Search Results

In-vitro diagnostic automated assay for the quantitative determination of the von Willebrand factor-GPIb-binding activity in human plasma collected from venous blood samples in 3.2% sodium citrate tubes on the BCS XP System.

As an aid used in the evaluation of patients with suspected or confirmed von Willebrand factor disorders.

Results of this test should always be interpreted in conjunction with the patient's medical history, clinical presentation and other laboratory findings.

The INNOVANCE VWF Ac assay is a particle enhanced turbidimetric assay based on binding of VWF to the recombinant GPIb (two gain-of-function mutations included). The assay provides quantitative VWF activity results on 3.2% citrated human plasma when used with Standard Human Plasma Calibrators (K023141).

The reagent kit consists of three components: Reagent I (containing polystyrene particles coated with anti-GPIb mouse monoclonal antibodies). Reagent II (buffer containing heterophilic blocking reagent) and Reagent III (containing recombinant GPIb).

The INNOVANCE VWF Ac assay is a device for the quantitative determination of von Willebrand factor-GPIb-binding activity. The device is intended to aid in the evaluation of patients with suspected or confirmed von Willebrand factor disorders.

Here's an analysis of its acceptance criteria and supporting studies:

1. Acceptance Criteria and Reported Device Performance

2. Sample sizes used for the test set and data provenance:

• Precision (Single Site):

  • Study 1 (reagent variability): 5 plasma pools + 2 control materials = 7 samples. Each tested 240 times (20 days x 2 runs x 2 replicates x 3 reagent lots or 3 calibrator lots for a total of 240 determinations, 80 per lot).
  • Study 2 (reproducibility multi-site): 5 plasma pools + 2 control materials = 7 samples. Each tested 90 times (3 external sites x 5 days x 2 runs x 3 replicates for a total of 90 determinations, 30 per site).
  • Study 3 (instrument/operator variability): 5 plasma pools + 2 control materials = 7 samples. Each tested 120 times (5 days x 2 runs x 4 replicates x 3 systems for a total of 120 determinations, 40 per system).
  • Provenance: Single site precision studies were conducted in Germany. Multi-site reproducibility and instrument/operator variability studies were conducted across three external sites (locations not specified for multi-site reproducibility study, but internal site for instrument/operator variability implies one location).

• Linearity (Test Sets):

  • Spiked Sample Linearity: 10 different concentrations of spiked samples.
  • Native Sample Linearity: 12 different concentrations of native samples.
  • Provenance: High pool for spiked samples prepared by spiking normal plasma with VWF concentrate. Low pool with VWF deficient plasma (internally produced). Native samples from contract blood plasma provider. Locations not specified but implied to be part of the testing facilities.

• Analytical Specificity/Interference (Test Sets):

  • Endogenous Interference: 4 VWF activity levels (low, 2 medical decision levels, high) tested with 5 test samples each. 80 measurements overall (4 VWF levels x 5 test samples x 4 replicates). Each sample prepared with native plasma and dilution/spiking.
  • Exogenous Interference: 4 VWF activity levels for each of the 30 interferent drugs. 32 measurements per interferent (4 VWF levels x 4 aliquots x 2 conditions - spiked/control).
  • Provenance: Samples prepared with native plasma and VWF deficient plasma or VWF concentrate.

• High Dose Hook (Test Set): 10 different dilution samples from a high VWF Ac activity plasma pool. Measured in 6 replicates.

  • Provenance: High pool prepared by spiking normal plasma with VWF concentrate, diluted with VWF deficient plasma.

• Traceability (Test Set): Three vials of WHO 6th International Standard for Factor VIII / von Willebrand Factor. Measured in single determination.

• Stability (Test Sets):

  • Reagent/Calibrator Shelf-Life: 5 plasma pools + 3 control materials. Measured 6-12 replicates at each of 6 time points over 15 months.
  • On-Board Stability: 5 plasma pools + 3 control materials. Measured 6 replicates at each of 5 time points.
  • Once Opened Stability: 5 plasma pools + 3 control materials. Measured 6-12 replicates at multiple time points (2,4,6 weeks for reagent I/II; 6,12,18 weeks for reagent III).
  • Reconstituted Stability (Calibrator): Aliquots of Standard Human Plasma (SHP) measured 6-12 replicates at various time points and conditions.
  • Ambient Temperature: 8 test samples. Each tested in 8 replicates on 3 days.
  • Transportation Study: 5 plasma pools + 3 control materials. Tested 3-6 replicates at various temperature stress conditions.
  • Freeze/Thaw Tolerance: 5 plasma pools + 3 control materials. Tested 3 replicates under stressed and unstressed conditions.
  • Sample Stability: At least 20 samples. Tested in quadruplicate over various time points and storage conditions.
  • Frozen vs. Fresh Samples: 60 fresh samples. Measured one replicate fresh, then one replicate after frozen storage.
  • Provenance: Plasma pools, control materials, patient samples. Specific origins (e.g., country) not detailed for all stability studies, but likely conducted at manufacturer's internal labs or controlled external sites.

