(85 days)
To qualitative aid in the identification by light microscopy of human cells of B-cell lineage, by recognizing CD20 antigen in normal and pathologic paraffin embedded tissues processed in neutral buffered formalin, B5, or Bouin's fixative. Positive results aid in the classification of lymphomas of B-cell origin and must be interpreted by a pathologist within the context of clinical data, gross and microscopic morphologic criteria and multiple chemical and immunohistochemical stains.
The ChemMate™ L26 (CD20) is comprised of a mouse monoclonal antibody, clone L26. and is of the IgG2a / Kappa light chain class of immunoglobulins. The antibody reacts with two non-covalently associated components (33kD and 30kD) present in the majority of B-cells. Positive staining occurs on the cells' plasma membrane. Recent studies demonstrated that the L26 antibody positively stained COS-1 cells transfected with CDNA encoding for the CD20 antigen (a pan B-cell marker). L26 has been classified as a CD20 antibody by the Fifth International Workshop on Human Leukocyte Differentiation Antigens.
A unique feature of the ChemMate™ L26 (CD20) is the antibody's ability to recognize epitopes on Ivmphocvtes that survive the rigors of processing and fixation. The resultant advantage in the use of this antibody is its applicability with formalin-fixed specimens.
The CD20 antigen is expressed in most B-cells present in peripheral blood and lymphoid tissue including germinal center blasts, mantle zone lymphocytes and scattered interfollicular lymphocytes. Staining also occurs in most non-Hodgkin's lymphomas of B-cell lineage.
The provided text is a 510(k) Summary for a medical device (ChemMate L26 (CD20) antibody reagent) and not a study on a new AI/software device. Therefore, it does not contain the information requested in the prompt regarding acceptance criteria, study design for AI performance, sample sizes for training/test sets, ground truth establishment methods, or expert qualifications in the context of an AI device evaluation.
The document describes the intended use and performance characteristics of an immunohistochemical antibody reagent. It summarizes the findings from various published studies and internal testing to demonstrate the antibody's reactivity and specificity in identifying B-cell lymphocytes in tissue samples.
Summary of what is present:
- Intended Use: To qualitatively identify human lymphocytes of B-cell lineage by light microscopy, recognizing CD20 antigen in normal and pathological paraffin-embedded tissues. Positive results aid in the classification of lymphomas as B-cell in origin and must be interpreted by a pathologist within the context of clinical data, morphological criteria, and multiple stains.
- Performance Characteristics (Immunoreactivity): The document provides extensive data from various published studies and internal testing on the reactivity of the L26 (CD20) antibody across a range of normal and cancerous tissues. This data essentially serves as the evidence for the device's performance, but it's not structured as a single "study" proving acceptance criteria in the way an AI/software device would be evaluated.
- Normal Reactive Tissues (Table I): Lymph Node (germinal center, mantle zone), Spleen (white pulp), Tonsil (germinal center, mantle zone), Thymus (medulla), Peripheral Whole Blood (B lymphocytes).
- Normal Non-Reactive Tissues (Table II): Various lymphoid (interfollicular, medullary areas of lymph node/spleen/tonsil, thymus cortex, bone marrow components) and non-lymphoid tissues (skin, thyroid, lung, brain, pancreas, liver, prostate, kidney, uterus, muscle, placenta).
- Non-Lymphoid Pathological Negative Tissues (Table III): A wide array of epithelial, endocrine, neuro/glial, mesenchymal, and small cell tumors were tested and reported as non-reactive.
- Non-Hodgkin's Lymphomas & Other Lymphoproliferative Disorders (Table IV):
- B-Cell Non-Hodgkin's Lymphoma: 397/430 (92%) were reactive. Subtypes like small cleaved, follicular & diffuse (100%), mixed small and large (100%), large cleaved (100%), large diffuse (98%), intermediate (100%), small non-cleaved (100%), large immunoblastic (96%), large lymphoblastic (pre B, 100%), malignant (96%) showed high reactivity.
- T-Cell Non-Hodgkin's Lymphoma: 1/89 (1.1%) were reactive, indicating high specificity against T-cell lymphomas.
- Other Lymphoproliferative Disorders: Hairy Cell Leukemia (87%), Idiopathic Thrombocythaemia (100%), Myelofibrosis/Osteomyelosclerosis (100%), with lower reactivity in Plasmacytoma/Plasma Cell Myeloma (10%).
- Hodgkin's Lymphoma (Table V): Lymphocyte Predominant (96%), Inconspicuous Nodularity (100%), Diffuse (100%) showed high reactivity, while Lymphocyte Depleted (11%), Mixed Cellularity (18%), and Nodular Sclerosing (6%) showed low reactivity.
- Reproducibility: Tested on serial sections of 281 tissue specimens (normal and tumor) run three times on different days. Consistent staining results were obtained.
Missing Information (as per the prompt's requirements for an AI/software device study):
The prompt specifically asks for details related to the evaluation of an AI/software device, which would include:
- Acceptance Criteria Table and Reported Device Performance: While the document reports performance for the antibody (reactivity/specificity rates), it does not define specific pre-set "acceptance criteria" that an AI device would need to meet (e.g., a specific AUC, sensitivity, or specificity threshold). The performance data is the reported performance.
- Sample Size for Test Set and Data Provenance: The document refers to various published studies and internal testing, using hundreds of tissue specimens. For example, 281 for reproducibility, and counts varying from 1 to 430 for different tissue types in immunoreactivity tables. However, it does not specify a single, defined "test set" sample size with data provenance (country, retrospective/prospective) in the way an AI study would. The data is aggregated from multiple sources and studies over time.
- Number of Experts & Qualifications for Ground Truth: Not applicable in the context of an antibody reagent validation. Ground truth is established by standard pathological diagnosis through histological examination and established classification systems, not by a specific number of experts reviewing data for an algorithm.
- Adjudication Method: Not applicable. Standard pathological diagnosis is assumed.
- MRMC Comparative Effectiveness Study: Not applicable, as this is an antibody reagent, not an AI device designed for human-in-the-loop assistance.
- Standalone Performance: The performance data is the standalone performance of the antibody in staining cells. There is no algorithm to run "without human-in-the-loop."
- Type of Ground Truth: Pathology (histological diagnosis of tissue types and disease status).
- Sample Size for Training Set: Not applicable. Antibodies are not "trained" in the machine learning sense. Their performance is inherent to their biological properties.
- How Ground Truth for Training Set Established: Not applicable.
In conclusion, the provided 510(k) summary thoroughly describes the performance of an immunohistochemical reagent based on published literature and internal testing, fulfilling the requirements for its specific product type. However, it does not fit the structure of an AI/software device study and therefore lacks most of the requested information pertinent to such a study.
§ 866.5130
Alpha -1-antitrypsin immunological test system.(a)
Identification. Analpha -1-antitrypsin immunological test system is a device that consists of the reagents used to measure by immunochemical techniques thealpha -1-antitrypsin (a plasma protein) in serum, other body fluids, and tissues. The measurements aid in the diagnosis of several conditions including juvenile and adult cirrhosis of the liver. In addition,alpha -1-antitrypsin deficiency has been associated with pulmonary emphysema.(b)
Classification. Class II (performance standards).