Comparative effectiveness of trastuzumab deruxtecan versus real-world treatment in HER2-amplified advanced solid tumors detected by cell-free DNA: An exploratory analysis of the HERALD trial and GOZILA study.

Abstract

833 Background: In the phase II basket trial HERALD, trastuzumab deruxtecan (T-DXd) showed favorable clinical outcomes for patients with HER2-amplified advanced solid tumors detected in cell-free DNA (cfDNA) (J Clin Oncol 2024). However, T-DXd benefit over real-world treatment is unclear. Randomized trials are challenging due to the rarity of this population. Using data from HERALD and the contemporaneous cfDNA-based screening study GOZILA, from which HERALD patients were recruited, we compared clinical outcomes of patients receiving T-DXd with real-world treatment. Methods: Patients with advanced solid tumors and HER2 amplification identified by Guardant360 in GOZILA were included. Those enrolled in HERALD comprised the T-DXd arm; those receiving physician’s choice treatment (PCT) were in the comparator arm. Patients included in the PCT arm had ≥1 prior line of therapy (≥2 for gastric/colorectal cancer) and initiated PCT within 63 days of cfDNA collection. Patients were excluded if they ever received T-DXd or had tissue-based HER2-positive gastric cancer (detection only in cfDNA was allowed). The primary endpoint was overall survival (OS), and the secondary endpoints were progression-free survival (PFS) and objective response rate (ORR). Results: Among 4,734 patients screened from December 2019 to January 2022, 252 had HER2 amplification; 104 patients with 18 cancer types were included (T-DXd, n = 62; PCT, n = 42). Common cancers were colorectal (n = 34), esophageal (n = 15), biliary tract (n = 9), and gastric (n = 7). The T-DXd and PCT arms were balanced for age (median, 64 vs 63 years), sex (male, 48% vs 67%), prior treatment line (median, 3 vs. 3), and plasma HER2 copy number (median, 8.64 vs 6.73). There were fewer gastrointestinal (GI) cancers in the T-DXd arm (55% vs 95%). In the PCT arm, 100% received chemotherapy, and 45.2% received HER2-targeted therapy. At a median follow-up of 10.8 months (mo) for all patients, OS was significantly longer with T-DXd than with PCT (median, 14.5 vs 4.6 mo; HR 0.44; 95% CI 0.28–0.68; P = 0.0003). PFS was prolonged with T-DXd (median, 6.6 vs 1.8 mo; HR 0.30; 95% CI 0.20–0.46; P <0.0001), and ORR was higher with T-DXd (56.5% vs 11.9%; P <0.0001). Similar results were observed in the subset of patients with GI cancers (Table). Conclusions: Despite potential biases and limitations, T-DXd demonstrated clinical benefits over real-world treatments for patients with advanced tumors with HER2-amplification detected in cfDNA. Efficacy Overall GI cancers T-DXd(n = 62) PCT(n = 42) T-DXd(n = 34) PCT(n = 40) Median OS, mo 14.5 4.6 10.8 4.6 HR (95% CI) 0.44 (0.28–0.68) 0.64 (0.39–1.06) Median PFS, mo 6.6 1.8 5.1 1.8 HR (95% CI) 0.30 (0.20–0.46) 0.44 (0.27–0.72) ORR, % 56.5 11.9 44.1 12.5