• Detection Limit (Test Sets):

  • LoB: 5 analyte-free samples. n=60 determinations per reagent lot per instrument.
  • LoD: 5 low-analyte samples. n=50 determinations per sample.
  • LoQ: 5 low-analyte samples. n=60 determinations per reagent lot per instrument.
  • Provenance: Samples prepared with VWF deficient plasmas and dilutions of normal plasma pools.

• Normal Range / Reference Interval (Test Set):

  • Adults: 302 apparently healthy individuals (150 blood group O, 152 blood group non-O) ≥ 18 years of age.
  • Pediatrics: 85 apparently healthy pediatric individuals (44 blood group O, 41 blood group non-O) > 4 weeks to < 18 years of age.
  • Provenance: Samples collected retrospectively/prospectively at three clinical sites in the U.S. at different geographic locations (for adults) and two U.S. clinical sites (for pediatrics).

• Diagnostic Accuracy (Method Comparison) (Test Set):

  • Vs. BC von Willebrand Reagent (current study): 144 samples (17 fresh, 124 frozen, 3 diluted). 102 samples included in statistical evaluation after filtering for AMI.
  • Vs. HemosIL VWF: 144 samples (17 fresh, 124 frozen, 3 diluted). 97 samples included in statistical evaluation after filtering for AMI.
  • Vs. BC von Willebrand Reagent (Patzke et al., 2014 re-analysis): 618 samples initially; 580 samples (29 diluted, no spiked) included in statistical evaluation.
  • Provenance: Patient samples from the U.S. (for current studies) and Germany, Switzerland, and USA (for Patzke et al. re-analysis). Patient cohort included those with suspected/confirmed VWD, hemophilia A, platelet dysfunction, and without VWD diagnosis.

• Diagnostic Concordance (Test Set): 138 patients (108 tested against Beckman Coulter vWF Reagent, 30 against HemosIL VWF Activity Reagent).

  • Provenance: Not explicitly stated but implies patient samples from clinical settings. Likely U.S. based on regulatory context.

3. Number of experts used to establish the ground truth for the test set and qualifications of those experts:

The device is an in-vitro diagnostic assay that measures a quantitative biomarker (VWF activity). The "ground truth" for several test sets, such as precision, linearity, interference, and detection limits, is based on the inherent properties of the samples themselves (e.g., known concentrations of VWF, presence of interferents) and measurements against a reference standard (WHO IS).

For the Reference Interval studies, "apparently healthy individuals" were used. Their healthy status would be determined by clinical criteria, likely involving medical professionals (physicians, nurses) at the collecting sites, but specific "experts" for ground truth adjudication are not mentioned as it's a physiological range determination.

For Diagnostic Accuracy (Method Comparison), the ground truth is established by comparing the device's performance against another established and cleared VWF activity assay (predicate devices: BC von Willebrand Reagent, HemosIL VWF). The performance of these predicate devices, in turn, has been established through their own regulatory clearance processes and clinical use. No direct "expert consensus" on individual patient cases for ground truth is explicitly described for method comparison. The patient cohort included those with previously diagnosed VWD, suggesting their diagnosis was established by treating physicians using standard clinical and laboratory methods.

For Diagnostic Concordance, the "ground truth" is defined by the "study site specific standard of care (SOC) VWF-activity assay" results in conjunction with other VWD testing panel results (VWF:Ag, FVIII activity). This is a comparison against established clinical practice rather than a 1:1 expert adjudicated ground truth for each case.

Therefore, direct explicit statements about "number of experts" and their "qualifications" for establishing individual patient-level ground truth (as might be seen in imaging studies) are generally not applicable or explicitly provided in this type of IVD submission for quantitative assays where the truth is often defined by:

• Reference to an international standard (WHO IS for VWF).
• Comparison to a legally marketed predicate device.
• Known concentrations in spiked/diluted samples.
• Clinical status (e.g., "apparently healthy individuals").

