Interpreting Epidemiologic Research: Lessons From Studies of Childhood Cancer

In recent years, the public has shown concern about trends in incidence rates, the occurrence of clusters, and the role of certain environmental exposures in the cause of childhood cancers. A front-page news story in the New York Times1 stimulated a dramatic upswing of public anxiety about these issues. Hearings by the US Senate Environment and Public Works Committee on a cluster of 11 childhood acute lymphoblastic leukemia (ALL) cases (since increased to 13) among the 8200 residents of a town in Nevada over a 3-year period led to a featured article in USA Today2 describing legislation under consideration to enhance the federal government’s role in responding to apparent cancer outbreaks in US communities.Compared with 1.22 million cancers (excluding non-melanoma skin cancers) diagnosed annually among adults in the United States (corresponding to an average annual incidence rate for all cancers of 398 per 100 000 person-years),3 there are only ∼8700 diagnosed per year among children younger than 15 years and 12 400 among children and adolescents younger than 20 years (corresponding to average annual incidence rates of 13.4 per 100 000 and 14.9 per 100 000 person-years, respectively).4 Carcinomas predominate among adults, and the major pediatric tumors are nonepithelial. The most common pediatric neoplasms are the leukemias (representing 30.2% of all cancers diagnosed in children younger than 15 years), brain and central nervous system cancers (21.7%), and lymphomas (10.9%); these 3 categories (together constituting 63%) and the remaining 37% of pediatric malignancies are characterized by substantial histologic and biological diversity.5–7 Instead of the anatomic site-based categories used for adult malignancies, a more appropriate classification system developed for pediatric neoplasms8 was recently updated and designated as the International Classification of Childhood Cancer.9This article includes 3 components. The first section focuses on terminology and criteria to evaluate whether statistical associations between risk factors and childhood cancer are causal in nature. The second section suggests a general approach for investigating possible pediatric cancer clusters. The third section considers how distinctive patterns and trends can be translated into new etiologic leads and summarizes potential causal factors (Tables 1–4).The major objectives of most epidemiologic studies are to determine whether a specific exposure or factor (eg, ionizing radiation, or a medical condition) is likely to cause a given disease and to quantify the strength of the relationship. Two major study designs are used to evaluate whether an exposure is linked with a given disease: the cohort and the case-control study designs. In a cohort study, exposed (eg, an occupational group, or people with a common environmental or medically related exposure) and unexposed (often the general population but sometimes a similar occupational group without the exposure) populations are ascertained then followed up (prospectively or retrospectively) to compare risks of developing particular disease outcomes. In an ideal case-control study, cases are those who have developed a particular disease in a specified population during the study period, and control subjects are a random sample of those in the population who have not developed disease; in practice, the investigator’s efforts to select control subjects may be affected by logistic issues. The case-control design is essential for economy in studies of rare diseases but requires retrospective collection of exposure information. An example of an ideal case-control study is one nested within a cohort, in which all cases are ascertained, but a randomly selected sample of the cohort is used for controls.Epidemiologists typically evaluate the association between exposure and disease by estimating the ratio of rates of disease in people who had previous exposure to the agent with unexposed people. By convention, an association between exposure and disease is considered to be statistically significant if the probability is less than an estimate of association as strong or stronger than the one observed that would arise if, in fact, there were no association; if the probability is 5% or greater, then the association is considered too likely to be attributable to random variation to be considered solid. Many scientists are unhappy with this evaluation criterion, but no satisfactory alternative has been widely adopted.Results or conclusions from different studies of a specific exposure and disease or from different investigators examining the same data sometimes seem to be contradictory. Pediatricians are better equipped to make an informed decision if they are familiar with key concepts and principles of interpretation particularly pertinent to epidemiologic studies of childhood cancer as described in this report.Critical to interpreting epidemiologic studies are the source population and the methods of selecting study subjects. In case-control studies, cases (ie, people with the disease of interest) and control subjects (ie, people without that specific disease) should be identified from the same population; ideally, control subjects should be chosen randomly from a complete list of the entire population from which cases arose.10–12 Examples of populations for which complete lists are available include the provincial-wide health insurance listings in Canada13; population-based lists of patients assigned to a general practitioner in the United Kingdom14; and the hospitalization, cancer, or other national registries in the Nordic countries.15–17 Population-based health care registries are limited in the United States, because even the nationwide Medicaid or Medicare lists are restricted to population groups defined by income or age. The rarity of childhood cancers limits the utility of large health maintenance organizations or most insurance plans for epidemiologic studies of pediatric tumors in the United States. Epidemiologic studies of childhood cancer have been conducted within US clinical trials consortia, because a high proportion of all children younger than 15 years (but not older adolescents) in whom cancer is diagnosed are seen by pediatric oncologists affiliated with these consortia.18,19 However, epidemiologic studies of pediatric cancer have not always included a substantial number of children from ethnic minorities, because regions with larger proportions of minorities are not always included, the proportion of pediatric cancer cases whose families agree to participate is smaller for minority than for nonminority children, and the proportion of minorities among control subjects has been lower than the percentage among cases.18,20Registration of patients who are treated by pediatric oncologists within the consortia often occurs within days of diagnosis, but the choice of control subjects is not so straightforward. One possibility might be selection of controls with other cancers or diseases from the same institution as cases if the exposures of interest do not cause the cancers or diseases in control subjects; if the exposure being evaluated is statistically or causally associated with the cancers or other diseases of control children, then the estimated risks using this control group tend to be lower than the actual risks. Because the major causes of most childhood cancers are unknown and the few known causes (high doses of ionizing radiation and certain inherited genetic disorders) are associated with more than 1 type of cancer or other serious pediatric disease, selecting control subjects with cancer or other serious pediatric disease is probably not a good choice. An alternative is to select otherwise healthy control subjects from the general population.For many years, control subjects for most US epidemiologic studies of childhood cancer have been selected by a telemarketing technique called random-digit dialing. Randomized listings of telephone numbers with the same area code and exchange as the cases are generated and systematically evaluated to identify households that contain children of similar age, or or ethnic group as the pediatric cancer in the United States, telephone has been this was not appropriate for in which substantial numbers of households the random-digit in the United States has been less than had been in previous because numbers of are used to telephone and there are more telephone per more to or in and more and of by control subjects. trends have led to between cases and control subjects; concern about the potential for selection has led to consideration of alternative for selecting control risk factor is a specific agent statistically associated with a factors can be as as high as or of as to particular or The to which the of a between a risk factor and a disease whether the of the should be considered as or factors may be associated (ie, or associated (ie, with the of exposure to a specific risk factor in or incidence of the disease, then is more of risk factor should be considered environmental or that may be in the cause of childhood of can include the or can be by or of identified as risk factors for childhood cancers include radiation ionizing and (eg, (eg, (eg, or a as (eg, as as specific of and and and genetic (eg, type is always particularly in case-control In of exposure for cases and control subjects more to an By if exposures be and the on then risk may be too high if exposures are systematically by cases but not control subjects or by control subjects but not case-control studies of children and adults, by the exposures many years Because childhood cancers are a study would to exposure from of if not of children over years to identify numbers of pediatric cancers for statistical a study would be too to be for case-control studies, methods are to estimate exposures and to for investigators should to or biological diagnosis, may not exposure during the period (eg, or may not be possible to a for to estimate exposures that may by or age. are not then on as data from or of subjects. data may be to or because of children with cancer to exposures that are often or only by of healthy exposures (eg, other to of childhood cancer, then may be for the to the exposures in methods used in epidemiologic of childhood cancer have with but studies to collection of substantial exposure from with there are few of different or for exposures to the cause of childhood In most efforts have on an approach with potential for and between cases and control most of the epidemiologic exposures and childhood cancer risk used exposure with about number of or year of have included this type of but the exposure used in epidemiologic studies were A recent that the risk for the different exposure the of environmental or biological more of a specific is to of a about a exposure to many or particularly because exposure and over with and to new methods for exposures are under should methods for for of over and among data for and of exposure and for appropriate control within data collection should as used in clinical to data about the disease or cancer and exposure of to the of exposure that are for environmental exposures (eg, or in from or clinical (eg, or should with control In the of of an exposure can be evaluated by with a known of a given for by the children, as in adults, there seem to be of to is from and epidemiologic studies of causal for in and exposures and cancer occurrence in One example is the statistical association between exposure to particularly during the of and in childhood leukemia risk in more used to estimate risk in most epidemiologic studies is defined as the ratio of the incidence of disease in exposed to the incidence of disease in unexposed The of risk is not an to most people. A risk of among exposed unexposed a over disease rates in unexposed a causal in a common disease would be However, an in a rare disease may not be particularly One to the of a risk is to this to the of a rare disease occurs in unexposed the annual incidence rate of 3 per 100 000 people and that the risk for that disease among people with a specified exposure is (eg, a in annual incidence of the rare disease to per 100 000 is observed in exposed with unexposed people in that The rate of developing the disease among exposed people would be per 100 000 people per year in to the rate of 3 per 100 000 per year among unexposed people. type of may be for interpreting the risk The same risk would arise if the rates in unexposed and exposed people were and per 100 000 per In case-control studies, typically the for risks of rare diseases as childhood cancer, the ratio and the risk are and the between these can be a statistically