but the fall in mortality has diverted attention from a relentless rise in incidence, a London conference will hear today (Monday 6 September).

"The marked disparity between incidence and mortality trends crystallises the problem posed by childhood leukaemia from a public health standpoint: we have become steadily better at treating it - at least in the sense of preventing children dying from it - but we have made little or no progress in preventing it. Rational approaches to prevention are difficult to formulate when so little is known about the cause," Professor Michel Coleman will tell experts at the first international scientific conference on the causes of childhood leukaemia.

The conference - Childhood leukaemia: incidence, causal mechanisms and prevention - has been organised by CHILDREN with LEUKAEMIA, Britain's leading charity devoted to the conquest of the disease in children.

Professor Coleman, who is Professor of Epidemiology and Vital Statistics at the London School of Hygiene and Tropical Medicine, said that between 480 and 500 children under the age of 15 are diagnosed with leukaemia in Britain each year and about 100 die from it. It accounts for about a third of all cancers in children.

The number of new cases diagnosed annually has been rising for at least 40 years, particularly in children under the age of five, he said. Importantly, by analysing the reliable mortality trends available between 1911 and the 1950s, together with directly measured incidence and survival data available since the 1960s, it was apparent that the rise in incidence had actually been going on much longer than that.

"Suggestions that part of the increase has been due to more children surviving infancy to reach the peak age of leukaemia incidence (1-4 years) and to improved registration and diagnosis of the disease, may be partly true, but they cannot plausibly explain the overall pattern of data now at our disposal," said Professor Coleman.

It is the worrying upward trend, an inability to pin down the causes and emerging concerns that cancer treatments are leaving survivors vulnerable to future serious health risks that has prompted the charity to summon top international experts from Europe, America, Asia and Australia to discuss the problem and, it is hoped, set an agenda for future research.

Conference chairman, Professor Denis Henshaw, explained: "If the increased risk facing today's children is at least partly caused by modern lifestyle factors, as is suggested by the increasing incidence, then it may be possible to take some preventive measures. But, first we need to determine what these factors are."

The conference, which will include many presentations of new data, will cover some of the 'usual suspects', such as radiation, parental smoking, viruses and air pollution. But other concepts that have received little attention to date will also be highlighted with new research findings. These include, for example, diet in early childhood, light pollution, melatonin and circadian rhythms, placental transfer to the foetus of damaging agents, medicines in pregnancy and delayed transgenerational susceptibility through the germline.