Epidemiological Trends of Histopathologically WHO Classified CNS Tumors in Developing Countries: Systematic Review

Eliciting patterns for the distribution of disease between different populations over time is important for the development and the realization of health policies (Chauvin et al., 2012). For tumors originating in the Central Nervous System (CNS), there is a collection of a large spectrum of neoplasms classified by World Health Organization (WHO), with each tumor having its own features, location, morphology, prognosis and treatment (DeAngelis, 2001; Louis et al., 2007; Katchy et al., 2011). The latest Globocan report for 2012 estimated a total incidence ASR of 3.4 per 100,000 and a mortality ASR of 2.5 per 100 000 worldwide (Parkin et al., 2001; GLOBOCAN, 2014). Incidence ASR of CNS tumors in the developing region was estimated to be 3.0 per 100,000, and for mortality ASR the estimation was 2.2 per 100,000. In addition, Globocan described gender-related data and showed that men have a higher CNS tumors incidence ASR 3.9 compared with women 3.0 worldwide. Unfortunately, however, trends for individually classified types of CNS tumors are not available from this database.


Introduction
Eliciting patterns for the distribution of disease between different populations over time is important for the development and the realization of health policies (Chauvin et al., 2012).For tumors originating in the Central Nervous System (CNS), there is a collection of a large spectrum of neoplasms classified by World Health Organization (WHO), with each tumor having its own features, location, morphology, prognosis and treatment (DeAngelis, 2001;Louis et al., 2007;Katchy et al., 2011).
The latest Globocan report for 2012 estimated a total incidence ASR of 3.4 per 100,000 and a mortality ASR of 2.5 per 100 000 worldwide (Parkin et al., 2001;GLOBOCAN, 2014).Incidence ASR of CNS tumors in the developing region was estimated to be 3.0 per 100,000, and for mortality ASR the estimation was 2.2 per 100,000.In addition, Globocan described gender-related data and showed that men have a higher CNS tumors incidence ASR 3.9 compared with women 3.0 worldwide.Unfortunately, however, trends for individually classified types of CNS tumors are not available from this database.official websites of cancer registries networks, such as Globocan, AFCRN (African Cancer-Registry Network) and EUCAN, or cancer registries for 149 individual countries.For PubMed search, four main limitation steps were applied.The first included year, tumor, incidence and country name, the second involved scanning retrieved abstracts for a list of key words, the third involved checking for authenticity of names (Sometimes a name of country overlapped with other elements for example the name Georgia stands for a European developing country and an American state) and finally a check on the date of the reported cases was applied (some papers were published in 2010 but were reporting cases prior to 2009).
Identified countries were grouped within six regions: Europe and Central Asia (ECA), Latin America and the Caribbean (LAC), East Asia and Pacific (EAP), Middle East and North Africa (MENA), South Asia (SA), and Sub-Saharan Africa (SSA), as per WHO cataloging.
Linear regression analyses for the relationship of Globocan reported incidence ASR (Ln) for all CNS tumors reported from 169 WHO registered countries, or 126 developing countries, verses latitude or longitude were run using SPSS Graduate Pack 16.0.A similar process was applied to test the significance of the relationship between PubMed reported Astrocytic tumors counts (Ln) for 25 countries with latitude or longitude.Natural logs of CNS tumors counts were chosen since values for counts had skewed distributions.For each country, latitude/ longitude coordinates were taken from Google Earth 6.0.2.(Google, 2014).

CNS tumor cases reported via cancer registries
Percentage ordered cases for all reported histopathologically distinguished types of CNS tumors identified through cancer registries showed an overwhelming majority of reported cases to be WHO unclassified (30311 cases, 96.1%), (Table 2).The data for unclassified tumor cases showed a huge variation of reported cases between countries, ranging between a 25-percentile of 77 to a 75-percentile of 2123 and a standard deviation of ±493.0.The top three reported types were astrocytic tumors (627 cases, 1.99%), gliomas (not who classified) (154 cases, 0.49 %) and embryonal tumors (107 cases, 0.34%).the least reported types were tumors of cranial and paraspinal nerves (1 case, 0.05%).Only 12 WHO classified categories were noted.

CNS tumor cases reported via pubmed
For CNS tumor cases identified via PubMed publications 17 WHO CNS tumor categories were noted, (Table 2).Collected data for all tumor cases showed a tighter variation of reported cases between countries ranging between a 25-percentile of 5.3 to a 75-percentile of 155 and a standard deviation of ±83.48.Under half of cases (3612 cases, 45.6 %) were unclassified and were labeled as brain and/or CNS tumors.The top three reported types were astrocytic tumors (1245 cases, 15.7%), pituitary tumors (792 cases, 10.0%) and meningeal tumors (544 cases, 6.88%).the least reported types were choroid plexus tumors (3 cases, 0.04%).

