Search Results

The CellSearch® Circulating Tumor Cell Kit is intended for the enumeration of circulating tumor cells (CTC) of epithelial origin (CD45-, EpCAM+, and cytokeratins 8, 18+, and/or 19+) in whole blood. The presence of CTC in the peripheral blood, as detected by the CellSearch® Circulating Tumor Cell Kit, is associated with decreased progression free survival and decreased overall survival in patients treated for metastatic breast, colorectal or prostate* cancer. The test is to be used as an aid in the monitoring of patients with metastatic breast, colorectal or prostate cancer. Serial testing for CTC should be used in conjunction with other clinical methods for monitoring metastatic breast, colorectal and prostate cancer. Evaluation of CTC at any time during the course of disease allows assessment of patient prognosis and is predictive of progression free survival and overall survival.

*Metastatic prostate cancer patients in this study were defined as having two consecutive increases in the serum marker PSA above a reference level, despite standard hormonal management. These patients are commonly described as having androgen-independent, hormone-resistant, or castration-resistant prostate cancer.

The CellSearch® Circulating Tumor Cell Kit contains a ferrofluid-based capture reagent and immunofluorescent reagents. The ferrofluid reagent consists of small particles with a magnetic core surrounded by a polymeric layer coated with antibodies targeting the EpCAM antigen for capturing CTC. After immunomagnetic capture and enrichment, fluorescent reagents are added for identification and enumeration of CTC. The fluorescent reagents include the following: anti-CK-Phycoerythrin (PE) specific for the intracellular protein cytokeratin (characteristic of epithelial cells), DAPI, which stains the cell nucleus, and anti-CD45-Allophycocyanin (APC) specific for leukocytes.

The reagent/sample mixture is dispensed by the CellTracks® AutoPrep® System into a cartridge that is inserted into a MagNest® cell presentation device. The strong magnetic field of the MagNest® device attracts the magnetically labeled epithelial cells to the surface of the cartridge. The CellTracks Analyzer II® or CellSpotter® Analyzer automatically scans the entire surface of the cartridge, acquires images and displays any event to the user where CK-PE and DAPI fluorescence are co-located. Images are presented to the user in a gallery format for final classification. An event is classified as a tumor cell when its morphological features are consistent with that of a tumor cell and it exhibits the phenotype EpCAM+, CK+, DAPI+ and CD45-.

The provided document (K103502) describes a 510(k) submission for a modified CellSearch® Circulating Tumor Cell Kit. However, it does not contain a study with acceptance criteria and a detailed report of device performance against those criteria in the way typically expected for a new or significantly modified device efficacy trial.

Instead, this 510(k) is a Special 510(k), which is submitted when there are minor changes to a legally marketed device that do not affect the intended use, fundamental scientific technology, mode of operation, or specimen type. In this particular case, the changes were labeling changes to assist the operator.

Therefore, the submission demonstrates equivalence to the predicate device (K073338) through verification of labeling specifications for the modified device and labeling process validation, rather than through a new clinical study with performance metrics against acceptance criteria.

Due to the nature of this particular 510(k) submission, many of the requested details about a clinical study, such as sample size, ground truth establishment for test and training sets, expert adjudication, and MRMC studies, are not present in the provided text.

Here's a breakdown of the available information:

1. A table of acceptance criteria and the reported device performance:

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

• Not applicable / Not provided. The submission focused on equivalence based on labeling and process validation, not a new clinical performance study.

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

• Not applicable / Not provided. No new clinical test set requiring expert ground truth was described.

4. Adjudication method for the test set:

• Not applicable / Not provided.

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. This device is for circulating tumor cell enumeration, not an imaging device requiring human reader interpretation in the context of an MRMC study comparing AI assistance. The CellTracks Analyzer II® or CellSpotter® Analyzer presents images to the user for final classification, but the 510(k) does not detail MRMC studies for this specific change.

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

• Not explicitly stated as a new study for this 510(k). The device involves an automated process for capture, enrichment, and initial scanning, with a user performing final classification. The submission focuses on the impact of labeling changes, not on a standalone algorithm performance study.

7. The type of ground truth used:

• Not applicable / Not provided for this specific 510(k). For the original predicate device, the "ground truth" for CTC identification would likely be based on the biological markers detected (EpCAM+, CK+, DAPI+, CD45-) and morphological features recognized by trained operators. However, this 510(k) does not present new performance data.

8. The sample size for the training set:

• Not applicable / Not provided. No new algorithm training was described for this submission.

