Predicting Treatment Tolerance and Survival in Patients with Gastroesophageal Adenocarcinoma
In 2020, gastric cancer was the fifth most common cancer worldwide, and esophageal cancer was the eighth most common cancer worldwide (Global Cancer Observatory). Strategies for treatment of both cancers range from curative to palliative. There is a growing evidence linking patient body composition with treatment tolerance (Hay et al., 2019; Miller et al., 2014). Therefore, the goal of this research was to find potentially significant biomarkers in predicting treatment tolerance and survival. This was done by investigating two aims. The first aim was to investigate the potential association between semantic radiomic features, clinical features, and demographic features with treatment outcome. The second aim was to investigate the potential association between agnostic radiomic features, clinical features, and demographic features with treatment outcome
We retrospectively identified 142 patients with gastric and esophageal cancer treated with neoadjuvant chemotherapy with some patients receiving radiation. Study outcomes were measures of treatment tolerance and survival. An existing segmentation model based on the nnU-Net architecture was used to derive cross-sectional masks from subcutaneous fat, skeletal muscle, and visceral fat at the L3 vertebral body level. Morphological and texture radiomic features were then extracted from the segmented cross-sections, and relationships between imaging-derived features and study outcomes were assessed.
On univariate analysis, skeletal muscle area was associated with a decrease in therapy break, and a decrease in Emergency Department admissions (p = 0.0072 and 0.0314, respectively). Increases in visceral fat area (p = 0.0285) and the ratio of visceral to subcutaneous fat areas (p = 0.0363) were both associated with chemotherapy dose reductions. An increase in skeletal muscle index (p = 0.0044) was associated with a decrease in therapy breaks. A combination of body mass index and skeletal muscle index was significant in predicting survival. Also, obesity increased the survival risk factor. Combining radiomic morphological features with clinical data increases the performance of a regularized logistic regression model in predicting the likelihood of surgical resection after neoadjuvant treatment.
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.
Rights for Collection: Masters Theses
Works are deposited here by their authors, and represent their research and opinions, not that of Duke University. Some materials and descriptions may include offensive content. More info