(49 days)
The ARCHITECT CA 125 Il assay is a Chemiluminescent Microparticle Immunoassay (CMIA) for the quantitative determination of CA 125 reactive determinants in human serum and plasma on the ARCHITECT i System. The ARCHITECT CA 125 II 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 II assay values should be used in conjunction with other clinical methods used for monitoring ovarian cancer.
The ARCHITECT CA 125 II assay is a two-step immunoassay to determine the presence of OC125 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 of the ARCHITECT CA 125 II assay, sample and OC125 coated paramagnetic microparticles are combined. CA 125 reactive determinants present in the sample bind to the OC125 coated microparticles. After washing, M11 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 CA 125 reactive determinants in the sample and the RLUs detected by the ARCHITECT i optical system.
Here's an analysis of the provided 510(k) summary, extracting the requested information about acceptance criteria and the supporting study:
The document describes the ARCHITECT CA 125 II Assay, a diagnostic device, and its 510(k) submission. It focuses on demonstrating substantial equivalence to a predicate device, the AxSYM CA 125 Assay.
1. Table of Acceptance Criteria and Reported Device Performance
| Performance Metric | Acceptance Criteria (Implied/Derived from Predicate) | Reported Device Performance (ARCHITECT CA 125 II Assay) |
|---|---|---|
| Reproducibility | Total precision %CV should be acceptable for a diagnostic assay (implicitly comparable to predicate). | Total precision %CV was determined to be less than or equal to 10%. |
| Method Comparison (Correlation with Predicate) | High correlation with the predicate device (AxSYM CA 125 assay). | Correlation coefficient of 0.985. |
| Method Comparison (Slope vs. Predicate) | Slope of approximately 1.0 when compared to the predicate device. | Slope of 1.06 (99% CI: 1.03, 1.11). |
| Method Comparison (Y-axis Intercept vs. Predicate) | Y-axis intercept close to 0 when compared to the predicate device. | Y-axis intercept of 4.0 U/mL (99% CI: 2.0, 4.9). |
| Reference Range (Healthy Females Pre-Menopausal) | (Implicitly, comparable to established CA 125 levels in healthy population). | 89.9% ≤ 35 U/mL; 6.1% 35.1-65 U/mL; 4.0% 65.1-100 U/mL; 0.0% > 100 U/mL. 94.4% of healthy females had values ≤ 35.0 U/mL. |
| Reference Range (Healthy Females Post-Menopausal) | (Implicitly, comparable to established CA 125 levels in healthy population). | 99.0% ≤ 35 U/mL; 1.0% 35.1-65 U/mL; 0.0% 65.1-100 U/mL; 0.0% > 100 U/mL. |
| Association between CA 125 Change and Disease State (Total Concordance) | (Implied: Demonstrates clinical utility in monitoring response to therapy). | 68.3% Total Concordance based on observation pairs. |
| Association between CA 125 Change and Disease State (Positive Concordance) | (Implied: Demonstrates clinical utility in monitoring response to therapy). | 76.6% Positive Concordance based on observation pairs. |
| Association between CA 125 Change and Disease State (Negative Concordance) | (Implied: Demonstrates clinical utility in monitoring response to therapy). | 61.4% Negative Concordance based on observation pairs. |
| Per-Patient Concordance (Positive) | (Implied: Demonstrates clinical utility in monitoring response to therapy on a per-patient basis). | 97.9% (95% CI: 88.7% - 99.9%) |
| Per-Patient Concordance (Negative) | (Implied: Demonstrates clinical utility in monitoring response to therapy on a per-patient basis). | 37.5% (95% CI: 15.3% - 64.5%) |
| Per-Patient Concordance (Total) | (Implied: Demonstrates clinical utility in monitoring response to therapy on a per-patient basis). | 82.5% (95% CI: 70.9% - 91.0%) |
Note on Acceptance Criteria: For a 510(k) submission seeking substantial equivalence, explicit quantitative acceptance criteria are often not stated in the summary as they would be for a PMA. Instead, the performance demonstrated is compared directly to the predicate device's known performance or to generally accepted diagnostic assay performance characteristics (e.g., precision limits, strong correlation). The document's structure implies that the demonstrated performance is considered sufficient to establish substantial equivalence.
2. Sample Size Used for the Test Set and Data Provenance
- Reproducibility (Precision): Three defibrinated plasma panel members were tested. The exact number of samples from these panel members is not specified, but it involved replicates of two, two separate times per day, for 20 days on 2 separate instruments (total of 3 x 2 x 2 x 20 x 2 = 480 discrete measurements).
- Data Provenance: Not explicitly stated, but likely laboratory-generated samples for method validation.
- Comparison Study: 280 serum specimens (120 of these from Ovarian cancer Patients).
- Data Provenance: Not explicitly stated (e.g., country of origin, retrospective/prospective). However, given the context of a 510(k) for a diagnostic assay, it would likely be retrospective anonymized or de-identified clinical samples.
- Reference Ranges (Apparently Healthy Population): 196 normal individual specimens (99 pre-menopausal females, 97 post-menopausal females).
- Data Provenance: Not explicitly stated.
- Reference Ranges (Patient Groups): 615 individual specimens from various patient groups (Breast Cancer: 50, Ovarian Cancer: 166, Colorectal Cancer: 50, Endometrial Cancer: 25, Lung Cancer: 50, Ovarian Disease: 100, Urogenital Disease: 49, Hypertension/CHD: 100, Benign Endometrial: 25).
- Data Provenance: Not explicitly stated.
- Ovarian Cancer Serial Specimens (Association between Change in Marker Value and Change in Disease State): 63 patients, resulting in 306 evaluable observation pairs.
- Data Provenance: Not explicitly stated, but implies clinical data from patients undergoing therapy for ovarian cancer. It is stated that staging information was available for 60 of the 63 women.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications
- This is a diagnostic assay measuring a biomarker (CA 125). The ground truth for this type of device is typically the actual concentration of the analyte, or the clinical status of the patient as determined by various clinical methods (e.g., physician diagnosis, pathology, imaging, follow-up over time).
- No "experts" in the sense of human readers/interpreters establishing ground truth for the assay values directly are mentioned. The CA 125 values themselves are the output of the device.
- For the "Association between Change in Marker Value and Change in Disease State" section, disease progression/no progression (W) would be the clinical ground truth. The document does not specify the number of experts or their qualifications used to determine this clinical ground truth for the 63 ovarian cancer patients. It refers to "Staging was available from the chart," suggesting that standard clinical records and physician determinations served as the basis for the disease state.
4. Adjudication Method for the Test Set
- No adjudication method is described in the conventional sense (e.g., 2+1, 3+1 for imaging studies). This is expected for an automated laboratory diagnostic assay where the output is a quantitative value, not subject to human interpretation in the same way an image would be.
- For the clinical disease state (W) used in the serial specimen analysis, the method for determining "Progression" or "No Progression" is not detailed, so no adjudication method is provided.
5. If a Multi Reader Multi Case (MRMC) Comparative Effectiveness Study was done
- No, an MRMC comparative effectiveness study was not done. This type of study compares the performance of human readers, with and without AI assistance, typically for interpreting medical images. This document describes an automated in-vitro diagnostic (IVD) assay, which does not involve human readers interpreting AI output. The comparison is between the new assay and a predicate assay, and clinical utility is assessed by correlating assay changes with patient disease progression.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
- Yes, the primary performance evaluation described is a standalone performance of the algorithm (the ARCHITECT CA 125 II Assay). This device is an automated immunoassay system that provides a quantitative CA 125 value without human subjective interpretation. All presented data (precision, method comparison, reference ranges, and association with disease state) represent the performance of the device itself.
7. The Type of Ground Truth Used
- For the quantitative CA 125 values: The ground truth is considered to be the true concentration of CA 125 reactive determinants in the sample. This is implicitly confirmed by the comparison to a predicate device (which aims to measure the same analyte) and by the internal precision studies.
- For the "Association between Change in Marker Value and Change in Disease State": The ground truth is the clinical disease state or progression of ovarian cancer based on clinical methods and chart review. This is not strictly "expert consensus pathology" or "outcomes data" but rather a clinical determination of disease progression over time.
8. The Sample Size for the Training Set
- The document does not explicitly mention a "training set" for an algorithm. This is an immunoassay, not a machine learning algorithm in the typical sense that requires a training phase. While the assay's reagents and calibration parameters would have been developed and optimized, the specific sample sizes used in that developmental phase are not provided in this 510(k) summary focused on validation.
- The "calibration curve" mentioned in the reproducibility section suggests that a set of calibrators is used to establish the standard curve for quantitative measurement, but this is distinct from an AI training set.
9. How the Ground Truth for the Training Set was Established
- As a traditional immunoassay, there isn't a "training set" in the AI sense.
- For calibration, the ground truth for the calibrators would be established by careful preparation of known concentrations of the CA 125 antigen, typically through gravimetric or volumetric methods, and confirmation by reference methods. The document states "6 levels (0 - 1000 U/mL)" for standards, implying these are known concentrations used to create the standard curve.
§ 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.