Estimated global cancer incidence in the oldest adults in 2018 and projections to 2050

Abstract Using GLOBOCAN estimates, we describe the estimated cancer incidence among adults aged 80 years or older at the regional and global level in 2018, reporting the number of new cancer cases, and the truncated age‐standardised incidence rates (per 100 000) for all cancer sites combined for this age group. We also presented the five most frequent cancers diagnosed by region and globally among females and males aged 65 to 79 years old and 80 years or older. We, finally, estimated the number of new cancer cases in 2050, the proportion of cases aged 80 years or older, and the proportional increase between 2018 and 2050 by region, by applying population projections to the 2018 incidence rates. In 2018, an estimated 2.3 million new cancer cases (excluding nonmelanoma skin cancers) were aged 80 years or older worldwide (13% of all cancer cases), with large variation in the profiles at regional levels. Globally, breast, lung and colon were the most common cancer sites diagnosed in the oldest females, while prostate, lung and colon were most frequent in the oldest males. In 2050, an estimated 6.9 million new cancers will be diagnosed in adults aged 80 years or older worldwide (20.5% of all cancer cases). Due to the complexity of cancer management in the oldest patients, the expected increase will challenge healthcare systems worldwide, posing a tangible economic and social impact on families and society. It is time to consider the oldest population in cancer control policies.

lowest cancer-specific survival relative to other age groups, [8][9][10] and the survival gap is widening, partially because the oldest patients do not benefit as much from advances in cancer treatment as younger patients. 8,9 In view of an unprecedented rising number of future patients in this age group and the challenges of their cancer management, a comprehensive description of the cancer burden in the oldest population is warranted. We thus provide a detailed profile of the current and future cancer estimated burden in the oldest adults worldwide and by world region, as a call for the design of dedicated and tailored cancer control programs for populations aged 80 years and above.

| METHODS
While the definition of an oldest-old person may vary according to the life expectancy of a given country, the United Nations' World

What's new?
The global population of aging persons is expected to increase significantly over the coming decades. To prepare for the forthcoming challenges in cancer care, more information is needed on cancer burden among the world's oldest patients. Here, assessment of global cancer burden in adults in 2018 shows that 2.3 million new cases occurred in persons aged 80 or older that year, accounting for 13 percent of all cases worldwide. Projections suggest that by 2050, 7 million new cancer cases will affect this age group.
The findings predict substantial challenges for healthcare and highlight the need for aging-specific cancer control initiatives.
T A B L E 1 Estimated number of new cancer cases in adults aged 80 or older, percentage of total cases (all age combined), percentage of the total population aged 80 years or older and truncated age-standardised incidence rates (TASRs), 2018 Regions  Truncated age-standardised incidence rates also varied greatly across regions, from 967 per 100 000 adults aged 80 years or older in Africa to 2557 in Oceania, and within regions (Table 1 and Figure 1).
As examples, regional TASRs ranged from 607 in South-Central Asia to 2351 in Eastern Asia (excluding China), and nationally from 275 in The Gambia to 3615 in Singapore (Table S1).

