Cancer incidence and mortality among young adults aged years worldwide in 2012: a population-based study

Transcription

1 Cancer incidence and mortality among young adults aged years worldwide in 2012: a population-based study Miranda M Fidler, Sumit Gupta, Isabelle Soerjomataram, Jacques Ferlay, Eva Steliarova-Foucher, Freddie Bray Summary Background To date, the burden of cancer among young adults has rarely been studied in depth. Our aim was to describe the scale and profile of cancer incidence and mortality worldwide among year-olds, highlighting major patterns by age, sex, development level, and geographical region. Methods We did a population-based study to quantify the burden of young adult cancers worldwide. We defined young adult cancers as those occurring between the ages of 20 and 39 years because these individuals will have passed puberty and adolescence, but not yet experienced the effects of hormonal decline, immune response deterioration, or organ dysfunction associated with chronic health conditions. Global, regional, and countryspecific (n=184) data estimates of the number of new cancer cases and cancer-associated deaths that occurred in 2012 among young adults were extracted in four 5-year bands from the International Agency for Research on Cancer s GLOBOCAN 2012 for all cancers combined and for 27 major types as defined by the International Classification of Disease, tenth revision. We report the number of new cancer cases and cancer-associated deaths overall and by sex alongside corresponding age-standardised rates (ASR) per people per year. We also present results using four levels of the Human Development Index (HDI; low [least developed], medium, high, and very high [most developed]), which is a composite indicator for socioeconomic development comprising life expectancy, education, and gross national income. Findings new cancer cases and cancer-associated deaths occurred among young adults worldwide in 2012, which equated to an ASR of 43 3 new cancer cases per people per year and 15 9 cancer-associated deaths per people per year. The burden was disproportionally greater among women and the most common cancer types overall in terms of new cases were female breast cancer, cervical cancer, thyroid cancer, leukaemia, and colorectal cancer; in terms of deaths, female breast cancer, liver cancer, leukaemia, and cervical cancer were the main contributors. When assessed by development level and geographical region, the cancer profile varied substantially; generally, the burden of infection-associated cancers was greater in regions under transition. Cancer incidence was elevated in very high-hdi regions compared with low-hdi regions (ASR 64 5 vs 46 2 cancer cases per people per year); however, the mortality burden was 3 times higher in low-hdi regions (ASR 25 4 vs 9 2 cancer-associated deaths per people per year), reflecting differences in cancer profiles and inferior outcomes. Lancet Oncol 2017; 18: Published Online October 27, S (17) See Comment page 1554 Section of Cancer Surveillance, International Agency for Research on Cancer, Lyon, France (M M Fidler PhD, I Soerjomataram PhD, J Ferlay MSc, E Steliarova-Foucher PhD, F Bray PhD); and Department of Haematology/Oncology, The Hospital for Sick Children, Toronto, ON, Canada (S Gupta MD) Correspondence to: Dr Miranda Fidler, Section of Cancer Surveillance, International Agency for Research on Cancer, Lyon, France fidlerm@fellows.iarc.fr Interpretation The global cancer burden among year-olds differs from that seen in younger or older ages and varies substantially by age, sex, development level, and geographical region. Although the cancer burden is lower in this age group than that observed in older ages, the societal and economic effects remain great given the major effects of premature morbidity and mortality. Targeted surveillance, prevention, and treatment are needed to reduce the cancer burden in this underserved age group. Funding International Agency for Research on Cancer (IARC) and European Commission s FP-7 Marie Curie Actions People COFUND. Copyright The Author(s). Published by Elsevier Ltd. This is an Open Access article published under the CC BY 3.0 IGO license which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. In any use of this article, there should be no suggestion that WHO endorses any specific organisation, products, or services. The use of the WHO logo is not permitted. This notice should be preserved along with the article s original URL. Introduction Cancer is a major cause of morbidity and mortality across all age groups in both developed and transitioning economies. 1 To date, research has focused on cancer in children and at older ages. This focus could be because a cure in children would lead to decades of productive life and because the burden of most epithelial cancers is increasing with age and cancer is becoming a leading cause of death in many countries. As a result, in relative terms, cancers that occur during young adulthood have been investigated to a lesser degree. These cancers are a bridge between paediatric and adult oncology, and represent a distinct spectrum of disease among young adults who have a Vol 18 December

