Breast Oncology Briefing

Weekly. Updated automatically every Monday morning.

Week of September 21 to September 28, 2026

Neo-Healer Phase III: Anbenitamab-Taxane Beats THP on Pathologic Response

Calendar note: no major breast congress is underway this week. ESMO Congress 2026 convenes in Madrid October 23 to 27 (per OncologyTube's pre-congress session tracker); company toplines are beginning to surface, but I could not verify any breast-specific late-breaker details against primary sources this week. This was otherwise a genuinely light window, and the edition is sized accordingly.

1. Practice-Relevant Clinical

Lead item: Neo-Healer phase III published in JAMA Oncology, anbenitamab plus albumin-bound docetaxel vs THP

The lead development this week is the publication of the multicenter phase III Neo-Healer trial in JAMA Oncology, covered by The ASCO Post on September 23. Systemic therapy-naive patients with stage II or III HER2-positive breast cancer from 61 centers in China (median age 52) were randomized 1:1 to neoadjuvant anbenitamab, an ERBB2-biparatopic antibody, plus HB1801, a solvent-free albumin-bound docetaxel formulation (n = 263), or trastuzumab plus pertuzumab with docetaxel as the taxane (THP, n = 258), each with or without carboplatin, every 3 weeks for six cycles. The primary endpoint was total pathologic complete response (tpCR) by blinded independent review.

tpCR was 62.4% with anbenitamab plus HB1801 vs 51.2% with THP (absolute difference 11.4 percentage points; 95% CI 3.2 to 19.6; P = .004). The advantage was consistent across subgroups: hormone receptor-positive 51.7% vs 44.4%, hormone receptor-negative 76.3% vs 59.5%, early-stage 63.8% vs 51.8%, locally advanced 59.6% vs 50.0%, with carboplatin 66.7% vs 54.5%, without carboplatin 59.2% vs 48.6%. Grade 3 or 4 treatment-related adverse events were 29.3% vs 28.3%, with no treatment-related deaths in either arm.

What this means in clinic: nothing changes in US practice today. Anbenitamab is not FDA approved, the trial was conducted exclusively in China, and the investigators themselves flag that long-term survival follow-up is warranted; a pCR endpoint with no event-free survival or overall survival readout keeps this a signal rather than a standard-of-care challenge. That said, this is a positive phase III against the THP backbone in a space where recent innovation has come mainly from de-escalation strategies, and it is worth watching whether the biparatopic-antibody approach travels outside China. No other phase 3 breast readouts or toplines were verified within the window in my scan.

2. FDA Landscape

Honest bottom line: I could identify no new breast-specific FDA action dated within September 21 to 28. The agency's most recent breast entries remain the September 4 camizestrant accelerated approval and the September 18 imlunestrant combination approval. Both fall just outside this window and were covered last week, but they are recapped here because coverage and practice uptake extended into this window, one carries a correction to a prior expectation, and the operational details (dosing, testing) still need to land in workflows.

Imlunestrant plus abemaciclib (September 18): full approval for ESR1-mutated disease after endocrine progression

Per the FDA approval notice, imlunestrant (Inluriyo) 400 mg orally once daily on an empty stomach plus abemaciclib 150 mg twice daily is fully approved for adults with ER-positive, HER2-negative, ESR1-mutated advanced or metastatic breast cancer, as detected by an FDA-authorized test, after progression on at least one line of endocrine therapy. Guardant360 CDx was concurrently approved as the companion diagnostic, limited to ligand-binding-domain ESR1 mutations on ctDNA.

