Abstract P4-11-04: Intrinsic subtypes and BCL2 as predictive and prognostic biomarkers in the TACT2 trial (CRUK/05/019)
Bibliographic record
Abstract
Abstract Introduction: TACT2, a multicentre randomized phase III trial in patients with node +ve or high risk node-ve invasive EBC with E-CMF as control tested two hypotheses in a 2x2 factorial design, presented results showing: i) no evidence of benefit from accelerated 2-weekly epirubicin (aE) compared to standard 3-weekly epirubicin (E) (Cameron 2012); and ii) capecitabine (X) gives equivalent efficacy but preferential side-effect profile to CMF (Canney 2014). Here we present prognostic and potential predictive value of translational biomarkers. We address two main hypotheses: i) aE is less effective in patients with luminal A than patients with other subtypes (Coates 2012) and ii) BCL2 is an independent prognosis marker (Callagy 2006). We also explore the relationship between CK5/6, EGFR and BCL2 and residual risk following chemotherapy. Methods: Tumour samples were collected prospectively from 3803 patients (86.6% of the 4391 TACT2 patients and 94.5% of those consenting). Tissue microarrays were constructed as per published guidelines and central ER, PgR, HER2, Ki67, CK5/6, EGFR and BCL2 staining performed and quantified by imaged analysis for ER, PgR, HER2, Ki67 and BCL2. EGFR and CK5/6 were dichotomised by light microscopy evaluation of cores. 94.5-97.6% of cases were stained and successfully analysed for individual biomarkers by IHC. Patients were categorised into 4 BC subtypes by central ER/PgR/HER2/Ki67 (Luminal A, B, HER2+ve, Triple-ve (TN)) with TN further divided into basal-like (CK5/6 or EGFR +ve) and 5-marker-ve (both CK5/6 and EGFR -ve). Log-rank tests assessed prognostic effect of each marker individually and cancer subtypes on time to tumour recurrence (TTR). Cox-regression models tested independent prognostic value of BCL2 in the presence of tumour size, grade, nodal status, and biological subtype. Results: No evidence of a difference in the efficacy of aE compared with E between the 4 patient subtypes was observed (Luminal A: n=608, HR (for aE compared with E) 0.80 (95% CI 0.47-1.38); Luminal B: n=1804, HR=0.97 (95% CI 0.76-1.23); HER2+ve: n=219, HR=1.03 (95% CI 0.53-1.99); TN: n=638, HR=1.00 (95% CI 0.72-1.39); Test for heterogeneity p=0.84). When subdividing the TN group into basal-like and 5-marker-ve, HRs were 0.66 (95% CI 0.45-0.98) and HR 1.56 (95% CI 0.82-2.96) respectively, post-hoc analysis for heterogeneity between these 2 subtypes p=0.025. No differential effect between subtypes was observed for the comparison of CMF and X. 82/434 (18.9%) patients with low BCL2 expression (≤10%) had a TTR event compared with 444/3158 (14.1%) with high BCL2 (>10%), univariate HR 1.39, 95% CI 1.10-1.76, p=0.006. However this difference was no longer seen after adjustment for clinical factors and biological subtype (HR 1.17 95% CI 0.90-1.52, p=0.25). Conclusion: We found no statistical evidence that luminal A cancers are associated with reduced benefit from aE vs E. A hypothesis generating observation that benefit from aE vs E might be different between basal-like and 5-marker-ve cancers should be interpreted with caution due to the small numbers of cases and the retrospective nature of the analysis. In this study BCL2 did not provide independent prognostic information when corrected for conventional histopathological features. Citation Format: James Morden, Judith Bliss, Jane Bayani, Robert Laing, Rajiv Agrawal, Jeremy Thomas, Andrew Goodman, Vivienne Loo, Peter Clark, Peter Canney, Peter Barrett-Lee, John Bartlett, David Cameron. Intrinsic subtypes and BCL2 as predictive and prognostic biomarkers in the TACT2 trial (CRUK/05/019) [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P4-11-04.
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 | 0.002 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 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 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".