Abstract PS6-07: Comparison of breast cancer staging models in patients after neoadjuvant chemotherapy
Bibliographic record
Abstract
Abstract Background: Neoadjuvant chemotherapy (NAC) is commonly utilized for breast cancer, however no consensus exists on the best way to stage these patients following treatment. Importantly, the American Joint Commission on Cancer (AJCC) 8th edition staging system did not specifically address staging after NAC. However, previous work by our group has shown that the pathologic prognostic stage does stratify patients with respect to outcomes. Our objective was to compare performance of the AJCC pathologic prognostic staging system to a second staging model, the Residual Cancer Burden (RCB) Index which takes into account residual tumor size, cellularity, lymph node status and size of any lymph node metastases. Methods: A retrospective review identified patients with stage I-III invasive breast cancer treated with NAC from 2004-2014 at Dana-Farber Cancer Institute. Patients were excluded if they did not have RCB reported on final pathology. Disease-free survival (DFS) was defined as any recurrence or death from any cause, and overall survival (OS) as death from any cause. DFS and OS were calculated using the Kaplan-Meier method for each staging model. Receiver operator characteristic (ROC) curves were used to assess model fit using the c-statistic and the Hanley method to compare c-statistics. Results: A total of 802 patients underwent NAC for stage I-III breast cancer. The median age was 48 years (range 22-86). Most patients presented with cT2 (n=470, 58.6%) or cT3 (n=188, 23.4%) and cN1 (n=422, 52.6%) disease. The majority (n=563, 70.2%) presented with grade 3 disease. In terms of subtype, 296 (36.9%) patients had hormone receptor-positive, HER2 negative, 261 (32.5%) HER2+, and 245 (30.5%) triple negative disease. Median follow up was 79.5 months (range 4-169). There were 176 recurrences including 32 local, 25 regional, and 145 distant recurrences. 676 (76.8%) patients were alive at last follow-up. The Table depicts the 7-year DFS and OS estimates for each of the staging models. The ROC c-statistics for DFS model fit were statistically similar, 0.72 for AJCC pathologic prognostic stage and 0.71 for RCB (p=NS). The c-statistics for OS were 0.74 and 0.71 respectively (p=NS). Conclusions: Our results provide additional external validation of the AJCC pathologic prognostic stage and RCB’s ability to stratify patients after NAC with respect to survival outcomes. These data can be used to inform subsequent revisions of the AJCC breast cancer staging system. Estimated 7-year DFS and OS in Potential Staging Models for Breast Cancer Patients after NAC (n=802)7yr-DFS7yr-OSPathologic Prognostic StageStage 0 (n=228)92.9%94.8%Stage IA (n=193)81.7%90.2%Stage IB (n=173)74.5%86.6%Stage IIA (n=105)62.2%71.5%Stage IIB (n=11)70.2%57.3%Stage IIIA (n=40)62.2%75.4%Stage IIIB (n=27)56.7%83.0%Stage IIIC (n=25)27.8%28.2%C-statistic (95% CI)0.72 (0.68-0.76)0.74 (0.69-0.79)RCB0 (n=226)93.5%94.8%I (n=118)83.0%90.0%II (n=278)75.9%85.0%III (n=180)55.1%69.9%C-statistic (95% CI)0.71 (0.67-0.75)0.71 (0.66-0.75) Citation Format: Olga Kantor, Alison Laws, Ricardo G Pastorello, Stephanie Wong, Tanujit Dey, Stuart Schnitt, Tari A King, Elizabeth A Mittendorf. Comparison of breast cancer staging models in patients after neoadjuvant chemotherapy [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS6-07.
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 | 0.010 |
| Meta-epidemiology (narrow) | 0.000 | 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) | 0.003 | 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".