Abstract P2-07-02: A newly derived combined clinical treatment score and immunohistochemical-4 prognostic tool
Bibliographic record
Abstract
Abstract AIM To determine whether a modified Clinical Treatment Score (CTS) based on continuous tumor size and 5 lymph node categories provided more prognostic information in an independent test set than the original CTS with and without the Immunohistochemical-4 (IHC4) algorithm for prediction of residual distant recurrence risk over 10-years. BACKGROUND Risk of recurrence information in patients with estrogen receptor-positive (ER+), early breast cancer informs decision-making on chemotherapy use. The CTS and IHC4 algorithms provide such information, particularly when used in combination (IHC4+C). Their derivation in the translational cohort of the Arimidex Tamoxifen Alone or in Combination trial (TransATAC) was described by Cuzick et al in 2011. In the original model tumor size and nodal status were each classified into three categories, causing prognostic information to be lost. METHODS We modeled a novel CTSn algorithm on outcome data from patients in the anastrozole and tamoxifen arms in ATAC incorporating tumor size as a continuous variable and sub-dividing nodal status into five categories. IHC4n was re-derived in the TransATAC cohort independent of CTSn. Patients were chemotherapy-naïve. We compared ability to predict risk of residual distant recurrence of the new IHC4n+Cn model with that of the original one when tested in a training cohort and in a validation set of chemotherapy-naïve patients from the Tamoxifen vs. Exemestane Adjuvant Multicentre (TEAM) trial using Cox regression models and the C-index. RESULTS The ATAC training set for CTSn comprised 4056 patients, the TransATAC training set for IHC4n comprised 1125 patients; 2591 patients were in the TEAM validation set. Patients in the TEAM set were older (median age in TransATAC: 63.5, TEAM: 68.3 years), had a higher nodal-burden (node-positive in TransATAC: 29.4%, TEAM: 51.8%) and had more Grade 3 tumors (TransATAC: 18.3%, TEAM: 32.2%). The new IHC4n+Cn was significantly prognostic, and non-significantly more prognostic than the original IHC4+C in both the training and validation cohorts. When assessed using the C-index statistic, IHC4n+Cn had a higher discriminatory ability than the original algorithm (Table 1). Table 1 TransATAC (N=1125)TEAM (N=2591) HR* (95% CI)C-indexHR* (95% CI)C-indexOld Models CTS2.26 (2.01-2.53)0.6811.88 (1.73-2.03)0.650IHC41.67 (1.46-1.91)0.6301.49 (1.35-1.63)0.604IHC4+C2.76 (2.40-3.18)0.7242.03 (1.87-2.21)0.671New Models CTSn2.64 (2.26-3.09)0.7212.16 (1.96-2.39)0.687IHC4n1.74 (1.52-2.01)0.6421.51 (1.36-1.68)0.603IHC4n+Cn2.91 (2.47-3.42)0.7382.28 (2.06-2.51)0.695(*Hazard Ratio for change in one Standard Deviation). CONCLUSION By separately remodellng the part of the IHC4+C score based on clinicopathological characteristics using the whole ATAC chemo-naïve cohort, and the part that uses IHC-derived information in chemo-naïve TransATAC patients, we increased the precision of the individual risk estimates produced by both CTSn and IHC4n compared to those given by the original algorithms. The new IHC4n+Cn shows a trend for improved prognostic ability compared to the original IHC4+C. Like its predecessor, it relies on information that is readily available to clinicians and integrates it in an evidence-based way to improve prognostication in ER+ early breast cancer. Citation Format: Dodson A, Sestak I, Bayani J, Dowsett M, Bartlett J, Cuzick J. A newly derived combined clinical treatment score and immunohistochemical-4 prognostic tool [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P2-07-02.
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 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 | 0.003 | 0.009 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 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".