Rates and Risk Factors
Rates and Risk Factors
Leukemias are cancers of the blood forming organs and are the most common childhood cancers.5 They are classified according to the type of cancer cells involved (e.g. lymphocytes or myelocytes), and according to whether they are acute or chronic. Acute leukemias have rapid onset and are characterized by accumulation in the blood of immature blood cells. Chronic leukemias progress more gradually and are characterized by more functional blood cells. Nearly three-quarters of childhood leukemias are classified as acute lymphocytic leukemia (ALL) and 15 percent as acute myeloid leukemia (AML). The classification developed by IACR and used in this report combines all lymphocytic leukemias, because in childhood the majority of lymphocytic leukemias are acute lymphocytic leukemias. Chronic myeloid leukemias are grouped separately.
As seen in Table 1a, the incidence rate of leukemia in NJ was 4.3 per 100,000 and in the US was 4.1 per 100,000. The incidence rate for lymphoid leukemia was 3.4 per 100,000 NJ children and 3.2 per 100,000 US children. Boys had higher rates than girls and white children experienced higher rates than black children.
Incidence and Mortality Trends
Figure 4 shows a decline in NJ leukemia incidence between 1982 and 1988 from 5.0 to 4.3 per 100,000. The rate has been essentially unchanged since 1988. Although NJ incidence rates were higher than the United States rates in the early 1980s, the gap between NJ leukemia incidence rates and US incidence rates has narrowed.
US and NJ mortality rates both declined throughout the period 1979-1994 and were similar in magnitude. In New Jersey, mortality rates for leukemia reflected changes in incidence rates; higher incidence of leukemia from 1981 to 1984 was reflected in a slight increase in mortality rates three years later while lower incidence of leukemia from 1985 to 1988 was reflected in decreasing mortality rates three years later.
For lymphocytic leukemia specifically, Figure 5 shows that the incidence declined from 4.0 to 3.0 per 100,000 between 1982 and 1986, and then increased until 1989 when the rate leveled off at about 3.7 per 100,000. The US incidence rate increased from 2.9 per 100,000 to a maximum of 3.5 per 100,000 in 1990 and then declined to 3.0 per 100,000 in 1993. The NJ and US mortality rates both show a gradual decline from 1.0 to 0.5 per 100,000 over the period.
Acute lymphocytic leukemia (ALL) incidence rates are diagnosed most frequently among children under age five, with a peak between ages three and five and a decline thereafter.6 Rates are higher among boys than girls, and the incidence of ALL among white children is nearly twice as high as among black children. Rates of acute myelocytic leukemia (AML) are also higher for white children than black children, but the incidence of AML in children is essentially equal among the sexes. AML is the predominant leukemia diagnosed in infants under four weeks of age.
During the last fifteen years, diagnostic methods for acute leukemias have greatly improved, thereby enabling more specific and effective treatment. Survival rates for the leukemias have improved substantially since the mid-1970's. The most recent five-year relative survival rate for children with ALL is 80 percent. For AML, the five-year survival rate is 43 percent.2, 5
Children with abnormal chromosomal conditions, such as Down Syndrome, are at increased risk of developing leukemia. Moreover, most acute leukemia cases have been found to have chromosomal abnormalities. The cause of chromosomal changes is largely unknown, but a few cases have been linked to ionizing radiation, benzene and other solvents, pesticides, and treatment with certain classes of chemotherapy drugs.1, 6
The risk of leukemia associated with exposure to ionizing radiation, such as alpha, beta, gamma rays and x rays, has been the focus of much research. Low levels of prenatal radiation have been associated with a slightly increased risk of leukemia in some studies; however, researchers are not in agreement as to whether this reflects a causal relationship.6, 7
The cancer causing effect of postnatal therapeutic radiation and radiation from an atomic bomb has been well established. Moderate to heavy doses of radiation in children increase the risk of leukemia and other cancers. The subsequent risk of leukemias was higher in children exposed at the youngest ages. Increased risk of both ALL and AML in children exposed to the atomic bomb blasts lasted for ten years after exposure, but declined sharply thereafter.7 Results of recent investigations in the United States suggest that there is a small increase in childhood leukemia among those exposed to radiation from fallout of nuclear weapons tests. Diagnostic radiation after birth has not been linked to increased risk of cancer in children.5-7
A number of studies have been conducted to investigate the incidence of childhood leukemia around nuclear plants, primarily in Britain. Researchers studied proximity, paternal occupational exposure to nuclear plants, and incidence in areas where nuclear facilities were planned but never built. They have concluded that leukemia in children around nuclear plants cannot be explained by radioactive discharges from the plant.7
Parental Occupational Exposures
Some parental occupational exposures have been suggested as risk factors for childhood leukemia. Paternal occupational exposures include solvents, plastics, petroleum products, and lead. Maternal exposures include paints and pigments, benzene, metal dust, and sawdust. However, exposure assessment is extremely difficult in parental occupational studies, especially in relation to timing of a pregnancy (preconceptual, pregnancy, or after birth), and few studies are large enough to detect separate effects of these periods.6, 7
Exposure to electromagnetic fields was first linked with childhood leukemia in a 1979 study in which children in Denver, Colorado living near electric wires with high capacity had two to three times greater risk of ALL than other children. While studies in Los Angeles County and Sweden corroborated these findings, other study results in Denmark and Finland did not. The most recent and comprehensive study of electromagnetic fields was conducted in the United States and failed to find any significant associations. This study did not have some of the methodological flaws of the previous ones. An National Institutes of Health and Department of Energy panel concluded that the epidemiological evidence that exposure to electromagnetic fields causes cancer in children is limited.8 While scientists continue to investigate the issue, the evidence to date indicates that exposure to electromagnetic fields is not likely to greatly increase the risk of leukemia among children.5, 9-11
Chemical Contamination of Community Drinking Water
In recent years there has been increasing concern about chemically contaminated drinking water and the risk of leukemia. In 1993, scientists of the Consumer and Environmental Health Services of this Department completed a study of certain cancers and contaminants in public drinking water for the years 1979-1987, before current drinking water regulations had been fully implemented. Results suggested associations between exposure to volatile drinking water contaminants such as trichloroethylene and leukemias and lymphomas in adults and children.12 Levels of trichloroethylene and many other volatile organic compounds in public drinking water have been greatly reduced in New Jersey since 1985.13
Studies in Woburn, Massachusetts have examined the relationship between contaminated drinking water and childhood leukemia.14-16 Community water supplies in Woburn contained elevated levels of the industrial chemicals trichloroethylene and tetrachloroethylene (also known as perchloroethylene). The follow-up study of Woburn concluded that the incidence of childhood leukemia was associated with the mothers' potential for exposure to contaminated water from specific wells, particularly for exposure during pregnancy. These findings should be interpreted with caution however, since small numbers of study subjects lead to imprecise estimates of risk and, as a result, the exact magnitude of the association between exposure and risk could not be stated.