Central nervous system (CNS) cancer in children and ...

Journal of the Neurological Sciences 359 (2015) 151–155

Contents lists available at ScienceDirect

Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns

Central nervous system (CNS) cancer in children and young people in the European Union and its involvements with socio-economic and environmental factors Agustín Llopis-González a,b,c, Teresa Alcaide Capilla a, Unai Chenlo Alonso a, Nuria Rubio-López a,b,c, Antoni Alegre-Martinez d, María Morales Suárez-Varela a,b,c,⁎ a

Public Health and Environmental Care Unit, Department of Preventive Medicine, University of Valencia, Valencia, Spain CIBER Epidemiologia y Salud Pública (CIBERESP), Spain c Center for Advanced Research in Public Health, CSISP-FISABIO, Valencia, Spain d Biomedical Sciences Department, Universidad Cardenal Herrera CEU, Valencia, Spain b

a r t i c l e

i n f o

Article history: Received 11 May 2015 Received in revised form 29 October 2015 Accepted 30 October 2015 Available online 31 October 2015 Keywords: Children Cancer Central nervous system (CNS) Environment Socio-economic Industrialization

a b s t r a c t Malignant central nervous system (CNS) tumors are the leading cause of death by cancer in children and the second commonest pediatric cancer type. Despite several decades of epidemiologic research, the etiology of childhood CNS tumors is still largely unknown. A few genetic syndromes and therapeutic ionizing radiation are thought to account for 5–10% of childhood cancer, but the etiology of other cases remains unknown. Nongenetic causes, like environmental agents, are thought to explain them. However, as very few epidemiologic studies have been conducted, it is not surprising that nongenetic risk factors have not been detected. The biggest difference between cancers for which there are good etiologic clues and those for which there are none could be the number of relevant studies. This study, which covers the 1980–2011 period, identified links between CNS cancer evolution and the socioeconomic and environmental indicators in the same space and time limits in the European Union. © 2015 Published by Elsevier B.V.

1. Introduction Cancer ranks second as a global cause of death [1]. The latest statistics have revealed an increase in the number of cases, which does not seem to be exclusively due to a better and more accurate diagnosis. It is, therefore, urgent to discover the risk factors and to treat them with appropriate preventive measures. According to the International Agency for Research on Cancer (IARC), cancer caused 7.6 million deaths (approximately 13% of all deaths) in 2008 [2]. The etiology of cancer indicates that genetic and environmental factors are interrelated [3]. There are different hypotheses about the causes that produce cancer based on changes that occur to different populations. Thus a relationship appears between certain agents and cancer development. The IARC classifies these agents according to scientific Abbreviations: CNS, Central nervous system; CO2, Carbon dioxide; EU, European Union; GDP, Gross Domestic Product; IARC, International Agency for Research of Cancer; ICD, international classification of diseases; ISIC, International Standard Industrial Classification; WHO, World Health Organization. ⁎ Corresponding author at: Public Health and Environmental Care Unit, Department of Preventive Medicine, University of Valencia, Avda. Vicente Andrés Estellés s/n, 46100 Burjasot, Valencia, Spain. E-mail address: [email protected] (M. Morales Suárez-Varela).

http://dx.doi.org/10.1016/j.jns.2015.10.055 0022-510X/© 2015 Published by Elsevier B.V.

evidence and informs us of carcinogenic agents, but not all substances have been assessed. Specifically, central nervous system (CNS) cancer is the second commonest pediatric cancer type that causes more deaths in childhood [4]. In Europe, it is estimated that 140 per million for children (0–14 years) and 157 per million for children (0–19 years) will develop cancer each year [5]. The etiology of CNS cancer is not welldefined, but has been described to between 5% and 10% of cases with known risk factors and genetic factors, such as therapeutic ionizing radiation. It is believed that environmental factors play a key role in the remaining cases, and that the roots of the remaining 90– 95% lie in the environment and lifestyle. Lifestyle factors include cigarette smoking, diet (fried foods, red meat), alcohol, sun exposure, environmental pollutants, infections, stress, obesity, and physical inactivity. There is evidence to indicate that of all cancer-related deaths, almost 25–30% are due to tobacco, as many as 30–35% are linked to diet, about 15–20% are owing to infections, and the remaining percentage is due to other factors: radiation, stress, physical activity, environmental pollutants, etc. [6], although the results obtained so far are contradictory. Many potential carcinogenics are liposolubles and are, therefore, bio-accumulative in fatty tissues [6]. The brain is a tissue with a high

