A high rate of recurrence after curative therapy is a major challenge for the management of hepatocellular carcinoma (HCC). Currently, no effective adjuvant therapy is available to prevent HCC recurrence. We designed a personalized neoantigen-loaded dendritic cell vaccine and neoantigen-activated T-cell therapy, and used it as adjuvant therapy to treat 10 patients with HCC who had undergone curative resection or radiofrequency ablation in the first stage of a phase II trial (NCT03067493). The primary outcomes were safety and neoantigen-specific immune response. Disease-free survival (DFS) was also evaluated. The immunotherapy was successfully administered to all the patients without unexpected delay and demonstrated a reasonable safety profile with no grade ≥3 treatment-related side effects reported. Seventy percent of patients generated de novo circulating multiclonal neoantigen-specific T-cell responses. Induced neoantigen-specific immunity was maintained over time, and epitope spreading was observed. Patients who generated immune responses to treatment exhibited prolonged DFS compared with nonresponders (P = 0.012), with 71.4% experiencing no relapse for 2 years after curative treatment. High expression of an immune stimulatory signature, enhanced immune-cell infiltration (i.e., CD8+ T cells), and upregulated expression of T-cell inflammatory gene profiles were found in the primary tumors of the responders. In addition, neoantigen depletion (immunoediting) was present in the recurrent tumors compared with the primary tumors (7/9 vs. 1/17, P = 0.014), suggesting that immune evasion occurred under the pressure of immunotherapy. Our study indicates that neoantigen-based combination immunotherapy is feasible, safe, and has the potential to reduce HCC recurrence after curative treatment.

Background and Aim: This study aims to construct a strategy that uses assistance from artificial intelligence (AI) to assist radiologists in the identification of malignant versus benign focal liver lesions (FLLs) using contrast‐enhanced ultrasound (CEUS).Methods: A training set (patients = 363) and a testing set (patients = 211) were collected from our institute. On four‐phase CEUS images in the training set, a composite deep learning architecture was trained and tuned for differentiating malignant and benign FLLs. In the test dataset, AI performance was evaluated by comparison with radiologists with varied levels of experience. Based on the comparison, an AI assistance strategy was constructed, and its usefulness in reducing CEUS interobserver heterogeneity was further tested.Results: In the test set, to identify malignant versus benign FLLs, AI achieved an area under the curve of 0.934 (95% CI 0.890–0.978) with an accuracy of 91.0%. Comparing with radiologists reviewing videos along with complementary patient information, AI outperformed residents (82.9–84.4%, P = 0.038) and matched the performance of experts (87.2–88.2%, P = 0.438). Due to the higher positive predictive value (PPV) (AI: 95.6% vs residents: 88.6–89.7%, P = 0.056), an AI strategy was defined to improve the malignant diagnosis. With the assistance of AI, radiologists exhibited a sensitivity improvement of 97.0–99.4% (P < 0.05) and an accuracy of 91.0–92.9% (P = 0.008–0.189), which was comparable with that of the experts (P = 0.904).Conclusions: The CEUS‐based AI strategy improved the performance of residents and reduced CEUS's interobserver heterogeneity in the differentiation of benign and malignant FLLs.

The tumor microenvironment is distinctive in primary and secondary liver cancer. B cells represent an important component of immune infiltrates. Here, we demonstrated that B cells are an important regulator in hepatocellular carcinoma (HCC) and colorectal cancer liver metastasis (CRLM) microenvironments. B cells displayed distinct developmental trajectories in HCC and CRLM. Single-cell analysis revealed that IgG+ plasma cells preferentially accumulated in HCC, whereas IgA+ plasma cells were preferentially enriched in CRLM. Mechanistically, IgG+ plasma cells in HCC were recruited by tumor-associated macrophages via the CXCR3–CXCL10 axis, whereas IgA+ plasma cells in CRLM were recruited by metastatic tumor cells via CCR10–CCL28 signaling. Functionally, IgG+ plasma cells preferentially promoted protumorigenic macrophages formation in HCC, and IgA+ plasma cells preferentially induced granulocytic myeloid-derived suppressor cells activation in CRLM. Clinically, increased infiltration of IgG+ plasma cells and macrophages in HCC was correlated to worse survival, whereas increased intratumoral IgA+ plasma cells and neutrophils in CRLM indicated poor prognosis. Taken together, this study demonstrated plasma and myeloid cell-mediated immunosuppression in HCC and CRLM, suggesting that selectively modulating primary or secondary tumor-related immunosuppressive regulatory networks might reprogram the microenvironment and provide an immunotherapeutic strategy for treating liver cancer. Significance: The immunomodulatory patterns of tumor-infiltrating B cells are distinct in primary and secondary liver cancer, with plasma cells mediating important physiologic processes that drive cancer progression.

Background and Aims: Scirrhous HCC (SHCC) is one of the unique subtypes of HCC, characterized by abundant fibrous stroma in the tumor microenvironment. However, the molecular traits of SHCC remain unclear, which is essential to develop specialized therapeutic approaches for SHCC. Approach and Results: We presented an integrative analysis containing single-cell RNA-sequencing, whole-exome sequencing, and bulk RNA-sequencing in SHCC and usual HCC samples from 134 patients to delineate genomic features, transcriptomic profiles, and stromal immune microenvironment of SHCC. Multiplexed immunofluorescence staining, flow cytometry, and functional experiments were performed for validation. Here, we identified SHCC presented with less genomic heterogeneity while possessing a unique transcriptomic profile different from usual HCC. Insulin-like growth factor 2 was significantly upregulated in SHCC tumor cells compared to usual HCC, and could serve as a potential diagnostic biomarker for SHCC. Significant tumor stromal remodeling and hypoxia were observed in SHCC with enrichment of matrix cancer–associated fibroblasts and upregulation of hypoxic pathways. Insulin-like growth factor 2 was identified as a key mediator in shaping the hypoxic stromal microenvironment of SHCC. Under this microenvironment, SHCC exhibited an immunosuppressive niche correlated to enhanced VEGFA signaling activity, where CD4⁺ T cells and CD8⁺ T cells were dysfunctional. Furthermore, we found that another hypoxic-related molecule SPP1 from SHCC tumor cells suppressed the function of dendritic cells via the SPP1-CD44 axis, which also probably hindered the activation of T cells. Conclusion: We uncovered the genomic characteristics of SHCC, and revealed a hypoxia-driven tumor stroma remodeling and immunosuppressive microenvironment in SHCC.