FDA 510(k) Search - By Innolitics (SaMD and AI Experts)

Immunological in vitro assay for quantitative determination of CA 15-3 in human serum and Li-heparin plasma to aid in the management of breast cancer patients. In conjunction with other clinical and diagnostic procedures, serial testing with this assay is an aid

· in the early detection of recurrence in previously treated stage II and III breast cancer patients
· for monitoring response to therapy in metastatic breast cancer patients
The electrochemiluminescence immunoassay "ECLIA" is intended for use on cobas e immunoassay analyzers.

Device Description

The Elecsys CA 15-3 II Test System is a two-step sandwich immunoassay with streptavidin microparticles and electrochemiluminescence detection for the quantitative determination of CA 15-3 in human serum and plasma. It is intended for use on the cobas e 411 immunoassay analyzer. The cobas e family of analyzers employs the electrochemiluminescence "ECLIA" technology.
Results are determined using a calibration curve that is generated specifically on each instrument by a 2-point calibration and a master curve provided with the reagent barcode.

AI/ML Overview

The provided text describes the 510(k) Premarket Notification for the Elecsys CA 15-3 II diagnostic test system. This is an in vitro immunoassay designed to quantitatively determine CA 15-3 in human serum and Li-heparin plasma to aid in the management of breast cancer patients. It is not an AI/ML powered device, therefore many of the requested elements pertaining to AI/ML specific studies (e.g., MRMC studies, training set details, expert ground truth establishment for AI) are not applicable.

However, I can extract information related to the performance studies conducted to prove the device meets acceptance criteria, focusing on the analytical performance characteristics relevant to an immunoassay.

Acceptance Criteria and Reported Device Performance

The document doesn't explicitly present a single "acceptance criteria" table with corresponding "reported device performance" values in a concise format. Instead, it describes various non-clinical performance evaluations and states that "specification was met" or "deviations were within predetermined acceptance criteria" for each. I will summarize the key performance parameters and the general findings related to meeting criteria.

Note: For each study, the acceptance criteria are implied by the statement that the "specification was met" or "deviations were within predetermined acceptance criteria." The document does not provide the numerical thresholds for these criteria.

Performance Parameter	Acceptance Criteria (Implied)	Reported Device Performance
1. Precision	Performance meets CLSI EP5-A3 guidelines.	Evaluated on one cobas e 801 Immunoassay Analyzer with one reagent lot. Performed 2 replicates of controls and 5 pooled human sera samples per run, 2 runs/day for 21 days. Implied: Met.
2. Analytical Sensitivity
2a. Limit of Blank (LoB)	Highest observed measurement values for samples free of analyte determined as 95th percentile.	Determined according to CLSI EP17-A2. Based on 1 run/day from 3 reagent lots on 1 cobas e 801 for 6 days (60 total blank replicates/lot). Reported: 1.0 U/mL (from comparison table). Implied: Met.
2b. Limit of Detection (LoD)	Lowest amount of analyte in a sample detectable with 95% probability.	Determined according to CLSI EP17-A2. Based on 1 run/day from 3 reagent lots on 1 cobas e 801 for 6 days (60 total replicates/sample/lot). Reported: 1.5 U/mL (from comparison table). Implied: Met.
2c. Limit of Quantitation (LoQ)	Performance meets CLSI EP17-A2 guidelines.	Based on 10 human serum samples (single donors) for 3 reagent lots. 1 run/day for 5 days (25 replicates/sample/lot). Reported: 3 U/mL (from comparison table). Implied: Met.
3. High-Dose Hook Effect	Hook concentration reported corresponds to the highest analyte concentration that generates a signal >10% above the upper limit of the measuring range.	Assessed on cobas e 801 in 3-fold determination with 2 spiked samples. Implied: Met for substances tested.
4. Linearity/Assay Reportable Range	All deviations within predetermined acceptance criteria. Performance meets CLSI EP6-A guidelines.	Evaluated using cobas e 801. Three high analyte serum and three high analyte plasma samples diluted into 15 concentrations (3-fold determination in single run). Implied: Met.
5. Endogenous Interferences	Non-interference at spiked concentrations for substances tested.	Tested effect of Bilirubin, Lipemia, Hemoglobin, Biotin, HAMA, and Rheumatoid Factors on cobas e 801 using low, mid, and high CA 15-3 concentrations. Implied: Met for all substances.
6. Exogenous Interferences – Anticoagulants	Performance consistent across different anticoagulants.	Determined by comparing values from serum and Li-Heparin plasma. 49 native samples tested on cobas e 801. Evaluation by Passing/Bablok regression. Implied: Met.
7. Method Comparison	Agreement with predicate device (cobas e 601 immunoassay analyzer). Performance meets CLSI EP09-A3 guidelines.	Performed with 190 human serum samples (single donors, native) covering the measuring range (5.14 - 290 U/mL for e 601; 4.24 – 273 U/mL for e 801). Evaluated by Passing/Bablok and linear regression. Implied: Met.
8. Reagent Stability	Onboard reagent stability and shelf-life stability claims supported.	Onboard: 16 weeks (new calibration every 28 days). Tested on cobas e 801. Real-time: Ongoing, tested at 0, 12/13/15, 17/19/25/30 months for 3 lots. Implied: Met.
9. Sample Stability	Sample stability maintained at various storage conditions.	2-8°C: 9 days (10 samples/type). Room Temp (20-25°C): 8 days (10 samples/type). -20°C: 53 weeks (10 samples/type). All tested on cobas e 601. Implied: Met.
10. Calibration Stability	Calibration lot stability (12 weeks) and on-board calibration stability (28 days) are maintained.	Lot: 12 weeks recommended. Tested for 13 weeks on cobas e 801 (5 HS samples). On-board: 28 days. Tested on cobas e 801 (6 HS samples). Implied: Met.

