K Number
K053497
Date Cleared
2006-04-14

(120 days)

Product Code
Regulation Number
864.5220
Panel
HE
Reference & Predicate Devices
AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
Intended Use

The EZCD4 Assay is intended to be performed on a Guava PCA System with CytoSoft 2.3 version software which includes three modules; EZCD4, Guava Check and Clean and Shutdown. The system is intended to identify and quantify the absolute counts of CD4 T-Lymphocytes in EDTA whole blood. The GuavaEZCD4 system is intended for the ongoing monitoring of patients with documented diagnosis of an immunodeficiency disease. The Guava EZCD4 system is intended for use only by trained laboratory professionals.

Device Description

The Guava EZCD4 System is an optimized cell analysis system consisting of the Guava PCA instrument, EZCD4 software for data acquisition and analysis and an EZCD4 Reagent Kit consisting of optimized reagents and protocols. The Guava EZCD4 Reagent Kit is a two-color direct immunofluorescence kit used for the enumeration of mature CD4 T lymphocytes in human blood. It consists of a monoclonal anti-human CD3 antibody conjugated to the tandem dye, phycoerythrin (PE)-Cy5 (PECy5) and a monoclonal anti-human CD4 antibody conjugated to PE. Guava 1X Lysing Solution and Guava Fixative are available in separate packaging and are added, after staining, to lyse erythrocytes and to preserve cells respectively. The Guava PCA instrument includes a laptop computer. The Guava PCA incorporates a new technology termed micro capillary cytometry. The device is a 3-parameter flow cytometer that utilizes a solid state 532 nm green laser, a fixed optical & fluidic system, a self-aligning flow cell and a micro capillary delivery system to perform cell analysis. The micro-syringe stepper pump allows the use of small sample volumes and volumetric measurements for the determination of absolute counts per microliter. The Guava PCA optic system detects forward scatter and two different wavelengths of fluorescence, 580/20 nm and 675/20 nm. It has a small foot print, 21.6 cm x 32 cm x 36.3 cm and is equipped with a network ready laptop computer with a 10GB hard drive and Windows 2000 operating system. The software is application specific with onscreen messaging enabling ease of use and facilitates short, effective training programs. Data is readily analyzed immediately after acquisition and is stored as electronic files on the hard drive.