Discussion
This study aimed to show epidemiological trends of individually reported CNS tumors subtypes in developing countries.Through a cyber-search approach, it was possible to elucidate trends of frequencies for individually classified CNS tumors and show patterns that were otherwise unavailable.Our data shows that the majority of reported cases from PubMed and cancer registries were for unclassified tumors.For classified tumors, Astrocytic tumors appear to be the most reported type.Interestingly, a relationship between geographical location and the occurrence of CNS tumors may exist.
Unfortunately, out of 149 developing countries only 66 registries representing 21 countries reported any information related to CNS tumors, confirming previous observations for the lack of reporting for CNS tumors in the developing region (Ferlay et al., 2010;Qaddoumi et al., 2011).The degree of reporting for tumors maybe linked to the level of socio -economic development for any particular country.The quality of data collected from the identified registries was poor compared to data retrieved thought published resources.We ensured that included PubMed publications were sourced by clearly individual hospitals or health sectors, and did not produce repetition or overlap of cases.Importantly, data produced from PubMed showed a larger range of tumor classification that was not detected in cancer registries.One problem was the lack of gender or age categorization for declared tumors in publications.Another caveat of retrieving data from such source is the hypothetical bias of results towards research interest in a particular type of tumors.However, our method of selecting published papers was intended to focus on country related data and not any particular tumor classification.
Alarmingly, the majority of reported cases were unclassified (Pubmed (45.6%), registries (96.1%)) consistent with the notion that reporting unclassified cancer variants is a growing phenomenon (Tavtigian et al., 2008).This could be a reflection of routed miscommunication between clinicians and statistical departments.It is possible that the epidemiological context for each of these bodies is different.While clinicians and pathologists are more concerned with understanding disease classifications to make crucial decisions involving management of treatment (Tavtigian et al., 2008), reporting statistical departments may be more concerned with "number of cases" in general.In addition, reporting classified tumors is more complex and involves extra documentations, such as pathological reports for classified tumors (Jensen and Storm, 1991).Unfortunately, such vague reporting systems hamper sharing of detailed medical knowledge and perhaps hinder the development of effective targeted medical management programs.
The retrieved data from PubMed shows that the top three reported types of CNS tumors were Astrocytic tumors, Pituitary tumors and Meningeal tumors.Reasons for the high frequency of Astrocytic tumors are generally not explained (Karipidis et al., 2007).It is possible that since Astrocytic tumors have devastating outcomes with poor survival, more researchers are interested to study them and thus there is more reporting in PubMed, however, Astrocytic tumors were also the most frequently reported type in cancer registries.Another consideration is the complexity of classification for astrocytic tumors, as several different histological categories of variants exist (Wen and Kesari, 2008;Adesina et al., 2010;Karsy et al., 2012) raising the possibility of effect due to increased size of grouping.
Interestingly, when analyzing the relationship between the natural logs of CNS tumors incidence ASR for all WHO registered countries including developing countries verses latitude, or for the natural logs of counts for Astrocytic tumors verses latitude, positive correlations emerged.A similar association has been observed for other tumors (Crocetti et al., 2012;Grant, 2012;Cuomo et al., 2013;van Leeuwen et al., 2013).Latitude effect may be associated with several factors such as UV irradiation, Vitamin D biosynthesis or geographical differences for regional health systems.
In conclusion, reporting epidemiological data for unclassified CNS tumors is a growing phenomenon in the developing world that contradicts the attitude of health professionals to deal with CNS tumors as a collection of independent tumors, of which each requires a targeted health management plan.Unfortunately, Globocan reports treated epidemiological data for CNS tumors within a single category, and only a few cancer registries for developing countries provided information related to the incidences of defined histological CNS tumors types.Through the retrieval of CNS tumor cases using ISI PubMed publications, overall frequencies for individually classified CNS tumors in 38 different countries were elucidated.Patterns associated with frequencies and ranking of classified tumors were deduced as well as novel data implicating an association between individual CNS tumors, such as astrocytic tumors, with geographical location.A consensus for future reporting of CNS tumors incidences need to be established in order to help progress prospects of CNS tumors' health management.DOI:http://dx.doi.org/10.7314/APJCP.2015.16.1.205 Epidemiology of Central Nervous System Tumors in Developing Countries Adeleye AO, Okolo CA, Akang EE, et al (2012)

Figure 1 .
Figure 1.A Flow Chart for Search Strategy and Data Extraction Approaches Used in this Study.Dashed-Box (---) represents limitations used for the search.Keywords list1: CNS tumors in [Country name] 149 developing countries, Cancer, Registry, Data, Statistics, Incidence rate, Mortality/death rate, Prevalence, Epidemiology, Registration, Quantitative study, Clinical profile, or ASR.Search Queries2: [Country name] 149 developing countries and Cancer registry or Cancer Society or Cancer report or Cancer statistics

Figure 2 .
Figure 2. Scatter Blots for the Natural Logs of CNS Tumors Counts Verses Geographical Coordinates of WHO Registered Countries.a) The natural logs ofGlobocan reported incidence ASR (Ln) for all CNS tumors from 169 countries versus i) Latitudes of countries south the equator, (-40 to 0°), ii) Latitudes of countries north the equator, (0 to 70°), or iii) Longitudes.b) The natural logs of Globocan reported incidence ASR (Ln) for all CNS tumors from 126 developing countries versus i) Latitudes of countries south the equator, (-40 to 0°), ii) Latitudes of countries north the equator, (0 to 70°), or iii) Longitudes.c) The natural logs counts of Astrocytic tumor collected from 25 developing countries versus i) Latitudes or ii) Longitudes.Equations for linear regression curves are shown above corresponding blots.The (*) indicates a significant linear regression relationship