The AuChildreChild HcoordinHealth)HospitaJ. ScotMurray
www.a
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
Cancer
Epidemiology,
iomarkersrevention
arch Article
osure to Diagnostic Radiological Procedures and the Risk
B& P
hildhood Acute Lymphoblastic Leukemia
D. Bailey1, Bruce K. Armstrong2, Nicholas H. de Klerk1, Lin Fritschi3, John Attia4,5,
Lockwood6, and Elizabeth Milne1; for the Aus-ALL Consortium
ractBac
lymphor chilmothe
Metic regrelatio
Resindex(OR) fand 1.concepabdom
Con
the agenty abou
s' Affiliationld Health Rn Australia,Health, Nelia, Westernrn Australiostatisticstment of Mrch Institun's Hospita
upplementaiomarkers
stralian Stun Consortiuealth Reseaating cente, Elizabeth Ml), Michellet (UniversityD. Norris
acrjourna
Downloa
kground: Diagnostic irradiation of the mother during pregnancy increases the risk of childhood acuteoblastic leukemia (ALL). There is inconsistent evidence on associations between ALL and other parentaldhood diagnostic irradiation. The aim of this analysis is to investigate whether diagnostic X-rays of ther before birth, of the father before conception, or of the child increased the risk of childhood ALL.thods: Data from 389 cases and 876 frequency-matched controls were analyzed using unconditional logis-ression, adjusting for studymatching factors andpotential confounders. Ameta-analysis of our findings inn to paternal X-rays before conceptionwith the published findings of previous studieswas also conducted.ults: There was no evidence of an increased risk with maternal abdominal X-rays before the birth of thechild or with the child having any X-rays more than 6 months before the censoring date. The odds ratioor any paternal abdominal X-ray before conception was 1.17 [95% confidence interval (95% CI), 0.88-1.55],47 (95% CI, 0.98-2.21) for more than one X-ray. The OR for any paternal intravenous pyelogram beforetion was 3.56 (95% CI, 1.59-7.98). The pooled OR for this study with previous studies of any paternalinal X-rays before conception was 1.17 (95% CI, 0.92-1.48).clusions: There was some evidence of an increased risk of ALL in the offspring if the father had morene abdominal X-ray before conception or had ever had an intravenous pyelogram.
than o
Impact: We plan to repeat this analysis by using pooled data to improve precision. Cancer Epidemiol
te lymphoblastic leukemia (ALL) is the most com-ype of childhood cancer in developed countries. Inrn countries, the age-standardized incidence ratesproximately 30 to 40 per million (1). It is moreon in boys, and the majority of cases are diagnosed
of 5 years (2). Little is known with anyt the causes of ALL, although it is likely
exposIn 19
s: 1University of Western Australia, Telethon Instituteesearch, Centre for Child Health Research, Perth,Australia; 2University of Sydney, Sydney School ofw South Wales, Australia; 3University of WesternAustralian Institute for Medical Research, Perth,
a, Australia; 4Centre for Clinical Epidemiology, University of Newcastle, Newcastle, NSW andedicine, John Hunter Hospital and Hunter Medicalte, New Lambton, NSW, Australia; and 6Royall, Brisbane, Queensland, Australia
ry data for this article are available at Cancer Epidemi-& Prevention Online (http://cebp.aacrjournals.org/).
dy of Causes of Acute Lymphoblastic Leukaemia inm conducted the study, and the Telethon Institute forrch (TICHR), University of Western Australia, was ther. Bruce K. Armstrong (Sydney School of Publicilne (TICHR), Frank M. van Bockxmeer (Royal PerthHaber (Children's Cancer Institute Australia), Rodneyof Newcastle), John Attia (University of Newcastle),
(Children's Cancer Institute Australia), Carol Bower
Research. on November 30, 202cebp.aacrjournals.org ded from
oth environmental and genetic factors play a roleecause of the early age at onset of ALL, parentalure before conception, maternal exposure duringancy, and exposure of the child to environmentals could all play a role.k of childhood ALL has previously been associatedexposure to diagnostic X-rays. The most studied
ure has been maternal X-rays during pregnancy.56, it was reported that children whose mothers
), Nicholas H. de Klerk (TICHR), Lin Fritschi (WA Institute forl Research), Ursula R. Kees (TICHR), Margaret Miller (EdithUniversity), and Judith R. Thompson (WA Cancer Registry) wereearch investigators. Helen D. Bailey (TICHR) was the projectator. The clinical investigators were as follows: Frank Alvarounter Hospital, Newcastle), Catherine Cole (Princess Margaretl for Children, Perth), Luciano Dalla Pozza (Children's Hospital atead, Sydney), John Daubenton (Royal Hobart Hospital, Hobart),ownie (Monash Medical Centre, Melbourne), Liane LockwoodChildren's Hospital, Brisbane), Maria Kirby (Women's andn's Hospital, Adelaide), Glenn Marshall (Sydney Children'sal, Sydney), Elizabeth Smibert (Royal Children's Hospital,rne), and Ram Suppiah (previously Mater Children's Hospital,e).
