(530 days)
The Lumipulse G B-Amyloid Ratio (1-42/1-40) is an in vitro cerebral spinal fluid (CSF) test that combines the results of Lumipulse G B-Amyloid 1-42 and Lumipulse & B-Amyloid 1-40 assays into a ratio of ß-amyloid 1-42 to ß-amyloid 1-40 concentrations using the LUMIPULSE G 1200 System. The Lumipulse G B-Amyloid Ratio (1-42/1-40) is intended to be used in adult patients, aged 55 years and older, presenting with cognitive impairment who are being evaluated for Alzheimer's disease (AD) and other causes of cognitive decline.
A test result ≥ 0.073 is a negative result which is consistent with a negative amyloid positron emission tomography (PET) scan result. A negative result reduces the likelihood that a patient's cognitive impairment is due to AD.
A test result ≤ 0.058 is a positive result which is consistent with a positive amyloid PET scan result. A positive result does not establish a diagnosis of AD or other cognitive disorder.
A test result between 0.059 and 0.072 is considered as a likely positive result as it is more likely consistent with a positive amyloid PET scan result. A likely positive result does not establish a diagnosis of AD or other cognitive disorders and has increased uncertainty in regard to amyloid PET positivity.
The Lumipulse G B-Amyloid Ratio (1-42/1-40) results must be interpreted in conjunction with other patient clinical information .
This test is not intended as a screening or stand-alone diagnostic test.
The Lumipulse G B-Amyloid Ratio (1-42/1-40) is an in vitro cerebral spinal fluid (CSF) test that calculates the ratio of two analytes, Lumipulse G ß-Amyloid 1-42 and Lumipulse & ß-Amyloid 1-40 assays to generate a numeric value between 0.001 to 1.000.
The test system consists of two component assays, Lumipulse & ß-Amyloid 1-42 and the Lumipulse G B-Amyloid 1-40 assay, running on LUMIPULSE G1200 system, and Lumipulse G B-Amyloid Ratio (1-42/1-40) Calculator Tool to calculate the Lumipulse G B-Amyloid Ratio (1-42/1-40). Lumipulse G B-Amyloid 1-42 and the Lumipulse G B-Amyloid 1-40 assays are packed individually. Results of individual assays have not been assessed to support the intended use except for determination of the Lumipulse G ß-Amyloid Ratio (1-42/1-40).
Acceptance Criteria and Device Performance for Lumipulse G ß-Amyloid Ratio (1-42/1-40)
This document describes the acceptance criteria and the study demonstrating the performance of the Lumipulse G ß-Amyloid Ratio (1-42/1-40) device.
1. Acceptance Criteria and Reported Device Performance
The core of the device's utility lies in its ability to predict amyloid PET scan results. The acceptance criteria and reported performance relate to the predictive values of the test for classifying patients as PET positive, likely positive, or negative.
| Performance Metric | Acceptance Criteria (Implied / Demonstrated) | Reported Device Performance (95% CI) |
|---|---|---|
| Predictive Value for Positive Result (Ratio ≤ 0.058) for PET positivity | High positive predictive value to indicate consistency with a positive amyloid PET scan result, reducing the likelihood of false positives. While not explicitly stated as a numerical threshold, the expectation is for a high percentage to support its aid in AD evaluation and to reduce unnecessary PET scanning. The clinical study results establish the demonstrated performance for this category. | 96.6% (171/177) (92.8% - 98.4%) - This indicates a strong correlation between a positive Lumipulse G B-Amyloid Ratio and a positive amyloid PET scan result. |
| Predictive Value for Likely Positive Result (0.059 ≤ Ratio ≤ 0.072) for PET positivity | A predictive value that indicates a higher likelihood of PET positivity than a negative result but with increased uncertainty, differentiating it from a clear positive or negative. The clinical study results establish the demonstrated performance for this category. | 59.1% (13/22) (38.7% - 66.7%) - This result confirms the "likely positive" interpretation, showing a majority are PET positive but with a wider confidence interval and lower value compared to the "positive" category, reflecting increased uncertainty. |
| Predictive Value for Negative Result (Ratio ≥ 0.073) for PET positivity | High negative predictive value to indicate consistency with a negative amyloid PET scan result, reducing the likelihood that cognitive impairment is due to AD. The clinical study results establish the demonstrated performance for this category. | 16.1% (15/93) (10.0% - 24.9%) - This value represents the likelihood of being PET positive given a negative test result. Conversely, this implies a strong negative predictive value for PET negativity (i.e., (93-15)/93 = 83.9% will be PET negative given a negative test result). |
Note on "Predictive Value for Negative Result (Ratio ≥ 0.073) for PET positivity": The table in the provided text for "Clinical Performance" presents the "Predictive Value %" as (Positive (n) / N), where N is the total for that Lumipulse G B-Amyloid Ratio category. Therefore, for the "Negative" category, "16.1% (15/93)" means that out of 93 patients with a negative Lumipulse G B-Amyloid Ratio, 15 (16.1%) were actually PET positive. This implicitly means that 93 - 15 = 78 (83.9%) were indeed PET negative when the Lumipulse test was negative.
