IJE Advance Access originally published online on June 10, 2005
International Journal of Epidemiology 2005 34(5):1100-1109; doi:10.1093/ije/dyi113
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cancer |
Vaccination history and risk of childhood leukaemia
1 Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, CT 06520-8034, USA
2 School of Public Health, University of California, Berkeley, CA 94720-7360, USA
3 Kaiser Permanente Northern California Region, Pediatric Oncology, 900 Kiely Blvd, Santa Clara, CA 95051, USA
* Corresponding author. E-mail: xiaomei.ma{at}yale.edu
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Background Previous studies on vaccination and childhood leukaemia generated inconsistent results.
Methods In the Northern California Childhood Leukaemia Study, a case–control study with incident cases and matched birth certificate controls, detailed written vaccination records were collected. A total of 323 cases aged 0–14 years at diagnosis and 409 controls were included in this analysis. All vaccinations were censored on the reference date (date of diagnosis for cases and the corresponding date for matched controls). Conditional logistic regression analysis was conducted, adjusting for potential confounding factors. A primary variable of interest is the number of administrations (doses) of various types of vaccines.
Results Vaccinations against diphtheria, pertussis, tetanus, poliomyelitis, measles, mumps, and rubella were not associated with the risk of leukaemia. The odds ratio for each dose of Haemophilus influenzae type b (Hib) vaccine was 0.81 (95% CI 0.68–0.96). Compared with children who received two or fewer doses of Hib vaccine, those who received three or more doses had a significantly reduced risk of childhood leukaemia (odds ratio = 0.55, 95% confidence interval 0.32–0.94). The number of doses of hepatitis B vaccine received was not associated with leukaemia risk.
Conclusions Hib vaccination is associated with a reduced risk of childhood leukaemia. Future studies with detailed exposure assessment and large sample sizes are needed to further address the role of vaccinations in the etiology of childhood leukaemia.
Keywords Case–control studies, child, leukaemia, vaccination
Accepted 5 May 2005
It is believed that immune function may play an important role in the etiology of childhood leukaemia since leukaemia is essentially a malignancy of the immune system. Vaccination, due to its role in stimulating the immune system, has been considered a potential modifier of the risk of childhood leukaemia. In a hospital series of childhood leukaemia cases in Brisbane, Australia, an excess of vaccinations against diphtheria, pertussis (whooping cough), and tetanus was reported in cases compared to the controls.1 Except for a study several decades later that observed an increased risk of various immunophenotypes of acute lymphoblastic leukaemia (ALL) associated with measles, mumps, and rubella (MMR) vaccination,2 no other reports of an increased risk of childhood leukaemia associated with vaccination have been seen. Instead, a few studies have reported a decreased risk of leukaemia (or unspecified childhood cancer) associated with vaccination (including small pox, diphtheria, tetanus, pertussis, measles, mumps, rubella, poliomyelitis, Haemophilus influenzae type b, and unspecified vaccinations), although the study design, types of vaccines studied, and subtypes of leukaemia varied substantially.3–10 In addition, several studies found no evidence of an association between vaccination and the risk of childhood leukaemia.11–13 The inconsistencies in existing findings with respect to vaccination and childhood leukaemia (Table 1) may be, in part, due to the variation in exposure assessment. Only two of the studies reported an exclusive reliance on written vaccination records,9,11 and most other studies relied on interviews with parents or self-administered questionnaires by parents.
