An infectious aetiology for childhood brain tumours? Evidence from space-time clustering and seasonality analyses

Clustering of Pediatric Brain Tumors in Texas, 2000-2017

Risk factors for pediatric brain tumors are largely unknown. Identifying spatial clusters of these rare tumors on the basis of residential address may provide insights into childhood socio-environmental factors that increase susceptibility. From 2000-2017, the Texas Cancer Registry recorded 4305 primary brain tumors diagnosed among children (≤19 years old). We performed a spatial analysis in SaTScan to identify neighborhoods (census tracts) where the observed number of pediatric brain tumors was higher than expected. Within each census tract, the number of pediatric brain tumors was summed on the basis of residential address at diagnosis. The population estimate from the 2007-2011 American Community Survey of 0- to 19-year-olds was used as the at-risk population. p-values were calculated using Monte Carlo hypothesis testing. The age-standardized rate was 54.3 per 1,000,000. SaTScan identified twenty clusters, of which two were statistically significant (p < 0.05). Some of the clusters identified in Texas spatially implicated potential sources of environmental risk factors (e.g., proximity to petroleum production processes) to explore in future research. This work provides hypothesis-generating data for further investigations of spatially relevant risk factors of pediatric brain tumors in Texas.

Cadmium (Cd) and Lead (Pb) topsoil levels and incidence of childhood leukemias

There are few well-established risk factors for childhood leukemias. While the frequency of childhood leukemias might be partially attributable to some diseases (accounting for a small fraction of cases) or ionizing radiation, the role of heavy metals has not been assessed. The objective of our study was to assess the potential association between levels of cadmium (Cd) and lead (PB) in soil and childhood leukemias incidence. We conducted a population-based case-control study of childhood leukaemia in Spain, covering 2897 incident cases gathered from the Spanish Registry of Childhood Tumours and including 14 Spanish Regions with a total population of 5,307,433 children (period 1996-2015). Cd and Pb bioavailable levels at every children's home address were estimated using data from the Geochemical Atlas of Spain. We used logistic regression to estimate odds ratios (ORs) and their 95% confidence intervals (95%CIs); we included as covariates: sex, rurality, employment rate and socioeconomic status. Metal levels were analysed according to two definitions: as continuous variable assuming linearity and as categorical variables to explore a potentially nonlinear association (quantiles). Increases in both Cd and Pb topsoil levels were associated with increased probability of childhood leukemias incidence. The results for the models with the continuous variables showed that a unit increase on the topsoil level was associated with an OR of 1.11 for Cd (95%CI 1.00-1.24) and an OR of 1.10 for Pb (95%CI 0.99-1.21). Our study may point towards a possible link between residential Cd and Pb topsoil levels and the probability of childhood leukemias incidence. Residing in a location with the highest concentrations of these heavy metals compared to those locations with the lowest could increase the risk around a 20%, for both Cd and Pb.

International trends in the incidence of brain tumours in children and young-adults and their association with indicators of economic development

INTRODUCTION

Childhood brain tumours (CBTs) are the second most common type of cancer in individuals aged 0-24 years globally and cause significant morbidity and mortality. CBT aetiology remains poorly understood, however previous studies found higher CBT incidence in high-income countries (HIC) compared to low-middle income countries (LMIC), suggesting a positive relationship between incidence and wealth.

MATERIALS & METHODS

Aggregated data from Cancer Incidence in Five Continents (CI5) were used to explore CBT epidemiology. Incidence rate ratios (IRR) compared CBT rates between twenty-five geographically and economically diverse countries. The relationship between incidence and economic development was explored using linear regression models and Spearman's rank correlation tests. Trends in CBT incidence between 1978 and 2012 were investigated using average annual percentage changes (AAPC).

RESULTS

CBT incidence was highest in North America and lowest in Africa. CBT incidence rates increased significantly with increasing GDP per capita (p = 0.006). Gini index was significantly negatively associated with CBT incidence. Incidence decreased with increasing income inequality within countries, indicated by higher Gini indices (p = 0.040). Increasing and decreasing CBT incidence trends were observed within individual countries, although only Italy (p = 0.02) and New Zealand (p < 0.005) experienced statistically significant changes over time.

CONCLUSIONS

The excess disease found in HIC may be explained by environmental risk factor exposure increasing CBT risk in wealthy populations. However, systematic limitations of substandard cancer detection and reporting in LMIC may mean incidence disparities result from misinformation bias rather than genuine differences in risk factor exposure. Further research is required to comprehensively describe CBT epidemiology and explain study findings.

ClustR: A Space-Time Cluster Analysis R Package for Individual-level Data

BACKGROUND

Until recently, large individual-level longitudinal data were unavailable to investigate clusters of disease, driving a need for suitable statistical tools. We introduce a robust, efficient, intuitive R package, ClustR, for space-time cluster analysis of individual-level data.

METHODS

We developed ClustR and evaluated the tool using a simulated dataset mirroring the population of California with constructed clusters. We assessed Cluster's performance under various conditions and compared it with another space-time clustering algorithm: SaTScan.

RESULTS

ClustR mostly exhibited high sensitivity for urban clusters and low sensitivity for rural clusters. Specificity was generally high. Compared with SaTScan, ClustR ran faster and demonstrated similar sensitivity, but had lower specificity. Select cluster types were detected better by ClustR than SaTScan and vice versa.

CONCLUSION

ClustR is a user-friendly, publicly available tool designed to perform efficient cluster analysis on individual-level data, filling a gap among current tools. ClustR and SaTScan exhibited different strengths and may be useful in conjunction.

High parental occupational social contact and risk of childhood hematopoietic, brain and bone cancers

BACKGROUND

The etiology of childhood cancer is largely unknown, though some research suggests an infectious origin of hematopoietic, central nervous system (CNS) and bone cancers.