152

A. Llopis-González et al. / Journal of the Neurological Sciences 359 (2015) 151–155

fat concentration, which makes it more vulnerable to toxics due to its high lipophilicity [7]. Cancer is classified in the international classification of diseases (ICD 10) from heading C00 to D48, and according to anatomical and histological criteria. In this classification, CNS cancer occupies header C72. As shown in Table 1, CNS cancer is classified into different subtypes according to the function of the cellular structure [8]. Studies that have been carried out on the role of potential environmental agents in the etiology of CNS cancer in children and young people have spark considerable controversy, probably because the environmental conditions in which these studies were performed differed from each other or belonged to distinct times where pollution levels varied and were higher in the last study years. Therefore, our objective was to know CNS cancer evolution in children and young people in a given area, Europe Union (EU) countries in our case, and its possible involvements with environmental and socio-economic factors, to shed some light on its prevention. 2. Material and methods 2.1. Design This ecological study was conducted during the 1980–2011 period in EU countries. 2.2. Study unit EU countries were the study area (except Cyprus because it had no available data). The CNS tumors (rubric C72) rate that corresponds to children and young people (aged 0–19 years) has been identified. To perform this comparative study among EU countries, age-standardized mortality rates were used by an indirect method, and by taking the World Health Organization (WHO) population in 1996 as a reference. 2.3. Study variables The selected socio-economic and environmental indicators were: Gross Domestic Product per capita (current US$), which indicates a country's development level; industrialization level (% of GDP), which comprises the value added in mining, manufacturing, construction, electricity, water and gas [9]; carbon dioxide (CO2) emissions (metric tons per capita); electricity consumed (kWh per capita); fossil fuel energy consumed (% of total); nuclear power energy production (% of total energy use); PM10 (micrograms per cubic meter, particulate matter concentrations refer to fine suspended particulates less than 10 μm in diameter, the state of a country's technology and pollution controls are an important determinant of particulate matter concentrations); use of chemical products (% of the value added in manufacturing); use of fertilizers (kilograms per hectare of arable land). All these data were collected for each studied country. The sources resorted to were the WHO (World Health Organization Statistical Information System) and the World Bank, with high data accuracy.

Table 1 Classification of childhood CNS cancer and its frequency. Histological subtypes

% of all CNS tumorsa

Ependymomas Astrocytomas Primitive neuroectodermal tumors/medulloblastomas Other gliomas Other specified and nonspecified CNS

10 40–50 25 10 8–13

a

Proportions vary according to different case series.

The variables were selected by considering the relationship between them and the association of socio-economic and environmental exposure with its effect on CNS tumors in youngsters aged 0–19 years. Both socio-economic and environmental variables were integrated into the CNS mortality rates. As this study is of an ecological type, it is not without its weak points and is, therefore, subject to ecological deceit. Thus specific studies on the obtained results are needed to confirm these findings. This study also has its strong points as it collected data from countries under the same information conditions (both exposure and effect), which increases its internal validity. Moreover, the evaluation made of environmental exposure allows some research hypotheses to be put forward. (http://datos.bancomundial.org/indicador).

