(225 days)
The PENTRA XLR is a quantitative multi-parameter automated hematology analyzer for in vitro diagnostic use in clinical laboratories to identify and enumerate the following parameters: WBC, RBC, HCT, MCV, MCH, MCHC, RDW, PLT, MPV, LYM (#, %), MON (#, %), NEU (#, %), EOS (#, %), BAS (#, %), as well as the Reticulocyte parameters RET (#, %), CRC, and IRF in K2EDTA and K3EDTA anticoagulated venous whole blood samples from patients ≥ 18 years of age.
The PENTRA XLR is a device modification to the ABX PENTRA 80 (K024002), which consists in the addition of the measurement of Reticulocytes (RET) parameters to the existing Complete Blood Count (CBC) and Differential (DIFF) counts already performed by the analyzer. The additional RET mode, based on the use of Thiazol orange reagent and fluorescence detection by optical bench, is independent from the modes already existing on the ABX PENTRA 80: the CBC and the CBC+DIFF modes, as there have been no modifications to the existing mechanical and analytical portions of the original device.
Here's an analysis of the acceptance criteria and study details for the HORIBA ABX SAS PENTRA XLR device, based on the provided text:
Important Note: The provided text is a 510(k) Summary, which typically focuses on demonstrating substantial equivalence to a predicate device rather than presenting a full, independent validation study in the format one might expect for a research paper. Therefore, some details like specific expert qualifications or adjudication methods for ground truth might not be explicitly stated for all aspects if they were part of the predicate device's original validation or are assumed by industry standards for this type of device.
1. Table of Acceptance Criteria and Reported Device Performance (Reticulocyte Parameters)
The document focuses on the newly added Reticulocyte (RET) parameters for the PENTRA XLR. Performance for existing CBC and DIFF parameters is stated to be unchanged from the predicate device (ABX PENTRA 80, K024002).
| Performance Metric | Parameter | Acceptance Criteria (PENTRA XLR Specified) | Reported Device Performance (PENTRA XLR) |
|---|---|---|---|
| Analytical Sensitivity (LoQ) | RET% | N/A (Lowest value with r² > 0.98 in linearity) | 0.7% |
| RET# | N/A (Lowest value with r² > 0.98 in linearity) | 0.01 x 10⁶/mm³ | |
| Precision (Repeatability) - CV% | RET% (Normal Level) | 0.95 for RET% and RET# stated in comparison table) | |
| Anticoagulant Comparability | Bias at Low, Mid, High points | Acceptance criteria met | Met for all measurands at all levels (K2EDTA vs K3EDTA) |
| RBC Mode Comparability | Bias at Low, Mid, High points | Acceptance criteria met | Met for RBC in RET vs CBC modes |
| Reference Interval | (Specific ranges for male/female) | Defined reference interval for 95% of values, 90% CI for limits | Established with 122 female and 120 male samples |
2. Sample Size Used for the Test Set and Data Provenance
- Precision (Repeatability): Minimum of 10 normal and 5 abnormal fresh whole blood samples collected into K2EDTA. Data Provenance: Fresh whole blood samples, collected at 3 different clinical sites (implicitly prospective).
- Precision (Reproducibility): One single lot of control material (Minotrol Retic).
- Linearity / AMR: Commercial Linearity kits.
- Carryover: Alternating high and low concentration samples.
- Interfering Substances:
- Addition Method: Samples with potential interfering substances added vs. control portions.
- Comparison Method: 10 to 20 patient specimens per group with known or potential reticulocyte interferents, and control samples.
- Sample Stability: 10 whole venous blood specimens collected in K2EDTA. Data Provenance: US, implicitly prospective.
- Comparability with Predicate Device: Total of 376 whole blood specimens collected in K2EDTA from adult patients. Data Provenance: US, from three test sites, collected prospectively or not (retrospective/prospective mix).
- Comparability between Anticoagulant types: Total of 90 normal and pathological blood specimens. Data Provenance: US, from two sites, specifically prospectively collected for this study.
- Comparability between CBC and RET modes for RBC: Total of 107 normal and pathological blood specimens collected in K2EDTA. Data Provenance: France, from one site.
- Reference Interval: 242 normal adult samples (122 female and 120 male) collected in K2EDTA. Data Provenance: US, from two test sites.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications
The document does not explicitly mention the use of experts to establish ground truth in the traditional sense (e.g., for image interpretation or diagnosis). For an automated hematology analyzer measuring quantitative parameters, the "ground truth" for method comparison and accuracy studies is typically a reference measurement procedure, sometimes a manual count (e.g., manual reticulocyte count) or a highly accurate predicate device or method.
In this case:
- Comparability with Predicate Device: The ABX PENTRA DX 120 (K991839) served as the reference method for evaluating the new RET parameters. This is an existing FDA-cleared automated hematology analyzer.
- Interfering Substances (Comparison Method): ABX PENTRA DX 120 (K991839) was also used as the comparative measurement procedure.
Since the device evaluates quantitative hematological parameters, human expert visual review (e.g., microscopic differentiation) or adjudication is typically not the primary "ground truth" for the measured values themselves, which are generated parametrically. Ground truth for diagnosis based on these values might involve experts, but that's beyond the scope of device performance.
4. Adjudication Method for the Test Set
No explicit adjudication method (like 2+1 or 3+1 consensus) is described or would typically be applicable for the direct measurement of quantitative hematology parameters by an automated analyzer, as outlined above. The evaluation relies on statistical comparison against a reference method or against established internal specifications.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No MRMC comparative effectiveness study was done. This type of study is more common for diagnostic devices where human readers interpret outputs (e.g., radiographs, pathology slides). The PENTRA XLR is an automated analyzer producing quantitative values, not interpretations requiring human reading.
6. Standalone (Algorithm Only Without Human-in-the-Loop Performance)
Yes, the studies described are essentially standalone performance studies for the algorithm (the automated analyzer). There is no "human-in-the-loop" component described for the measurement of the parameters. The output is a direct quantitative reading from the instrument.
7. Type of Ground Truth Used
The ground truth used depends on the specific study:
- Analytical Sensitivity, Precision, Linearity, Carryover: Internal specifications, commercial linearity kits with "expected values" considered "true values," and statistical analysis against these benchmarks.
- Interfering Substances: Comparison against a predicate device (ABX PENTRA DX 120) and acceptable bias limits.
- Sample Stability: Comparison against an initial measurement (T0) of the same sample.
- Comparability with Predicate Device: Measurements from the legally marketed predicate device, ABX PENTRA DX 120 (K991839).
- Comparability between Anticoagulant types & RBC Mode Comparability: Statistical comparison between measurements obtained from different conditions on the PENTRA XLR itself.
- Reference Interval: Statistical analysis of normal adult samples according to CLSI EP28-A3 guidelines.
8. The Sample Size for the Training Set
The document does not provide a specific sample size for a "training set." This is an automated diagnostic device, not a machine learning model in the contemporary sense that would undergo explicit "training" with a labeled dataset in the way a deep learning algorithm might. The device's "training" and calibration would be part of its engineering development, using internal standards and optimization, rather than a publicly reported "training set." The focus of this 510(k) is on the validation and verification of its analytical performance.
9. How the Ground Truth for the Training Set Was Established
As noted above, a distinct "training set" with established ground truth as understood in machine learning contexts is not directly applicable or described here. The instrument's operational principles are based on established impedance and fluorescence cytochemistry, not a learned model from a large, expert-labeled dataset for classification/detection. Therefore, the concept of "ground truth for the training set" isn't explicitly addressed in this type of submission for this kind of device.
§ 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.”