How to fine tune cancer prognosis

Other illnesses may give clues to cancer prognosis

Doctors can more accurately predict outcomes for cancer patients if they have information on other illnesses the patient may have, say scientists.

They believe data about patients' non-cancer ailments should be included in cancer statistics.

This would improve the validity of research into cancer and the accuracy of survival estimates, they say.

The Washington University team report their findings in the Journal of the American Medical Association.

At present, trial data used by cancer specialists tends to focus on tumour size and aggressiveness.

Dr Jay Piccirillo and colleagues say adding information on patients' non-cancer ailments, called comorbidities, improves the accuracy of trial data.

The bottom line is that comorbidity data provides critical additional information for determining a patient's prognosis Dr Jay Piccirillo, head of the study at Washington University

It also enables doctors to give better estimations of likely survival to their patients, they say.

Dr Piccirillo's team checked the medical records of more than 17,700 patients being treated for one form or another of cancer.

Patients with more severe comorbidities had poorer survival chances than those with less severe or no comorbidities.

Reduced survival

Generally, patients with severe comorbidities were 24% less likely to be alive after three years than patients with no comorbidities.

Their chance of being alive after five years was 33% less.

These patients were also 15% more likely to have their cancer return.

COMMON COMORBITIES IN CANCER PATIENTS Cardiovascular conditions Respiratory diseases Renal failure Diabetes

Dr Piccirillo said: "The bottom line is that comorbidity data provides critical additional information for determining a patient's prognosis."

He believes comorbidity data should be incorporated into computer-based programs used to estimate survival.

Information on cancer patients' other illnesses would also be helpful for research into cancer, he said.

For example, it may not be appropriate to include all cancer patients in a given drug trial because their comorbidities may influence the drug's effectiveness.

Similarly, including only patients with no illness other than their cancer might make the results of a study look over-positive.

Improved accuracy

Dr Piccirillo and colleagues have developed a training programme to help doctors collect comorbidity data. They are now turning this in to a web-based system.

He said, here in the UK, the NHS plans to incorporate this programme into its current cancer registry system.

Dr John Toy, medical director of Cancer Research UK, said: "Experience tells us that factors such as other medical conditions in patients are very important in estimating both probability of survival and response to treatment and this is why they are tightly controlled in trials.

"This work could lead towards greater accuracy in predicting outcome in cancer patients in the general population, not just those in trials, information that would be invaluable to cancer health professionals and patients," he said.