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

• Not applicable / Not provided.

In summary: K103502 is a Special 510(k) for minor labeling changes to an already cleared device. It establishes substantial equivalence by demonstrating that these changes do not affect the device's fundamental performance, rather than by presenting new clinical performance data from a dedicated study with specific acceptance criteria and detailed methodologies for establishing ground truth, sample sizes, or reader studies. The "study" proving the device meets acceptance criteria in this context is the verification of labeling specifications and validation of the labeling process, demonstrating that the modified device remains functionally identical to the predicate.

Ask a specific question about this device

The CellSearch™ Circulating Tumor Cell Kit is intended for the enumeration of circulating tumor cells (CTC) of epithelial origin (CD45-, EpCAM+, and cytokeratins 8, 18+, and/or 19+) in whole blood.

The presence of CTC in the peripheral blood, as detected by the CellSearch™ Circulating Tumor Cell Kit, is associated with decreased progression free survival and decreased overall survival in patients treated for metastatic breast, colorectal or prostate* cancer. The test is to be used as an aid in the monitoring of patients with metastatic breast, colorectal or prostate cancer. Serial testing for CTC should be used in conjunction with other clinical methods for monitoring metastatic breast, colorectal and prostate cancer. Evaluation of CTC at any time during the course of disease allows assessment of patient prognosis and is predictive of progression free survival and overall survival.

*Metastatic prostate cancer patients in this study were defined as having two consecutive increases in the serum marker PSA above a reference level, despite standard hormonal management. These patients are commonly described as having androgen-independent, hormone-resistant, or castration-resistant prostate cancer.

The CellSearch" Circulating Tumor Cell Kit contains a ferrofluid-based capture reagent and immunofluorescent reagents. The ferrofluid reagent consists of nanoparticles with a magnetic core surrounded by a polymeric layer coated with antibodies targeting the EpCAM antigen for capturing CTC. After immunomagnetic capture and enrichment, fluorescent reagents are added for identification and enumeration of CTC. The fluorescent reagents include the following: anti-CK-Phycoerythrin (PE) specific for the intracellular protein cytokeratin (characteristic of epithelial cells), DAPI which stains the cell nucleus, and anti-CD45-Allophycocyanin (APC) specific for leukocytes.

The reagent/sample mixture is dispensed by the CellTracks® AutoPrep® System into a cartridge that is inserted into a MagNest® cell presentation device. The strong magnetic field of the MagNest® device attracts the magnetically labeled epithelial cells to the surface of the cartridge. The CellTracks Analyzer II or CellSpotter® Analyzer automatically scans the entire surface of the cartridge, acquires images and displays any event to the user where CK-PE and DAPI fluorescence are co-located. Images are presented to the user in a gallery format for final classification. An event is classified as a tumor cell when its morphological features are consistent with that of a tumor cell and it exhibits the phenotype EpCAM+, CK+, DAPI+ and CD45 -.

Here's an analysis of the acceptance criteria and the study that proves the device meets them, based on the provided text:

Device: CellSearch™ Circulating Tumor Cell Kit (expanded indications for use in Metastatic Prostate Cancer)

1. Table of Acceptance Criteria and Reported Device Performance:

Notes on Acceptance Criteria: The document does not explicitly state numerical acceptance criteria for most performance metrics. Instead, it presents the results of studies and statistical analyses, implying that the observed performance (e.g., strong linearity, good reproducibility, and statistically significant associations with clinical outcomes) met the internal and regulatory expectations for device approval. For clinical effectiveness, the key acceptance criterion is the demonstration of statistically significant associations between CTC counts and clinical outcomes (PFS and OS), and that CTC provides prognostic information, even in cases of discordance with PSA.

