258 - Exploratory Analysis of Plasma Protein Abundance Kinetics from I-SABR, A Randomized Phase II Trial of Stereotactic Ablative Radiotherapy with or without Nivolumab in Stage I-IIA or Recurrent Non-Small

A. Seo¹, S. H. Lin², Z. Liao³, S. Gandhi², M. M. Mirza³, R. Maguire⁴, S. G. Swisher⁵, J. Heymach⁶, and J. Y. Chang²; ¹Division of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, ²Department of Thoracic Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, ³The University of Texas MD Anderson Cancer Center, Houston, TX, ⁴Department of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, ⁵Department of Thoracic and Cardiovascular Surgery, The University of Texas MD Anderson Cancer Center, Houston, TX, ⁶Department of Thoracic/Head and Neck Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX

Purpose/Objective(s): In early-stage non-small cell lung cancer (NSCLC), combined immunotherapy with stereotactic ablative radiotherapy (I-SABR) showed a significant 58% improvement in 4-year event-free survival (EFS) in our recent randomized phase 2 trial (NCT03110978). In this exploratory study, we aimed to characterize the systemic immune response changes with I-SABR and identify markers associated with EFS from our trial. Materials/

Methods: From June 30, 2017 to March 22, 2022, patients with treatment-naïve, Stage I-IIA NSCLC or isolated parenchymal recurrences <= 7 cm were randomly assigned to SABR or SABR with 4 cycles of nivolumab (first dose on the same day or within 36 hours after the first SABR fraction). Primary endpoint results were previously reported. Plasma samples were collected at three timepoints: pre-treatment, post-treatment, and in 2-3 month follow-up. Abundances of 250 inflammatory proteins were quantified using NULISAseq (Alamar Biosciences). Linear regressions were done to assess differential abundance across timepoints and across treatment arms. Cox proportional hazards regression and Kaplan-Meier survival analyses were done to correlate protein abundances with EFS.

Results: In total, 228 samples were tested. From pre- vs. post-treatment timepoints, 55 proteins were significantly differentially abundant in I-SABR but not in SABR after false discovery rate (FDR) adjustment, including an increase in immune activation markers such as IL2, IFNG, CXCL9, and CXCL10. The abundance of CXCL9 and CXCL10, attractants for CD8+ T, NK cells, and M1 macrophages, continued to rise in follow-up with I-SABR. The abundance of GZMB, important for NK and CD8+ T-cell effector function, was initially stable in I-SABR pts post-treatment but decreased with SABR, with abundance significantly increased in I-SABR but not SABR at follow-up. The abundance of PD-1, the target of nivolumab, significantly decreased in I-SABR patients and subsequently rose in follow-up compared to stably unchanged levels in SABR patients. Immunosuppressive markers such as CTLA4, LAG3, IL10, and IL12 rose post-treatment with I-SABR but were stable with SABR. CTLA4 and LAG3 continued to rise at follow-up with I-SABR. Increased GZMB abundance from post-treatment to follow-up in I-SABR was associated with lower EFS (HR 5.74, 95% CI 2.46-13.42).

Conclusion: Markers of immune activation and exhaustion dynamically change during treatment and in follow-up. Increased CTLA4 and PD-L1 suggest a regulatory response after I-SABR by follow-up. The association between increased GZMB abundance at follow-up with I-SABR and lower EFS may indicate CD8+ T and NK cells are more exhausted. Evaluating the exhaustion landscape of peripheral blood cells may provide an avenue to further improve patient outcomes.