Search Results

EasySep Human Bone Marrow CD138 Positive Selection Kit is an in vitro diagnostic device intended to enrich CD138+ cells from bone marrow collected from patients diagnosed with multiple myeloma. The CD138+ cells are enriched by immunomagnetic positive selection for use in validated downstream assays. The end-user is responsible for validation of this kit for use with the assay. For in vitro diagnostic use by laboratory professionals.

Device Description

The EasySep CD138 Human Bone Marrow CD138 Positive Selection Kit contains reagents to enrich CD138 positive (CD138+) human bone marrow cells. The kit contains reagents for up to 60 mL of bone marrow purifications. The kit contents are listed in Table 1. Materials required but not provided are shown in Table 2.

AI/ML Overview

Acceptance Criteria and Device Performance for EasySep Human Bone Marrow CD138 Positive Selection Kit

This document summarizes the acceptance criteria and a detailed account of the studies conducted to demonstrate the EasySep Human Bone Marrow CD138 Positive Selection Kit meets these criteria.

1. Table of Acceptance Criteria and Reported Device Performance

Parameter	Acceptance Criteria	Reported Device Performance
Reproducibility with Contrived Samples	Within Lot (across site) CV for Medium/High CD138+ panel members: < 15% Within Lot (across site) CV for Low CD138+ panel members: < 35% Within Site (across lot) CV for Medium/High CD138+ panel members: < 15% Within Site (across lot) CV for Low CD138+ panel members: < 35%	Within Lot (across site) CV: - High (>15%): 4.0% (2.8 - 6.0%) - Med (3-15%): 4.5% (2.2 - 7.0%) - Low (<3%): 11.9% (8.2 - 18.3%) Within Site (across lot) CV: - All medium and high CD138+ panel members: < 15% - All low CD138+ panel members: < 35% Fold Enrichment: - Low CD138+: 85.7-fold (20.2-206.0) - Med CD138+: 9.3-fold (5.0-15.4) - High CD138+: 3.8-fold (2.6-5.5) Purity (Enriched Mean): - Low CD138+: 56.2% - 79.6% - Med CD138+: 85.5% - 92.6% - High CD138+: 88.5% - 94.0%
Reproducibility with Clinical Specimens	Normal/Abnormal disposition: Concordant between kit lots for a given FISH probe.	For all 9 BMA tested, the percentage of abnormal cells for each probe were very similar between kit lots and the same normal/abnormal disposition for a given probe was concordant between the pairs of EasySep Human Bone Marrow CD138 Positive Selection Kit lots.
Repeatability	Flow Cytometry Analysis (enriched CD138+ purity CV): - Low initial CD138+ frequency: -- Repeatability: N/A (Implicitly met by overall within-laboratory) -- Within-laboratory: < 35% - Medium initial CD138+ frequency: -- Repeatability: N/A (Implicitly met by overall within-laboratory) -- Within-laboratory: < 15% FISH Analysis: All enriched samples for low and medium initial CD138+ frequency meet abnormal FISH patterns.	Flow Cytometry Analysis (enriched CD138+ purity CV): - Low initial CD138+ frequency: -- Repeatability: 0.81% to 1.56% -- Within-laboratory: 0.84% to 1.77% - Medium initial CD138+ frequency: -- Repeatability: 0.17% to 1.41% -- Within-laboratory: 0.65% to 0.96% FISH Analysis: All enriched samples sent for FISH analysis (n=3 per panel member, one replicate per run) gave abnormal FISH patterns.
Analytical Specificity (Interference - Excess Heparin)	Enriched CD138+ purity from test samples (with excess heparin) should be > 90% of enriched CD138+ purity from control samples (without excess heparin). Statistical difference: Not statistically different from control specimens (for low frequency specimens).	Three clinical specimens (initial 0.4%, 5.0%, 31.3%): - Specimen #1: 91.0% of control - Specimen #2: 103% of control - Specimen #3: 98.0% of control Three low frequency clinical specimens (initial <0.2%): - Specimen #1: 99% of control - Specimen #2: 100% of control - Specimen #3: 100% of control - P-value for one-tailed t-test: 0.2740 (not statistically different).
Specimen Stability (FISH)	All enriched samples should have conclusive FISH results. Each of the five FISH probes should result in the same FISH disposition (normal or abnormal) across time points (baseline, 44-54h, ~73h) when initial CD138 frequencies were above the putative LLoD for the flow assay (0.14%).	Specimen #1 (initial 10.6%): FISH results for three probes analyzed did not change across time points (baseline -> 73h). Specimen #2 (initial 0.5%): FISH results consistent with baseline and 44-54h. More variable (~73h) due to insufficient cells. Specimen #3 (initial 0.4%): FISH disposition for three probes analyzed consistent across time points (baseline -> ~73h). Overall, FISH disposition was consistent between baseline and ~73h for all three specimens when initial CD138 frequencies were above the putative LLoD for the flow assay.
Kit Stability (Shelf-Life and In-Use)	Ability to enrich CD138+ cells to a purity of >95% and >70% recovery.	Met all acceptance criteria at long-term and in-use storage (5°C ± 3°C) for up to 7 months, and accelerated storage (25°C ± 2°C/60% ± 5% RH) for up to 7 days. Supports a 6-month shelf life at 5°C ± 3°C.
Kit Stability (Shipping)	CD138+ cell purity had to be >95% with at least 70% of cells recovered after exposure to summer and winter temperature profiles for 48 hours.	All acceptance criteria were met at the long-term storage condition of 5°C ± 3°C for up to 7 months, demonstrating product performance unaffected by transport stress. Supports a 6-month shelf-life following transport stress.
Limit of Detection (Contrived Samples)	Lowest initial CD138+ frequency that gives consistent abnormal FISH results in a contrived specimen after CD138+ cell enrichment.	Consistently achieved abnormal t(11;14) disposition in FISH analysis for contrived specimens targeting as low as 0.08% initial CD138+ frequency. Final LoD set at 0.05% CD138+ (limit of detection of the flow cytometer).
Limit of Detection (Clinical Samples)	Overall enriched purity CV ≤ 35%. Enriched samples' FISH disposition same as undiluted clinical specimen's FISH disposition for all probes.	Specimen enriched from 0.05% to 4.1% - 7.4%. Overall enriched purity CV was 19.7% (meeting ≤35%). All enriched samples had FISH disposition identical to the undiluted clinical specimen (100% abnormal for Chromosome 13q deletion and TP53).
Clinical Study (FISH)	Increased percentage of abnormal nuclei in enriched samples compared to unenriched samples. Successful enrichment of a range of bone marrow CD138+ cells (low, intermediate, and high) with viable cells for FISH.	CD138 purity increased from 12.9% ± 15.9% (pre-enrichment) to 79.6% ± 19.9% (post-enrichment). Fold enrichment > 1 for all specimens with CD138 frequencies <40% prior to enrichment. In 14 patients, abnormal nuclei were only detected after enrichment. For all patients, the percentage of abnormal nuclei detected was higher in the enriched sample compared to the unenriched sample. Demonstrates utility for detection of multiple FISH probes.

