A Phase II Study of Stereotactic Ablative Radiotherapy plus Atezolizumab plus Tiragolumab in Treatment-Naïve Patients with Advanced Non–Small Cell Lung Cancer
Abstract
PURPOSE: SKYROCKET is a single-center, single-arm phase II study that evaluated whether the addition of stereotactic body radiotherapy (SBRT) to atezolizumab plus tiragolumab enhances antitumor efficacy in PD-L1-positive non-small cell lung cancer (NSCLC). PATIENTS AND METHODS: Patients received SBRT to all metastatic sites and subsequently received atezolizumab and tiragolumab every 3 weeks on day 1 of each 21-day cycle with 7 days of completion of SBRT. The primary endpoint was investigator-assessed progression-free survival (PFS) from the start of SBRT. Exploratory endpoints included single-cell RNA sequencing (scRNA-seq) obtained from tumor biopsies at baseline (BL) and on-treatment (OT) and were further characterized as clinical benefit [CB; partial response or stable disease (SD) ≥ 6 months] and non-clinical benefit (NCB; progressive disease or SD < 6 months). RESULTS: Forty-one patients were enrolled. At a median duration of follow-up of 9.8 months, median PFS at the start of SBRT was 9.3 months [95% confidence interval (CI), 6.0-NR]. No new safety signal was observed. scRNA-seq of paired BL and OT (n = 3) and single-time point (n = 2) revealed that CD8 progenitor-exhausted (TPEX) cells markedly expanded after treatment, with high T-cell immunoreceptor with Ig and ITIM (TIGIT) and PD-1 expression in CD8 TPEX/TEX subsets. CD4 regulatory T cells (Treg) were reduced in the CB group but increased in the NCB group. The CB group showed reduced Treg suppression and NECTIN-TIGIT signaling, whereas NCB maintained proliferative CTLA4high/TIGIThigh Treg supported by ICAM/galectin-CTLA4 pathways. CONCLUSIONS: SBRT followed by atezolizumab plus tiragolumab showed encouraging clinical activity with manageable safety in PD-L1-positive metastatic NSCLC.
Fields of Study & Subjects
Cancer Immunotherapy and Biomarkers
Single-cell and spatial transcriptomics
Radiomics and Machine Learning in Medical Imaging
Publication details
- Journal / source
- Clinical Cancer Research
- Publisher
- American Association for Cancer Research (AACR)
- Volume
- 32
- Issue
- 14
- Pages
- 2898–2910
- Publication date
- 20 April 2026
- Language
- EN