Leukemia in children exposed to multiple risk factors

Meta-analysis of standardized incidence and mortality rates of childhood leukaemia in proximity to nuclear facilities

The meta-analysis combined and statistically analysed studies of childhood leukaemia and nuclear facilities. Focus was on studies that calculated standardized rates for individual facilities. Due to variability between study designs, eight separate analyses were performed stratified by age and zone. One hundred and thirty-six sites were used in at least one analysis. Unadjusted, fixed effects and random effects models were used. Meta-rates greater than one were found in all models at all stratification levels often achieving statistical significance. Caution must be used when interpreting these results. The meta-analysis was able to show an increase in childhood leukaemia near nuclear facilities, but does not support a hypothesis to explain the excess. Each type of model utilized has limitations. Fixed effects models give greater weight to larger studies; however, population density may be a risk factor. Random effects models give greater weight to smaller studies that may be more likely to be affected by publication bias. A limitation of the overall study design is that standardized rates must be available for individual sites which led to exclusion of studies that only calculated rates for multiple sites and those that presented other statistical methods. Further, dose-response studies do not support excess rates found near nuclear facilities. However, it cannot be ignored that the majority of studies have found elevated rates, although not usually statistically significant.

Examination of gender effect in birth weight and miscarriage associations with childhood cancer (United Kingdom)

BACKGROUND

Higher birth weight and maternal history of miscarriage has been associated with an increased risk of childhood leukemia. The possibility that this association may be sex-specific has not been explored in detail in previous studies.

METHODS

In a retrospective case-control study, 732 childhood (< or =14 years) cancer cases from a population-based Registry in Northern England whose hospital birth records could be accessed and 3,723 controls matched for date and hospital of birth to the cases were compared. We examined birth weight for sex-specific associations with childhood cancer. Conditional logistic regression analysis was used for statistical evaluation of associations.

RESULTS

In acute lymphoblastic leukemia (ALL) (225 cases and 1,163 matched controls), birth weight and sex showed a strong interaction (P = 0.003). In boys with ALL, but not in girls, there was a nonlinear association with birth weight (P for trend = 0.008; OR = 3.05 for the highest quintile compared to the second lowest quintile, 95% CI = 1.40-6.64; P = 0.005). When birth weights were adjusted using UK standards for gestational age and sex, the risk associations were similar in statistical significance and magnitude. Maternal history of miscarriage showed an association with all cancers and individually with ALL. The miscarriage association with ALL was statistically significant in boys only (OR = 1.91, 95% CI = 1.07-3.42; P = 0.03). A multivariable model for ALL containing other examined maternal and reproductive variables confirmed the independence of the birth weight and miscarriage associations. There was no birth weight or miscarriage associations in other cancers.

CONCLUSIONS

This study confirmed the risk associations with birth weight and miscarriages in childhood ALL. Statistically significant association of size at birth suggested marked differences in etiology between girls and boys.

Maternal pregnancy loss, birth characteristics, and childhood leukemia (United States)

OBJECTIVE

The authors evaluated the relation between maternal pregnancy loss, birth characteristics, and childhood leukemia in the Northern California Childhood Leukemia Study.

METHODS

Incident cases of childhood leukemia (age 0-14 years) were rapidly ascertained, and controls were randomly selected from birth records and individually matched to cases. A total of 366 cases [313 acute lymphoblastic leukemia (ALL) and 53 acute myeloid leukemia (AML)] and 460 controls were included in this analysis. The biological mothers of all subjects provided detailed reproductive history and birth characteristics of the index children during a personal interview. Odds ratio (OR) and 95% confidence intervals (CI) were estimated using conditional logistic regression. Data on maternal pregnancy loss and birth characteristics were also available from the birth certificates of 96.3% of all subjects.

RESULTS

History of miscarriage was associated with a significantly increased risk of AML (OR = 2.94, 95% CI: 1.03, 8.34), but not ALL. Neither birth weight, birth order, or parental ages appeared to be an important predictor of the risk of ALL or AML. A comparison between data from two different sources (interview versus birth certificate) indicated good reproducibility and offered some evidence against recall bias.

CONCLUSION

Maternal history of miscarriage is associated with an increased risk of childhood AML.

