The researchers studied patterns of gene expression that indicate high levels of abnormal chromosomes or chromosomal instability (CIN) in cancer.

They showed that a subset of these genes are needed by CIN cancer cells to survive - and that these cells have higher levels of these genes, making them more resistant to a drug called paclitaxel.

They then investigated whether CIN can help doctors identify which patients are most likely to respond to the chemotherapy drugs paclitaxel or carboplatin in a prospective clinical trial of ovarian cancer patients.

Patients with high levels of the CIN gene pattern were more resistant to paclitaxel.

Crucially, patients with high levels of CIN responded well to carboplatin - another commonly used ovarian cancer drug.

In contrast, tumours with low levels of CIN were resistant to carboplatin but responded to paclitaxel.

The researchers believe clinical tests for CIN - which can be identified by dye-stained cells under a microscope - may be available within five years, saving patients from ineffective chemotherapy and leading to personalised treatment.

Senior author Dr Julian Downward, based at Cancer Research UK's London Research Institute, said: "Our work suggests that resistance to paclitaxel is driven by distinct irregularities in the cancer cell that lead to abnormal cancer cell division. These changes may also render cancers more sensitive to carboplatin treatment."

Each year in the UK over 6,600 cases of ovarian cancer are diagnosed, and around 4,500 women die from the disease.

Senior co-author Dr James Brenton, based at Cancer Research UK's Cambridge Research Institute and an ovarian cancer clinician at Addenbrooke's Hospital said: "All ovarian cancer patients are treated with carboplatin but most who have more aggressive cancers are treated with paclitaxel as well. As we don't know which patients will benefit from having additional paclitaxel it is very important that we develop better tests to choose the right drugs for individual patients."

Lead author Dr Charles Swanton, based at Cancer Research UK's London Research Institute and at the Royal Marsden Hospital said: "Until now it has been difficult to predict which patients may benefit from individual chemotherapy drugs before treatment. We hope that simple tests based on these findings will save patients from unnecessary treatments with paclitaxel that can have severe side-effects with limited benefit."

Dr Lesley Walker, director of cancer information at Cancer Research UK, said: "This is an excellent example of how basic science can translate into discoveries that improve treatments given to patients. One of Cancer Research UK's goals is to create more targeted treatments with fewer side effects, and this is a crucially important step in achieving this - potentially personalising the treatment of thousands of people with cancer."

infoncerresearchuk/