Network-Based Analysis for the Quantification of Brain and Body Immune Axes with Total-Body PET Imaging

The brain has long been viewed as an isolated immune-privileged organ. However, growing evidence suggests a critical role of the interplay between central and peripheral inflammation in pathologic conditions. This highlights the urgent need for novel system-level analysis approaches to study the complex brain–body cross-talk during immune responses. This study aims to validate network analysis of total-body 18-kDa translocator protein (TSPO) PET imaging for studying brain and body immune axes. Methods: Two graph-based analysis frameworks were tested with total-body PET imaging and the third-generation TSPO radioligand [18F]LW223 in 2 mouse models of lipopolysaccharide-induced systemic infection (n = 31) and pharmacologic blocking (n = 19) with LW223. First, a perturbation covariance approach was adopted in the study of individualized deviation from the normative/reference interorgan immune covariance network; then, an interscan connectivity analysis was used to investigate intragroup and between-group similarities in whole-body TSPO expression patterns. Results: Application of the perturbation covariance approach showed increased deviations from the reference interorgan covariance network 2 h (t-statistic = 3.74; P = 8.88 × 10−4) and 24 h (t-statistic = 4.02; P = 4.23 × 10−4) after lipopolysaccharide challenge, with a 7-d recovery after infection (t-statistic = 0.03; P = 0.98), reflecting previous general evidence of a time-dependent immune response to lipopolysaccharide infection. Additionally, the perturbation approach revealed a widespread dose-specific increase in deviations after blocking agent administration, with a magnitude of deviations and several extreme deviations increasing with dose (F-statistic = 37.67; P = 2.33 × 10−6). Interscan connectivity analysis confirmed strong alterations from the physiologic whole-body TSPO expression pattern 24 h after the lipopolysaccharide challenge and after administration of high-dose blocking agent. Conclusion: To our knowledge, this study represents the first application of network analysis to total-body TSPO PET data, supporting its adoption in the study of systemic immune responses for diagnosing immune-related conditions and evaluating immune therapies.