The evidence bears emphasis because of its structure: in EMBER-3 (874 patients, randomized 1:1:1 to imlunestrant, investigator's-choice endocrine therapy, or the combination), the combination beat imlunestrant monotherapy on investigator-assessed PFS in the overall population, but imlunestrant monotherapy did not beat investigator's-choice endocrine therapy in the overall or ESR1-not-detected populations; the label is therefore restricted to ESR1-mutated disease. The key effect sizes come from an exploratory subgroup of 159 ESR1-mutated patients: median PFS 11.1 months (95% CI 7.4 to 13.7) with the combination vs 5.5 months (95% CI 3.8 to 7.2) with imlunestrant alone (HR 0.53; 95% CI 0.35 to 0.80); ORR 35% (95% CI 22 to 48) vs 15% (95% CI 7 to 23). Overall survival was immature at interim analysis (35% deaths in the ESR1-mutated subgroup). Two practical notes: PARP-inhibitor-eligible patients were excluded from EMBER-3, so germline BRCA carriers remain sequenced toward a PARP inhibitor, and per Medscape's September 21 coverage, grade 3/4 toxicities with the combination included diarrhea (9%), neutropenia (21%), ALT elevation (5%), and AST elevation (2.5%), with roughly 3% ILD/pneumonitis and just under 5% venous thromboembolism.

Camizestrant (September 4): a correction to the ODAC expectation

Worth flagging as an explicit correction: in April 2026 the Oncologic Drugs Advisory Committee voted 6-3 against this application, with panelists arguing there was no evidence that switching from an aromatase inhibitor to a SERD before documented disease progression improves clinical outcomes. The FDA approved it anyway, as an accelerated approval with required confirmatory trials, as detailed in the agency's press announcement. If your working assumption after April was that the switch-before-progression paradigm was dead, correct it: it is now a labeled, ctDNA-triggered strategy. Per the FDA notice, SERENA-6 randomized 315 patients 1:1 (all at least 6 months into first-line AI plus CDK4/6 inhibition without progression) to switch to camizestrant 75 mg daily while continuing the same CDK4/6 inhibitor at the same dose, vs continuing the AI. PFS measured from ESR1 mutation detection was 16 months (95% CI 12.7 to 18.2) vs 9.2 months (95% CI 7.2 to 9.5); HR 0.44 (95% CI 0.31 to 0.60); p < 0.00001. OS immature. The label carries a boxed warning for QTc prolongation with concomitant QTc-prolonging drugs plus bradycardia and embryo-fetal toxicity warnings. Note the regulatory asymmetry: this one is accelerated and contingent; the imlunestrant combination above is a full approval.

Landscape and pending actions

3. Survivorship / Supportive Care

JNCI: chemotherapy and prediagnosis lifestyle shape functional aging trajectories in older survivors

The strongest supportive-care publication of the window is a Women's Health Initiative analysis published in JNCI on September 23, comparing 15-year RAND SF-36 physical function trajectories among women with localized (n = 6,462) or regional (n = 1,872) breast cancer against 37,196 age-matched cancer-free comparators (full cohort n = 45,530; median age at diagnosis 72). Survivors showed acute postdiagnosis declines with incomplete recovery and a persistent deficit vs comparators: regional disease carried larger deficits (-6.8 points at year 1, -5.3 at year 10), and chemotherapy recipients had the steepest early declines (-7.1 at year 1), consistent across agent classes. Grip strength and walking pace trends in a subset (1,045 survivors, 1,970 comparators) mirrored the SF-36 findings. Deficits were greater and more sustained in older women, while women younger than 70 showed smaller differences after the early drop. Healthier prediagnosis lifestyle was associated with greater functional reserve and better recovery.

For a breast oncology APP, this is useful anticipatory-guidance material for older patients weighing chemotherapy: the functional cost can be durable and detectable a decade out, and modifiable prediagnosis lifestyle is an intervention target. The observational design cannot isolate treatment causality, but the cohort size, matched comparators, and 15-year horizon make it one of the more concrete functional-aging datasets we have.

Secondary: surgical extent and quality of life after adjuvant radiotherapy

A small in-window cross-sectional study published September 23 in the European Journal of Breast Health compared 98 women after adjuvant radiotherapy (58 breast-conserving surgery, 40 modified radical mastectomy). Breast conservation was associated with better QLQ-C30 global health status (66.66 vs 37.50; p < 0.001), functional scores (60.00 vs 46.66; p = 0.005), and symptom summary scores (67.94 vs 58.97; p = 0.001). These are unadjusted, single-center comparisons without clinical covariates, so treat this as confirmatory texture for shared decision-making about surgical extent rather than new evidence.