Study Details:

Sample Size Used for the Test Set and Data Provenance:
- Precision: 2 replicates of controls and 5 pooled human sera samples per run, 2 runs per day for 21 days.
- LoB/LoD: 60 total blank replicates per reagent lot (LoB) and 60 total replicates per sample per reagent lot (LoD).
- LoQ: 25 replicates per sample per reagent lot.
- High-Dose Hook Effect: 2 samples.
- Linearity: 3 high analyte serum samples and 3 high analyte plasma samples, diluted to 15 concentrations.
- Endogenous Interferences: For each interfering substance, three human serum samples (low, mid, high CA 15-3 concentrations).
- Exogenous Interferences (Anticoagulants): 49 native samples (serum and Li-Heparin plasma).
- Method Comparison: 190 human serum samples (single donors, native samples).
- Reagent Stability:
  - Onboard: Six human serum samples.
  - Real-time: PreciControl Tumor Marker Level 1 and 2 (3 lot numbers over various time points).
- Sample Stability: 10 samples for each sample type (Serum, Li-Heparin plasma) for each storage condition.
- Calibration Stability:
  - Lot: Five human serum samples.
  - On-board: Six human serum samples.
- Data Provenance: The document states "human serum" and "human plasma" samples were used, and refers to Roche Diagnostics GmbH in Mannheim, Germany, and Penzberg, Germany for establishment registration. It appears to be primarily laboratory-generated data for analytical validation, not specific patient cohorts with country of origin. The term "native samples" implies real patient samples, but no specifics on their origin are given beyond being "human." The studies are retrospective analytical investigations to characterize the device's technical performance.
Number of Experts Used to Establish the Ground Truth for the Test Set and the Qualifications of Those Experts:
- This is an in vitro diagnostic (IVD) immunoassay, not an AI/ML device relying on interpretation of medical images or clinical data by experts. The "ground truth" for these analytical studies is the quantitative concentration of CA 15-3 in the sample, measured using reference methods or established spiked concentrations. No human experts were involved in establishing the "ground truth" in the way it would be for an AI device. The ground truth for analytical performance is based on the inherent properties of the samples and reference materials.
Adjudication Method for the Test Set:
- Not applicable. This is an IVD immunoassay, not an AI/ML interpretation study requiring human adjudication.
If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done:
- No, an MRMC study was not done. This is an IVD immunoassay. MRMC studies are typically performed for imaging interpretation/AI devices to assess the impact of AI on human reader performance.
If a Standalone (i.e., Algorithm Only Without Human-in-the-Loop Performance) Was Done:
- Yes, the entire study is essentially a "standalone" performance evaluation of the immunoassay system (Elecsys CA 15-3 II on cobas e analyzers). Its output is a quantitative value (CA 15-3 concentration), not an interpretation that a human would then use as "assistance." The performance metrics evaluated (precision, linearity, limits, interference, method comparison) are intrinsic to the device's analytical function without human interaction for result generation.
The Type of Ground Truth Used:
- The ground truth for this device's performance studies is based on:
  - Reference materials/controls: For precision, stability studies (e.g., PreciControl Tumor Marker).
  - Known (spiked) concentrations: For linearity, hook effect, interference studies, where samples are prepared with precisely known amounts of analyte or interfering substances.
  - Native samples: For method comparison and sample type evaluations, where the "ground truth" is established by the predicate device or a trusted reference measurement.
  - Blank samples: For Limit of Blank determination.
The Sample Size for the Training Set:
- Not applicable in the context of an AI/ML algorithm. This device is a biochemical immunoassay. Its "training" or "calibration" involves internal calibration procedures (e.g., 2-point calibration and master curve provided with reagent barcode), which are distinct from the data-driven training of a machine learning model. The reference to "calibration curve" and "master curve" relates to the instrument's operational setup, not a machine learning training dataset.
How the Ground Truth for the Training Set Was Established:
- Not applicable in the AI/ML sense. Calibration is performed using "CalSet" (calibration materials) and is standardized against previous Elecsys CA 15-3 assays, which were in turn standardized against Enzymun Test CA 15-3 and CA 15-3 RIA by Fujirebio Diagnostics. This establishes traceability to recognized reference methods for the quantitative measurement.

Summary

Regulation Number and Section

§ 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.