|
In the Northern California Childhood Leukaemia Study (NCCLS) we evaluated the potential etiologic roles of a variety of genetic and environmental factors, including vaccinations. We addressed the quality of exposure assessment by obtaining written vaccination records with detailed information about the types of vaccines a child received, the number of doses of each vaccine, and the exact date each vaccine was administered. The potential etiologic role of diphtheria, pertussis, tetanus (DPT), poliomyelitis, MMR, hepatitis B, and Haemophilus influenzae type b (Hib) vaccination is assessed in the present analysis.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Study population and data collection
The NCCLS commenced in 1995 and is currently ongoing. The study area includes 17 counties in the Greater San Francisco Bay Area (1995 to the present) and in 1999 was expanded to a total of 35 counties in Northern and Central California. In the NCCLS, incident cases of newly diagnosed childhood leukaemia (age 0–14 years) are rapidly ascertained from major pediatric clinical centres, usually within 72 h after diagnosis. Although case ascertainment is hospital-based, a comparison with all population-based cases ascertained by the statewide California Cancer Registry (2000) confirms that the NCCLS protocol successfully identified 95% of all age-eligible newly diagnosed childhood leukaemia cases among residents of the 5-county San Francisco metropolitan area and 76 percent of such cases in the other 30 counties. Controls are randomly selected from the statewide birth files maintained by the Center for Health Statistics in the California Department of Health Services and individually matched to cases on date of birth, sex, mother's race (white, African American, or other), and Hispanic status (a child is considered Hispanic if either parent is Hispanic). For cases diagnosed before December 1, 1999, the case:control ratio was 1:1, and mother's county of residence at the time of child's birth was used as an additional matching criterion. For cases diagnosed on or after December 1, 1999, the case:control ratio is 1:2. A detailed protocol for control selection has been reported elsewhere, and it is indicated that the birth certificate controls selected in the NCCLS are comparable to ideal controls who would have been enrolled under optimum circumstances with respect to parental age, parental education level, and maternal reproductive history.14
To be eligible, each case or control had to (i) reside in the study area at the time of diagnosis; (ii) be under 15 years of age at the reference date (date of diagnosis for cases and the corresponding date for matched controls); (iii) have at least one parent or guardian who speaks English or Spanish; and (iv) have no previous history of malignancy. The study protocol was approved by the Institutional Review Boards of University of California, Berkeley and all collaborating institutions, and a written informed consent was obtained for all participating subjects.
An in-home personal interview with the primary care taker of each case or control subject—usually the biological mother (97.5% of all respondents)—was scheduled as soon as the written informed consent was obtained for a case or a verbal consent was obtained for a control subject (a written consent for controls was subsequently obtained at the time of interview). The respondent (>99% of respondents are parents) was asked to provide a copy of the child's complete vaccination record. If the parents did not have a written record, did not know whether the record was complete, or the record kept by the parents was considered incomplete by the parents themselves or by the NCCLS study staff, the parents were asked to sign an authorization form so a copy of the child's vaccination record could be requested directly from the primary care physician(s) by the study staff. The study staff evaluated the completeness of the written records by comparing the records with recommended immunization schedules based on the birth year and age of each child, as well as discussing with parents and/or providers. By the end of 2002, a total of 382 cases and 482 matched controls had been interviewed in the NCCLS. Among the subjects, 39 (10.2%) cases and 33 (6.8%) controls were excluded because no complete written vaccination records could be obtained from either parents or physicians. Among the remaining 792 subjects with complete vaccination data, 20 cases and 40 controls who were not in matched sets were excluded. Included in the present analysis are a total of 237 matched pairs (one case and one control) and 86 matched triplets (one case and two controls). Written vaccination records for 70.1% of cases and 66.0% of controls were provided by parents, and the remainder by physicians.
Approximately 85% of the eligible cases consented to participate. To enroll the 409 controls included in the study, a total of 847 potential controls were searched, among which 143 (16.7%) could not be located, and 126 (14.7%) refused without providing eligibility information. Of the 578 potential controls who were successfully located and whose eligibility was assessed, 477 (82.5%) were eligible. If the same percentage of eligible subjects among those whose eligibility was assessed (n = 578) was assumed in those whose eligibility could not be assessed (n = 143 + 126 = 269), a total of 699 potential controls were presumed eligible [n = 477 + (269 x 82.5%) = 699]. Therefore, the overall control participation rate in the study was 
58.5 percent (409/699 = 58.5%).
