In current clinical practice, there are no valid pre-operatively established algorithms or biomarkers that can reliably predict the surgical and clinical outcome of patients with ovarian cancer. Patients may undergo high risk ultra-radical surgery, with a morbidity rate of 20%, and one fifth of these patients will relapse early. If an algorithm predictive of surgical outcome can be created, then patients’ cancer treatment could be individualised to best suit their predicted outcome, and those who will not benefit from ultra-radical procedures can avoid them. In this feasibility study, it was assessed to determine if the biopotential properties of paired macroscopically normal and cancerous omentum tissue measured directly during cytoreductive surgery for epithelial ovarian cancer could act as surrogate markers for ovarian tumour profiling and hence be predictive of surgical and clinical outcome.

A bespoke apparatus was designed and developed by the Department of Bioengineering at Imperial College to measure biopotential in tissues ex vivo. The setup was optimised through a series of experiments determine the optimal methodology to accurately measure the tissue biopotential. The biopotential of paired macroscopically normal and cancerous omentum tissue was measured ex vivo in 36 patients who underwent primary ovarian debulking for stage III or IV ovarian cancer. The biopotential readings were correlated with various clinical outcomes. For all 36 patients, the cancerous omentum produced a lower biopotential reading than its paired macroscopically normal tissue (p<0.0001). When stratified at the median biopotential, it was observed that patients with a lower biopotential for their cancerous omentum had a poorer progression-free survival (p=0.0179 for all histotypes and p=0.0143 for high grade serous).

The omentum samples were sectioned and immunostained, and the tumour cellularity of the samples was calculated. A negative correlation was observed between biopotential readings and higher stromal content, which is also associated with poorer PFS. There was also a negative correlation observed between the biopotential readings and matrix/cellular density when tested on 3D models.