Paediatric cancer in low-income and middle-income countries

Elsevier Editorial System(tm) for The Lancet Oncology Manuscript Draft Manuscript Number: THELANCETONCOLOGY-D-12-01512R1 Title: Pediatric Cancer in Countries with Limited Resouces Article Type: Review (Invited) Keywords: Paediatric, Cancer, developing countries, limited resources Corresponding Author: Dr. Ian Magrath, Corresponding Author's Institution: First Author: Ian T Magrath, DSc (Med) Order of Authors: Ian T Magrath, DSc (Med); Eva Steliarova-Foucher, PhD; Sidnei Epelman, MD; Raul C Ribiero, MD; M'Hammed Harif, MD; Chi-Kong Li, MD; Rejin Kebudi, MD; Scott MacFarlane, MD; Scott C Howard, MD Manuscript Region of Origin: BELGIUM Abstract: Patterns of cancer across the world have undergone dramatic changes as a consequence of the industrial revolution, which was firmly established in Great Britain by 1850. The spread of industrialization, however, has been slow, and the economies of most countries remain in an early or intermediate industrial phase. Lower levels of development are characterized by poverty, limited access to education and health care due to scattered rural populations, too few health care providers, and weak health systems. Less developed countries also have younger populations and therefore a higher proportion of cancer in children than more developed countries. The majority of these children die. Exposure to the technological revolution is minimal in the rural regions of developing countries, and chronic infections and infestations, which remain the leading cause of disease-related death in all except the high income countries, may also be major risk factors for childhood cancer in these regions. Here, we discuss childhood cancer in relationship to development at a global level and propose strategies that could result in improved care. Education of the public, more and better trained health professionals, strengthened cancer services, locally relevant research, regional hospital networks, international collaboration and health insurance are essential components of an improved model of care.

INCTR, SJCRH

Paediatric Cancer in Countries with Limited Resources Manuscript 3 Revised Version

Ian Magrath et al. Revised Version 12/31/2012

Prof. Ian Magrath,1 Eva Steliarova-Foucher,2 Sidnei Eppelman, Prof. Raul Ribeiro,3 Prof. M’Hamed Harif,4 Prof. Chi-Kong Li, Prof. Rejin Kebudi,6 Prof. Scott MacFarlane,7 Prof. Scott C. Howard3 1 International Network for Cancer Treatment and Research, Brussels, Belgium 2 International Agency for Research on Cancer, Lyon, France 3 Santa Marcelina Hospital, Sao Paolo, Brazil 4 St. Jude Children’s Research Hospital, Memphis, USA 5 Centre Hospitalier Universitaire Mohhamed VI, Marrakech, Morocco 6 Prince of Wales Hospital, Hong Kong, China 7 Cerrahpasa Institute, Istanbul Turkey 7 Starship Children’s Health, Auckland, New Zealand

Paediatric Cancer in Countries with Limited Resources_REVISED VERSION.docx

2

Conflict of Interest

IM None

ESF

SE

RCR

MH

CKL

RK

SM

SH

Authors’ Contributions

IM Design and writing of manuscript, creating tables and figures, identification of many

references

ESF

SE

RCR

MH

CKL

RK

SM

SH

3

Abstract

Patterns of cancer across the world have undergone dramatic changes as a consequence of the industrial revolution, which was firmly established in Great Britain by 1850. The spread of industrialization, however, has been slow, and the economies of most countries remain in an early or intermediate industrial phase. Lower levels of development are characterized by poverty, limited access to education and health care due to scattered rural populations, too few health care providers, and weak health systems. Less developed countries also have younger populations and therefore a higher proportion of cancer in children than more developed countries. The majority of these children die. Exposure to the technological revolution is minimal in the rural regions of developing countries, and chronic infections and infestations, which remain the leading cause of disease-related death in all except the high income countries, may also be major risk factors for childhood cancer in these regions. Here, we discuss childhood cancer in relationship to development at a global level and propose strategies that could result in improved care. Education of the public, more and better trained health professionals, strengthened cancer services, locally relevant research, regional hospital networks, international collaboration and health insurance are essential components of an improved model of care.

4

Epidemiology of Childhood Cancer in Developing Countries

Childhood cancer (considered here as cancer in 0-14 year olds) constitutes only a small fraction of the global cancer burden, but 84% of childhood cancers occur in the low (LIC) and middle income (MIC) countries (LMIC) where 90% of the world’s children live and where access to care is limited (Table 1) [1,2,3]. In 2010, LMIC accounted for 5.7 billion (83%) of the world’s population (6.9 billion). These countries have younger median ages and higher proportions of children in their populations (27% in MIC and 40% in LIC) than high income countries (HIC), in which 17% are less than 15 years [1]. In 2008, Globocan, a project of the International Agency for Research on Cancer (IARC) [3] estimated that approximately148,000 cancers in children aged 0-14 years occurred in less developed regions, a term which corresponds well to the World Bank’s LMIC, and includes Africa, Asia (excluding Japan), Latin America and the Caribbean, Melanesia, Micronesia and Polynesia, comprising a total population of 5.5 billion. The equivalent Globocan estimate for children with cancer living in more developed regions, defined as Europe, Northern America, Australia, New Zealand and Japan, was 28,000. These regions, together, have a total population of 1.2 billion people and differ little from HIC as defined by the World Bank. Relative importance of cancer as a cause of death in HIC and LMIC Relatively few countries, accounting for 15.9% of the world’s population, fall into the HIC category, yet these countries account for 31.6% of global deaths from cancer, largely because of the high proportion of older people in their populations [1,4]. World Health Organization (WHO) mortality data for the year 2008 (latest available) indicates that cancer is the leading cause of disease-related death in children between the ages of 5 and 14 years living in HIC (Table 2), with approximately 5.4 times as many deaths from cancer than from communicable diseases [4]. In LMIC, however, deaths from communicable diseases continue to outweigh cancer deaths in all age groups, although the ratio between communicable and non-communicable diseases becomes progressively smaller as the level of development increases, ranging from 20 times as many deaths from infections and parasitic diseases in low income countries (LIC) to close to twice as many in upper middle income countries (UMIC) [4]. In children aged 0-4, communicable diseases cause more deaths than cancer in all income groups, including HIC, but in HIC, there are approximately twice as many deaths from infections and parasitic diseases than from cancer. In LMIC, the figure is closer to 150 fold higher (Table 3). The demographic transition and development The age structures of developing countries (Figure 1) arise from their high birth rate, relatively high death rate and low life expectancy. This is consistent with a model of changes in population structure resulting from the demographic transition brought about by the industrial revolution (Figure 1). Great Britain, where the industrial revolution began, had, by 1850, made the transition from an economy based on agriculture to one based on manufacturing, with its attendant and progressive urbanization, new lifestyles, new exposures to a variety of chemicals, cheap manufactured cigarettes, more food (often processed or chemically preserved) and, at least initially, massive pollution. These dramatic changes were factors in the rapid increase in cancer that occurred at all ages, even while improved living conditions, safe water, better sanitation and eventually, markedly improved health care (including cancer care) led to an increased lifespan. However, most countries are still undergoing a demographic transition, although many of the accoutrements of the industrial and technological revolutions have reached them as global trade has increased. The degree of development can, then, be

5

seen primarily as a reflection of the degree to which countries have undergone a demographic transition, although many local factors, political and otherwise have also influenced socioeconomic progress. In turn, the degree to which they have progressed through the demographic transition influences their pattern of disease, as described above, creating, in addition to the high, if falling, rates of communicable diseases, from which they have long suffered [4], an additional, rising mortality from cancer – the so-called “double burden of disease.” However, although the global cancer burden is projected to increase as populations grow and age, the childhood cancer burden is expected to increase only marginally under the assumption of medium population growth. This is because fertility is falling, and the global childhood population, according to the United Nations, has reached a peak [1]. In addition, aging does not apply to childhood populations. But there is a caveat. While demographic factors influencing the cancer burden are likely to have a minimal impact on childhood cancer, (although predictions can prove to be wrong!), development could result in increased exposure to risk factors and a larger than anticipated increase in childhood cancer. Burden and spectrum of childhood cancer Deaths from infection and parasitic diseases predominate in the 0-4 years age group in all countries, regardless of income, such that the percentage of disease-related deaths caused by cancer is very small compared to 5-14 year olds (0.5% and 14.6% respectively). However, this does not indicate that cancer is less common in 0-4 year olds since the mortality rate from cancer (6.3 per 100,000) is higher than in 5-14 year olds (4.3 per 100,000) (Tables 2 and 3). Children in the 0-4 year age group have a somewhat different spectrum of cancers than older children (although there is overlap); the majority of cases of retinoblastoma, neuroblastoma, Wilms tumor, embryonal rhabdomyosarcoma, hepatoblastoma and yolk sac tumors occur in this age group [5]. Some of these tumours may originate from from embryonic cells persisting after birth, or from oncogenic events in utero. In children between 5 and14 years of age, sarcomas are more common, while haematological malignancies and brain tumours (both encompassing many different cancers), which are the most common tumour types in children, occur in both age groups. While cancer incidence rates in 0-14 year olds are higher in more developed than in less developed countries (Table 4), the percentage of all cancers in the childhood age group is higher – often considerably higher - in the less developed countries. This is because of the higher proportion of children in their populations [1]. On average, 2% of cancers in less developed countries and 0.5% of cancers in more developed countries are in the 0-14 age group [3], but these figures vary in different regions and countries. In Africa as a whole, 4.8% of cancers are in children below the age of 15 years, compared to 0.4% in Europe, indicating that the proportion of cancers occurring in children is approximately 12-fold higher in Africa than in Europe (Figure 2). Information of this kind is important in cancer control planning. Incidence of childhood cancer in LMIC Based on Globocan estimates, the incidence of childhood cancer varies from just over 5 to close to 20 per 100,000 people per year in different geographical regions and in countries at different socioeconomic levels (Figure 2). It should be recognized that because of the paucity of cancer registries in developing countries, and the use of cancer rates in all ages, Globocan estimates for LMIC might not be of sufficient quality for some tasks, particularly with respect to childhood cancer. But, as stated in the description of data sources used to create Globocan estimates [3], they are still of unique importance as this information, which also incorporates WHO mortality data, often remains the only relatively unbiased source of information available

6

on the profile of cancer. Nonetheless, there is an urgent need for more cancer registries in developing countries and particularly in rural regions, since most cancer registries are in urban areas, and may not accurately reflect the national picture, particularly in countries where the population is predominantly rural (e.g., equatorial Africa [6]. The reason for the variability in total childhood cancer incidence across the world is largely unknown, although it must relate to exposure to risk factors. In some cases the very high incidence of one cancer may be enough to account for an unexpectedly high overall incidence. In equatorial Africa, for example, Burkitt lymphoma (BL) accounts for approximately half of all childhood cancers [7,8] while the AIDS epidemic in many African countries led to a marked increase in Kaposi sarcoma (KS),even to the point of exceeding the incidence of BL [9]. Other countries appear to have a very low overall incidence of childhood cancer, particularly countries in the southern and western parts of Africa [3]. One reason for low overall incidence rates may be that many childhood cancers are not diagnosed, particularly leukaemias and brain tumours, which appear from the limited registry data available to be generally uncommon in LIC [5,10]. But while under-diagnosis and under-reporting doubtless occur, the magnitude of the incompleteness of the data is difficult to estimate because of the lack of a source of independent data, such as death certificates. Moreover, it is unlikely that under-reporting is the sole reason for low overall incidence rates, since some cancers such as retinoblastoma and Wilms tumor have comparable or even higher incidence rates in LMIC than in HIC, while diseases with similar clinical characteristics may vary in incidence in different regions, e.g., precursor B and precursor T cell acute lymphoblastic leukaemias (ALL) [5,10,11]. In contrast to incidence rates, mortality rates are much higher in children in the less developed countries suggesting much lower survival rates. For example, in 2008, the mortality rate in Africa was 6.9 per 100,000 children compared to 3.1 per 100,000 in Europe. Unsurprisingly, in Africa, an average of 5.13% of cancer deaths were in children while In Europe, children accounted for only 0.2% of cancer deaths (Figure 3). Although survival data is not available, a mortality: incidence ratios give an idea of the correlation between development and survival (Figure 4). Risk factors for childhood cancer in LMIC Differences in incidence of individual childhood cancers reflect differences in exposure to risk factors coupled to genetic factors that influence the likelihood of a cancer developing. There is little recent data on the incidence of childhood cancer sub-types among LMIC, but earlier data strongly suggested that there is considerable geographic variability (Figure 5). Leukemias and lymphomas are the most common childhood cancers throughout the world, with ALL accounting for approximately 30% of all cases of childhood cancer, although its incidence does appear to vary. (Figure 5). In equatorial Africa, ALL is uncommon, its place being taken by BL, except where the HIV epidemic has led to a marked increase in the incidence of KS [9]. Some cancers may vary in incidence according to the level of socioeconomic development. Support for this possibility is provided by the decline in incidence of BL and increase in incidence of ALL in the Gaza strip that coincided with a period of rapid socioeconomic development [12]. There is additional evidence that the incidence of common (precursor B) ALL is low in LMIC and increases markedly in incidence as nations undergo socioeconomic development [12,13]. For example, ALL has a low incidence in India and several institution-based studies have shown that the proportion of T cell cases is high, particularly in rural regions, but has been decreasing over time [11-15]. This is consistent with data from Egypt [16,17] and evidence from death certificates in England and Wales that precursor-B ALL, using the early age peak (at 2-5 years)

7

as a marker, started to increase in incidence in the early 20th century, first in England and Wales, and subsequently in various other population groups [13,17,18]. Retinoblastoma and BL, in contrast, may be cancers that are inversely associated with socioeconomic development and appear to be more common in rural regions [7,8,10]. Retinoblastoma, is sometimes the most common solid tumour of childhood in LMIC and may have a higher incidence than neuroblastoma, which is invariably more frequent in HIC [5,10]. Epstein-Barr virus (EBV) is associated with several childhood cancers, particularly in LMIC [19,20]. Almost 100% of BL in Africa and over 50% of BL in LMIC from which data is available contain EBV [20]. Although EBV is ubiquitous, infection occurs in infancy or early childhood in LMIC and this could account for its more frequent association with specific cancers in these countries [20]. In contrast, in the wealthiest countries, where initial EBV infection occurs in adolescents and young adults, a majority of BLs are EBV negative. The Berlin-Frankfurt-Münster (BFM) cooperative group, for example, reported that only 11% of BL cases in their series were associated with EBV [21]. EBV is also present in almost all cases of Hodgkin lymphoma in children less than 10 years of age in LMIC [22,23], and has been reported in NK/T cell lymphomas children in Peru [24]. T cell lymphohistiocytosis in Asia (associated with hemophagocytic syndrome) [25] as well as in nasopharyngeal carcinoma, which is relatively common in children in North Africa, South and South East Asia [25,26]. BL is probably also predisposed to by malaria, at least in regions holoendemic for P. falciparum, which include equatorial Africa and Papua New Guinea [20]. Hepatocellular carcinoma, which is strongly associated with Hepatitis B (HBV) and C (HCV) viruses is rare in children, but occurs at higher incidence in some regions, such as Taiwan, where it can be effectively prevented by vaccination against HBV [27,28] The increased risk of KS in children co-infected with HIV (Human Immunodeficiency Virus) and the KS herpesvirus has already been noted [10,29]. Although HIV infection has been shown to predispose to B cell lymphoma in many countries, the most recent evidence suggests that this is not so in the case of BL in children in equatorial Africa [30]. Radiation increases the risk of leukaemias and solid tumors in children as shown by epidemiological studies conducted in children exposed pre or post natal to radiation from the atomic bombs dropped on Japan, or following therapeutic radiation or diagnostic tests [31,32]. A high incidence of thyroid cancer was observed after the Chernobyl nuclear accident [33,34] and background environmental radiation also appears to increase the risk for leukemia, probably with no lower dose threshold [5,10,35].

Access to Care in LMIC

With the exception of hepatocellular carcinoma, where a causal factor (HBV) has been identified and an effective vaccine exists, prevention of childhood cancer is not possible, and mortality rates can only be reduced by improving treatment outcomes. Treatment requires many experts, infrastructure, equipment and drugs and can be costly, but a child with cancer should have the same rights to health care as any other child. The stark reality in LMIC, unfortunately, is that access to care is extremely limited. An unknown fraction of children with potentially curable cancers never receive treatment – not even palliative care and most of those who do die. While minimizing symptoms throughout the entire clinical course of cancer is of primary importance, palliative care services are in their infancy in LMIC and the expansion of hospital and

8

community-based palliative care services in LMIC, as well as access to pain medication, must be a high priority for these countries [36].