2. Sample Sizes Used for the Test Set and Data Provenance:

• Recovery: 30 samples (5 different spike levels x 5 donors + unspiked control, processed once). Data provenance: Blood samples from a single healthy donor pooled and spiked with cultured breast cancer cells (SK-BR-3), repeated for four additional donors. Not explicitly stated if prospective or retrospective, but likely prospective lab study.
• Linearity/Reportable Range: Same data as Recovery study (30 samples).
• Limits of Detection: Not a separate sample set, derived from Recovery data and system specifications.
• System Reproducibility (Control): N=99 for both low and high controls. Not explicitly stated if prospective or retrospective or country of origin, but implies controlled laboratory conditions over 30 days.
• System Reproducibility (MBC Patient Samples): 163 duplicate blood samples from 47 metastatic breast cancer patients. Data provenance: Not explicitly stated country of origin, appears to be retrospective analysis of samples collected during a clinical study.
• System Reproducibility (MCRC Patient Samples): 1,627 duplicate blood samples from 430 MCRC patients. Data provenance: Not explicitly stated country of origin, appears to be retrospective analysis of samples collected during a clinical study.
• Clinical Study (Metastatic Prostate Cancer Patients):
  • Test set for clinical efficacy: 231 metastatic prostate cancer patients enrolled.
  • PFS/OS analyses:
    • Baseline: 219 patients
    • 2-5 Weeks: 199 patients (PFS), 203 patients (OS)
    • 6-8 Weeks: 141 patients (PFS), 163 patients (OS)
    • 9-12 Weeks: 134 patients (PFS), 149 patients (OS)
    • 13-20 Weeks: 116 patients (PFS), 143 patients (OS)
  • Concordance with PSA: 197, 159, 146, and 138 patients at 2-5, 6-8, 9-12, and 13-20 weeks respectively had paired CTC and PSA data.
  • Data Provenance: Multi-center prospective clinical trial. Country of origin not specified, but typically such trials would involve multiple sites in the US and/or Europe.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications:

The document does not mention the use of experts to establish a "ground truth" for the test set in the traditional sense of image adjudication or pathological review for the clinical study.

• For the technical performance studies (Recovery, Linearity, LOD, Reproducibility), the "ground truth" for cell counts was effectively established by the engineered spiking of known numbers of cells (Recovery, Linearity) or by established analytical methods for the control materials. The CellTracks Analyzer II or CellSpotter Analyzer automatically scans and displays events, with a user performing final classification based on EpCAM+, CK+, DAPI+, CD45- phenotype and morphological features. While this involves human review, the expertise and number of individuals are not specified as a "ground truth" panel.
• For the clinical study (Metastatic Prostate Cancer), the "ground truth" for clinical outcomes (Progression-Free Survival and Overall Survival) was established through:
  • Disease Progression: Determined by the clinical sites using PSA, imaging, and/or clinical signs and symptoms. This implies a standard clinical assessment by the patient's treating physicians and/or study investigators, not a centralized expert panel for ground truth labeling of images.
  • Overall Survival: Determined from the date of death or date of last contact (for those still alive) from the study follow-up. This is an objective outcome.

Essentially, the device itself generates the "CTC count," and the study evaluates how these counts correlate with established clinical endpoints, rather than evaluating the accuracy of the CTC count against a separate "expert ground truth" for each specific scan.

4. Adjudication Method for the Test Set:

• Clinical Outcomes (PFS/OS): As noted above, disease progression was determined by the clinical sites using a combination of PSA, imaging, and clinical signs/symptoms. It is not described as a formal "adjudication panel" in the sense of multiple independent reviewers resolving discrepancies. Survival is an objective endpoint.
• CTC Enumeration: The CellTracks Analyzer II or CellSpotter Analyzer displays images of potential CTC events to a user for "final classification." The document does not detail an adjudication method (e.g., 2+1, 3+1) for these user classifications of CTCs, implying a single user's classification is considered the final count.

5. If a Multi Reader Multi Case (MRMC) Comparative Effectiveness Study Was Done:

No, a Multi Reader Multi Case (MRMC) comparative effectiveness study was not reported. The study focused on the prognostic value of the device's output (CTC counts) as a standalone marker, not on how the device assists human readers in making a diagnosis or prognosis. The device enumerates CTCs, and a human user reviews the images for final classification, but the study design does not involve comparing human readers with and without AI assistance on a case set.

6. If a Standalone Study Was Done (Algorithm only without human-in-the-loop performance):

The device is described as having a human-in-the-loop component where the "CellTracks Analyzer II or CellSpotter® Analyzer automatically scans the entire surface of the cartridge, acquires images and displays any event to the user where CK-PE and DAPI fluorescence are co-located. Images are presented to the user in a gallery format for final classification." Therefore, the reported performance throughout the document (e.g., clinical efficacy) represents the performance of the system including the final human classification, not a purely standalone algorithm.

7. The Type of Ground Truth Used:

• For Technical Performance (Recovery, Linearity): Known spike concentrations of cultured cancer cells.
• For Clinical Efficacy (PFS, OS):
  • Progression-Free Survival (PFS): Clinical determination of disease progression (based on PSA, imaging, and clinical signs/symptoms) or death.
  • Overall Survival (OS): Date of death or last contact.
    These are real-world clinical outcomes and are considered the definitive "ground truth" for prognostic value.

