A Validated T Cell Radiomics Score Is Associated With Clinical Outcomes Following Multisite SBRT and Pembrolizumab

Purpose: Combining immune checkpoint blockade (ICB) with stereotactic body radiation therapy (SBRT) may improve the local response to radiation and the systemic response to immunotherapy. However, no prognostic markers exist to identify patients likely to benefit from combined therapy. The degree of T cell–mediated immunity, which can be quantified with radiomics using computed tomography (CT) imaging, is predictive of immunotherapy response. Herein we investigated whether a validated T cell radiomics score (RS) is correlated with clinical outcomes after multisite SBRT and pembrolizumab (SBRT + P). Methods and Materials: The RS was quantified for 68 patients with metastatic treatment-refractory adult solid tumors who received SBRT (30-50 Gy, 3-5 fractions) and pembrolizumab ≤7 days after SBRT. RS was calculated using 8 variables, including 5 radiomics features extracted from pretreatment CT scans. At a prespecified cutoff of the 25th percentile, we assessed the association between RS and clinical outcomes. The Kaplan-Meier method was used to estimate survival outcomes. The prognostic effect of RS was assessed via logistic regression or Cox proportional hazards models. In an exploratory analysis, RS was also analyzed as a continuous variable. Results: One hundred thirty-nine tumors were analyzed. At the 25th percentile cutoff, high-RS patients were more likely to exhibit irradiated tumor responses to SBRT + P (odds ratio [OR] 10.2; 95% confidence interval [CI], 1.76-59.17; P =.012). High-RS was associated with improved TMC compared with low-RS tumors (hazard ratio [HR] 0.18; 95% CI, 0.04-0.74; P =.018). Furthermore, high-RS patients had improved PFS (HR 0.47, 95% CI, 0.26-0.85; P =.013) and OS (HR 0.39, 95% CI, 0.20-0.75; P =.005). As a continuous variable, higher RS was associated with improved PFS (HR 0.12, 95% CI, 0.03-0.51; P =.004) but did not reach statistical significance for TMC (HR 0.36, 95% CI, 0.02-7.02; P =.502) or OS (HR 0.28, 95% CI, 0.05-1.55; P =.144). Conclusions: We demonstrated the clinical validity of RS (at the 25th percentile cutoff) as a prognostic biomarker in patients treated with SBRT + P. Future validation of the prognostic value of RS in larger similarly treated patient cohorts is warranted.