Human resource deficiencies, limited finances and poverty In the 36 LIC listed by the World Bank, although heterogeneous in many ways, there are more similarities than differences with respect to the factors that influence access to care. The vast majority of treatment centers are within the public sector, and referrals predominantly come through primary or secondary health care providers, many of whom see very few childhood cancers in their entire professional lives. WHO strongly advocates the development of networks of primary health care providers (where the access point to the health system is via an ambulatory generalist rather than a hospital) on the grounds that this would result in greater equity with regard to access to care [37]. However, there are often insufficient primary care providers in LIC and in many MIC or regions of MIC to serve the needs of the country, and the majority of the population has little knowledge of cancer, particularly in children. Many families consult traditional healers, who are more accessible in rural regions, but may lead to delays in obtaining a medical consultation. Even primary care providers or specialists may not consider the possibility of cancer and months may be wasted in futile treatment of a condition the child does not have; delay in diagnosis results in more advanced disease and a worse prognosis. When children are sufficiently ill, parents may take them directly to a tertiary care center, but the paucity of such hospitals, either in the country as a whole (as in LIC particularly), or in large rural regions, means that for much of the population of LMIC, tertiary care centers are far from patient’s homes. This creates multiple problems, including transportation costs, absence of parents from the home for long periods of time with detriment to the rest of the family and an unaffordable loss of income. A free hostel near the treatment center where parents can stay while away from home can reduce abandonment (premature termination) of therapy [38], but the cost of treatment may deter families from accepting care. With the exception of the few countries which have insurance coverage, the family must pay out-of-pocket for diagnosis and much or all of therapy. Parents sometimes stop therapy when their money runs out or if they feel that they must return home, even if to do so could result in the sick child’s death. Ultimately, reaching a tertiary care center is only of value if trained staff with experience in the treatment of childhood cancer and the necessary equipment and drugs (nearly all are listed on WHO’s Essential Medicines List), as well as standard treatment protocols are available. Even cheap, commonly used drugs such as cyclophosphamide and vincristine may not be imported for periods of time, and radiation therapy facilities, along with the staff required to deliver radiotherapy, are totally insufficient in many countries – for example, it has been estimated by the International Atomic Agency’s Program of Action for Cancer Therapy (PACT) that 80% of Africans lack access to radiation therapy [39]. In 30 African and Asian countries, radiation therapy is completely absent and the various experts – medical or otherwise - required to ensure well organized health services are also lacking, or very few in number and therefore inaccessible to much of the population. Needless to say, the deficits are worse in rural areas and cancer centers vary considerably with respect to the quality of care provided. Different ethnic and income groups may also experience inequities in care – even in HIC, although differences are particularly large in LMIC. Often the wealthy are treated in well equipped private hospitals in their own or other countries (not necessarily HIC) for treatment or obstetric care - frequently paid for by their government [40]. Although capacity for pediatric cancer care inevitably improves with development, such that overall, access to care is better in MIC than LIC, there are enormous variations in the latter,

9

both within and between countries. Per capita income, for example, differs 10 fold among the total of 108 economies presently classified by the World Bank as MIC, and in any event, does not correlate precisely with the quality or extent of health care. It is not possible to meaningfully summarize access to care in MIC as a whole, but, as with LIC, there are some commonalities. Access to care is much better in the large urban regions that are developing rapidly in MIC, but in other parts of the country the situation may differ little from that described for LIC (e.g., in China, India and Brazil. This is the probable explanation for the absence of a change in mortality rates for all childhood cancers in 11 MIC in Latin America between 1985 and 2007, a period when mortality rates were steadily declining in North America and Japan [41]. Even now, mortality rates in Latin America are twice those of North America [41]. Many non-medical issues, such as access to secondary educational facilities, limitations in university places and incomplete coverage of academic disciplines (including medical specialities), influence the availability of human capital for childhood cancer care [42]. These non-medical factors also relate to the level of development and vary markedly among and within countries. The migration of health workers to countries able to offer better professional and personal opportunities either before or after specialist training is another oft-debated issue. Many assume that training in HIC is advantageous, but many trainees stay abroad permanently. This is sufficiently common in LMIC that the WHO has issued guidelines on recruitment policies to be used by HIC [43]. Sometimes health care providers are tempted to perform fee-for-service work in their own country, or to work in well-equipped private hospitals, which provide care only for those who can afford it (costs in for-profit private systems are generally higher). This further reduce the work force that provides for the poor. Professional societies, even if very small initially, can help to counteract the drain on resources for the poor by enhancing job satisfaction, e.g., by facilitating professional discussions among their members, promoting continuing education, fostering the development of research networks, and collaborating with similar organizations in adjacent countries. For example, the Turkish Pediatric Oncology Group has evolved into a strong national force for education, research collaboration and cancer registration [44]. Financing health systems Universal insurance coverage with pooling of funds is strongly recommended by WHO as a means of financing health systems [45], since it creates equity with respect to access to care. Insurance also overcomes the obstacle of the cost of diagnosis and therapy that is a frequent reason for late diagnosis and, consequently, advanced disease or refusal of care (WHO states that 100 million people are pushed into poverty each year because they must pay “out-of-pocket” for health care they cannot afford). But more countries are beginning to develop national insurance programs [46-48]. In Mexico, the introduction by the government in 2006 of insurance for the most socioeconomically deprived families (50% of the population), the development of standard treatment protocols and the accreditation of 49 programs for childhood cancer resulted in a reduction of abandonment from 35% to 4% although by 2010, a reduction in mortality had still not been observed [49-52]. Nigeria, and Rwanda have also introduced national insurance programs with tangible benefits [46,47] and China initiated an insurance program for certain leukemias treated at government-approved centers in 2010, along with many additional measures to improve healthcare coverage [53,54] These examples demonstrate that more equitable access to care and reduction in abandonment of therapy by families can be achieved when there is the political will. But insurance alone is not enough, and must be accompanied by improvements in health systems and the quality of care. These changes may require many years.

10

Models of Care and the Need for Research Childhood cancer is one of the success stories in the history of cancer treatment, with 5-year survival rates of 80% or more now being achieved in HIC (article 1). And because children have their entire lives before them, saving the lives of children has a much greater impact on economic development than cancer care in the elderly. Although cure rates are generally very low in LIC, treatment of curable childhood cancers is highly cost-effective even if only possible in a very few institutions [55]. Essentially all countries, however poor, make some attempt to care for at least some children with cancer (e.g., a subset of cases, such as BL or ALL) even though they lack appropriately trained specialists. But when trained medical and nursing staff are so few (e.g., in LIC and many regions in MIC) the diagnosis made and care given may be poor or non-existent, even in major university hospitals. Logical steps in addressing this situation are to first improve the quality of care in existing institutions treating children with cancer, then use these centers to train staff from centers elsewhere in the country, perhaps as part of a national cancer control plan (Panel). An assessment of the current national situation is a first essential, and a decision made as to whether to first focus on a single center and/or a single disease is all that is possible. Even one center can provide a model, serve as a focal point for services and teaching activities and initiate hospital networks that improve health systems [56]. Such plans may differ according to the available resources, and their success is more likely if supported by the government and/or a local champion as well as by external support [57]. In MIC, when high quality facilities already exist, the approach is not different, except that the initial resources are likely to be better and external aid may not be required [58]. Prompt referral and early diagnosis are essential to reduce the fraction of patients with advanced disease, and in some cases, e.g. retinoblastoma, to preserve function [59-61]. Diagnosis Correct diagnosis is the foundation on which all subsequent management rests. Unfortunately, in the poorest countries, there are too few pathologists to diagnose all cases in the country and even many MIC have limited diagnostic capabilities – because of lack of training in pediatric cancer diagnosis or the unavailability of special pathological techniques. Although tissue blocks or slides are sometimes sent to centers in other countries for diagnosis, this does not represent a sustainable solution. Approaches to improving diagnosis may also need to be focused initially, as with clinical care, on one or a small number of centers. Web-based tools (telepathology) can facilitate this process and digital connections between local pathologists and international experts, or the use of telepathology in-country can help greatly. On-line educational tools can also be used to improve access (of patients and doctors) to information and add value to formal education [62,63]. Management Pediatric oncology programs may be located in cancer centers or hospitals with pediatric departments. If the latter, it is critical to physically separate in-patient oncology patients from general hospital patients who may have infectious diseases that are readily transmitted to neutropenic patients, resulting in a hiigh likelihood of death [64-68]. Overcrowding because of the limited locations where children can be treated is a frequent problem and hygienic facilities and standardized procedures for disposal of contaminated materials, avoidance of re-use of needles, as well as prompt recognition and treatment of the common toxic effects of chemotherapy are basic elements of care without which toxic deaths will be frequent. A mechanism for assuring the availability of chemotherapy is essential, but all too often, in LIC,

11

drug supplies are variable. In some countries, it may be preferable for haematologists to treat haematological malignancies. Psychosocial and social assistance provides family support and may decrease the risk of abandonment. Much of the research conducted in the USA and Europe in the second half of the 20th century involved identifying subsets of a disease (staging) that may be treated with less therapy without detriment, or in whom present therapy is inadequate. Unfortunately, in many LMIC, staging is imprecise because of lack of access to imaging techniques or specialists in their use; ultrasound imaging is more widely available. However, the lack of precise staging does not mean that patients cannot be effectively treated, particularly when local therapy (surgery or radiation) is not involved. The same chemotherapy may be given to all patients and the results analyzed in order to try to identify clinical or easily measured biochemical risk factors that can be used to triage future patients into different treatment groups [11,70]. Local staff and external experts should, in any event, be aware that prognostic factors validated in HIC may be less predictive in LMIC for a number of reasons: a) the majority of patients have advanced disease, b) the biology (genetics) of tumour and patient may differ and 3) co-morbidities such as malnutrition and infections, including hepatitis, tuberculosis, virus and fungal infections, which lead to poor therapy tolerance are more common [11,66-68,71,72]. Remaining in hospital for long periods may also increase the risk of nosocomial infection and the likelihood of antibiotic resistance [11,67]. Some treatment approaches developed in HIC may be too intensive or toxic for use in LMIC, and must be adapted to local resources. Pilot studies – with good follow up - are essential to document the toxicity and efficacy of such adapted treatment protocols. Institutional and national collaboration The degree of cooperation within an institute or between institutes is characteristic of low resource settings. At the lower end of the scale, specialists may act independently, as they always have, rarely discussing a care plan, or the cause of death with colleagues. Being unaccountable, they are often out-of-date and have no knowledge of the outcome of their interventions. In many major centers however, interdisciplinary meetings are routine. Evidence-based approaches to care - a major factor in therapeutic advances made in Europe and the USA in the last several decades - may not be possible because of the limited (or complete absence) of training in research methodology, the cost of research infrastructure and the lack of importance given to research for professional advancement. Until recently, limited information was available regarding clinical trials in LMIC. The development of the need to register clinical trials in data bases, e.g. Clinical Trials. Gov (US) has provided more objective data, if incomplete [73]. Most clinical studies that are undertaken in LMIC by academic institutions tend to be retrospective case series, usually exploring the feasibility and efficacy of treatment protocols designed in HIC. Exceptions exist in some major universities and cancer institutes, particularly national cancer institutes in MIC, where randomized studies may be performed [67,71,74-77]. Yet unique research opportunities exist in LMIC, particularly with respect to epidemiological studies (that could lead to prevention) and the development of alternative approaches to treatment elements that are not feasible in LIC or MLIC. There is some evidence that the degree of motivation within the country, particularly of policy makers and senior health care professionals is important. A survey of treatment outcome in 10 developing countries suggested that better results were achieved in countries in which GDP, GNI per capita, the number of physicians and nurses per 1000 population, and most significantly, per capita annual government expenditure on health care, were all associated with a better survival rate [78].

12

International Collaboration While LIC have yet to develop organized approaches to childhood cancer, the greatest obstacle they face is the lack of human capital. This means that both service provision and education are generally rudimentary. Most MIC have some pediatric cancer programs, although much of the development that has taken place has come ultimately from direct or indirect collaboration with foreign institutions [11,70, 79-83] and international experts can still help greatly, particularly by visiting the country to provide education (e.g., via a “twinning program” or collaboration with an international organization) or by undertaking frequent on-line meetings or providing tools for distance learning. Improvements will initially be slow, but will accelerate as resources improve. Results achieved in HIC provide a target to aim for. In the interim, measurable five-year survival rates, even if well below those of HIC, will help to create a dynamic that is not easily ignored by politicians. International collaborators can also help with more general aspects of health system strengthening, including hospital management, health care infrastructure and financing and may also create opportunities to participate in research at an international level, leading to more rapid completion of research studies and an expansion of global research capacity [84-85]. In some cases, they can help to provide capital funding, e.g., for the purchase of radiotherapy equipment [86]. An increasing number of clinical trials relating to drug development include both HIC and LMIC because the requisite number of patients can be more quickly accrued, and products developed more rapidly, but such trials should have relevance to the local population. More research infrastructure in LMIC will doubtless be required as the already sizeable pharmaceutical industries of India, China and many other countries develop more products [87]. Globalization of pharmaceutical development, however, raises numerous questions regarding comparability of results in different populations, intellectual property rights and ethics (especially with respect to minors and ensuring access of populations to those products they helped develop) [84,85]. A Model of Care The first need of LMIC in developing improved care of children with cancer is capacity building, taking account of existing capacity and focusing initially on ensuring that at least some patients are treated effectively. Care, however, begins in the community and a plan to improve childhood cancer care should link the public, primary care providers, a wide range of specialists, health workers and policy-makers. The goal should be to incorporate, over time, the various elements listed in the Panel at a rate and in a sequence that is tailored to the country’s needs and includes the government and a wide range of stake-holders, including, if necessary, international experts. Unlike many other health challenges, pediatric cancer is a largely tractable problem that is, relatively speaking, quite small, although the number of patients is offset by the potential complexity and cost of diagnosis and treatment. Interestingly, the cost of treating one ALL patient in Holland and Nordic countries has been reported to be in the region of $100,000 [88,89], whereas in major cities in China, the cost is much lower $4-12,000 [90,91]. While such calculations should be interpreted with caution, these figures suggest that total treatment costs for the 84% of childhood cancers occurring in LMIC are likely to be significantly lower than the cost of treating the remaining 16% of childhood cancers in HIC. But HIC may be “overspending”: the Institute of Medicine recently reported that a third ($765 billion in 2009) of health care costs in the USA is “wasted” [92]). In contrast, governments of LMIC will surely

13

need to Increase expenditure on education and health care, with or without external aid, in order to create the first essential of sustainable socioeconomic development – human capital [93].

References

1. World Population Prospects, the 2010 Revision. United Nations Department of Economic and Social Affairs. Population Division, Population Estimates and Projections Section. http://esa.un.org/unpd/wpp/Sorting-Tables/tab-sorting_ageing.htm. and

http://esa.un.org/unpd/wpp/Excel-Data/population.htm . Accessed on 31-12-2012

2. The World Bank. Data. http://data.worldbank.org/about/country-classifications. Accessed on 12-12-2012

3 Ferlay, J., Shin, H. R., Bray, F., Forman, D., Mathers, C., and Parkin, D. M. GLOBOCAN 2008 v2.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 10. http://globocan.iarc.fr. Accessed on 31-12-2012

4. World Health Organization Global Health Observatory Data Repository. http://apps.who.int/ghodata/?vid=10012. Accessed on 31-12-2012

International Incidence of Childhood Cancer, Volume 2 . Parkin DM, Kramarova E, Draper GJ, Masuyer E, Michaelis J, Neglia JP et al., editors. IARC Scientific Publications No. 144. International Agency for Research on Cancer. 1998. Lyon.

6. Valsecchi MG, Steliarova-Foucher E. Cancer registration in developing countries: luxury or necessity? Lancet Oncol 2008; 9:159-167

7. Aka P, Kawira E, Masalu N, Emmanuel B, Brubaker G, Magatti J, Mbulaiteye SM. Incidence and

trends in Burkitt lymphoma in northern Tanzania from 2000 to 2009. Pediatr Blood Cancer. 2012

;59:1234-8.

8. Amir H, Kaaya EE, Manji KP, Kwesigabo G, Biberfeld P. Kaposi's sarcoma before and during a

human immunodeficiency virus epidemic in Tanzanian children. Pediatr Infect Dis J. 2001;20:518-

21.

9, Reid BC, Ghazarian AA, Demarini DM, Sapkota A, Jack D, Lan Q et al. Research opportunities for cancer associated with indoor air pollution from solid-fuel combustion. Environ Health Perspect 2012; 120:1495-1498

10. Howard SC, Metzger ML, Wilimas JA, Quintana Y, Pui CH, Robison LL et al. Childhood cancer epidemiology in low-income countries. Cancer 2008; 112:461-472

11. Magrath I, Shanta V, Advani S, Adde M, Arya LS, Banavali S et al. Treatment of acute lymphoblastic leukaemia in countries with limited resources; lessons from use of a single protocol in India over a twenty year period. Eur J Cancer 2005; 41:1570-1583

14

12. Ramot B, Magrath I. Hypothesis: the environment is a major determinant of the immunological sub-type of lymphoma and acute lymphoblastic leukaemia in children. Br J Haematol 1982; 50:183-189

13. Brown WM, Doll R. Leukaemia in Childhood and Young Adult Life. Br Med J 1961; 1:981-988

14. Arya LS, Padmanjali KS, Sazawal S, Saxena R, Bhargava M, Kulkarni KP et al. Childhood T-lineage acute lymphoblastic leukemia: management and outcome at a tertiary care center in North India. Indian Pediatr 2011; 48:785-790

15. Rajalekshmy KR, Abitha AR, Anuratha N, Sagar TG. Time trend in frequency of occurrence of major immunophenotypes in paediatric acute lymphoblastic leukemia cases as experienced by Cancer Institute, Chennai, south India during the period 1989-2009. Indian J Cancer 2011; 48:310-315

16. Hussein H, Sidhom I, Naga SA, Amin M, Ebied E, Khairy A et al. Outcome and prognostic factors of acute lymphoblastic leukemia in children at the National Cancer Institute, Egypt. J Pediatr Hematol Oncol 2004; 26:507-514

17. Kamel AM, Assem MM, Jaffe ES, Magrath I, boul Enein MI, Hindawy DS. Immunological phenotypic pattern of acute lymphoblastic leukaemia in Egypt. Leuk Res 1989; 13:519-525

18. Greaves MF. Aetiology of acute leukaemia. Lancet 1997; 349:344-349

19. Rao CR, Gutierrez MI, Bhatia K, Fend F, Franklin J, Appaji L et al. Association of Burkitt's lymphoma with the Epstein-Barr virus in two developing countries. Leuk Lymphoma 2000; 39:329-337

20. Magrath I. Epidemiology: clues to the pathogenesis of Burkitt lymphoma. Br J Haematol 2012; 156:744-756

21. Karajannis MA, Hummel M, Oschlies I, Anagnostopoulos I, Zimmermann M, Stein H et al. Epstein-Barr virus infection in Western European pediatric non-Hodgkin lymphomas. Blood 2003; 102:4244

22. Rao CR, Franklin J, Lalitha N et al. Frequency of Epstein Barr Virus (EBV) association and p53 expression in childhood and adult Hodgkin’s disease in India. Int J Pediatr Hematol.Oncol 6: 377-386, 2000.