ponding Author: Helen Bailey, Telethon Institute for Child Healthch, Centre for Child Health Research, The University of Westernia, P.O. Box 855, Perth, WA 6872, Australia. Phone: 61-8-9489-ax: 61-8-9489-7700. E-mail: [email protected]
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
ed having an abdominal X-ray during pregnancytwice as likely to die of leukemia as children whosers did not recall having an X-ray (4). These findingssupported by those of a continuation of that studyd others that used medical records to assess expo-uring pregnancy (6, 7). Since the 1950s and 1960s,he radiation dose delivered in obstetric X-rays andported risk of childhood cancer associated withnal X-rays during pregnancy have decreased (8).ver, a recent review of case-control studies pub-since 1990 concluded that there remains an excesslbeit small, of childhood leukemia from maternals during pregnancy (9).re is little evidence of any association of childhoodwith maternal abdominal or pelvic X-rays beforeption (10-12). Findings on the risk of paternalure before conception are inconsistent, with twots suggesting an association with abdominals (12, 13) and two reports finding no association). There is some evidence in animal models, how-that paternal exposure to high-dose radiationto conception is associated with leukemia in theing (14, 15).association between diagnostic X-rays in childrene risk of later ALL has also been investigated withcting results. Since 1990, three case-control studies) have reported an increasing risk with increasinger of X-rays, whereas four case-control studies1, 13, 19) and one cohort study (20) have foundociation with ever having an X-ray.previous study of childhood ALL has investigatede of computerized tomography (CT) scans specif-Because CT scans involve exposure to much high-els of radiation than plain X-rays, concerns haveraised about the possible carcinogenic effects ofuse (21, 22). Since the 1980s, the use of CT scanscreased exponentially in the United Kingdomhe United States (21), which probably reflectsce in most other developed countries; thus, furtherch into the risk of exposure to diagnostic radiationrranted.Australian Study of Causes of Acute Lympho-Leukaemia in Children (Aus-ALL) was a nationalthat began in 2003 and investigated genetic,y, and environmental causes of childhood ALL.im of this current analysis is to investigate whetherostic X-rays, including CT scans, of the parents orin key periods increased the risk of childhoodSpecifically, we investigated whether maternalinal or pelvic region X-rays before or during theancy, paternal abdominal or pelvic region X-raysconception of the child, or any X-rays of the childsed the risk of ALL, and whether the relationshipby the type of procedure or the part of the body
ed. For parents, we limited our analyses to abdom-r pelvic region X-rays as we thought these were the
ures most likely to affect the reproductive organs oroping fetus.
in thedone
r Epidemiol Biomarkers Prev; 19(11) November 2010
Research. on November 30, 202cebp.aacrjournals.org Downloaded from
rials and Methods
populations-ALL was a national, population-based, case-l study of childhood ALL conducted in Australiaen 2003 and 2007, which recruited 416 cases andcontrols with ages younger than 15 years (23, 24).child to be eligible, the biological mother needed toailable and have sufficient English skills to completeuestionnaires. Case families were identified andited through all 10 pediatric oncology centers inalia. Cases were eligible if they had been diagnoseden July 1, 2003, and December 31, 2006, and hadd remission. The study had Human Research Ethicsittee approval from participating hospitals. Controls
recruited by random digit dialing between 2003 andndwere frequency matched to cases by age, sex, andf residence in an approximate ratio of 3:1. Full detailsstudy population, including control recruitmentds, have been previously published (23, 24).
collectionh parents were mailed questionnaires that includedions about a range of potentially leukemogenicures. The parents were asked whether they hadny X-rays of the abdomen (including stomach), pel-ips, lower back, or kidney before the birth of theFathers were asked about exposures up until theof the child as it was thought that they would recallmore accurately than exposures in the time beforeption. If they had had any X-rays, they were askedtheir age at the time of each X-ray and the type ofdure [plain X-rays, CT scans, intravenous pyelo-s (IVP), or barium studies]. In addition, mothersasked if any X-ray was done before or during theancy and, if the latter, the trimester of each X-ray.ers were also asked whether their child had anys of any part of their body. For each X-ray, they wereto choose the part of the body from the followingead (including dental), chest, abdomen (includingch or hips), arm(s), leg(s), or whole body. They weresked the type of procedure and the child's age at thef the procedure. Both parents were also asked fors of any radiotherapy to treat cancer.
sure metricseach X-ray reported by the parents, we identifiedpe of X-ray, parts of the body examined, and thency of exposure in the time period of interest.determine whether the father had an X-ray doneconception, the approximate conception date was
ated by subtracting the child's gestational age at birthis or her date of birth. The father's age in years at theption date was calculated. If the X-ray was done atge or earlier, it was classed as being done beforeption. X-rays done after this age were not included
analyses. The association ofALLwith paternalX-raysin the year of conception was also investigated.
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
nalyses of the child's exposures, we included alls done before diagnosis for cases and all X-raysbefore the date of return of the written question-for controls. There were several reasons for thisach: The control questionnaire asked about expo-“since your child was born.” As controls were fre-y matched, there was no corresponding case whosef diagnosis could have been used. The analysesthen repeated including only X-rays done more6 months before the above censoring dates toe procedures done when cases were showing signsess but had yet to be diagnosed with ALL.
nophenotype and cytogenetic classificationobtained information about immunophenotypeytogenetic subtypes of participating cases fromedical record details provided by clinicians. Thewas determined using metaphase cytogeneticsuorescence in situ hybridization screening.
tical analysisds ratios (OR) and 95% confidence intervals (95%CI)estimated using unconditional logistic regressionSS for Windows version 15 (SPSS, Inc.). Separatels were fitted to assess the associations betweennal, paternal, and filial X-rays and ALL. All modelsadjusted for the study matching factors: age atosis, sex, and state of residence. Variables consid-priori to be potential confounders of the associationen X-rays and risk of ALL were assessed individu-r inclusion in the models. Different variables wereered for analyses of maternal exposures (maternaltion, family income, ethnicity, birth order, birtht, maternal age at birth, maternal height, maternalmass index, maternal smoking in the year before, paternal exposures (paternal education, familye, ethnicity, paternal age at birth, paternal smoking2 years before birth), and filial exposures (maternaltion, family income, ethnicity, maternal age at birth,rder, birth weight, gestation, plurality, birth defect,nal smoking in the year before birth, and paternaling in the 2 years before birth). Variables that metpirical criteria for confounding (independently
ated with both the exposure and the outcome) wereed in the final models, as follows: for maternalures, maternal age at the child's birth, birth order,nal education, and family income; for paternal expo-paternal age at the child's birth, paternal education,mily income; and for filial exposures, maternal ageh, birth defect, and family income.also analyzed the data by immunophenotype andenetic subtypes.
