Abstract 1030: Concordance between the DESTINY-Breast04 clinical trial assay (4B5[CDx]) and other HER2 IHC assays for HER2-low breast cancer in real-world practice: First phase of a large-scale, multicenter global ring study
Bibliographic record
Abstract
Abstract Background: DESTINY-Breast04 (DB-04) showed that trastuzumab deruxtecan improves survival vs chemotherapy in patients with human epidermal growth factor receptor 2 (HER2)-low (immunohistochemistry [IHC] 1+ or IHC 2+ with in situ hybridization negative) metastatic breast cancer (BC). The VENTANA 4B5(CDx) IHC assay, now approved in the United States (US) as a companion diagnostic, was used as the clinical trial assay (CTA) in DB-04. Real-world differentiation of HER2-low vs HER2 IHC 0 BC with non-CTAs is difficult. This is the first phase of a study assessing agreement between CTA and non-CTA scores. Methods: 50 clinical BC samples with HER2 IHC scores of 0, 1+, 2+, and 3+ were sent to laboratories in Europe, Canada, and the US for HER2 IHC testing per ASCO/CAP 2018 guidelines. After routine protocol scoring, pathologists were trained and rescored the samples 2 weeks later. Post training scores were compared with reference CTA scores determined by a central laboratory and a panel of experts. The primary endpoint was positive percentage agreement (PPA) and negative percentage agreement (NPA) between CTA and non-CTA scores, assessed after training, considering HER2-low as positive and HER2 IHC 0 as negative. Results: A total of 3449 post training non-CTA scores were available for analysis. PPA for HER2-low vs HER2 IHC 0 was 87.5% and NPA was 61.9%, with Cohen κ of 0.51 (Table). HER2-low vs HER2 IHC 0 agreement between CTA and non-CTA varied across subgroups. PPA >95% was shown with 4B5 Laboratory Developed Tests (LDTs) and HercepTest (Omnis), NPA of 80% with Leica, and overall agreement >80% with HercepTest (Link48), 4B5 LDTs, and German, French, and Italian laboratories. Conclusions: The degree of concordance between the CTA and non-CTAs varied among assay types and laboratory locations. A low NPA for some non-CTAs suggests the need for further assay optimization for HER2 IHC 0 evaluation. Post training non-CTA results HER2-low vs HER2 IHC 0 BC Post training scores PPA (95% CI), % NPA (95% CI), % Overall percentage agreement (95% CI), % Cohen κ (95% CI) Overall non-CTA test results (N = 3449) 87.5 (86.0-89.0) 61.9 (58.9-64.9) 78.7 (77.1-80.1) 0.51 (0.48-0.54) Non-CTA subgroups HercepTest (Omnis) (n = 467) 95.5 (92.3-97.7) 36.9 (28.9-45.4) 75.2 (70.8-79.4) 0.37 (0.28-0.46) HercepTest (Link48) (n = 790) 88.5 (85.2-91.2) 64.3 (57.8-70.4) 80.3 (77.2-83.2) 0.55 (0.48-0.61) Leica (n = 96) 59.3 (45.0-72.4) 80.0 (61.4-92.3) 66.7 (55.5-76.6) 0.35 (0.17-0.53) Non-4B5 LDTs (n = 1609) 83.8 (81.2-86.1) 67.8 (63.5-72.0) 78.2 (75.9-80.3) 0.52 (0.47-0.57) 4B5 LDTs (n = 487) 96.0 (93.0-98.0) 58.8 (50.4-66.8) 83.0 (79.1-86.4) 0.59 (0.51-0.68) Country subgroups Germany (n = 340) 93.1 (88.5-96.3) 64.4 (54.2-73.6) 83.1 (78.3-87.2) 0.61 (0.51-0.70) France (n = 342) 95.4 (91.5-97.9) 64.4 (54.4-73.6) 84.7 (80.1-88.6) 0.64 (0.55-0.73) Italy (n = 395) 89.9 (85.2-93.5) 72.5 (63.6-80.3) 83.9 (79.6-87.6) 0.64 (0.55-0.72) Spain (n = 567) 83.0 (78.4-86.9) 64.2 (56.6-71.3) 76.3 (72.3-80.0) 0.48 (0.40-0.56) US/Canada (n = 1029) 86.6 (83.6-89.2) 61.8 (56.1-67.3) 78.2 (75.4-80.8) 0.50 (0.44-0.56) Europe, other (n = 776) 85.0 (81.3-88.2) 52.8 (46.2-59.3) 73.8 (70.3-77.1) 0.40 (0.32-0.47) Citation Format: Hans-Ulrich Schildhaus, Sunil Badve, Corrado D’Arrigo, Gelareh Farshid, Annette Lebeau, Vicente Peg, Fréderique Penault-Llorca, Josef Rueschoff, Wentao Yang, Neil Atkey, Jessica Baumann, Elisabeth Beyerlein, Amy Hanlon Newell, Alexander Penner, Akira Moh, Guiseppe Viale. Concordance between the DESTINY-Breast04 clinical trial assay (4B5[CDx]) and other HER2 IHC assays for HER2-low breast cancer in real-world practice: First phase of a large-scale, multicenter global ring study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1030.
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 | 0.024 | 0.023 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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".