23. Naresh KN, Johnson J, Srinivas V, Soman CS, Saikia T, Advani SH et al. Epstein-Barr virus association in classical Hodgkin's disease provides survival advantage to patients and correlates with higher expression of proliferation markers in Reed-Sternberg cells. Ann Oncol 2000; 11:91-96

24. Rodriguez-Pinilla SM, Barrionuevo C, Garcia J, Martinez MT, Pajares R, Montes-Moreno S et al. EBV-associated cutaneous NK/T-cell lymphoma: review of a series of 14 cases from Peru in children and young adults. Am J Surg Pathol 2010; 34:1773-1782

15

25. Subramaniam K, D'Rozario J, Pavli P. Lymphoma and Other Lymphoproliferative Disorders in IBD: A Review. J Gastroenterol Hepatol 2012;

26. Toumi N, Frikha M, Siala W, Khabir A, Karray H, Boudawara T et al. [Juvenile nasopharyngeal carcinoma: anatomoclinic, biologic, therapeutic and evolutive aspects]. Bull Cancer 2010; 97:427-433

27. Ni YH, Chang MH, Wu JF, Hsu HY, Chen HL, Chen DS. Minimization of hepatitis B infection by a 25-year universal vaccination program. J Hepatol 2012; 57:730-735

28. McMahon BJ, Bulkow LR, Singleton RJ, Williams J, Snowball M, Homan C et al. Elimination of hepatocellular carcinoma and acute hepatitis B in children 25 years after a hepatitis B newborn and catch-up immunization program. Hepatology 2011; 54:801-807

29. Tukei VJ, Kekitiinwa A, Beasley RP. Prevalence and outcome of HIV-associated malignancies among children. AIDS 2011; 25:1789-1793

30. Parkin DM, Garcia-Giannoli H, Raphael M, Martin A, Katangole-Mbidde E, Wabinga H et al. Non-Hodgkin lymphoma in Uganda: a case-control study. AIDS 2000; 14:2929-2936

31. Wakeford R. Childhood leukaemia following medical diagnostic exposure to ionizing radiation in utero or after birth. Radiat Prot Dosimetry 2008; 132:166-174

32. Mole RH. Childhood cancer after prenatal exposure to diagnostic X-ray examinations in Britain. Br

J Cancer. 1990;62:152-68.

33. Piciu D, Piciu A, Irimie A. Thyroid cancer in children: a 20-year study at a Romanian oncology institute. Endocr J 2012; 59:489-496

34. Steliarova-Foucher E, Stiller C, Kaatsch P, Berrino F, Coebergh JW, Lacour B et al. Geographical patterns and time trends of cancer incidence and survival among children and adolescents in Europe since the 1970s (the ACCISproject): an epidemiological study. Lancet 2004; 364:2097-2105

35. Little MP, Wakeford R, Kendall GM. Updated estimates of the proportion of childhood leukaemia

incidence in Great Britain that may be caused by natural background ionising radiation. Radiol

Prot. 2009;29:467-82.

36. Silbermann M, Al-Hadad S, Ashraf S, Hessissen L, Madani A, Noun P et al. MECC regional initiative in pediatric palliative care: Middle Eastern course on pain management. J Pediatr Hematol Oncol 2012; 34 Suppl 1:S1-11

37. The World Health Report 2008- Primary Health Care Now More Than Ever. http://www.who.int/whr/2008/en/index.html Last Accessed 31-12-2012

16

38. Howard SC, Pedrosa M, Lins M, Pedrosa A, Pui CH, Ribeiro RC et al. Establishment of a pediatric oncology program and outcomes of childhood acute lymphoblastic leukemia in a resource-poor area. JAMA 2004; 291:2471-2475

39. Potterton L. Access to affordable radiotherapy. IAEA Bulletin 2010; 52:10-11

40. Health Tourism. http://www.health-tourism.com/pediatric-oncology/philippines/ Accessed 31-

12-2012

41. Chatenoud, L, Bertuccio P, Bosetti, C, Levi, F et al., et. al., Childhood Cancer Mortality in America, Asia, and Oceania, 1970 Through 2007Cancer 2010;116:5063–74.

42. Education for sustainable development. UNESCO http://www.unesco.org/new/en/education/themes/leading-the-international-agenda/education-for-sustainable-development/browse/2/

43. WHO practice on the international recruitment of health personnel. http://www.who.int/hrh/en/.

Accessed on 31-12-2012

44. Kebudi R. Pediatric oncology in Turkey. J Pediatr Hematol Oncol 2012; 34 Suppl 1:S12-S14

45. Health Systems Financing: the Path to Universal Health Coverage. Plan of Action. World Health Organization. http://www.who.int/health_financing/Health_Systems_Financing_Plan_Action.pdf Accessed 31-12-2012.

46. Obalum DC, Fiberesima F. Nigerian National Health Insurance Scheme (NHIS): An overview. Niger Postgrad Med J 2012; 19:167-174

47. Lu C, Chin B, Lewandowski JL, Basinga P, Hirschhorn LR, Hill K et al. Towards universal health coverage: an evaluation of Rwanda Mutuelles in its first eight years. PLoS One 2012; 7:e39282

48. Perez-Cuevas R, Doubova SV, Zapata-Tarres M, Flores-Hernandez S, Frazier L, Rodriguez-Galindo C et al. Scaling up cancer care for children without medical insurance in developing countries: The case of Mexico. Pediatr Blood Cancer 2012;

49. Knaul FM, Gonzalez-Pier E, Gomez-Dantes O, Garcia-Junco D, Arreola-Ornelas H, Barraza-Llorens M et al. The quest for universal health coverage: achieving social protection for all in Mexico. Lancet 2012; 380:1259-1279

50. Perez-Cuevas R, Chertorivski S, Cortes-Gallo G, Rodriguez-Ortega E, Caballero F, Munoz-Hernandez O. Lessons learned from the first evaluation of the Medical Insurance for a New Generation: bridging research and policy. Salud Publica Mex 2012; 54 Suppl 1:S90-S97

51. Ribeiro RC. Impact of the Mexican government's system of social protection for health, or Seguro Popular, on pediatric oncology outcomes. Pediatr Blood Cancer 2012;

17

52. Rivera-Luna R, Correa-Gonzalez C, Altamirano-Alvarez E, Sanchez-Zubieta F, Cardenas-Cardos R, Escamilla-Asian G et al. Incidence of childhood cancer among Mexican children registered under a public medical insurance program. Int J Cancer 2012; Pediatr Infect Dis J. 2012 Sep 25. [Epub ahead of print]

53. Tang S, Tao J, Bekedam H. Controlling cost escalation of healthcare: making universal health coverage sustainable in China. BMC Public Health 2012; 12 Suppl 1:S8

54. Li Y, Wu Q, Xu L, Legge D, Hao Y, Gao L et al. Factors affecting catastrophic health expenditure and impoverishment from medical expenses in China: policy implications of universal health insurance. Bull World Health Organ 2012; 90:664-671

55. Bhakta N, Martiniuk AL, Gupta S, Howard SC. The cost effectiveness of treating paediatric cancer in low-income and middle-income countries: a case-study approach using acute lymphocytic leukaemia in Brazil and Burkitt lymphoma in Malawi. Arch Dis Child 2012;

56. Howard SC, Ribeiro RC, Pui CH. Strategies to improve outcomes of children with cancer in low-

57. Epelman S. Pediatric oncology in Brazil. Pediatr Hematol Oncol 1998; 15:377-383

58. Luna-Fineman S, Barnoya M, Bonilla M, Fu L, Baez F, Rodriguez-Galindo C. Retinoblastoma in Central America: report from the Central American Association of Pediatric Hematology Oncology (AHOPCA). Pediatr Blood Cancer 2012; 58:545-550

59. Howard SC, Wilimas JA. Delays in diagnosis and treatment of childhood cancer: where in the world are they important? Pediatr Blood Cancer 2005; 44:303-304

60. Leander C, Fu LC, Pena A, Howard SC, Rodriguez-Galindo C, Wilimas JA et al. Impact of an education program on late diagnosis of retinoblastoma in Honduras. Pediatr Blood Cancer 2007; 49:817-819

61. Epelman S. Preserving vision in retinoblastoma through early detection and intervention. . Curr Oncol Rep. 2012;14:213-9.

62. Qaddoumi I, Mansour A, Musharbash A, Drake J, Swaidan M, Tihan T et al. Impact of telemedicine on pediatric neuro-oncology in a developing country: the Jordanian-Canadian experience. Pediatr Blood Cancer 2007; 48:39-43

63. Qaddoumi I, Ribeiro RC. Can communication technologies reduce inequalities in cancer care? Lancet Oncol 2011; 12:224

64. Gavidia R, Fuentes SL, Vasquez R, Bonilla M, Ethier MC, Diorio C et al. Low socioeconomic status is associated with prolonged times to assessment and treatment, sepsis and infectious death in pediatric Fever in el salvador. PLoS One 2012; 7:e43639

18

65. Gupta N, Seth T, Mishra P, Mahapatra M, Rathi S, Kapoor R et al. Treatment of acute myeloid leukemia in children: experience from a tertiary care hematology centre in India. Indian J Pediatr 2011; 78:1211-1215

66. Gupta S, Bonilla M, Valverde P, Fu L, Howard SC, Ribeiro RC et al. Treatment-related mortality in children with acute myeloid leukaemia in Central America: incidence, timing and predictors. Eur J Cancer 2012; 48:1363-1369

67. Raje NS, Rao SR, Iyer RS, Kelkar RS, Pai SK, Nair CN et al. Infection analysis in acute lymphoblastic leukemia: a report of 499 consecutive episodes in India. Pediatr Hematol Oncol 1994; 11:271-280

68. Gupta S, Antillon FA, Bonilla M, Fu L, Howard SC, Ribeiro RC et al. Treatment-related mortality in children with acute lymphoblastic leukemia in Central America. Cancer 2011;

69. Spinetta JJ, Jankovic M, Masera G, Ablin AR, Barr RD, Ben Arush MW et al. Optimal care for the child with cancer: A summary statement from the SIOP Working Committee on Psychosocial Issues in Pediatric Oncology. Pediatr Blood Cancer 2009; 52:904-907

70. Ngoma T, Adde M, Durosinmi M, Githang'a J, Aken'Ova Y, Kaijage J et al. Treatment of Burkitt lymphoma in equatorial Africa using a simple three-drug combination followed by a salvage regimen for patients with persistent or recurrent disease. Br J Haematol 2012; 158:749-762

71. Somjee S, Pai S, Parikh P, Banavali S, Kelkar R, Advani S. Passive active prophylaxis against Hepatitis B in children with acute lymphoblastic leukemia. Leuk Res 2002; 26:989-992

72. Bakhshi S, Padmanjali KS, Arya LS. Infections in childhood acute lymphoblastic leukemia: an

analysis of 222 febrile neutropenic episodes. Pediatr Hematol Oncol. 2008;25:385-92.

73. Califf, MD; Zarin MD, Kramer JM et al. Characteristics of Clinical Trials Registered in ClinicalTrials.gov, 2007-2010. JAMA. 2012;307:1838-1847.

74. Boschmonar MG, Alvarez YG, Garcia AM, Soto TN, Roger MC, Garrote LF. Childhood cancer survival in Cuba. Eur J Epidemiol 2000; 16:763-767

75. Swaminathan R, Rama R, Shanta V. Childhood cancers in Chennai, India, 1990-2001: incidence and survival. Int J Cancer 2008; 122:2607-2611

76. Wiangnon S, Veerakul G, Nuchprayoon I, Seksarn P, Hongeng S, Krutvecho T et al. Childhood cancer incidence and survival 2003-2005, Thailand: study from the Thai Pediatric Oncology Group. Asian Pac J Cancer Prev 2011; 12:2215-2220

77. Bao PP, Zheng Y, Wu CX, Peng P, Gong YM, Huang ZZ et al. Population-based survival for childhood cancer patients diagnosed during 2002-2005 in Shanghai, China. Pediatr Blood Cancer 2012; 59:657-661

19

78. Ribeiro RC, Steliarova-Foucher E, Magrath I, Lemerle J, Eden T, Forget C et al. Baseline status of paediatric oncology care in ten low-income or mid-income countries receiving My Child Matters support: a descriptive study. Lancet Oncol 2008; 9:721-729

79. Israels T, Moreira C, Scanlan T, Molyneux L, Kampondeni S, Hesseling P et al. SIOP PODC: Clinical guidelines for the management of children with Wilms tumour in a low income setting. Pediatr Blood Cancer 2013; 60:5-11

80. Hesseling P, Israels T, Harif M, Chantada G, Molyneux E. Practical recommendations for the management of children with Endemic Burkitt Lymphoma (BL) in a resource limited setting. Pediatr Blood Cancer 2012;

81. Molyneux E, Davidson A, Orem J, Hesseling P, Balagadde-Kambugu J, Githanga J et al. The management of children with Kaposi sarcoma in resource limited settings. Pediatr Blood Cancer 2012;

82. Naresh KN, Raphael M, Ayers L, Hurwitz N, Calbi V, Rogena E et al. Lymphomas in sub-Saharan Africa - what can we learn and how can we help in improving diagnosis, managing patients and fostering translational research? Br J Haematol 2011;

83. Gopal S, Wood WA, Lee SJ et al., Meeting the challenge of hematologic malignancies in sub-Saharan Africa. Blood. 2012;119:5078-87.

84. Dunne J, Murphy MD, Rodriguez WJ. The Globalization of Pediatric Clinical Trials. Pediatrics 2012;

85. Lu F, Suda KJ, Marks KE. Geographic location of antineoplastic agent clinical trials conducted in developed and developing countries. Int J Clin Pharm 2012;

86. Salminen E, Izewska J, Andreo P et al. IAEA’s role in the global management of cancer-focus on

upgrading radiotherapy services. Acta Oncologica, 2005; 44: 816--824

87. Perkovic V, Patil V, Wei L, Lv J, Petersen M, Patel A. Global randomized trials: the promise of India and China. J Bone Joint Surg Am 2012; 94 Suppl 1:92-96

88. van Litsenburg RR, Uyl-de Groot CA, Raat H, Kaspers GJ, Gemke RJ. Cost-effectiveness of

treatment of childhood acute lymphoblastic leukemia with chemotherapy only: the influence of

new medication and diagnostic technology. Pediatr Blood Cancer. 2011;57:1005-10.

89. Rahiala J, Riikonen P, Kekäläinen L, Perkkiö M. Cost analysis of the treatment of acute childhood

lymphocytic leukaemia according to Nordic protocols. Acta Paediatr. 2000;89:482-7.

90. Luo XQ, Ke ZY, Guan XQ, Zhang YC, Huang LB, Zhu J. The comparison of outcome and cost of three

protocols for childhood non-high risk acute lymphoblastic leukemia in China. Pediatr Blood

Cancer. 2008;51:204-9.

20

91. Liu Y, Chen J, Tang J, Ni S, Xue H, Pan C. Cost of childhood acute lymphoblastic leukemia care in

Shanghai, China. Pediatr Blood Cancer. 2009;53(4):557-62.

92 Best Care at Lower Cost: The Path to Continuously Learning Health Care in America. A report of

the Institute of Medicine, USA. Released, Sept 6th 2012. http://www.iom.edu/Reports/2012/Best-

Care-at-Lower-Cost-The-Path-to-Continuously-Learning-Health-Care-in-America.aspx. Accessed

31-12-2012

93. Martin G, Grant A, D'Agostino M. Global health funding and economic development. Global

Health. 2012 Apr 10;8:8. doi: 10.1186/1744-8603-8-8.

21

Table 1. Selected development indicators in economies classified by the World Bank according to

gross national income (GNI) per capita, using the Atlas method (current US dollars). Data from

http://databank.worldbank.org/ddp/home.do?Step=1&id=4 for 2010

Indicator High High

(OECD)

Upper

Middle

Lower

Middle

Low

Number of countries/economies in World Bank income groups 70 31 54 54 36

GNI per capita (Atlas Method, current USD) 38764 40142 5865 1595 533.99

GDP per capita (Current USD) 41062 42220 7326 1882 579

Population (millions) 1127 1033 2472 2494 799

Population (% Rural) 19.7 20 40.2 61.6 72.4

Population (% 0-14) 17.3 17.0 21.9 31.9 39.25

Age dependency ratio (percentage of non-working age population) 49 49.7 42.9 59.2 77.0

Poverty headcount ratio at <$1.25 per day, percent of population - - 9 30 48

Health expenditure per capita, (current USD) 4876 5093 379 70.9 26.9

Health expenditure, total (percentage of GDP) 12.5 12.9 6.0 4.3 5.3

Health expenditure, public (% GDP) 8.2 8.4 3.3 1.7 2.1

Health expenditure, private (% GDP) 4.4 4.5 2.8 2.6 3.2

Health expenditure, public (% total health expenditure) 65 65 54.4 39.5 38.8

Out-of-pocket health expenditure (% total health expenditure) 13.7 13.6 33.3 52.8 48.1

Number of physicians per 1000 people 2.7 2.8 1.7 0.8 0.2

Number of nurses or midwives per 1000 people 7.0 7.2 2.6 1.5 0.5

Literacy rate (% females 15 and above) 97.9 - 91.3 62.3 56

Literacy rate (% males 15 and above) 98.7 - 93.5 80 69.7

Income groups used in the table are based on the 2011 GNI per capita calculated according to the World Bank Atlas Method (http://data.worldbank.org/about/country-classifications). The groups are: Low Income: $1,025 or less: Low Low Middle Income: $1,026-$4,035: Upper Middle Income: $4,036-

$12,475: High income: $12,476 or more. OECD: Organization for Economic Cooperation and

Development is a group of 34 countries founded in 1961 to promote economic progress and world

trade. Most of these countries are high income countries with a very high index of development (UNDP

classification; see http://hdr.undp.org/en/media/HDR_2011_EN_Complete.pdf)

22

Table 2. Deaths by Cause in Children (both sexes) aged 5 to 14 years, 2008

Data from http://apps.who.int/ghodata/?vid=10012 Global Burden of Disease: Causes of Death 2008.2 Percentages relate to the disease related deaths (except disease related deaths, which relate to deaths from all causes)

World

%

High Income %

Upper Middle Income

% Lower Middle Income

% Low Income %

Population 5-14 yrs

1,231,558,000 127,027 166,082,000 715,046,000 206,201,000

All Causes 1,265,306 17,397 66,973 721,871 459,065

Disease related 1,059,408 10,251 42,987 508,492 360,305

Infections, Parasites

496,605 46.9 608 5.9 14,738 11.0 269,783 53.0 211,475 58.7

Non-communicable diseases

222,784 21.0 9,126 89.0 24,030 55.9 136,270 26.8 53,357 14.8

Malignant Neoplasms

53,464 14.6 3,317 32.0 7,984 18.6 30,541 6.0 11,622 3.2

23

Table 3. Deaths by Cause in Children (both sexes) aged 0-4 years, 2008

World %

High Income

% Upper Middle Income

% Lower Middle Income

% Low Income %

Population 0-14 yrs

634,176,000 62,609,000 83,258,000 382,776,000 121,531,000

All Causes 8,327,552 85,237 405,329 4,586,599 3,249,547

Disease related

8,044,722 77,766 384,684 4,441,085 3,140,383

Infections/Parasitic

3,152,673 39.2 3,849 4.9 82,902 23.4 1,568,202 35.3 1,497,720 47.7

Non-Communicable Diseases

614,759 7.6 31,769 40.8 87,542 22.7 347,009 7.8 148,439 4.7

Malignant Neoplasms

40,061 0.50 1,901 2.4 4,728 1.2 23,341 0.5 10,091 0.3

Data from http://apps.who.int/ghodata/?vid=10012 Global Burden of Disease: Causes of Death 2008.2 Percentages relate to the disease related deaths

24

Table 4. Age specific incidence rates of cancer by Human Development Index (HDI) 1

Population Group

Incident cases in 0-14 year olds

Age-specific incidence

(0-14 years) per 100,000

Cumulative risk*

%

World 175,058 10.4 0.16

Very High HDI 24,855 14.9 0.22

High HDI 25,080 12.6 0.19

Medium HDI 110,027 9.5 0.14

Low HDI 15,035 9.7 0.14

1 The United Nations HDI is based on two educational indicators , life expectancy at birth and the logarithm of income which are aggregated into a single index (HDI). See http://hdr.undp.org/en/statistics/hdi/ for details. *Risk of developing cancer during a lifetime.