-analysisconducted a meta-analysis of our findings on pater-dominal or pelvic region X-rays before conception
he published findings of previous studies. We diderform meta-analyses of either maternal or filial
Thehood
Caacrjournals.org
Research. on November 30, 202cebp.aacrjournals.org Downloaded from
exposures as they have been reviewed recently, 26). To be included in the meta-analysis, eachwas required to (a) be a cohort or case-controlthat presented ORs and the corresponding 95% CIse association between paternal abdominal or pelvicn X-ray before conception and risk of childhoodor provide data that allowed them to be calculated)b) be population based. We searched PubMed on13, 2010, for original studies of X-rays beforeption and risk of childhood ALL published fromto 2010. The following search terms were used:emia,” “(X-ray or radiation)” and “(child, childool, or infant),” and “preconception.” All journals were read to identify studies that specifically in-ated paternal abdominal or pelvic region X-rays.case-control studies were identified. Two of theses restricted their investigations to subpopulationsldren with ALL; one was restricted to children withsyndrome (10); and the other was restricted to
s (17). The study by Meinert and colleagues (13)ncluded despite using a case definition of “acuteood leukemia” as ALL accounts for 75% to 80%s group (2). We extracted the most appropriate OReach study and used fixed-effects, precision-basedting (27) to calculate summary ORs with the mostpriate result from Aus-ALL. Statistical hetero-ty among studies was assessed using the Cochraneand expressed using I2. A forest plot was producedSTATA version 10 (StataCorp).
lts
were notified of 568 incident cases of ALL, of49 were ineligible to participate: 30 from non-
sh-speaking backgrounds, 12 overseas visitors, 3e biological mother was unavailable, and 4 whoot reach remission. Of 519 eligible cases, parents(80.2%) cases consented to participate in the study88 mothers (74.8% of eligible) and 328 fathersof eligible) returned the questionnaire.
the 2,947 eligible control families identified throughm digit dialing, 2,071 (70.3%) agreed to take part.se of age and sex frequency-matching quotas,ly recruited 1,361 of these families to the study.e recruited families, 870 mothers (63.9%) andthers (55.1%) returned the written questionnaires.demographic characteristics of cases and controls
returned the written questionnaire were generallyr (Table 1). Case children were more likely to be aorn (47.8%) and to have a birth defect (5.7%) thancontrols (41.7% and 3.1%, respectively). Controlts were more likely than case parents to be tertiaryted, have a higher income, and to have been
ars old or older when the child was born.
rnal X-rays before or during pregnancy
re was no evidence of an increased risk of child-ALL with maternal abdominal or pelvic region
ncer Epidemiol Biomarkers Prev; 19(11) November 2010 2899
reviations: ACT, Australian Capital Territory; NSW, New South Wales; NT, Northern Territory; QLD, Queensland; SA, Southtralia; TAS, Tasmania; VIC, Victoria; WA, Western Australia; pa, per annum.opean = at least three European grandparents; 50% European = two European grandparents; at least 50% non-European andknown if 50% European = two non-European grandparents and the ethnicity of the other two grandparents is unknown;terminate = no two grandparents with the same ethnicity (i.e., European or non-European) and more than two grandparents
Table 2. ORs and 95% CIs for risk of ALL with parental diagnostic X-rays of the abdominal or pelvic region at any time before or during theindex pregnancy
Any kidney X-ray 3.9 4.5 0.74 (0.40-1.39) 0 0.1 Not estimated 6.7 3.3 2.13 (1.15-3.94)Any plain kidney X-ray or CT 1.8 1.5 0.98 (0.38-2.52) 0 0 Not estimated 2.5 2.3 1.16 (0.48-2.82)Any IVP 3.1 3.0 0.92 (0.45-1.89) 0 0.1 Not estimated 4.9 1.5 3.56 (1.59-7.98)
NOTE: The abdominal or pelvic region is defined as the abdomen (including stomach), lower back, pelvis, hips, or kidneys.*There is one missing value as one mother did not know how the timing of the X-ray was related to the index pregnancy.†There are two missing values for cases and one for controls because the fathers did not know either if they had had a relevant X-ray or their age at the time of the X-ray.‡Adjusted for matching variables (age group, sex, and state), birth order, maternal education, and maternal age at birth.§The reference groups are those who did not have diagnostic radiological X-ray of the abdomen, lower back, pelvis, hips, or kidneys before or during pregnancy (286 cases and582 controls).∥Adjusted for matching variables (age group, sex, and state), paternal education, and paternal age at birth.¶The reference groups are those who did not have diagnostic radiological X-ray of the abdomen, lower back, pelvis, hips, or kidneys before the conception of the index child.**Diagnostic X-rays are defined as any plain X-ray, CT scan, IVP or barium study.
*The†Ch‡Adbirth§Thediag∥Diagnostic X-rays are defined as any plain X-ray, CT scan, IVP, or barium study.
Bailey et al.
Cance2902
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
s at any time before the pregnancy (Table 2). Therelso no increased risk with any specific type of X-ray,ing CT scans, or having more than one X-ray2). Relatively few women (4 cases and 16 controls)ted having an abdominal or pelvic region X-rayg pregnancy, and there was no evidence of an in-d OR associated with any body site being X-rayed2). There were insufficient numbers of exposed
ers to analyze the data by trimester of exposurets not shown).
nal X-rays before conceptionre was little evidence of an increased risk of child-
ALL with paternal abdominal or pelvic regions at any time before the conception of the child
moreand th
r Epidemiol Biomarkers Prev; 19(11) November 2010
Research. on November 30, 202cebp.aacrjournals.org Downloaded from
1.17; 95% CI, 0.88-1.55; Table 2) or with any suchs done when the father was the same age as het conception (OR, 1.64; 95% CI, 0.79-3.42), althoughely few men (13 cases and 20 controls) reportedn X-ray (results not tabulated). The OR for havingbdominal or pelvic region X-ray at any time beforeption was 1.03 (95% CI, 0.73-1.44) and the OR forthan one was 1.47 (95% CI, 0.98-2.21; P for trend; Table 2). The OR for having an IVP was 3.56CI, 1.59-7.98; Table 2). A smaller increase wasor any barium study (OR, 1.59; 95% CI, 0.81-3.13;2). Among the 27 fathers who had an IVP beforeption, two case fathers and one control father had
rend = 0.10 P for rend = 0.53 f X-ray plain X-ray 45.6 39.2 1.22 (0.94-1.59) 41.7 36.7 1.15 (0.88-1.52) CT 1body and type of X-ray
2
0.90 0.36- 2.27) 7 1 0.87 0.32- 2.34)
abdominal X-ray 1
10 1.03 0.67- 1.61) 4 10 3 0.90 0.56- 1.44) ncluding stomach or hips)ny barium study 1 ny abdominal plain 8
17
0.990.98
0.36- 2.72)0.60- 1.60)
7 19 7
0.990.81
0.36- 2.72)0.47-1.38)
y or CT scanny IVP 1
X-ray to the head 1
19
0.771.00
0.22- 2.68)0.62- 1.62) 1
1 10 8
0.771.15
0.22-2.68)0.70- 1.88)
ncluding dental X-rays)
ny plain head X-ray 9 8 1.06 0.64- 1.75) 9 7 1.22 0.72- 2.05) ny head CT 1 1 0.77 0.26- 2.26) 4 1 0.80 0.27- 2.37) plain chest X-ray 1 16 1.26 0.89- 1.77) 1 3 15 3 1.24 0.86- 1.78) ny 9.2 .1 ( 8. . (
ny plain arm X-ray 16.6 11.7 1.49 (1.00- 2.22) 13.1 10.2 1.21 (0.77-1.91)ny plain leg X-ray 8.3 7.7 1.20 (0.74-1.96) 5.3 7.2 0.76 (0.42-1.36)
re is one missing value as the child's age when the X-ray was done was not known.ildren who only had an X-ray within 6 months of the censoring date were excluded from this analysis.justed for matching variables (age group, sex, and state), birth order, maternal education, maternal age at birth, and child hasdefect.reference group is children who did not have any diagnostic radiological X-rays at any time before the censoring date (date ofnosis for cases and date of questionnaire return for controls).
