Abstract Background: TUXEDO-3 (NCT05865990) is an international, multicenter, multicohort, single-arm, phase II trial that enrolled patients with metastatic breast cancer (mBC) and newly diagnosed brain metastases (BMs) or BMs progressing after local therapy (active BMs) (cohort 1), advanced non-small cell lung cancer and active BMs (cohort 2), and leptomeningeal disease (LMD) from any solid tumors (cohort 3) to evaluate the efficacy and safety of the antibody-drug conjugate (ADC) patritumab deruxtecan (HER3-DXd). The treatment showed central nervous system (CNS) activity in all three cohorts. In cohort 1, HER3-DXd activity was observed irrespective of the BC subtype in a heavily pretreated population, showing that HER3-DXd may be a promising novel therapeutic option for patients with breast cancer and CNS involvement. The number of available ADCs across stages and lines of therapy in breast cancer is growing, but evidence on their optimal sequencing remains limited, highlighting an important unmet need, especially in the context of CNS involvement. Here, we performed an unplanned post-hoc, descriptive, pooled efficacy analysis of all patients with HER2-positive mBC and CNS involvement from the TUXEDO-3 trial who had received prior therapy with T-DXd. Methods: The study design and results of the TUXEDO-3 trial have been recently reported (Preusser et al., Nat Med, 2025) and presented at ASCO 2025. Patients with HER2-positive mBC with active BMs and CNS involvement that received prior ADC were included in this post hoc analysis. We evaluated progression-free survival (PFS) for intracranial disease as per RANO-BM criteria, PFS for extracranial and bicompartmental (intracranial and extracranial) disease as per RECIST v.1.1, and overall survival (OS). PFS and OS were estimated using the Kaplan-Meier method, reporting the median with 95% confidence intervals (CI) based on the log-log method. Results: Between 12 December 2023 and 2 July 2024, 35 patients with mBC and CNS involvement active BMs (n=21) and LMD (n=14) were enrolled across cohorts 1 and 3. Of them, 10 (28.6%) had HER2-positive mBC, including 9 (90.0%) with active BMs and 1 (10%) with LMD, who also had parenchymal BMs. Median age was 54.5 (range: 35.0-75.0) years. The number of previous treatment lines for advanced disease in the HER2-positive mBC population was 4.0 (range: 2.0-7.0); all patients had received prior T-DXd, three (30.0%) had additional T-DM1, and six (60%) tucatinib as well. Three patients (30%) received T-DXd as last treatment line prior to enrollment. At data cutoff (28 February 2025), one patient remained on treatment and six continued in follow-up. Median follow-up was 4.7 (range: 2.0-10.3) months, and median treatment duration was 3.2 (range: 0.0-9.7) months. Median intracranial and bicompartmental PFS were 6.3 months (95%CI: 1.4-Not reached), whereas median extracranial PFS and OS were not reached. Conclusions: In this post hoc analysis of the prospective TUXEDO-3 trial, HER3-DXd showed promising PFS and OS in patients with HER2-positive mBC and active BMs or LMD, previously treated with T-DXd. Despite the small sample size, these findings suggest a potential strategy for sequential ADC treatment in a patient population with no established standard of care. With a median PFS exceeding 6 months, further investigation of HER3-DXd in patients progressing on T-DXd is warranted. Citation Format: R. Bartsch, A. Llombart-Cussac, M. Gión, J. Garde-Noguera, R. Greil, M. Ruiz-Borrego, M. Valero, J. García-Mosquera, M. Arumi, J. Cortés, M. Campolier, J. Guerrero, J. Rodríguez-Morató, P. González-Alonso, I. Mas-Rincón, M. Vaz-Batista, M. Marhold, F. Oberndorfer, J. Furtner, T. Fuereder, A. Berghoff, M. Preusser. Outcome of patritumab deruxtecan (HER3-DXd) in patients with HER2-positive metastatic breast cancer and CNS involvement previously treated with T-DXd: A subanalysis of TUXEDO-3 abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PD13-10.
Bartsch et al. (Tue,) studied this question.