NCI Disputes Assertion of a Kids’ Cancer Epidemic

BETHESDA, Md—The National Cancer Institute has strongly challenged the notion, initiated at an Environmental Protection Agency (EPA) conference and widely dispersed by a newspaper article, that the United States is in the midst of an epidemic of childhood cancers.

A special ongoing analysis has found “no evidence of a consistent increase in the incidence during 1973 to 1994 for specific types of childhood cancer,” epidemiologist Martha Linet, MD, told the National Cancer Advisory Board (NCAB). She chairs a 12-member working group of leading NCI epidemiologists named, after reports of the “epidemic” surfaced, to evaluate patterns and trends over the past quarter century in childhood cancers.

An EPA meeting in mid-September declared that cancer in children had increased steadily for the past 25 years and that the situation demanded a focused program to determine the causes of and ways to prevent these malignancies.

A front-page article in the Sept. 26, 1997, New York Times stated that “the rate of cancer among American children has been rising for decades” and that, according to NCI statistics, “over a few decades, that has meant striking double-digit increases.”

However, if indeed such a dramatic situation exists, the NCI has failed to find it. “There is a lot we don’t understand about these childhood cancers, but it is far from clear that there is an epidemic of childhood cancers,” NCI director Richard D. Klausner, MD, told the NCAB.

J. Michael Bishop, MD, of the University of California, San Francisco, NCAB’s interim chair, asked Dr. Klausner “how two federal agencies in the same city could reach such different conclusions” and whether there was any communications between the two agencies before EPA made its assertion.

“I must say it was surprising,” Dr. Klausner replied. He noted that while there was some NCI participation in the EPA conference, “there was no communication with my office from the leadership of the EPA.” He said he tried to contact EPA administrator Carol M. Browner after the conference “but was unsuccessful.”

Childhood cancers account for a tiny portion of new cancers diagnosed each year—an estimated 8,800 out of 1,382,000 in 1997. Such a small percentage (0.0064) adds to the complexity of accurately assessing the patterns and trends in malignancies among youngsters, Dr. Linet said.

The primary source for evaluating childhood cancers is data gathered and published by NCI’s Surveillance, Epidemiology, and End Results (SEER) program, whose data go back to 1973. The NCI team led by Dr. Linet is reanalyzing those data.

She noted that the international class-ification system includes 12 major categories of childhood cancers, divided in-to subtypes, which themselves contain subgroups. The leukemia category, for example, consists of two major cell types, lymphocytic and myeloid; the myeloid leukemias are divided into acute and chronic, and each of these categories is split further.

The latest SEER data show that the top two malignancies account for nearly half and the leading five categories account for more than three-quarters of all childhood cancers in the United States: Leukemias (31.4%), brain and other central nervous system cancers (17.6%), lymphomas (12.4%), sympathetic nervous system cancers (8.1%), and soft tissue sarcomas (7.1%).

Leukemias, Brain Tumors

“While variations have occurred from year to year, overall there were no significant increases or decreases in rates for any of the specific cancers other than leukemia and brain tumors,” Dr. Linet said. And these two leading causes of childhood cancers have shown significant differences in their trends.

“For all childhood leukemias combined, incidence rose from 1973 to approximately 1989, when the incidence peaked,” she said. “After 1989, the incidence of childhood leukemia declined at an even more rapid rate than the earlier rate of increase.” Lymphocytic leukemia rose 2% annually until 1989 and since has declined at a rate of 4% annually.

Brain and other CNS tumors, however, have risen throughout the period. “The most dramatic increase occurred during 1984 to 1986, with smaller increases after and to a lesser extent before,” she said.

Interpreting incidence trends has been complicated by changes over the past few decades, such as new classifications and diagnostic tools, she said. Analysis shows that “some of the increases in lymphocytic leukemia and nonlymphocytic or myeloid leukemias were due to improvements in diagnostic specificity that paralleled the need for subtype specification to guide treatment decisions,” Dr. Linet explained.

Malcolm A. Smith, MD, PhD, a member of the NCI working group, has led an analysis indicating that the increase in the incidence of childhood brain tumors since 1973 was due in large measure to an increase seen in 1984-1985 and occurred at the same time as the surge in the use of diagnostic magnetic resonance (MR) imaging. Said Dr. Linet: “It was difficult to visualize tumors in the brain stem prior to the widespread availability and use of MR scans, which occurred in the mid-1980s coinciding with the period of greatest increase in brain tumor incidence.”

She also noted the significant drop in childhood cancer death rates over the last 25 years. “Based on death certificates and population data, rates have declined dramatically for childhood leukemia and similarly for most childhood cancers, although to a lesser extent for brain and other central nervous system cancers,” she said. She attributed this to “the dramatic advances in treatment, which now result in cure or long-term remission for a substantial portion of children.”