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
1-32 years) and 12 years for controls (range 1-16. There was little evidence of an increased risk ofood ALL with paternal CT scans of the abdominalvic region at any time before the conception of the(Table 2).
s in childhoodre was little evidence of an increased risk of ALLany childhood X-rays before the censoring date1.21, 95% 0.93-1.57; Table 3) or when the analysesrestricted to X-rays done more than 6 months beforensoring date (OR, 1.15; 95% CI, 0.88-1.51). There
ome evidence of an increased risk with having moren
fers mainly to somatic (not constitutive) changes in tumor cells. Only
Caacrjournals.org
Research. on November 30, 202cebp.aacrjournals.org Downloaded from
arm (OR, 1.49; 95% CI, 1.00-2.22) at any time beforensoring date, but no such increase when the analy-ere restricted to X-rays done more than 6 monthse the censoring date (Table 3). There was nonce of an increased risk associated with X-rays ofther part of the body or other type of X-rays, includ-T scans (Table 3).
nophenotype and cytogenetic subtypesre was some evidence of an association betweennal X-rays before conception and risk of t(12;21)-Runx-1) subtype of ALL (OR, 1.68; 95% CI, 0.92-
able 4), and X-rays of the child before the censoring
e X-ray (OR, 1.34; 95% CI, 0.93-1.92) or an X-ray date and ALL involving translocations other than t(12;21)
4. ORs for any parenta
diagnostic radiological X-ray to the abdominal or pelvic region before theenotype and cytogenetic subtype
E: Eight cases with no information on genetic subtype and seven cases with unknown/other lineage were excluded. Subgroupsnot mutually exclusive.justed for matching variables (age group, sex, and state), birth order, maternal education, and maternal age at birth.e reference category is mothers who did not have an abdominal or pelvic X-ray at any time before or during the indexnancy.justed for matching variables (age group, sex, and state), paternal education, and paternal age at birth.reference category is fathers who did not have an abdominal or pelvic X-ray at any time before the conception of the index.
r the analyses of maternal exposure, the numbers of cases are as follows: TCF3/PBX1 t(1;19), n = 7; BCR-ABL t(9;22), n = 6;r translocations, n = 35; for the analyses of paternal exposure: TCF3/PBX1 t(1;19), n = 7; BCR-ABL t(9;22), n = 6; otherslocations, n = 31.her structural change” is defined as any structural change that is neither a translocation nor a deletion, such as isochromo-es or abnormal derivative chromosomes.ere were less controls (maternal analyses 680; paternal analyses 601) included in these analyses because one state had noes of ALL in this subcategory; therefore, controls from that state were excluded from the analyses (maternal analyses 179;rnal analyses 148).
NOTare*Adbirth†Thediag‡TC§Othor a∥Thcon¶Refers mainly to somatic (not constitutive) changes in tumor cells. Only one of these cases had Down syndrome.
Bailey et al.
Cance2904
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
LL rearrangements (OR, 2.59; 95% CI, 1.30-5.16;5). In the “other translocations” group, there weres with TCF3/PBX1 t(1;19), 6 with BCR-ABL t(9;22),3 with mostly individual translocations. The OR for-ray of the child before the censoring date for thesenumbers were 2.88 (95% CI, 0.51-16.39), 4.95 (95%4-45.14), and 2.27 (95% CI, 1.03-5.00), respectively.
tal radiotherapycase father reported receiving radiotherapy before
ption of the index child, as did three control fathers.asemother and three control mothers reported havingherapy before conception.No statistical analyseswerebecause of the small numbers of these exposures.
-analysesails of previous studies considered for inclusion ineta-analyses of paternal abdominal or pelvic X-rays
mmarized in Supplementary Table S1. There weretudies that investigated the risk of paternal X-rays
analywith t
r Epidemiol Biomarkers Prev; 19(11) November 2010
Research. on November 30, 202cebp.aacrjournals.org Downloaded from
to conception (11, 13). The larger of these, by Shuolleagues (11), studied the 2 years before concep-whereas the other, by Meinert and colleagues (13),d the 2 years before the child's birth. The ORs fromtwo studies and ours were all above unity, and thed OR was 1.17 (95% CI, 0.92-1.48; Fig. 1). For thisanalysis, we used our OR for any paternal X-rayswhen the father's age at the time of the X-ray wasthe same as at conception or 1 year younger, butooled estimates were similar when we used ourr X-rays done when the father's age at the time of-ray was the same as at conception, or the OR foraving an X-ray before conception. Another studyvestigated the risk of paternal X-rays for a longerof exposure (5 years before conception) in children
Down syndrome. The pooled OR for that studyus-ALL for any paternal X-rays within 5 years ofption from was 1.22 (95% CI, 0.85-1.74). The meta-
E: Eight cases with no information on genetic subtype and seven cases with unknown/other lineage were excluded. Subgroupsnot mutually exclusive.justed for matching variables (age group, sex, and state), birth order, maternal education, maternal age at birth, and child hasdefect.reference group is children who did not have any diagnostic radiological X-ray at any time before the censoring date (date ofnosis for cases and date of questionnaire return for controls).F3/PBX1 t(1;19), n = 7; BCR-ABL t(9;22), n = 6; and other translocations, n = 33.er structural change is defined as any structural change that is neither a translocation nor a deletion, such as isochromosomesbnormal derivative chromosomes.ere were 687 controls included in these analyses because one state had no cases of ALL in this subcategory; therefore, 183trols from that state were excluded from the analyses.
