Dataset Information

Intratumor graph neural network recovers hidden prognostic value of multi-biomarker spatial heterogeneity.

ABSTRACT: Biomarkers are indispensable for precision medicine. However, focused single-biomarker development using human tissue has been complicated by sample spatial heterogeneity. To address this challenge, we tested a representation of primary tumor that synergistically integrated multiple in situ biomarkers of extracellular matrix from multiple sampling regions into an intratumor graph neural network. Surprisingly, the differential prognostic value of this computational model over its conventional non-graph counterpart approximated that of combined routine prognostic biomarkers (tumor size, nodal status, histologic grade, molecular subtype, etc.) for 995 breast cancer patients under a retrospective study. This large prognostic value, originated from implicit but interpretable regional interactions among the graphically integrated in situ biomarkers, would otherwise be lost if they were separately developed into single conventional (spatially homogenized) biomarkers. Our study demonstrates an alternative route to cancer prognosis by taping the regional interactions among existing biomarkers rather than developing novel biomarkers.

SUBMITTER: Qiu L

PROVIDER: S-EPMC9307796 | biostudies-literature | 2022 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Intratumor graph neural network recovers hidden prognostic value of multi-biomarker spatial heterogeneity.

Qiu Lida L Kang Deyong D Wang Chuan C Guo Wenhui W Fu Fangmeng F Wu Qingxiang Q Xi Gangqin G He Jiajia J Zheng Liqin L Zhang Qingyuan Q Liao Xiaoxia X Li Lianhuang L Chen Jianxin J Tu Haohua H

Nature communications 20220722 1

Biomarkers are indispensable for precision medicine. However, focused single-biomarker development using human tissue has been complicated by sample spatial heterogeneity. To address this challenge, we tested a representation of primary tumor that synergistically integrated multiple in situ biomarkers of extracellular matrix from multiple sampling regions into an intratumor graph neural network. Surprisingly, the differential prognostic value of this computational model over its conventional non ...[more]

PMID: 35869055

Similar Datasets

Project description:PurposeIntratumor metabolic heterogeneity parameters on 18F-2-fluoro-2-deoxy-D-glucose (18F-FDG) positron emission tomography-computed tomography (PET-CT) have been proven to be predictors of the clinical prognosis of cancer patients. The study aimed to examine the correlation between 18F-FDG PET-CT-defined heterogeneity parameters and the prognostic significance in patients with colorectal cancer.MethodsThe study included 188 patients with colorectal cancer who received surgery and 18F-FDG PET/CT examinations. Preoperative 18F-FDG PET/CT conventional and metabolic heterogeneity parameters were collected, including maximum, peak, and mean standardized uptake value (SUVmax, SUVpeak, and SUVmean), metabolic tumor volume (MTV), total lesion glycolysis (TLG), heterogeneity index-1 (HI-1) and heterogeneity index-2 (HI-2), and clinicopathological information. Correlations between these parameters and patient survival outcomes were inferred.ResultsThe associations between 18F-FDG PET/CT parameters and clinical outcomes were analyzed. Tumor thrombus (P < 0.001), tumor stage (P=0.001), MTV (P=0.003), HI-1 (P=0.032), and HI-2 (P=0.001) differed between the two groups with and without recurrence. Multivariate analysis showed that, in the radical surgery group, HI-2 (HR = 1.10, 95% CI: 1.04-1.17, P=0.001), tumor stage (HR = 20.65, 95% CI: 4.81-88.62, P < 0.001), and regional lymph nodes status (HR = 0.16, 95% CI: 0.04-0.57, P=0.005) were independent variables significantly correlated with progression-free survival (PFS) and HI-2 (HR = 1.16, 95% CI: 1.07-1.26, P < 0.001) was an independent variable affecting overall survival (OS). In the palliative surgery group, HI-2 (HR = 1.03, 95% CI: 1.01-1.06, P=0.020) was an independent variable affecting PFS, and all the parameters were not statistically significant for OS.ConclusionHI-2, tumor stage, and regional lymph nodes status might predict the outcomes of colorectal cancer more effectively than other 18F-FDG PET/CT defined parameters.

Dataset Information

Intratumor graph neural network recovers hidden prognostic value of multi-biomarker spatial heterogeneity.

Publications

Intratumor graph neural network recovers hidden prognostic value of multi-biomarker spatial heterogeneity.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets