© 1998 by Oxford University Press
Researchers Hunt For Elusive Environmental Causes of Leukemia
(This is the second of two articles on research into the causes of childhood leukemia.)In the midst of growing public concern about environmental exposures that may increase risks for childhood leukemia and other cancers, a number of far-reaching epidemiological studies are under way, and researchers say results in the next few years should provide important clues if not definitive answers.
Environmental causes of childhood cancer have long been suspected by many scientists, and the public seems even more convinced that these risks exist. But the causes have been difficult to pin down, partly because cancer in children is rare and partly because potential risk factors are so many and various.
Last September, the U.S. Environmental Protection Agency, spurred by concerns that childhood cancer rates may be increasing, held the first-ever national conference to explore possible links between childhood cancer and environmental causes and to recommend a federal strategy to guide prevention research in this area.
"The world that our children are born into now includes tens of thousands of new chemicals that simply were not around just a few decades ago - substances that are present in our air, in our water, in our homes, on our foods," EPA Administrator Carol M. Browner said at the conference.
"Children, because their bodies and minds are still developing, are more susceptible than adults to environmental threats. Proportionate to body weight, they eat more of certain types of food, drink more fluids, and breathe more air than adults do. The young ones crawl on the floor or the ground; the older ones spend a lot of time outdoors - and, thus, children are often more exposed to potentially harmful pollutants in the soil, around the house, or in the air."
Martha A. Linet, M.D., of the National Cancer Institute's Radiation Epidemiology Branch, agreed that additional work is needed to discover what causes cancer in children. But she said some statements about cancer incidence rates in children that were made at the EPA meeting and in subsequent news reports could mislead because very different types of childhood cancer were all combined together.
"Many epidemiologists believe that total childhood cancer is not an etiologically meaningful entity," she said. "Each of the 12 major internationally recognized categories of childhood cancer has its own very clear descriptive epidemiology; they differ by age, by gender, by race, and many other factors. There's not a totally consistent picture, and the only [childhood cancer] that appears to be continuing to go up is brain tumors, where there was a big increase in the mid-1980s, followed by a much slower increase since then."
(According to NCI's Surveillance, Epidemiology, and End Results Program, the annual incidence rate for brain tumors in children up to age 14 rose from 2.4 per 100,000 in 1973-1974 to 3.3 in 1993-1994. For acute lymphoblastic leukemia, the rate rose from 2.7 per 100,000 in 1973-1974 to a peak of 3.8 in 1989, followed by a steady decline to 2.9 in 1993-1994.)
Preventable Causes
Linet also noted that efforts to identify preventable causes of childhood cancer are not new, and that the first large-scale studies in the United States, Canada, the United Kingdom, Germany, and other countries, planned a decade or more ago, have now begun to report findings.
Linet led a collaborative NCI/Children's Cancer Group study that evaluated residential measurement of power-line electromagnetic field exposures ("EMF") and risk of ALL. That study found "little evidence" of a link between EMF and ALL, although the issue remains controversial (see sidebar).
Whatever the final verdict on the risks of EMF, most scientists agree that it could cause no more than a small fraction of all childhood leukemia cases - which leaves hundreds of other potential risks, most of which have been less thoroughly explored. These include inherited or random genetic mutations; exposures to infectious agents at unusual times (see News, Jan. 7, 1998); parental occupational exposures to radiation or chemicals; parental dietary or behavioral patterns during pregnancy or before conception; and parental, fetal, or childhood exposures to environmental toxins such as pesticides.
A study by E. George Knox, Ph.D., a retired professor at the University of Birmingham, England, received wide attention when it was published in the April 1997 Journal of Epidemiology and Community Health. Knox examined proximity of childhood cancer deaths to dozens of potential hazards including factories, power stations, railroads, and motorways. He found higher-than-expected numbers of cancer deaths among children living near sources of petroleum product emissions and kiln and furnace smoke and gases.
