Abstract P5-01-01: Predicting OncoDX Recurrence Scores with Immunohistochemical Markers
Bibliographic record
Abstract
Abstract Background: Standard immunohistochemistry (IHC) performed for invasive breast carcinoma includes ER, PR and Her2/Neu status, and these markers are used in conjunction with other patient and tumour factors to determine prognosis and guide treatment. Many, but not all, low stage, lymph node (LN) negative, ER positive patients have a good prognosis without chemotherapy. Thus a demand exists for predictive tools to stratify patient risk within this subgroup. It has been reported that, for a subset of ER positive Her2/Neu negative patients, the 21 gene OncotypeDX recurrence score (Onco-RS) adds independent prognostic information to that obtained from these standard IHC markers (1). As several genes analyzed for the Onco-RS relate to ER, PR, HER2/neu and proliferative status, it is reasonable to try to incorporate clinical-pathological variables and these IHC scores into a predictive model. Indeed, recent studies suggest that most of the additional information provided by the OncoDX-RS may be obtained more cost effectively using the Ki-67 IHC based proliferation percentage combined with a semi-quantitative assessment of standard IHC markers including ER and PR and Her2/neu (2). The aim of this study is to assess the ability of a simple combined IHC recurrence score (IHC-RS) to predict Onco-RS. The IHC-RS was derived from a simple semi-quantitative assessment of ER and PR combined with Ki-67 proliferation percentage. Design: A cohort of 159 women aged 27–78 with ER positive, HER2/neu negative breast cancer completed Oncodx testing between March 2010 and May 2012. This sample reflects the population selected at our institution for Oncotype testing. The variables investigated for inclusion in a model to predict RS score included tumor grade, stage, patient age, Allred ER & PR and Ki-67 percentage. Results and Discussion: A predictive model was developed to generate a recurrence score (IHC-RS) using stepwise multiple regression incorporating Allred ER score, Allred PR score and Ki-67 percentage. The best subset model (Schwartz BIC) accounted for 60.7% of the Onco-RS variability (adjusted R2 = 60.7, p = 0.05). In addition, analysis of individual cases where the IHC-RS was not in agreement with the Onco-RS reveals that the Onco-RS, although technically highly reproducible, may suffer from sampling error. The IHC-RS is more robust with respect to sampling error, owing to the retention of tumor architecture inherent in IHC. IHC, however, can lack the technical reproducibility and transportability inherent in the Onco-RS methodology. There are clearly advantages to an ICH derived multi-score such as IHC-4 (3) or the simpler IHC-RS proposed in this study. Full utility of any IHC-based recurrence score will require standardization of testing and scoring both within and across different testing laboratories. In addition, full utility of any IHC based model will require direct correlation to patient outcome, rather than to a surrogate marker such as the Onco-RS used in this study. 1. M Dowsett et al. J Clin Oncol. 2010 Apr 10; 1829–1834 2. Cuzick J et al. J Clin Oncol. 2011 Nov 10;4273-8. 3. S Barton et al. British Journal of Cancer (2012) 106, 1760–1765. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P5-01-01.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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 source (direct Gemma or distilled Codex), 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".