2. Sample Size for Test Set and Data Provenance
- Sample Size for Test Set: 292 patients.
- Data Provenance: The data for the clinical performance study (test set) was obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) sample bank.
- Country of Origin: Not explicitly stated, but ADNI is a large-scale North American-based research study.
- Retrospective or Prospective: The study utilized "banked CSF samples," indicating a retrospective approach to sample collection for the purpose of this device's validation. However, the PET evaluation was conducted, and the time interval between CSF sampling and PET evaluation was analyzed, suggesting that while samples were banked, their correlation to PET was part of the study design.
3. Number of Experts and Qualifications for Ground Truth
- Number of Experts: A minimum of three trained independent readers.
- Qualifications of Experts: The experts were "trained independent readers" who were blinded to all other clinical data. Their specific professional qualifications (e.g., radiologist, neurologist) or years of experience are not explicitly stated in the provided text.
4. Adjudication Method for the Test Set
- Adjudication Method: 2+1 adjudication method. The amyloid PET status for each patient was determined by a minimum of three trained independent readers. If the first two readers disagreed, an adjudicator's reading was used.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
- Was an MRMC comparative effectiveness study done? No, the provided text does not describe an MRMC comparative effectiveness study comparing human readers with AI assistance versus human readers without AI assistance (i.e., the effect size of how much human readers improve with AI vs without AI assistance). The device is an in-vitro diagnostic (IVD) assay that produces a numerical ratio result, not an AI assisting human image interpretation.
6. Standalone Performance Study
- Was a standalone (algorithm only without human-in-the loop performance) done? Yes, the clinical performance study evaluates the performance of the Lumipulse G B-Amyloid Ratio (1-42/1-40) device in isolation against the ground truth (visual amyloid PET read). The device generates a numerical ratio which is then categorized (positive, likely positive, negative) and compared directly to the PET results, without direct human intervention in interpreting the device's output or integrating it into a diagnostic workflow with human readers.
7. Type of Ground Truth Used
- Type of Ground Truth: The ground truth used was visual amyloid PET scan results. This was determined by "Florbetapir (18F) PET evaluation," with images analyzed by visual read and scored as either amyloid PET positive or amyloid PET negative by trained independent readers with adjudication.
8. Sample Size for the Training Set
- Sample Size for Training Set: 235 patients were used to determine the assay cutoff. This group was from the Amsterdam Dementia Cohort (ADC). It is important to note that this was explicitly stated as "distinct from the subjects ... evaluated in the pivotal clinical validation study (test set)."
9. How the Ground Truth for the Training Set was Established
- How Ground Truth for Training Set was Established: For the 235 patients from the Amsterdam Dementia Cohort (ADC) used to determine the assay cutoff, the ground truth was visual amyloid PET scan status.
- PET Tracers Used: Florbetaben (18F), Florbetapir (18F), and Flutemetamol (18F).
- Interpretation: Images were analyzed by visual read and scored as either amyloid PET positive or amyloid PET negative.
- Protocol: CSF collection, processing, and handling were conducted according to a "standardized ADC protocol." A "pre-analytical bridging study was conducted using fresh prospectively collected individual CSF samples" to account for variations between ADC and ADNI protocols.
N/A