(393 days)
The AxSYM® CA 125™ assay is a Microparticle Enzyme Immunoassay (MELA) for the quantitative measurement of CA 125 assay values in human serum. The AxSYM® CA 125™ assay is to be used as an aid in monitoring response to therapy for patients with epithelial ovarian cancer. Serial testing for patient CA 125 assay values should be used in conjunction with other clinical methods used for monitoring ovarian cancer.
AxSYM CA 125 is a microparticle enzyme immunoassay on the AxSYM System for the quantitative measurement of CA 125 assay values in human serum. AxSYM CA 125 employs Abbott Calibrators and Controls.
Here's an analysis of the provided text regarding the Abbott AxSYM CA 125 assay, structured according to your request.
Acceptance Criteria and Study for Abbott AxSYM CA 125 Assay
The provided document describes the Abbott AxSYM CA 125 assay, a microparticle enzyme immunoassay (MEIA) intended for the quantitative measurement of CA 125 in human serum, specifically as an aid in monitoring the response to therapy for patients with epithelial ovarian cancer. The study demonstrating its performance focuses on proving substantial equivalence to a predicate device, the ABBOTT CA 125 II™ RIA assay.
1. Table of Acceptance Criteria and Reported Device Performance
Since this is a submission demonstrating substantial equivalence to a predicate device, the "acceptance criteria" are intrinsically linked to the predicate's performance and the statistical measures used to show comparable results.
| Acceptance Criteria Category (Implied for Substantial Equivalence to Predicate) | Target/Predicate Performance (ABBOTT CA 125 II™ RIA) | AxSYM CA 125 Performance (Reported) | Met? |
|---|---|---|---|
| Correlation Coefficient (Linear Regression vs. Predicate) | Not explicitly stated as a target, but high correlation expected | 0.997 | Yes |
| Slope (Linear Regression vs. Predicate) | Expected to be close to 1.0 | 1.01 | Yes |
| Y-intercept (Linear Regression vs. Predicate) | Expected to be close to 0.0 | -2.6 U/mL | Yes |
| Dynamic Range | 0 - 500 U/mL | 0 - 600 U/mL | Yes |
| Sensitivity | 0.4 U/mL | 2 U/mL | Yes (within acceptable range for equivalence) |
| Area Under the Curve (ROC Analysis) | 0.80 | 0.84 | Yes |
| Sensitivity (at 35 U/mL reference value) | 66.0% | 65.0% | Yes |
| Specificity (at 35 U/mL reference value) | 92.0% | 95.4% | Yes |
| Concordance (at 35 U/mL) | Not explicitly stated as a predicate target, but high concordance expected | 96.2% (healthy) | |
| 97.8% (benign) | |||
| 98.0% (cancer) | |||
| 95.2% (total) | Yes | ||
| Comparable Trending Results (Serial Data) | Assumed for predicate | Good agreement with clinical status for 39/44 patients | Yes |
Notes on Acceptance Criteria: The primary acceptance criterion here is "substantial equivalence" to the predicate device. This is demonstrated by showing that the new device performs similarly across various analytical and clinical metrics. The specific numeric targets for the AxSYM CA 125 assay would have been set by Abbott based on the predicate's known performance and what is considered clinically acceptable for a CA 125 assay.
2. Sample Sizes Used for the Test Set and Data Provenance
- Linear Regression Analysis: 533 specimens
- Data Provenance: Not specified (e.g., country of origin). The study is retrospective, as it uses existing specimens with measured CA 125 values.
- ROC Analysis and Sensitivity/Specificity:
- 130 apparently healthy females
- 45 patients with benign gynecologic disease
- 197 patients with ovarian cancer
- Total for ROC/Sensitivity/Specificity: 372 specimens
- Data Provenance: Not specified (e.g., country of origin). The study is retrospective.
- Concordance (Total Subjects): 542 total subjects (which seems to encompass the 130 healthy, 45 benign, and 197 ovarian cancer patients, plus potentially other samples used in the linear regression that weren't categorized for ROC analysis).
- Data Provenance: Not specified (e.g., country of origin). The study is retrospective.
