LogoFDA 510(k) Search
K Number
DEN070004
Device Name
XPERT EV, MODEL GXEV-100N-10; GENEXPERT DX SYSTEM, MODELS P/N 900-0144, 900-0145, 900-0146, 900-0065
Manufacturer
CEPHEID
Date Cleared
2007-03-16

(4 days)

Product Code
OAI
Regulation Number
866.3225
Type
Post-NSE
Panel
MI
Reference & Predicate Devices
N/A
AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
Intended Use

The Cepheid® Xpert EV assay is a reverse transcription polymerase chain reaction (RT-PCR) using the GeneXpert® Dx System for the presumptive qualitative detection of enterovirus (EV) RNA in cerebrospinal fluid (CSF) specimens from individuals with signs and symptoms of meningitis. This test, in conjunction with other laboratory results and clinical information, may be used as an aid in the laboratory diagnosis of enterovirus infection in patients with a clinical suspicion of meningitis or meningoencephalitis.

Assay performance characteristics have not been established for immunocompromised or immunosuppressed patients.

CAUTION: The results obtained with the Xpert EV assay should be used only as an adjunct to clinical observation and other information available to the physician. Positive Xpert EV results do not rule out other causes of meningitis, including bacteria, mycobacteria, other viruses (e.g. herpes family viruses, arboviruses, mumps virus, etc) and fungi.

Device Description

The Xpert EV assay is designed to detect enterovirus (EV) RNA (enterovirus genome 5' untranslated region (UTR) between nucleotide 452 and 596) in CSF samples from patients exhibiting meningitis like symptoms. The assay includes reagents, primers, and probes for the simultaneous detection of nucleic acid from the target EV and the sample-processing control/internal control (SPC/IC). The assay includes the SPC/IC to verify adequate processing of the target virus and monitors the presence of inhibitors in the RT-PCR assay to avoid a false negative result. The assay also includes a probe check control to verify reagent rehydration, probe integrity, and reaction-tube filling in the cartridge.

The assay is run on the GeneXpert Dx System. The GeneXpert Dx System automates and integrates sample purification, nucleic acid amplification, and detection of the target sequence using real-time PCR and RT-PCR assays. The system consists of an instrument, personal computer, and preloaded software for running tests on collected samples and viewing the results. The system requires the use of single-use disposable GeneXpert cartridges that hold the PCR reagents and host the PCR process. Because the cartridges are self-contained, cross-contamination between samples is eliminated. The above described sample-processing control/internal control (SPC/IC) is named CIC in the GeneXpert Dx System software.

To run a test, the CSF sample and four reagents are transferred into designated chambers of the Xpert EV cartridge. The GeneXpert Dx System performs sample preparation by lysing the virus and SPC (armored pseudovirus RNA), binding the RNA to the capture matrix, and eluting the RNA. The RNA is mixed with dry RT reagents and transferred into the reaction tube for preparation of cDNA. The cDNA is then mixed with dry PCR reagents and transferred into the reaction tube for real-time PCR and detection. The EV primers and probe amplify and detect a consensus region of the enterovirus 5' untranslated region (UTR). The test takes approximately 2.5 hours.

AI/ML Overview

This document outlines the acceptance criteria and supporting studies for the Xpert EV assay, designed for the presumptive qualitative detection of enterovirus (EV) RNA in cerebrospinal fluid (CSF) specimens for individuals with meningitis symptoms.

Acceptance Criteria and Reported Device Performance

CriteriaReported Device Performance
Analytical Performance
Precision/ReproducibilityFirst Study (Multi-center, blind):
  • Negative samples: 100% agreement across 3 sites (90/90)
  • Positive samples (near LoD): 100% agreement for CVA6 (89/89) and CVA9 (90/90), 100% agreement for CVA17 (89/89). Total agreement for positive samples was 99.6% (268/268).
  • Overall agreement: 99.7% (358/358).
    Second Study (stressed system, 31 instruments, 121 ICORE modules):
  • Negative: 100% agreement (78/78)
  • CA9 2X LoD: 90% agreement (18/20)
  • CA9 4X LoD: 100% agreement (19/19)
  • EV70 2X LoD: 100% agreement (20/20)
  • EV70 4X LoD: 100% agreement (20/20) |
    | Linearity/Assay Reportable Range | Linear over an 8-log dynamic range for CVA9, with an R² value of 0.9989 (Ct range 10-40). |
    | Detection Limit (LoD) | CVA9: 80.0 TCID50/mL
    EV70: 1.3 TCID50/mL
    PV1: 4.0 TCID50/mL
    CVA17: 1.0 TCID50/mL
    CVA6: 33.0 TCID50/mL |
    | Analytical Reactivity (60 EV serotypes) | All 60 tested enterovirus serotypes were detected at their presumed LoD (replicates of 3). Estimated TCID50/mL values provided for each. |
    | Analytical Specificity | No detectable amplicons generated for 22 non-EV pathogens (bacteria including Group B Streptococcus, H. influenzae, E.coli, N. meningitidis, C. freundii, C. koseri and viruses including EBV, HSV-1, HSV-2, HHV-6, HHV-7, Adenovirus-2, Measles, Mumps, Parainfluenza 1-3, Influenza A, Influenza B, VZV, CMV) at high concentrations. No cross-hybridization with purified human genomic RNA from WBC. |
    | Interfering Substances | Positive enterovirus results obtained even with highest levels of potentially interfering substances: 1071 mg/dL protein, 7,140 cells/mm³ WBC, 2.5% v/v blood (bloody taps), and 3.6 g/dL hemoglobin. |
    | Clinical Performance (against Clinical Diagnosis) | Prospective Samples:
  • Sensitivity: 96.3% (26/27)
  • Specificity: 97.2% (103/106)
    Banked Samples:
  • Sensitivity: 100% (23/23)
  • Specificity: 97.0% (96/99) |
    | Clinical Performance (against Viral Culture) | Prospective Samples:
  • Positive Agreement: 100.0% (8/8)
  • Negative Agreement: 89.4% (110/123)
    Banked Samples:
  • Positive Agreement: 95.7% (22/23)
  • Negative Agreement: 81.4% (153/188) |

