K Number
K062473
Manufacturer
Date Cleared
2007-04-12

(231 days)

Product Code
Regulation Number
866.3830
Reference & Predicate Devices
Predicate For
N/A
AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
Intended Use

The LIAISON® Borrelia burgdorferi assay uses chemiluminescent immunoassay (CLIA) technology on the LIAISON® Analyzer for the qualitative presumptive detection of IgG and IgM antibodies to VIsE (variable major protein-like sequence, expressed) protein antigen of Borrelia burgdorferi in human serum. This assay should be used only on samples from patients with signs and symptoms that are consistent with Lyme disease. Positive or equivocal results should be supplemented by testing with a standardized Western blot procedure. Positive supplemental results provide evidence of exposure to Borrelia burgdorferi and can be used to support a clinical diagnosis of Lyme disease. Negative results by LIAISON® Borrelia burgdorferi should not be used to exclude Lyme disease.

The LIAISON® Borrelia burgdorferi Serum Controls contains two assayed quality control sera (negative and positive) that are used to monitor the performance of the LIAISON® Borrelia burgdorferi assay.

Device Description

The method for the qualitative determination of IgG and IgM antibodies to Borrelia burgdorferi is an indirect chemiluminescence immunoassay (CLIA). All assay steps and incubations are performed by the LIAISON® Analyzer, with the exception of the initial magnetic particle resuspension. Recombinant antigens specific for Borrelia (VIsE antigens) are used for coating the magnetic particles (solid phase). Two mouse monoclonal antibodies (anti-human IgG and anti-human IqM) are linked to an isoluminol derivative (isoluminol-antibody coniugate). During the first incubation, anti-Borrelia burgdorferi antibodies present in calibrators, samples or controls bind to the solid phase magnetic particles. Unbound material is removed by a wash cycle and isoluminol-antibody conjugates are added. During the second incubation, the antibody conjugates react with anti-Borrelia burgdorferi IgG and IgM antibodies that are already bound to the solid phase. The second incubation is followed by a wash cycle to remove unbound conjugate. Subsequently, the start reagents are added and a flash chemiluminescence reaction is thus induced. The light signal and hence the amount of isoluminol-antibody coniuqate is measured by a photomultiplier as relative light units (RLU) and is indicative of the presence of Borrelia burgdorferi antibodies present in calibrators, controls or samples.

AI/ML Overview

Here's a detailed breakdown of the acceptance criteria and the study proving the device meets them, based on the provided text:

Device Name: DiaSorin LIAISON® Borrelia burgdorferi and DiaSorin LIAISON® Borrelia burgdorferi Serum Controls

Intended Use: Qualitative presumptive detection of IgG and IgM antibodies to Borrelia burgdorferi in human serum, to be used only on samples from patients with signs and symptoms consistent with Lyme Disease. Positive/equivocal results need supplemental Western Blot testing.


1. Table of Acceptance Criteria and Reported Device Performance:

The document implicitly uses performance metrics like Percent Agreement (Sensitivity and Specificity related) and Reproducibility (%CV) as indicators of acceptable performance. Explicit acceptance criteria in terms of specific thresholds for these metrics are not stated in the provided text as 'acceptance criteria'. Instead, the study results are presented as evidence of adequate performance for substantial equivalence.

Performance MetricAcceptance Criteria (Implicit from context)Reported Device Performance
Comparative Clinical Trials
Lyme Disease Testing PatientsHigh agreement with comparator and Western Blot for positive/negative samplesPositive Agreement: 70.0% (35/50) (95% CI: 57.6 - 80.5%)
Negative Agreement: 99.1% (973/982) (95% CI: 98.4 - 99.5%)
Overall Agreement: 97.1% (1008/1038) (95% CI: 96.1 - 97.9%)
(Vs. Comparator Immunetics® C6 ELISA™ Kit)
LIAISON vs. Western Blot (Positive/Equivocal Samples)Proportion of W. Blot positive results for LIAISON positive/equivocal samplesLIAISON Positive: 41 samples, 22 W. Blot positive
LIAISON Equivocal: 4 samples, 0 W. Blot positive
Apparently Healthy Adult Blood DonorsLow prevalence in healthy populationsEndemic Area: 0.7% Positive Prevalence (2/300)
Non-Endemic Area: 0.3% Positive Prevalence (1/300)
Characterized Lyme Disease Samples (CDC)High agreement with characterized CDC samplesPositive Agreement: 93.2% (41/44) (95% CI: 83.3 – 98.1%)
Negative Agreement: 81.3% (13/16) (95% CI: 58.3 – 94.5%)
Overall Agreement: 90.0% (54/60) (95% CI: 83.3 – 95.5%)
LYMErix™ Vaccine RecipientsNo false positives for pre-vaccine negative samples; agreement for post-vaccinePre-vaccine Negative: 3/3 (100%) (95% CI: 36.9 - 100%)
Post-vaccine Negative: 11/11 (100%) (95% CI: 76.1 - 100%)
(Compared to Immunetics® C6 Lyme ELISA™ for pre/post vaccine status)
CDC Lyme serum panel (Time from Onset)Reasonable agreement across different stages of disease onsetNormals: 100% agreement (5/5)
0-1 Month: 83.3% agreement (5/6)
1-2 Months: 75.0% agreement (6/8)
3-12 Months: 53.3% agreement (8/15)