Nothing else substantive surfaced in the survivorship literature within the window; the remainder of my scan returned qualitative work published before the window.

4. Translational / Horizon

Guardrail: nothing in this section is practice-changing today. These are early signals, exploratory analyses, and small prospective studies intended to sketch where the field is moving. Sourcing note: exact publication dates could not be confirmed against primary sources for the items below; each surfaced in this week's literature scan and is included for awareness rather than as a dated event.

Causal AI applied to UNIRAD: Ataraxis Breast CTX

OncoDaily's review describes a prospective-retrospective analysis of the phase III UNIRAD trial (adjuvant everolimus plus endocrine therapy in high-risk, node-positive HR-positive/HER2-negative early breast cancer, negative overall) testing Ataraxis Breast CTX, a causal multimodal AI model integrating H&E-extracted features with clinical variables to estimate residual recurrence risk and predicted chemotherapy benefit. Among 556 clinically high-risk patients, the model separated a low-risk group with 93% 5-year DFS from a high-risk group at 80%. The provocative, and purely exploratory, finding was a treatment interaction: in the low predicted-chemotherapy-benefit group, 5-year DFS was 95% with everolimus vs 85% without (significant by log-rank), while no everolimus benefit was seen in the high-benefit group; the treatment-by-score interaction was HR 1.68 (95% CI 1.12 to 2.53; P = 0.01). The limitations matter more than the headline: only 56% of UNIRAD had evaluable slides with modest baseline imbalances, neoadjuvant-treated patients were excluded, everolimus discontinuation was high, and the model has no validation against CDK4/6 inhibitor escalation, which is the actual contemporary adjuvant decision. Keep this in the "interesting framework, unvalidated predictive claim" bucket.

Dual ctDNA/CTC minimal residual disease score

A prospective study in npj Breast Cancer (n = 58, early-stage disease across HR-positive/HER2-negative, TNBC, and HER2-positive subtypes) tracked a single truncal mutation per patient by droplet digital PCR in both plasma ctDNA and circulating tumor cells, sampled before treatment, about 1 month post-surgery, then every 6 months in high-risk patients. A composite L-MRD score integrating six clinical and molecular variables anticipated all relapses (100% sensitivity; median lead time 27.95 months) and stratified 5-year recurrence risk at 2.2% vs 41.6% (HR 6.6; P = 0.04), with AUC 0.891 and NPV 97.8%. Baseline ctDNA and/or CTCs were detectable in 67.2% and persisted post-surgery in 53.6%. The numbers are striking but the cohort is tiny, the P value fragile, and single-mutation ddPCR tracking is a different technology from the tumor-informed WGS panels dominating MRD trials; the dual-analyte concept is the transferable idea.

DETECT IVb: early CTC clearance as a response readout on eribulin

The phase II DETECT IVb trial, reported in Breast Cancer Research and Treatment (109 women with HER2-negative metastatic breast cancer receiving eribulin), found that CTC clearance at the first on-treatment assessment (at least 21 days in, typically cycle 2 or 3) identified responders early: 34.7% of the 75 evaluable patients cleared CTCs entirely, and those patients had median PFS of 5.7 vs 2.9 months (adjusted HR 0.56 for PFS), while dropping below the 5-cell threshold associated with better OS (HR 0.45 for death). Baseline counts of 5 or more CTCs per 7.5 mL remained independently prognostic (63% higher progression risk after adjustment). This is a biomarker-association study in a single-arm phase II, not an interventional test of CTC-guided switching, and the authors call for exactly that. It is a useful data point for the early-response-monitoring conversation.

Cuts Made (held back for length or out of window; ask for a deeper dive on any):

Archive