2 Research in context Evidence before this study Although substantial research on cancer in children and older age groups has been undertaken, the burden of cancer among young adults (20 39 years) has rarely been studied in depth, and is often overlooked by cancer researchers and policy makers alike. To determine the current evidence describing young adult oncology globally we searched PubMed for publications in English, without any date restrictions, using the terms young adult, cancer, global, incidence, and mortality. Because there is no internationally agreed definition for young adults, the term adolescent and young adult was also included, because these groups are sometimes combined. From our search it was clear that there was a paucity of research on young adult cancer in countries with limited resources, with nearly all of the literature focusing on high-income countries, particularly the USA, Canada, and several European countries. Thus, we sought to describe the magnitude and patterns of cancer in young adults globally to improve the understanding of cancer in this age group. Added value of this study To our knowledge, the present study is the first to explore the global burden of cancer among young adults for the given period and age group. By assessing all cancers combined and 27 major types, we provide evidence on the cancer burden distribution by age, sex, development level, and geographical region. Furthermore, our study explores barriers to improving cancer information and outcomes in young adults in low-income, middle-income, and high-income countries, and highlights future opportunities for improvement. Implications of all the available evidence Our data serve as a status report to aid young adult oncology researchers and increase awareness of cancers among this underserved age group. The considerable societal and economic effects of cancer in young adults as a major cause of premature morbidity and mortality requires targeted, resource-dependent interventions. Through continuous surveillance, vaccination, early detection programmes, and curative treatment the cancer burden can be reduced in this underserved age group. For more on GLOBOGAN see large proportion of their expected lifespans remaining, contribute substantially to the economy, and have an important role in caring for their families. 2 Thus, international research investigating the specific issues unique to this age group of cancer patients is needed to improve cancer-associated outcomes. In this study, we assess the scale and profile of young adult cancers globally. We report the estimated number of new cases and deaths in 2012, as well as the corresponding incidence and mortality, describing variations by human development level and geographical region. Methods Study design Although 0 14 years is broadly accepted as the age range for childhood cancer and years is broadly accepted for adolescent cancer, the age range for young adult cancer is less clear because there is no uniform opinion on the upper age limit for this group, 3 with the ages 24 years, 4 25 years, 5 29 years, 6,7 34 years, 8 and 39 years 9 all currently being used. For this study, we define young adult cancers as those that occur between the ages of 20 and 39 years, in line with that suggested by the Adolescent and Young Adult Oncology Progress Review Group. 9 As the group states, the rationale for using this age range relates to biological and physiological maturity, with individuals having passed puberty, but not yet experienced the effects of hormonal decline, immune response deterioration, or organ dysfunction associated with chronic health conditions. 9 Thus, using this definition, we provide an inclusive assessment of the scale and profile of cancers in this age group that overlaps partially or entirely with that used by other organisations involved in young adult oncology. Data sources and statistical analyses To quantify the cancer burden for this age group, global, regional, and country-specific (n=184) estimates of the number of new cancer cases and cancer-associated deaths that occurred in 2012 among young adults were extracted in four 5-year bands from the International Agency for Research on Cancer s GLOBOCAN To make comparisons with younger and older age groups, the estimated number of new cancer cases and deaths among individuals aged 0 19 years, years, and 60 years and older was also extracted. Further information on the methods used to create the incidence, mortality, and population estimates used in GLOBOCAN 2012 is provided elsewhere. 1 We present the results for all cancer types included in GLOBOCAN 2012, defined by the International Classification of Disease, tenth revision (ICD-10): 10 lip and oral cavity (C00 08), nasopharynx (C11), other pharynx (C09 10, C12 14), oesophagus (C15), stomach (C16), colorectal (including anus; C18 21), liver (C22), gallbladder including biliary tract (C23 24), pancreas (C25), larynx (C32), lung including trachea (C33 34), melanoma of the skin (C43), Kaposi s sarcoma (C46), female breast (hence forth referred to as breast; C50), cervix uteri (C53), corpus uteri (C54), ovary (C56), prostate (C61), testis (C62), kidney including renal pelvis and ureter (C64 66), bladder (C67), brain and CNS (C70 72), thyroid (C73), Hodgkin s lymphoma (C81), non-hodgkin lymphoma (C82 85, C96), multiple myeloma and immunoproliferative diseases (C88, C90), Vol 18 December 2017