2.4. Statistical analysis With this information, a database was created using countries as a unit of ecological analyses with standardized age and sex rates. To know the correlation between death by CNS cancer and the socioeconomic and environmental factors, a Pearson's correlation analysis was used. A comparison was made between the death rates of a specific year and the factors of the previous year due to the lag period that cancer has [10], which we considered to be at least 1 year. Correlations were taken as significant if p-values were below 0.05. The clustering process is a statistical exploration technique that performs the partitioning of a data set into subsets (called clusters) so that the data in each subset share some common trait (most often the proximity according to some defined similarity measure). Clustering is used in many fields, including image analysis and bioinformatics [11,12]. Data clustering algorithms can be hierarchical or partitional [13]. The hierarchical cluster analysis (HCA) was used in this paper, whereas partitional algorithms determine all the clusters at once. The traditional representation of the hierarchical approach is a tree, called a dendrogram, with the different types of energy use in each country in the EU. Clustering is based on a matrix that contains the “distances” (dissimilarities) between multivariate observations. We used a multivariate comparison of death by CNS cancer in childhood in the EU by a cluster analysis-comparison of the different uses of energies (use of chemical products, CO2 emissions, fossil fuel energy consumed, industrialization level and nuclear energy) and GDP (Gross Domestic Product) in each EU country. Finally, in order to know the relation between the two variables from another perspective, a linear multiple regression analysis was used to assess the socio-economic and environmental factors as a whole. Regarding mortality, we identified the weight of each one with a saturated repeated model by introducing steps. In this way, the factors that better explained the evolution of CNS tumors in children and young people in the EU were selected. Data were analyzed by the Statistical Product and Service Solutions software (SPSS).

Table 2 Pearson's correlation of GDP per capita with mortality by CNS tumors in children and its pvalue in EU countries. Year

Pearson's product

p-Value

1981–1982 1993–1994 1994–1995 1995–1996 1996–1997 1997–1998 1998–1999 2000–2001 2001–2002 2008–2009

−0.520 −0.615 −0.589 −0.434 −0.469 −0.480 −0.394 −0.460 −0.609 −0.450

0.032 0.002 0.003 0.030 0.018 0.015 0.042 0.018 0.001 0.018

A. Llopis-González et al. / Journal of the Neurological Sciences 359 (2015) 151–155 Table 3 Pearson's correlation of the industrialization level with mortality by CNS tumors in children and its p-value in EU countries. Year

Pearson's product

p-Value

1983–1984 1985–1986 1989–1990 1990–1991 1993–1994

−0.719 0.494 0.542 0.472 0.461

0.001 0.037 0.025 0.041 0.031

3. Results 3.1. Pearson's correlation analysis In Table 2 we observe a negative correlation between mortality by CNS tumors in children and GDP. The higher the GDP per capita per EU country, the lower mortality by CNS tumors. However Table 3 indicates that the higher a country's industrialization level, the higher the mortality caused by CNS tumors in children; that is, the more industrial a country is, the more people will be exposed to all the pollutants emitted. The industrialization level (% of GDP) comprises the value added in mining, manufacturing, construction, electricity, water and gas. The added value is the net output of a sector after adding up all the outputs and subtracting the intermediate inputs. It was calculated without deductions for depreciation of fabricated assets, or for depletion and degradation of natural resources.

3.2. Multivariate analysis As a result of the hierarchical cluster multivariate analysis, in which we valued use of chemical products, CO2 emissions, fossil fuel energy consumed, industrialization level, nuclear energy and PM10, once again chemical products and industrialization level were the most closely associated factors with CNS cancer tumors in children and young people in EU countries (Fig. 1).

153

3.3. Multiple linear regression analysis We used this analysis to evaluate all the socio-economic and environmental factors as a whole in relation to mortality, rather than independently, as in previous analyses. This can offer a more realistic vision of the etiology of CNS cancer because exposure to environmental substances is simultaneous. We assessed all the indicators, except use of fertilizers for which not enough data were available. We observed an ecological association. The group of factors that better predicted the CNS cancer death rates in children and young people in the EU was associated with CO2 emissions, fossil fuel energy consumed and industrialization level. It is noteworthy that the environmental factors included in this result were chemical agents. So we once again highlight the importance of chemical agents, as in previous analyses. See Table 4.

4. Discussion The incentive of this study came from observing a cluster of childhood CNS tumors, and socio-economic and environmental exposure. From the results obtained with the analysis done based on this approach, we highlight that industrialization level related closely to death by CNS cancer in children and young people because it stood out in the three analyses done. This result reiterates that CNS cancer is more frequent in industrialized countries, as observed by Hiroko [14] and McKinney [15] in similar studies. This fact can lead to confusion when it comes to checking the Pearson's correlation results, where the GDP (measured in euros) stood out because more levels of significance were under 0.05, but most presented negative correlations. This can be explained by the amount of money invested in health services—the higher a country's GDP, the more it invests in health services—, which may be why mortality lowered. The World Bank classifies countries according to their GNI per capita income: low income ($995 or less); lower middle income ($996–$3945); developing countries with upper middle income ($3946–$12,195); developed countries with high

Fig. 1. Hierarchical cluster: dendogram using childhood CNS cancer and environmental exposure per chemical product and industrialization level in the EU.