8. The Sample Size for the Training Set:

The document does not explicitly describe a separate "training set" in the context of developing the CellSearch™ Circulating Tumor Cell Kit. The studies described are primarily performance validation and clinical validation studies. The CellSearch system and its CTC enumeration method likely underwent internal development and optimization using various samples, but a formally documented "training set" for an AI/algorithm component is not presented. It refers to a "long run method of NCCLS guideline EP5-A2" for reproducibility with control samples, which implies standardized internal quality control, not an AI training set.

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

Since a formal "training set" for an AI/algorithm in the sense of machine learning is not described, the method for establishing its ground truth is also not detailed. The "ground truth" for the technical aspects of the device's operation (e.g., successful capture and identification of cells) would have been established through microscopy, immunostaining controls, and other laboratory techniques during the device's design and analytical validation.

Ask a specific question about this device

The CellSearch™ Circulating Tumor Cell Kit is intended for the enumeration of circulating tumor cells (CTC) of epithelial origin (CD45-, EpCAM+, and cytokeratins 8, 18+, and/or 19+) in whole blood.

The presence of CTC in the peripheral blood, as detected by the CellSearch™ Circulating Tumor Cell Kit, is associated with decreased progression free survival and decreased overall survival in patients treated for metastatic breast or metastatic colorectal cancer. The test is to be used as an aid in the monitoring of patients with metastatic breast or metastatic colorectal cancer. Serial testing for CTC should be used in conjunction with other clinical methods for monitoring breast and colorectal cancer. Evaluation of CTC at any time during the course of disease allows assessment of patient prognosis and is predictive of progression free survival and overall survival.

The CellSearch" Circulating Tumor Cell Kit contains a ferrofluid-based capture reagent and immunofluorescent reagents. The ferrofluid reagent consists of nanoparticles with a magnetic core surrounded by a polymeric layer coated with antibodies targeting the EpCAM antigen for capturing CTC. After immunomagnetic capture and enrichment, fluorescent reagents are added for identification and enumeration of CTC. The fluorescent reagents include the following: anti-CK-Phycoerythrin (PE) specific for the intracellular protein cytokeratin (characteristic of epithelial cells), DAPI which stains the cell nucleus, and anti-CD45-Allophycocyanin (APC) specific for leukocytes.

The reagent/sample mixture is dispensed by the CellTracks® AutoPrep® System into a cartridge that is inserted into a MagNest® cell presentation device. The strong magnetic field of the MagNest® device attracts the magnetically labeled epithelial cells to the surface of the cartridge. The CellTracks® Analyzer II or CellSpotter® Analyzer automatically scans the entire surface of the cartridge, acquires images and displays any event to the user where CK-PE and DAPI fluorescence are co-located. Images are presented to the user in a gallery format for final classification. An event is classified as a tumor cell when its morphological features are consistent with that of a tumor cell and it exhibits the phenotype EpCAM+, CK+, DAPI+ and CD45 -.

Here's a breakdown of the acceptance criteria and the study details for the CellSearch™ Circulating Tumor Cell Kit, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The document describes several performance characteristics without explicitly stating "acceptance criteria" for each. However, the study results clearly demonstrate the device's performance in these areas.

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

The primary clinical study supporting the expanded indication for use involved Metastatic Colorectal Cancer (MCRC) patients.

• Sample Size (Test Set/Clinical Study): 430 MCRC patients with measurable disease.
  • PFS Analysis: 413 patients for baseline, decreasing for follow-up time points (e.g., 356 for 1-2 weeks, 180 for 13-20 weeks).
  • OS Analysis: 413 patients for baseline, decreasing for follow-up time points (e.g., 357 for 1-2 weeks, 193 for 13-20 weeks).
  • CTC/Imaging Correlation: 366 MCRC patients for "patient-wise" comparison; 815 observations for "observation-wise" comparison.
• Data Provenance:
  • The study was a multi-center prospective, clinical trial.
  • The document does not explicitly state the country of origin, but the context of an FDA 510(k) submission for a US market device implies that the data likely includes US sites.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

The term "ground truth" here refers to the clinical outcomes and radiological assessments used to validate the predictive power of CTC counts.

