Radiomic CT accurate in predicting spread of common lung cancer, enhancing surgical decisions

A team of researchers out of China found the CT-based radiomic signature of primary tumors can be used to quantiatively and noninvasively predict the spread of a common form of lung cancer to lymph nodes (LN) in the chest.

The retrospective study was published in the American Journal of Roentgenology.

“The radiomics signature performed much better than clinical-histopathologic features did, showing its incremental value in preoperative prediction of the pathologic status lung cancer in patients who are candidates for surgery, thus aiding in accurate decision making,” wrote Tong-Fu Yu with the Department of Radiology at First Affiliated Hospital of Nanjing Medical University in China and colleagues.

A total of 492 patients with lung adenocarcinoma who underwent preoperative unenhanced chest CT were included—and 300 radiomics features quantifying tumor intensity, texture and wavelet were taken from scan data.

Results showed the radiomics signature more accurate in predicting mediastinal LN metastasis than the clinicohistopathologic model. Radiomics achieved 91.1 percent in ROC analysis, area under the curve of 0.97, 94.8 percent sensitivity and 92 percent specificity.

“The major finding of our study was that multifeature-based radiomics biomarkers aided in noninvasive preoperative prediction of mediastinal LN metastasis in patients with lung adenocarcinoma and successful stratification of patients into high-risk and low-risk groups with high accuracy,” Yu et al. wrote.