25

PANEL: Elements of a model of pediatric cancer control in low- and middle-income countries

Public health policy

National Cancer Control Plan

Governmental decision to control cancer, including in children. Definition of priorities in wide consultation with the field and international experts.

Cancer Registration Provision of data on cancer incidence from population-based cancer registry relevant to local policy and research needs International collaboration to make comparisons and improve methodology

Routine national statistics

Regular population counts and estimates and publication of detailed results, including for children and infants (age 0) Death certification and reporting of causes of death to WHO

Health insurance coverage

Health insurance program in which pre-payment and pooling of costs allows the family is not deterred from a medical consultation on account of cost. Universal insurance is an important element of sustainability Subsidized travel in order that lack of money does not prevent diagnosis at the earliest possible time

Primary health care

Sufficient coverage by primary care providers to permit good access to primary care

Referral pathway Prompt referral to a tertiary care center when necessary

Centres of excellence

Centralisation of treatment, especially for rare cases

Health professionals

Awareness of childhood cancer

A population educated in the signs and symptoms that could be caused by cancer and free of stigmata associated with cancer

Primary care Primary care providers educated in the signs and symptoms that could be caused by cancer and who know where to refer patients for further investigations

Oncology team A tertiary care facility with the necessary specialists, including pediatric oncology, radiation therapy, surgical oncology, oncology nursing, as well as non-oncologists who deal with discrete sets of diseases (haematology, neurology, orthopedics, neurosurgery etc.). Multidisciplinary communication among all specialists through regular “tumor board” meetings and patient rounds to assure that each patient gets the best possible treatment available. The tumor board can also serve additional functions such as education of junior staff and trainees.

Training and professional support

Professional organizations that provide continuing educational opportunities and discussion of issues that need to be addressed within the region or country, with a special emphasis on local issues and concerns, such as abandonment of treatment, cancer registration, etc.

Diagnosis and risk stratification

Access to prompt biopsy of suspected cancer

Secondary care facilities capable of performing biopsy where necessary and who know the importance of prompt pathological diagnosis

Pathology service Access to a pathology service with well-trained pathologists, the necessary equipment, and reagents to make an accurate diagnosis. Prompt reporting of pathology results to the clinician in charge. Consultative network.

26

Staging and risk stratification

Access to necessary imaging procedures to identify the stage or degree of spread of the cancer. This will vary markedly according to socioeconomic status of the country

Treatment

Treatment protocols adapted to local conditions

Standardized protocols, previously agreed upon for all pediatric cancers, and ideally, partnership in a disease-specific network or cooperative group capable of conducting clinical research, even if primarily designed to assess suitability of a regimen in the local setting.

Access to medications and supportive care

An efficient pharmacy which procures drugs prior to running out, and a national or regional system that ensures that all needed drugs are available, including those required for pain control, such as opioids

Support infrastructure

Psychosocial and financial

Psychosocial support with mechanisms to avoid lost earnings, local lodging when home is far away, and access to social assistance and advice. Presence of a patient advocate to identify non-medical family and patient needs. Subsidized travel and lodging to make access to tertiary care feasible.

Outcome evaluation and quality improvement

A system of data collection comprising a hospital registry, as well as necessary data bases and data managers able to efficiently collect data. A systematic approach and appropriate resources (tracking officer) to ensure follow up of all patients to determine the outcome of therapy Collection of clinically relevant data Evaluating of delays in cancer diagnosis and referral, abandonment pattern, etc. Long-term follow-up of cancer patients and evaluation of their survival

Advocacy and sustainability

Awareness of childhood cancer

A population educated in the signs and symptoms that could be caused by cancer and free of stigmata associated with cancer and that early diagnosis is important for the outcome

Non-governmental organizations

Local or national NGOs that advocate for pediatric cancer and may assist in providing funding for necessary drugs, salaries, construction, etc. Such NGO’s may play an important role in ensuring sustainability and advocating for improvements in the health system and the development of cancer control plans, and the means to carry them out

External collaboration

Associations with international professional groups or organizations in order to ensure access to relevant new information concerning the efficiency of the health system, and who may be in a position to affect, for example, drug prices through bulk purchasing and provide access to participation in international studies or highly skilled trainers and educators who can help improve “medical” education as well as advise governments on the development of the national health system and the public on healthy living styles. Twinning programs with other centers can facilitate development of specific services and treatment protocols. Participation in international research studies.

NGO, non-governmental organization NB. The above elements of a pediatric cancer control program represent an ideal list which will need to be adapted to the socioeconomic status of the country.

27

Figure Legends

Figure 1. Model of the marked changes in population age-structure resulting from the Industrial

Revolution. Developing countries are in phases 2 and 3 and more developed countries in

phases 4 and 5. Stage 1: High birth and death rates: pre-industrial; Stage 2: Decline in death

rate: improved food and water supply, better public health and hygiene; Stage 3: Decline in birth

rate: urbanization; Stage 4: Low birth and death rates; Stage 5: Contraction: birth rate lower

than death rate. Population pyramids are shown below.

Figure 2. Incidence rates of childhood cancer (0-14 years) and proportions (%) of childhood

cancer, males and females combined, in selected populations. Data from ref 3.

Figure 3. Mortality rates of childhood cancer (0-14 years) and percentage of cancer deaths,

males and females combined, that occur in childhood in selected populations. Data from ref 3.

Figure 4. Figure 4. Childhood cancer incidence and mortality in different parts of world in 2008.

Data from Ref 3.

Figure 5. Incidence rates of childhood cancer registered in the countries shown in the 1980s and assembled in an international comparative study. An update is presently in progress. ASR (World). Data from ref. 5

Slide 1.jpgClick here to download high resolution image

Slide2.JPGClick here to download high resolution image

Slide3.JPGClick here to download high resolution image

Slide4.JPGClick here to download high resolution image

Slide5.JPGClick here to download high resolution image

INCTR, SJCRH

Paediatric Cancer in Countries with Limited Resources Manuscript 3

Ian Magrath et al. 11/23/2012

Ian Magrath,1 Eva Steliarova-Foucher,2 Raul Ribiero,3 M’Hamed Harif,4 Chi-Kong Li,5

Rejin Kebudi, 6 Scott MacFarlane,7 Scott C. Howard3 1 International Network for Cancer Treatment and Research, Brussels, Belgium 2 International Agency for Research on Cancer, Lyon, France 3 St Jude Children’s Research Hospital, Memphis, USA 4 Centre Hospitalier Universitaire Mohhamed VI, Marrakech, Morocco 5 Prince of Wales Hospital, Hong Kong, China 6 Cerrahpasa Institute, Istanbul Turkey 7 Starship Children’s Health, Auckland, New Zealand

Pediatric Cancer in Countries with Limited Resources FINALRevised2.docx

2

Abstract (Word Count 207 310)

Patterns of cancer across the world have undergone dramatic changes as a consequence of the industrial revolution, which was firmly established in Great Britain by 1850. The spread of industrialization, however, has been slow, and the economies of most countries remain in an early or intermediate industrial phase. This situation is reflected in the World Bank’s classification of countries by gross national income (GNI) per capita and the UNDP’s Human Development Index (HDI), which combines four indicators of health, education and wealth into a composite measure of development. Lower levels of development are characterized by poverty, limited access to education and health care, due in large part to scattered rural populations, too few health care providers, and weak health systems and poverty. And since mortality declines before birth rates as countries industrialize, Lless developed countries also have faster growing, younger populations and therefore a higher proportion of their cancer in children and adolescents than more developed countries. Exposure to the consequences of industrialization, however, is minimal in the rural regions of the poorest countries, and chronic infections and infestations, which remain a majorthe leading cause of disease-related death in all except the high income countries, are probablymay also be major determinants of childhood cancer in these regions. This is fertile ground for research. A component of development is urbanization, which results in changes in lifestyle and new exposures, with a resultant increase in the incidence of many cancers and other non-communicable diseases. However, only in high income countries has cancer overtaken communicable diseases as the leading cause of disease-related death. Here, we examine discuss childhood cancer in relationship to development at a global level, discuss needs and propose approaches strategies that to could result in improveding the care of children with cancer in countries with limited resources., whose health services are both weak and ill-adapted to cancer care. Education of the public, more and better trained health professionals, strengthened cancer services, locally relevant research, regional hospital networks, international collaboration and health insurance are essential components of an improved policy designed to meet these challenges.model of care.

Formatted: Left, Line spacing: single

3

Epidemiology of Childhood Cancer in Developing Countries Childhood cancer (generally considered by descriptive epidemiologists as cancer in 0-14 year olds) constitutes only a small fraction of the global cancer burden, but the majority of childhood cancers occur the burden falls mainly on in the low and middle income countries (LMIC), where 90% of the world’s children live and access to care is limited (Table !) [1-3]. In 2010, LMIC accounted for 5.7 billion (83%) of the world population (6.9 billion). They have lower median ages and higher proportions of children (20 to 50%) in their populations than high income countries (HIC) in which children comprise less than 20% of the population.where. , iIn 2008, estimates provided by Globocan, a data base developed by the International Agency for Research on Cancer (IARC), indicated that approximately 148,000 (84%) of cancers in children aged cancers (0-14 years occurred in less developed regions, which correspond well to the World Bank’s LMIC [2] and include were estimated to have occurred in the 0-14 year age group. In the same year, there were almost 148,000 incident cases of childhood cancer in less developed countries Africa, Asia, excluding Japan, Latin America and the Caribbean, Melanesia, Micronesia and Polynesia, comprising a total population of 5.5 billion. The equivalent Globcan estimate for children living in more developed regions, which are defined as Europe, Northern America, Australia, New Zealand and Japan, was 28,000. compared to 28,000 in more developed countries These regions, together, have a total population of 1.2 billion and differ little from the HIC as defined by the World Bank.1 [Manuscript 1 of this series]

. A major reason for this is that 84% of the world’s children live in LMIC.1-3 In 2008, only 1.45% of all cancer deaths were in the 0-29 year age group3, due to the fact that cancer is predominantly a disease (or set of diseases) of people above the age of 65 years. The cancer burden of any country, then, is coupled to the age-structure of its population and does not exclusively depend on cancer incidence. In LMIC, populations are comprised primarily of children, adolescents and young adults, with estimated median ages in 2010 ranging from 15 (Uganda) to 30 years or so. High income countries (HIC), with some exceptions, have median ages from 30 to 50 (Monaco).2,4 Relative importance of cancer as a cause of death Relatively few countries and only 15.9% of the world’s population fall into the high income (HIC) category, yet these countries account for 31.6% of global deaths from cancer, largely because of the high proportion of older people in their populations.2,3 World Health Organization mortality data for the year 2008 (latest available) indicates that cancer is the leading cause of disease-related death in children between the ages of 5 and 14 years living in HIC (Table 21), with approximately 5.4 times as many deaths from cancer than as from communicable diseases.3 In low and middle income countries (LMIC), however, deaths from communicable diseases continue to outweigh cancer deaths in all age groups, although the ratio between communicable and non-communicable diseases becomes progressively smaller as the level of development increases, ranging from 20 times as many deaths from infections and parasitic diseases in low income countries (LIC)s to close to twice as many in upper middle income countries (UMIC). In children aged 0-4, communicable diseases cause more deaths than cancer in all income groups, including HIC, but in HIC, there are approximately twice as many deaths from infections and parasitic diseases than from cancer, while in LMIC, the figure is closer to 150 fold higher (Table 32).3 Cancer in adolescents and young adults The definition of “childhood” is quite arbitrary, the and varies somewhat among different health professionals but most cancer registries consider childhood cancer to refer to cancer in the 0-14

4

year age group. The major types of cancer in adolescents and young adults, (between 15 and 29), however, are quite specific for this age group and are rare in adult life. Tumours of older adults such as carcinomas, for example, rarely occur in this age range but sarcomas and haematological malignancies are relatively common. This provides some justification for expanding the age group normally taken to indicate childhood cancers. According to WHO, mortality data,3 in 2008, there were a total of 14,590 deaths from cancer in persons aged 5-29 years in HIC, which accounted for 28% of disease-related deaths in this age range. In contrast, there were only 3,187 deaths from communicable diseases in people aged 5-29 years in HIC, accounting for 5.9% of disease-related deaths. In HIC, then, cancer causes 74% of all disease-related deaths in the 5-29 year age group. This is not the case for LMIC, where in 2008, there were more deaths in the 5-29 age group from infection and parasitic diseases (1,310,858) than from cancer (187,222).3 Cancer in young childrenChildhood cancer and age

Perhaps not surprisingly, although deaths from infection and parasitic diseases predominate in all World Bank income groups in the 0-4 years age group, the number of cancer deaths in the world was of the same order of magnitude as that of in the 5-14 age group (40,061 compared to 53,464) even though the 0-4 year old population is about half the size of that of 5-14 year olds (Tables 21 and 32).2,3 This suggests that cancer is somewhat more common, and survival

rates particularly poor in younger children. Children in the 0-4 age group also have a somewhat different spectrum of cancers from older children (although there is overlap), since the majority of cases of retinoblastomas, neuroblastomas, Wilms tumor, hepatoblastomas and yolk sac tumors, for example, occur in this age group, probably reflecting an origin from persistent embryonic cells in many cases, or potentially oncogenic events in utero. Between 5-14 years of age, sarcomas are more common while haematological malignancies occur in both age groups. Unfortunately, there is very little recent data on tumour type in LMIC in childhood, but earlier data strongly suggested that there is considerable geographic variability (Fig 4?). [ ] Proportion of cancers in the age-range 0-14 years

While cancer incidence rates in 0-14 year olds are higher in more developed than in less developed countries (Table 43), the percentage of all cancers that are in the childhood age group is higher – often considerably higher - in the less developed countries because of the higher proportion of children in their populations (Figure 1a).1,2 For example, the percentage of all cancers occurring in African children below the age of 15 years (4.8% - sometimes double this figure in individual countries) is approximately 10 fold higher than the corresponding figure in EU27, i.e., the 27 countries of the European Union (0.46%). In contrast to incidence rates, mortality rates are much higher in children in the less developed countries3. For example, in 2008, the mortality rate in Africa was 6.9 per 100,000 children compared to 3.1 per 100,000 in EU27, and, unsurprisingly, in Africa, 5.31% of cancer deaths were in children (although, again, in individual countries, the percentage may can be twice as high) (Figure 1b). In EU27, children accounted for only 0.17% of cancer deaths. The demographic transition and development The very different age structures of developing countries arise from their high birth rate, relatively high death rate and low life expectancy. This is consistent with a model of changes in population structure resulting from the demographic transition brought about by the industrial revolution (Figure 2). Great Britain, where the industrial revolution began, had by 1850 made

Formatted: Highlight

5

the transition from an economy based on agriculture to one based on manufacturing, with its attendant and progressive urbanization, new lifestyles, and new exposures to a variety of chemicals, cheap manufactured cigarettes, more food (often processed or chemically preserved) and, at least initially, massive pollution. These dramatic changes were factors in the rapid increase in cancer that occurred, even while improved living conditions, safe water and better sanitisation led to an increased lifespan. However, most countries have still not undergone a similar transition, although many of the accoutrements of the industrial and technological revolutions have reached them as global trade has increased. The degree of development can, then, be seen as a general reflection of the degree to which countries have undergone a demographic transition, although many local factors, political and otherwise have also affected influenced the socioeconomic progress made. The growing burden of cancer The burden of cancer is growing in almost all countries as populations age, and increase in size. In sub-Saharan Africa, for example, Globocan data1 indicate that incident cases of cancer at all ages will almost double between 2008 and 2030, from 0.55 to 1.0 million from demographic effects alone. It has been assumed that mortality:incidence ratios will remain stable at 0.76, such that the actual number of deaths from cancer will inevitably rise unless major advances in prevention and/or treatment can be introduced. However, there are good reasons to anticipate a continued and even more marked reduction in the incidence and mortality of infectious diseases. Tuberculosis incidence and prevalence, for example, have been falling since 1990, and global mortality from this disease has fallen by more than one third (excluding HIV+ patients) since 1990, although there has been little change in Africa.5 Unfortunately, LMIC suffer from a “double burden of disease” since they still have high, if falling rates of the communicable diseases they have long suffered from, but an additional rising mortality from cancer. Taking into consideration only demographic changes, the cancer burden in developing countries is projected to increase from its present 7.1 million to 12.7 million in 2030, and mortality to increase from 4.8 to 9 million (accounting for some 70% of all cancer deaths) ,1 the In contrast, the childhood cancer burden is expected to increase marginally under the assumption of medium population growth. Incidence of childhood cancer in different geographic regions In the most technologically advanced countries, an average of 0.5% of all patients with cancer are aged 0-14 whereas in less developed countries taken as a whole, 2.0% of all cancer patients fall into this age group.1 (Figure 3). As is apparent from Figure 1a, these average figures hide a quite marked variability in both the incidence and proportion of childhood cancer in different countries at different levels of socioeconomic development and in different geographic regions. The incidence, which is more important from the epidemiological perspective than the fraction of childhood cases, varies from just over 5 to over close to 20 per 100,000 around the world, although for LMIC the quality of available statistics (e.g., Globocan estimates) is poor because of the paucity of cancer registries as well as pathologists in these countries. The level of coverage achieved by the registry clearly influences incidence and mortality rates. It provides a measure of the size of the problem, and a measure of the impact of cancer control measures that have been implemented. It may also contribute to awareness, and provides information that can lead to testable hypotheses. In LMIC, the coverage is currently particularly low, and global estimates, which are based on less than 5% coverage of the population may be imprecise. Investing in one or more cancer registries is not a luxury, but a wise and inexpensive political decision.34 Mandatory death certification in conjunction with cancer registration provides valuable confirmation of mortality rates from cancer.

