The study included 29 patients with locally advanced gastric adenocarcinoma who were enrolled in the Neo-PLANET trial at Zhongshan Hospital between 2018 and the present analysis. All participants underwent pre-operative [18F]FDG PET/CT imaging before starting a treatment regimen that combined immune checkpoint inhibitors with chemotherapy and radiation given before surgery. Using semiautomated image analysis, the team extracted a battery of metabolic parameters from the primary tumor, regional lymph nodes, and spleen, including maximum and mean standardized uptake values adjusted for lean body mass, metabolic tumor volume, and total lesion glycolysis. These measurements were then tested for associations with pathological response, disease-free survival, and overall survival.

What the researchers found was striking precisely because it was not where they expected it. The metabolic activity of the primary tumor itself, measured by SUVmax and SUVmean, did not reliably predict which patients would live longer. Nor did the metabolic activity of involved lymph nodes. Instead, it was the spleen, an organ that filters blood and houses large reservoirs of immune cells, that emerged as the most informative signal. Patients whose spleens showed higher mean standardized uptake values before treatment experienced significantly longer disease-free survival, with a hazard ratio of 0.19 and a 95 percent confidence interval of 0.037 to 0.94, and longer overall survival, with a hazard ratio of 0.12 and a 95 percent confidence interval of 0.016 to 0.83. Both associations reached statistical significance.

This finding ties into a growing body of evidence that the spleen plays an active role in the antitumor immune response, a concept the researchers refer to as the tumor-spleen immunity cycle. When immune checkpoint inhibitors unleash T cells to attack cancer cells, the spleen may act as a critical hub where activated immune cells are generated, expanded, and deployed. A metabolically active spleen on FDG PET may therefore reflect a heightened state of systemic immune activation, one that is already primed to respond when immunotherapy is administered. The researchers also found that higher S-SUVmean was linked to an increased combined positive score of PD-L1, a well-established biomarker that predicts response to immune checkpoint blockade, providing a biological bridge between the imaging finding and known immunology.

The study also examined whether pathological response, the traditional gold standard for judging neoadjuvant treatment efficacy, correlated with survival outcomes. Surprisingly, neither pathological complete response nor major pathological response was associated with disease-free survival or overall survival in this cohort of patients receiving immunotherapy plus chemoradiotherapy. This disconnect raises important questions about how treatment response should be assessed in the immunotherapy era, where the immune system may continue to exert antitumor effects long after visible tumor burden has been reduced, and where conventional pathological staging may not fully capture the depth of immune-mediated clearance.

Tumor size, however, did matter for one outcome. When researchers measured the length, width, and radius vector of the primary tumor on PET imaging, larger dimensions were significantly associated with shorter disease-free survival, with hazard ratios approaching 9 or higher and confidence intervals that excluded unity. Metabolic tumor volume, a three-dimensional measure of metabolically active tumor burden, also predicted shorter disease-free survival with a hazard ratio of 4.53. These parameters did not, however, reach significance for overall survival, suggesting that tumor bulk primarily influences the risk of recurrence rather than death from any cause.

When the researchers combined spleen metabolic activity with pathological response, the predictive power increased further. Patients who had either a higher S-SUVmean or achieved a major pathological response showed a dramatic survival advantage, with a hazard ratio of 0.09 for disease-free survival and 0.08 for overall survival, both statistically significant. This composite finding suggests that integrating imaging-based immune signatures with conventional pathological assessment could yield a more comprehensive picture of treatment efficacy than either measure alone.

The implications of this work extend beyond gastric cancer. Immune checkpoint inhibitors have transformed the treatment landscape for many solid tumors, yet only a subset of patients respond durably. Identifying reliable predictive biomarkers remains one of the most pressing challenges in oncology. While tumor PD-L1 expression, microsatellite instability, and tumor mutational burden are commonly used, they capture only part of the complex interplay between cancer and the immune system. A noninvasive imaging marker that reflects systemic immune activation, obtainable from a scan that patients already undergo for staging, would represent a practical and accessible addition to the biomarker toolkit.

The concept of the tumor-spleen immunity cycle is gaining traction among immunologists and oncologists alike. Animal studies have shown that splenectomy can impair the efficacy of immune checkpoint inhibitors in some models, and clinical observations have suggested that splenic metabolic changes on PET imaging correlate with response to immunotherapy in melanoma and lung cancer. This study is among the first to extend that framework to gastric cancer, a disease where immunotherapy is increasingly used in the perioperative setting but where predictive biomarkers remain poorly defined.

The authors caution that their analysis is based on a small number of patients from a single phase II trial and that the findings require validation in larger, independent cohorts. The confidence intervals, while statistically significant, are wide, reflecting the limited sample size. Nevertheless, the consistency of the spleen signal across both disease-free and overall survival, its biological plausibility, and its correlation with a known immunotherapy biomarker lend credibility to the result. If confirmed, S-SUVmean could become a simple, cost-effective tool for identifying patients most likely to benefit from neoadjuvant immunotherapy, helping clinicians personalize treatment decisions before a single dose is given and potentially sparing others from ineffective and toxic regimens.

Subject of Research: Predictive value of baseline FDG PET/CT spleen metabolic parameters for immunotherapy efficacy in locally advanced gastric adenocarcinoma

Article Title: Associations of baseline [18F]FDG PET/CT metabolic parameters with the efficacy of neoadjuvant immunotherapy plus concurrent chemoradiotherapy in patients with locally advanced gastric adenocarcinoma: a post hoc analysis of the Neo-PLANET phase II trial

Article References: Wang, J., Lv, J., Tang, Z., Wang, X., Wang, Q., Sun, Y., Shi, H., & Liu, T. (2026). Associations of baseline [18F]FDG PET/CT metabolic parameters with the efficacy of neoadjuvant immunotherapy plus concurrent chemoradiotherapy in patients with locally advanced gastric adenocarcinoma: a post hoc analysis of the Neo-PLANET phase II trial. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02751-z

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02751-z

Keywords: FDG PET/CT, gastric adenocarcinoma, immunotherapy, spleen, biomarker, neoadjuvant chemoradiotherapy, disease-free survival, overall survival, PD-L1, tumor-spleen immunity cycle, Neo-PLANET trial, metabolic tumor volume