2. Sample Size Used for the Test Set and Data Provenance

Reproducibility with Contrived Samples:
- Test Set: 16 panel members (contrived samples at low, medium, and high CD138+ levels). More panel members (6 each) at low and medium levels, and 4 at the high level.
- Data Provenance: Not explicitly stated, but "healthy donor whole blood" was used as a matrix for spiking, implying healthy human blood products. The study was performed at "three independent sites," suggesting multi-site internal validation rather than external country-specific data. Retrospective/prospective not specified for sample collection, but the study design is prospective.
Reproducibility with Clinical Specimens:
- Test Set: 9 clinical Multiple Myeloma (MM) bone marrow aspirates (BMA). Each split to test two different kit lots.
- Data Provenance: Multiple Myeloma patient bone marrow aspirates. The document does not specify countries of origin, but it is implied to be clinical samples from patients. A prospective study design for testing these samples is implied.
Repeatability:
- Test Set: 6 panel members (contrived samples) generated by spiking MM cell line SK-MM-2 into healthy donor bone marrow. Two panel members (one low, one medium frequency) were generated each day over 3 non-consecutive days, resulting in 6 unique panel members for the study. Each operator performed 3 runs of enrichments in quadruplicate (n=12 total replicates per panel member for flow cytometry, n=3 selected replicates for FISH).
- Data Provenance: "Healthy donor bone marrow" and "CD138+ cell line SK-MM-2." Similar to reproducibility, likely internal multi-site data. Prospective study design.
Analytical Specificity (Interference - Excess Heparin):
- Test Set (Study 1): 3 fresh clinical multiple myeloma bone marrow specimens.
- Test Set (Study 2, for low frequency specimens): 3 clinical specimens prepared by adding MM bone marrow mononuclear cells into healthy donor bone marrow at <0.2% initial CD138 frequency.
- Data Provenance: Clinical Multiple Myeloma bone marrow specimens and spiked healthy donor bone marrow. Country of origin not specified. Prospective study design.
Specimen Stability:
- Test Set: ~six MM patient BMAs. Data is shown for 3 specimens specifically.
- Data Provenance: Multiple Myeloma patient bone marrow aspirates. Country of origin not specified. Prospective study design for testing.
Kit Stability (Shelf-Life and In-Use, Shipping):
- Test Set: Three kit lots for shelf-life, one lot for in-use stability, one batch for shipping stability.
- Data Provenance: Internal testing of kit components. Prospective study design.
Limit of Detection (Contrived Samples):
- Test Set: Whole Blood (WB) or Bone Marrow (BM) samples from healthy donors spiked with SK-MM-2 cell line at 9 different serially diluted target percentages (from ~20% down to 0%). Each sample enriched in duplicate.
- Data Provenance: Healthy donor WB or BM and CD138+ cell line SK-MM-2. Likely internal multi-site data. Prospective study design.
Limit of Detection (Clinical Multiple Myeloma Bone Marrow Specimens):
- Test Set (Study 1): 5 low CD138 frequency clinical specimens (pre-enrichment frequencies 0.2%, 0.5%, 0.7%, 1.5%, 1.7%).
- Test Set (Study 2, Verification): 1 clinical specimen produced by spiking MM patient BMA into healthy donor BMA at 0.05% initial CD138 frequency. Split into 6 fractions (2 kit lots, 3 replicates per lot).
- Data Provenance: Clinical Multiple Myeloma bone marrow specimens and spiked healthy donor BMA. Country of origin not specified. Prospective study design.
Clinical Study:
- Test Set: 33 Multiple Myeloma patient bone marrow aspirates.
- Data Provenance: Multiple Myeloma patient bone marrow aspirates. The document refers to "two studies" where these samples were collected but doesn't specify if they were from multiple countries or regions. Prospective analysis of these collected samples.