of botprocesure.ported(95%mostan asbecauin Auhavinin somotheFin
risksbeenno ashoodALL
Figureconcepwas the
X-rays and Childhood ALL
www.a
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
the results are not presented. There was no studyarable with ours that presented results on thency of X-rays at any time before conception; thus,ere unable to do a meta-analysis of dose response.
ssion
re was little evidence of any increased risk of child-ALL associated with maternal X-rays before org the pregnancy or with X-rays of the child.vious studies have also reported no evidence of ansed risk of childhood ALL with maternal abdo-or pelvic region X-rays before pregnancy (10-12).isks associated with maternal exposure duringancy have been well studied over the last 50 years.ugh at least six studies published before 1990sted an increased risk of ALL (5, 28) or leukemia-31) associated with abdominal or pelvic X-rays
g pregnancy, it is probably more appropriate toare our findings with more recent studies because
creasebut a
same or 1 year younger than his age at conception.
Caacrjournals.org
Research. on November 30, 202cebp.aacrjournals.org Downloaded from
h the reduction in the radiation doses incurred fromdures (8) and the decreasing prevalence of expo-A meta-analysis of studies published since 1990 re-that the pooled relative risk of leukemia was 1.16
CI, 1.00-1.36; ref. 9). However, none of the threerecently published studies of ALL (10, 16, 32) foundsociation. Like us, their estimates lacked precisionse of the low prevalence of exposure; for example,s-ALL, less than 2% of control mothers reportedg an abdominal X-ray during pregnancy, whereasme early studies about 15% to 20% of controlrs reported exposure during pregnancy (30, 33).dings from case-control studies investigating theassociated with X-ray exposures after birth havemore inconsistent. Like ours, four studies foundsociation between having any X-ray during child-and risk of either leukemia (13, 29) or specifically(10, 19). Three other studies found little or no in-
d risk of ALL with having a low number of X-raysn increased risk with having a “higher” number,
1. Forest plot showing study-specific and summary ORs for paternal abdominal or pelvic X-rays before conception. *The time period beforetion varied by study: Meinert et al., within 2 years of birth; Shu et al., within 2 years of conception; Bailey et al., father's age at the time of the X-ray
ncer Epidemiol Biomarkers Prev; 19(11) November 2010 2905
ing inconcegermoffsppaternmatidleukeequivtion (male(43) aspringnal reinvestof ALsuresincludevidethe rigenomnewbALL,natallsult frNo
ciatedduresobserIVP icauseoffsprrepor0.92-3diagnassociderlyiurinato thegatedare rethinkof thethe yALL,suchnal orcollecor the
Bailey et al.
Cance2906
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
gh the definition of higher varied from “more thanto “more than five” (11, 16, 28). A Chinese study ofia cases diagnosed in the mid-1980s and a morelarge U.S. study of ALL both found an increased
f disease with X-rays before diagnosis, with a dosense (17, 18). Results within studies have also beensistent, which makes them difficult to interpret. Ina, Infante-Rivard and colleagues analyzed risk byype and found an increased risk of ALL with X-rayure in boys with polymorphisms of one of the DNAgenes (APE ex 5) but a decreased risk among girlsthe same polymorphism (16). Similarly, one smallese study, which used hospital-based controls,an association between dental X-rays and non–ALL, but not with X-rays of other parts of the bodying higher doses of radiation (34).rall, there was little evidence in our case-controlof an increased risk of childhood ALL associatedaternal X-rays before conception, which was con-t with the results of our meta-analyses. However,was some evidence that the father having had morene abdominal or pelvic region X-ray at any timeconception was associated with an increased risk.was also evidence of increased risk with the fatherg an IVP or barium study before conception. Onlyther studies have specifically investigated paternalure to abdominal or pelvic region X-rays beforeption, and these have had inconsistent results). In the first study, which was restricted to childrenosed with leukemia by 18 months of age, there wasnce of increased risk, including a dose response,both paternal lower and upper abdominal X-raystime before conception (12). The ORs were highestdominal X-rays done in the year before conception.s-ALL, there were only 18 cases diagnosed bynths of age; thus, we did not have adequate powersimilar analyses. However, excluding those diag-by 18 months made little difference to our find-The second study, conducted in Germany, alsoan increased risk of acute leukemia associatedaternal X-rays of the abdomen or gastrointestinal
in the 2 years leading up to the child's birth (13).the analyses were restricted to children diagnosedmonths, no association was found. A limitation ofudy was the inclusion of irrelevant paternal expo-between conception and birth. The two most recents, both conducted in the United States, found noiation with paternal abdominal X-rays and ALL1). The first of these (11) was the study with thecases and investigated the risk of any lower abdom--rays in the 2 years before conception, whereas thetudy investigated the risk of paternal abdominals within the 5 years leading up to conception in aation of children with Down syndrome (10).dies of any paternal X-rays before conception, re-ss of the site of the X-ray, have reported conflicting
s, with two suggesting an association (28, 35) andinding no association (12, 19). The risk of ALL in
assocthere
r Epidemiol Biomarkers Prev; 19(11) November 2010
Research. on November 30, 202cebp.aacrjournals.org Downloaded from
ffspring of male radiation workers has also beentigated using estimated dose of radiation in theths before conception (36-40). The initial report ofreased risk of ALL and evidence of a dose responseg children of men employed at Sellafield nuclear(36) has not been replicated in subsequent, largers (37-40).ernal exposure to radiation damages sperm, result-DNA double-strand breaks (41). Therefore, it is
ivable that exposure close to conception damagescell DNA and thus increases the risk of ALL in thering. There is evidence from animal models thatal exposure to radiation in the spermatozoa or sper-stages of germ cell development increases the risk ofmia in the offspring (14). In humans, this would bealent to exposure about 30 to 47 days before concep-42). In addition, in the later stages of development,germ cells lose the ability to repair DNA damagend the risk of chromosomal aberrations in the off-may depend on the efficiency of subsequent mater-pair of the DNA of the zygote (44). Therefore, toigate if paternal X-rays are associated with the riskL in the child, it may be necessary to identify expo-in the fewmonths before conception and, perhaps, toematernal genotype in the analyses.We foundweaknce that paternal X-rays before conception increasedsk of the ETV6-Runx-1 t(12;21) ALL. ETV6-Runx-1ic fusion sequences have been identified in dried