The exact dates that a child received any vaccinations against diphtheria, pertussis, and tetanus (DPT), DT, Td, poliomyelitis, MMR, hepatitis B, or Hib were recorded. Some vaccination records contained only information on month and year for each vaccination, but not the exact day. A day of 15 was assigned to these dates (7% of all vaccination dates). All vaccination data were censored on the reference date (i.e. date of diagnosis for cases and the corresponding date for controls). Vaccinations administered after the reference date were not included. Since it is not well understood how long pre-leukemogenic events may occur before diagnosis, or exactly when leukemogenesis becomes irreversible, we censored the vaccination data on different dates, including 1, 6, and 12 months before the reference date. Symptoms of childhood leukaemia such as fatigue or loss of appetite may lead to the case parents seeking medical advice with a subsequent doctor's visit that could result in vaccinations right before diagnosis. In addition, a parent or physician may decide to postpone a scheduled vaccination because the case child is exhibiting pre-diagnostic symptoms such as fever. Since no meaningful differences in results were observed with the different censoring points, we decided to report results based on data censored only on the reference date.
Statistical analysis
Pearson's 
2 was utilized to compare demographic and socioeconomic characteristics between cases and controls. Conditional logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) as approximations of relative risks, adjusting for maternal education and annual household income, which appeared to differ between cases and controls (the P values for maternal education and household income were 0.064 and 0.002, respectively). To avoid a reduced number of matched sets in adjusted analyses, the middle category for annual household income (30–44.9 thousand dollars) was assigned to 20 subjects (8 cases and 12 controls) whose income data were missing.
In addition to maternal education and household income, we also evaluated adjusting for birth weight and for daycare attendance (The California School Immunization Law requires that children be current on their vaccination status, including poliomyelitis, DTP, MMR, hepatitis B, and Hib, to attend school or child care facilities). However, because the adjustment of neither variable introduced a noticeable difference in risk estimates, these two variables (birth weight and daycare attendance) were dropped from the models. Statistical analyses were carried out separately for total leukaemia (n = 323) and ALL (n = 282). Due to biological evidence that leukaemia occurring in infants (<12 months) may have a distinctive etiology, analyses were also conducted excluding infant leukaemia cases (n = 15). Since analyses including or excluding infant leukaemia generated extremely similar results (which may be due to the small number of infant leukaemia cases), we decided to report results from all cases.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Although controls were matched to the cases born in the study area based upon ethnic information recorded on the birth certificates, the ethnic data shown in Table 2 were obtained from the in-home personal interviews. There was only minimal discrepancy on race and Hispanic status between cases and controls, indicating that the matching process was successful. Control children were from households with a higher annual income than cases. Mother's educational level differed between cases and controls, but the difference did not reach statistical significance (Table 2).
|
Both DPT and poliomyelitis vaccines are recommended to be administered as early as 2 months after birth. In the study population, 99.1% of the cases and 99.0% of the controls received one or more DPT vaccinations prior to the reference dates, and the percentage was 100% for cases and 99.7% for controls who were at least one year old at the reference dates (Table 3). Similar percentages were observed for vaccinations against poliomyelitis. Table 4 shows the risk of childhood leukaemia associated with each type of vaccination, measured as the number of doses before the reference date for overall leukaemia and ALL. DPT vaccination was not associated with the risk of overall leukaemia or ALL, neither was poliomyelitis vaccination. The possible role of receiving three or more doses of DPT vaccines during infancy (<12 months), which is consistent with immunization recommendations in recent years, was also examined, but no significant association was observed (data not shown).
|
|
The proportion of MMR vaccination was slightly lower (92.6 in cases and 93.2% in controls), which is likely due to the older age of recommended initial administration (12 months) (Table 3). For subjects who were at least 2 years old at the reference dates, 98.6% cases and 98.3% controls had received MMR vaccination (data not shown). In this analysis, MMR vaccination, measured as the number of doses, was not associated with the risk of overall leukaemia or ALL (Table 4).
In the study population, 94.7% of the cases and 95.1% of the controls had received Hib vaccination prior to the reference date (Table 3). Hib vaccination was associated with a significantly reduced risk of leukaemia. The OR associated with each dose of Hib vaccine was 0.81 (95% CI 0.68–0.96) for overall leukaemia and similar for ALL (Table 4). Compared with children who had received two or fewer doses of Hib vaccine prior to the reference date, those who had received three or more doses (which are recommended) had a significantly decreased risk of overall leukaemia (OR = 0.55, 95% CI 0.32–0.94) (Table 4). Results pertaining to Hib vaccination hardly changed when the doses of DPT, polio, MMR, and hepatitis B vaccination were adjusted for in addition to maternal education and household income,.
Among the 323 cases, 10.5, 3.7, 7.7, 75.8% and 2.2, had received 0, 1, 2, 3, or 4 doses of hepatitis B vaccines, respectively. The percentage of controls (n = 409) who had received 0, 1, 2, 3, or 4 doses of hepatitis B vaccines was similar (Table 3). Overall, hepatitis B vaccination, measured by the number of doses, was not associated with the risk of childhood leukaemia (Table 4). When the doses of DPT, polio, MMR, and Hib vaccination were adjusted for in addition to maternal education and household income, receiving three or more doses of hepatitis B vaccines during infancy appears to be associated with a significantly increased risk of overall leukaemia (OR = 1.63, 95% CI 1.02–2.59) and ALL (OR = 1.66, 95% CI 1.01–2.72).
When analyses were conducted by year of birth, the protective effect of Hib vaccination was observed among children who were born in or before 1995 but not among children who were born after 1995. Receiving three or more doses of hepatitis B vaccines during infancy appears to be associated with an increased risk of overall leukaemia and ALL among children who were born in or before 1995, but the associations were only of borderline significance. For example, the OR for overall leukaemia was 1.81 (95% CI 0.99–3.30) among children who were born in or before 1995 (Table 5).
|
The total number of administrations of the five types of vaccines (treating DT and Td the same as DPT) was similar in the 323 cases (mean 14.56 with a standard error of 0.17) and 409 controls (mean 14.81 with a standard error of 0.15).
No significant differences in results were observed for vaccination records from different sources (parents vs physicians).
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Since there were very few children who were not vaccinated in the present study, it is not informative to classify vaccination as ‘ever/never’. Rather, the possible etiologic role of vaccinations measured by the number of doses was evaluated. Each dose of Hib vaccination was associated with a significantly reduced risk of childhood leukaemia, while the history of DPT, poliomyelitis, and MMR vaccinations did not differ between cases and controls. Although the OR for having received three or more doses of hepatitis B vaccine during infancy was elevated for children born in or before 1995, it was only of borderline significance. The interpretation of such a finding is unclear given that the number of doses of hepatitis B vaccine before the reference date was not associated with leukaemia risk.
In contrast to several earlier studies,3,4,6,7 results from the present study did not support a protective effect of DTP, poliomyelitis, or MMR vaccinations, but are consistent with three more recent studies.9,12,13 It is possible that the present study had limited power in detecting an effect even if one was present because almost all children received these vaccines and followed very similar schedules.
The finding regarding the protective effect of Hib vaccination is consistent with the result from a recent US study by Groves et al.,9 although the positive association they observed was with presumptive conjugate Hib vaccination measured as ever/never. The study by Groves et al.9 reported that a smaller percentage of ALL cases (47%) than controls (53%) received Hib vaccination. In the present study, the percentage of cases and controls who had been administered Hib vaccine was much higher (94.7 and 95.1%, respectively), which reflects the recent conduct of the NCCLS. The possible mechanism for the observed protective effect is unknown. It is not clear whether the finding could have been due to the vaccine itself or the prevention of related infections, including meningitis, pneumonia, and infections of the blood, bones, and joints. In a Hib vaccination trial conducted in Finland, the incidence of childhood leukaemia was lower in the early vaccination arm (in which several doses of Hib vaccine was administered starting from the age of three months) than the late vaccination arm (in which a single dose of Hib vaccine was administered at the age of 2 years), but the difference did not reach statistical significance.15 More recently, the effect of different formulations of a conjugate Hib vaccine was evaluated in this trial, and incidence of childhood leukaemia was not different between the polysaccharide-diptheria toxoid conjugate arm and the oligosaccharide-CRM197 conjugate arm.16
The present study is one of the first to explore the relationship between hepatitis B vaccination and childhood leukaemia. Petridou et al.12 studied the role of hepatitis B vaccine combined with several other viral vaccines and did not present results pertaining to the independent effect of hepatitis B vaccine. Dockerty et al.13 also assessed the etiologic role of hepatitis B vaccination. The exposure was characterized as ‘yes/no’ (OR = 0.93, 95% CI 0.49–1.76). When we analysed the NCCLS data in a similar way, the OR for hepatitis B vaccination (ever/never) was 0.76 (95% CI 0.36–1.59). Data are available which link impaired responsiveness to neonatal hepatitis B vaccination and human leukocyte antigen types prone to autoimmune conditions.17,18 Certain polymorphisms of the human leukocyte antigen are reported to be associated with the risk of childhood leukaemia.19,20 It may be that vaccinations affect the way the immune system reacts to other infectious agents.
In an earlier evaluation of the potential etiologic role of vaccinations using 167 matched case–control pairs from the NCCLS, we found a significantly increased risk of leukaemia in subjects who had received three or more doses of hepatitis B vaccines (OR = 2.6), especially those who received three or more doses during infancy (OR 
5.0).21 The present analysis includes children from the earlier evaluation (who were enrolled during 1995–1999) and children who were enrolled during 1999–2002. Although we still observed an elevated OR associated with the receipt of three or more doses of hepatitis B vaccine during infancy, the magnitude of association was much smaller (OR = 1.41, Table 4). In addition, the association was not statistically significant. Possible explanations for the apparent discrepancy in our earlier and current observations are at least 2-fold. The small sample size of our earlier evaluation might have given rise to an unstable estimate. In addition, the new cases who had been added to the present analysis were diagnosed and born later than cases included in the earlier evaluation. In the US, universal infant hepatitis B vaccination was first recommended in 1991.22 The average year of birth was 1992 for cases included in the earlier evaluation, and 1995 for cases added to the current analysis. Children who were born after 1992 have a higher chance of receiving three or more doses of hepatitis B vaccine during infancy than children who were born earlier. Variation in the timing of hepatitis B vaccine administration is less likely among children who were born later. When we conducted separate analyses for those who were born in or before 1995 and those born after 1995, receiving three or more doses of hepatitis B vaccine during infancy appeared to be a possible risk factor for the former group (OR = 1.81, 95% CI 0.99–3.30) but not for the latter (OR = 1.05, 95% CI 0.51–2.16) (Table 5). A similar difference by birth cohort was also observed for the relation between receiving three or more doses of Hib vaccine and leukaemia risk. No difference by birth cohort was observed for the relationship between DPT, poliomyelitis, MMR vaccines and childhood leukaemia (Table 5). A cohort effect was only apparent for hepatitis B and Hib vaccinations.
To date, reported findings with respect to vaccinations and the risk of childhood leukaemia have been mixed. The inconsistency may be due, in part, to the variation in exposure assessment. In most of the studies, data on vaccination history were collected through interviews3–6,12,13 or self-administered questionnaires.2 The types of vaccines studied were not specified in several studies.2,6–8 Moreover, in some of the studies,2,4,5,10–12 there was no indication of data censoring to ensure that vaccinations reported occurred before diagnosis (or a corresponding date for controls) and cases and controls had a comparable time window for exposure.
This analysis of data from the NCCLS is distinguished by the reliance on written vaccination records, rather than parental recall, and matching of cases and controls on date of birth and geographic region. Although written vaccination records are not always error free and may not be regarded as the ‘gold standard’, they are probably more reliable than parental recall. Parents in California are required to provide doctor-issued vaccination records to schools, daycares, summer camps, and essentially any other facility that their children may attend, or to provide specific (e.g. religious) reasons for exemption. Otherwise, their children may not be admitted. If a parent believes that a specific vaccination is not recorded, the parent will most likely contact the doctor's office to have the information added to the written records. Under these circumstances, it is uncommon for a parent to have additional information that is not included in written vaccination records. While some previous studies only asked whether a subject had or had not received a specific vaccination, the NCCLS collected data regarding number of doses, as well as date of each vaccination. As shown here, these detailed data are important for a more informative analysis. For example, with exact date of administration for each vaccine, censoring at different time points is possible. The decision to present only data censored on the reference date was made after no meaningful differences were observed for data censored at 1, 6, or 12 months before the reference date. The relationship between Hib vaccination and ALL can serve as an example. If vaccinations were censored on the reference date, the OR associated with each dose of Hib was 0.81 (95% CI 0.68–0.96) (Table 4). If vaccinations were censored at 1, 6, or 12 months prior to the reference date, the OR would be 0.79 (95% CI 0.64–0.98), 0.78 (95% CI 0.63–0.97) and 0.77 (95% CI 0.61–0.97), respectively. The matching on date of birth is important since vaccination practice in the US is largely determined by recommendations made by the Advisory Committee on Vaccination Practices and other authorities, and these recommendations vary over time.
Lastly and importantly, the timing of the NCCLS made it possible to evaluate the potential etiologic role of hepatitis B and Hib vaccination. In the US, universal infant hepatitis B vaccination was first recommended in 1991.22 The first Hib vaccine was licensed in 1985, but it was not administered on a large scale until several years later. Studies conducted in the US of children born before 1990 may not have sufficient power to address the potential effect of hepatitis B and Hib vaccinations simply because there were very few, if any, who had received either vaccine. Similarly, studies conducted in the future may have difficulty assessing the role of hepatitis B and Hib vaccines as most children would have received these vaccines following a similar schedule. Receipt of these vaccines appears to be a unique time-sensitive exposure, and the timing of NCCLS coincides with the window of opportunity to study the potential etiologic role of hepatitis B and Hib vaccinations. This is where Table 5 assumes importance.
While the present study has a moderate sample size, the matched design and detailed exposure assessment serve to increase the power of the statistical analyses.23 In addition, complete written vaccination records were not available for 8.3% of subjects, a percentage considerably lower than what was reported in a recent US study (>20%).9 To further address the issue of missing data, we tried to keep the subjects without written records in the analysis by coding vaccination exposures as three-category variables: no, yes, and unknown, and then we estimated the ORs for having the exposure and for the unknown. The ORs for the unknown did not differ substantially from the ORs for having the exposure.
We could not distinguish between different formulations of Hib vaccines, which is a limitation of the study. However, cases and controls were from similar geographic regions and matched on date of birth, i.e. age and calendar time simultaneously. Therefore, cases and their matched controls were likely subjected to the same formulation of Hib vaccines and similar Hib immunization recommendations. Another limitation of the NCCLS is the lack of data on family history of hepatitis B infection or status of maternal hepatitis B surface antigen (HBsAg), which may influence vaccination-seeking behavior. The NCCLS is currently ongoing, and we have added to the protocol a question regarding the maternal HBsAg status at the time of index child's birth.
The lack of information on why some children had received fewer than the recommended number of doses for various vaccines is also of concern. The first dose of hepatitis B vaccine (the only vaccine recommended to be given to newborns), when administered in the hospital right after birth as opposed to being administered in the office of a primary care pediatrician, may not have been recorded consistently for all the subjects. However, it is unlikely that underreporting would have affected cases and controls in systematically different ways. In general, some subjects could have stopped or postponed receiving certain vaccines after experiencing allergies and/or other adverse effects. If this experience was more common in one group than in the other, it would bias the results.
In the present study, the socioeconomic differences between cases and controls, as reflected by differences in annual household income and maternal education level, were of concern. Although both variables were adjusted for in the analyses, there might have been residual confounding. We conducted an analysis of matched cases and controls with the same annual household income (about a third of the overall study population). In this analysis, the results with regard to vaccinations persisted, which offered assurance that the findings are not completely due to socioeconomic differences between cases and controls.
In the present study, Hib vaccination was associated with a reduced risk of childhood leukaemia, although the underlying mechanism is unclear. A possible role of the receipt of three or more doses of hepatitis B vaccine during infancy is also suggested, but the data available are far from conclusive. Considering the limited understanding of the etiologic role of vaccinations, future studies with both detailed exposure assessment and larger sample sizes are needed. This topic will be pursued further in the NCCLS. Furthermore, exploration of the relationship between these two vaccines and childhood leukaemia may be pursued by using the Vaccine Safety Data Link maintained by the Centers for Disease Control and Prevention, and in large-scale vaccine efficacy trials conducted in other countries, provided that outcomes such as childhood leukaemia are recorded in a reliable manner.
KEY MESSAGES
|
|
Acknowledgments |
|---|
The study was supported by grants from the National Institute of Environmental Health Sciences (R01 ES09137 and PS42 ES04705), USA. We thank clinical collaborators at the participating hospitals for help with recruiting patients, and Dr Silvia Franceschi of the International Agency for Research on Cancer for valuable comments. Dr Xiaomei Ma was supported in part by Hope Street Kids, a non-profit organization based in Virginia, USA.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
1 Innis MD. Immunisation and childhood leukaemia. Lancet 1965;1:605.[Medline]
2 Buckley J, Buckley C, Ruccione K et al. Epidemiological characteristics of childhood acute lymphocytic leukemia. Analysis by immunophenotype. Leukemia 1994;8:856–64.[Web of Science][Medline]
3 Kneale GW, Stewart AM, Wilson LM. Immunizations against infectious diseases and childhood cancers. Cancer Immunol Immunother 1986;21:129–32.[Web of Science][Medline]
4 McKinney PA, Cartwright RA, Saiu JM et al. The inter-regional epidemiological study of childhood cancer (IRESCC): a case control study of aetiological factors in leukaemia and lymphoma Arch Dis Child 1987;62:279–87.
5 Hartley AL, Birch JM, McKinney PA et al. The Inter-Regional Epidemiological Study of Childhood Cancer (IRESCC): past medical history in children with cancer. J Epidemiol Community Health 1988;42:235–42.
6 Nishi M, Miyake H. A case-control study of non-T cell acute lymphoblastic leukaemia of children in Hokkaido, Japan. J Epidmiol Community Health 1989;43:352–55.
7 Kaatsch P, Kaletsch U, Krummenauer F et al. Case control study on childhood leukemia in Lower Saxony, Germany. Klin Pädiatr Basic considerations, methodology, and summary of results. 1996;208:79–85.