3. Number of Experts Used to Establish Ground Truth and Qualifications

For all studies involving FISH analysis: "Specimen slides were enumerated by a minimum of two readers."
Qualifications of Experts: The document does not explicitly state the qualifications of the "readers" who enumerated the FISH slides (e.g., "radiologist with 10 years of experience"). Based on the context of FISH analysis, these would typically be trained cytogenetic technologists or pathologists with expertise in FISH interpretation.

4. Adjudication Method for the Test Set

For FISH analysis in the Clinical Study and LoD studies: "Specimen slides were enumerated by a minimum of two readers; 50 nuclei per reader were analyzed, for a total of 100 nuclei. If counts for a reader were below or equal to the cutoff corresponding to each probe, then the result from the reader was determined to be normal and the percentage of abnormal nuclei would be indicated as '0'."
This suggests a form of consensus or independent reading with a pre-defined rule for determining "normal" based on individual reader counts. It is not explicitly stated if there was an adjudication procedure for discordant readings between the two readers, or if a third reader was involved (e.g., 2+1 method).

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

No, an MRMC comparative effectiveness study was not performed as described. The studies focus on the analytical and clinical performance of the device itself (enrichment efficacy, reproducibility, LoD) rather than comparing human reader performance with and without AI assistance. The device is a "Hematopoietic cell enrichment kit," meaning it's a tool for sample preparation, not an interpretive AI algorithm intended to assist human readers directly.

6. Standalone Performance

Yes, a standalone performance was done for the device's function. The entire document details the performance of the "EasySep Human Bone Marrow CD138 Positive Selection Kit" in enriching CD138+ cells. This is the standalone performance of the device (a physical kit for cell separation), not an algorithm. The device's output (enriched cells) is then used in downstream assays (flow cytometry, FISH), which are performed by humans and/or other laboratory instruments. The performance metrics reported (purity, fold enrichment, CV, FISH disposition) are directly attributable to the standalone function of the kit.

7. Type of Ground Truth Used

Flow Cytometry: For quantification of CD138+ cell purity and initial frequencies. This is a widely accepted and quantitative method in immunology.
FISH Analysis: For detecting chromosomal abnormalities. This is a standard cytogenetic method, with "normal/abnormal disposition" and "percentage of abnormal nuclei" serving as the ground truth for genetic status, based on enumeration by expert readers (as described in section 3).
Contrived Samples: For reproducibility and LoD studies, the "ground truth" for the initial CD138+ frequency was established by spiking a known cell line (SK-MM-2) into healthy donor blood/bone marrow. The t(11;14) translocation characteristic of SK-MM-2 cells then serves as the ground truth for "abnormal" FISH results in these contrived samples.
Clinical Specimens: For analytical specificity and clinical studies, the "ground truth" for abnormal FISH patterns (e.g., 13q deletion, t(11;14)) comes from the FISH analysis itself, which is a gold standard for detecting these abnormalities in patient samples, as enumerated by expert readers.

8. Sample Size for the Training Set

Not applicable. This device is a physical kit for cell enrichment, not a machine learning model that requires a "training set." The studies performed are verification and validation studies to demonstrate the kit's performance characteristics.

9. How the Ground Truth for the Training Set Was Established

Not applicable. As the device is not an AI/ML algorithm, there is no "training set" or corresponding ground truth to establish for it in the context of machine learning. The studies described are for analytical and clinical validation of a physical diagnostic device.

Ask a Question

Ask a specific question about this device

Page 1 of 1