154

A. Llopis-González et al. / Journal of the Neurological Sciences 359 (2015) 151–155

Table 4 All the indicators that better predicted the CNS cancer death rates in children and young people. Factors

ß

p-Value

Mortality (constant) CO2 emissions Fossils fuels Nuclear energy Industrialization level PM10

−4.90469 0.00001 0.17902 −0.19076 −0.19991 0.05577

0.418 0.028 0.049 0.074 0.009 0.087

income (above $11,906) [16]. The fact that death by CNS cancer in children and adolescents increases in more industrialized countries (with a higher proportion of industries) in our study could be due to the stronger chemical pollution that industries generate, and which implies worse environmental quality. Developed countries (those with a higher GDP index) present less mortality, which could be interpreted as their wealth originating from other non-polluting activities or as greater environmental control of industrial activities; indeed greater wealth is associated with better healthcare, which would justify these results. Certainly these results suggest the need for specific studies on chemical contamination, death by CNS in children, and on developing specific policies to protect against these contaminants. According to the multivariate analysis, a close link was found with use of chemicals products; here chemical products are understood as those used in the production of the industries classified as Category 3 in the International Standard Industrial Classification (ISIC). A relation between CNS cancer in children and young people and use of chemicals products has been found in the studies of Gomes, Al Zayadi and Guzman [17], Alexander et al. [18] and Haidar et al. [19]. In all these cases, CNS cancer in children and young people was related with their parents' occupation; the closest links appeared with parents who worked in chemical and oil industries. The studies of Cordier et al. [20] and Peters et al. [21] also showed a specific bond between parents who worked and were exposed to solvents, and an increase in astroglial tumors in their children. The fact that use of chemical products seemed to be closely related with death by CNS cancer, according to multivariate analyses, does not mean that it is the most directly probable cause of CNS cancer compared to other considered factors. Nevertheless, it is a secondary cause given its extensive use, and because exposure to chemical products is easy and frequent, which confers it more importance than the other indicators studied in this ecological study. A very general factor is chemical products, and this result can help us to know that such products are more relevant to the etiology of CNS tumors than physical agents. However, many studies are required to study all the chemical substances in depth and to ascertain what the possible causes of CNS tumors can be. The multiple linear regression analysis results showed the selected indicators that best predicted mortality by CNS cancer. At the same time, these indicators better explain evolution of death by CNS cancer in children and young people in EU countries during the 1980–2011 period. It is noteworthy that none of these indicators was related with mortality when analyzed independently in the correlations, except industrialization level. This analysis was more realistic because the population presented simultaneous exposure to all these environmental factors. As mentioned in the Introduction, cancer is not a consequence of only one factor, but of multiple factors. These results show the necessity of controlling these environmental indicators because of the impact they may have on healthy people, and given the need for future studies to confirm these results and to enable us to identify the etiology of CNS cancer in children and young people more specifically. An association between CNS cancer mortality and nuclear energy production was found in the study of Chang et al. [22]. Nevertheless, these authors studied only this indicator, and not its possible effects as a whole with other indicators. The same can be stated of CO2 emissions or fossil fuel energy use in the studies of Ferrís and Torjada [23], Ferrís and Torjada [24] and Zayas Mujica and Cabrera Cárdenas [25].

We found no relation herein between using electricity and death by CNS cancer in children and young people. Electricity is understood as the production of power plants and cogeneration plants that generate fewer electrical losses in not only the transmission, distribution and transformation rates, but also the consumption rates, of cogeneration plants. This indicator was chosen because it was the most appropriate one to explain the number of electrical power stations, electrical lines and electrical objects in each country. We considered it an important indicator because, based on the studies of Saito et al. [26] and Schüz [27], which indicated an association between CNS cancer in children and electromagnetic fields of extremely low frequency, many have been conducted and have obtained very different results. In the present work, we found no relation like Lagroye [28], and Pearlman [29], and unlike the works of Baldi [30] or Hardell [31]. Nor did we identify a relation with fertilizers. However, fertilizers are chemical substances, and as previously mentioned, we found a close link with such substances. Fertilizers are quite new products, which is why we still have insufficient information about their consequences for human health, and not many data have been registered. Nevertheless, an association with increased mortality by CNS cancer in children and young people has been found in the studies of Ward [32] and Gurney et al. [33]. 5. Conclusions Numerous different environmental exposures have been hypothesized to contribute to the development of CNS cancer, and have been explored in epidemiology research in recent decades. The fact that very few associations have been consistently replicated in studies by different researchers suggests that many distinct etiologies may be involved. The present study into death by CNS cancer in children and young people in EU countries during the 1980–2011 period and its involvements with socio-economic and environmental agents concludes that CNS cancer is more frequent in industrialized countries. The following stand out: chemical products, specifically CO2 emissions, PM10, and use of fossils fuels and nuclear sources to produce energy on the whole. However, no association was found with the use of electric energy and fertilizers in this EU-based ecological study. CNS cancer may have nongenetic etiology factors, but more studies have to be conducted to provide an ultimate answer. Disclosure The authors report no conflicts of interest in this work. References [1] M. Heron, Deaths: leading causes for 2010. National Vital Statistics reports: from the centers for disease control and prevention, National Center for Health Statistics, Natl. Vital Stat. Rep. 62 (2013) 1–97. [2] J. Ferlay, H.R. Shin, F. Bray, D. Forman, C.D. Mathers, D. Parkin, GLOBOCAN 2008, cancer incidence and mortality worldwide: IARC CancerBase no. 10, International Agency for Research on Cancer, Lyon, France, 2010 (Available at: http://globocan.iarc.fr). [3] J. Olsen, K. Overvad, The concept of multifactorial etiology of cancer, Pharmacol. Toxicol. 1 (2009) 33–38. [4] E. Ward, C. DeSantis, A. Robbins, B. Kohler, A. Jemal, Childhood and adolescent cancer statistics, CA Cancer J. Clin. 64 (2014) 83–103. [5] E. Steliarova-Foucher, C. Stiller, P. Kaatsch, F. Berrino, J.W. Coebergh, B. Lacour, M. Parkin, Geographical patterns and time trends of cancer incidence and survival among children and adolescents in Europe since the 1970s (the ACCIS project): an epidemiological study, Lancet 364 (2004) 2097–2105. [6] P. Anand, A.B. Kunnumakkara, A.B. Kunnumakara, C. Sundaram, K.B. Harikumar, S.T. Tharakan, et al., Cancer is a preventable disease that requires major lifestyle changes, Pharm. Res. Dordr. 25 (2008) 2097–2116. [7] P. Irigaray, J.A. Newby, S. Lacomme, D. Belpomme, Overweight/obesity and cancer genesis: more than a biological link, Biomed. Pharmacother. 61 (2007) 665–678. [8] E. Kramarova, C.A. Stiller, International classification of childhood cancer, Int. J. Cancer 68 (1996) 759–765. [9] World Bank, World databankRetrieved from http://data.worldbank.org/indicator/ NV.IND.TOTL.ZS 2015. [10] D.B. Richardson, S.R. Cole, H. Chu, B. Langholz, Lagging exposure information in cumulative exposure-response analyses, Am. J. Epidemiol. 174 (2011) 1416–1422. [11] M.R. Anderberg, Cluster Analysis for Applications, Academic Press, New York, 1973.