METHODS

We examined parental occupational social contact as a proxy for exposure to infectious agents and risk of childhood cancer. This population-based case-control study utilized a linkage of four Danish data-registries, and included 3581 cases (<17 years, diagnosed 1973-2012) and 358,100 age-matched controls. We examined the risks of leukemia, lymphoma, CNS and bone cancer related to high occupational social contact from (1) conception to birth and (2) birth to diagnosis.

RESULTS

Acute lymphoblastic leukemia (ALL) and bone cancer were inversely associated with high maternal social contact from conception to birth (OR: 0.86, 95% CI: 0.67-1.10) and birth to diagnosis (OR: 0.54, 95% CI: 0.34-0.86). Children of fathers with high social contact from birth to diagnosis had an increased risk of bone cancers, particularly in rural areas (OR: 1.65, 95% CI: 1.03-2.63). Parental social contact was associated with increased risk of astrocytoma, with strongest associations found in first-born children (maternal: OR: 1.54, 95% CI: 1.02-2.32; paternal: OR: 1.82, 95% CI: 1.05-3.17).

CONCLUSION

Our results support the notion of a role of infections for some cancer types.

Space-time clustering of childhood cancers: a systematic review and pooled analysis

The aetiology of childhood cancers remains largely unknown. Space–time clustering of cases might imply an aetiological role of infections. We aimed to review the evidence of space–time clustering of specific childhood cancers. We searched Medline and Embase for population-based studies that covered a pre-defined study area, included cases under 20 years of age and were published before July 2016. We extracted all space–time clustering tests and calculated the proportion of positive tests per diagnostic group. In a pooled analysis, we performed a Knox test of the number of pairs of cases close to each other in time and space pooled across studies. 70 studies met our eligibility criteria, 32 of which reported Knox tests. For leukaemia, the proportion of positive tests was higher than expected by chance at both time of diagnosis (26%) and birth (11%). The pooled analysis showed strong evidence of clustering at diagnosis for children aged 0–5 years for a spatial and temporal lag of 5 km and 6 months, respectively (p < 0.001). The evidence was mixed for lymphoma and tumours of the central nervous system. The current study suggests that leukaemia cases cluster in space–time due to an aetiological factor affecting children under 5 years of age. The observed pattern of clustering of young children close to time of diagnosis is compatible with Greaves’ delayed-infections-hypothesis.

Childhood cancer in small geographical areas and proximity to air-polluting industries

AIM

Pediatric cancer has been associated with exposure to certain environmental carcinogens. The purpose of this work is to analyse the relationship between environmental pollution and pediatric cancer risk.

METHOD

We analysed all incidences of pediatric cancer (<15) diagnosed in a Spanish region during the period 1998-2015. The place of residence of each patient and the exact geographical coordinates of main industrial facilities was codified in order to analyse the spatial distribution of cases of cancer in relation to industrial areas. Focal tests and focused Scan methodology were used for the identification of high-incidence-rate spatial clusters around the main industrial pollution foci.

RESULTS

The crude rate for the period was 148.0 cases per 1,000,0000 children. The incidence of pediatric cancer increased significantly along the period of study. With respect to spatial distribution, results showed significant high incidence around some industrial pollution foci group and the Scan methodology identify spatial clustering. We observe a global major incidence of non Hodgkin lymphomas (NHL) considering all foci, and high incidence of Sympathetic Nervous System Tumour (SNST) around Energy and Electric and organic and inorganic chemical industries foci group. In the analysis foci to foci, the focused Scan test identifies several significant spatial clusters. Particularly, three significant clusters were identified: the first of SNST was around energy-generating chemical industries (2 cases versus the expected 0.26), another of NHL was around residue-valorisation plants (5 cases versus the expected 0.91) and finally one cluster of Hodgkin lymphoma around building materials (3 cases versus the expected 2.2) CONCLUSION: Results suggest a possible association between proximity to certain industries and pediatric cancer risk. More evidences are necessary before establishing the relationship between industrial pollution and pediatric cancer incidence.

Perinatal light imprinting of circadian clocks and systems (PLICCS): A signature of photoperiod around birth on circadian system stability and association with cancer

Recent findings from animal models suggest that plasticity of human circadian clocks and systems may be differentially affected by different paradigms of perinatal photoperiod exposure to the detriment of health in later life, including cancer development. Focusing on the example of cancer, we carry out a series of systematic literature reviews concerning perinatal light imprinting of circadian clocks and systems (PLICCS) in animal models, and concerning the risk of cancer development with the primary determinants of the perinatal photoperiod, namely season of birth or latitude of birth. The results from these systematic reviews provide supporting evidence of the PLICCS and cancer rationale and highlight that investigations of PLICCS in humans are warranted. Overall, we discuss findings from experimental research and insights from epidemiological studies. Considerations as to how to "test" PLICCS in epidemiological studies and as to the potential for non-invasive preventative measures during perinatal periods close our synthesis. If the PLICCS rationale holds true, it opens the exciting prospect for amenable, early-life, preventative measures against cancer development (and other disorders) in later life. Indeed, non-invasive anthropogenic light exposure may have enormous potential to alleviate the public health and economic burden of circadian-related diseases.

Space-time clustering of childhood cancers in Switzerland: A nationwide study

The aetiology of childhood cancers remains largely unknown. It has been hypothesized that infections may be involved and that mini-epidemics thereof could result in space-time clustering of incident cases. Most previous studies support spatio-temporal clustering for leukaemia, while results for other diagnostic groups remain mixed. Few studies have corrected for uneven regional population shifts which can lead to spurious detection of clustering. We examined whether there is space-time clustering of childhood cancers in Switzerland identifying cases diagnosed at age <16 years between 1985 and 2010 from the Swiss Childhood Cancer Registry. Knox tests were performed on geocoded residence at birth and diagnosis separately for leukaemia, acute lymphoid leukaemia (ALL), lymphomas, tumours of the central nervous system, neuroblastomas and soft tissue sarcomas. We used Baker's Max statistic to correct for multiple testing and randomly sampled time-, sex- and age-matched controls from the resident population to correct for uneven regional population shifts. We observed space-time clustering of childhood leukaemia at birth (Baker's Max p = 0.045) but not at diagnosis (p = 0.98). Clustering was strongest for a spatial lag of <1 km and a temporal lag of <2 years (Observed/expected close pairs: 124/98; p Knox test = 0.003). A similar clustering pattern was observed for ALL though overall evidence was weaker (Baker's Max p = 0.13). Little evidence of clustering was found for other diagnostic groups (p > 0.2). Our study suggests that childhood leukaemia tends to cluster in space-time due to an etiologic factor present in early life.

