Search Results

The Prosigna™ Breast Cancer Prognostic Gene Signature Assay is an in vitro diagnostic assay which is performed on the NanoString nCounter® Dx Analysis System using FFPE breast tumor tissue previously diagnosed as invasive breast carcinoma. This qualitative assay utilizes gene expression data, weighted together with clinical variables to generate a risk category and numerical score, to assess a patient's risk of distant recurrence of disease.

The Prosigna Breast Cancer Prognostic Gene Signature Assay is indicated in female breast cancer patients who have undergone surgery in conjunction with locoregional treatment consistent with standard of care, either as:

1. A prognostic indicator for distant recurrence-free survival at 10 years in post-menopausal women with Hormone Receptor-Positive (HR+), lymph node-negative, Stage I or II breast cancer to be treated with adjuvant endocrine therapy alone, when used in conjunction with other clinicopathological factors.

2. A prognostic indicator for distant recurrence-free survival at 10 years in post-menopausal women with Hormone Receptor-Positive (HR+), lymph node-positive (1-3 positive nodes), Stage II breast cancer to be treated with adjuvant endocrine therapy alone, when used in conjunction with other clinicopathological factors. The device is not intended for patients with 4 or more positive nodes.

Special Conditions for Use: Prosigna is not intended for diagnosis, to predict or detect response to therapy, or to help select the optimal therapy for patients.

Used together, the Prosigna™ Breast Cancer Prognostic Gene Signature Assay and nCounter Dx Analysis System are a nucleic acid hybridization and image analysis system based upon coded probes designed to detect the messenger RNA transcribed from 58 genes. The test input is purified RNA from FFPE breast tumor specimens which are acquired from surgical resection. The Prosigna assay uses gene-specific probe pairs that hybridize directly to the mRNA transcripts in solution. The nCounter Dx Analysis System delivers direct, multiplexed measurements of gene expression through digital readouts of the relative abundance of the mRNA transcripts. Specifications are included as part of the Prosigna Assay to control for sample quality, RNA quality, and process quality. Prosigna simultaneously measures the expression levels of 50 genes used in the PAM50 classification algorithm (Parker et al., 2009), 8 housekeeping genes used for signal normalization, 6 positive controls, and 8 negative controls in a single hybridization reaction, using nucleic acid probes designed specifically to those genes. The Prosigna assay utilizes prototypical expression profiles (centroids) which are associated with and define each of the four PAM50 molecular subtypes of breast cancer. The software algorithm produces a Prosigna Score (referred to as ROR Score or Risk of Recurrence Score in the literature (Dowsett et al., 2013)) based on the similarity of the expression profile to each PAM50 molecular subtype, as well as the gross pathological tumor size and a proliferation score computed from a subset of genes. Three risk categories (low, intermediate and high) were defined based on a study with over 1007 patient samples associating Prosigna score with longterm outcome.

The required components for the Prosigna Assay include the RNA Isolation kit (manufactured by Roche), Prosigna reagents (Reference Sample, CodeSet, Prep Pack, Cartridge(s) and Prep Plate) and the instruments that comprise the nCounter Dx Analysis System; the Prep Station and Digital Analyzer.

The test output is a patient specific report which includes a Prosigna score (0-100) and risk category (low/intermediate/high).

Let's break down the acceptance criteria and the study that proves the device meets those criteria for the Prosigna™ Breast Cancer Prognostic Gene Signature Assay, based on the provided FDA 510(k) summary.

1. Table of Acceptance Criteria and Reported Device Performance

The FDA 510(k) summary for the Prosigna assay demonstrates clinical performance relative to its intended use as a prognostic indicator for distant recurrence-free survival (DRFS) at 10 years. The acceptance criteria are implicit in the statistical significance and magnitude of the prognostic information provided by the Prosigna Score, both as a continuous variable and when categorized into risk groups.

