tested must be sufficient to obtain unbiased estimates of device performance. Analytical validation must include data demonstrating acceptable, as determined by FDA:(A) Agreement with a comparator method(s) determined to be acceptable by FDA. This evidence must demonstrate the accuracy for detecting the size of the nucleotide repeats and the diagnostic categorical calls in DNA in the indicated specimen type(s) from patients that are representative of the intended use population. Accuracy must be assessed for both diagnostic and carrier subsets independently.
(B) Device precision including repeatability and reproducibility, using clinical samples. The study must evaluate all possible sources of variability including, as appropriate, between-site and between operator at a minimum of three sites of which two must be external with a minimum of two operators per site, between-day on a minimum of 3 non-consecutive days, between-run, within-run, between-lot in a minimum of three lots, and between instrument on a minimum of three instruments. Precision must be demonstrated per specimen and determine for both categorical call and by the size of the repeat (
i.e., the percentage of replicates for which the allele fell within the target precision size range). Precision data must be calculated and presented with and without results determined to be invalid.(C) Device performance at the limit of detection of each allele across the range of sizes and as a function of the indicated DNA input for the assay.
(D) Specificity of the reagents for their targets, absence of cross-reactivity, evaluation of sources of interference relevant to the specimen type, and a demonstration of the absence of cross contamination.
(E) Performance of the pre-analytical methods, including DNA extraction methods.
(F) Performance of the device across the range of indicated DNA input concentrations for the assay.
(G) Specimen stability throughout indicated specimen storage ranges, including under expected storage and transport conditions.
(v) Robust evidence demonstrating that the number and frequency of incorrect results due to mosaicism are clinically acceptable, as determined by FDA.
(vi) An appropriate traceability plan to minimize the risk of incorrect results over time, including a description of the molecular size standards and other reagents that may be required for result interpretation, as applicable, that demonstrate the reliable interpretation of the size of the fragments.
(vii) Acceptable, as determined by FDA, device stability protocols and acceptance criteria, that are sufficient to ensure indicated analytical and clinical performance throughout the indicated device stability period. The protocols and acceptance criteria must be adequate to demonstrate that there is no degradation in signal intensity of full mutations when testing a specimen at the latest indicated time point within the indicated device stability that is comprised of the lowest indicated DNA input that can be used.