orn blood spots of children later diagnosed withsuggesting that this subtype of ALL is initiated pre-y (45). Thus, it is plausible that these aberrations re-om unrepaired paternal germ cell DNA damage.previous study has reported the risk of ALL asso-with the father having specific types of proce-before conception, such as the increased risk weved with the father having an IVP. If having ans carcinogenic, it would be plausible that it couldcancer in pelvic tissues as well as cancer in theing through damage to germ cells. A recent studyted an OR for prostate cancer of 1.67 (95% CI,.03) with having an IVP at least 5 years beforeosis, suggestive of an increased risk (46). Anyation may be due to some factor related to the un-ng condition—that is, a condition of the kidney orry tract, or its subsequent treatment—rather thanIVP itself. To our knowledge, no one has investi-whether paternal renal or urinary tract conditionslated to ALL in the offspring, and it is difficult toof a plausible reason why they should be. In onefew studies of the effects of paternal medication inear before conception and the risk of childhoodno association was seen with any medications,as antibiotics, that may have been used to treat re-urinary tract disease (47). In Aus-ALL, we did nott any information on why the X-rays were needed,subsequent diagnosis or treatment. The observed
iation could also be due to recall bias; however,is no reason why case fathers should selectively
childotherthe sm(9;22)is plauhoodthoug(9;22)moreour rewithpotenfurtheTo o
invesCTs whavinfromdoseappro220 pilar prin botbeen sbe spe(52). Nan incabdomever hpretedless thcase aCT scnot abmaterexposlenceAustrple, anumb10% athan 1∼20%indivibearinAu
ascertchildreligib
of eligagreetrol p93% oof seltion fmoreand,controthe geing ancationprioriand eied in otion bA li
was crecordbe veas it iswouldlocatiselvesas witfor botempquestnot redeeplfrequethe caand, iderrepA powhelmto recway tby a pwhichundemethothat wthe likconceIn
creaseX-rayhavinincreamoretimeincreapelvihavinin a C
X-rays and Childhood ALL
www.a
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
having this procedure and not other types of pro-es. However, our findings were based on smallers of exposed subjects. In addition, the lack oflear temporal relationship between the timing ofP and conception is at odds with the evidenceanimal models of a critical window for paternalure to radiation (14); thus, our findings may bechance.found evidence of an association between thehaving any X-ray and ALL involving translocationsthan t(12;21) andMLL rearrangements. The ORs forall subgroups of TCF3/PBX1 t(1;19), BCR-ABL t
, and other rare translocations were all elevated. Itsible that these translocations are caused by child-X-rays. TCF3/PBX1 t(1;19) translocations areht to occur postnatally (48) and the frequency of ttranslocations may increase with age as they arecommon in adult ALL (49), which is consistent withsults. This evidence of an increased risk of ALLtranslocations associated with X-ray exposure is atially important finding and should be examinedr in larger studies.ur knowledge, this is the first study to specifically
tigate the risks associated with having CT scans.ere of particular interest, as the radiation dose fromg a CT scan is many times larger than that receivedhaving a plain X-ray; for example, the radiationfrom a routine CT of the abdomen or pelvis isximately equivalent to that received from havinglain chest X-rays (50). In addition, the dose for sim-ocedures may vary significantly across institutionsh adult (50) and pediatric populations (51). It hasuggested that a large number of future cancers willcifically related to abdominal and pelvis CT scansonetheless, in Aus-ALL, we found no evidence ofreased risk with maternal or paternal CTs of theen or pelvis before conception, or with the childaving a CT. However, these results should be inter-in light of the small number of exposed subjects;an 5% of case and control parents and 2% or less ofnd control children in this study had a relevantan; thus, the estimates lacked precision. We werele to investigate the risk of ALL associated withnal CTs during pregnancy as there were only twoed controls and no exposed cases. Such low preva-may not be found in other populations, or indeed inalia, if we were to repeat the study now. For exam-U.S. study found that between 1993 and 2006, theer of CTs done annually increased by more thanyear, whereas the population increased by less% a year (53). In addition, they also estimated thatof abdominal and pelvic CT scans are done onduals ages 18 to 44 years (approximately the child-g years; ref. 53).s-ALL had strengths and limitations. Cases wereained from oncology centers that treat almost all
en with ALL in Australia and 75% of parents ofle cases consented to participate in Aus-ALL; 70%
popularger
Caacrjournals.org
Research. on November 30, 202cebp.aacrjournals.org Downloaded from
ible controls recruited by random digit dialing alsod to participate. However, 64% of participating con-arents returned the questionnaires compared withf participating case parents, raising the possibilityection bias. We had information on parental educa-or all Aus-ALL participants. Control parents werelikely than case parents to have a tertiary educationusing area-based measures, we have shown thatl parents were of higher socio-economic status thanneral Australian population (23). Furthermore, hav-X-ray was positively associated with parental edu-level among the controls. SES was considered ato be a potential confounder of this association,thermaternal or paternal education levelwas includ-ur analytic models. It is therefore unlikely that selec-ias related to SES had amajor impact on our findings.mitation of Aus-ALLwas that information on X-raysollected from the parents rather than from medicals. However, apart from during pregnancy, it wouldry difficult to collect evidence from medical recordslikely that over the lifetime of an individual, X-rayshave been taken at multiple locations, and these
ons could only be identified by the parents them-. Because we collected information from parents,h most case-control studies, there was the potentialth exposure misclassification and recall bias. We at-ted to minimize this using standardized writtenionnaires. Nonetheless, using these methods wouldmove the potential for case parents to think morey about past exposures and thus to report themmorently. However, our results suggest that this was notse as there were comparatively fewORs above unityf anything, in our study, case mothers may have un-orted having an X-ray before or during pregnancy.