• Number of Experts: Not explicitly stated as a specific number of individual experts for "ground truth" establishment, but rather implied through standard clinical practice.
• Qualifications of Experts:
  • Treating Oncologists: Determined the method of imaging and treatment for each patient.
  • Certified Radiologist: Performed image interpretation at the participating site using RECIST uni-dimensional criteria.
  • Clinical Data: Survival data (PFS and OS) are objective clinical endpoints collected as standard of care.

4. Adjudication Method for the Test Set

The document does not describe an explicit adjudication method (e.g., 2+1, 3+1) for establishing the "ground truth" from experts for the MCRC clinical study.

• Imaging Interpretation: Performed by a single "certified radiologist at the participating site using RECIST uni-dimensional criteria." This suggests individual expert reads rather than a consensus or adjudication panel for each image.
• Clinical Outcomes (PFS/OS): These are objective endpoints based on clinical assessments (CT scans, clinical signs/symptoms for progression; death for overall survival) monitored over time.

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

No, a multi-reader multi-case (MRMC) comparative effectiveness study comparing human readers with AI assistance versus without AI assistance was not reported in this summary.

The device is an enumerating system for circulating tumor cells, intended to provide a quantitative biomarker, not to assist human readers in interpreting medical images like a CAD system would. The clinical study evaluated the prognostic value of the CTC count itself, as determined by the system.

6. Standalone Performance Study

Yes, a standalone performance study was done. The CellSearch™ Circulating Tumor Cell Kit is designed to enumerate CTCs automatically (with user image verification). The studies on recovery, linearity, limits of detection, and reproducibility directly assess the standalone performance of the algorithm and system in generating CTC counts.

• The system uses CellTracks® AutoPrep® System for capture and fluorescent labeling, and CellTracks® Analyzer II or CellSpotter® Analyzer to scan, acquire images, and display events.
• "Images are presented to the user in a gallery format for final classification." This implies a human-in-the-loop for final classification of events as tumor cells if the automated system generates "events" for review, but the core enumeration process is largely automated and its performance characteristics (recovery, linearity, LOD) are measured on the system's output. The clinical outcome studies then evaluate the prognostic value of these counted CTCs.

7. Type of Ground Truth Used for the Test Set

The ground truth for the clinical effectiveness study in MCRC patients comprised:

• Clinical Outcomes/Survival Data:
  • Progression Free Survival (PFS): Measured from baseline blood draw to diagnosis of progression by CT scans and/or clinical signs and symptoms.
  • Overall Survival (OS): Measured from baseline blood draw to the time of death.
• Radiological Assessment: Disease status (Complete Response, Partial Response, Stable Disease, Progressive Disease) determined by a certified radiologist using RECIST uni-dimensional criteria.

8. Sample Size for the Training Set

The document does not provide information on a specific "training set" sample size for the algorithm within the device.

As this is a 510(k) submission for an expanded indication for an existing device (K062013), the underlying algorithm was likely developed prior to this submission. The summary focuses on the validation of the device's performance for the new indication and its analytical characteristics. If the algorithm uses machine learning components, its training set details are not included in this summary. The current document describes the validation (test) set for the expanded clinical indication.

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

Since no "training set" is explicitly mentioned or detailed for the algorithm's development in this document, the method for establishing its ground truth is not provided. The reported studies primarily focus on the analytical performance and clinical utility of the already developed device.

Ask a specific question about this device

The CellSearch™ Circulating Tumor Cell Kit is intended for the enumeration of circulating tumor cells (CTC) of epithelial origin (CD45-, EpCAM+, and cytokeratins 8, 18+, and/or 19+) in whole blood.

The presence of CTC in the peripheral blood, as detected by the CellSearch™ Circulating Tumor Cell Kit, is associated with decreased progression free survival and decreased overall survival in patients treated for metastatic breast cancer. The test is to be used as an aid in the monitoring of patients with metastatic breast cancer. Serial testing for CTC should be used in conjunction with other clinical methods for monitoring breast cancer. A CTC count of 5 or more per 7.5 mL of blood at any time during the course of the disease is predictive of shorter progression free survival and overall survival.

The CellSearch™ Circulating Tumor Cell Kit contains a ferrofluid-based capture reagent and immunofluorescent reagents. The ferrofluid reagent consists of nanoparticles with a magnetic core surrounded by a polymeric layer coated with antibodies targeting the EpCAM antigen for capturing CTC. After immunomagnetic capture and enrichment, fluorescent reagents are added for identification and cnumeration of CTC. The fluorescent reagents include the following: anti-CK-Phycoerythrin (PE) specific for the intracellular protein cytokeratin (characteristic of epithelial cells), DAPI which stains the cell nucleus, and anti-CD45-Allophycocyanin (APC) specific for leukocytes.

