ESMO Congress 2026

The druggability of TROP2 as an ADC target has a clear foundation in the literature. Trop2 is a transmembrane glycoprotein encoded by Tacstd2, overexpressed in many cancers with prognostic significance, and involved in signaling for self-renewal, proliferation, invasion, and survival ^[3]; its overexpression quantitatively stimulates tumor growth in proportion to expression levels, and tumor cell growth is abrogated by knockdown ^[6]. On the targeting side, datopotamab deruxtecan (Dato-DXd) binds TROP2 specifically, is internalized and trafficked to lysosomes with DXd release inducing DNA damage and apoptosis in preclinical studies, and delivered tumor regression across TROP2-expressing models including NSCLC PDX ^[4]; an earlier review also documented an ongoing clinical study of IMMU-132 (anti-TROP2 antibody linked to SN38) in epithelial cancers ^[3].

Two parallel TROP2 readouts appeared at ESMO26. One is the TROPION-Breast02 brain metastases analysis reported by Dent: among 61 first-line mTNBC patients with brain metastases, Dato-DXd versus chemotherapy showed CNS PFS of 9.5 vs 2.6 months, HR 0.27; icORR in measurable lesions was 62.5% vs 37.5%; in patients with untreated brain metastases, OS HR was 0.35, and the longest icDoR was ~27 months. The other is the phase II study of BAT8008 (TROP2 ADC) plus BAT1308 (PD-1) in first-line nonAGA NSCLC: 104 patients, China only, ORR 61%, stratified by PD-L1 status as 47% negative, 59% low, 83% high; mPFS 12 months, with 7.7 months in squamous and NR in non-squamous; TRAEs include myelosuppression and stomatitis, and outcomes look better in non-squamous disease.

Relative to prior knowledge, three steps forward. The previous evidence for Dato-DXd centered on systemic antitumor activity and mechanism of action ^[4]; this readout provides directly quantified intracranial efficacy, and the OS HR 0.35 in untreated brain metastases is a harder endpoint. The BAT8008 combination sketches a chemotherapy-free first-line option, and its ORR gradient rising with PD-L1 expression aligns with Egbert Smit's position at the meeting—ADCs in NSCLC need biomarker-driven patient selection and mechanistic explanation of outcome differences across payloads, and SCC histology should be excluded from TROP2 ADCs given detrimental efficacy—while the split of 7.7 months in squamous versus NR in non-squamous for BAT8008 offers corroboration at the same venue.

The limitations are equally plain: the brain metastases analysis is exploratory with n=61, and the posting experts themselves flagged it as exploratory, small n, but a real CNS signal; the BAT8008 data come only from a 104-patient phase II study in China, with limited evidence grade, and extrapolation warrants caution.