The use of radiomics features can effectively predict recurrence-free survival (RFS) and benefits of postsurgical adjuvant therapy in patients with an intermediate- or high-risk gastrointestinal stromal tumor (GIST), according to the results of a recently published study in La Radiologia Medica, the official journal of the Italian Society of Medical Radiology.

Recognizing that GISTs are the most common mesenchymal tumor of the digestive tract, and that the stomach and small intestine are the most common sites, the researchers sought to explore the development and validation of a radiomics nomogram for the prediction of RFS and benefits of adjuvant treatment in patients with high- or intermediate-risk GISTs according to the use of computed tomography (CT) radiomics features.

The current retrospective analysis enrolled a total of 220 patients with intermediate- or high-risk stratified GISTs who underwent surgical resection at Fujian Medical University Union Hospital in China between July 2009 and September 2015. Study inclusion criteria were as follows: (1) complete patient data available; (2) routine contrast-enhanced abdominal CT performed within 30 days prior to surgery; (3) GISTs in which macroscopic complete resection was performed; and (4) postoperative pathologically confirmed GIST.

Continue Reading

Among the 220 participants, 114 were men and 106 were women. Overall, 102 patients underwent intermediate-risk stratification and 118 underwent high-risk stratification. The size of all tumors was more than 2.0 cm, with 55 of the tumors more than 10.0 cm.

Learn more about experimental therapies for gastrointestinal stromal tumors  

The study cohort of 220 participants was randomized in a 6:4 ratio to a training cohort of 131 individuals and a validation cohort of 89 individuals. In the first year following surgery, abdominal CT follow-up and laboratory tests were performed every 3 months. Follow-up occurred every 6 months in the initial 1 to 3 years after surgery, every 6 to 12 years in the 3 to 5 years after surgery, then annually thereafter.

RFS was defined as “the period from the first surgical treatment to the first documented tumor recurrence (imaging-confirmed).” Imatinib was administered routinely postsurgery, at a 400 mg daily dose. Radiomics analysis comprised CT image acquisition, tumor segmentation, feature selection, feature extraction, model construction, and assessment.   

Ultimately, 8 radiomic signatures were chosen. The area under the curve (AUC) of the radiomics signature model for predicting 3-year, 5-year, and 7-year RFS in the training and validation cohorts were as follows: training cohort AUC=0.80, 0.84, and 0.76, respectively; validation cohort AUC=0.78, 0.80, and 0.76, respectively.

The constructed radiomics nomogram was more accurate than the clinicopathologic nomogram in the prediction of RFS in GIST (95% CI, 0.817-0.911 vs 95% CI, 0.675-0.791, respectively).

The researchers concluded that “a nomogram constructed based on preoperative CT radiomics features could be used for RFS prediction in high/intermediate-risk GISTs and assist the clinical decision-making for GIST patients.”


Wang F-H, Zheng H-L, Li J-T, et al. Prediction of recurrence-free survival and adjuvant therapy benefit in patients with gastrointestinal stromal tumors based on radiomics features. Radiol Med. Published online September 4, 2022. doi:10.1007/s11547-022-01549-7