3 Both sexes Men Women New cases % ASR New cases % ASR New cases % ASR All cancers except non-melanoma skin cancer % % % % % % % 9 9 Thyroid % % % % % % 2 0 Colorectal % % % % % % 0 8 Brain and CNS % % % 1 6 Non-Hodgkin lymphoma % % % 1 5 Testis % % 2 7 Ovary % % 2 6 Stomach % % % 1 1 Melanoma of skin % % % 1 4 Lip or oral cavity % % % 0 8 Hodgkin s lymphoma % % % 1 0 Lung % % % 0 9 Kaposi s sarcoma % % % 0 7 Corpus uteri % % 1 4 Nasopharynx % % % 0 5 Kidney % % % 0 5 Esophagus % % % 0 3 Bladder % % % 0 2 Other pharynx % % % 0 2 Pancreas % % % 0 2 Gallbladder % % % 0 2 Larynx % % % 0 1 Multiple myeloma % % % 0 1 Prostate % % 0 1 ASR=age-standardised rate. Table 1: Estimated new cancer cases and ASR per people per year and by sex, for all cancers combined and 27 cancer types, among year-olds worldwide in 2012 leukaemia (C91 95), and all sites combined excluding non-melanoma skin cancer (C00 97, except C44). We calculated age-standardised incidence and mortality rates (ASR) per people per year at the global and country level using the World Standard Population as proposed by Segi 11 and modified by Doll and colleagues. 12 Case fatality, a measure of the severity of a disease, was approximated by dividing the mortality ASR by the incidence ASR (M:I). To document key patterns in the cancer burden, we draw comparisons between low-income and middle-income countries (LMICs) and high-income countries (HICs). Furthermore, we present the results by geographical region, using 19 UN-defined geographical sub-regions, and development level, using the Human Development Index (HDI; low [least developed], medium, high, and very high [most developed]; appendix p 1); 13 the HDI is a socioeconomic development indicator comprising life expectancy, education, and gross national income. All analyses were undertaken using Stata 13.1 statistical software. Role of the funding source The funder had no role in the study design, collection, analysis or interpretation of the data, or writing of the report. JF had full access to all raw data used in the study. The corresponding author had full access to all clean data and the final responsibility to submit for publication. Results new cancer cases and cancer-associated deaths were estimated to have occurred among year-olds worldwide in 2012, with the disease more frequent among women (male:female ratios of 0 5 for incidence and 0 8 for mortality; tables 1 and 2). The incidence ASR was 43 3 new cancer cases per people per year and mortality ASR was 15 9 cancer-associated deaths per people per year. The annual burden was markedly higher among young adults relative to children and adolescents (0 19 year-olds) at approximately four times greater in See Online for appendix Vol 18 December

4 Both sexes Men Women New cases % ASR New cases % ASR New cases % ASR All cancers except non-melanoma skin cancer % % % % % % % % % % % % % 2 5 Brain and CNS % % % 0 8 Colorectal % % % 0 9 Stomach % % % 0 8 Non-Hodgkin lymphoma % % % 0 7 Lung % % % 0 6 Lip or oral cavity % % % 0 2 Ovary % % 0 8 Kaposi s sarcoma % % % 0 3 Oesophagus % % % 0 3 Hodgkin s lymphoma % % % 0 2 Nasopharynx % % % 0 2 Testis % % 0 3 Pancreas % % % 0 1 Other pharynx % % % 0 1 Kidney % % % 0 1 Melanoma of skin % % % 0 1 Gallbladder % % % 0 1 Bladder % % % 0 0 Corpus uteri % % 0 1 Larynx % % % 0 0 Thyroid % % % 0 1 Prostate % % 0 1 Multiple yeloma % % % 0 0 ASR=age-standardised rate. Table 2: Estimated cancer-related deaths and ASR per people per year and by sex, for all cancers combined and 27 cancer types, among year-olds worldwide in 2012 terms of new cancers and approximately three times greater in terms of cancer-associated deaths, but considerably lower than cases and deaths in middle-age (40 59 years) or in late-adulthood (60 years or older; figure 1). The spectrum of cancers observed among year-olds also differed from those diagnosed at younger and older ages. Common tumour types in children and adolescents, including leukaemia and cancers of the brain and CNS were among the most common cancers at ages years (figure 1). However, common epithelial tumours such as breast cancer (ranked first for new cases and deaths overall), cervical cancer (ranked second for new cases and fourth for deaths overall), and colorectal cancer (ranked fifth for new cases and sixth for deaths overall) were more frequently observed among young adults than children or adolescents, though still to a lesser extent than those observed in older ages. cancer or cervical cancer were the most common cancer types in young adults for most countries in terms of incidence and mortality (figure 2), and together accounted for (30 9%) of the total estimated new cases and (21 4%) of the total estimated cancerassociated deaths (tables 1 and 2). The ASR for incidence of breast cancer was 17 0 per people per year and 9 9 per people per year for cervical cancer. The ASR for mortality for breast cancer was 4 4 per people per year and 2 5 per people per year for cervical cancer. Other frequently diagnosed cancers included thyroid cancer ( [8 1%] cases; ASR 3 5 per people per year), leukaemia ( [5 1%] cases; ASR 2 2 per people per year), and colorectal cancer ( [4 2%] cases; ASR 1 8 per people per year). Notably, thyroid cancer was 4-times more common in women than in men, with an ASR of 5 7 per women per year and 1 4 per men per year. In terms of deaths, leukaemia Vol 18 December 2017

5 ( [10 1%] deaths; ASR 1 6 per people per year), liver cancer ( [10 1%] deaths; ASR 1 6 per people per year), and brain or CNS cancer ( [5 8%] deaths; ASR 0 9 per people per year) were additional large contributors to the burden. Even among young adults, the cancer profile according to 5-year age intervals was heterogeneous (figure 1). The proportion of leukaemia, lymphomas, testicular cancer, and thyroid cancer reduced with increasing age; whereas, the proportions of cervical, breast, liver, and colorectal cancer increased. A similar transition was observed when deaths were assessed, with cancers more commonly observed in children or adolescents decreasing in frequency as age at diagnosis increased, with these cancer types being replaced by cancers more common in adulthood. The global burden of young adult cancers in terms of new cases of cancer varied according to the four-level HDI (appendix pp 2 13). Although the absolute number of new cases was greatest among countries in the high- HDI level, the incidence was greatest at the very high- HDI level (ASR 64 5 new cases per people per year), followed by the low-hdi level (ASR 46 2 new cases per people per year). With respect to the cancer profile across the HDI levels (appendix pp 14 15), breast cancer was the most common cancer and cervical cancer was the second most common cancer, in the low, medium, and high-hdi levels, while at the very high- HDI level these cancers ranked first and fifth, respectively (figure 3). Thyroid cancer, melanoma of the skin, and testicular cancer were more frequent in very high-hdi regions, whereas cancers associated with infection were more frequent in countries indexed within the low-hdi level; indeed, one in three cancers (33 2%) were linked to major infectious agents, including human papillomavirus (HPV), human herpesvirus-8 (HHV-8), hepatitis B virus (HBV), hepatitis C virus (HCV), and Helicobacter pylori, in low-hdi countries relative to one in nine (11 3%) in very high-hdi settings (appendix p 16). Notably, across all HDI levels, the top five incident cancers accounted for more than 50% of the total estimated number of new cases (figure 3). Mortality was reduced with increasing HDI (appendix pp 17 28). The overall case fatality of 55 0% (M:I 25 4/46 2) and 49 1% (M:I 16 1/32 8) in low and medium-hdi countries, respectively, was approximately 41 and 35 percentage points higher than that observed in very high-hdi populations (14.3%; M:I 9 2/64 5; appendix pp 29 32). In terms of the cancer profile, breast and cervical cancer were among the top five causes of cancer-associated deaths across all HDI levels, as was leukaemia (figure 3). cancer and Kaposi s sarcoma were the remaining largest contributors to the mortality burden in low-hdi regions. At the medium-hdi level, colorectal cancer and stomach cancer were among the most fatal cancers, while liver cancer was the leading cause of cancer-associated death in high-hdi regions. Age (years) Age (years) A B New cases or cancer-related deaths (%) Kaposi s sarcoma Prostate Other pharynx Non-Hodgkin lymphoma Melanoma of skin Gallbladder Brain and CNS Oesophagus Larynx Hodgkin s lymphoma Lip or oral cavity Bladder Kidney Colorectum Ovary Multiple myeloma Thyroid Finally, for very high-hdi regions, brain and CNS cancer and colorectal cancer were the remaining largest causes of cancer death. Briefly, for cancer-specific case fatality, an increasing step-wise gradient with decreasing human development was observed for most cancer sites (appendix pp 29 32); the greatest disparities were observed for Hodgkin s lymphoma, melanoma of the skin, and cancers of the prostate, testis, thyroid, kidney, and breast all of which had an estimated case fatality at least 5-times greater in low-hdi settings compared with very high-hdi settings. With regard to geographical variations, the incidence among year-olds was greatest in Australia and New Zealand, and most countries in northern America and Europe, whereas parts of Africa, western Asia, and southern Asia had the lowest incidence (figure 4A). When assessed on a broader regional scale, the cancer profiles were diverse (appendix pp 2 13). For example, the ASR for breast cancer varied from 6 6 per people per year in Central America to 14 0 per people per year in Australia and New Zealand. Greater variations were observed for cervical cancer, with the incidence varying 15-fold; it was lowest in northern Africa and highest in southern Africa (figure 5). Other marked geographical differences included those of n n Lung Pancreas Corpus uteri Testis Stomach Nasopharynx Other Figure 1: Cancer type distribution for estimated (A) new cancer cases and (B) cancer-related deaths in Vol 18 December

