Abstract PO4-16-02: Can high progesterone receptor (PgR) expression identify tumours with low-risk tumour gene expression scores?
Bibliographic record
Abstract
Abstract Background Strong PgR expression predicts favorable outcomes for ER+ve HER2-ve breast cancer and has been proposed as a surrogate marker to distinguish between IHC-defined luminal A and luminal B subtypes. It is therefore possible that strong PgR expression may be able to predict tumor gene expression test results and independently identify tumors that are unlikely to be chemotherapy sensitive. PgR expression is traditionally determined by immunohistochemistry (IHC). Several validated RNA-based PGR expression tests have been developed that may outperform IHC. Methods We compared Oncotype DX RS with Oncotype DX reported PGR in 4 independent datasets which included 407 cases from the OPTIMA prelim trial. We further analyzed 251 OPTIMA prelim cases which had additional tumor gene expression data. All gene expression assays were performed by the test vendor, including PGR gene expression determined using the Mammatyper assay. PgR IHC was determined in a single laboratory on triplicate tissue micro-arrays using quantitative image analysis including a 10% manual quality control check. We analyzed PGR expression using cutoffs that correspond to approximately 20% staining; the standard Oncotype DX PGR assay is reported as positive if the score is >5.4, corresponding to approximately 1% staining by IHC. We used Spearman’s rank correlation coefficient to compare PGR data. Results The four Oncotype DX datasets consistently demonstrated that high Oncotype PGR expression was associated with a low RS (table). Combining the 3 validation data sets consisting of 633 cases, 70.9% had high PGR expression of which 92.7% had a an Oncotype RS ≤25. Approximately 50% of cases with low Oncotype PGR expression had an Oncotype RS >25. Mammatyper and Oncotype PGR were highly correlated (Rs = 0.9258, P< 0.001) in the OPTIMA prelim dataset (n=251), with only 8.4% of tumors having discordant high/low Mammatyper and Oncotype PGR scores. 93.2% of 176 Mammatyper high PGR expression cases had an RS ≤25. The Mammatyper PGR and PgR IHC correlation was weaker (Rs=0.763, P< 0.001); 87.2% of 211 cases with >20% staining had RS ≤25. PgR IHC staining had a bimodal distribution and there was little effect on the prediction of low RS score up to a 67% cut-off. Mammatyper and Oncotype PGR scores both appear to have a normal distribution. We took advantage of this to perform an exploratory analysis using a higher Mammatyper PGR cutoff. We were able to show superior prediction of a low RS (96.8%) but with a reduced proportion (50.2%) of high PGR score tumors. High PGR gene expression was weakly associated with low (≤60) Prosigna ROR_PT score and MammaPrint low risk (72.2% and 65.9% respectively) and with Prosigna and MammaPrint luminal A subtype (both 64.8%). Conclusion High progesterone receptor gene expression measured using locally performed RNA-based assays may allow the reliable prediction of Oncotype DX low-risk tumours. This analysis provides additional information for the clinical utility of PGR measurement. Additional data will be presented on the optimal PGR cutoff. OPTIMA prelim is registered as ISRCTN42400492 and funded by the UK NIHR Health Technology Assessment Programme, award number 10/34/01. Views expressed are those of the authors and not those of the HTA Programme, NIHR, NHS or the Department of Health. Table. Oncotype DX RS and PGR in 4 datasets Distribution of RS according to %cases with high or low PGR Citation Format: Robert Stein, Ralph Wirtz, Andrea Marshall, Jane Bayani, Sebastian Eidt, Claudia Schumacher, Hans-Peter Sinn, Andreas Schneeweiss, Andreas Makris, Iain Macpherson, Luke Hughes-Davies, Tammy Piper, Monika Sobol, Georgina Dotchin, Helen Higgins, Sarah Pinder, Abeer Shaaban, Janet Dunn, John MS Bartlett. Can high progesterone receptor (PgR) expression identify tumours with low-risk tumour gene expression scores? [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-16-02.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".