Allergy and infectious disease histories and the risk of childhood acute lymphoblastic leukaemia

Infectious disease histories were evaluated in a population-based case-control study of childhood acute lymphoblastic leukaemia (ALL) as it has been hypothesised that delays in early infections are associated with an increased risk of disease. Allergy histories were also assessed as part of a broader evaluation of the role of immune factors in ALL. Cases (n = 255) were diagnosed between 1980 and 1991 at one of four referral centres in a 31-county area of New York State; controls (n = 760) were a random sample of live births from the same region, frequency matched to cases by sex, race and birth year. Data were collected by mailed questionnaire, completed by case and control parents in 1995. Allergy and infectious histories before the age at leukaemia diagnosis for cases and an equivalent age for controls were evaluated. The adjusted odds ratio and 95% confidence interval [CI] associated with a positive history of any allergy was 0.58 [95% CI 0.38, 0.88] compared with a negative allergy history. The occurrence of several common childhood illnesses before 25 months of age and ALL were assessed, with both weak positive and weak inverse associations observed. Overall, these analyses provide little support for the hypothesis that infection delay in early life is associated with an increased risk of ALL. Children with positive allergy histories reported significantly more infections than those with negative histories; however, effect modification of the infection-ALL associations by child allergy history was not observed. Nonetheless, these observations suggest the importance of assessing both allergy and infectious histories and their possible interactions when evaluating the association between these immune factors and childhood ALL.

Early child-care and preschool experiences and the risk of childhood acute lymphoblastic leukemia

An infectious etiology for childhood acute lymphoblastic leukemia (ALL) has been suggested, yet few studies have focused on the role of early child care. Day-care histories were examined in a case-control study of ALL in New York State. Cases (n = 255) were diagnosed at one of four referral centers between 1980 and 1991; controls (n = 760) were randomly selected from livebirths in the 31 counties served by the referral centers. Self-administered questionnaires were mailed to the parents of cases and controls in 1995. Day-care histories were censored at the age of diagnosis for cases and at an equivalent date for controls. The odds ratio for children who stayed at home compared with those who attended day care for >36 months was 1.32 (95% confidence interval (CI): 0.70, 2.52); the odds ratios for 1-18 and 19-36 months of day care were 1.74 (95% CI: 0.89, 3.42) and 1.32 (95% CI: 0.64, 2.71), respectively. Elimination of cases with T-cell ALL enhanced the risk. Starting care at an earlier age was not associated with a decreased risk of ALL. These findings do not support the hypothesis that infrequent contact with peers during early childhood could delay exposure to infectious diseases and increase the risk of ALL.

A cluster of childhood leukemia near a nuclear reactor in northern Germany

Between February 1990 and December 1995, professionals diagnosed six cases of childhood leukemia among residents of the small rural community of Elbmarsch in northern Germany. Five of these cases were diagnosed in only a 16-mo period between February 1990 and May 1991. All cases lived in close proximity (i.e., 500-4,500 m) to Germany's largest capacity nuclear boiling-water reactor. We calculated standardized incidence ratios and exact 95% confidence intervals for a 5-km-radius circular area around the plant. The standardized incidence ratio for the time period 1990-1995 was 460 (95% confidence interval: 210, 1,030). The analysis was restricted further to the years 1990 and 1991, and the standardized incidence ratio increased to 1,180 (95% confidence interval: 490, 2,830). Presently, this cluster of childhood leukemia cases cannot be explained in terms of established and putative risk factors--including radiation from medical sources--for childhood leukemia.

History of maternal fetal loss and increased risk of childhood acute leukemia at an early age. A report from the Childrens Cancer Group

BACKGROUND

Maternal reproductive history of fetal loss previously has been reported to be associated with an increased risk of leukemia in subsequent offspring. Data from a Childrens Cancer Group (CCG) case-control study were analyzed to test the hypothesis that this association was dependent on the number of previous fetal losses and age at leukemia diagnosis.

METHODS

A case-control study using a large Childrens Cancer Group database examined maternal history of fetal loss as a risk factor for childhood leukemia in subsequent offspring. One thousand seven hundred fifty-three patients with childhood acute leukemia were compared with 839 community control subjects s and 2081 nonleukemia cancer control subjects.

RESULTS

A modest increase in risk was found to be associated with a history of fetal loss. Stratification by age at diagnosis of leukemia showed that this association was significant only for those patients diagnosed before 4 years of age and most significant in those patients diagnosed before 2 years of age. When comparing community controls with patients acute lymphocytic leukemia diagnosed before 2 years of age, one previous fetal loss was associated with a five-fold increased risk (P < 0.001) whereas two or more fetal losses were associated with a relative risk of 24.8 (P < 0.001). Similarly, patients with acute myelocytic leukemia diagnosed before 2 years of age demonstrated 5-fold and 12-fold increased risks associated with the previous fetal loss and 2 or more previous fetal losses, respectively.

CONCLUSIONS

Childhood acute leukemia occurring at younger ages may be associated with an underlying genetic abnormality or chronic environmental exposure, which can be either lethal to the developing fetus or mutagenic and result in the development of acute leukemia.

Birth characteristics of childhood cancer cases, controls, and their siblings

Reproductive characteristics of childhood cancer cases, controls, and their siblings were examined using data from a case-control study in the Denver, Colorado metropolitan area. Childhood cancer patients (n = 356) diagnosed from 1976 to 1983 were identified, and 242 were interviewed. Controls were selected by random digit dialing, with 212 interviews being completed (60% of eligibles). Extremes of birth weight were more common only among brain cancer cases. Patients were more often born preterm, particularly those with brain tumors [odds ratio (OR) = 6.1; 95% confidence interval (CI), 1.6-23.4] and were more likely to have birth defects (OR = 2.1; 95% CI, 0.9-5.0). Twins were more common among case siblings (OR = 2.6; 95% CI, 0.8-8.2). Low birth weight and preterm delivery among siblings were related only to soft tissue sarcoma. Birth defects were more common among case siblings, particularly leukemia cases (OR = 3.2; 95% CI, 1.3-7.7). Previous reports of elevated birth weight among cases and increased risk of miscarriage in case mothers were not corroborated, but associations with preterm delivery, high birth order, and birth defects among cases and birth defects and twinning among cases siblings encourage additional evaluation.

Viruses, clusters and clustering of childhood leukaemia: a new perspective?

Clusters of childhood leukaemia have, during a lengthy and controversial history, focussed attention on two alternative putative aetiological agents: infections and localised environmental pollution. In the United Kingdom emphasis is currently placed on the latter because of reports of localised clusters in the vicinity of two nuclear reprocessing plants. Now the most recent studies of spatial clustering in the United Kingdom also support the hypothesis that a substantial population of cases of childhood acute lymphoblastic leukaemia (ALL) arise as a rare host response to certain patterns of exposure to common infectious agents--the aberrant response model. Relevant aspects of the epidemiology of ALL are reviewed from this perspective and the hypothesis shown to be capable of unifying reported associations with different types of risk factor. It is probable that specific agent(s) are involved though none have been identified and these may share many epidemiological characteristics of herpes viruses. The possible relevance of these results to associations with prenatal parental occupational exposures to dusts and ionising radiation is explored.

Major histocompatibility complex, t-complex, and leukemia

In experimental models, leukemia was the first disease shown to have an association with the major histocompatibility complex (MHC) genes. In humans, several allelic human-leukocyte antigen (HLA) associations also have been recognized. In addition to allelic associations, atypical HLA segregation patterns have been observed in leukemic families. These include a higher frequency of HLA-identical unaffected siblings, increased HLA homozygosity and increased maternal HLA-DR identity. These observations suggest preferential transmission of disease-associated haplotypes and a male transmission bias in leukemic families. The lack of disease-specific segregation, however, supports the idea that the HLA system is not directly relevant in leukemogenesis. Therefore, the existence of another genetic region linked to the MHC, causing segregation distortion, and containing recessive leukemia susceptibility genes may be postulated. The mouse t-complex would fit this model. This gene complex has recessive (semi-) lethal genes, is transmitted preferentially through fathers, and both the mouse t-complex and its rat homolog, growth and reproduction complex grc, confer susceptibility to carcinogenesis. This model could also explain the increased spontaneous abortion rate in mothers of leukemic patients, epidemiologic associations of leukemia with oral clefts and neuroectodermal tumors, and the transmission of a radiation-induced leukemia risk through fathers. Such segregation distortion might be the reason behind the maintenance of a gene(s) with a lethal effect in the population.

