Contrast-Enhanced Computed Tomography–Based Radiogenomics Analysis for Predicting Prognosis in Gastric Cancer

Abstract

ObjectiveThe aim of this study is to identify prognostic imaging biomarkers and create a radiogenomics nomogram to predict overall survival (OS) in gastric cancer (GC).MaterialRNA sequencing data from 407 patients with GC and contrast-enhanced computed tomography (CECT) imaging data from 46 patients obtained from The Cancer Genome Atlas (TCGA) and The Cancer Imaging Archive (TCIA) were utilized to identify radiogenomics biomarkers. A total of 392 patients with CECT images from the Nanfang Hospital database were obtained to create and validate a radiogenomics nomogram based on the biomarkers.MethodsThe prognostic imaging features that correlated with the prognostic gene modules (selected by weighted gene coexpression network analysis) were identified as imaging biomarkers. A nomogram that integrated the radiomics score and clinicopathological factors was created and validated in the Nanfang Hospital database. Nomogram discrimination, calibration, and clinical usefulness were evaluated.ResultsThree prognostic imaging biomarkers were identified and had a strong correlation with four prognostic gene modules (P < 0.05, FDR < 0.05). The radiogenomics nomogram (AUC = 0.838) resulted in better performance of the survival prediction than that of the TNM staging system (AUC = 0.765, P = 0.011; Delong et al.). In addition, the radiogenomics nomogram exhibited good discrimination, calibration, and clinical usefulness in both the training and validation cohorts.ConclusionsThe novel prognostic radiogenomics nomogram that was constructed achieved excellent correlation with prognosis in both the training and validation cohort of Nanfang Hospital patients with GC. It is anticipated that this work may assist in clinical preferential treatment decisions and promote the process of precision theranostics in the future.

Methodological quality of radiomic-based prognostic studies in gastric cancer: a cross-sectional study

“…For the training cohort, the C-index ranged from 0.654 ( 25 ) to 0.880 ( 37 ), and the AUC ranged from 0.722 ( 30 ) to 0.965 ( 26 ). For the validation cohort, the C-index ranged between 0.610 ( 25 ) and 0.810 ( 20 ), and the AUC ranged between 0.744 ( 32 ) and 0.878 ( 33 ).…”

“…For domain 1, all studies followed the well-documented imaging protocol criteria. Fourteen studies (64%) used multiple segmentation methods (by different physicians/algorithms/software) ( 19 , 21 – 26 , 28 , 29 , 31 , 32 , 34 , 38 , 39 ), and 10 studies (45%) used images obtained at different time points ( 19 , 20 , 24 , 27 – 29 , 35 , 36 , 38 , 40 ). None of the studies conducted phantom studies to assess the feature reliability of the different CT scanners.…”

“…For domain 2, all 22 studies conducted feature reduction or adjustment of multiple tests and validation. Fourteen studies (64%) only performed internal validation ( 19 – 21 , 24 , 26 – 32 , 34 , 35 , 37 ), and 1 of the studies used the training cohort to validate the model ( 30 ). Only 8 studies (36%) were validated using both internal and external datasets ( 22 , 23 , 25 , 33 , 36 , 38 – 40 ).…”

“…For domain 4, multivariate analysis of non-radiomics features was performed in almost all studies (n=21, 95%). Biological correlations were involved in 2 (9%) studies ( 26 , 32 ). The performance of the radiomics models was assessed by comparing the results with “gold standards” in 13 (59%) studies ( 19 – 25 , 27 , 31 , 32 , 35 , 36 , 39 ).…”

Radiogenomics study to predict the nuclear grade of renal clear cell carcinoma

“…Radiogenomics focus on discovering correlations between radiomic features and genomics [1] . Based on the principle that medical images are the outcome of complex interactions at the genetic and molecular levels [2] , radiogenomic studies have revealed associations between imaging features and genetic information of tumours in glioblastoma, gastric, ovarian and other cancers [3] , [4] , [5] . Nuclear grade of ccRCC is considered to affect prognosis and surgical benefits, and its preoperative evaluation can provide effective guidelines for clinical treatments [6] .…”