6

Some African countries, such as Malawi and Uganda have a higher overall incidences due predominantly to Kaposi sarcoma related to the HIV epidemic (Figure 4), but other countries,

such as Kazakhstan and Turkmenistan also have relatively high rates which remain largely unexplained, although highly polluted areas in these countries have sometimes been blamed,. Pollution, however, is generally thought unlikely to be a major factor in the causation of cancer, although few studies have been done in very highly polluted regions.6-8 Other countries appear to have a very low incidence of childhood cancer, particularly countries in the southern and western parts of Africa.1 One reason for low overall incidence rates may be that many childhood cancers are not diagnosed, particularly leukemias and brain tumours, which appear, from the limited registry data available, to be are generally uncommon in LIC.9,10 But wWhile under-diagnosis and under-reporting doubtless occur, the magnitude of the incompleteness of the data is difficult to estimate because of the lack of a source of independent data such as death certificates. Moreover, it is unlikely that under-reporting is the sole reason for low overall incidence rates, since some cancers such as retinoblastoma and Wilms tumor have comparable or even higher incidence rates in LMIC than in HIC, while diseases with similar clinical characteristics may vary in incidence in different regions e.g., pre B and pre T cell leukemias (Figure 4).10 [39] Risk factors for childhood cancer in LMIC

Leukemias and lymphomas are the most common childhood cancers throughout most of the world, with ALL accounting for approximately 30% of all cases of childhood cancer (Figure 4).

This is not the case in equatorial Africa, where BL still predominates and accounts for 40-50% of all childhood cancers, except in countries where HIV-associated epidemic has led to a marked increase in the incidence of Kaposi sarcoma (KS). Other factors, somehow related in some way to development, may also be at play. Support for this possibility is provided by the reduced incidence of BL and increased incidence of ALL in the Gaza strip that coincided with a period of rapid socioeconomic development.11 There is additional evidence that the incidence of common (precursor B) ALL is low in LMIC and increases markedly in incidence as nations undergo socioeconomic development.11,12 For example, ALL has a low incidence in India and several institution-based studies have shown that the proportion of T cell cases is high, particularly in rural regions, but has been decreasing over time,which is consistent with a low incidence of precursor B ALL.13 [insert Rajalekshmy,[39] Magrath],[ A] There This is consistent with alsodata from Egypt [A,C] and evidence from death certificates in England and Wales that precursor-B ALL, using the early age peak (at 2-5 years) as a marker, started to increase in incidence in the early 20th century, first in England and Wales, and subsequently in various other population groups.12,14,15 Retinoblastoma may be a cancer that is inversely associated with socioeconomic development. It is sometimes the most common solid tumor in LMIC and may have a higher incidence than neuroblastoma, which is invariably more frequent in HIC. 10,32 The incidence of different childhood cancers in selected populations at different levels of socioeconomic development is shown in Figure 4. Epstein-Barr virus (EBV) is among the potentially important infections that predispose to childhood cancer, particularly in LMIC.16,17 In aAlmost 100% of BLs in Africa and over 50% of BLs in LMIC from which data is available most countries of the world contain EBV.`[17] Although EBV is ubiquitous, infection occurs in infancy or early childhood in LMIC and this may be one of the reasons that BL is more often associated with EBV in these countries. In contrast, in the wealthiest countries, where initial EBV infection occurs in adolescents and young adults, a majority of BLs are EBV negative. The Berlin-Frankfurt-Münster (BFM) cooperative group, for example, reported that that only 11% of BL cases in their series were associated with EBV.18

Formatted: Highlight

Formatted: Highlight

7

EBV is also present in almost all cases of Hodgkin lymphoma in children younger less than 10 years of age in LMIC,19 and has been reported in NK/T cell lymphomas children in Peru,20 T cell lymphohistiocytosis in Asia (associated with hemophagocytic syndrome),21 as well as in nasopharyngeal carcinoma, which is relatively common in children in North Africa, South and South East Asia.22 BL is probably also predisposed to by malaria, at least in regions holoendemic for P. falciparum, which include equatorial Africa and Papua New Guinea.17,23 Improved control of malaria may could, therefore, well also result in a reduction in the incidence of BL. Hepatocellular carcinoma, which is strongly associated with Hepatitis B (HBV) and C viruses (HCV) is rare in children, but occurs at higher incidence in some regions, such as Taiwan, where it can be effectively prevented by vaccination against HBV.24,25 HIV infected children who are also infected with Kaposi sarcoma (KS) virus (KSV or HHV8) have a markedly increased chance of developing KS and in several African countries, KS has become the most common cancer.10,26 Although HIV infection has been shown to predispose to B cell lymphoma, the most recent evidence suggests that in contrast to the situation in HICs, HIV does not predispose to BL in children in equatorial Africa.27 Radiation increases the risk of leukaemias and solid tumors in children as shown by epidemiological studies conducted in children exposed pre or post natal to radiation from the atomic bombs dropped on Japan, or during therapeutic radiation or diagnostic tests [31,32]. A high incidence of thyroid cancer was observed after the Chernobyl nuclear accident [33,34] and background environmental radiation also increases the risk for leukemia, probably with no lower dose threshold [5,10,35]. Radiation is one of the few other known risk factors for childhood cancer, as demonstrated by the studies by Doll and colleagues reporting increased rates of leukemia in children who had received therapeutic radiation28 and the very high incidence of thyroid carcinoma after the Chernobyl nuclear accident.29-31 However, in the absence of an event of this kind, radiation accounts for a minor proportion of cancer in children, especially in developing countries. Retinoblastoma and neuroblastoma may also be cancers in which factors associated with socioeconomic development are important. Retinoblastoma, for example, is often the most frequent childhood solid tumor in LMIC, with a higher incidence than neuroblastoma - the reverse of highly developed countries.10,32

Cancer Registration

Cancer registration is an important prerequisite to cancer control plans and childhood cancer should be included in the registry.33,34 The level of coverage achieved by the registry clearly influences incidence and mortality rates. It provides a measure of the size of the problem, and a measure of the impact of cancer control measures that have been implemented. It may also contribute to awareness, and provides information that can lead to testable hypotheses. In LMIC, the coverage is currently particularly low, and global estimates, which are based on less than 5% coverage of the population may be imprecise. Investing in one or more cancer registries is not a luxury, but a wise and inexpensive political decision.34 Mandatory death certification in conjunction with cancer registration provides valuable confirmation of mortality rates from cancer.

8

Access to Care in LMIC

While few would disagree that in spite of the high cost and complexity of therapy, children with cancer should have the same rights to health care as any other child, the stark reality in LMIC is that access to care is extremely limited and an unknown fraction of children with potentially curable cancers never receive care – not even palliative care. Palliative care should be available everywhere, even if no other treatment is available, Yet Eeven when functioning pediatric hematology oncology departments are present in major medical centers are present, as is the case throughout the Middle East, and pain medication is available, pediatric palliative care services are still in their infancy. and there is an urgent need to develop hospital-based and community-based palliative and supportive care services. In many other LMIC, even access to pain medication poses a major problem.35

The large (but variable) difference between 5-year survival rates in LMIC and HIC is shown in Figure 5. These differences are due to a variety of factors discussed below. In the 36 LIC

(even more economies; Table 1) listed by the World Bank, there are more similarities than differences, with many shared problems that influence access to care. Tthe vast majority of treatment centers are within the public sector, and referrals predominantly come through primary or secondary health care providers. WHO strongly advocates the development of networks of primary health care providers (where the access point to the health system is via an ambulatory generalist rather than a hospital) on the grounds that this would result in greater equity in health care in general.36 However, there are often insufficient primary care providers in LIC to serve the needs of the country, and the majority have little knowledge of cancer, particularly in children. Many families, therefore, first consult traditional healers, who are more accessible, but may lead to delays in obtaining a conventional consultation. Even primary care providers or specialists may not consider the possibility of cancer and many months are often wasted in futile treatment of a benign condition the child does not have. And the greater the delay in referral, the more advanced the disease and the worse the prognosis. Education of primary care physicians is essential to promote referral, but to be effective it is also necessary to educate the public with respect to the symptoms of cancer and the possibility of cure if referred promptly for specialist care. When children are sufficiently ill, parents may take them directly to a secondary or tertiary care center directly, but the paucity of such hospitals means that for much of the population in LIC (often predominantly rural) tertiary care centers are far away, raising creating multiple problems including transportation costs, absence from the home for long periods with detriment to siblings and an unaffordable loss of income. A hostel near the treatment center where parents can stay free of charge while away from home can sometimes reduce abandonment of (prematurely terminating) therapy,37 but does not overcome the problem of treatment cost. With the exception of the few countries which have reasonably good insurance coverage, the family must pay for diagnosis and much or all of therapy. Parents may stop therapy when their money runs out or if they feel that they must return home, even if to do so could result in the sick child’s death. Ultimately, reaching a tertiary care center is only of value if trained staff are available with experience in the treatment of childhood cancer and the necessary drugs (nearly all are listed on the WHO’s Essential Medicines List), as well as standard treatment protocols feasible and effective in resource-poor settings and equipment are available – which is often not the case, especially in LIC. Even cheap, commonly used drugs such as cyclophosphamide and vincristine may not be imported for periods of time, and radiation therapy facilities, along with

9

the staff required to deliver radiotherapy, are totally insufficient in many countries – for example, it has been estimated by the International Atomic Agency’s Program of Action for Cancer Therapy (PACT) that 80% of Africans lack access to radiation therapy.38 At worst, radiation therapy is completely absent in some 30 countries along with the many specialists required to diagnose and care for children with cancer. Needless to say, the deficits are worse in rural rather than urban areas and cancer centers vary considerably with respect to the quality of care provided. Different ethnic and income groups may also experience inequities in the care provided.39-45 Often the wealthy are treated in private hospitals or even go abroad for treatment or obstetric care Although capacity for pediatric cancer care will inevitably improves with development, such that overall, access to care is better in MIC, the pace of improvement will depend upon government decisions on how to spend extra income. there are enormous variations in the latter, both within and between countries; per capita income, for example, differs 10 fold among the total of 108 countries and dependencies presently classified by the World Bank as MIC (these are further divided into upper and lower MIC), it is not possible to meaningfully summarize access to care across this broad range of countries (see Table 1). However, UMIC often provide good access to care, especially in urban regions, although there may also be large regions of the country where the situation differs little from that described for LIC (e.g., in China, India and Brazil). Many non-medical issues such as aAccess to secondary educational facilities and limitations in university places, incomplete coverage of academic disciplines or the absence of education in specific disciplines, (including medical specialities), are all important factors that influence the availability of human capital for childhood cancer care. These non-medical factors also relate to the level of development [ ] and vary markedly among and within countries, especially in MIC. The migration of health workers to countries able to offer better professional and personal opportunities either before or after specialist training is another oft-debated issue. Many assume that training in HIC is advantageous, but many trainees stay abroad permanently. This is sufficiently common in LMIC that the WHO has issued guide lines guidelines on recruitment policies used by HIC.46 Sometimes health care providers are tempted to perform fee-for-service work in their own country, or work in well equipped private hospitals, which provide care only for those who can afford it (costs in for-profit private systems are generally higher) and further reduce the work force that provides for the poor. Governments may choose not to emphasize cancer care, or may not have the financial resources to provide for the poor, especially if not convinced that patients can be cured. Professional societies, even if very small initially, can help to counteract the drain on resources for the poor be powerful motivating forces by enhancing job satisfaction, e.g., by facilitating discussion among their members exchanges of problems encountered and potential solutions, promoting continuing education, fostering the development of research networks, and collaborating with similar organizations in adjacent countries. For example, from humble beginnings, the Turkish Pediatric Oncology Group has grown into a strong national force for education, collaboration on clinical research projects, and a facilitator of cancer registration.48 Universal insurance coverage with pooling of funds is strongly recommended by WHO,36 since it creates equity with respect to access to care. Insurance also overcomes the obstacle of the financial cost of diagnosis and therapy that is a frequent reason for late diagnosis and, consequently, advanced disease, or abandonment of therapy (WHO states that 100 million people are pushed into poverty each year because they must pay “out-of-pocket” for health care which they cannot afford). Although health insurance is sometimes available in LIC, it is often

10

through private companies, expensive, and therefore only purchased by the wealthier members of the community – the minority that is least affected by the cost of care.47-49 In Mexico, the introduction, by the government, of insurance for the most socioeconomically deprived families (50% of the population), the development of standard treatment protocols and accreditation of 49 programs for the care of childhood cancer resulted in a reduction of abandonment from 35% to 4% and an improvement in survival rates in childhood cancer.50-53 Nigeria, and Rwanda have also introduced national insurance programs with tangible benefits,47,48 and China initiated an insurance program for certain leukemias treated at Government- approved centers in 2010, along with many additional measures to improve healthcare coverage.54,55 These examples demonstrate that more equitable access to care and reduction in of therapy abandonment of therapy can be achieved when there is the political will, but insurance alone is not enough, and must be accompanied by all of the components of care discussed in the next following section,although the development of a total care model may require many years..

Patient Management, Clinical Trials and Models of Care and the Need for Research Childhood cancer is one of the success stories in the history of cancer treatment, with survival rates of 80% or more at 5 years now being achieved in the most technologically advanced countries HIC (article 1, this edition). And because children have their entire lives before them, saving the lives of children may have a much greater impact on economic development than cancer care in the elderly. Unfortunately In LIC, where, cure rates are often low in LIC, due to presentation with advanced disease (resulting in a greater likelihood of complications and higher treatment costs), and the unavailability of the necessary medical expertise factors described above, essentially all countries, however poor, make some attempt to care for some children with cancer (e.g., a subset of cases, such as BL or ALL) although they may have no or very few pediatric oncologists, pathologists or medical oncologists. When trained medical and nursing staff are so few (e.g., in the poorest LIC Failure to perform pre-treatment investigations, to provide emergency care and or to documentdocumentation of findings, including the diagnosis made and treatment given, may be poor or non-existent, sometimes occurs even in major university hospitals.56 While this may be a consequence of too many patients to care for, the absence of systems to ensure that patients reach specified regional centers promptly, good medical record keeping, access to a broad range of knowledgeable) specialists, access to drugs on the WHO Essential Medicines list and to a essential equipment, the use of standard treatment protocols, continuing education and consequences (e.g., retraining) when staff are clearly negligent, the results of treatment will be poor.49 It seems logical, therefore, to develop a national plan for pediatric oncology plan for pediatric oncology, the first element of which should be an assessment of the current national situation through the collection of relevant information. Even if the plan entails a single center and/or a single disease to start with, this can both provide a model and serve as a focal point for services and teaching activities within the country, very often in conjunction with external aid [56]. Such a plan will of necessity take account of the available resources and may be very different for LIC compared to MIC [57]. In the latter, where some high quality facilities already exist, the emphasis is likely to be on continuous expansion of capacity, ideally, through the creation of hospital networks [58], with an emphasis on training and education relating to the prompt referral and diagnosis such that access is improved, reducing the fraction of patients with advanced disease and improving the results of treatment [59,60]. Clearly, the strategy must go well beyond simply training pediatric oncologists.

11

Diagnosis Correct diagnosis and accurate staging are is the foundations on which all subsequent management rests. Unfortunately, in the poorest countries, the few existing pathologists are insufficient to diagnose all cases in the country and even many MIC have limited diagnostic capabilities – in part because of limited training in the diagnosis of pediatric cancers, and in part because of the lack of reagents andcost of special pathological techniques such as immunohistochemistry and FISH analysis[23 ]. Although tissue blocks or slides are sometimes sent to centers in other countries for diagnosis, this does not represent a sustainable solution and approaches to improving diagnosis may need to be performed initially in the few centers that have at least basic requirements. There is no substitute for introducing at least immunohistochemistry into one or two pathology laboratories in the country and providing training and quality control by external pathologists although somebody, usually the patient has to pay for the reagents and initial training. On-line case discussions help patients in real time and provide real-world training and mentoring for local pathologists. Web-based tools have been developed to facilitate this process. Such digital systems could also be used exclusively in-country, assuming there is a well trained technician to prepare samples and internet access. On-line educational tools can also be used to improve access (of patients and doctors) to information and add value to formal education.59-61 One limiting factor to improving diagnosis is the cost of diagnostic reagents, which is usually born by the patient, so that even when definitive testing, such as immunohistochemistry, is available, patients may not be able to afford it. Along with treatment costs, this could be solved by health insurance, or by working with local NGO’s to develop additional funding and advocate to local or federal governments for improved resources.. Management One of the first elements of the creation of a pediatric oncology program is the physical separation of in-patient oncology care from general hospital patients who may have infectious diseases. In such a setting In the absence of a separate physical location, leukemia patients, in particular, are likely to die during induction therapy and all patients receiving chemotherapy will also be at risk for serious infection. Overcrowding is a frequent problem in major centers, but the importance of clean bathrooms, hand washing hygienic facilities and standardized procedures for disposal of contaminated materials, needles and waste, as well as prompt recognition and treatment of fever and neutropenia and prevention or recognition of other frequent complications of chemotherapy cannot be overemphasized.62-66 If chemotherapy is made up in an adjacent pharmacy, appropriate protective hoods should be available. At higher income levels, a critical care unit would help save the lives of a small number of patients, but prior to this, care providers should understand the basic elements of standard supportive care and ensure that this does not compromise sevices elsewhere because of the high staff: patient ratios required.Psychosocial and social assistance provides much needed family support and may decrease the likelihood of abandonment of therapy [D] Adapted treatment regimens

An important principle in clinical care is the adaptation of treatment to the stage or extent of disease and/or to the presence of markers indicating good or poor risk. Much of the research

12

conducted in the USA and Europe in the second half of the 20th century involved identifying not only better therapy for specific disease entities, but also the best therapy for subsets of the same disease that may be treated with less therapy without detriment. Unfortunately, in many LMIC, staging is imprecise because of lack of access to advanced imaging techniques and specialists in their use and interpretation, which may be unavailable or unaffordable; ultrasound imaging is, generally more widely, available, even in district hospitals. The use of available resources to diagnose patients earlier through public and professional education may be more important than the purchase of capital equipment that requires technical expertise as well as the necessary skills to interpret results.57,58 However, the lack of precise staging is not always essential to a good outcome, particularly when local therapy is not involved [72]. Local staff and external experts should also be aware that prognostic factors validated in HIC may be less predictive in LMIC because the majority of patients have advanced disease, and toxic death rates are also more likely to suffer from, malnutrition, and delays caused by intercurrent infections (e.g. hepatitis, tuberculosis, viral and fungal infections), which lead to poor therapy tolerance and can obscure the indentification of risk factors true treatment failures.39,62-66

Sometimes, the conditions in which patients and their families live lead to a high rate of comorbidities, including malnutrition and infections. On the other hand, Rremaining in the hospital for long periods can may also increase the risk of infection nosocomial infection and the likelihood of antibiotic resistance.39 Psychosocial and social workers are essential members of the team. They can help children better understand their disease and its treatment while assisting families financially and psychologically. Some treatment approaches developed in HICs may be too intensive in LMIC, or require drugs which are not always available, or too expensive, although in pediatric cancer, almost all diseases can be treated effectively with generic drugs. or must must be modified for use in countries according to the resources available. Some do require radiation therapy, and the invariable deficiency in insufficiency of machines and expertise in LMIC iscan be a major problem in poorly served regions.38 While its absence does not preclude development of oncology services, it remains an important part of therapy for many cancers, and the provision of radiotherapy services should be a part of any national cancer program involving cancer care. Institutional and national collaboration

The degree of cooperation within an institute or between institutes in low resource settings also varies approximately with the level of development. At the lower end of the scale, specialists may act in isolation. Multidisciplinary meetings, whether to discuss care, or why a patient died, may be impossible because not all disciplines are available, or there because of lack of needed experts, is limited time or lack of incentive to discuss cases. Evidence-based approaches to care - the model of care that was so successful in Europe and the USA - may not be possible in LIC especially, because of deficiencies in patient care and infrastructure for documentation of all aspects of diagnosis andn treatment, including the rate of Poor documentation of disease extent, abandonment premature tereminatio n of therapy, or lack of follow up after completion of treatment. The limited training in research methodology and the added cost of research infrastructure are doubtless, important contributors to the paucity of research from LMIC, although MIC, especially the larger countries, usually have at least some centers that are capable of undertaking clinical and basic research [70 ]. and non-standard therapy may eliminate all possibility of evidence-based approaches to care - the model of care that was so successful in Europe and the USA, In the end, most patients are treated with modified western protocols, Yet there are unique research opportunities in LMIC, particularly to undertake epidemiological studies (that could lead to prevention) and determining the extent to which decreasing intensity of therapy and substituting alternatives for those elements of treatment that

Field Code Changed

13

are not feasible in LIC or MLIC, such as radiation or infusional therapy, can be done, at least in some circumstances, with minimal compromise of efficiency with largely unknown outcomes in LIC 67-69 and thoseA very high fraction of clinical trials that do take place are product-driven and initiated by the international pharmaceutical industry, which often brings its own infrastructure and may provide little training to local health care providers. In contrast, national pediatric oncology groups can increase awareness of pediatric cancer in the community, leading to earlier diagnosis, support education of health professionals, implement pediatric cancer registries, and develop locally feasible treatment protocols.70 There is some evidence that the degree of motivation within the country, particularly of policy makers and senior health care professionals is important. and external cooperation, including training for health professionals, is beneficial. A survey of treatment outcome in 10 developing countries suggested that better results were achieved in countries in which per capita annual total and governmental health-care expenditure, GNI per capita, the number of physicians and nurses per 1000 population, and most significantly, government expenditure on health care were all associated with a better survival rate.71 It is also true that key senior clinical personnel can do much to move education and research forward in their country. Sometimes, the conditions in which patients and their families live lead to a high rate of comorbidities, including malnutrition and infections. On the other hand, remaining in the hospital can also increase the risk of infection and the likelihood of antibiotic resistance.39 Psychosocial and social workers are essential members of the team. They can help children better understand their disease and its

treatment while assisting families financially and psychologically. International Collaboration While the numerous difficulties that have prevented LIC from developing organized approaches to childhood cancer, perhaps the greatest obstacle, commonly encountered in the LIC, is the absence of any specialists trained in pediatric oncology or in other specialties essential to care. In this casecircumstance, collaboration with one or more countries that have the requisite expertise is essential. Ideally, international experts should visit the country to provide training (e.g., via a “twinning program” or collaboration with an international organization) without the loss of the few potential experts for several years – and sometimes permanently. International experts can also provide many teaching tools, including and frequent contact via on-line meetings information and distance learning can facilitate the care of individual patients and implementation of adapted protocols. It may will take many years to achieve the standards of the most developed countries, but the latter provide a target – survival rates close to those attained in the most advanced countries, as well as the means to achieve the standards of the most developed countries, but the latter provide a target – 5 year survival rates close to those attained in the most advanced countries (Fig.5%). as well as the means to achieve it. Measurable five-year survival rates of 30-40% overall may be an acceptable first step and one even if lower than in HIC that will help to convince the public, health care providers and policy makers that cure is possible, creating a dynamic that is not easily ignored by policy makers avoided. International collaborators can also help with more general aspects of health system strengthening, hospital management and other aspects of health care infrastructure that are presently lacking to various degrees in LMIC and may create opportunities to participate in research at an international level, leading to more rapid completion of research studies and an expansion ding of global research capacity.67-69 International non-governmental organizations working with local health professionals and public health officials can participate in educating the public and providing services at cancer treatment centers that the government is not in a position to fund (play facilities and therapists, for example), The International Confederation of

Field Code Changed

14

Childhood Cancer Parent Organizations (http://icccpo.org/index.cfm) helps develop and sustain parent groups, and includes 158 organizations in 85 countries as of October 30, 2012. A Model of Care

The first need of LMIC in developing improved capacity for the care of children with cancer is capacity building, which touches every aspect from early diagnosis to the documentation of late effects. Care, however, begins in the community. , and pPublic education and education of primary care providers, as well as a wide range of specialists as well as and policy -makers is essential to developing a national plan for pediatric cancer care (NPPC). The various elements that of an NPPC might include, although if necessary introduced in a stepwise fashion such a plan might include are shown in Table 45. The introduction of research, often in conjunction with clinical care has a long and successful tradition in pediatric oncology. Those who perform research learn the importance of accurate information and protocol discipline, as well as the necessity of good follow up. ALL and lymphomas are suitable diseases on which to focus attention in the context of research, since these tumors are the most common of all childhood cancers, and have the advantage that surgery and radiation therapy are not standard components of treatment. National and multinational studies, usually conducted in conjunction with external collaborators (e.g. GFAOP, INCTR, AHOPCA), or by local, trained investigators, have been shown to improve survival rates, with on occasion, an impact at the population level.39,72 Survival rates currently being achieved in different world regions are shown in Figure 5. In MIC, basic research may be affordable, at the larger institutions. While the research could

cover a wide range, identifying causal factors and developing preventive approaches are of particular importance. Another step towards sustainability and constant improvement is encouraging physicians and scientists in developing countries to compete for grants at an international level, providing, at the same time, a goal to be achieved and a potential source of external funding. The “My Child Matters” program supported and coordinated by the Sanofi Espoir Foundation (http://www.fondation-sanofi-espoir.com/en/) and the UICC, which provides grants for investigators in selected LMICs along with a “mentor” to assist grantees to reach their goals provides an excellent model for this approach.71 Clearly, providing good pediatric cancer care involves a large number of people, disciplines, medicines and technologies. Unlike many other health challenges, however, pediatric cancer has been shown to be is a largely tractable problem that is, in the scheme of things, quite small. Child health is an important element of the millennium development goals, although cancer is not mentioned. Many Nonetheless, policy makers often shy away from addressing acancer - a complex set of diseases that they find difficult to understand, and which require relatively expensive and potentially highly toxic and costly treatment approaches. In such circumstances it is not surprising that , and which is cancer, particularly in LIC is still often considered by health care providers and policy makers to be incurable. by many. The latter may be true in many instances in the LMIC. B ut it has proven to be untrue elsewhere. And, Iironically, the financial cost of the resources presently being used being spent to cure approximately on 80% of the 20% of childhood cancers that occur in the HIC (published costs for overall care vary greatly, but suggest that costs may by at least 10-fold less in the LMIC [G, H, I]). Thus, treatment costs for the 4-fold greater number of pediatric cancers in the LMIC are likely to be less than treatment costs in the HIC [J]. Moreover, the enormous amount of money “wasted” in the affluent health care systems (for example, $750 billion per year in the USA, according to the Institute of Medicine [E])- those in highly developed countries -– would be more than enough to provide a sustainable solution to children with cancer throughout the world (including the building of many hospitals and the training of needed experts).

15

References

1. , Ferlay, J., Shin, H. R., Bray, F., Forman, D., Mathers, C., and Parkin, D. M. GLOBOCAN 2008 v2.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 10. http://globocan.iarc.fr. Accessed on 10-31-2012

2 World Population Prospects, the 2010 Revision. United Nations Department of Economic and Social Affairs. Population Division, Population Estimates and Projections Section. http://esa.un.org/unpd/wpp/Sorting-Tables/tab-sorting_ageing.htm. Accessed on 10-31-

2012

3 World Health Organization Global Health Observatory Data Repository. http://apps.who.int/ghodata/?vid=10012. Accessed on 10-31-2012

4 The World Factbook, Central Intelligence Agency. https://www.cia.gov/library/publications/the-world-factbook/fields/2177.html. Accessed on 10-31-2012

5 World Health Organization Global Report. 1.http://www.who.int/tb/publications/global_report/2011/gtbr11_full.pdf.

6 Weng HH, Tsai SS, Chen CC, Chiu HF, Wu TN, Yang CY. Childhood leukemia development and correlation with traffic air pollution in Taiwan using nitrogen dioxide as an air pollutant marker. J Toxicol Environ Health A 2008; 71:434-438

7 van Leeuwen DM, Pedersen M, Hendriksen PJ, Boorsma A, van Herwijnen MH, Gottschalk RW et al. Genomic analysis suggests higher susceptibility of children to air pollution. Carcinogenesis 2008; 29:977-983

8 Reid BC, Ghazarian AA, Demarini DM, Sapkota A, Jack D, Lan Q et al. Research opportunities for cancer associated with indoor air pollution from solid-fuel combustion. Environ Health Perspect 2012; 120:1495-1498

9 Howard SC, Metzger ML, Wilimas JA, Quintana Y, Pui CH, Robison LL et al. Childhood cancer epidemiology in low-income countries. Cancer 2008; 112:461-472

10 International Incidence of Childhood Cancer, Volume 2 . Parkin DM, Kramarova E, Draper GJ, Masuyer E, Michaelis J, Neglia JP et al., editors. IARC Scientific Publications No. 144. International Agency for Research on Cancer. 1998. Lyon.

11 Ramot B, Magrath I. Hypothesis: the environment is a major determinant of the immunological sub-type of lymphoma and acute lymphoblastic leukaemia in children. Br J Haematol 1982; 50:183-189

12 Brown WM, Doll R. Leukaemia in Childhood and Young Adult Life. Br Med J 1961; 1:981-988

16

13 Arya LS, Padmanjali KS, Sazawal S, Saxena R, Bhargava M, Kulkarni KP et al. Childhood T-lineage acute lymphoblastic leukemia: management and outcome at a tertiary care center in North India. Indian Pediatr 2011; 48:785-790

14 Greaves MF. Aetiology of acute leukaemia. Lancet 1997; 349:344-349

15 Kamel AM, Assem MM, Jaffe ES, Magrath I, boul Enein MI, Hindawy DS. Immunological phenotypic pattern of acute lymphoblastic leukaemia in Egypt. Leuk Res 1989; 13:519-525

16 Rao CR, Gutierrez MI, Bhatia K, Fend F, Franklin J, Appaji L et al. Association of Burkitt's lymphoma with the Epstein-Barr virus in two developing countries. Leuk Lymphoma 2000; 39:329-337

17 Magrath I. Epidemiology: clues to the pathogenesis of Burkitt lymphoma. Br J Haematol 2012; 156:744-756

18 Karajannis MA, Hummel M, Oschlies I, Anagnostopoulos I, Zimmermann M, Stein H et al. Epstein-Barr virus infection in Western European pediatric non-Hodgkin lymphomas. Blood 2003; 102:4244

19 Naresh KN, Johnson J, Srinivas V, Soman CS, Saikia T, Advani SH et al. Epstein-Barr virus association in classical Hodgkin's disease provides survival advantage to patients and correlates with higher expression of proliferation markers in Reed-Sternberg cells. Ann Oncol 2000; 11:91-96

20 Rodriguez-Pinilla SM, Barrionuevo C, Garcia J, Martinez MT, Pajares R, Montes-Moreno S et al. EBV-associated cutaneous NK/T-cell lymphoma: review of a series of 14 cases from peru in children and young adults. Am J Surg Pathol 2010; 34:1773-1782

21 Subramaniam K, D'Rozario J, Pavli P. Lymphoma and Other Lymphoproliferative Disorders in IBD: A Review. J Gastroenterol Hepatol 2012;

22 Toumi N, Frikha M, Siala W, Khabir A, Karray H, Boudawara T et al. [Juvenile nasopharyngeal carcinoma: anatomoclinic, biologic, therapeutic and evolutive aspects]. Bull Cancer 2010; 97:427-433

23 Naresh KN, Raphael M, Ayers L, Hurwitz N, Calbi V, Rogena E et al. Lymphomas in sub-Saharan Africa - what can we learn and how can we help in improving diagnosis, managing patients and fostering translational research? Br J Haematol 2011;

24 Ni YH, Chang MH, Wu JF, Hsu HY, Chen HL, Chen DS. Minimization of hepatitis B infection by a 25-year universal vaccination program. J Hepatol 2012; 57:730-735

25 McMahon BJ, Bulkow LR, Singleton RJ, Williams J, Snowball M, Homan C et al. Elimination of hepatocellular carcinoma and acute hepatitis B in children 25 years after a hepatitis B newborn and catch-up immunization program. Hepatology 2011; 54:801-807

17

26 Tukei VJ, Kekitiinwa A, Beasley RP. Prevalence and outcome of HIV-associated malignancies among children. AIDS 2011; 25:1789-1793

27 Parkin DM, Garcia-Giannoli H, Raphael M, Martin A, Katangole-Mbidde E, Wabinga H et al. Non-Hodgkin lymphoma in Uganda: a case-control study. AIDS 2000; 14:2929-2936

28 Doll R. Effects of small doses of ionising radiation. J Radiol Prot 1998; 18:163-174

29 Darby SC, Olsen JH, Doll R, Thakrar B, Brown PD, Storm HH et al. Trends in childhood leukaemia in the Nordic countries in relation to fallout from atmospheric nuclear weapons testing. BMJ 1992; 304:1005-1009

30 Piciu D, Piciu A, Irimie A. Thyroid cancer in children: a 20-year study at a Romanian oncology institute. Endocr J 2012; 59:489-496

31 Steliarova-Foucher E, Stiller C, Kaatsch P, Berrino F, Coebergh JW, Lacour B et al. Geographical patterns and time trends of cancer incidence and survival among children and adolescents in Europe since the 1970s (the ACCISproject): an epidemiological study. Lancet 2004; 364:2097-2105

32 Onwasigwe CN, Aniebue PN, Ndu AC. Spectrum of paediatric malignancies in eastern Nigeria (1989-1998). West Afr J Med 2002; 21:31-33

33 Curado MP, Voti L, Sortino-Rachou AM. Cancer registration data and quality indicators in low and middle income countries: their interpretation and potential use for the improvement of cancer care. Cancer Causes Control 2009; 20:751-756

34 Valsecchi MG, Steliarova-Foucher E. Cancer registration in developing countries: luxury or necessity? Lancet Oncol 2008; 9:159-167

35 Silbermann M, Al-Hadad S, Ashraf S, Hessissen L, Madani A, Noun P et al. MECC regional initiative in pediatric palliative care: Middle Eastern course on pain management. J Pediatr Hematol Oncol 2012; 34 Suppl 1:S1-11

36 The world health report 2008: primary health care now more than ever.

http://www.searo.who.int/LinkFiles/Reports_whr08_en.pdf. Accessed on 10-31-2012

37 Howard SC, Pedrosa M, Lins M, Pedrosa A, Pui CH, Ribeiro RC et al. Establishment of a pediatric oncology program and outcomes of childhood acute lymphoblastic leukemia in a resource-poor area. JAMA 2004; 291:2471-2475

38 Potterton L. Access to affordable radiotherapy. IAEA Bulletin 2010; 52:10-11

39 Magrath I, Shanta V, Advani S, Adde M, Arya LS, Banavali S et al. Treatment of acute lymphoblastic leukaemia in countries with limited resources; lessons from use of a single protocol in India over a twenty year period. Eur J Cancer 2005; 41:1570-1583

18

40 Howard SC, Ribeiro RC, Pui CH. Strategies to improve outcomes of children with cancer in low-income countries. Eur J Cancer 2005; 41:1584-1587

41 Cunha KC, Oliveira MC, Gomes AC, de Castro LP, Viana MB. Clinical course and prognostic factors of children with Burkitt's lymphoma in a developing country: the experience of a single centre in Brazil. Rev Bras Hematol Hemoter 2012; 34:361-366

42 Klumb CE, Hassan R, De Oliveira DE, De Resende LM, Carrico MK, De Almeida DJ et al. Geographic variation in Epstein-Barr virus-associated Burkitt's lymphoma in children from Brazil. Int J Cancer 2004; 108:66-70

43 Queiroga EM, Gualco G, Weiss LM, Dittmer DP, Araujo I, Klumb CE et al. Burkitt lymphoma in Brazil is characterized by geographically distinct clinicopathologic features. Am J Clin Pathol 2008; 130:946-956

44 Epelman S. Pediatric oncology in Brazil. Pediatr Hematol Oncol 1998; 15:377-383

45 Gupta N, Seth T, Mishra P, Mahapatra M, Rathi S, Kapoor R et al. Treatment of acute myeloid leukemia in children: experience from a tertiary care hematology centre in India. Indian J Pediatr 2011; 78:1211-1215

