Exploratory Study of Apparent Diffusion Coefficient Histogram Metrics in Assessing Pancreatic Malignancy.

Area of Science:

• Radiology
• Oncology
• Medical Imaging

Background:

• Pancreatic cancer diagnosis relies on imaging and biopsy.
• Apparent diffusion coefficient (ADC) metrics offer insights into tissue microstructure.
• Histogram analysis of ADC values can provide quantitative biomarkers.

Purpose of the Study:

• To assess the utility of whole-lesion 3D histogram apparent diffusion coefficient (ADC) metrics in differentiating pancreatic malignancies.
• To evaluate the performance of ADC histogram metrics in determining lesion histology, T-stage, N-stage, and grade.

Main Methods:

• Abdominal MRI with diffusion-weighted imaging was performed on 42 pancreatic malignancies.
• Two radiologists independently calculated whole-lesion histogram ADC metrics using 3D volumes of interest.
• Statistical analyses included Mann-Whitney tests and receiver operating characteristic (ROC) analyses.

Main Results:

• Whole-lesion ADC histogram metrics were significantly lower in pancreatic ductal adenocarcinomas (PDACs) compared to pancreatic neuroendocrine tumors (PanNETs).
• The mean of the 10th-25th percentile (mean10-25) demonstrated the highest area under the curve (AUC) for distinguishing PDAC from PanNET (AUC = 0.884-0.917).
• ADC histogram metrics, particularly mean0-10 and mean10-25, showed diagnostic performance for nodal status (N-stage).
• No significant association was found between ADC histogram metrics and tumor T-stage or grade.

Conclusions:

• Volumetric ADC histogram metrics show promise as non-invasive biomarkers for pancreatic malignancy.
• The mean10-25 metric outperformed standard mean ADC for assessing lesion histology and nodal status.
• Histogram analysis of ADC values is a valuable tool in the non-invasive assessment of pancreatic cancer.