Differential implications of tumor endothelial cell and lymphocyte densities in advanced hepatocellular carcinoma patients treated with immunotherapy - AI-based tumor microenvironment profiling and predictive outcomes

Sep 30, 2026

Differential associations between tumor microenvironment features and immunotherapy efficacy were evaluated in unresectable hepatocellular carcinoma patients. Pretreatment H&E images from 163 cases were retrospectively analyzed with an AI/deep-learning model to quantify spatial immune and non-immune cell distributions. High tumor endothelial cell density predicted longer progression-free survival with atezolizumab plus bevacizumab, while an inflamed immune phenotype with abundant intratumoral TILs predicted improved outcomes after anti-PD-1 monotherapy. Results suggest clinically actionable biomarkers for treatment selection in advanced HCC.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Differential associations with immunotherapy efficacy

Patient cohort and AI-based spatial profiling

Tumor endothelial cell density and atezo-bev outcomes

Inflamed immune phenotype and anti-PD-1 outcomes

Implications for biomarker-driven treatment selection

npj | precision oncology Article | Published in partnership with The Hormel Institute, University of Minnesota | | | | --- | --- | --- | | [https://doi.org/10.1038/s41698-025-01207-x](https://doi.org/10.1038/s41698-025-01207-x) | | | | Differential implications of tumor endothelial cell and lymphocyte densities in advanced hepatocellular carcinoma patients treated with immunotherapy
Check for updates | | | | Gwangil Kim1,8, Beodeul Kang2,8, Jung Yong Hong3,8, Haeyoun Kang1, Jung Sun Kim2, Sohyun Hwang1,4, Sung Hwan Lee5, Sang Hoon Jung6, ChansikAn6, Won Suk Lee2, Chiyoon Oum7, Gahee Park7,
Mingu Kang7, Yoojoo Lim7, Jin Woo Oh7, Siraj M. Ali7, Chan-Young Ock7, Chan Kim2 , Ho Yeong Lim3 & Hong Jae Chon2 | | | | We investigated whether artiﬁcial intelligence (AI)-based tumor microenvironment proﬁling correlates with treatment efﬁcacy in unresectable hepatocellular carcinoma (HCC) patients treated with immune checkpoint inhibitor (ICI) therapies. Spatial distribution of immune/non-immune cells from pretreatment H&E images of 163 patients was retrospectively analyzed using an AI/deep-learning model. High tumor endothelial cell (TEC) density was associated with signiﬁcantly longer progressionfree survival (PFS) in the atezolizumab plus bevacizumab (atezo-bev) cohort (HR 0.51 [0.27-0.97]; p = 0.037) but not in the anti-PD-1 monotherapy cohort (HR 1.02 [0.59-1.77]; p = 0.935) . Conversely, inﬂamed immune phenotype, characterized by high intratumoral TIL densities, predicted longer PFS after anti-PD-1 monotherapy (HR 0.50 [0.25-0.99]; p = 0.042) but not after atezo-bev (HR 0.92 [0.50- 1.69]; p = 0.762) . Our exploratory analysis using AI/deep-learning model demonstrated high TEC density predicted superior outcomes with atezo-bev, while TIL presence correlated with improved anti-PD-1 monotherapy efﬁcacy in HCC patients, suggesting potential clinical applicability in treatment selection. | | | | Until recently, systemic therapy for hepatocellular carcinoma (HCC) was limited to sorafenib, with subsequent approval of additional tyrosine kinase inhibitors—lenvatinib, regorafenib and cabozantinib—providing only modest improvement in clinical outcomes1–4. However, recent advancements have signiﬁcantly expanded treatment options, speciﬁcally immune checkpoint inhibitors(ICIs)in combination regimens, such asatezolizumab plus bevacizumab (atezo-bev)5–8. These developments have marked a turning point in the management of advanced HCC, particularly with the phase 3 IMbrave 150 trial demonstrating the superior efﬁcacy ofatezo-bev over sorafenib5. The trial reported a progression-free survival (PFS) of 6.8 months, with a hazard ratio (HR) for death of 0.58 (95% CI 0.42–0.79) for | atezo-bev compared to sorafenib, establishing a new standard of care for untreated, unresectable HCC.
Despite these advancements, substantial variation in patient response to atezo-bev persists, underscoring the need for predictive biomarkers to both guide therapy selection and delineate patients who are unlikely to derive much beneﬁt from atezo-bev. Unlike other cancers, where biomarkers such as PD-L1 or tumor mutation burden are routinely used to predict ICI efﬁcacy, these markers have limited utility in HCC due to its unique tumor biology. For example, HCC is characterized by relatively low tumor mutational burden, with high mutation burden (>10 mutations/megabase) being rare. Expression of | | 1Department of Pathology, CHA Bundang Medical Center, Seongnam, Korea. 2Medical Oncology, Department of Internal Medicine, CHA Bundang Medical Center, CHA University School of Medicine, Seongnam, Korea. 3Division of Hematology-Oncology, Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea. 4CHA Future Medicine Research Institute, CHA Bundang Medical Center, Seongnam, Korea. 5Department of Surgery, CHA Bundang Medical Center, CHA University, Seongnam, Korea. 6Department of Radiology, CHA Bundang Medical

FAQs

How was tumor microenvironment profiling performed in this study?

Pretreatment H&E images from 163 unresectable HCC patients were retrospectively analyzed using an AI/deep-learning model to quantify spatial distributions of immune and non-immune cells.

What does high tumor endothelial cell (TEC) density indicate for atezolizumab plus bevacizumab?

High TEC density was associated with significantly longer progression-free survival in the atezolizumab plus bevacizumab cohort.

Which biomarker was linked to improved outcomes after anti-PD-1 monotherapy?

An inflamed immune phenotype characterized by high intratumoral TIL densities predicted longer progression-free survival after anti-PD-1 monotherapy.

Sep 30, 2026

Uploaded by Sophia Brooks