Study Details

  1. Sample Size for Test Set and Data Provenance:

    • Clinical Study:
      • Initial enrollment: 475 patients.
      • Analyzable subjects after exclusion: 434.
      • Subjects with all relevant test results for clinical diagnosis evaluation: 255 (133 prospective, 122 banked).
      • Subjects with viral culture results: 320 (131 prospective, 211 banked).
    • Provenance: Samples were collected retrospectively and prospectively through a multi-site investigational study at six institutions. The specific country of origin is not explicitly stated, but the multi-site nature suggests collection within the United States or a similar regulatory environment.

  2. Number of Experts and Qualifications for Ground Truth:

    • The document does not specify the number of individual experts (e.g., radiologists, pathologists) used to establish ground truth for the clinical diagnosis. Instead, it defines "Clinical Diagnosis" based on a set of objective laboratory results and clinical evidence, effectively using a consensus of clinical and laboratory findings rather than individual expert adjudication of individual cases.
    • For viral culture, it mentions a "designated central laboratory" using "Super E-Mix Shell Vials for viral culture" and "pan enterovirus antibody" staining, followed by "indirect immuno-fluorescence antibody for enterovirus identification." This implies expert laboratory personnel performing and interpreting these specialized tests, but specific qualifications are not detailed. Enrolling sites used their "own standard procedure for viral culture," also implying laboratory expertise.

  3. Adjudication Method for Test Set:

    • No formal adjudication method like "2+1" or "3+1" is described for the clinical ground truth.
    • The "Clinical Diagnosis" ground truth was established by pre-defined criteria: "clinical evidence consistent with meningitis, laboratory results for CSF Gram stain, CSF bacterial culture, CSF glucose, CSF-blood glucose ratio, CSF total protein concentration, CSF leukocyte count, and either detection of an EV genome in CSF and/or positive CSF EV culture." This represents a rule-based consensus from a combination of clinical and laboratory data rather than expert adjudication of interpretations.
    • For viral culture, a comparison was made between enrolling sites and a central laboratory, with 56 out of 57 subjects showing concordant results, and one discrepant result. This suggests a form of cross-validation or comparison, but not a formal adjudication process.

  4. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:

    • No MRMC comparative effectiveness study was done. This device is a standalone diagnostic assay for pathogen detection, not an AI-assisted human reader system. Therefore, the effect size of human readers improving with AI assistance is not applicable.

  5. Standalone (Algorithm Only) Performance:

    • Yes, the entire performance evaluation presented in the document represents the standalone performance of the Xpert EV assay (algorithm/device only) without human-in-the-loop assistance for interpretation beyond standard laboratory procedures and clinical decision-making. The "Xpert EV result" (positive or negative) is the direct output of the instrument.

  6. Type of Ground Truth Used:

    • Clinical Diagnosis: A composite "Clinical Diagnosis" ground truth was used, based on a combination of:
      • Clinical evidence consistent with meningitis.
      • Standard laboratory results from CSF (Gram stain, bacterial culture, glucose, CSF-blood glucose ratio, total protein, leukocyte count).
      • Detection of an EV genome in CSF or positive CSF EV culture.
    • Viral Culture: Positive or negative results from viral culture performed by designated central and local laboratories were also used as a ground truth for comparison.

  7. Sample Size for Training Set:

    • The document does not explicitly mention a "training set" in the context of device development or specific algorithm training. The performance characteristics studies (analytical and clinical) are presented as validation studies for the device. If an internal algorithm training phase occurred, its sample size is not detailed in this summary.

  8. How Ground Truth for Training Set was Established:

    • As no explicit "training set" is described for an algorithm, the method for establishing ground truth for such a set is not provided. The described studies focus on the validation of the device's performance against established clinical and laboratory reference methods.

§ 866.3225 Enterovirus nucleic acid assay.

(a)
Identification. An enterovirus nucleic acid assay is a device that consists of primers, probes, enzymes, and controls for the amplification and detection of enterovirus ribonucleic acid (RNA) in cerebrospinal fluid (CSF) from individuals who have signs and symptoms consistent with meningitis or meningoencephalitis. The detection of enterovirus RNA, in conjunction with other laboratory tests, aids in the clinical laboratory diagnosis of viral meningitis caused by enterovirus.(b)
Classification. Class II (special controls). The special control is FDA's guidance document entitled “Class II Special Controls Guidance Document: Nucleic Acid Amplification Assay for the Detection of Enterovirus RNA.” See § 866.1(e) for the availability of this guidance document.