Background Tertiary lymphoid structures (TLS) within the tumor microenvironment have been associated with cancer prognosis and therapeutic response. However, the immunological pattern of a high peritumoral TLS (pTLS) density and its clinical potential in hepatocellular carcinoma (HCC) remain poor. This study aimed to elucidate biological differences related to pTLS density and develop a radiomic classifier for predicting pTLS density in HCC, offering new insights for clinical diagnosis and treatment. Methods Spatial transcriptomics (n=4) and RNA sequencing data (n=952) were used to identify critical regulators of pTLS density and evaluate their prognostic significance in HCC. Baseline MRI images from 660 patients with HCC who had undergone surgery treatment between October 2015 and January 2023 were retrospectively recruited for model development and validation. This included training (n=307) and temporal validation (n=76) cohorts from Xiangya Hospital, and external validation cohorts from three independent hospitals (n=277). Radiomic features were extracted from intratumoral and peritumoral regions of interest and analyzed using machine learning algorithms to develop a predictive classifier. The classifier’s performance was evaluated using the area under the curve (AUC), with prognostic and predictive value assessed across four independent cohorts and in a dual-center outcome cohort of 41 patients who received immunotherapy. Results Patients with HCC and a high pTLS density experienced prolonged median overall survival (p<0.05) and favorable immunotherapy response (p=0.03). Moreover, immune infiltration by mature B cells was observed in the high pTLS density region. Spatial pseudotime analysis and immunohistochemistry staining revealed that expansion of pTLS in HCC was associated with elevated CXCL9 and CXCL10 co-expression. We developed an optimal radiomic-based classifier with excellent discrimination for predicting pTLS density, achieving an AUC of 0.91 (95% CI 0.87, 0.94) in the external validation cohort. This classifier also exhibited promising stratification ability in terms of overall survival (p<0.01), relapse-free survival (p<0.05), and immunotherapy response (p<0.05). Conclusion We identified key regulators of pTLS density in patients with HCC and proposed a non-invasive radiomic classifier capable of assisting in stratification for prognosis and treatment.

Background: Intratumoral tertiary lymphoid structures (iTLS) in hepatocellular carcinoma (HCC) are associated with improved survival and may influence treatment decisions. However, their non-invasive detection remains challenging in HCC. We aim to develop a non-invasive model using baseline contrast-enhanced MRI to predict the iTLS status. Methods: A total of 660 patients with HCC who underwent surgery were retrospectively recruited from four centers between October 2015 and January 2023 and divided into training, internal test, and external validation sets. After features dimensionality and selection, corresponding features were used to construct transfer learning radiomic (TLR) models for diagnosing iTLS, and model interpretability was explored with pathway analysis in The Cancer Genome Atlas-Liver HCC. The performances of models were assessed using the area under the receiver operating characteristic curve (AUC). The log-rank test was used to evaluate the prognostic value of the TLR model. The combination therapy set of 101 patients with advanced HCC treated with first-line anti-programmed death 1 or ligand 1 plus antiangiogenic treatment between January 2021 and January 2024 was used to investigate the value of the TLR model for evaluating the treatment response. Results: The presence of iTLS was identified in 46.0% (n=308) patients. The TLR model demonstrated excellent performance in predicting the presence of iTLS in training (AUC=0.91, 95% CI: 0.87, 0.94), internal test (AUC=0.85, 95% CI: 0.77, 0.93) and external validation set (AUC=0.85, 95% CI: 0.81, 0.90). The TLR model-predicted iTLS group has favorable overall survival (HR=0.66; 95% CI: 0.48, 0.90; p=0.007) and relapse-free survival (HR=0.64; 95% CI: 0.48, 0.85; p=0.001) in the external validation set. The model-predicted iTLS status was associated with inflammatory response and specific tumor-associated signaling activation (all p<0.001). The proportion of treatment responders was significantly higher in the model-predicted group with iTLS than in the group without iTLS (36% vs 13.73%, p=0.009). Conclusion: The TLR model has indicated accurate prediction of iTLS status, which may assist in the risk stratification for patients with HCC in clinical practice.