ESMO Congress 2026

Anti-HER2 therapy is a template of targeted treatment: HER2 gene amplification and protein overexpression are present in 25% to 30% of breast cancers and make the tumor more aggressive; trastuzumab plus chemotherapy extended median time to disease progression from 4.6 to 7.4 months in metastatic breast cancer ^[2], and concurrent trastuzumab with chemotherapy also established benefit in the adjuvant setting for operable HER2-positive early breast cancer ^[5]. At the molecular level, the HER2-enriched subtype is characterized by activation of the HER2/phosphorylated HER2/EGFR pathway ^[1]; in the neoadjuvant setting, pCR has been validated through pooled analysis as associated with EFS and OS and usable as a surrogate for long-term benefit ^[6], which provides a reference frame for reading another pCR-based neoadjuvant dataset presented at this meeting.

The main HER2 storyline of the ESMO26 breast cancer oral and proffered paper sessions centered on ADCs in HER2-low breast cancer: Barbara Pistilli discussed the first three abstracts, which each tested a different HER2 ADC yet shared the same target, the same linker, and the same payload class. The most watched was China's TQB2102 (biparatopic HER2 TOP1 ADC), which reported PFS of 16.4 vs 5.5 months, HR 0.29, p<0.01 in mostly HR+/HER2-low metastatic breast cancer; the posting expert also noted that whether the biparatopic design adds anything remains unclear, that the HER2 TOP1 field is already quite crowded with excellent agents, and that the ADC landscape in HR+/HER2- metastatic breast cancer is likewise crowded, with optimal sequencing still unresolved.

Another HER2-related readout at the same meeting was negative: a meta-analysis pooling 8 studies and 2,529 patients evaluated adding carboplatin to neoadjuvant taxane plus dual HER2 blockade, with pCR rates of 62.2% vs 61.7%, showing no significant improvement, including across the evaluated clinical subgroups. Against the background that pCR is validated as associated with long-term outcomes ^[6], this suggests routine platinum intensification in the dual-blockade era lacks support. The neoHIP discussion also indicated that predictive biomarkers for immune sensitization are not yet settled: benefit from pembro in the ER+/HER2+ population was seen only in high TIL or PD-L1+ cases, differing in direction from IMpassion050, where benefit was seen in PD-L1-/HR+ patients.

Taken together, TQB2102 extends the reach of anti-HER2 agents from the classic amplified/overexpressed population ^[2] to HR+-predominant HER2-low advanced breast cancer; the limitations are equally clear: the sample size, trial phase, and the specific comparator regimen of this readout were not disclosed in the conference posts, the added value of the biparatopic design is in question, and optimal sequencing among multiple same-class ADCs, as well as the place of carboplatin in current neoadjuvant therapy, await more mature data.