A. Llopis-González et al. / Journal of the Neurological Sciences 359 (2015) 151–155 [12] B. Everitt, Cluster Analysis, Heinemann Educ Books, London, 1974. [13] J.A. Hartigan, Clustering Algorithms, Wiley, New York, 1975. [14] O. Hiroko, Epidemiology of brain tumors. En: methods of molecular biology, Cancer Epidemiol. (2009) 472. [15] P. McKinney, Central nervous system tumours in children epidemiology and risk factors, Bioelectromagnetics 7 (2005) 60–68. [16] World Bank, World databankRetrieved from http://databank.worldbank.org/ddp/ home.do?Step=12&id=4&CNO=2 2011. [17] J. Gomes, A. Al Zayadi, A. Guzman, Occupational and environmental risk factors of adult primary brain cancers: a systematic review, Int. J. Occup. Environ. Med. 2 (2011) 82–111. [18] B.H. Alexander, J.H. Mandel, L.L.F. Scott, G. Ramachandran, Y.C. Chen, Brain cancer in workers employed at a specialty chemical research facility, Arch. Environ. Occup. Health 68 (2013) 218–227. [19] S. Haidar, C. Burns, K. Birdsong, K. Bodner, E. Salazar, J.J. Collins, Mortality among chemical workers at Texas City plant: 1940–2001, Int. Arch. Occup. Environ. Health 82 (2009) 145–151. [20] S. Cordier, L. Mandereau, S. Preston-Martin, J. Little, F. Lubin, B. Mueller, et al., Parental exposure to polycyclicaromatic hydrocarbons and the risk of childhood brain tumors: the SEARCH International Childhood Brain Tumors study, Am. J. Epidemiol. 159 (2004) 1109–1116. [21] S. Peters, D.C. Glass, K.R. Greenop, B.K. Armstrong, M. Kirby, E. Milne, et al., Childhood brain tumours: associations with parental occupational exposure to solvents, Br. J. Cancer 111 (2014) 998–1003. [22] C. Chang, A. Dave, E. Feinberg, M. Frimer, J.J. Marango, J. Sherman, Elevated childhood cancer incidence proximate to U.S. nuclear power plants, Arch. Environ. Health 58 (2003) 74–82. [23] J. Ferrís i Torjada, J. García i Castell, J.A. López, F. García i Domínguez, J.A. Ortega, O. Berbel, et al., Enfermedades asociadas a la polución atmosférica por combustibles fósiles. Aspectos pediátricos, An. Esp. Pediatr. 56 (2001) 213–225.

155

[24] J. Ferrís i Torjada, J.A. Ortega, J. Aliaga, A. Ortí, J. García i Castell, Introducción: el niño y el medio ambiente, An. Esp. Pediatr. 56 (2002) 353–359. [25] R. Zayas, U. Cabrera, Los tóxicos ambientales y su impacto en la salud de los niños, Rev. Cuba. Pediatr. 79 (2007). [26] T. Saito, H. Nitta, O. Kubo, S. Yamamoto, N. Yamaguchi, S. Akiba, et al., Power-frequency magnetic fields and childhood brain tumors: a case–control study in Japan, J. Epidemiol. 20 (2010) 54–61. [27] J. Schüz, Exposure to extremely low-frequency magnetic fields and the risk of childhood cancer: update of the epidemiological evidence, Prog. Biophys. Mol. Biol. 107 (2011) 339–342. [28] I. Lagroye, Y. Percherancier, J. Juutilainen, F.P. De Gannes, B. Veyret, ELF magnetic fields: animal studies, mechanisms of action, Prog. Biophys. Mol. Biol. 107 (2001) 369–373. [29] M. Pearlman, R. Frye, I. Butler, A.C. Papanicolaou, E. Castillo, J. Slopis, Patterns of electromagnetic activity recorded from neoplastic tissue, J. Neuro-Oncol. 105 (2011) 445–449. [30] I. Baldi, G. Coureau, A. Jaffré, A. Gruber, S. Ducamp, D. Provost, et al., Occupational and residential exposure to electromagnetic fields and risk of brain tumors in adults: a case–control study in Gironde, France, Int. J. Cancer 129 (2011) a1477–a1484. [31] L. Hardell, M. Carlberg, F. Söderqvist, K.H. Mild, Case–control study of the association between malignant brain tumours diagnosed between 2007 and 2009 and mobile and cordless phone use, Int. J. Oncol. 43 (2013) 1833–1845. [32] M.H. Ward, Too much of a good thing? Nitrate from nitrogen fertilizers and cancer, Rev. Environ. Health 24 (2009) 357–363. [33] J.G. Gurney, M.A. Smith, A.F. Olshan, S.S. Hecht, C.M. Kasum, Clues to the etiology of childhood brain cancer: N-nitroso compounds, polymaviruses, and other factors of interest, Cancer Investig. 19 (2001) 630–640.