Spatial analysis of childhood cancer: a case/control study

Background

Childhood cancer was the leading cause of death among children aged 1-14 years for 2012 in Spain. Leukemia has the highest incidence, followed by tumors of the central nervous system (CNS) and lymphomas (Hodgkin lymphoma, HL, and Non-Hodgkin’s lymphoma, NHL). Spatial distribution of childhood cancer cases has been under concern with the aim of identifying potential risk factors.

Objective

The two objectives are to study overall spatial clustering and cluster detection of cases of the three main childhood cancer causes, looking to increase etiological knowledge.

Methods

We ran a case-control study. The cases were children aged 0 to 14 diagnosed with leukemia, lymphomas (HL and NHL) or CNS neoplasm in five Spanish regions for the period 1996-2011. As a control group, we used a sample from the Birth Registry matching every case by year of birth, autonomous region of residence and sex with six controls. We geocoded and validated the address of the cases and controls. For our two objectives we used two different methodologies. For the first, for overall spatial clustering detection, we used the differences of K functions from the spatial point patterns perspective proposed by Diggle and Chetwynd and the second, for cluster detection, we used the spatial scan statistic proposed by Kulldorff with a level for statistical significance of 0.05.

Results

We had 1062 cases of leukemia, 714 cases of CNS, 92 of HL and 246 of NHL. Accordingly we had 6 times the number of controls, 6372 controls for leukemia, 4284 controls for CNS, 552 controls for HL and 1476 controls for NHL. We found variations in the estimated empirical D(s) for the different regions and cancers, including some overall spatial clustering for specific regions and distances. We did not find statistically significant clusters.

Conclusions

The variations in the estimated empirical D(s) for the different regions and cancers could be partially explained by the differences in the spatial distribution of the population; however, according to the literature, we cannot discard environmental hazards or infections agents in the etiology of these cancers.

Population mixing for leukaemia, lymphoma and CNS tumours in teenagers and young adults in England, 1996-2005

Background

Little aetiological epidemiological research has been undertaken for major cancers occurring in teenagers and young adults (TYA). Population mixing, as a possible proxy for infectious exposure, has been well researched for childhood malignancies. We aimed to investigate effects of population mixing in this older age group using an English national cancer dataset.

Methods

Cases of leukaemia, lymphoma and central nervous system (CNS) tumours amongst 15–24 year olds in England (diagnosed 1996–2005) were included in the study. Data were obtained by ward of diagnosis and linked to 1991 census variables including population mixing (Shannon index); data on person-weighted population density and deprivation (Townsend score) were also used and considered as explanatory variables. Associations between TYA cancer incidence and census variables were investigated using negative binomial regression, and results presented as incidence rate ratios (IRR) with 95% confidence intervals (CI).

Results

A total of 6251 cases of leukaemia (21%), lymphoma (49%) and CNS tumours (30%) were analysed. Higher levels of population mixing were associated with a significant decrease in the incidence of CNS tumours (IRR = 0.83, 95% CI = 0.75-0.91), accounted for by astrocytomas and ‘other CNS tumours’; however, there was no association with leukaemia or lymphoma. Incidence of CNS tumours and lymphoma was 3% lower in more deprived areas (IRR = 0.97, 95% CI = 0.96-0.99 and IRR = 0.97, 95% CI =0.96-0.98 respectively). Population density was not associated with the incidence of leukaemia, lymphoma or CNS tumours.

Conclusions

Our results suggest a possible role for environmental risk factors with population correlates in the aetiology of CNS tumours amongst TYAs. Unlike studies of childhood cancer, associations between population mixing and the incidence of leukaemia and lymphoma were not observed.

UK case control study of brain tumours in children, teenagers and young adults: a pilot study

BACKGROUND

Tumours of the central nervous system are the second most common group of childhood cancers in 0-14 year olds (24% of total cancers) and represent a major diagnostic group in 15-24 year olds. The pilot case-control study aimed to establish methodologies for a future comprehensive aetiological investigation among children and young adults.

METHODS

Eligible cases were newly diagnosed with an intracranial tumour of neuroepithelial tissue aged 0-24 years. The pilot recruited patients through Leeds and Manchester Principal Treatment Centres. Controls were drawn from general practice lists. Controls were frequency matched by age and gender.

RESULTS

We interviewed 49 cases and 78 controls comprising 85% of the target sample size. Response rates were 52% for cases and 32% for controls. Completion of the questionnaire was successful, with a very small proportion of missing data being reported (5-10%). The age distribution of cases and controls was similar with around three-quarters of interviewed subjects aged 0-14. Half of cases and almost two-thirds of controls reported using a mobile phone with the majority starting between 10-14 years of age. Prevalence of breastfeeding was lower in cases than controls (Odds Ratio 0.4; 95% CI 0.2-1.2), whilst cases were more likely to be delivered by caesarean section (OR 1.6; 95% CI 0.6-4.4). Cases were significantly more likely to have a birthweight > 3.5 kg compared to controls. Cases were also more likely to come from a family with 3 or more siblings than controls (OR 3.0; 95% CI 0.7-13.6). The majority of participants (>80%) were in favour of taking either blood or saliva to aid molecular epidemiological research.

CONCLUSIONS

Successful methods were established for identifying and recruiting a high proportion of case subjects, exploiting strong links with the clinical teams at the treatment centres. Control procedures proved more difficult to implement. However, working closely with national clinical and professional research networks will enable improved control identification and recruitment, with good prospects for collecting biological samples in the future.

Cross-space-time clustering of childhood cancer in Great Britain: evidence for a common aetiology

Previously, we identified space-time clustering in certain childhood cancers. This study aimed to determine whether there was cross-space-time clustering between different diagnostic groups. A total of 32,295 cases were diagnosed during 1969–1993. Cross-space-time clustering was analyzed by a second-order procedure based on Diggle's method. Locations were birth and diagnosis addresses. The following space-time combinations were examined: address and date of birth; address at birth and date of diagnosis; address and date of diagnosis. Cross-space-time clustering analyses considered clustering pairs of cases from two different diagnostic groups. Formal statistical significance was taken as p < 0.00067 and marginal significance 0.01 > p ≥ 0.00067. Based on address at birth and date of diagnosis, there was statistically significant cross-clustering between cases of HL and intracranial and intraspinal embryonal tumors (IIET), both aged 0–14 years (p < 0.0001). Based on address and date of birth, there was marginally significant cross-clustering between cases of lymphoid leukemia (LL) aged 5-14 years and Hodgkin lymphoma (HL) aged 0–14 years (p = 0.0019). Based on address and date of diagnosis there was marginally significant cross-clustering between cases of LL aged 1–4 years and soft tissue sarcoma (STS) aged 0–14 years (p = 0.0041). Findings from this study are consistent with possible common aetiological factors between different diagnostic groups. They suggest a common aetiology for the following pairs of diagnostic groups: HL and IIET; older cases of LL and HL; younger cases of LL and STS. The possibility of common infectious mechanisms should be explored.

Case-control study of paternal occupation and social class with risk of childhood central nervous system tumours in Great Britain, 1962-2006

BACKGROUND

Paternal occupational exposures have been proposed as a risk factor for childhood central nervous system (CNS) tumours. This study investigates possible associations between paternal occupational exposure and childhood CNS tumours in Great Britain.

METHODS

The National Registry of Childhood Tumours provided all cases of childhood CNS tumours born and diagnosed in Great Britain from 1962 to 2006. Controls without cancer were matched on sex, period of birth and birth registration sub-district. Fathers' occupations were assigned to one or more of 33 exposure groups. A measure of social class was also derived from father's occupation at the time of the child's birth.

RESULTS

Of 11 119 cases of CNS tumours, 5 722 (51%) were astrocytomas or other gliomas, 2 286 (21%) were embryonal and 985 (9%) were ependymomas. There was an increased risk for CNS tumours overall with exposure to animals, odds ratio (OR) 1.40 (95% confidence intervals (CIs) 1.01, 1.94) and, after adjustment for occupational social class (OSC), with exposure to lead, OR 1.18 (1.01, 1.39). Exposure to metal-working oil mists was associated with reduced risk of CNS tumours, both before and after adjustment for OSC, OR 0.87 (0.75, 0.99).Risk of ependymomas was raised for exposure to solvents, OR 1.73 (1.02,2.92). For astrocytomas and other gliomas, risk was raised with high social contact, although this was only statistically significant before adjustment for OSC, OR 1.15 (1.01,1.31). Exposure to paints and metals appeared to reduce the risk of astrocytomas and embryonal tumours, respectively. However, as these results were the result of a number of statistical tests, it is possible they were generated by chance.Higher social class was a risk factor for all CNS tumours, OR 0.97 (0.95, 0.99). This was driven by increased risk for higher social classes within the major subtype astrocytoma, OR 0.95 (0.91, 0.98).

CONCLUSION

Our results provide little evidence that paternal occupation is a significant risk factor for childhood CNS tumours, either overall or for specific subtypes. However, these analyses suggest that OSC of the father may be associated with risk of some childhood CNS cancers.

Space-time clustering of childhood central nervous system tumours in Yorkshire, UK

BACKGROUND

We specifically tested the aetiological hypothesis that a factor influencing geographical or temporal heterogeneity of childhood central nervous system (CNS) tumour incidence was related to exposure to a transient environmental agent.

METHODS

Information was extracted on individuals aged 0-14 years, diagnosed with a CNS tumour between the 1st January 1974 and 31st December 2006 from the Yorkshire Specialist Register of Cancer in Children and Young People. Ordnance Survey eight-digit grid references were allocated to each case with respect to addresses at the time of birth and the time of diagnosis, locating each address to within 0.1 km. The following diagnostic groups were specified a priori for analysis: ependymoma; astrocytoma; primitive neuroectodermal tumours (PNETs); other gliomas; total CNS tumours. We applied the K-function method for testing global space-time clustering using fixed geographical distance thresholds. Tests were repeated using variable nearest neighbour (NN) thresholds.

RESULTS

There was statistically significant global space-time clustering for PNETs only, based on time and place of diagnosis (P = 0.03 and 0.01 using the fixed geographical distance and the variable NN threshold versions of the K-function method respectively).

CONCLUSIONS

There was some evidence for a transient environmental component to the aetiology of PNETs. However, a possible role for chance cannot be excluded.

Space-time clustering in childhood nervous system tumors in the Region of Murcia, Spain, 1998-2009

AIMS

The aims of this study are the following: first, to analyze incidence, trends, and survival of nervous system tumors in children under the age of 15 in the Region of Murcia, Spain, during the years 1998-2009 and second, to evaluate if certain environmental exposures may be involved in the etiology of childhood nervous system tumors. The study was performed on the spatial and temporo-spatial distribution of the observed cases.

METHODS

The Environment and Pediatric Cancer in the Region of Murcia is an ongoing research project aimed at carefully collecting pediatric environmental history (PEH) and to use geographical information systems to map the incidence and to analyze the geographical distribution of pediatric cancer incidence in our region. Between 1998 and 2009, 125 patients were diagnosed with nervous system tumors. The spatial and temporal space clusters were evaluated using Kulldorff's statistics. Address at diagnosis was the main feature evaluated.

RESULTS

The incidence (cases/million children) for central nervous system (CNS) tumors was 34.2, that for sympathetic nervous system tumors was 10.9, and that for retinoblastoma was 1.9. There was evidence of space clustering for medulloblastoma and space-time clustering for all tumors, CNS tumors, astrocytoma, and neuroblastoma.

CONCLUSIONS

The incidence and survival for each type and subtype of nervous system tumors were within the reported values for the European region. There is evidence that spatial and spatial-temporal distribution in these cases is not random. The development of a careful PEH in these patients will help to reinforce geographical information system studies and to ascertain the importance of associated risk factors.

Risk of primary childhood brain tumors related to season of birth in Kumamoto Prefecture, Japan

OBJECTIVE

Environmental factors present around the time of birth may induce the development of childhood cancer. Some studies suggested an excess of brain tumors in children born during the winter months. The aim of this study was to look for evidence of the seasonality of birth in children who were younger than 15 years at the time of brain tumor diagnosis in the Kumamoto Prefecture, Japan.

METHODS

We surveyed 115 patients younger than 15 years who were diagnosed with primary intracranial tumors. All patients were born between 1989 and 2003.

RESULTS

We found a statistically significant difference between the season of their birth and the expected distribution of birth dates in the Kumamoto Prefecture (p = 0.028). Among the different diagnostic groups there was a statistically significant winter peak in the birth of patients with germ cell tumor (p = 0.001). No statistically significant seasonal patterns were detected in the birth season of patients with astrocytoma, malignant glioma, and medulloblastoma.

CONCLUSIONS

Although our data provide modest support for a winter peak in the birth of children with brain tumors, we posit that there may be yet unknown, complex biological mechanisms that account for these putative seasonal patterns.

Knox meets Cox: adapting epidemiological space-time statistics to demographic studies

Many important questions and theories in demography focus on changes over time, and on how those changes differ over geographic and social space. Space-time analysis has always been important in studying fertility transitions, for example. However demographers have seldom used formal statistical methods to describe and analyze time series of maps. One formal method, used widely in epidemiology, criminology, and public health, is Knox 's space-time interaction test. In this article, we discuss the potential of the Knox test in demographic research and note some possible pitfalls. We demonstrate how to use familiar proportional hazards models to adapt the Knox test for demographic applications. These adaptations allow for nonrepeatable events and for the incorporation of structural variables that change in space and time. We apply the modified test to data on the onset offertility decline in Brazil over 1960-2000 and show how the modified method can produce maps indicating where and when diffusion effects seem strongest, net of covariate effects.

Infectious exposure in the first years of life and risk of central nervous system tumours in children: analysis of birth order, childcare attendance and seasonality of birth

Background:

An infective, mostly viral basis has been found in different human cancers. To test the hypothesis of a possible infectious aetiology for central nervous system (CNS) tumours in children, we investigated the associations with proxy measures of exposure to infectious disease.

Methods:

In a large case–control study nested in the populations of Denmark, Norway, Sweden, and Finland of 4.4 million children, we studied the association of birth order and seasonal variation of birth with subsequent risk for CNS tumours. We identified 3983 children from the national cancer registries, and information on exposure was obtained from the high-quality national administrative health registries. We investigated the association between childcare attendance during the first 2 years of life and the risk for CNS tumours in a subset of Danish children with CNS tumours, using information from the Danish Childcare database.

Results:

We observed no association between birth order and risk of CNS tumours overall (odds ratio (OR) for second born or later born vs first born, 1.03; 95% confidence interval (CI), 0.96–1.10) or by histological subgroup, and children with CNS tumours did not show a seasonal variation of birth that was distinct from that of the background population. Childcare attendance compared with homecare showed a slightly increased OR (1.29; 95% CI, 0.90–1.86) for CNS tumours, with the highest risk observed in children attending a crèche. The strongest association was observed for embryonal CNS tumours. We found no effect of age at enrolment or duration of enrolment in childcare.

Conclusion:

These results do not support the hypothesis that the burden of exposure to infectious disease in early childhood has an important role in the aetiology of paediatric CNS tumours.

Season of birth and diagnosis for childhood cancer in Northern England, 1968-2005

The aim of this study was to investigate seasonal variation in the incidence of cancer in children aged 0-14 years. Details of 2959 primary malignant cases (1659 males, 1300 females), diagnosed during the period 1968-2005, were extracted from a specialist registry (the Northern Region Young Persons' Malignant Disease Registry). Seasonal variation was analysed with respect to month of birth and diagnosis. The chi-squared heterogeneity test was used to test for non-uniform variation. Poisson regression analysis was used to fit sinusoidal (harmonic) models to the data, using month of birth and month of diagnosis, respectively, as covariates in separate models. There was significant sinusoidal variation based on month of birth for acute lymphoblastic leukaemia (ALL) aged 1-6 years (P = 0.04; peak in March). For 0- to 14-year-old boys, there was statistically significant sinusoidal variation in month of birth for acute non-lymphocytic leukaemia (P = 0.04; peak in September) and astrocytoma (P = 0.03; peak in October). Based on month of diagnosis, there was statistically significant sinusoidal variation in girls for all lymphomas (P = 0.048; peak in March) and Hodgkin lymphoma (HL) (P = 0.005; peak in January), and in boys for osteosarcoma (P = 0.049; peak in October). This study confirms previous findings of seasonal variation around the month of birth for childhood ALL (at the peak ages) and provides further evidence of seasonal variation around month of birth for astrocytoma and around month of diagnosis for HL. The results are consistent with a role for environmental factors in the aetiology of these diagnostic groups. Further studies are needed to examine putative candidate agents.

Season of birth and risk for adult onset glioma

Adult onset glioma is a rare cancer which occurs more frequently in Caucasians than African Americans, and in men than women. The etiology of this disease is largely unknown. Exposure to ionizing radiation is the only well established environmental risk factor, and this factor explains only a small percentage of cases. Several recent studies have reported an association between season of birth and glioma risk. This paper reviews the plausibility of evidence focusing on the seasonal interrelation of farming, allergies, viruses, vitamin D, diet, birth weight, and handedness. To date, a convincing explanation for the occurrence of adult gliomas decades after a seasonal exposure at birth remains elusive.

The association between birth order, sibship size and glioma development in adulthood

The etiology of brain tumors is still largely unknown. Previous research indicates that infectious agents and immunological characteristics may influence adult glioma risk. The purpose of our study was to evaluate the effects of birth order and sibship size (total number of siblings), as indicators of the timing and frequency of early life infections, on adult glioma risk using a population of 489 cases and 540 cancer-free controls from the Harris County Brain Tumor Study. Odds ratios for birth order and sibship size were calculated separately from multivariable logistic regression models, adjusting for sex, family history of cancer, education, and age. Each one-unit increase in birth order confers a 13% decreased risk of glioma development in adulthood (OR = 0.87, 95% CI = 0.79-0.97). However, sibship size was not significantly associated with adult glioma status (OR = 0.97, 95% CI = 0.91-1.04). Our study indicates that individuals who were more likely to develop common childhood infections at an earlier age (those with a higher birth order) may be more protected against developing glioma in adulthood. More biological and epidemiological research is warranted to clarify the exact mechanisms through which the timing of common childhood infections and the course of early life immune development affect gliomagenesis.

Ependymal tumors

Ependymomas represent a heterogeneous group of glial tumors whose biological behavior depends on various histological, molecular, and clinical variables. The scope of this chapter is to review the clinical and histo-logical features as well as the molecular genetics of ependymomas with special emphasis on their influence on tumor recurrence and prognosis. Furthermore, potential molecular targets for therapy are outlined.

[Epidemiology of primary brain tumor]

Two main approaches are generally used to study the epidemiology of primary brain tumors. The first approach is to identify risk factors, which may be intrinsic or related to external causes. The second main approach is descriptive. Intrinsic factors potentially affecting risk include genetic predisposition and susceptibility, gender, race, birth weight and allergy. Radiation exposure is the main extrinsic factor affecting risk. A large body of work devoted, among others, to electromagnetic fields and especially cellular phones, substitutive hormonal therapy, pesticides, and diet have been published. To date, results have been discordant. Descriptive epidemiological studies have reported an increasing annual incidence of primary brain tumors in industrialized countries. The main reasons are the increasing age of the population and better access to diagnostic imaging. Comparing incidences from one registry to another is difficult. Spatial and temporal variations constitute one explanation and evolutions in coding methods another. In all registries, weak incidence of primary brain tumors constitute a very important limiting factor. Renewed interest from the neuro-oncological community is needed to obtain pertinent and essential data which could facilitate improved knowledge on this topic.

The epidemiology of neuroblastoma: a review

Neuroblastoma is the most common tumour in children less than 1 year of age. The goal of this review was to summarise the existing epidemiological research on risk factors for neuroblastoma. A comprehensive search of the literature was undertaken using PubMed for epidemiological studies on neuroblastoma risk factors. We ascertained 47 articles which examined the risk factors. Ten studies employed population-based case-control designs; six were hospital-based case-control studies; two were cohort studies; and five employed ecological designs. Studies ranged in size from 42 to 538 cases. Three studies showed evidence of an increased risk of disease with use of alcohol during pregnancy (OR range 1.1, 12.0). Protective effects were seen with maternal vitamin intake during pregnancy (OR range 0.5, 0.7) in two studies, while risk of disease increased with maternal intake of diuretics (OR range 1.2, 5.8) in three studies. Three studies reported a decrease in risk for children with a history of allergic disease prior to neuroblastoma diagnosis (OR range 0.2, 0.4). The rarity of neuroblastoma makes this disease particularly challenging to study epidemiologically. We review the methodological limitations of prior research and make suggestions for further areas of study.

Seasonality of birth in children with central nervous system tumours in Denmark, 1970-2003

We investigated possible seasonal variation of births among children <20 years with a central nervous system tumour in Denmark (N=1640), comparing them with 2,582,714 children born between 1970 and 2003. No such variation was seen overall, but ependymoma showed seasonal variation.

A method to model season of birth as a surrogate environmental risk factor for disease

Environmental exposures, including some that vary seasonally, may play a role in the development of many types of childhood diseases such as cancer. Those observed in children are unique in that the relevant period of exposure is inherently limited or perhaps even specific to a very short window during prenatal development or early infancy. As such, researchers have investigated whether specific childhood cancers are associated with season of birth. Typically a basic method for analysis has been used, for example categorization of births into one of four seasons, followed by simple comparisons between categories such as via logistic regression, to obtain odds ratios (ORs), confidence intervals (CIs) and p-values. In this paper we present an alternative method, based upon an iterative trigonometric logistic regression model used to analyze the cyclic nature of birth dates related to disease occurrence. Disease birth-date results are presented using a sinusoidal graph with a peak date of relative risk and a single p-value that tests whether an overall seasonal association is present. An OR and CI comparing children born in the 3-month period around the peak to the symmetrically opposite 3-month period also can be obtained. Advantages of this derivative-free method include ease of use, increased statistical power to detect associations, and the ability to avoid potentially arbitrary, subjective demarcation of seasons.

Potential mechanisms of the human polyomavirus JC in neural oncogenesis

The human polyomavirus JC (JCV) is a small DNA tumor virus and the etiologic agent of the progressive multifocal leukoencephalopathy. In progressive multifocal leukoencephalopathy, active JCV replication causes the lytic destruction of oligodendrocytes. The normal immune system prevents JCV replication and suppresses the virus into a state of latency so that expression of viral proteins cannot be detected. In a cellular context that is nonpermissive for viral replication, JCV can affect oncogenic transformation. For example, JCV is highly tumorigenic when inoculated into experimental animals, including rodents and monkeys. In these animal tumors, there is expression of early T-antigen but not of late capsid proteins, nor is there viral replication. Moreover, mice transgenic for JCV T-antigen alone develop tumors of neural tube origin. Detection of JCV genomic sequences and expression of viral T-antigen and agnoprotein suggest a possible association of this virus with a variety of human brain and non-CNS tumors. Here, we discuss the mechanisms involved in JCV oncogenesis, briefly review studies that do and do not support a causative role for this virus in human CNS tumors, and identify key issues for future research.

Space-time clustering of glioma cannot be attributed to specific histological subgroups

We previously showed that infectious exposures may be involved in the aetiology of adult glioma, by analysing for space-time clustering using population-based data from the South of the Netherlands. Here we extended these analyses and describe in detail the space-time clustering patterns in glioma subgroups, gender and age-categories. Knox tests for space-time interactions between cases were applied with fixed thresholds of close in space, <5 km, and close in time, <1 year apart. We used the spatial coordinates of the addresses at diagnosis in the analyses. Tests were repeated replacing geographical distance with distance to the Nth nearest neighbour. N was chosen such that the mean distance was 5 km. Data were also analysed by a second order procedure based on K-functions. There was only statistically significant space-time clustering for oligodendroglioma. Clustering was present for adults aged 30-54 years and was more pronounced among males. Given the low prior probability of an infectious aetiology for this specific subgroup, these results should probably be interpreted as false-positive. We conclude that space-time clustering of glioma cannot be attributed to a specific glioma subgroup. The observed clustering in our previous study is therefore probably an overall effect within and between glioma subgroups.

Breastfeeding and risk of childhood CNS tumours

We investigated infant feeding habits in relation to risk of childhood central nervous system tumours among 633 cases in the UK Childhood Cancer Study (UKCCS). No significant effect of breastfeeding was detected overall (odds ratio 1.01, confidence interval: 0.85-1.21) nor in any morphological subgroup. Similarly, no effect for the duration of breastfeeding or any other feeding practices was observed.

Space-time clustering analyses of testicular cancer amongst 15-24-year-olds in Northern England

There has been speculation that environmental exposures may be involved in the aetiology of testicular cancer in adolescent boys and young men. Indirect evidence for this hypothesis would be provided by the finding of space-time clustering. To examine this we have looked for evidence of space-time clustering using data from a population-based cancer registry from Northern England. All cases of testicular cancer diagnosed in males aged 15-24 years during the period 1968-2002 were included in the study. Tests for space-time interactions between cases were applied with fixed thresholds of close in space and close in time. Addresses at birth and diagnosis were used in the analyses. To adjust for the effect of varying population density tests were repeated replacing fixed geographical distances with nearest neighbour thresholds. A total of 257 cases of testicular cancer were identified for analysis. Overall there was no evidence for space-time clustering. However, there was statistically significant space-time clustering for 15-19-year-old based on time of birth and place of diagnosis (p<0.001). The very limited finding of space-time clustering may provide tentative evidence for an environmental, or infectious component to aetiology. However, it may well be a chance finding. A larger study based on national data is required.

Winter birth in association with a risk of brain tumor among a Finnish patient population

The aim of this study was to analyze whether winter birth is related to risk of brain tumor in a clinical sample of patients from northern Finland. The study group comprised 101 patients suffering from a primary brain tumor. When comparing births in winter to births in other seasons, a 1.39-fold (95% CI 1.01-1.77) excess of winter births among patients was observed compared to respective births in the general population (p = 0.026). Especially patients with pituitary adenomas exhibited a 2.5-fold (95% CI 1.5-4.4) excess of winter births. The authors conclude that the season-of-birth effect in brain tumor patients should not be neglected when the actual and important tumorigenesis is investigated.

Infection, immune responses and the aetiology of childhood leukaemia

Childhood leukaemia is the principal subtype of paediatric cancer and, despite success in treatment, its causes remain enigmatic. A plethora of candidate environmental exposures have been proposed, but most lack a biological rationale or consistent epidemiological evidence. Although there might not be a single or exclusive cause, an abnormal immune response to common infection(s) has emerged as a plausible aetiological mechanism.

Childhood leukaemias and CNS tumours: correlation of international incidence rates

Childhood leukaemia has a potential infectious aetiology whilst infections may also be linked to paediatric central nervous system (CNS) tumours. Using data from 29 countries we investigated the correlation between international incidence rates of childhood leukaemia and CNS tumours, focusing on acute lymphoblastic leukaemia (ALL), astrocytoma and ependymoma-subtypes that are hypothesised to have an infectious aetiology. Relationships between incidence rates and national demographic factors were also examined using Pearson's correlation coefficient to quantify associations. Comparing two diagnostic categories of leukaemia with four groups of CNS tumours, a highly significant positive correlation was found between ALL and astrocytoma (r = 0.57, P = 0.002). Higher rates of ALL and CNS tumours were associated with increased affluence, with the strongest correlation for Gross Domestic Product per capita and CNS tumours (r = 0.70, P < 0.001). National incidence rates of childhood ALL and astrocytomas were highly correlated and this may reflect a common environmental cause whose origin may be infectious in nature. International incidence of ALL and CNS tumours were also correlated with economic related factors. Variation in levels of ascertainment may partially explain this, although childhood environmental exposures related to infections will also be affected by levels of affluence.

Bayesian methods for early detection of changes in childhood cancer incidence: Trends for acute lymphoblastic leukaemia are consistent with an infectious aetiology

Published data on time trends in the incidence of childhood leukaemia show inconsistent patterns, with some studies showing increases and others showing relatively stable incidence rates. Data on time trends in childhood cancer incidence from the Childhood Cancer Registry of Piedmont, Italy were analysed using two different approaches: standard Poisson regression and a Bayesian regression approach including an autoregressive component. Our focus was on acute lymphoblastic leukaemia (ALL), since this is hypothesised to have an infectious aetiology, but for purposes of comparison we also conducted similar analyses for selected other childhood cancer sites (acute non-lymphoblastic leukaemia (AnLL), central nervous system (CNS) tumours and neuroblastoma (NB)). The two models fitted the data equally well, but led to different interpretations of the time trends. The first produced ever-increasing rates, while the latter produced non-monotonic patterns, particularly for ALL, which showed evidence of a cyclical pattern. The Bayesian analysis produced findings that are consistent with the hypothesis of an infectious aetiology for ALL, but not for AnLL or for solid tumours (CNS and NB). Although sudden changes in time trends should be interpreted with caution, the results of the Bayesian approach are consistent with current knowledge of the natural history of childhood ALL, including a short latency time and the postulated infectious aetiology of the disease.

Space-time clustering of childhood cancer in great Britain: A national study, 1969–1993

Previous studies have provided evidence that infections may play a part in the aetiology of certain childhood cancers. The finding of space-time clustering indicates the presence of an environmental component to aetiology and is especially supportive of a role for infections. Space-time clustering occurs when excess numbers of cases of a disease are observed within small geographical locations at limited periods of time and this cannot be explained in terms of general excesses in those locations or at those times. To investigate whether infections may be involved in the aetiology of childhood cancer, we have analysed for space-time clustering using a large set of national population-based data from Great Britain for the period 1969-1993. Data were examined by a second-order procedure based on K-functions, with fixed thresholds of closeness in space (0.5-7.5 km) and closeness in time (0.1-1.5 years). Locations were addresses at diagnosis. Tests were repeated, replacing geographical distances with distances to the 19th-33rd nearest neighbours and this provided the primary result for each analysis. There were a total of 32,295 cases of childhood cancer. The analyses showed statistically significant evidence of space-time clustering for acute lymphoblastic leukaemia over the whole age range (p = 0.04), but especially for ages 1-4 years (p = 0.03). There was less statistically significant evidence for total leukaemia (p = 0.048). Significant space-time clustering was also evident for soft tissue sarcomas (p = 0.03) and osteosarcomas (p = 0.02). Results support other evidence suggesting a role for infections in the aetiology of these particular diagnostic groups.

Space–time clustering patterns of gliomas in the Netherlands suggest an infectious aetiology

To test the hypothesis that infectious exposures may be involved in glioma aetiology, we have analysed space-time clustering and seasonal variation using population-based data from the South of The Netherlands between 1983 and 2001. Knox tests for space-time interactions between cases were applied, with spatial coordinates of the addresses at time of diagnosis, and with distance to the Nth nearest neighbour. Data were also analysed by a second order procedure based on K-functions. Tests for heterogeneity and Edwards' test for sinusoidal variation were applied to examine seasonal variation of incidence. There was statistically significant space-time clustering in the Eastern, but not in the Western part of the region. Clustering was only present in adults, particularly in less densely populated areas. There was no evidence for seasonal variation. The results support a role for infectious exposures in glioma aetiology that may act preferentially in certain geographical areas.

Is there a common aetiology for certain childhood malignancies? Results of cross-space-time clustering analyses

We previously demonstrated significant space-time clustering amongst cases of childhood leukaemia (in particular acute lymphoblastic leukaemia (ALL)), central nervous system (CNS) tumour (especially astrocytoma), soft tissue sarcoma and Wilms' tumour. We hypothesised that there may be common aetiological mechanisms between some of these diagnostic groups. To test this hypothesis we analysed for cross-space-time clustering between these diagnostic groups, using population-based data from north-west England. Data were examined by a second-order procedure based on K-functions. Reference points in time and space were dates and addresses at birth and diagnosis. The results showed statistically significant (P < 0.05) cross-clustering between cases of leukaemia and CNS tumour and between cases of ALL and astrocytoma. There was no statistically significant cross-clustering of Wilms' tumours and soft tissue sarcomas with any other malignancy. In conclusion, these findings are consistent with common, possibly infectious, aetiological mechanisms for childhood leukaemia (particularly ALL) and CNS tumours (particularly astrocytoma).

Central nervous system tumours in children: Epidemiology and risk factors

In children, a variety of distinct tumour types arise in the central nervous system (CNS) but they are frequently considered by epidemiologists as a single entity. CNS tumours are the second most common childhood malignancy and occur more frequently in developed countries than developing nations. Approximately 50% of paediatric CNS tumours are gliomas, with astrocytomas of the piloctyic type predominating. Distributions of subtypes by age show that primitive neuroectodermal tumours (PNET) and ependymomas mainly occur in children less than 10 years. There is an ongoing debate with regard to the apparent increases in incidence over recent decades, which, it has been suggested, are accounted for by improved diagnostic technologies implemented in the mid 1980s. CNS tumours differ between children and adults with respect to the proportion of all cancers (20% in children, less than 2% in adults), their cellular origins, classification scheme, the proportions of pathological subtypes, and anatomic site. The causes of childhood CNS tumours are largely unknown; and although an estimated 5% or more may be explained by genetic predisposition, investigations of environmental aetiology have not been fruitful. Whilst high dose ionising radiation is an established risk factor for this group of tumours, reported associations with dietary N-nitroso compounds have not been consistent. Exposure to ELF EMF, at any level, has not been associated with childhood CNS tumours, but the current evidence base is inadequate for complete evaluation. The rarity of childhood CNS tumours necessitates careful attention being paid to the design of future aetiological studies.