Search Results

The ARCHITECT CA 15-3 assay is a Chemiluminescent Microparticle Immunoassay (CMIA) for the quantitative determination of DF3 defined antigen in human serum and plasma on the ARCHITECT i System. The ARCHITECT CA 15-3 assay is to be used as an aid in the management of Stage II and Stage III breast cancer patients. Serial testing for patient CA 15-3 assay values should be used in conjunction with other clinical methods for monitoring breast cancer.

The ARCHITECT CA 15-3 assay is a two-step immunoassay to determine the presence of DF3 reactive determinants in human serum or plasma, using Chemiluminescent Microparticle Immunoassay (CMIA) technology with flexible assay protocols, referred to as Chemiflex™. In the first step, sample, wash buffer and 115D8 coated paramagnetic microparticles are combined. DF3 reactive determinants present in the sample bind to the 115D8 coated microparticles. After washing, DF3 acridinium-labeled conjugate is added in the second step. Pre-Trigger and Trigger Solutions are then added to the reaction mixture; the resulting chemiluminescent reaction is measured as relative light units (RLUs). A direct relationship exists between the amount of DF3 reactive determinants in the sample and the RLUs detected by the ARCHITECT i optical system.

Acceptance Criteria and Device Performance Study for ARCHITECT® CA 15-3® Assay

This document summarizes the acceptance criteria and the study performed to demonstrate that the ARCHITECT® CA 15-3® Assay meets these criteria, based on the provided 510(k) summary.

1. Table of Acceptance Criteria and Reported Device Performance

The provided document details various performance characteristics, which serve as the acceptance criteria for the device. The reported performance for the ARCHITECT CA 15-3 Assay is compared against these criteria.

2. Sample Size Used for the Test Set and Data Provenance

• Reproducibility (Precision):
  • Sample Size: Five defibrinated plasma-based panel members were tested. Each was tested in replicates of two, at two separate times per day, for 20 days. This amounts to 5 samples * 2 replicates * 2 times/day * 20 days = 400 individual measurements.
  • Data Provenance: The study was performed at Fujirebio Diagnostics, Inc. (FDI). It is a prospective study as samples were repeatedly tested under defined conditions.
• Comparison Study:
  • Sample Size: 402 serum specimens.
    • 250 serum specimens were from patients diagnosed with breast cancer (stage I through stage IV).
    • The remaining 152 specimens were presumably from other patient populations or healthy individuals to provide a broader range of concentrations.
  • Data Provenance: Not explicitly stated, but implies clinical samples acquired for testing. Likely retrospective, as samples were "tested using" both assays, suggesting pre-existing specimens. Country of origin not specified.
• Reference Ranges (Apparently Healthy Population):
  • Sample Size: 396 normal individual specimens.
    • 99 healthy pre-menopausal females.
    • 100 healthy post-menopausal females.
    • 197 healthy males.
  • Data Provenance: Not specified, but generally, reference range studies use prospectively collected healthy donor samples. Country of origin not specified.
• Reference Ranges (Patient Groups - Various Diseases):
  • Sample Size: 569 specimens from various disease states:
    • 120 Ovarian Cancer
    • 50 Colorectal Cancer
    • 50 Lung Cancer
    • 100 Breast Disease (Non-malignant)
    • 100 Ovarian Disease (Non-malignant)
    • 49 Urogenital Disease
    • 50 Pregnancy
    • 100 Hypertension/CHD
  • Data Provenance: Not specified, but likely retrospective collection of patient samples. Country of origin not specified.
• Breast Cancer Serial Specimens (Monitoring Effectiveness):
  • Sample Size:
    • Patients: 74 evaluable breast cancer patients.
    • Observations: 377 evaluable observations (average 5.1 observations per patient).
    • Observation Pairs: 303 observation pairs were used for the 2x2 table analysis.
  • Data Provenance: Likely retrospective collection of serial samples from breast cancer patients undergoing monitoring. The average age and stage distribution suggest these are patient cohorts from clinical practice. Country of origin not specified.

3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications

This being an immunoassay for a biomarker, the "ground truth" is typically defined by established clinical diagnostic criteria, pathology, or disease progression status determined by a physician. The document does not describe the use of independent experts to establish ground truth in the same way an imaging or diagnostic AI system might for interpretation tasks.

• For the Monitoring Effectiveness study, the "Change in Disease State" (Progression vs. No Progression) would be considered the ground truth. This is generally determined by attending clinicians based on a comprehensive assessment (imaging, clinical examination, pathology, other biomarkers, etc.). The document indicates "Stage was available from the charts for 61 women," implying physician diagnosis and staging. No specific number of experts or their qualifications for establishing this "ground truth" is provided, as it's assumed to be part of standard clinical practice.

4. Adjudication Method for the Test Set

Adjudication methods (like 2+1, 3+1) are typically used when multiple human readers interpret data to resolve disagreements and establish a consensus ground truth. For this type of immunoassay, the readouts are quantitative (U/mL), and the "ground truth" for disease status is derived from clinical outcomes rather than interpretation of the assay itself by multiple experts.

• Therefore, no specific adjudication method for the test set is mentioned or appears applicable in the context of this device. The values generated by the device are objective measurements, and the clinical outcomes for ground truth are determined by standard medical practice.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

• No MRMC comparative effectiveness study was performed or is applicable for this device. This is an immunoassay designed to quantitatively measure a biomarker, not to be interpreted by human readers in the same way as an imaging device. The device's output is a numerical value, not an image or data requiring human interpretation for diagnosis.

6. Standalone (Algorithm Only Without Human-in-the-Loop) Performance Study

• Yes, a standalone performance study was conducted. The entire testing described (reproducibility, method comparison, reference ranges, and serial monitoring) evaluates the performance of the ARCHITECT CA 15-3 Assay in generating quantitative results for CA 15-3. The device operates as an automated system to produce these measurements. The "human-in-the-loop" component is the clinician's interpretation of these quantitative results in conjunction with other clinical methods for patient management, not in the direct operation or result generation of the assay itself.

7. Type of Ground Truth Used

• Method Comparison: The ground truth for this comparison was the quantitative result provided by the predicate device, AxSYM CA 15-3 assay. This is a comparative "ground truth" rather than an absolute clinical outcome.
• Reference Ranges: The ground truth was based on the health status (apparently healthy vs. various disease states) of the individuals from whom the specimens were collected, as determined by their medical history/diagnosis.
• Monitoring Effectiveness (Serial Specimens): The ground truth was the "Change in Disease State" (Progression or No Progression) of the breast cancer patients. This would be determined by objective clinical assessments (e.g., imaging reports, biopsy results, clinical examination findings, and other physician observations) over time. This falls under outcomes data / clinical diagnosis.

8. Sample Size for the Training Set

• The document does not explicitly describe a separate "training set" in the context of machine learning or algorithm development. For an immunoassay, the development process involves reagent formulation, assay optimization, and calibration, which are typically iterative processes internal to the manufacturer. The data presented here are for validation and verification of the final commercial assay's performance.

9. How the Ground Truth for the Training Set was Established

• Since a distinct "training set" and its associated ground truth for algorithmic development are not outlined in the provided 510(k) summary (as it's not an AI/ML-based diagnostic device in the modern sense), this question is not applicable here. The "ground truth" for calibrating the assay would be established using known concentrations of CA 15-3 antigen.

Ask a specific question about this device