The reagent/sample mixture is dispensed by the CellTracks AutoPrep® System into a cartridge that is inserted into a MagNest® cell presentation device. The strong magnetic field of the MagNest® device attracts the magnetically labeled epithelial cells to the surface of the cartridge. The CellTracks® Analyzer II or CellSpotter® Analyzer automatically scans the entire surface of the cartridge, acquires images and displays any event to the user where CK-PE and DAPI fluorescence are co-located. Images are presented to the user in a gallery format for final classification. An event is classified as a tumor cell when its morphological features are consistent with that of a tumor cell and it exhibits the phenotype EpCAM+, CK+, DAPI+ and CD45-.

The CellSearch™_Circulating Tumor Cell Kit is intended for the enumeration of circulating tumor cells (CTC) of epithelial origin (CD45-, EpCAM+, and cytokeratins 8, 18+, and/or 19+) in whole blood. The presence of CTC in the peripheral blood, as detected by the CellSearch™_Circulating Tumor Cell Kit, is associated with decreased progression free survival and decreased overall survival in patients treated for metastatic breast cancer. The test is to be used as an aid in the monitoring of patients with metastatic breast cancer. Serial testing for CTC should be used in conjunction with other clinical methods for monitoring breast cancer. A CTC count of 5 or more per 7.5 mL of blood at any time during the course of the disease is predictive of shorter progression free survival and overall survival.

ACCECPTANCE CRITERIA

STUDY DETAILS

1. Sample size used for the test set and the data provenance:

   • Recovery and Linearity: 30 samples from 5 healthy donors. Spiked with cultured breast cancer cells (SK-BR-3).
   • Reproducibility (Patient Samples): 163 duplicate samples from 47 patients.
   • Expected Values (Control Subjects): 345 total samples (145 healthy volunteers, 101 women with non-malignant breast disease, 99 women with other non-malignant diseases).
   • Clinical Study (Metastatic Breast Cancer): 177 patients with metastatic breast cancer. Follow-up analyses had varying sample sizes at different time points (e.g., 126 at 3-5 weeks, 138 for imaging comparison).
   • Data Provenance: Retrospective for recovery, linearity, and reproducibility. Prospective, multi-center clinical trial for metastatic breast cancer patients. Country of origin not specified, but the applicant is Veridex, LLC, located in Warren, NJ, USA.

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

   • Clinical Study (Imaging Ground Truth): Two expert radiologists initially, blinded to clinical information. A third independent radiologist adjudicated disagreements. Qualifications are not specified beyond "expert radiologists."

3. Adjudication method for the test set:

   • Imaging Ground Truth:
     • If one radiologist categorized as Indeterminate (I) and the other as Stable Disease (S), Partial Response (PR), or Progressive Disease (PD), the latter radiologist's classification was used (n=11).
     • If both radiologists categorized as Indeterminate (I), the data was not used (n=3).
     • A third independent radiologist adjudicated disagreements between the two primary readers regarding PD and NPD (n=27). If the third radiologist rendered "Indeterminate (I)", the data was not used (n=2).

4. 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:

   • The study did not describe an MRMC comparative effectiveness study involving AI or human readers being assisted by AI. The comparison was betweenCTC measurements and radiological assessments, primarily on their prognostic value and variability. The device is not an AI-assisted diagnostic tool for human readers but rather an automated system for CTC enumeration.

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

   • Yes, the performance of the CellSearch™ system, which includes automated scanning and image acquisition (CellTracks® Analyzer II or CellSpotter® Analyzer) with user classification of events, was evaluated. The "user classification" aspect suggests a human-in-the-loop for final interpretation of events presented by the system, but the core processing and initial identification are automated. The recovery, linearity, limits of detection, and reproducibility studies assess the standalone performance of the system for enumerating CTCs.

6. The type of ground truth used:

   • Recovery & Linearity: Expected cell counts from spiked cultured breast cancer cells.
   • Reproducibility: Comparison of duplicate samples processed by the system.
   • Clinical Study:
     • Prognosis (PFS/OS): Clinical outcomes data (progression-free survival and overall survival) based on CT scans, clinical signs and symptoms, and time to death.
     • Imaging Comparison: Expert consensus of radiologists based on World Health Organization (WHO) bi-dimensional criteria for disease assessment (NPD vs. PD).