AI/ML Overview
{
  "acceptance_criteria_and_performance_table": {
    "title": "Summary of Guava EZCD4 System Performance Studies",
    "headers": [
      "Study Type",
      "Acceptance Criteria",
      "Reported Device Performance"
    ],
    "rows": [
      [
        "Correlation Study (vs. Predicate Devices)",
        "Not explicitly stated as numerical acceptance criteria, but implied demonstration of substantial equivalence via linear regression (R-squared, Slope, Intercept).",
        "**Site 2:** N=92, R-squared=0.95, Slope=+1.00, Intercept=18.64, Range=13-1465\n**Site 3:** N=91, R-squared=0.93, Slope=+0.96, Intercept=35.51, Range=17-1175\n**Site 4:** N=88, R-squared=0.98, Slope=+1.17, Intercept=18.46, Range=47-1439\n**Site 5:** N=94, R-squared=0.98, Slope=+0.95, Intercept=13.29, Range=8-1076\n(Conclusion: Results are substantially equivalent)"
      ],
      [
        "Linearity and Sensitivity Study",
        "Not explicitly stated as numerical acceptance criteria, but implied demonstration of linearity across a wide range of CD4+ T cell counts.",
        "Linear regression equation: y = 1.0118x - 45.237\nR-squared = 0.9866\n(Graph shows strong linearity from 0 to >2000 CD4+ T cells/µL)"
      ],
      [
        "Intra-Laboratory Reproducibility",
        "Not explicitly stated as numerical acceptance criteria (e.g., max CV%), but implied low variability across sites and ranges.",
        "**Site 2:** Low CV=13.50%, Mid CV=8.06%, High CV=4.75%\n**Site 3:** Low CV=10.45%, Mid CV=7.20%, High CV=4.92%\n**Site 4:** Low CV=11.05%, Mid CV=3.87%, High CV=3.56%\n**Site 5:** Low CV=4.69%, Mid CV=3.69%, High CV=3.23%"
      ],
      [
        "Carryover Study",
        "Absence of significant sample carryover, as determined by statistical comparison.",
        "Wilcoxon-Mann-Whitney test 1-sided p-value = 0.1474 (Conclusion: No significant sample carryover demonstrated)"
      ],
      [
        "Specificity Study (Guava anti-CD3)",
        "Expected specificity for peripheral blood lymphocytes.",
        "Mean % positive in Lymphocytes: 47.80%\nMean % positive in Monocytes: 3.52%\nMean % positive in Granulocytes: 1.52%\n(Conclusion: Expected specificity for lymphocytes)"
      ],
      [
        "Specificity Study (Guava anti-CD4)",
        "Expected specificity for peripheral blood lymphocytes, allowing for known weak positive staining of monocytes.",
        "Mean % positive in Lymphocytes: 23.58%\nMean % positive in Monocytes: 81.58%\nMean % positive in Granulocytes: 4.39%\n(Conclusion: Expected specificity for lymphocytes, with known weak positive for monocytes)"
      ],
      [
        "Blood Sample Aging (Pre-Staining)",
        "Storage of unstained EDTA whole blood for specific periods without significant impact on results.",
        "Unstained EDTA anti-coagulated whole blood specimens may be stored for periods up to and including 72 hours (at 18-25°C) prior to staining and analysis."
      ],
      [
        "Blood Sample Aging (Post-Staining)",
        "Storage of stained and fixed EDTA whole blood for specific periods without significant impact on results.",
        "Stained EDTA anti-coagulated whole blood specimens may be stored for periods up to and including 24 hours (at 18-25°C or 2-8°C) prior to analysis."
      ]
    ]
  },
  "study_details": {
    "test_set_sample_size": {
      "correlation_study": "365 abnormal whole blood samples (approximately 90 per site) from three CD4+ absolute count ranges (0-200, 201-500, 501-2000).",
      "linearity_sensitivity": "22 aliquots (11 pairs) prepared from bulk blood samples.",
      "intra_laboratory_reproducibility": "10 replicate whole blood samples from each of three abnormal donors (low, mid, high ranges) per site, across 4 sites. Total 120 samples (3 donors x 10 replicates x 4 sites).",
      "carryover_study": "21 low range sample replicates and 18 high range sample replicates in a specific alternating sequence.",
      "specificity_study": "5 healthy subjects, EDTA anti-coagulated whole blood samples, analyzed for 15,000 events per blood component for each donor.",
      "blood_sample_aging": "Pre-Staining: 5 healthy normal donors, 5 blood specimens each (total 25 specimens). Post-Staining: 5 healthy normal donors, 7 stained samples per blood sample, pooled (total 35 samples for 18-25°C and 35 for 2-8°C)."
    },
    "data_provenance": {
      "correlation_study": "Prospective, collected at four clinical trial sites in the US.",
      "specificity_study": "Prospective, collected from healthy subjects.",
      "blood_sample_aging": "Prospective, collected from healthy normal donors."
    },
    "number_of_experts_ground_truth_test_set": "Not applicable, as these studies primarily assessed device performance against predicate devices or expected biological/analytical characteristics rather than expert-derived ground truth.",
    "qualifications_of_experts": "Not applicable.",
    "adjudication_method": "Not applicable.",
    "mrmc_comparative_effectiveness_study": "No, a multi-reader multi-case (MRMC) comparative effectiveness study was not explicitly described, as this is a device for quantifying CD4+ T cells, not for image interpretation by human readers. The correlation study compared the device's output to predicate flow cytometry systems.",
    "standalone_performance_done": "Yes, various standalone performance metrics were evaluated, including linearity, reproducibility (intra-laboratory precision), carryover, and specificity. The correlation study also assessed its performance against predicate devices.",
    "type_of_ground_truth_test_set": {
      "correlation_study": "Reference measurements from predicate flow cytometry systems (Becton Dickinson FACSCalibur with MultiTest or TriTest reagents).",
      "linearity_sensitivity": "Known 'Expected values' derived from gravimetric preparation of blood cell aliquots with known high and low CD4+ T cell counts.",
      "intra_laboratory_reproducibility": "The device's own measurements across replicates and sites, assessing internal consistency; no external 'ground truth' in the context of diagnostic accuracy.",
      "carryover_study": "The device's own measurements of alternating low and high concentration samples, showing lack of influence.",
      "specificity_study": "The characteristic immunological staining patterns of known cell populations (lymphocytes, monocytes, granulocytes) as observed on a reference flow cytometer (FACSCalibur) with the study reagents.",
      "blood_sample_aging": "The device's own measurements at different time points, comparing to the initial (0 hour) measurement value as a reference for stability."
    },
    "training_set_sample_size": "Not specified. As this is a flow cytometer system with reagents and software for quantitative measurements, it's unlikely to have a 'training set' in the machine learning sense. The software's algorithms would likely be based on established flow cytometry principles and calibrated/validated during development.",
    "how_ground_truth_training_set_established": "Not applicable. The system's operation and algorithms for data acquisition and analysis are based on established principles of flow cytometry and are likely calibrated and validated through internal development and testing, rather than an external 'ground truth' for a machine learning training set."
  }
}

§ 864.5220 Automated differential cell counter.

(a)
Identification. An automated differential cell counter is a device used to identify one or more of the formed elements of the blood. The device may also have the capability to flag, count, or classify immature or abnormal hematopoietic cells of the blood, bone marrow, or other body fluids. These devices may combine an electronic particle counting method, optical method, or a flow cytometric method utilizing monoclonal CD (cluster designation) markers. The device includes accessory CD markers.(b)
Classification. Class II (special controls). The special control for this device is the FDA document entitled “Class II Special Controls Guidance Document: Premarket Notifications for Automated Differential Cell Counters for Immature or Abnormal Blood Cells; Final Guidance for Industry and FDA.”