1 year: 100% agreement (8/8)
Total: 73.0% agreement (27/42) |
| Reproducibility (Index & RLU) | Low variability (%CV) across runs, days, sites, and instruments | Overall %CV for Index: Ranges from 8.66% to 17.57% across different samples and sites.
Overall %CV for RLU: Ranges from 9.58% to 14.31% across different samples and sites.
Internal Reproducibility (Index): Overall %CV ranges from 8.06% to 14.20% across different samples. |


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

  • Lyme Disease Testing Patients:
    • Sample Size: 1038
    • Data Provenance: Prospective samples tested at two external US laboratories and DiaSorin, Inc.
  • Apparently Healthy Adult Blood Donors:
    • Sample Size: 600 (300 from endemic areas, 300 from non-endemic areas)
    • Data Provenance: Prospective samples.
  • Characterized Lyme Disease Samples:
    • Sample Size: 60
    • Data Provenance: Retrospective samples provided by the Centers for Disease Control and Prevention (CDC), implying US origin.
  • LYMErix™ Vaccine Recipients:
    • Sample Size: 11
    • Data Provenance: Retrospective samples.
  • CDC Lyme Serum Panel:
    • Sample Size: 42
    • Data Provenance: Samples from the CDC, implying US origin.

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

The document does not explicitly state the number or qualifications of experts used to establish a "ground truth" for the test set in the traditional sense of consensus reading for medical imaging. The ground truth for the clinical samples appears to be established through:

  • Comparator Assay: Immunetics® C6 B. burgdorferi (Lyme) ELISA™ Kit (K003754)
  • Standardized Western Blot Procedure: MarDx Marblot Western Blot IgG and IgM assay for confirming positive/equivocal results.
  • Clinical Diagnosis: For the CDC Lyme serum panel, samples had a "clinical diagnosis of Lyme Disease" and were categorized as "Normals (no Lyme Disease)". This implies a ground truth based on clinical assessment and established panels, rather than expert adjudication of individual cases.
  • CDC Characterized Samples: The "Characterized Lyme Disease Samples" from the CDC likely have their status (positive/negative) established by the CDC through a comprehensive approach including clinical, laboratory, and epidemiological data.

Therefore, there isn't a stated number of human "experts" directly establishing ground truth via case-by-case review as would be seen in AI studies for classification tasks.


4. Adjudication Method for the Test Set:

No explicit adjudication method (like 2+1, 3+1) is described for the test set. The ground truth relies on:

  • Comparison with a legally marketed predicate device (Immunetics® C6 B. burgdorferi (Lyme) ELISA™ Kit).
  • Confirmation of positive/equivocal results by a standardized Western Blot procedure.
  • Use of characterized panels from the CDC.
  • "Clinical diagnosis of Lyme Disease" for some samples in the CDC panel.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done, and the effect size of how much human readers improve with AI vs without AI assistance:

No, an MRMC comparative effectiveness study was not done. This device is an in-vitro diagnostic (IVD) assay, not an AI software intended to assist human readers (e.g., radiologists). The study evaluates the performance of the assay itself compared to other diagnostic methods, not human performance with and without AI.


6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:

Yes, the studies presented are standalone performance evaluations of the LIAISON® Borrelia burgdorferi assay. The "device" in this context is the automated chemiluminescent immunoassay (CLIA) system, which produces a result (qualitative presumptive detection of antibodies) without direct real-time human interpretation or assistance during the assay process. The results are then read and potentially followed up by a clinician, but the assay itself functions independently.


7. The Type of Ground Truth Used:

The ground truth used in the studies is a combination of:

  • Comparator Assay Results: The Immunetics® C6 B. burgdorferi (Lyme) ELISA™ Kit was used as a comparator.
  • Standardized Western Blot Results: Used as a confirmatory test for positive/equivocal results.
  • Characterized Samples/Panels: Samples provided by the CDC which are stated as "Characterized Lyme Disease Samples" or have a "clinical diagnosis of Lyme Disease," indicating a ground truth established by a combination of clinical, laboratory, and epidemiological data.

8. The Sample Size for the Training Set:

The document does not provide information on a training set sample size or how the device was "trained." This is because the device is an immunoassay kit and an analyzer, not a machine learning or AI algorithm in the contemporary sense that would require a distinct training set for algorithm development. The "training" of such a device primarily involves assay development, optimization, and validation procedures, not iterative learning from a labeled dataset.


9. How the Ground Truth for the Training Set Was Established:

As explained in point 8, there is no explicit "training set" in the context of an AI/ML algorithm. Therefore, information on how a ground truth for a training set was established is not applicable and not provided in the document. The device's performance is driven by its biochemical reagents and optical detection system, which are developed and optimized through traditional laboratory methods rather than machine learning on a labeled dataset.

§ 866.3830

Treponema pallidum treponemal test reagents.(a)
Identification. Treponema pallidum treponemal test reagents are devices that consist of the antigens, antisera and all control reagents (standardized reagents with which test results are compared) which are derived from treponemal sources and that are used in the fluorescent treponemal antibody absorption test (FTA-ABS), theTreponema pallidum immobilization test (T.P.I.), and other treponemal tests used to identify antibodies toTreponema pallidum directly from infecting treponemal organisms in serum. The identification aids in the diagnosis of syphilis caused by bacteria belonging to the genusTreponema and provides epidemiological information on syphilis.(b)
Classification. Class II (performance standards).