7. The sample size for the training set:

   • The document does not explicitly describe a separate "training set" for the device's development or a machine learning algorithm. The studies detailed are primarily validation studies of the device's technical performance and clinical utility.

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

   • Not applicable as no specific training set for a machine learning algorithm is described. The device operates based on immunomagnetic capture and fluorescent labeling with rule-based classification principles, followed by user classification, rather than being a trained AI model in the modern sense.

Ask a specific question about this device

The CellSearch™ Circulating Tumor Cell Kit is intended for the enumeration of circulating tumor cells (CTC) of epithelial origin (CD45-, EpCAM+ and cytokeratin 8 & 18+, and/or cytokeratin 19+) in whole blood in conjunction with the CellTracks® AutoPrep System, the CellSpotter® Analyzer or CellTracks® Analyzer II, and the CellSearch™ Circulating Tumor Cell Control Kit.

The presence of CTC in the peripheral blood, as detected by the CellSearch™ Circulating Turnor Cell Kit, is associated with decreased progression free survival and decreased overall survival in patients treated for metastatic breast cancer. A CTC count of 5 or more per 7.5 mL of blood generally is predictive of shorter progression free survival and shorter overall survival.

Epithelial cells are immunomagnetically labeled by targeting the Epithelial Cell Adhesion Molecule (EpCAM) antigen. Anti-EpCAM monoclonal antibodies conjugated to ferrofluid particles are colloidal and, when mixed with a sample containing the target epithelial cells, bind to the EpCAM antigen associated with the epithelial cells. After immunomagnetic selection of epithelial cells from 7.5 mL of blood. Fluorescent reagents are added at this time to discriminate between the immunomagnetically selected cells. Anti-Cytokeratin - Phycoerythrin (CK-PE) stains the intracellular cytoskeleton cytokeratin proteins expressed in cells of epithelial origin, anti-CD45-Allophycocyan (CD45-APC) stains leukocytes and DAPI stains DNA present in the cell nucleus. A strong magnetic field is applied to the processed reagent/sample mixture that causes the labeled target cells to move to the cartridge surface. The cartridge is then placed on the CellTracks® Analyzer II for data acquisition and analysis. The CellTracks® Analyzer II acquires images of PE, APC and DAPI fluorescence staining of the entire viewing surface.

After data acquisition is completed, the images are analyzed for any event where cytokeratin-PE and DAPI are within a specified space in the cartridge, i.e. indicating the possible presence of a cell with a nucleus that expresses cytokeratin. Images from each fluorescent color as well as a composite image of the cytokeratin staining (green) and the nuclear staining (purple) are presented to the user in a gallery for final cell classification. A cell is classified as a tumor cell when it its EpCAM+ (i.e., it is captured), CK+, DAPI+ and CD45-. A check mark placed by the operator next to the composite images classifies the event as a Circulating Tumor Cell (CTC) and the software tallies all the checked boxes to obtain the CTC count.

Our data demonstrate that metastatic breast cancer patients with 5 or more CTC/per 7.5 mL of blood have a significantly greater probability for shorter progression free and overall survival than patients who have fewer than 5 CTC per 7.5 mL of blood.

Response:

1. Table of Acceptance Criteria and Reported Device Performance

2. Sample Size and Data Provenance

• Test Set (Clinical Study - Metastatic Breast Cancer Patients):

  • Sample Size: 177 patients for baseline analysis (N=177). 154 patients for 1st follow-up PFS analysis (23 not evaluable), 163 patients for 1st follow-up OS analysis (evaluation based on available data, specific N for OS is 163 as per Figure 5 and related text).
  • Data Provenance: A multi-center, prospective, longitudinal clinical trial. The location (country of origin) is not explicitly stated in the provided text, but it is implied to be within the US given the FDA submission.

• Test Set (Non-Clinical Studies - Recovery, Linearity, Reproducibility):

  • Recovery: Blood samples from 5 healthy donors, each with 6 aliquots (5 spiked, 1 unspiked). Total 30 samples analyzed for regression.
  • Linearity/Reportable Range: Data from the recovery study, analyzed as a dilution series (no new samples).
  • Limits of Detection: Based on inherent system capabilities and recovery calculations, not a separate test set.
  • Reproducibility (Control Samples): 99 measurements each for low and high control samples, over 30 days.
  • System Reproducibility (Patient Specimens): 163 duplicate samples collected from 47 patients.