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

Given that this is an in vitro diagnostic quantitative assay for a biomarker, the "adjudication method" in the sense of multiple experts reviewing and reaching consensus on findings (like in an imaging study) is not explicitly detailed or typically performed.

• Analytical Studies (Precision, Linearity, Interference, Detection Limits, Stability, Carryover): These studies involve objective measurements according to standardized protocols (CLSI guidelines). Data analysis is statistical, not based on expert adjudication of individual results.
• Reference Interval Studies: These involve collecting data from a "healthy" population. The "ground truth" of their healthy status would be confirmed by clinical screening, not through formal adjudication of individual VWF results by multiple experts.
• Diagnostic Accuracy (Method Comparison): This involves statistical comparison (e.g., Passing-Bablok regression) of the new device's results against a predicate device. Discrepancies are analyzed statistically, not through expert adjudication of individual patient results.
• Diagnostic Concordance: This study compares the diagnostic outcome classification (e.g., "any type of VWD," "VWD excluded") from the subject device to the "site-specific standard of care" diagnostic classification. While the SOC classification would have involved clinical expertise, the comparison itself is statistical, not an adjudication process.

Therefore, for this device, the adjudication method described is none in the traditional sense of expert consensus on individual cases. The ground truth relies on established methodologies, reference standards, and predicate device performance.

5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

No MRMC study was done. This device is an in vitro diagnostic assay for measuring a biomarker (von Willebrand factor activity) in a blood sample. It is an automated test performed by an instrument, not a device that assists human readers (e.g., radiologists) in interpreting medical images or other complex data. Therefore, the concept of "human readers improving with AI assistance" does not apply to this product.

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

The INNOVANCE VWF Ac assay is an automated, standalone in-vitro diagnostic device. It provides quantitative VWF activity results on the BCS XP System. This means the analytical and diagnostic performance data presented (precision, linearity, interference, detection limits, method comparison, diagnostic concordance percent agreements) represents the algorithm/device-only performance without direct human intervention in interpreting the VWF activity value generated by the instrument for each sample. Clinicians then interpret these quantitative results in conjunction with other clinical information to make a diagnosis.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

The "ground truth" utilized varies depending on the specific study type:

• Analytical Performance (Precision, Linearity, Interference, Detection Limits, Stability, Carryover):

  • Known concentrations: For linearity and detection limits, samples were created with known concentrations of VWF by dilution or spiking.
  • Reference materials: For traceability, the WHO 6th International Standard for Factor VIII / von Willebrand Factor was used as the ground truth.
  • Absence of interferents/analytes: For interference and limit of blank, samples were prepared to be analyte-free or with defined amounts of interferents.
  • "Healthy" population status: For reference intervals, individuals identified as "apparently healthy" through clinical screening were used.

• Diagnostic Accuracy (Method Comparison):

  • Predicate Device Results: The results of established, legally marketed VWF activity assays (BC von Willebrand Reagent and HemosIL VWF) served as the comparator or "ground truth" for evaluating the performance of the new device.

• Diagnostic Concordance:

  • Site-Specific Standard of Care (SOC) Diagnosis: The ground truth was based on the diagnostic classification ("any type of VWD," "VWD excluded") determined by the study site's existing clinical practice, which involved their SOC VWF-activity assay along with other elements of the initial VWD testing panel (VWF:Ag, FVIII activity). This is a form of clinical consensus based on established standard diagnostic panels.

• Training Set Ground Truth: (See question 9 below, as this information is not explicitly provided in this document but would be distinct from the test set ground truth).

8. The sample size for the training set:

The document does not explicitly state the sample size for a training set. This is common for this type of IVD, where the analytical performance is established through rigorous validation studies against reference methods and spiked samples, rather than relying on machine learning models that require distinct training and test sets in the same way imaging AI does. The device's foundational principle is particle-enhanced turbidimetry, not a machine learning algorithm requiring a separate training pipeline for its core function.

9. How the ground truth for the training set was established:

As the document does not explicitly mention a "training set" in the context of machine learning, the concept of establishing ground truth for such a set is not applicable here. The INNOVANCE VWF Ac assay is a biochemical assay. Its development and optimization would rely on established laboratory principles, analytical validation, and calibration against reference standards, rather than a machine learning training process with a ground truth dataset. The calibration process uses "Standard Human Plasma" (SHP) which is traceable to the WHO 6th International Standard. This calibration material, and its assigned values, would be the closest analogue to "ground truth" used in setting up the device for accurate measurements.

Ask a specific question about this device