significant association is is possible that the association may be attributable to study of the data collection or the of factors related to used to whether an association is a statistical association or a causal association with biological or public health include the of the risk risks between and or are with whether the risk increased with exposure studies, the appropriate between the exposure and the disease (ie, the exposure the disease, with a appropriate for between first exposure to a agent and of the first of a and the biological of the of these factors should be but for not that be large risks (eg, the or of cancer among acute leukemia among children with is less likely that or the entire can be to a from a or all other for the be in risk to be as many studies of that the statistical association in populations to the criteria are statistical associations pertinent The first a statistical association likely to be and the second a for which of is the large epidemiologic studies from different in risk risks from to with an estimate of risks that were than of leukemia in of who were during of the was conducted the of a (eg, to evaluate potential during of the of radiation exposure is from studies of cancer incidence and among the in and and studies of children and adults who high doses of ionizing radiation from environmental and have been associated with of childhood as the of the risk associated with lower doses of ionizing radiation, as that from during is to possibility that the of or may a statistical association between the and disease be considered in the and of a medical that of cancer in of exposed to during may have been a of or health than the ionizing radiation In the to for a of many to the health of the that cancer risks were increased even among children with no of health in health as the likely cause of the increased incidence of childhood of the increased childhood cancer risk among of during 3 led to a in in in using lower radiation of during and of and other with Epidemiologic studies a in childhood cancer risks between and in between and in the United and between and in the United of with increased risk of childhood leukemia to most of the criteria for have about the of that during has been linked with of pediatric tumors in to leukemia (ie, of that the association is restricted to case-control but not or (ie, of that there was an of risks among children exposed in to radiation from the in and (ie, of increased risks associated with doses or of and that there is no from cancer risks in with radiation exposures in (ie, of biological include that there was of increased and neoplasms of and risks among in than in those that there is a of of associations in cohort the and studies by limited statistical in and and that there is an of about in from childhood leukemia was from to because the were systematically only from second example a between an environmental exposure and childhood leukemia that not the criteria for of from or of with in risk of childhood data from large epidemiologic using more exposure in the United and the United not a causal (ie, for and the strength of the statistical associations observed not data from epidemiologic studies were or childhood leukemia risks not with or code (ie, no risks not with exposure estimated exposures In the a proportion of children with high exposures had risks of leukemia risk estimated as for children whose estimated exposures were and for those with exposures children whose estimated exposures were the strength of the association was The of studies not the biological of the to not to cancer occurrence in and radiation from has not been shown to cause to or other of of of biological of the in risk among US children was likely attributable to selection that among families that in with high or code those with a who developed leukemia were more likely to participate in the large US epidemiologic study than those with a the were more likely to participate only in the the of a study, a of or a evaluate the of the whether statistical associations are likely to be A similar should be of medical particularly is given to a from a using not included in the statistical that are on criteria for data should be given interpreting than that are from of should be and because can be on that would the most of studies can be with a technique called or The dramatic in of is concern among of data clinical trials has particularly to whether there is a and to quantify the for a a is seen in many but not all of data from epidemiologic studies to with a number can be the studies have similar methods and However, this is the because epidemiologic studies often in study of control subjects population methods used for exposure and other Because there are no to studies to or those studies that do not a of the of a risk estimate studies with and are because even a study of can have a large on the of a may be particularly to whether an exposure of public concern (eg, environmental of ionizing radiation, as a in is linked with a specific type of childhood cancer, particularly the association is and observed in different epidemiologic to be about to a of to a risk there of or statistical for systematically a of epidemiologic this is an area of statistical with a of methods under methods have been efforts should be with appropriate is between a specified exposure and a disease, is to the that the number of who the disease or as a of the A recent example is of pediatric has in by and the of possible by the the number of for in children increased between and and the number of and among children in a major increased from by has been that of the for of children and the and of evaluation of the pediatric in acute particularly in and a proportion of all in children, the to a radiation is substantial because of the risk per of radiation for children, with in pediatric of all but of the radiation from cancer risks per using and estimated that risk of from cancer was 1 in or increased in a a of the risk of from cancer was estimated to be increased in a estimated cancer risks were 1 of than for adults million of the and are annually to children younger than 15 years in the United States. a study were conducted to the causes of among all children younger than 15 years in the United States, estimated that of the 000 from cancer in this would be attributable to childhood radiation exposure from