ssible reason for this is that they were so over-ed with details of their child's current treatment
all their own medical history; however, there is nohat we can verify this. This theory is not supportedrevious study of ALL and maternal prenatal X-rays,reported that case and population control parents
restimated exposure to a similar extent (54). Ourd of asking about X-rays before conception meante were unable to restrict the period investigated toely critical times, such as the final fewmonths beforeption for fathers.conclusion, there was little evidence of any in-d risk with maternal abdominal or pelvic regions before or during the pregnancy or with the childg any X-rays. There was some evidence of ansed risk of ALL in the offspring if the father hadthan one abdominal or pelvic region X-ray at anybefore conception. There was no evidence of anysed risk with maternal or paternal abdominal orc CTs before pregnancy or with the child everg a CT. Given the high dose of radiation deliveredT, it is important to repeat these analyses in other
lations for which the prevalence is higher or in asample to improve the precision. Pooled analyses
ncer Epidemiol Biomarkers Prev; 19(11) November 2010 2907
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
lanned in the Childhood Leukemia Internationalrtium (55).
sure of Potential Conflicts of Interest
otential conflicts of interest were disclosed.
owledgmentsRece
d other environmental factors in the etiology of leukemias andn-Hodgkin's lymphomas in childhood: a case-control study.mori 1990;76:413–9.
20. Hachexp
21. HaBr
22. Roca
23. Baco20
24. Michca
25. Scdiarev
26. WaexDo
27. FleRe
28. Shtro63
29. Agep18
30. Kame19
31. Muto
32. RoPema
33. Po19
34. NilymCo
35. GranMo
36. GaReyo19
37. Kinexlym
38. McPaca
39. Soof
r Epidemiol Biomarkers Prev; 19(11) November 2010
Research. on November 30, 202cebp.aacrjournals.org Downloaded from
Support
onal Health andMedical Research Council (NHMRC) grant 254539LL), NHMRC Post-Graduate Scholarship 513934 (H. Bailey),C Career Development Award 513910 (E. Milne), and NHMRChip 513706 (L. Fritschi).costs of publication of this articlewere defrayed in part by the paymentcharges. This article must therefore be hereby marked advertisementdance with 18 U.S.C. Section 1734 solely to indicate this fact.
hank Dr. Rick Thompson for his advice about X-rays. published OnlineFirst 09/22/2010.
rencesParkin DM, Kramarova E, Draper GJ, et al, editors. Internationalidence of childhood cancer, vol. II. 1998. International Agency forearch on Cancer, World Health Organization, Lyon (France).le J. Epidemiology of childhood cancer. Lyon (France): Internationalency for Research on Cancer, World Health Organization; 1999.Nally RJ, Parker L. Environmental factors and childhood acutekemias and lymphomas. Leuk Lymphoma 2006;47:583–98.wart A, Webb J, Giles D, Hewitt D. Malignant disease in childhooddiagnostic irradiation in utero. Lancet 1956;271:447.
hell JF, Stewart AM. Prenatal irradiation and childhood malignancyeview of British data from Oxford survey. Br J Cancer 1975;31:–87.rd DD, Paterson JCS, Treuting WL. Fetal exposure to diagnosticays, and leukemia and other malignant diseases in childhood.atl Cancer Inst 1959;22:1093–104.cmahon B. Prenatal x-ray exposure and childhood cancer. J Natlncer Inst 1962;28:1173.ll R, Wakeford R. Risk of childhood cancer from fetal irradiation.J Radiol 1997;70:130–9.keford R. On pre- or postnatal diagnostic x-rays as a risk factorchildhood leukaemia. Radiat Environ Biophys 2009;48:237–9.abery AM, Olshan AF, Gamis AS, et al. Exposure to medical testdiation and acute leukemia among children with Down syndrome:eport from the Children's Oncology Group. Pediatrics 2006;118:99–508.u XO, Potter JD, Linet MS, et al. Diagnostic x-rays and ultrasoundosure and risk of childhood acute lymphoblastic leukemia byunophenotype. Cancer Epidemiol Biomarkers Prev 2002;11:–85.u XO, Reaman GH, Lampkin B, Sather HN, Pendergrass TW,bison LL. Association of paternal diagnostic-x-ray exposure withof infant leukemia. Cancer Epidemiol Biomarkers Prev 1994;3:
5–53.inert R, Kaletsch U, Kaatsch P, Schuz J, Michaelis J. Associationstween childhood cancer and ionizing radiation: results of apulation-based case-control study in Germany. Cancer Epidemiolmarkers Prev 1999;8:793–9.mura T. Paternal exposure to radiation and offspring cancer ince—reanalysis and new evidences. J Radiat Res (Tokyo) 1991;:64–72.her A, Varin M, Lamontagne Y, Oth D. Effect of pre-conceptionalternal or internal irradiation of N5 male mice and the risk ofkemia in their offspring. Carcinogenesis 1998;19:1553–8.nte-Rivard C. Diagnostic x rays, DNA repair genes and childhoodte lymphoblastic leukemia. Health Phys 2003;85:60–4.u XO, Jin F, Linet MS, et al. Diagnostic X-ray and ultrasoundosure and risk of childhood cancer. Br J Cancer 1994;70:–6.rtley K, Metayer C, Selvin S, Ducore J, Buffler P. Diagnostic x-rayosures and risk of childhood acute leukemia. Am J Epidemiol9 (supp 11);169:S48.gnani C, Pastore G, Luzzatto L, Terracini B. Parental occupation
