Professor Julian Peto, head of Epidemiology at the Institute of Cancer Research in London, was speaking at the opening of a major cancer conference in Lisbon.

While the total numbers of cases was increasing, he said, this was due largely to population increase, an ageing population and the elimination of "competing diseases".

He said: "Take away the cancer deaths caused by tobacco and the reality is that deaths from cancers linked to other causes are stable or falling."

Data from national statistics shows that the incidence of cancer has risen by around 20% in men and 30% in women since 1970.

Top priority for doctors, he said, was to continue persuading their patients to give up smoking.

"In 1970 Britain had the worst total death rates in the world from tobacco.

"Then half the adults stopped smoking and tobacco deaths in middle age halved from 80,000 a year to 35,000 a year in 2000."

Take away the cancer deaths caused by tobacco and the reality is that deaths from cancers linked to other causes are stable or falling

Professor Julian Peto, Institute of Cancer Research

There have been increases in some cancers over the past decade - for example, testicular cancer and Hodgkin's lymphoma, and some cancers, such as prostate cancer, are being increasingly detected.

The most important risk factor for cancer after smoking is obesity, responsible for approximately 7% of all cancer deaths among non-smokers in Europe.

Cervical success

However, some cancers have been effectively controlled by screening, prevention and treatment, said Professor Peto.

For example, the epidemic of human papillomavirus (HPV), a sexually transmitted infection, during the past few decades, would have caused many more deaths from cervical cancer without the intervention of the cervical screening programme.

Without screening in the UK, the number of deaths might have risen to 5,000 a year - instead, the figure is stable at 1,000 a year.

"In many European countries cervical cancer would also have increased dramatically without screening.

"Prevention of this epidemic is a major achievement that isn't widely recognised."

However, Professor Peto warned that an ethical clampdown by governments meant that vital data needed by researchers was becoming increasingly difficult to obtain.

He said: "Under the European Data Protection Act it may even be illegal to use historical personnel records to study the mortality of factory workers.

"We are already seeing doctors and hospitals withholding data for fear of legal action. This nonsense protects no-one and is now starting to damage the very fabric of research."