the The that the of pediatric the in cancer but the for to the radiation the same as with health are with of high of childhood cancers. childhood cancer are defined as of cases in or with similar and a statistically with rates for the and period or A childhood cancer are those as a of a specific statistical the childhood cancer incidence in a particular can be (ie, during a given period but with or (ie, with approach and for investigating possible childhood cancer include between and of pediatric cancers in the whether the cluster includes diagnosed cases only or a of and cases and of the and of the of is often used to the of cases that a potential a more may be of during the Because the causes the are known for most childhood the of the cases to may be may be in the period as investigators in and suggests an cause for childhood this are a number of incidence of childhood in of population (eg, developing new and regions with increased population or attributable to new in of or related to major or during has been to be related to childhood but have not been and specific have not been during and has been linked with increased and sometimes risks of childhood The possible role of during and childhood has been using in number and of and other of exposure to cancer have been linked with environmental ionizing or or other or or to known or in or environmental or (eg, in the or in are no for a but cluster are led by and health with from federal and The is for of and statistical the are on the for and Two for for a cluster include the for and and a recent by the about a potential the key first should include of the of the of cases pediatric oncologists and other cancer and other and collection of and for data should the investigators in and the of The investigators to for with of and to the the then include of of selection of an appropriate area for of and to include the during which all potential cases can be and should include that the and of data on cases likely be to the data available for cases in populations used to to to of childhood cancer incidence in the regions used to rates and of the of cases in the study to include the to the to a possible in estimating the population of statistical and that evaluation of a large number of exposures in statistical associations that by The number of cases that a cluster is but the rarity of childhood cancer suggests that numbers be the investigators determine that the cluster a significant then potential causes be should that epidemiologic methods are limited numbers of particularly no exposure can the occurrence of the childhood cancer of childhood cancer and a on childhood cancer and in the regions by the and have of trends in these for the period and have to in patterns by age, or ethnic group, and histologic within major cancer to compare incidence trends in childhood cancers among populations can be because of in population and of childhood cancer the rarity of childhood cancer, and and variation in and International childhood cancer incidence data have been systematically in by the International for on for the and 1 to of the major categories a limited number of are of childhood cancers can be of childhood leukemia include the to 3 years of for the common of the lower incidence and of a to 3 years of in with US the trends for common in with of a the in and the in the United and the incidence of acute leukemia with the only apparent in The of a among and of a among may a role for genetic factors in occurrence of common but the of an among in the followed by of a first in and in the United States unknown or environmental exposures in a In to ethnic or are apparent for nervous system cancers in tumors in and in may be linked with genetic factors or exposures that by or ethnic or ethnic in genetic of or other may be the incidence is for all of leukemias and the ratio is for similar for and disease and for acute leukemia The incidence among the of central nervous system with the ratio apparent for and tumors but there is between and for and other The major categories of and other tumors are characterized by among than among are unknown for the in incidence of and the among for cancer and and the of in of acute and other but etiologic leads to include exposures that by of and genetic of nervous system tumors is during incidence is evaluated to by during the first the rate is seen in the first and with age, a for these tumors of tumors is in the of with a during for than for particularly for this may a role for in the cause of this type of The for incidence of is during and the incidence for other of occurs during The for incidence of is but incidence of not to The variation in of patterns for other of and for may studies of pediatric cancers have evaluated a large number of risk is known about the cause of childhood particularly the of these cancers. and genetic factors seem to in no more than 5% to of different categories of childhood environmental exposures and factors to of the occurrence of childhood risk factors are known to cause specific of childhood and other exposures have been statistically linked with of childhood cancers (Tables of pediatric cancers have increased in children with genetic or to high doses of ionizing radiation are associated with increased risks of acute lymphoblastic and central nervous system and or limited data certain occupational and and other exposures with increased risk of of childhood but number of environmental or risk factors have been linked with childhood cancer in the from large and and epidemiologic studies of brain and other childhood cancers. the from these and other recent has new the causes of most childhood cancers studies in those that on childhood are not The of the entire of epidemiologic in the and of the of studies are to interpreting the Epidemiologic studies, of the into a of and all in by genetic The of is as new methods are developed for exposure and data the used by scientists or even the methods used in trials in data collection efforts in epidemiologic studies, particularly those with the of childhood can be on a by or the specific exposures under evaluation with the childhood cancer other serious who for childhood cancers a between as and of exposures without to disease with families and an of requires to and public but not of the key of the causes of childhood

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.