CT-Based Radiomics Showing Generalization to Predict Tumor Regression Grade for Advanced Gastric Cancer Treated With Neoadjuvant Chemotherapy

Abstract

Objective: The aim of this study was to develop and validate a radiomics model to predict treatment response in patients with advanced gastric cancer (AGC) sensitive to neoadjuvant therapies and verify its generalization among different regimens, including neoadjuvant chemotherapy (NAC) and molecular targeted therapy.

Materials and methods: A total of 373 patients with AGC receiving neoadjuvant therapies were enrolled from five cohorts. Four cohorts of patients received different regimens of NAC, including three retrospective cohorts (training cohort and internal and external validation cohorts) and a prospective Dragon III cohort (NCT03636893). Another prospective SOXA (apatinib in combination with S-1 and oxaliplatin) cohort received neoadjuvant molecular targeted therapy (ChiCTR-OPC-16010061). All patients underwent computed tomography before treatment, and thereafter, tumor regression grade (TRG) was assessed. The primary tumor was delineated, and 2,452 radiomics features were extracted for each patient. Mutual information and random forest were used for dimensionality reduction and modeling. The performance of the radiomics model to predict TRG under different neoadjuvant therapies was evaluated.

Results: There were 28 radiomics features selected. The radiomics model showed generalization to predict TRG for AGC patients across different NAC regimens, with areas under the curve (AUCs) (95% interval confidence) of 0.82 (0.76~0.90), 0.77 (0.63~0.91), 0.78 (0.66~0.89), and 0.72 (0.66~0.89) in the four cohorts, with no statistical difference observed (all p > 0.05). However, the radiomics model showed poor predictive value on the SOXA cohort [AUC, 0.50 (0.27~0.73)], which was significantly worse than that in the training cohort (p = 0.010).

Conclusion: Radiomics is generalizable to predict TRG for AGC patients receiving NAC treatments, which is beneficial to transform appropriate treatment, especially for those insensitive to NAC.

Keywords: gastric cancer; generalization; neoadjuvant therapy; radiomics; tumor regression grade.

Figure 1

Flowchart and patient enrollment of this study. AGC, advanced gastric cancer. For the regimens, EOX (epirubicin, oxaliplatin, and capecitabine), SOXA (apatinib in combination with S-1 and oxaliplatin).

Figure 2

Importance ranking for 28 selected radiomics features using random forest The length of the bin and the depth of the color blue represent the important degree of the radiomics features. The feature name is ordered by the following rule: phase (p or d represents the venous-portal phase or the delayed phase) _ pre-processing_feature category_feature name. For example, for the first feature, i.e., d_logarithm_firstorder_Median indicates that the feature is named as Median from the first-order category, with transformation by logarithm.

Figure 3

The receiver operator characteristic (ROC) curves for models in each cohort. (A–E) Represents ROC curves of models for patients in the training cohort, the internal validation cohort, the Dragon III cohort, the external validation cohort, and the SOXA cohort. Rmodel, the radiomics model; Cmodel, the clinical model; Recmodel, the RECIST model; Combmodel, the combined model. The SOX cohort is defined the patients receiving the regimen of S-1 and oxaliplatin; the SOXA cohort is defined as the patients receiving the regimen of apatinib in combination with SOX (S-1 and oxaliplatin). AUC, area under the curve.

Figure 4

The predictive value of the individualized nomogram to predict tumor regression grade (TRG) for two patients receiving the EOX (epirubicin, oxaliplatin, and capecitabine) regimen. The two patients had similar clinical baseline information but different insensitivity to neoadjuvant chemotherapy treatment (both were men, 65 years old, cT4aN2M0, and similar tumor size). The individualized nomogram integrated radiomics score, age, and tumor size successfully predicted the outcomes of the patients, which mainly relied on the performance of radiomics score.