High mechanical conditioning by tumor extracellular matrix stiffness is a predictive biomarker for anti-fibrotic therapy in HER2-negative breast cancer

Abstract Purpose: Tumor progression has been linked to stiffening of the extracellular matrix caused by fibrosis. Cancer cells can be mechanically conditioned by stiff extracellular matrix, exhibiting a 1,004-gene signature [mechanical conditioning (MeCo) score]. Nintedanib has demonstrated antifibrotic activity in idiopathic pulmonary fibrosis. This study explores nintedanib’s antifibrotic effect on breast cancer outcomes. Experimental Design: We present long-term follow-up and analysis of a neoadjuvant randomized phase II trial in early HER2-negative breast cancer. Patients (N = 130) underwent a baseline biopsy and received 12 paclitaxel courses alone (control arm) or in combination with nintedanib (experimental arm). The tumor MeCo score was determined by RNA sequencing. The primary aim was to assess nintedanib’s impact on event-free survival based on MeCo scores. Results: Follow-up data were retrieved from 111 patients; 75 baseline and 24 post-run-in phase samples were sequenced. After median follow-up of 9.67 years, median event-free survival was not statistically different between arms (P = 0.37). However, in the control arm, high- versus low-MeCo patients had a statistically higher relapse risk: HR = 0.21; P = 0.0075. This risk was corrected by nintedanib in the experimental arm: HR = 0.37; P = 0.16. Nintedanib demonstrated pharmacodynamic engagement, reducing the MeCo score by 25% during the run-in phase (P < 0.01). Patients with low MeCo after run-in had the best long-term prognosis (HR = 0.087; P = 0.03). Conclusions: High MeCo is predictive of poor outcomes in HER2-negative early breast cancer, although this risk can be mitigated by nintedanib, which is able to specifically reduce MeCo.