- Serial Tracking Data: 44 patients with ovarian cancer
- Data Provenance: Not specified (e.g., country of origin). This part of the study is prospective or retrospective depending on how the "clinical status" data was collected in relation to the serial testing. Given it's a 510(k) summary, it's likely retrospective use of existing banked samples with known clinical follow-up.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications
- The document does not explicitly state the "number of experts" or their qualifications for establishing ground truth for the test set.
- For the ROC analysis and concordance, patient categorization (healthy, benign gynecologic disease, ovarian cancer) would have been based on clinical diagnosis, which inherently involves expert medical judgment (e.g., oncologists, gynecologists, pathologists). However, the specific details of this process are not provided.
- For the serial tracking data, "clinical status" would have been determined by treating physicians, also representing expert medical opinion.
4. Adjudication Method for the Test Set
- The document does not describe a specific adjudication method (e.g., 2+1, 3+1).
- Given that the ground truth appears to be based on established clinical diagnoses for patient groups (healthy, benign, cancer) and clinical status for serial tracking, it's likely that the diagnoses were well-documented and perhaps not subject to a specific adjudication process for the purpose of this assay validation, but rather accepted clinical records.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was Done
- No, an MRMC comparative effectiveness study was not done.
- This study evaluates a laboratory assay (an in-vitro diagnostic device), not an imaging device or a decision-support AI that directly assists human readers/interpreters in making diagnoses from complex data. Therefore, the concept of "human readers improving with AI vs. without AI assistance" does not apply here. The device itself is the "AI" (automated assay system), not an aid to human interpretation of another data source.
6. If a Standalone Study (Algorithm Only Without Human-in-the-Loop Performance) was Done
- Yes, this entire study is essentially a standalone (algorithm only) performance evaluation. The AxSYM CA 125 assay operates automatically to measure CA 125 levels. Its "performance" is its ability to accurately and precisely quantify CA 125 in serum and to align with existing clinical classifications/outcomes. The comparisons are between the new automated assay and a predicate manual/semi-automated assay (RIA), and against clinical categories, demonstrating its standalone analytical and clinical utility.
7. The Type of Ground Truth Used
- Clinical Diagnosis: For the ROC analysis, sensitivity, specificity, and concordance, the ground truth was based on the clinical diagnosis of the patient populations (apparently healthy, benign gynecologic disease, and ovarian cancer).
- Clinical Status/Outcome Data: For the serial tracking data, the ground truth was the "clinical status" of the ovarian cancer patients, which would encompass progression, remission, or stability based on various clinical assessments.
8. The Sample Size for the Training Set
- The document does not specify a separate "training set". This type of assay validation for substantial equivalence typically uses collected patient samples (the "test set") to directly compare the new device against the predicate and against clinical outcomes. Immunoassays are "trained" during their development phase to optimize reagent concentrations, reaction conditions, and calibration curve parameters, but this is a different concept from a machine learning "training set". The 533 and 372 specimens mentioned are for the validation/comparison study.
9. How the Ground Truth for the Training Set Was Established
- As no explicit "training set" is mentioned in the context of machine learning, this question is not directly applicable. If "training set" refers to the samples used during the development and optimization of the immunoassay itself (e.g., for calibrator development), the document does not provide details on how ground truth was established for those early-stage samples. However, for an immunoassay, the "ground truth" for training would typically involve using samples with known analyte concentrations determined by reference methods or highly characterized pools.
§ 866.6010 Tumor-associated antigen immunological test system.
(a)
Identification. A tumor-associated antigen immunological test system is a device that consists of reagents used to qualitatively or quantitatively measure, by immunochemical techniques, tumor-associated antigens in serum, plasma, urine, or other body fluids. This device is intended as an aid in monitoring patients for disease progress or response to therapy or for the detection of recurrent or residual disease.(b)
Classification. Class II (special controls). Tumor markers must comply with the following special controls: (1) A guidance document entitled “Guidance Document for the Submission of Tumor Associated Antigen Premarket Notifications (510(k)s) to FDA,” and (2) voluntary assay performance standards issued by the National Committee on Clinical Laboratory Standards.