Maternal reproductive history and birth characteristics in childhood acute lymphoblastic leukemia

Using birth-registration data, a case-control study was done to investigate the possible associations of childhood acute lymphoblastic leukemia (ALL) with birth characteristics and maternal reproductive history. The data included cases born and diagnosed in Minnesota since 1969. Matched analyses were conducted using 337 cases and 1336 birth year-matched controls. There was a statistically significant increased odds of ALL for birth to older (greater than 35 years) mothers (odds ratio (OR) = 2.14, 95% confidence interval (CI) = 1.28, 3.58), older fathers (OR = 1.62, 95% CI = 1.14, 2.30), mothers with at least a high school education (OR = 1.61, 95% CI = 1.05, 2.48), and larger intervals (greater than 5 years) between the birth of the proband and the preceding sibling (OR = 1.86, 95% CI = 1.12, 3.09). The increased odds of ALL for birth by Caesarean section approached significance (OR = 1.42, P = 0.06). No overall association was found for: gender, race, paternal education, fetal-loss history, birth order, prenatal care history, pregnancy complications, inducement of labor, multiple birth, gestational age, or birth weight. Age at diagnosis was an important effect modifier of some analyses. For cases diagnosed before age 2 years, there was a 2.7-fold increased odds of ALL if the last pregnancy had resulted in a fetal loss (P = 0.03). For cases diagnosed before age 4 years, birth weight greater than 3800 g was associated with a significant 2.05-fold increased odds of ALL. These data strengthen the hypothesis that prenatal events may play a causative role in childhood ALL, particularly in those cases diagnosed at a younger age.

Epidemiology of the childhood acute leukemias

Epidemiologic studies of the childhood leukemias have provided information relevant to several aspects of the care and follow-up of these children. The observations made regarding in utero radiation and ALL risk have certainly curtailed the use of routine obstetric diagnostic radiographs; observations regarding the association between birth weight, fetal loss, and other gestational events provide added enthusiasm for further research into basic biologic events occurring during fetal development; and the genetic patterns of disease supply critical information for genetic counseling and follow-up of affected patients and families. Additionally, the continued epidemiologic surveillance of children with cancer serves to form the foundation from which we will assess any future changes in childhood cancer incidence or pattern. Although not discussed here, the epidemiology of late effects, including second malignancies, reproductive function, and neuropsychologic functioning will assume a more prominent role as more children survive ALL and move into adulthood. While analytic studies have yet to yield an association as strong as the lung cancer/cigarette association in adults, future research designed to isolate biologically homogeneous disease populations for study may lead us to new and important associations. The continued cooperation of large pediatric oncology groups and private physicians is crucial as these future investigations are undertaken.

Cancer and leukemia risks after low level radiation--controversy, facts and future

Quantification of delayed low dose radiation (LDR) effects is still controversial. The current concept of the shape of the dose-response curve, particularly at the very low levels, is derived primarily by extrapolation from high doses and is affected by economic, social and political implications of cancer yield. Evidence based on epidemiological studies of populations exposed to fallout, occupational, intrauterine or background LDR is limited, due to methodological drawbacks and the need for extremely large sample sizes. Nevertheless, recent data indicate that LDR-induced childhood leukemia and thyroid cancer may exceed the rates predicted on the basis of the linear quadratic curve. The high yield in utero and in early childhood could be associated with low cumulative load of background radiation, and a consequently more effective radiation increment. A long term follow up of children exposed to 90 mSv after scalp X-irradiation revealed a relative risk of 3.8 and an excess risk of about 1.08 per 1000 man-sievert per year for thyroid cancer. Application of these findings to the post-Chernobyl state of events suggests that an increment of up to 20% in thyroid cancer might occur in a population exposed to 5 mSv as an aftermath of a similar accident. Prediction of future risk estimates should therefore be made with alertness and an open mind.

Birth weight as a risk factor for childhood acute lymphoblastic leukemia

Increased birth weight previously has been reported to be associated with childhood acute leukemia although the etiologic importance of this finding remains unclear. To further assess birth weight and associated parameters as a risk factor for childhood leukemia, a case/control study was performed using children with acute lymphoblastic leukemia (ALL) born in the state of Minnesota and diagnosed since 1969. Data obtained from birth registrations of 219 cases were compared with two control groups matched on date and county of birth (group I) or year of birth (group II). No significant differences were observed in mean birth weights of cases and controls. Statistically significant associations with birth weights greater than 3800 g were identified in cases diagnosed within the first 4 years of life. No associations were found between birth weight and ALL for case children diagnosed after 4 years of age. Factors that might be associated with increased birth weight, including maternal age, birth order, length of gestation, and socioeconomic status as measured by paternal education, were not found to be associated with an increased risk for ALL. The significance of the finding of high birth weight as a risk factor for childhood ALL remains unknown but suggests that pregnancy-related events may be of importance in the etiology of ALL in young children.