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

• Test Set (Clinical Validation):
  • The primary clinical validation study was ABCSG-8, with 1,478 patients available for analysis after QC failures (out of 1,620 tissues initially available).
  • A prior study, TransATAC, was used to select the Prosigna Score cut-offs for risk categories. This study also demonstrated continuous relation to DRFS.
• Data Provenance: The 510(k) summary states that the validation population for the Prosigna Assay utilized "Treatment arms from a randomized trial conducted in Europe" with a "prospective retrospective study design". Both TransATAC and ABCSG-8 samples were independent from those used to train the Prosigna algorithm.

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

The document does not explicitly state the number or qualifications of experts used to establish the ground truth for the test set. However, a "pre-defined pathology review criteria for adequate tumor" was applied to the ABCSG-8 tissue samples, which would typically involve expert pathologists. The clinical endpoints (distant recurrence, distant recurrence-free survival) are outcome data derived from patient follow-up within the clinical trial, rather than ground truth established by expert consensus on images.

4. Adjudication Method for the Test Set

The document does not specify an adjudication method for establishing specific diagnoses or outcomes on a case-by-case basis. The outcomes (distant recurrence, death) are clinical events recorded during patient follow-up, typical of a clinical trial. Pathology review criteria were applied to screen samples, but this is a quality control step, not an adjudication of a device output.

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 MRMC comparative effectiveness study involving human readers and AI assistance is mentioned. The Prosigna assay is a gene expression profiling test system, providing a numeric score and risk category directly. It is not designed to assist human readers in interpreting images or other data in the way an AI diagnostic tool might. Its primary purpose is to add prognostic information to clinical variables.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

Yes, the clinical performance described is a standalone (algorithm only) performance. The Prosigna Assay generates a risk category and numerical score directly from gene expression data and clinical variables. The studies evaluate the prognostic value of this score and risk category on patient outcomes (DRFS) without human interpretation of the assay's raw output. The output is intended for healthcare professionals to consider alongside other clinicopathological factors in treatment decisions.

7. The Type of Ground Truth Used (expert consensus, pathology, outcomes data, etc.)

The ground truth used for the clinical validation studies (TransATAC and ABCSG-8) was patient outcomes data, specifically:

• Distant Recurrence-Free Survival (DRFS) at 10 years.
• Incidence of distant recurrence (events).
  The clinical endpoints were derived from long-term follow-up of patients in these retrospective clinical trials.

8. The Sample Size for the Training Set

The document explicitly states that "Both the TransATAC and ABCSG-8 study samples were independent from those samples used to train the Prosigna algorithm." However, the exact sample size for the training set (i.e., the cohort(s) used to develop the PAM50 classification algorithm and subsequently to define the Prosigna Score and risk categories prior to its validation) is not specified in this 510(k) summary. It references "Parker et al., 2009" for the PAM50 classification algorithm and "Dowsett et al., 2013" for the Prosigna Score literature, which would contain details on the training cohorts used for algorithm development and initial risk stratification. The summary notes that "Three risk categories (low, intermediate and high) were defined based on a study with over 1007 patient samples associating Prosigna score with long-term outcome." This "study with over 1007 patient samples" likely refers to the TransATAC study, which was used for defining cut-offs and therefore might be considered part of an extended training/calibration phase before independent validation in ABCSG-8.

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

Similar to the validation set, the ground truth for the (unspecified) training set would have been established using patient outcomes data, specifically clinical endpoints such as:

• Distant Recurrence-Free Survival
• Overall Survival
• Other breast cancer-specific endpoints.

The "PAM50 classification algorithm" itself was "based on intrinsic subtypes" (Parker et al., 2009), implying that molecular profiling of tumors was correlated with clinical outcomes to identify and define these subtypes, which then formed the basis of the gene signature. The "study with over 1007 patient samples associating Prosigna score with long-term outcome" (likely TransATAC) would also have used patient outcomes data for this association and for defining risk category cut-offs.

Ask a specific question about this device