6 A Most common cancer type (incidence) (104) (48) Kaposi s sarcoma (11) Thyroid (8) Testis (6) Melanoma of skin (3) (2) Colorectum (1) No data Non-Hodgkin lymphoma (1) Not applicable B Most common cancer type (mortality) (84) (30) (18) (18) Brain and nervous system (17) Kaposi s sarcoma (9) Stomach (3) Non-Hodgkin lymphoma (2) Colorectum (1) Lung (1) No data Melanoma of skin (1) Not applicable Figure 2: Global map depicting the most common cancer type by country in terms of estimated (A) new cases and (B) cancer-related deaths among year-olds in Numbers in parenthesis are the number of countries that have this cancer as their most common cancer. testicular cancer and melanoma of the skin, both of which were concentrated in northern America, Europe, and Australia and New Zealand (appendix pp 2 13). Kaposi s sarcoma was most common in eastern Africa and southern Africa, where (84 5%) of the estimated new cases occurred. Finally, the greatest burden of thyroid cancer was observed in northern America (ASR per people per year), where the incidence was 45-times higher than that observed in middle Africa (ASR 0 3 per people per year); the burden was also substantial in Australia and New Zealand (ASR 7 7 per people per year), western Europe (ASR 6 1 per people per year), western Asia (ASR 5 6 per people per year), and eastern Asia (ASR 5 4 per people per year). Despite the incidence being greatest in the most developed countries, these regions conversely had the lowest mortality, with the mortality burden greatest in parts of Africa and Asia (figure 4B). Indeed, cancer fatality was greatest in western Africa (64 3%; M:I 25 4/39 5), Vol 18 December 2017

7 with proportions greater than 50% also seen in middle Africa, eastern Africa, and southern Asia all substantially higher than the case fatality proportions in Australia and New Zealand (11 3%; M:I 8 6/76 4), western Europe (12 5%; M:I 8 6/68 9), and northern America (12 6%; M:I 9 6/76 0; appendix pp 29 32). Although not as substantial, variation in cancer mortality was also seen in HICs (figure 5). In terms of the cancer profile, the greatest mortality burden for breast cancer was observed in western Africa (ASR 6 5 per people per year), middle Africa (ASR 4 7 per people per year), and eastern Africa (ASR 4 7 per people per year; appendix pp 17 28). For cervical cancer, the mortality burden varied 31-fold, ranging from an ASR of 0 2 per people per year in northern Africa and western Asia to an ASR of 5 1 per people per year in Melanesia, Micronesia, and Polynesia. Again, the case fatality for specific cancer sites varied substantially worldwide; however, cancers with known worse survival prospects showed less heterogeneity (appendix pp 29 32). Kaposi s sarcoma Non-Hodgkin lymphoma Colorectum Lip or oral cavity Cases 54 9% 50 6% 45 1% 49 4% Low HDI cases deaths Medium HDI cases deaths Kaposi s sarcoma Colorectum Stomach Deaths 51 0% 47 6% 49 0% 52 4% Discussion To our knowledge, this is the first study to investigate cancer incidence and mortality among young adults worldwide. Overall, the global incidence of cancer in year-olds in 2012 was 43 3 per people per year, and the corresponding mortality was 15 9 per people per year. We report considerable variations in the scale of incidence and mortality worldwide, and illustrate the heterogeneity of cancer types in this age group when stratified by age, sex, development level, and geographical region. Furthermore, our study highlights the need to increase awareness and resources for this neglected subpopulation. Cancers of the breast and cervix uteri were the major contributors to the cancer burden among young adults globally, although there was wide variability in the scale and profile of cancer according to HDI level and geographical region. Broadly, we observed a greater proportion of infection-associated cancers at lower HDI levels. These findings are consistent with established epidemiological transitions and cancer transitions as countries become increasingly societally and economically developed. 14 Variations in the young adult cancer profile globally also relate to differences in screening or detection practices, genetic predispositions, and exposure to other risk factors. 2 For example, the burden of thyroid cancer is substantially greater in very high-hdi countries, with incidences highest in North America. This association with human development level probably corresponds to changes in diagnostic practice and overdiagnosis in the USA and Canada, as well as in Australia, New Zealand, Europe, and some HICs in Asia. 15 Similarly, a substantial proportion of breast cancers diagnosed before the age of 40 years are linked with genetic factors, and thus the elevated rates in North America, Australia, New Zealand, Thyroid Thyroid Melanoma of skin Testis 50 9% 58 6% 49 1% 41 4% High HDI cases deaths Very high HDI cases deaths and Europe could relate to a higher incidence of BRCA1/2 mutations in individuals of European ancestry. 16 However, breast cancer incidence in west Africa are strikingly similar to these western countries. Previous research has shown that women from west Africa develop breast cancer at younger ages, with a mean age at presentation of between 35 and 45 years, years earlier than women in HICs; 17 these cancers, which could include a disproportionate number of tumours with poor prognosis, 17 are probably associated with a mix of genetic and environmental factors unique to this subpopulation. 18 Such differences in the distribution of cancer types, with more fatal cancers generally more prominent in LMICs, was in turn responsible for the worse cancer Brain and CNS Brain and CNS Colorectum 53 8% 44 1% 46 2% 55 9% Figure 3: Proportion of five most frequent cancer types for estimated new cases and cancer-related deaths among year-olds in 2012, by HDI level HDI=Human Development Index. Vol 18 December

