Diagnostic Accuracy of Liquid Biopsy in Endometrial Cancer

Simple Summary The number of endometrial cancer (EC) cases is constantly growing. However, the current diagnostic approach is still rather imprecise, leaving 1/3 of patients temporarily undiagnosed. Moreover, final diagnosis is made after the surgery. That mean the histology of tumor, which influences scope of resection, is uncertain during procedure. This results in over- and undertreatment of EC patients. Those diagnostic problems might be solved by liquid biopsy—a new, minimally invasive method to obtain tumor biomarkers. Therefore, this study aimed to evaluate the usefulness of information obtained from liquid biopsy components (tumor educated platelets and circulating tumor DNA) coupled with random forest algorithm and deep neural networks to diagnose EC patients and evaluate tumor histology preoperatively. Abstract Background: Liquid biopsy is a minimally invasive collection of a patient body fluid sample. In oncology, they offer several advantages compared to traditional tissue biopsies. However, the potential of this method in endometrial cancer (EC) remains poorly explored. We studied the utility of tumor educated platelets (TEPs) and circulating tumor DNA (ctDNA) for preoperative EC diagnosis, including histology determination. Methods: TEPs from 295 subjects (53 EC patients, 38 patients with benign gynecologic conditions, and 204 healthy women) were RNA-sequenced. DNA sequencing data were obtained for 519 primary tumor tissues and 16 plasma samples. Artificial intelligence was applied to sample classification. Results: Platelet-dedicated classifier yielded AUC of 97.5% in the test set when discriminating between healthy subjects and cancer patients. However, the discrimination between endometrial cancer and benign gynecologic conditions was more challenging, with AUC of 84.1%. ctDNA-dedicated classifier discriminated primary tumor tissue samples with AUC of 96% and ctDNA blood samples with AUC of 69.8%. Conclusions: Liquid biopsies show potential in EC diagnosis. Both TEPs and ctDNA profiles coupled with artificial intelligence constitute a source of useful information. Further work involving more cases is warranted.