SUMMARY OF INVENTION
Investigators have developed a novel software program to analyze preoperative CT images to predict clinical outcomes of hepatopancreatobiliary (HPB) cancer patients. Patients with HPB who are candidates for surgery routinely undergo preoperative CT imaging, to assess for metastatic disease and determine resectability based on disease burden. This novel software analyzes preoperative CT images to stratify hepatopancreatobiliary (HPB) cancer patients at risk for poor surgical outcomes as well as predict response to chemotherapy.
Texture analysis is an established and promising imaging modality, which involves automated measurement of pixel intensity variation. A smooth (homogenous) image has uniform pixel intensities, and an irregular (heterogeneous) image has varied pixel intensities. Within the context of computed tomography (CT) images, texture analysis can quantify regional variations in pixels which are unable to be quantified by visual inspection.
In oncology research, texture analysis has been shown to be a useful tool to supplement lesion diagnosis and characterization, as well as to classify liver tumors, and to predict survival of colorectal cancer patients. However, this technique has not previously been used preoperatively to predict clinical outcomes.
- By identifying liver metastases undetectable with conventional preoperative imaging, this technology can predict the risk of liver cancer recurrence, thereby identifying patients who could benefit from adjuvant chemotherapy at the time of surgical resection
- The software can also help predict overall survival after surgical resection in patients with certain types of hepatopancreatobiliary cancers
- Significantly, this technology can identify patients at risk of postoperative liver insufficiency, a major cause of morbidity and mortality in patients undergoing liver resection
Hepatopancreatobiliary (HPB) cancers include cancers of the pancreas, liver, gallbladder, and bile ducts. While these types of cancers are second among cancer-related deaths, prognosis remains poor. Particularly for pancreatic cancer, the 5-year overall survival rate is only 6%.
AREAS OF APPLICATION
Initial application in hepatopancreatobiliary cancers
STAGE OF DEVELOPMENT
Ready to use
U.S. National Application 15/577,101 published
Amber Simpson, PhD, Laboratory Head, Department of Surgery, Memorial Hospital Research Laboratories, Memorial Hospital, MSK