8 A Age-standardised incidence rate (per people per year) B > <34 6 No data Not applicable Age-standardised mortality rate (per people per year) > <9 2 No data Not applicable Figure 4: Global map depicting the estimated age-standardised (A) incidence and (B) mortality per people per year for overall cancer by country among year-olds in 2012 outcomes noted in our results among young adults from these regions. However, even within the same cancer type, mortality was higher in LMICs; this was particularly apparent for kidney, testicular, and breast cancer, leukaemia, and lymphoma, implying that these variations in case fatality are probably due to fractured health infrastructures, 2 the detection of cancers at a later stage 2 because of ineffective screening tests or no early detection procedures, 19 and poor access and availability of treatment. 20 With nearly one million cancers occurring among young adults worldwide in 2012, efforts are urgently needed to address the cancer burden in this age group. This population has, however, been neglected by cancer researchers and policy makers, often being referred to as a lost tribe. 9,21,22 Throughout the past decade, young adult oncology has attracted increasing attention in some HICs, given that cancer represents the most common cause of disease-associated death for this age group 23 and survival improvements have lagged behind those of both children and older adults. 6,21 Globally, however, much remains unknown in this area of oncology. Expanding the young adult cancer agenda beyond HICs remains a challenge, but one of great importance as young adults represent a large proportion of the population with an inherent potential for economic growth in many LMICs. An opportunity to reduce the cancer burden in young adults lies in prevention because unlike most cancers occurring at younger ages, a substantial proportion of young adult cancers are preventable. In view of the substantial contribution of cervical and liver cancer to the global young adult cancer burden, national HPV vaccination programmes of HPV-naive people and HBV vaccination programmes for neonates would probably have a sizable effect in reducing cancer incidence in young adults worldwide. Similarly, adequate treatment Vol 18 December 2017

9 East Africa Middle Africa North Africa South Africa West Africa Caribbean Central America South America North America East Asia Southeast Asia South Asia West Asia East Europe North Europe South Europe West Europe Australia/New Zealand Melanesia/Micronesia/Polynesia Age-standardised incidence (per people per year) East Africa Middle Africa North Africa South Africa West Africa Caribbean Central America South America North America East Asia Southeast Asia South Asia West Asia East Europe North Europe South Europe West Europe Australia/New Zealand Melanesia/Micronesia/Polynesia Age-standardised mortality (per people per year) Thyroid Colorectum Brain and CNS Non-Hodgkin lymphoma Testis Ovary Stomach Melanoma of skin Hodgkin s lymphoma Lip or oral cavity Lung Kaposi s sarcoma Corpus uteri Nasopharynx Kidney Oeosophagus Bladder Other pharynx Pancreas Gallbladder Larynx Multiple myeloma Prostate Other Figure 5: Estimated (A) age-standardised incidence and (B) mortality per people for each cancer type among year-olds in 2012, by geographical region for HIV could further reduce the cancer burden of Kaposi s sarcoma, particularly in sub-saharan Africa where incidence was greatest. Although lifestyle and environmental factors drive some of the burden, many young adult cancers are not strongly influenced by the major risk factors associated with the onset of certain cancers at older ages, such as tobacco, alcohol, or nutrition. To target cancers representing smaller contributors to the overall burden, additional relatively inexpensive preventive strategies of known efficacy should be considered, including those targeting melanoma of the skin (eg, limiting exposure to direct sunlight, increased sunscreen use, and avoidance of tanning salons). 9 Another relevant opportunity is ensuring timely diagnosis, with previous research suggesting that young adults experience more delays compared with children, adolescents, and older age groups. 2,21 In HICs, delays in diagnosis have been associated with psychological and social factors; whereas, in LMICs, cultural norms and geographic accessibility could represent the greatest obstacles. Affordability of and access to care also represent substantial barriers to cancer care worldwide. Although no age-specific screening tests are currently available for young adults, 24 our data indicate that screening and early detection programmes might have a significant effect at a limited cost particularly those targeting cervical cancer (eg, visual inspection with acetic acid screening) and breast cancer (eg, self-awareness through self-examination and enhanced screening for BRCA1/2 carriers), with lesser but still potentially worthwhile effects for testicular cancer, colorectal cancer, Vol 18 December