mmer GP, Seidenbusch MC, Schneider K, et al. A cohort study ofildhood cancer incidence after postnatal diagnostic x-rayosure. Radiat Res 2009;171:504–12.ll EJ, Brenner DJ. Cancer risks from diagnostic radiology.J Radiol 2008;81:362–78.bbins E. Radiation risks from imaging studies in children withncer. Pediatr Blood Cancer 2008;51:453–7.iley HD, Milne E, de Klerk N, et al. Representativeness of childntrols recruited by random digit dialing. Paediatr Perinat Epidemiol10;24:293–302.lne E, Royle JA, de Klerk NH, et al. Fetal growth and risk ofildhood acute lymphoblastic leukemia: results from an Australianse-control study. Am J Epidemiol 2009;170:221–8.hulze-Rath R, Hammer GP, Blettner M. Are pre- or postnatalgnostic X-rays a risk factor for childhood cancer? A systematiciew. Radiat Environ Biophys 2008;47:301–12.keford R. Childhood leukaemia following medical diagnosticposure to ionizing radiation in utero or after birth. Radiat Protsimetry 2008;132:166–74.iss JL. The statistical basis of meta-analysis. Stat Methods Meds 1993;2:121–45.u XO, Gao YT, Brinton LA, et al. A population-based case-con-l study of childhood leukemia in Shanghai. Cancer 1988;62:5–44.er EA, Schuman LM, Wallace HM, Rosenfie AB, Gullen WH. Anidemiological study of childhood leukemia. J Chronic Dis 1965;:113.plan HS. An evaluation of the somatic and genetic hazards of thedical uses of radiation. Am J Roentgenol Radium Ther Nucl Med58;80:696–706.rray R, Heckel P, Hempelmann LH. Leukemia in children exposedionizing radiation. N Engl J Med 1959;261:585–9.man E, Simpson J, Ansell P, Lightfoot T, Mitchell C, Eden TOB.rinatal and reproductive factors: a report on haematologicallignancies from the UKCCS. Eur J Cancer 2005;41:749–59.lhemus DW, Koch R. Leukemia and medical radiation. Pediatrics59;23:453–61.shi M, Miyake H. A case-control study of non-T-cell acutephoblastic-leukemia of children in Hokkaido, Japan. J Epidemiolmmunity Health 1989;43:352–5.aham S, Levin ML, Lilienfeld AM, et al. Preconception, intrauterine,d postnatal irradiation as related to leukemia. J Natl Cancer Instnogr 1966;19:347–71.rdner MJ, Snee MP, Hall AJ, Powell CA, Downes S, Terrell JD.sults of case-control study of leukemia and lymphoma amongung people near Sellafield Nuclear Plant in West Cumbria. BMJ90;300:423–9.len LJ, Clarke K, Balkwill A. Paternal preconceptional radiationposure in the nuclear industry and leukemia and non-Hodgkinsphoma in young people in Scotland. BMJ 1993;306:1153–8.Laughlin JR, King WD, Anderson TW, Clarke EA, Ashmore JP.ternal radiation exposure and leukemia in offspring—the Ontario
se-control study. BMJ 1993;307:959–66.rahan T, Haylock RG, Muirhead CR, et al. Cancer in the offspringradiation workers: an investigation of employment timing and a
Published OnlineFirst September 22, 2010; DOI: 10.1158/1055-9965.EPI-10-0542
nalysis using updated dose information. Br J Cancer 2003;89:5–20.rahan T, Roberts PJ. Childhood cancer and paternal exposure toizing radiation: preliminary findings from the Oxford Survey ofildhood Cancers. Am J Ind Med 1993;23:343–54.gh NP, Stephens RE. X-ray-induced DNA double-strand breaks inan sperm. Mutagenesis 1998;13:75–9.
ler ID. Current issues in mutagenesis and carcinogenesis. 69.mparison of the duration of spermatogenesis between maleents and humans. Mutat Res 1996;352:169–72.rchetti F, Wyrobek AJ. DNA repair decline during mouse spermio-esis results in the accumulation of heritable DNA damage. DNApair (Amst) 2008;7:572–81.rchetti F, Essers J, Kanaar R, Wyrobek AJ. Disruption of maternalA repair increases sperm-derived chromosomal aberrations. Proctl Acad Sci U S A 2007;104:17725–9.emels JL, Cazzaniga G, Daniotti M, et al. Prenatal origin of acutephoblastic leukaemia in children. Lancet 1999;354:1499–503.les P, Evans S, Lophatananon A, et al. Diagnostic radiation
cedures and risk of prostate cancer. Br J Cancer 2008;98:1852–6.n WQ, Shu XO, Potter JD, et al. Parental medication use and risk ofldhood acute lymphoblastic leukemia. Cancer 2002;95:1786–94.
Am55. Ch
Ju
Caacrjournals.org
Research. on November 30, 202cebp.aacrjournals.org Downloaded from
emels JL, Leonard BC, Wang YX, et al. Site-specific translocationd evidence of postnatal origin of the t(1;19) E2A-PBX1 fusion inildhood acute lymphoblastic leukemia. Proc Natl Acad Sci U S A02;99:15101–6.vandi F, Kebriaei P. Philadelphia chromosome-positive acutephoblastic leukemia. Hematol Oncol Clin North Am 2009;
:1043–63.ith B. Radiation dose associated with common computed tomog-hy examinations and the associated lifetime attributable risk ofncer. Arch Intern Med 2009;169:2078–86.Lean D, Malitz N, Lewis S. Survey of effective dose levels fromical paediatric CT protocols. Australas Radiol 2003;47:135–42.Gonzalez B. Projected cancer risks from computed tomographicns performed in the United States in 2007. Arch Intern Med 2009;9:2071–7.ttler FA, Thorriadsen BR, Bhargavan M, et al. Medical radiationosure in the US in 2006: preliminary results. Health Phys 2008;
:502–7.ante-Rivard C, Jacques L. Empirical study of parental recall bias.J Epidemiol 2000;152:480–6.
ildhood Leukemia International Consortium [Internet]. [cited 2010
l 13]; Available from: http://clic.berkeley.edu/.
ncer Epidemiol Biomarkers Prev; 19(11) November 2010 2909
2010;19:2897-2909. Published OnlineFirst September 22, 2010.Cancer Epidemiol Biomarkers Prev Helen D. Bailey, Bruce K. Armstrong, Nicholas H. de Klerk, et al. of Childhood Acute Lymphoblastic LeukemiaExposure to Diagnostic Radiological Procedures and the Risk
Updated version
10.1158/1055-9965.EPI-10-0542doi:
Access the most recent version of this article at:
