[18F]F-AraG Uptake in Vertebral Bone Marrow May Predict Survival in Patients with Non–Small Cell Lung Cancer Treated with Anti-PD-(L)1 Immunotherapy

Despite the systemic impact of both cancer and the associated immune response, immuno-PET is predominantly centered on assessment of the immune milieu within the tumor microenvironment. The aim of this study was to assess the value of [¹⁸F]F-AraG PET imaging as a noninvasive method for evaluation of system-wide immune status of patients with non–small cell lung cancer before starting immunotherapy. Methods: Eleven patients with advanced non–small cell lung cancer were imaged with [¹⁸F]F-AraG before starting immunotherapy. Diagnostic [¹⁸F]FDG PET/CT scans were analyzed to assess differences in the extent of disease among patients. SUV_max, SUV_mean, and total SUV (SUV_total) from all tumor lesions, active lymph nodes, spleen, vertebral bone marrow, liver, thyroid, heart, and bowel were extracted from the baseline [¹⁸F]F-AraG scans, and discriminant and Kaplan–Meier analyses were performed to test their ability to predict patient response and overall survival. Results: The extent of the disease was variable in the patient cohort, but none of the [¹⁸F]FDG biomarkers associated with tumor burden (SUV_max, total metabolic tumor volume, and total lesion glycolysis) was predictive of patient survival. The differences in the [¹⁸F]F-AraG and [¹⁸F]FDG distribution were observed both within and between lesions, confirming that they capture distinct aspects of the tumor microenvironment. Of the 3 SUV parameters studied, [¹⁸F]F-AraG SUV_total provided a dynamic range suitable for stratifying tumors or patients according to their immune activity. [¹⁸F]F-AraG SUV_total measured in the lumbar and sacral vertebrae differentiated between patients who progressed on therapy and those who did not with 90.9% and 81.8% accuracy, respectively. The Kaplan–Meier analysis revealed that patients with high [¹⁸F]F-AraG SUV_total in the lumbar bone marrow had significantly lower probability of survival than those with a low signal (P = 0.0003). Conclusion: This study highlights the significance of assessing systemic immunity and indicates the potential of the [¹⁸F]F-AraG bone marrow signal as a predictive imaging biomarker for patient stratification and treatment guidance.