Radiomic mapping model for prediction of Ki-67 expression in adrenocortical carcinoma

Area of Science:

• Radiology
• Oncology
• Medical Imaging Analysis

Background:

• Adrenocortical carcinoma (ACC) is a rare endocrine malignancy.
• Ki-67 expression is a key prognostic marker in ACC, indicating tumor proliferation.
• Accurate preoperative prediction of Ki-67 expression is crucial for treatment planning.

Purpose of the Study:

• To evaluate the utility of contrast-enhanced computed tomography (CT)-derived radiomic features for preoperative prediction of Ki-67 expression in ACC.
• To identify specific radiomic features significantly associated with Ki-67 expression levels in ACC.

Main Methods:

• Retrospective analysis of 53 patients with histopathologically confirmed ACC.
• Extraction of radiomic features from preoperative contrast-enhanced abdominal CT scans.
• Statistical analysis, including multivariate linear regression and correlation analysis, to assess the relationship between radiomic features and Ki-67 index.

Main Results:

• Ten radiomic features demonstrated statistically significant differences between high and low Ki-67 expression subgroups.
• A predictive model based on radiomic signature showed a significant association with Ki-67 expression status (R²=0.67, p=0.002).
• Shape flatness, elongation, and grey-level long run emphasis were independently correlated with Ki-67 expression (p=0.002, 0.01, 0.04).
• Area under the curve values for shape flatness and elongation were 0.78 and 0.7, respectively, for identifying high Ki-67 expression.

Conclusions:

• Radiomic features from contrast-enhanced CT offer promising preoperative prediction of Ki-67 index in ACC.
• Morphological radiomic features, specifically shape flatness and elongation, are superior predictors of high Ki-67 expression in ACC.