A GENETIC modelling algorithm that predicts patient response to three standard chemotherapy drugs used to treat ovarian cancer could extend patients’ lives by 21 months.

A retrospective study used data from an ovarian cancer registry to genetically profile over 3000 ovarian tumour samples from patients already on one of three common ovarian cancer drugs – paclitaxel, cyclophosphamide and topotecan – to discover differences between tumours that responded to treatment and tumours that didn’t.

The Canadian researchers said site-specific cancers have traditionally been considered to be homogenous, but increasingly, evidence is pointing to significant heterogeneity within the disease.

They also noted that “because it has been so difficult to predict which ovarian cancers will respond to each of the three available drugs, doctors have largely been forced to guess which will work best”, which can lead to treatment failure.
The model, called COXEN (CO-eXpresssion gENe analysis), allowed the researchers to identify the heterogeneity and genetic signatures of tumours that responded to each drug.

They were also able to show that patients who had, by chance, been given the drug that the COXEN model would have picked for them lived 21 months longer than patients who had been initiated on a different drug.

The researchers said their study shows that biomarker-based personalised chemotherapy selection could improve survival of patients with advanced ovarian cancer.

While the model would need to be validated with a prospective clinical trial before use in a clinical setting, the authors said that COXEN led to similar results in patients with bladder cancer in a previous study.

A prospective clinical trial of COXEN in bladder cancer is underway.

PLOS One 2014; 5 February