10 For Cancer Incidence in Five continents see and melanoma of the skin in high-risk populations. 9,24 Research into such programmes specifically in young adults is needed, to determine the benefits and potential risks, particularly for breast and testicular self-examination because these recommendations 24 are not currently supported by evidence. In general, increasing awareness of cancer in this age group at both public and professional levels is needed to bring attention to the aforementioned primary and secondary prevention measures, reduce delays, and lead to improvements in the cancer burden across all resource levels. Finally, once diagnosed, young adults with cancer must be able to access high quality care. Currently, care for these individuals is scattered; even in HICs, young adult oncology expertise is rare, with 90% of North American young adults with cancer receiving their care in community practices. 25,26 In many cases, which treatment protocols offer this population the best chance of cure is still unclear, with extrapolations made from younger and older populations. This uncertainty is largely due to the low participation in the relatively few clinical trials available for this age group; 2,21 thus, increasing the availability of and participation in clinical trials for this group is of paramount importance. 9 Even when effective treatment protocols for young adults are known, further work is necessary to ensure that all patients can access such treatment. For example, despite evidence proving the superiority of paediatric treatment protocols for young adults with acute lymphoblastic leukaemia (ALL), a study 25 found that the percentage of Californian young adults with ALL treated with such protocols reduced from 31% in to 21% in Access to appropriate treatment is even more crucial in countries under transition. Although freely available guidelines for young adult oncology, such as that produced by the National Comprehensive Cancer Network, 24 might be of help, it remains uncertain whether such guidelines are appropriate for more resource-limited settings. Finally, our data show that even within HICs cancer mortality varies substantially. Several HICs have built alternative models of young adult cancer care delivery, such as a nationwide network of young adult cancer units in the UK; 27 whether these different models improve cancer outcomes is unknown. Implementation of the above recommendations will require substantial investment in young adult oncology. Although such investments in LMICs might seem too complex or costly, the economic advantages are clear given the number of potential years of life saved; this is exemplified by the fact that this population has approximately years of their life expectancy remaining (average global life expectancy at birth for individuals aged years in 2012 was 63 3 years 28 ), which is substantially greater than that observed in older adults who have not only a shorter remaining life expectancy but also poorer survival prospects after cancer diagnosis. Because individuals in this age group are the most financially productive members of their societies, it is clear that improving equity in prevention, diagnosis, and care will not only have large societal effects, but might also be cost-effective. 21 Studies undertaking cost benefit analyses of improving young adult cancer care, particularly in LMICs, are needed to inform policy makers and advocates. Our results provide a comprehensive estimation of young adult cancers worldwide in 2012 but have several limitations. The GLOBOCAN cancer groupings do not directly correspond with the recommended classification scheme for tumours diagnosed in young adults by Birch and colleagues. 29 Furthermore, our results are estimates from the GLOBOCAN database, which were compiled using a hierarchy of methods; thus, accuracy was dependent on the quality and availability of the source information at a given time. 1 Clearly, the inclusion of uncertainty intervals for the estimates provided in this study would be useful, and collaborative work is currently underway to create such intervals that take into account deficiencies in the quality and availability of the source information at the national level, as well as inherent statistical uncertainty. To address both of these limitations, future research should be undertaken using available population-based cancer registry data through Cancer Incidence in Five Continents, which would allow an assessment of histologically defined cancers, including those that we were not able to address in this paper (eg, bone cancer and soft tissue sarcomas), and trends, albeit for a limited number of countries or subnational regions. In summary, the global cancer burden among year-olds differs from that in younger or older ages, but also varies substantially by age, sex, development level, and geographical region. Although cancer is less frequently observed in young adults than at older ages, its effects remain considerable because these individuals have a large proportion of their expected lifespans remaining, contribute substantially to the economy, and play a major part in caring for their families. Because young adult cancer patients exhibit a combination of features observed in younger and older patients, progress needs to be achieved through a combination of the methods that led to improvements in these other groups: advancement of risk stratification and treatment protocols through clinical trials in children and implementation of effective prevention and early detection at older ages. Contributors MMF conceived and designed the study. JF acquired the data. MMF analysed the data. MMF, SG, IS, JF, ES-F, and FB interpreted the data. MMF drafted the manuscript. MMF, SG, IS, JF, ES-F, and FB critically reviewed the manuscript. Declaration of interests We declare no competing interests. Acknowledgments We gratefully acknowledge the many population-based cancer registries worldwide and their staff for their willingness to contribute their data, from which GLOBOCAN 2012 is built. The work completed in this Vol 18 December 2017