3. Number, Qualifications, and Adjudication of Experts for Test Set Ground Truth

For the non-clinical studies (recovery, linearity, reproducibility, interfering substances), the ground truth was established by in vitro spiking of known numbers of cultured breast cancer cells (SK-Br-3, MCF-7, PC3-9) or by preparing control samples with known cell concentrations. Therefore, no human experts were used for establishing ground truth in these sections.

For the clinical study with metastatic breast cancer patients:

• Number of Experts: Not specified.
• Qualifications of Experts: Not specified. The clinical endpoints were Progression Free Survival (PFS) and Overall Survival (OS), often diagnosed by standardized clinical and imaging assessments. PFS was determined by CT scans and/or clinical signs and symptoms. OS was time to death. These are considered objective outcomes.
• Adjudication Method: Not specified for diagnosis of progression or death. However, standard clinical trial practices usually involve blinded assessment of imaging and clinical events for progression, and death is an unambiguous endpoint.

4. Adjudication Method for the Test Set

As noted above, for the non-clinical studies, ground truth was based on known quantities/concentrations, so no human adjudication was required. For the clinical study, the adjudication method for progression (CT scans and/or clinical signs and symptoms) and death (recorded as an event) is not explicitly detailed in terms of expert consensus or arbitration processes (e.g., 2+1), but implies standard clinical practice for objective endpoints.

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

No MRMC comparative effectiveness study was done. The device, CellSearch™ Circulating Tumor Cell Kit, is an immunomagnetic circulating cancer cell selection and enumeration system. It primarily involves automated cell selection, staining, image acquisition, and then user classification based on presented images. While a user does "classify" the event as a CTC, the study focuses on the analytical performance of the kit and its ability to predict clinical outcomes, rather than a comparative effectiveness study of human readers with or without AI assistance. The summary explicitly states: "All of the above new studies with the CellSearch™ Circulating Cell Kit/ AutoPrep/CellTracks® Analyzer II system demonstrate that the detection of turnor cells by the CellSearch™ Circulating Cell Kit/ AutoPrep/CellTracks® Analyzer II system is substantially equivalent to the oreaticate system."

6. Standalone (Algorithm Only) Performance

Yes, a standalone performance was done for the analytical portions of the device. The CellTracks® Analyzer II acquires images, and then "images are analyzed for any event where cytokeratin-PE and DAPI are within a specified space... A cell is classified as a tumor cell when it its EpCAM+ (i.e., it is captured), CK+, DAPI+ and CD45-." This initial analysis and identification of potential CTCs is an algorithmic/standalone function. However, the final classification involves a human-in-the-loop: "Images from each fluorescent color as well as a composite image... are presented to the user in a gallery for final cell classification. A check mark placed by the operator next to the composite images classifies the event as a Circulating Tumor Cell (CTC) and the software tallies all the checked boxes to obtain the CTC count." So it's a semi-automated system with a significant algorithm component preceding human review. The non-clinical studies on recovery, linearity, and limits of detection would largely represent this standalone performance.

7. Type of Ground Truth Used

• Non-Clinical Studies (Recovery, Linearity, Reproducibility, Interfering Substances): Known Spike Concentration and Control Sample Value. For recovery and linearity, a known number of cultured cancer cells were spiked into blood. For reproducibility, control samples with known cell concentrations were used.
• Clinical Studies (Metastatic Breast Cancer Patients): Outcomes Data (Progression Free Survival based on clinical and imaging assessments, and Overall Survival based on time to death).

8. Sample Size for the Training Set

The provided summary does not explicitly describe a separate "training set" with established ground truth for an AI/algorithm in the modern sense. The modifications were "to optimize its performance on the CellTracks® AutoPrep System to deliver the same performance as previously established using the CellPrep™ Sample Preparation System." It appears to be an optimization/calibration based on reference values and predicate device performance, rather than an AI training process using a distinct training dataset.

The clinical data from the multi-center prospective trial (N=177) "were used to determine whether the number of CTC predict disease progression and survival" and to establish the predictive value of CTC counts (e.g., 5 or more CTC/7.5 mL being predictive of shorter survival). This clinical data served as the basis for the intended use claims and clinical correlations.

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

As noted above, a distinct "training set" for an AI algorithm is not explicitly defined. The system's performance optimization aims to match the predicate device's established performance, and its clinical utility is established via the prospective clinical study with predefined clinical endpoints (PFS and OS), which serve as the "ground truth" for those claims.

Ask a specific question about this device