Cancer following medical irradiation

Several generalizations about radiation carcinogenesis can be made: 1) a single exposure is sufficient to elevate cancer incidence many years later: 2) radiation-induced cancer cannot be distinguished from naturally occurring cancer, i.e., there is not unique radiogenic cancer; 3) all cancers appear to be increased after irradiation with the exception of chronic lymphocytic leukemia, and possibly Hodgkin's disease, cervical cancer, and a few others; 4) the breast, thyroid, and bone marrow appear especially radiosensitive; 5) leukemia is the most prominent radiogenic tumor and shows a wave-like pattern of excess incidence over time, and the excess begins within two to four years, peaks about six to eight years, and decreases to normal levels about 25 years later; 6) solid tumors have a minimum latent period of about ten years, and for several cancers, the temporal pattern of incidence appears to follow the natural incidence, i.e., the cancers do not occur before the ages normally associated with increased incidence, implying that age-dependent factors influence the expression of disease; 7) age at exposure is perhaps the most important host factor influencing subsequent cancer risk; 8) the percentage increase in cancer incidence per rad is not the same for all cancers, i.e., some cancer of high natural incidence, e.g., colon, have low "relative risks" and some cancers of low natural incidence, e.g., thyroid, have high "relative risks;" 9) dose-effect curves are often linear, but curvilinearity is also observed and is possibly associated with the need for "two ionizing events" for transformation to occur at low doses, the influence of cell sterilization at moderate doses, the likelihood of "wasted" dose at high doses, and/or the influence of factors that effect the expression of disease.

Transplacental carcinogens and mutagens: childhood cancer, malformations, and abortions as risk indicators

Childhood cancer, malformations, and spontaneous abortions in Finland were analyzed according to the parents' occupations. Children of women working in the food industry and farming and of men working in motor vehicle driving and farming appeared to have an elevated risk of cancer. Women in industrial and construction occupations had an increased risk of having malformed children and spontaneous abortions.

A dosage response curve for the one rad range: adult risks from diagnostic radiation

Most exposure to low-level ionizing radiation, both diagnostic x-rays and nuclear radiation, occur in the range between 100 milirads and 10 rads--the "one rad range". In the past, the estimates of hazards in this range have been obtained by linear extrapolation from data on persons who were exposed to much higher dosages, generally in the centirad range used in radiotherapy of non-malignant disease. This article presents the first dosage response curve for the one rad range ever to be developed directly from data on men exposed to ordinary diagnostic radiation. The findings are based on approximately 220 men with non-lymphatic leukemia and more than 270 random-sample controls from the Tri-State Survey. The new findings suggest that the estimates previously obtained by extrapolation from high dosage levels to low dose levels underestimate the actual hazards by an order of magnitude. The new dosage response curves indicate that linear extrapolation fails because it disregards the subgroups in the general population that are particularly vulnerable to x-ray. There are immediate implications concerning the use of medical x-rays in screening or for routine purposes. The past risk-benefit calculations are based on extrapolative estimates and require drastic revision. Uses of x-ray which were previously marginal are now clearly counterindicated.

Fertility and irradiation: a preconceptional investigation in teratology

The effects of preconceptional irradiation on female fertility were investigated in six groups of 35 nulliparous Swiss-Webster mice whole-body irradiated with single doses of 25, 50, 75, 100, 150, and 300 rads of 60Co. They were mated with nontreated males and killed on gestational day 18. The litters were recorded in the chronologic order of their occurrence within each group. The percentage of pregnancies decreased with the dose, while the prenatal loss increased. Up to 100 rads, the middle group of pregnancies was devoid of resorptions, fetal death, and congenital anomalies, thus giving rise to a "maximum viability enclave." Beyond this level those three adverse effects were more or less homogeneously distributed over the whole period for which pregnancies were recorded. Exencephaly, microcephaly, and eye defects were the most frequent malformations. Among other sequelae, hypertrophy of the higher pole occurred in the left kidney in the surviving infertile females killed 5 months after treatment.