Mali Barbi, MD MSc | Breast & Gyn Oncologist

Recent #RxPONDER analysis (the Oncotype one) added AMH analysis. and it made me rethink my own practice. The premenopausal chemo benefit tracks ovarian reserve, not menopause. AMH ≥10, chemo helps (HR 0.46). AMH <10, it doesn’t (HR 1.27). OPTIMA then swapped ovarian suppression for chemo in low-risk tumors and matched it. “Premenopausal” was never the real variable. Measure, don’t assume. #bcsm #MedOnc #breastcencer Ovarian reserve as a measure of adjuvant chemotherapy benefit in hormone receptor positive (HR-positive), HER2-negative, node-positive breast cancer in SWOG S1007 (RxPONDER)☆ - Annals of Oncology https://t.co/FWLQBTmOgc

#OPTIMA addresses a real and long-standing tension in how we treat premenopausal node-positive ER+ breast cancer. The signal is meaningful: PAM50 ROR&lt;60 patients who received adequate OFS did not appear to benefit from chemotherapy. That is a hypothesis worth taking seriously. https://t.co/m0YBV3LIzh

#ASCO26 | We've been sparing node-negative #breastcancer patients from chemo using #Oncotype. #OPTIMA just did it in the population we've been most afraid to touch. 4,400+ patients. ER+/HER2−. Node-positive, up to pN2. #Prosigna_ROR: ≤60 → no chemo. 5-year IBCFS: 90.4% vs

#ESMO2026, my read. What changes my practice, and what doesn’t yet. #BreastCancer OptiTROP-Breast01: OS HR 0.60, and the gap held despite crossover to TROP2 ADCs. In TNBC, the ADC goes early. BRIDGET: isolated brain progression in HER2+, local therapy plus tucatinib, keep the backbone. Single arm, historical control, but I’d use it. @drsarahsam NATALEE OS: HR 0.81, nominal. iDFS remains the proven benefit. Paired biopsies after T-DXd: HER2 rarely lost, efflux and uptake change instead. A repeat biopsy that only re-checks HER2 explains little. #Endometrialcancer TroFuse-005: sac-TMT beat doxorubicin or weekly paclitaxel after chemo-IO, PFS HR 0.59, with an OS benefit reported. First ADC with survival in endometrial cancer. I want the subgroups before I call it for everyone. KEYNOTE-C93: pembro alone vs chemo in dMMR, PFS HR 0.57, OS not significant. Real option for patients avoiding chemo. But about 20% progress early, and we can’t identify them upfront. GY020: negative and underpowered. Radiation alone stays my standard in early-stage dMMR. DUO-E: in pMMR, durvalumab OS HR 0.95. Olaparib bought PFS with twice the grade ≥3 toxicity. Never tested against durvalumab alone. #Ovariancancer RAINFOL-02: rina-S OS HR 0.43 without FRα selection. Press release only so far. If it holds, FRα stops being the gate. The next question is whether it works earlier. I am bringing the first-line maintenance trial of rina-S to @NorthwellHealth, for newly diagnosed non-HRD ovarian cancer, where PARP inhibitors help least. BEHOLD-1: response tracked with linker cleavage and TOP1, not B7-H4 level. Exploratory, but it changes what the next biomarker should measure. #ctDNA Positive means risk. One negative draw means little. >>>Most of what we saw improves PFS. Fewer trials improved survival, and almost none told us who benefits. That’s the work now. @myESMO @OncoAlert #ESMO26 #bcsm #gyncsm

#ESMO2026. The ADC question I am watching this week is not which target. It is which payload, and in what order.
Monday, one breast session: SWITCH (4RO), ADC after ADC. OptiTROP-Breast01 (4639RO), final OS for sac-TMT in pretreated TNBC. Bulumtatug fuvedotin (4640RO), MMAE after a TopoI ADC.
Same day, Presidential Symposium: TroFuse-005 (LBA11), sac-TMT in previously treated endometrial cancer.
TROP2 is expressed in most endometrial tumors. If sac-TMT works, the target was never the hard part. The hard part is what comes after. I give the best treatment first. The tumor that progresses is then already resistant to that payload, and the next line needs a different mechanism. That is where we need more small molecule targeted drugs, not another antibody carrying the same payload. #gyncsm #bcsm