46 WHO practice on the international recruitment of health personnel. http://www.who.int/hrh/en/. Accessed on 10-31-2012

47 Obalum DC, Fiberesima F. Nigerian National Health Insurance Scheme (NHIS): An overview. Niger Postgrad Med J 2012; 19:167-174

48 Lu C, Chin B, Lewandowski JL, Basinga P, Hirschhorn LR, Hill K et al. Towards universal health coverage: an evaluation of Rwanda Mutuelles in its first eight years. PLoS One 2012; 7:e39282

49 Perez-Cuevas R, Doubova SV, Zapata-Tarres M, Flores-Hernandez S, Frazier L, Rodriguez-Galindo C et al. Scaling up cancer care for children without medical insurance in developing countries: The case of Mexico. Pediatr Blood Cancer 2012;

50 Knaul FM, Gonzalez-Pier E, Gomez-Dantes O, Garcia-Junco D, Arreola-Ornelas H, Barraza-Llorens M et al. The quest for universal health coverage: achieving social protection for all in Mexico. Lancet 2012; 380:1259-1279

51 Perez-Cuevas R, Chertorivski S, Cortes-Gallo G, Rodriguez-Ortega E, Caballero F, Munoz-Hernandez O. Lessons learned from the first evaluation of the Medical Insurance for a New Generation: bridging research and policy. Salud Publica Mex 2012; 54 Suppl 1:S90-S97

52 Ribeiro RC. Impact of the Mexican government's system of social protection for health, or Seguro Popular, on pediatric oncology outcomes. Pediatr Blood Cancer 2012;

19

53 Rivera-Luna R, Correa-Gonzalez C, Altamirano-Alvarez E, Sanchez-Zubieta F, Cardenas-Cardos R, Escamilla-Asian G et al. Incidence of childhood cancer among Mexican children registered under a public medical insurance program. Int J Cancer 2012;

54 Tang S, Tao J, Bekedam H. Controlling cost escalation of healthcare: making universal health coverage sustainable in China. BMC Public Health 2012; 12 Suppl 1:S8

55 Li Y, Wu Q, Xu L, Legge D, Hao Y, Gao L et al. Factors affecting catastrophic health expenditure and impoverishment from medical expenses in China: policy implications of universal health insurance. Bull World Health Organ 2012; 90:664-671

56 Howard SC, Wilimas JA. Delays in diagnosis and treatment of childhood cancer: where in the world are they important? Pediatr Blood Cancer 2005; 44:303-304

57 Luna-Fineman S, Barnoya M, Bonilla M, Fu L, Baez F, Rodriguez-Galindo C. Retinoblastoma in Central America: report from the Central American Association of Pediatric Hematology Oncology (AHOPCA). Pediatr Blood Cancer 2012; 58:545-550

58 Leander C, Fu LC, Pena A, Howard SC, Rodriguez-Galindo C, Wilimas JA et al. Impact of an education program on late diagnosis of retinoblastoma in Honduras. Pediatr Blood Cancer 2007; 49:817-819

59 Wayne State University Ghana Rural Health Project. http://prognosis.med.wayne.edu/article/student-leads-wsu-telemedicine-project-in-rural-ghana. Accessed on 10-31-2012

60 Qaddoumi I, Mansour A, Musharbash A, Drake J, Swaidan M, Tihan T et al. Impact of telemedicine on pediatric neuro-oncology in a developing country: the Jordanian-Canadian experience. Pediatr Blood Cancer 2007; 48:39-43

61 Qaddoumi I, Ribeiro RC. Can communication technologies reduce inequalities in cancer care? Lancet Oncol 2011; 12:224

62 Gavidia R, Fuentes SL, Vasquez R, Bonilla M, Ethier MC, Diorio C et al. Low socioeconomic status is associated with prolonged times to assessment and treatment, sepsis and infectious death in pediatric Fever in el salvador. PLoS One 2012; 7:e43639

63 Gupta S, Bonilla M, Valverde P, Fu L, Howard SC, Ribeiro RC et al. Treatment-related mortality in children with acute myeloid leukaemia in Central America: incidence, timing and predictors. Eur J Cancer 2012; 48:1363-1369

64 Gupta S, Antillon FA, Bonilla M, Fu L, Howard SC, Ribeiro RC et al. Treatment-related mortality in children with acute lymphoblastic leukemia in Central America. Cancer 2011;

65 Somjee S, Pai S, Parikh P, Banavali S, Kelkar R, Advani S. Passive active prophylaxis against Hepatitis B in children with acute lymphoblastic leukemia. Leuk Res 2002; 26:989-992

20

66 Raje NS, Rao SR, Iyer RS, Kelkar RS, Pai SK, Nair CN et al. Infection analysis in acute lymphoblastic leukemia: a report of 499 consecutive episodes in India. Pediatr Hematol Oncol 1994; 11:271-280

67 Dunne J, Murphy MD, Rodriguez WJ. The Globalization of Pediatric Clinical Trials. Pediatrics 2012;

68 Lu F, Suda KJ, Marks KE. Geographic location of antineoplastic agent clinical trials conducted in developed and developing countries. Int J Clin Pharm 2012;

69 Perkovic V, Patil V, Wei L, Lv J, Petersen M, Patel A. Global randomized trials: the promise of India and China. J Bone Joint Surg Am 2012; 94 Suppl 1:92-96

70 Kebudi R. Pediatric oncology in Turkey. J Pediatr Hematol Oncol 2012; 34 Suppl 1:S12-S14

71 Ribeiro RC, Steliarova-Foucher E, Magrath I, Lemerle J, Eden T, Forget C et al. Baseline status of paediatric oncology care in ten low-income or mid-income countries receiving My Child Matters support: a descriptive study. Lancet Oncol 2008; 9:721-729

72 Ngoma T, Adde M, Durosinmi M, Githang'a J, Aken'Ova Y, Kaijage J et al. Treatment of Burkitt lymphoma in equatorial Africa using a simple three-drug combination followed by a salvage regimen for patients with persistent or recurrent disease. Br J Haematol 2012; 158:749-762

73 Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010; 127:2893-2917

74 Baade PD, Youlden DR, Valery PC, Hassall T, Ward L, Green AC et al. Population-based survival estimates for childhood cancer in Australia during the period 1997-2006. Br J Cancer 2010; 103:1663-1670

75 Bao PP, Zheng Y, Wu CX, Peng P, Gong YM, Huang ZZ et al. Population-based survival for childhood cancer patients diagnosed during 2002-2005 in Shanghai, China. Pediatr Blood Cancer 2012; 59:657-661

76 Boschmonar MG, Alvarez YG, Garcia AM, Soto TN, Roger MC, Garrote LF. Childhood cancer survival in Cuba. Eur J Epidemiol 2000; 16:763-767

77 Wiangnon S, Veerakul G, Nuchprayoon I, Seksarn P, Hongeng S, Krutvecho T et al. Childhood cancer incidence and survival 2003-2005, Thailand: study from the Thai Pediatric Oncology Group. Asian Pac J Cancer Prev 2011; 12:2215-2220

78 Swaminathan R, Rama R, Shanta V. Childhood cancers in Chennai, India, 1990-2001: incidence and survival. Int J Cancer 2008; 122:2607-2611

A. Rajalekshmi KR,Abitha AR, Anuratha N and Sagar TG., Time trend in frequency of

occurrence of major immunophenotypes in paediatric acute lymphoblastic leukemia cases as

Formatted: title

21

experienced by Cancer Institute, Chennai, south India during the period 1989-2009. Indian J

Cancer. 2011 Jul-Sep;48(3):310-5. doi: 10.4103/0019-509X.84932.

B. Wakeford R. Childhood leukaemia following medical diagnostic exposure to ionizing radiation in

utero or after birth. Radiat Prot Dosimetry. 2008;132(2):166-74. doi: 10.1093/rpd/ncn272. Epub 2008

Oct 15.

C. Hussein H, Sidhom I, Naga SA, Outcome and prognostic factors of acute lymphoblastic leukemia in

children at the National Cancer Institute, Egypt. J Pediatr Hematol Oncol. 2004 Aug;26(8):507-14.

D. Spinetta JJ, Jankovic M, Masera G et al. Optimal care for the child with cancer: A summary statement

from the SIOP Working Committee on Psychosocial Issues in Pediatric Oncology. Pediatr Blood Cancer.

2009 Jul;52(7):904-7. doi: 10.1002/pbc.21863.

E. Best Care at Lower Cost: The Path to Continuously Learning Health Care in America. A report of the

Institute of Medicine, USA. Released, Sept 6th 2012. http://www.iom.edu/Reports/2012/Best-Care-at-

Lower-Cost-The-Path-to-Continuously-Learning-Health-Care-in-America.aspx

F. Gopal S, Wood WA, Lee SJ et al., Meeting the challenge of hematologic malignancies in sub-Saharan

Africa. Blood. 2012 May 31;119(22):5078-87. doi: 10.1182/blood-2012-02-387092.

.

e 1. Selected development indicators in economies classified by the World Bank according to gross

national income (GNI) per capita, using the Atlas method (current US dollars). Data from

http://databank.worldbank.org/ddp/home.do?Step=1&id=4 for 2010

Formatted: Normal

Field Code Changed

Formatted: No underline, Font color:Auto

Field Code Changed

Formatted: No underline, Font color:Auto

Formatted: No underline, Font color:Auto, Superscript

Formatted: Normal

Field Code Changed

Formatted: No underline, Font color:Auto

22

Indicator High High

(OECD)

Upper

Middle

Lower

Middle

Low

Number of countries/economies in World Bank income groups 70 31 54 54 36

GNI per capita (Atlas Method, current USD) 38764 40142 5865 1595 533.99

GDP per capita (Current USD) 41062 42220 7326 1882 579

Population (millions) 1127 1033 2472 2494 799

Population (% Rural) 19.7 20 40.2 61.6 72.4

Population (% 0-14) 17.3 17.0 21.9 31.9 39.25

Age dependency ratio (percentage of working age population) 49 49.7 42.9 59.2 77.0

Poverty headcount ratio at <$1.25 per day, percent of population - - 9 30 48

Health expenditure per capita, (current USD) 4876 5093 379 70.9 26.9

Health expenditure, total (percentage of GDP) 12.5 12.9 6.0 4.3 5.3

Health expenditure, public (% GDP) 8.2 8.4 3.3 1.7 2.1

Health expenditure, private (% GDP) 4.4 4.5 2.8 2.6 3.2

Health expenditure, public (% total health expenditure) 65 65 54.4 39.5 38.8

Out-of-pocket health expenditure (% total health expenditure) 13.7 13.6 33.3 52.8 48.1

Number of physicians per 1000 people 2.7 2.8 1.7 0.8 0.2

Number of nurses or midwives per 1000 people 7.0 7.2 2.6 1.5 0.5

Literacy rate (% females 15 and above) 97.9 - 91.3 62.3 56

Literacy rate (% males 15 and above) 98.7 - 93.5 80 69.7

Income groups used in the table are based on the 2011 GNI per capita calculated according to the World Bank Atlas Method (http://data.worldbank.org/about/country-classifications). The groups are: Low Income: $1,025 or less Low Middle Income: $1,026-$4,035 Upper Middle Income: $4,036-$12,475 High: $12,476 or more OECD: Organization for Economic Cooperation and Development is a group of 34 countries founded in

1961 to promote economic progress and world trade. Most of these countries are high income

countries with a very high index of development (UNDP classification; see

http://hdr.undp.org/en/media/HDR_2011_EN_Complete.pdf)

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

Formatted ...

23

24

Table 21. Deaths by Cause in Children (both sexes) aged 5 to 14 years, 2008

Data from http://apps.who.int/ghodata/?vid=10012 Global Burden of Disease Causes of Death 2008.2 Percentages relate to the disease related deaths (except disease related deaths, which relate to deaths from all causes)

World

%

High Income %

Upper Middle Income

% Lower Middle Income

% Low Income %

Population 5-14 yrs

1,231,558,000 127,027 166,082,000 715,046,000 206,201,000

All Causes 1,265,306 17,397 66,973 721,871 459,065

Disease related 1,059,408 83.7 10,251 59.0 42,987 64.1 508,492 70.4 360,305 78.4

Infections, Parasites

496,605 46.9 608 5.9 14,738 11.0 269,783 53.0 211,475 58.7

Non-communicable diseases

222,784 21.0 9,126 89.0 24,030 55.9 136,270 26.8 53,357 14.8

Malignant Neoplasms

53,464 14.6 3,317 32.0 7,984 18.6 30,541 6.0 11,622 3.2

25

Table 32. Malignant NeoplasmsDeaths by Cause in Children (both sexes) aged 0-4 years, 2008

World %

High Income

% Upper Middle Income

% Lower Middle Income

% Low Income %

Population 0-14 yrs

634,176,000 62,609,000 83,258,000 382,776,000 121,531,000

All Causes 8,327,552 85,237 405,329 4,586,599 3,249,547

Disease related

8,044,722 96.6 77,766 91.2 384,684 86.0 4,441,085 96.8 3,140,383 96.6

Infections/Parasitic

3,152,673 39.2 3,849 4.9 82,902 23.4 1,568,202 35.3 1,497,720 47.7

Non-Communicable Diseases

614,759 7.6 31,769 40.8 87,542 22.7 347,009 7.8 148,439 4.7

Malignant Neoplasms

40,061 0.50 1,901 2.4 4,728 1.2 23,341 0.5 10,091 0.3

Data from http://apps.who.int/ghodata/?vid=10012 Global Burden of Disease Causes of Death 2008.2 Percentages relate to the disease related deaths (except disease related deaths, which relate to deaths from all causes)

26

Table 43. Age specific incidence rates of cancer by Human Development Index (HDI) 1

Population Group

Incident cases in 0-14 year olds

Age-specific incidence

(0-14 years) per 100,000

Cumulative risk*

World 175,058 10.4 0.16

Very High HDI 24,855 14.9 0.22

High HDI 25,080 12.6 0.19

Medium HDI 110,027 9.5 0.14

Low HDI 15,035 9.7 0.14

1 See http://hdr.undp.org/en/media/HDR_2011_EN_Complete.pdf for definitions. *Risk of developing cancer during a lifetime – expressed here as a fraction of 1. Multiply by 100 for percentage of the population expected to develop cancer.

27

Table 54. The way forward forElements of a model of care for pediatric cancer care in low- and

middle-income countries

Society

Awareness of childhood cancer

A population educated in the signs and symptoms that could be caused by cancer and free of stigmata associated with cancer

Effective access to primary care

Primary care providers educated in the signs and symptoms that could be caused by cancer and who know where to refer patients for further investigations

Sufficient primary care providers in the family’s region to permit good access to primary care

Health insurance program in which pre-payment and pooling of costs allows the family is not deterred from a medical consultation on account of cost. Universal insurance is an important element of sustainability

Diagnosis and risk stratification

Access to prompt biopsy of suspected cancer

Secondary care facilities capable of performing biopsy where necessary and who know the importance of prompt pathological diagnosis

Pathology service Access to a pathology service with well-trained pathologists, the necessary equipment, and reagents to make an accurate diagnosis. Prompt reporting of pathology results to the clinician in charge.

Referral pathway Prompt referral to a tertiary care center when necessary

Staging and risk stratification

Access to necessary imaging procedures to identify the stage or degree of spread of the cancer

Treatment

Treatment protocols adapted to local conditions

Standardized protocols, previously agreed upon for all pediatric cancers, and ideally, partnership in a disease-specific network or cooperative group capable of conducting clinical research, even if primarily designed to assess suitability of a regimen in the local setting

Access to medications and supportive care

An efficient pharmacy which procures drugs prior to running out, and a national or regional system that ensures that all needed drugs are available, including those required for pain control, such as opioids

Infrastructure

Psychosocial and financial support

Access to subsidized travel in order that lack of money does not prevent diagnosis at the earliest possible time

Psychosocial support with mechanisms to avoid lost earnings, local lodging when home is far away, and access to social assistance and advice. Presence of a patient advocate to identify non-medical family and patient needs

Outcome evaluation and quality improvement

A system of data collection comprising a hospital registry, as well as necessary data bases and data managers able to efficiently collect data. A systematic approach and appropriate resources (tracking officer) to ensure follow up of all patients to determine the outcome of therapy Collection of clinically relevant data Evaluating of delays in cancer diagnosis and referral, abandonment pattern, etc. Long-term follow-up of cancer patients and evaluation of their survival

28

Personnel and a multidisciplinary approach

Oncology team A tertiary care facility with the necessary specialists, including pediatric oncology, radiation therapy, surgical oncology, oncology nursing, etc. Multidisciplinary communication among all specialists through regular “tumor board” meetings and patient rounds to assure that each patient gets the best possible treatment available. The tumor board can also serve additional functions such as education of junior staff and trainees.

Training and professional support

Professional organizations that provide continuing educational opportunities and discussion of issues that need to be addressed within the region or country, with a special emphasis on local issues and concerns, such as abandonment of treatment, cancer registration, etc.

Advocacy and sustainability

Cancer registry Data on cancer incidence (and mortality) Provision of the pertinent information driven by local needs International collaboration to make comparisons and improve methodology

Non-governmental organizations

Local or national NGOs that advocate for pediatric cancer and may assist in providing funding for necessary drugs, salaries, construction, etc. Such NGO’s may play an important role in ensuring sustainability and advocating for improvements in the health system and the development of cancer control plans, and the means to carry them out

External collaboration

Associations with international professional groups or organizations in order to ensure access to relevant new information concerning the efficiency of the health system, and who may be in a position to affect, for example, drug prices through bulk purchasing and provide access to participation in international studies or highly skilled trainers and educators who can help improve “medical” education as well as advise governments on the development of the national health system and the public on healthy living styles. Twinning programs with other centers can facilitate development of specific services and treatment protocols.