Knox's study has been widely criticized on methodological grounds, including its reliance on postal codes rather than exact addresses. But some scientists say Knox may nevertheless be onto something, particularly in light of his earlier research showing spatial clustering of cancer cases suggesting focal environmental hazards.
Both the CCG study and the U.K. Childhood Cancer Study, headed by Sir Richard Doll, M.D., of the University of Oxford - the largest childhood leukemia studies in their respective countries - are casting wide nets that the researchers hope will snare some environmental culprits. And in doing so, they hope also to identify subgroups of leukemia that might have different genetic and immunologic characteristics as well as different etiologic mechanisms.
Confining Risk Factors
"One of the overriding things our study was designed to determine is whether risk factors are confined to biologically defined subgroups," said Leslie L. Robison, Ph.D., of the Department of Pediatrics at the University of Minnesota, Minneapolis, and principal investigator of the CCG study, which includes more than 1,900 children with ALL. Subgroups may be defined by cell type (B versus T cells), by the presence of specific cell surface markers (or immune phenotypes), or by cytogenetic characteristics.
Subsets of the subjects in the CCG study have been used to examine in greater depth questions of particular interest. Besides the EMF study, these include an investigation of indoor residential radon radiation risks for ALL led by Jay Lubin, Ph.D., of NCI, and an analysis of pesticides, polyaromatic hydrocarbons, and heavy metals, led by Jonathan D. Buckley, M.D., Ph.D., at the University of Southern California, Los Angeles.
In Buckley's study, household dust samples were collected with special high-powered vacuum cleaners to be analyzed for dozens of commonly used pesticides and other known or suspected carcinogenic compounds. Studies by EPA have shown that pesticide levels indoors on carpets - where young children crawl and play - are higher than those typically found on the lawns and gardens where the chemicals were applied. The study will also estimate automobile traffic density in the children's neighborhoods.
Other components of the CCG study expected to see publication soon, Robison said, are investigations of prenatal diagnostic x-rays (one of the few risk factors considered already established), parental smoking (see News, March 5, 1997), breast feeding, and military service.
The U.K. investigators have collected information on all children diagnosed with cancer in England and Wales since 1992. They will test five hypotheses: that childhood cancer may be caused by radiation exposures in utero or after birth, by in utero or postnatal chemical exposures, by occupational chemical or radiation exposures to fathers' sperm, by postnatal EMF exposures, and as a rare abnormal response to infection.
Blood samples have been taken from all children and their families for determination of genetic and immunologic characteristics. Doll said data collection from parents was completed in late 1997; medical record data are still under way, and preliminary results should be available by the end of 1998.
At the University of California, Berkeley, Patricia A. Buffler, Ph.D., dean of the School of Public Health, and Martyn T. Smith, Ph.D., professor of toxicology and director of the Superfund Research Center there, are collecting information on all cases of childhood leukemia diagnosed in Northern and Central California, including bone marrow, blood, and cheek swab samples (for genetic tests).
The study's primary aim is to assess chemical exposures of the child and the parents to pesticides and other chemicals, dietary exposures, and genetics, and like their U.K. and CCG counterparts, the Berkeley researchers are interested in identifying subtypes of leukemia that may help elucidate causes.
"We are characterizing the cases on a molecular and cytogenetic basis: Do they contain things like a mutation in the ras gene, or translocations . . . or changes in chromosome number?" Smith explained. "And independently of leukemia type on a pathological basis, can we subgroup these cases by asking, are there similar things about those that have a loss of a particular chromosome, for example, and is there a common exposure or a common genetics associated with that? Because we think we're dealing with a group of diseases that might look similar clinically, but are actually quite different in their etiology. And we're trying to understand if the genetic changes occurred before birth, trying to find the window when exposure would have been critical."
The study is funded by the federal Superfund program although it does not focus on Superfund sites. Smith said studies focused on suspected cancer clusters near such sites have not been productive, and useful answers are more likely to emerge from large and broad-based studies such as the one Buffler and he are doing.
-- Tom Reynolds
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||