Gastrointestinal stromal tumors (GIST) constitue the largest category of primary non-epithelial neoplasms of the stomach and small bowel. They are characterized by a remarkable cellular variability and their malignant potential is sometimes difficult to predict. Very recent studies, using mitotic count and tumor size as the best determinants of biological behavior, divide GISTs into three groups: benign, borderline and malignant tumors. We report on a male patient who underwent a right hepatectomy for a large metastasis 11 years after the surgical treatment of an antral-pyloric gastric neoplasm, histologically defined as leiomyoblastoma and with clinical, morphological and immunohistochemical features of benignity (low mitotic count, tumor size < 5 cm, low cellular proliferation index). Histological and immunohistochemical analysis of the hepatic metastasis showed the cellular proliferation index (Ki-67) to be positive in 25% of neoplastic cells, as opposed to the primary gastric tumor in which Ki-67 was positive in only 5% of neoplastic cells. In conclusion, although modern immunohistochemical techniques are now available to obtain useful prognostic information, the malignant potential of GISTs is sometimes difficult to predict: neoplasms clinically and histologically defined as benign could metastasize a long time after oncologically correct surgical treatment. Therefore, benign GISTs also require consistent, long-term follow-up.
Development of a Malignancy Potential Binary Prediction Model Based on Deep Learning for the Mitotic Count of Local Primary Gastrointestinal Stromal Tumors