NGO, non-governmental organization

29

Figure Legends Figure 1a. Incidence rate of childhood cancer (0-14 years) and percentage of all incident cancer cases in various populations and geographical regions. The 10 countries included are the 5 with the highest fraction of cancers in the 0-14 age group- and the 5 with the lowest fraction. Only countries with more than 1000 children with cancer per year are included.1 Figure 1b. Mortality rate of childhood cancer (0-14 years) and percentage of all cancer deaths in the same populations as in Figure 1a. HDI, human development index Figure 2. A model of the demographic transition as countries move from pre-industrial revolution agricultural economies to industrial economies. The population pyramids show how the population structure changes during the transition period. Most developing countries are in phases 2 and 3, but some regions may still be essentially in phase I, where the high birth rate is balanced by the high death rate. In phase 5, the population may begin to contract as the birth rate continues to fall below replacement level (i.e., 2 children per family). Figure 3. Incidence of cancer in age 0-14 years as a percentage of the total cancer incidence.73 More developed regions include all regions of Europe, Northern America, Australia, New Zealand and Japan. Less developed regions include Africa, Asia (excluding Japan), Latin America and the Caribbean, Melanesia, Micronesia and Polynesia. Figure 4. Incidence rates of childhood cancer registered in the countries shown in 1980s and assembled in an international comparative study.10 ASR, age-standardised incidence rate (world). Figure 5. Five-year survival rates of children (age 0-14 years) with cancer diagnosed in countries, regions and periods shown.74-78

Senior Editor, The Lancet Oncology Email: [email protected] ----- Dear Dr McLarnon, Here are the responses to the reviewers comments. Reviewers' comments: Reviewer #1: A well written contribution with some concrete and realistic suggestions. I learned a lot from this paper. Some minor grammer that will be sub-edited out but no real change suggested ----- Reviewer #2: This is a scholarly overview of the issues surrounding childhood cancer in LMIC Comments MAJOR SUGGESTIONS: 1. p 3 para 3 While the importance of AYA cancer is not to be ignored , the bulk of this review addresses cancer in the 0 - 14 age group, and the paragraph on cancer in adolescents and young adults is superficial in comparison to the major focus of the review. Its content shifts the frame of reference for analysis of both incidence and mortality, and demands different analysis and a different framework. I would recommend removing this paragraph and focusing the review on the 0 -14 age group, consistent with the title. If retained, the paragraph needs expansion and a supporting table - but that would be the less preferable choice. Paragraph three has been removed. 2. p 3 para 1 - this paragraph would benefit from clarification of some of the quoted statistics which are confusing. If 84% of incident cancers occurred in the 0-14 age group ( for which a reference is needed) and for this group n = 148000 in LMIC, then the total cancer incidence in all age groups in this group of countries would be ~ 209000 which seems unlikely, but if true should be backed by a reference. In addition, line 6 indicates that "...only 1.45% of all cancer deaths were in the 0- 29 age group, due to the fact that cancer is a disease or group of diseases oaf people above the age 65 years" - this too seems incompatible with the description above of 84% of incident cases occurring in the 0 - 14 year age group. We agree that the first sentence is confusing – perhaps it was not clear that 84% referred to proportion of LMIC children to total children and not to all ages cancers. A second cause for confusion is the switch from the World Bank definitions to the definitions of more and less developed countries used by Globocan, from whom these data derive. The problem is that data relating to the number of incident cases of cancer is hard to come by the lower down the scale of development one goes, hence estimates are the only option, and Globocan is the only source of such estimates. This would take several lines to explain, so to avoid using up more space, the switch in definition is mentioned and the Globocan definition provided. In effect, therefore, the entire paragraph has been rewritten.

Letter to Editor.docx

3. p5 para 2 While the incidence data and its interpretation is critical, it might be valuable to identify the limitations of existing registries prior to the section on incidence. While the last seven lines of para 2 address that in part, a critical statement is made in paragraph 5 on page 6 " The level of coverage achieved by the registry clearly influences incidence and mortality rates" is a legitimate caveat and precursor to any discussion about incidence and mortality rates. It would also be of value, given the noted variability in incidence distribution by diagnosis demonstrated in figure 4 to comment on whether there may be any systematic variation in reporting of particular disease entities, e.g central nervous system tumors, or biases resulting from twinning, outreach or other development programs. The bulk of paragraph 5 on page 6 has been transposed to the section on incidence so that the caveats now precede the incidence data. Comments have been slightly expanded with respect to the issue on systematic variation. It should be said that while cancer registration has its limitation, there is NO other relevant data source. In addition, the data presented in what is now Figure 5 have been peer reviewed for comparability and this is why they were selected for this presentation. 4. p10 para 1 Adapted treatment regimes - while this the heading for this section, there is no clear recommendation for adapted regimes scaled to the particular economic, health status (e.g malnutrition etc) and geographic conditions - te focus is more on stage and risk adapted therapies adopted in HICs. It would be perhaps valuable to clarify what the main message of this section is. In response to this comment, reference to adaptation has been divided into two parts – the possibility that good results can be obtained without extensive imaging, and the need to adapt western protocols to the local resources. The text has been modified accordingly. MINOR 5. p5 para 2 "which are generally uncommon in LIC" - does this imply true low incidence or would it be better framed as uncommonly reported given the statement that they are underdiagnosed that precedes this statement. Which of these two apply is unknown – it may well be a mixture of both. The text has been slightly modified to make it clearer that there is, at present, no good way to distinguish between them and both probably apply, but to varying degrees in different countries. Given the absence of precise information, we would rather not state “uncommonly reported” since this avoids the issue rather than addressing it. 6. p 6 para 1 "Almost 100% of of BLs in Africa and over 50% of BLs in most countries of the world contain EBV" id succeeded in the same paragraph by the statement that in the wealthiest countries.... a majority of BLs are EBV negative. and a further statement that in the BFM series only 11% were associated

with EBV. THese two statements merit harmonization or interpretation as they are intuitively contrary. The statements are not, in fact, contrary, since the majority of countries in the world fall into the LMIC and available data indicates that such countries have a frequency of EBV association in BL of greater than 50%. Only in the wealthy countries (as exemplified by the BFM group) is the frequency of EBV association low. However, since this did not come across clearly to the reviewer, the text has been modified and should now be clear. In addition a review of the epidemiology of BL is cited again so that a source of additional information is available to those interested in this issue. 7. p6 para 2 lines 2 - 3 - grammatical correction required. It is not clear what the reviewer is referring to, but minor changes which probably improve 2 sentences in this part of the manuscript have been made in the hope that we have identified his/her concern. 8. p7 para 2 - it would be valuable to more robustly clarify the concept of abandonment - it does not emerge clearly until the last line of the paragraph but is a key concept of the access to care issue. How about “premature termination” of therapy? This has been added in parenthesis at the first instance of the phrase “abandonment of therapy.” The term “abandonment” is in the process of being added as a formal MESH term, but this is not yet the case. 9. p 7 para 3 - please spell out PACT. Done, although this also necessitated spelling out IAEA 10. p9 para 1 - many typographical errors - line 6 "negligent" implies intent - would a better descriptor be "erroneous" or "errors in judgement" or some other more neutral term. This sentence has been removed completely, although we do not believe that any intent is implied by the word “negligent” – rather, it indicates carelessness irresponsibility, total lack of concern etc. – in fact, precisely what was meant. 11. p 9 para 5 - Clarify that the risk of infection exists in a setting where physical separation has NOT occurred. Similar attention to detail around the statements about "prior to this care providers should understand the basic elements of supportive care" - in time sequence or by priority? Point 1 has been clarified, point 2 omitted. 12. p 20 Table 2 - Should the title of this table not be similar to the title of Table 1 " Deaths by Cause in Children ( both sexes) aged 0-4 years 2008". Yes: this error has been corrected.

13. p 22 Table 4 This table appears to describe the ideal plan for optimizing the care of children with cancer in MLICs rather than the "Way forward" which seems to suggest a developmental plan - perhaps a different title would be appropriate. A subtle difference, but we have offered an alternative title. ----- Reviewer #3: Thank you for giving me the opportunity to read this paper. 1. There is no "Search Strategy and Selection Criteria" presented. A search strategy is being prepared. 2. Clarification of the definition of HDI is required; do the authors use the post-2011 definition? The UNDP currently uses four categories to describe development status based upon three measurables, not four. The World Bank Classification of countries by GNI is extensively used by the authors and should be defined. Definitions and references have been provided. The definition of HIC on the UNDP website is used and cited. This is the definition used in the 2011 UNDP report. There are in fact 2 indicators for education: mean number of years of schooling for adults aged 25 years and expected years of schooling for children of school entering age. The other indicators are life expectancy at birth and the logarithm of income. The World Bank classification of countries by GNI is given in Table 1. 3. The authors fail to acknowledge that developments in HIC's have taken 250years, and a ruinously expensive financial investment over the last 60-70 years, to achieve. The authors present no alternative to encouraging LIC's to replicate the philosophies of HIC's rather than devising their own culturally and economically appropriate strategies. It is difficult to know precisely what the reviewer is suggesting. It is not clear what is added by a discussion of the time taken to reach the HIC level of development, or to which particular ruinously expensive financial investment over the last 60-70 years he/she is referring to. It is also unclear how the reviewer thinks that extremely poor countries are going to be able to develop an alternative to the HIC’s “philosophy” and in any event, they are a considerable way along this track, the only alternative being traditional medical systems, for which there is no evidence of efficacy in cancer and it would seem highly unlikely that they will move away from the medicine of HIC. Perhaps the reviewer is arguing against providing aid to LMIC, since they may become dependent upon it. This is certainly true of the heavily indebted countries, but would seem to be beyond the scope of this paper. In effect, we don’t see an alternative, unless we have misunderstood the reviewer. While “developments” in HICs have taken centuries, this was because of a lack of pre-existing knowledge. This is no longer the case, and as shown by S. Korea, development can take place in a matter of a decade or so if all required elements are in place (particularly good governance, emphasis on education, public health etc.).

4. The revolution in many LIC's is likely to be stepwise and the authors could usefully lay greater emphasis on what they consider "first steps". This, I venture to suggest, is likely to include data collection and data management, and here "international expertise" might indeed have a role. Data are required to convince governments that children with cancer merit priority over children with other life-threatening disease, and to justify the authors' extensive recommendations. We do not believe that children with cancer merit priority over children with other life-threatening diseases, but that all have the same right to the best available therapy for their disease – albeit, this right cannot be exercised in many circumstances. Cancer is an expensive disease to treat, but some children with cancer are successfully treated today even in LIC and some survive. We are unaware of any country that does not attempt to provide some pediatric cancer care, however minimal. Improving what there is already is step number one. Step number two is to build capacity. These points have been added to the text along with other remarks relating to stepwise progression. It has been made clear that the long list of recommendations are unlikely to be undertaken simultaneously, but that what can be done depends upon the present level of development and what should be done must be decided by the government, aided by advice from organizations such as the WHO and the many NGOs which focus on this area. We have also made this clear in the manuscript. One of the difficulties with which the authors have clearly grappled is that not all LIC's have a similarly LI!! Thus there is a world of difference between Mexico and Mali, between Honduras and Haiti, between Bolivia and Burkina Faso. One accepts that in a review of this nature generalizations are inevitable but the inequity in national economies makes such generalizations particularly dangerous. Of the 50 countries with the lowest GNI, 36 are in sub-Saharan Africa. It is easy to be cynical but the fact that pathologists with immunohistochemistry resources are essential, that new radiotherapy units including physical resources, radiotherapists and radiographers are essential, that hostels at each treatment centre are essential, that surgeons are essential, that educated primary care physicians are essential, that drugs on the WHO essential drugs list are essential, that treatment centres with clean bathrooms are essential, that competent nurses are essential, that paediatric oncologists are essential and that psychosocial and social workers are essential, is not news. Neither are these ambitions credible within a reasonable time frame throughout most of the not-developing countries of sub-Saharan Africa. Perhaps there is a need to split the paper into discussion of MIC's and LIC's seperately. What is immediately required in LIC's is a strategy to provide adequate care in the absence of these "essentials". It is not clear what danger the reviewer is referring to in the first part of his critique, or what strategy in LICs could provide adequate care in the absence of these essentials. One can only do the possible. This would appear to be impossible. Whatever can be done may take a considerable time, but is there an alternative? Similarly, it is not possible to discuss the vast number of differences among the LIC in the world and to discuss what could and could not be

done in each of these situations, but we can try to make more people aware of the inequities that exist in the modern world, and to empathize with the plight of the poor. At least some may benefit. As for separating LICs and MICs, these are somewhat arbitrary distinctions and there are vast differences in each of these groups. We have tried, however, to indicate that MICs have at least some quite high quality resources while they often also have regions that differ little from LIC. Thus, MIC face similar problems to LIC, although have more resources with which to deal with them and are not so dependent upon outside aid/collaboration. Incidentally, we have not advocated simply giving aid to LICs, and assistance would be more in the context of knowledge transfer than simply money. Although we have not created separate sections for MIC and LIC, we have tried to point out similarities and differences throughout the text. 6. The submission is very long (approximately 6,000 words), occasionally repetitive and there are frequent errors of grammar and syntax. The Abstract is also longer than recommended. Proof reading has been perfunctory and personal reminders have been included in the text. We have tried to remove repetition as much as possible. The entire manuscript has been revised and it has been shortened by several hundred words. 7. The number of figures could be reduced. Figure 5 could comfortably be replaced by a brief text description. In figure 3 Africa is included as a "less developed" region thereby conflating the incidence data from this group. The disparity between Africa and the rest of the world is therefore greater than illustrated. Table 4 replicates the text from "Cancer Registration" (p6) et seq. and includes all of the authors' proposals. There seems to be no justification for the inclusion of both and the Table is easier to assimilate. We have tried to add some country data as well as regional data in the figures, and have been at pains to point out that agglomerated data such as ”Africa and Asia” is provided to give a general idea, but masks quite major differences that exist within these large regions. One simply cannot do justice to the global situation in an article of this type, and whatever one does is open to criticism. We have kept replication of information from the table and text to a minimum, but to some text is necessary to create a background for the table. Editorial points to be addressed: 1. Reviewer 4 comments that there are considerable variations in income and resources amongst low-income countries and, as such, there are big differences between very low income countries and middle income countries. Please ensure this is reflected in the manuscript with discussion of the specific problems facing those categories of countries where appropriate. These distinctions have been highlighted, while still emphasizing that the interventions to improve the situation (training, international collaboration, data management) are common to LIC and MIC. 2. Figure 2: details of the various stages should be in the figure legend.

We have changed the figure and legend as requested. 3. There are too many figures and tables at the moment. As Reviewer 4 notes, figure 3 can be replaced with a short text description. Table 4 should be a panel; please remove redundancy between this panel and the text (as per reviewer 4's comments). We have removed figure 3 as requested, along with figure 5, but substituted other figures – one at the last minute since it made an important point. Hopefully, this will be acceptable, but if not, we would be pleased to follow the editors recommendations. We have done our best to remove redundancy between the panel and text but some repetition is inevitable, since the text provides more information. 4. Reviewers commented on the flow of the article, which made it difficult to read. Please bear this in mind while you are revising the paper. We hope that the extensive revisions have made the paper easier to follow and more readable. 5. We require signed conflict of interest statements from all authors. If there are no conflicts of interest, authors should state that there are "none". An extract from the journal's policy on conflict of interest is shown below: "The Lancet Oncology requires all authors to declare a Conflict of Interest if they, or their institution, has financial or personal relationships with other people or organisations that may inappropriately influence their actions. The information declared in these statements will be used to determine acceptability. In particular, authors who, within the past 3 years and with a relevant company or competitor have stock, equity, a contract of employment, or a named position on a company board; or hold (or is applying for) a relevant patent (this restriction is for the life of the patent); or have (or will do) a fee from any organisation other than The Lancet Oncology to write, be named on, or submit a paper, will be ineligible to offer a manuscript to the journal". 6. We require signed statements from all authors stating their contribution to the manuscript. This will be done. For author contribution and conflicts of interest ststements, please use the Author Statement form at: http://download.thelancet.com/flatcontentassets/authors/tlo-author-signatures.pdf. 7. We require written consent from any individuals who are cited in acknowledgments or personal communications. The following format can be used: "I permit <corresponding author> et al to list my name in the acknowledgments section of their manuscript and I have seen a copy of the paper <full article title>" There are none.

Formatted: Font: Bold

"I permit <corresponding author> et al to cite a personal communication from me in their manuscript <full article title>" 8. We require confirmation that the paper has not been submitted to another journal, and has not been published in whole or in part elsewhere previously. We confirm both. 9. For papers listed in references that are "in press" we need to see a galley proof and letter from the publisher stating that it is 'in press' as well as the full expected citation (ie, publication date/volume/issue etc). There are no papers in press 10. Images that have been published previously should be accompanied by a statement indicating permission to reproduce the image. If required, further assistance can be obtained from the editorial team. If you have borrowed published images from colleagues, you must obtain permission from the publisher of the paper, not just from the authors. If all the figures are your own and have not been published before then this requirement does not apply. Figure 5 was published in a Chapter on Childhood Cancer authored by ESF and others in World Cancer Report 2008 of the International Agency of Research on Cancer P Edited by Peter Boyle and Bernard Levin. Permission to publish is almost certain and in process. 11. Please provide: one preferred degree qualification per author and indicate any full professors; affiliation details (department, institute, city, state, country) for each author; full institutional correspondence address for corresponding author. 12. Please supply two copies of any annotated figures as separate jpgs (one copy annotated and one non-annotated) with a resolution of 300 dpi at a width of 107 mm. Our in-house illustrators will annotate according to journal style using the annotated figures as a guide. We have no annotated figures 13. Figure legends should be a maximum of 30 words; please ensure they are within this limit where amended. This is so for all except for Figure 1 – explanation above. 14. References should be in the Vancouver style and numbered in the order in which they first appear in the manuscript. If the references "move" from the body text into tables or figures, please maintain the sequence of citation. Please ensure tables and figures are cited correctly in the body text to prevent the need for renumbering of references should the table and figure citations subsequently move. References relevant to the figures are cited according to the main list, but are all also in the text, so there should be no need for renumbering.

15. Please keep your abstract to no more than 200 words. I believe it is 204 words 16. Please ensure that reference numbering throughout the manuscript is not inserted with electronic referencing software, such as Endnote. All reference numbers are Word numbers. Text citations have been put into square brackets. 17. Please supply a section entitled "Search strategy and selection criteria". This should state clearly the sources (databases, journals, or book reference lists, etc) of the material covered and the criteria used to include or exclude studies. Please state which search terms, languages and date ranges were used. This issue is discussed above and by e mail 18. Please provide a "Conflict of interest" section. These statements should match exactly those given on the Authors' signed conflict of interest forms. Authors should be referred to by their initials in this section. If there are none, then please state "The authors declared no conflicts of interest" or "The other authors declared no conflicts of interest". Will do, although needs to be completed 19. Please provide a "Contributors" section stating the contribution each author has made to the paper. These statements should match exactly those given on the Authors' signed forms. Authors should be referred to by their initials in this section. Will do – again, not quite finished --- End of editorial comments ---

Many thanks

Ian Magrath and Scott Howard