8 Kaatsch P, Kaletsch U, Meinert R et al. German case control study on childhood leukaemia—basic considerations, methodology and summary of the results. Klin Padiatr 1998;210:185–91.[Web of Science][Medline]
9 Groves FD, Gridley G, Wacholder S et al. Infant vaccinations and risk of childhood acute lymphoblastic leukaemia in the USA. Br J Cancer 1999;81:175–78.[CrossRef][Web of Science][Medline]
10 Schüz J, Kaletsch U, Meinert R, Kaatsch P, Michaelis J. Association of childhood leukaemia with factors related to the immune system. Br J Cancer 1999;80:585–90.[CrossRef][Web of Science][Medline]
11 Salonen T, Saxen L. Risk indicators in childhood malignancies. Int J Cancer 1975;15:941–46.[Web of Science][Medline]
12 Petridou E, Trichopoulos D, Kalapothaki V et al. The risk profile of childhood leukaemia in Greece: a nationwide case–control study. Br J Cancer 1997;76:1241–47.[Web of Science][Medline]
13 Dockerty JD, Skegg DC, Elwood JM, Herbison GP, Becroft DM, Lewis ME. Infections, vaccinations, and the risk of childhood leukaemia. Br J Cancer 1999;80:1483–89.[CrossRef][Web of Science][Medline]
14 Ma X, Buffler PA, Layefsky M, Does MB, Reynolds P. Control selection strategies in case–control studies of childhood diseases. Am J Epidemiol 2004;159:915–21.
15 Auvinen A, Hakulinen T, Groves F. Haemophilus influenzae type B vaccination and risk of childhood leukaemia in a vaccine trial in Finland. Br J Cancer 2000;83:956–58.[CrossRef][Web of Science][Medline]
16 Groves F, Sinha D, Auvinen A. Haemophilus influenzae type b vaccine formulation and risk of childhood leukaemia. Br J Cancer 2002;87:511–12.[CrossRef][Web of Science][Medline]
17 Martinetti M, Cuccia M, Daielli C et al. Anti-HBV neonatal immunization with recombinant vaccine. Part II. Molecular basis of the impaired alloreactivity. Vaccine 1995;13:555–60.[CrossRef][Web of Science][Medline]
18 Martinetti M, De Silvestri A, Belloni C et al. Humoral response to recombinant hepatitis B virus vaccine at birth: role of HLA and beyond. Clin Immunol 2000;97:234–40.[CrossRef][Web of Science][Medline]
19 Taylor G, Dearden S, Payne N et al. Evidence that an HLA-DQA1-DQB1 haplotype influences susceptibility to childhood common acute lymphoblastic leukaemia in boys provides further support for an infection-related aetiology. Br J Cancer 1998;78:561–65.[Web of Science][Medline]
20 Taylor GM, Dearden S, Ravetto P et al. Genetic susceptibility to childhood common acute lymphoblastic leukaemia is associated with polymorphic peptide-binding pocket profiles in HLA-DPB1*0201. Hum Mol Genet 2002;11:1585–97.
21 Ma X, Does M, Buffler PA, Wiencke JK. Hepatitis B vaccination and the risk of childhood leukemia. Poster presented at the 93rd Annual Conference of the American Association for Cancer Research, 2002, San Francisco, CA, USA.
22 Hepatitis B virus: a comprehensive strategy for eliminating transmission in the United States through universal childhood vaccination. Recommendations of the Immunization Practices Advisory Committee (ACIP). MMWR Recomm Rep 1991;40:1–25.[Medline]
23 Cochran W. Planning and Analysis of Observational Studies. New York: John Wiley and Sons, 1983.
24 Little J. Epidemiology of Childhood Cancer. Lyon, France: International Agency for Research on Cancer, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E Roman, J Simpson, P Ansell, S Kinsey, C. Mitchell, P. McKinney, J. Birch, M Greaves, T Eden, and on behalf of the United Kingdom Childhood Cancer S Childhood Acute Lymphoblastic Leukemia and Infections in the First Year of Life: A Report from the United Kingdom Childhood Cancer Study Am. J. Epidemiol., March 1, 2007; 165(5): 496 - 504. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. H. Rathore No Association Between Childhood Vaccines and Risk of Leukemia AAP Grand Rounds, January 1, 2006; 15(1): 3 - 4. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||