Three new ADCs at ESMO 2026, three targets: HER2 (TQB2102, APEX-Breast01), EGFR x HER3 (iza-bren, PANKU-Breast01), nectin-4 (SHR-A2102, cervix). One payload: TOPO1 inhibitor. In cervix, the approved ADC is tisotumab, an MMAE payload, so a TOPO1 ADC after it is a real mechanism switch. In breast after T-DXd, a new target on the same payload is a different bet, and cross-resistance is the open question. The trial I want randomizes by payload sequence, not by target. #gyncsm #ESMO26 @myESMO

#PANKUBreast01: iza-bren, an EGFR x HER3 bispecific ADC with a TOPO1 payload, vs chemo in HR+/HER2- MBC n=383. PFS 8.2 vs 4.0 mo, HR 0.41. ORR 40.5% vs 20.2%. A real result, and it enrolled HER2 0, 1+ and 2+, so no HER2 test gates it. In clinic it lands in a space already holding T-DXd, Dato-DXd and sacituzumab. All TOPO1. Iza-bren is a new target, not a new mechanism. The number I want is how many patients had a prior ADC, and how they did. #bcsm #ESMO26 @myESMO

FDA approved #HER2CLIMB05. Metastatic HER2+ #breastcancer is now crowded, and there is no sequencing trial. Until there is one, here is how I think about it. Most of my patients will see T-DXd in the neoadjuvant or adjuvant setting. When they recur, HER2CLIMB-05 with tucatinib comes into play. If she is HR+, palbociclib per PATINA. If she has not seen T-DXd, I start with T-DXd. Brain metastases are the hard part. If I use T-DXd up front and she develops brain metastases, tucatinib is what I have left. So do I hold T-DXd for the brain, or use it first and keep tucatinib for the brain? I use it first as soon as I can and hope to better sequencing trials or new agents for subsequent lines

#PATINA got @FDAOncology approved today. #Palbociclib + #trastuzumab ± #pertuzumab + ET as 1L maintenance in HR+/HER2+ mBC. 44.3 vs 29.1 months PFS. A 15-month gain in a chemo-free maintenance phase. Including patients with asymptomatic CNS mets. And I'm sitting here thinking

We're all hearing about the #camizestrant approval, so here are my two cents. The FDA cleared it in HR+/HER2- advanced breast cancer for patients who develop an ESR1 mutation in ctDNA on first-line endocrine therapy, before the scan progresses. It cleared because switching at that molecular signal, instead of waiting for radiographic progression, roughly doubled PFS, 16 vs 9.2 months. Accelerated approval, so no survival data yet, and part of that PFS gap may just reflect leaving the control arm on a failing drug. Where I see it landing: the trigger to change therapy moves from the scan to the blood. That is the real shift, bigger than one drug. But two practical questions decide whether it works in clinic, and neither is settled. 1- How often would I draw the ctDNA? SERENA-6 monitored every two to three months. That is a real testing burden on patients who feel well and look stable on imaging. 2- And in whom? You would monitor everyone on first-line AI plus CDK4/6i, but only the fraction who turn ESR1 positive get the drug. In SERENA-6 that meant screening thousands to randomize a few hundred. The burden of testing lands on a much larger group than the treatment does. So the open question is not really the drug. It is whether serial monitoring of stable patients earns its place, when the testing falls on many and the benefit reaches few. I am not convinced yet. #bcsm

The FDA-EMA split on SERENA-6 perfectly captures the current tension in breast oncology: molecular vs. radiographic progression. Interestingly, we are seeing a complete role reversal here, with the FDA wearing the conservative hat, which is highly atypical for them in the oncology space. The EMA embraced ctDNA ESR1 detection as a valid trigger for early intervention, securing a massive PFS win. Meanwhile, the FDA and ODAC are holding the line, demanding mature OS data before officially shifting the treatment paradigm. Two valid regulatory philosophies, but a complex landscape to navigate for our patients

#ESMO2026, my read. What changes my practice, and what doesn’t yet. #BreastCancer OptiTROP-Breast01: OS HR 0.60, and the gap held despite crossover to TROP2 ADCs. In TNBC, the ADC goes early. BRIDGET: isolated brain progression in HER2+, local therapy plus tucatinib, keep the backbone. Single arm, historical control, but I’d use it. @drsarahsam NATALEE OS: HR 0.81, nominal. iDFS remains the proven benefit. Paired biopsies after T-DXd: HER2 rarely lost, efflux and uptake change instead. A repeat biopsy that only re-checks HER2 explains little. #Endometrialcancer TroFuse-005: sac-TMT beat doxorubicin or weekly paclitaxel after chemo-IO, PFS HR 0.59, with an OS benefit reported. First ADC with survival in endometrial cancer. I want the subgroups before I call it for everyone. KEYNOTE-C93: pembro alone vs chemo in dMMR, PFS HR 0.57, OS not significant. Real option for patients avoiding chemo. But about 20% progress early, and we can’t identify them upfront. GY020: negative and underpowered. Radiation alone stays my standard in early-stage dMMR. DUO-E: in pMMR, durvalumab OS HR 0.95. Olaparib bought PFS with twice the grade ≥3 toxicity. Never tested against durvalumab alone. #Ovariancancer RAINFOL-02: rina-S OS HR 0.43 without FRα selection. Press release only so far. If it holds, FRα stops being the gate. The next question is whether it works earlier. I am bringing the first-line maintenance trial of rina-S to @NorthwellHealth, for newly diagnosed non-HRD ovarian cancer, where PARP inhibitors help least. BEHOLD-1: response tracked with linker cleavage and TOP1, not B7-H4 level. Exploratory, but it changes what the next biomarker should measure. #ctDNA Positive means risk. One negative draw means little. >>>Most of what we saw improves PFS. Fewer trials improved survival, and almost none told us who benefits. That’s the work now. @myESMO @OncoAlert #ESMO26 #bcsm #gyncsm

TROPION-Breast02, brain mets subgroup at ESMO 2026. First-line Dato-DXd vs chemo in mTNBC, n=61. CNS PFS 9.5 vs 2.6 mo, HR 0.27. Intracranial ORR 62.5% vs 37.5%. Longest intracranial response about 27 mo. In TNBC, brain mets come early and we usually treat them locally first. This makes me ask whether a patient with small asymptomatic lesions could start Dato-DXd and defer /combine SRS Not ready for clinic. This is a post hoc subgroup, n=30 with untreated brain mets. And even on Dato-DXd, median OS was 11.6 mo with brain mets vs 23.7 mo without. The CNS benefit is real; the prognosis is still poor. The answer needs a dedicated CNS trial. #bcsm #ESMO26 @myESMO

Not my disease, but DeLLphi-305 is probably the result of ESMO 2026. #Tarlatamab, a DLL3 x CD3 T-cell engager, added to durvalumab as first-line maintenance in extensive-stage small cell lung cancer. Median OS 21.1 vs 12.7 mo, HR 0.42. 12-month OS 79% vs 53%. Why this matters: for 30 years, median survival in ES-SCLC sat around 10 to 13 months. Adding IO to chemo moved it about 2 months. This moves it about 8, with an HR rarely seen in any metastatic disease. Three points beyond lung:
A T-cell engager works in a solid tumor.
Maintenance is where you add the new drug: after induction, on lower disease burden.
DLL3 is a neuroendocrine target. Cost: CRS and ICANS need monitoring infrastructure. #lcsm #ESMO26 @myESMO