Radiomics analysis of contrast-enhanced CT predicts lymphovascular invasion and disease outcome in gastric cancer: a preliminary study

Abstract

Background: To determine whether radiomics features based on contrast-enhanced CT (CECT) can preoperatively predict lymphovascular invasion (LVI) and clinical outcome in gastric cancer (GC) patients. Methods: In total, 160 surgically resected patients were retrospectively analyzed, and seven predictive models were constructed. Three radiomics predictive models were built from radiomics features based on arterial (A), venous (V) and combination of two phase (A + V) images. Then, three Radscores (A-Radscore, V-Radscore and A + V-Radscore) were obtained. Another four predictive models were constructed by the three Radscores and clinical risk factors through multivariate logistic regression. A nomogram was developed to predict LVI by incorporating A + V-Radscore and clinical risk factors. Kaplan-Meier curve and log-rank test were utilized to analyze the outcome of LVI. Results: Radiomics related to tumor size and intratumoral inhomogeneity were the top-ranked LVI predicting features. The related Radscores showed significant differences according to LVI status (P < 0.01). Univariate logistic analysis identified three clinical features (T stage, N stage and AJCC stage) and three Radscores as LVI predictive factors. The Clinical-Radscore (namely, A + V + C) model that used all these factors showed a higher performance (AUC = 0.856) than the clinical (namely, C, including T stage, N stage and AJCC stage) model (AUC = 0.810) and the A + V-Radscore model (AUC = 0.795) in the train cohort. For patients without LVI and with LVI, the median progression-free survival (PFS) was 11.5 and 8.0 months (P < 0.001),and the median OS was 20.2 and 17.0 months (P = 0.3), respectively. In the Clinical-Radscore-predicted LVI absent and LVI present groups, the median PFS was 11.0 and 8.0 months (P = 0.03), and the median OS was 20.0 and 18.0 months (P = 0.05), respectively. N stage, LVI status and Clinical-Radscore-predicted LVI status were associated with disease-specific recurrence or mortality. Conclusions: Radiomics features based on CECT may serve as potential markers to successfully predict LVI and PFS, but no evidence was found that these features were related to OS. Considering that it is a single central study, multi-center validation studies will be required in the future to verify its clinical feasibility.

Contrast-Enhanced CT-Based Radiomics Analysis in Predicting Lymphovascular Invasion in Esophageal Squamous Cell Carcinoma

“…In addition, two additional radiomics features were specifically extracted, namely the maximum 3D diameter and the Mesh Volume (Table 1), even though the two radiomics features were not independent predictors. The result showed that patients with LVI-positive had greater maximum 3D diameter and Mesh Volume than patients without LVI (p < 0.001), which was consistent with previous studies on the prediction of LVI in gastric and hepatocellular carcinoma (15,25). Since there was no reliable individual factor to predict LVI, a predictive model combining radiomics and clinical features would be viable.…”

“…We performed three sequential steps for feature selection. First, we evaluated the inter-observer and intra-observer agreement of radiomic features and selected features with ICC values greater than 0.75 (15,(33)(34)(35). Second, Wilcoxon rank sum test (36,37) was used to select features with P value less than 0.05.…”

“…Currently, LVI can be diagnosed only by postoperative histopathology, and preoperative prediction is extremely difficult (14,15). Compared with conventional CT, CECT can better distinguish normal tissues from tumors, and perform better in detecting tumors, showing tumor extent and staging (16,17).…”

“…Nie et al (28) developed a radiomics nomogram incorporating Rad-score, clinical and PET/ CT parameters to predict LVI in lung adenocarcinoma, which showed good predictive performance. Chen et al (15) found that radiomics features based on CECT could serve as potential markers for predicting LVI and PFS in gastric cancer. The model established by radiomics features combined with clinical features has high diagnostic efficiency.…”

Integrative nomogram of intratumoral, peritumoral, and lymph node radiomic features for prediction of lymph node metastasis in cT1N0M0 lung adenocarcinomas

“…After the ICC selected the repeatable features, spearman correlation analysis (SPM) combined with the least absolute shrinkage and selection operator (LASSO) method were utilized to select the most useful predictive features in the training cohort. The threshold of the Spearman correlation coefficient was 0.9 to reduce feature redundancy 35 , and the LASSO was used to further select the features with penalty parameter tuning that was conducted by tenfold cross-validation based on minimum criteria.…”