11 manuscript was undertaken while MMF was a postdoctoral fellow at the International Agency for Research on Cancer, which was partially funded by the European Commission s Seventh Framework Programme Marie Curie Actions People COFUND. References 1 Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN Int J Cancer 2014; 136: E Magrath I, Epelman S. Cancer in adolescents and young adults in countries with limited resources. Curr Oncol Rep 2013; 15: No authors listed. What should the age range be for AYA oncology? J Adolesc Young Adult Oncol 2011; 1: Gatta G, Zigon G, Capocaccia R, et al. Survival of European children and young adults with cancer diagnosed Eur J Cancer 2009; 45: CanTeen, Cancer Australia. National service delivery framework for adolescents and young adults with cancer canceraustralia.gov.au/sites/default/files/publications/national_ service_delivery_framework_for_adolescents_and_young_adults_ with_cancer_teen_52f301c25de9b.pdf (accessed April 12, 2017). 6 Bleyer A, O Leary M, Barr R, Ries L. Cancer Epidemiology in Older Adolescents and Young Adults 15 to 29 Years of Age including SEER Incidence and Survival: gov/archive/publications/aya/aya_mono_complete.pdf (accessed March 6, 2017). 7 De P, Ellison LF, Barr RD, et al. Canadian adolescents and young adults with cancer: opportunity to improve coordination and level of care. CMAJ 2011; 183: E Kero AE, Järvelä LS, Arola M, et al. Late mortality among 5-year survivors of early onset cancer: a population-based register study. Int J Cancer 2015; 136: Adolescent and Young Adult Oncology Progress Review Group. Closing the gap: research and care imperatives for adolescents and young adults with cancer research/ayao-august-2006.pdf (accessed April 12, 2017). 10 WHO. International Statistical Classification of Diseases and Related Health Problems (ICD) 10th Revision 2010 Edition (accessed April 12, 2017). 11 Segi M. Cancer mortality for selected sites in 24 countries ( ). Sendai, Japan: Department of Public Health, Tohoku University of Medicine, Doll R, Payne P, Waterhouse J, (eds). Cancer incidence in five continents: a technical report. Springer; Berlin; UN Development Programme. Human Development Report development_report_1.pdf (accessed April 12, 2017). 14 Fidler MM, Soerjomataram I, Bray F. A global view on cancer incidence and national levels of the human development index. Int J Cancer 2016; 139: Vaccarella S, Franceschi S, Bray F, Wild CP, Plummer M, Dal Maso L. Worldwide thyroid-cancer epidemic? The increasing impact of overdiagnosis. N Engl J Med 2016; 375: Janavičius R. Founder BRCA1/2 mutations in Europe: implications for hereditary breast-ovarian cancer prevention and control. EPMA J 2010; 1: Vanderpuye V, Grover S, Hammad N, et al. An update on the management of breast cancer in Africa. Infect Agent Cancer 2017; 12: Brinton LA, Sherman ME, Carreon JD, Anderson WF. Recent trends in breast cancer among younger women in the United States. J Natl Cancer Inst 2008; 100: Sankaranayananan R, Black R, Swaminathan R, Parkin D. An overview of cancer survival in developing countries. Cancer survival in developing countries. Lyon: International Agency for Research on Cancer, V1.pdf (accessed July 20, 2017). 20 Gelband H, Sankaranarayanan R, Gauvreau CL, et al. Costs, affordability, and feasibility of an essential package of cancer control interventions in low-income and middle-income countries: key messages from Disease Control Priorities, 3rd edn. Lancet 2016; 387: Bleyer A. Young adult oncology: the patients and their survival challenges. CA Cancer J Clin 2007; 57: Stevens MCG, Elfenbein DS, Boxer DA, et al. The Lost Tribe and the need for a promised land: The challenge of cancer in teenagers and young adults. Eur J Cancer 2006; 42: Bleyer WA, Barr RD. Cancer in adolescents and young adults. Berlin, Germany: Springer-Verlag, Coccia PF, Altman J, Bhatia S, et al. Adolescent and young adult oncology. Clinical practice guidelines in oncology. J Natl Compr Canc Netw 2012; 10: Muffly L, Lichtensztajn D, Shiraz P, et al. Adoption of pediatric-inspired acute lymphoblastic leukemia regimens by adult oncologists treating adolescents and young adults: A population-based study. Cancer 2017; 123: Albritton KH, Wiggins CH, Nelson HE, Weeks JC. Site of oncologic specialty care for older adolescents in Utah. J Clin Oncol 2007; 25: Carr R, Whiteson M, Edwards M, Morgan S. Young adult cancer services in the UK: the journey to a national network. Clin Med 2013; 13: The World Bank. Life expectancy at birth, total (years) (accessed June 21, 2017). 29 Birch JM, Alston RD, Kelsey AM, Quinn MJ, Babb P, McNally RJQ. Classification and incidence of cancers in adolescents and young adults in England Br J Cancer 2002; 87: Vol 18 December