| Cancer profile in the oldest adults
Among females aged 80 years or older, breast, lung and colon cancers were the most common cancer sites globally and in most regions  Figure S1). The five most common cancers represented 52% of the total number of new cancer cases that occurred in the oldest females in 2018 worldwide, ranging from 45% in Africa to 60% in China.
Among males aged 80 years or older, lung and prostate cancers were the leading cancer sites at the global level ( Figure 3). Colon cancer was frequent in almost all regions. Stomach cancer was frequent in Asia and Latin America and the Caribbean and ranked fourth at the global level. Contrary to other regions, liver cancer was common in Africa (second) and Asia (fifth). The cancer profile in the 80 years or older age group is also similar to that observed in the 65 to 79 age group ( Figure S2). Globally, the five most common cancers represented 59% of all cancers diagnosed in the oldest males, varying from 58% in Northern America and Oceania to 64% in China. bean (+253%), and Africa (+228%), while the lowest increase will be seen in Europe (+87%, Figure 4).  15 Because the oldest adults are heterogeneous in terms of health status, and fitness, chronologic age alone is often a poor indicator of an individual's physiological or functional status, and should not be used as the sole criterion for treatment decision-making. 16,17 Cancer survival is lower in this age group compared to other age groups [8][9][10]18 ; which is mainly explained by excess mortality in the first months after diagnosis, 9,18 possibly due to suboptimal treatment, higher postoperative mortality, inappropriate risk stratification. 16,19,20 Because some older patients may benefit from surgery or chemotherapy, [21][22][23] individualised cancer management (including, but not limited to, modification of treatment schedules and/or dosing and implementation of geriatric-specific supportive measures) is, therefore, essential in the oldest adults. The utilisation of comprehensive geriatric assessments is crucial to identifying patients that will optimally benefit from treatment, and in countering other deficits that could lead to improved treatment tolerance. 16,24 Less developed countries face specific challenges, given a substantial increase in the number of new cancer diagnoses among their oldest populations. In addition to a lack of specialised infrastructure, oncologists, pathologists and surgeons, and lack of availability of, or accessibility to radiotherapy and chemotherapy, 25,26 there is also a generalised lack of cancer care personnel with geriatric expertise, and geriatric oncology is vastly underdeveloped. 27 Because of competing health demands, the oldest adults may not currently represent a priority in cancer plans and programs in resource-limited countries. However, due to the predicted rise in the number of patients diagnosed with cancer, it is clear that these countries should consider the specific needs of the oldest adults when developing and implementing cancer control programs.

| DISCUSSION
Supportive and palliative care, including pain management, are crucial to relieve unnecessary pain and suffering for patients and their families, and are, therefore, essential components of cancer care management. Investment in palliative care is cost-effective for the healthcare system and society regardless of the level of country development. 28 Currently, supportive/palliative care and pain medications are not available in many of the least developed countries, 29 and even where they are, older age is a barrier to their implementation. 30 Although pain medications, such as opioids, can be delivered to the majority of patients regardless of chronological age, the oldest patients with cancer warrant careful clinical consideration when managing their pain because of age-related physiologic changes, immunosuppression, polypharmacy, comorbidity and frailty. 30 Though international organisations have issued guidelines for palliative care and pain management, 31 specific considerations regarding older adults still need to be addressed, and this represents a relevant gap for future research.
Alongside the challenges for healthcare systems, the projected rise in the number of patients with cancer aged 80 years or older will have a major social and economic impact on families and society. The oldest adults with cancer often experience a decline in functional status 32,33 ; therefore, they need support, usually from within the family, to take care of them to undertake daily activities. The caregiver role is recognised as central in cancer care management. 34 Yet caregiver burden defined as "the strain or load borne by a person who cares for a chronically ill, disabled, or elderly family member", 35 is common in persons who care for older patients, 36,37 and is associated with higher psychosocial and physical morbidity. 38 46 However, the proportion of prostate cancer cases among men aged 80 years or older which are a result of overdiagnosis has been estimated to be of only 6% (compared to 32% in those aged 70-74 years). 47 In addition, recent data suggests that global prostate cancer incidence has decreased, which might reflect declines PSA screening. 48 Therefore, we believe that the effect of overdiagnosis on our results may not be as significant, and this may apply not only to prostate cancer, but also for other screen-detected tumours such as colon and breast cancer. Irrespectively, increasing access to breast, colorectal and cervical cancers screening for the current targeted age group (ie, up to 74-years-old for colorectal cancer in some countries) might also benefit the oldest cohorts in a long-term perspective.
Our study has limitations. In countries with no incidence data, GLOBOCAN estimates were computed from cancer specific mortality data that may be less accurate for older age groups; where vital statistics system exists, mortality data rely on the accuracy of the cause of death reported on the death certificate. Identifying the underlying cause of death may be challenging in older patients who may present with fatal comorbidities. In many low resource countries, verbal autopsy is the only method currently available to obtain estimates of the distribution of causes of death but the method may be less reliable in older age groups. 49 Furthermore, there is a higher probability of under-ascertainment of cancer cases at older ages because of comorbidities and frailty, as well as limited histological verification of cancer diagnoses among the oldest-old patients. 4 The actual incidence may then be higher than our estimation. Besides, our projections neither took into account historical trends in cancer incidence or potential changes in risk in younger cohorts, nor preventive actions taken to tackle the cancer burden in each region. Cancer-specific projections for 2050 and the impact of various cancer prevention interventions on 2050 projections were beyond the scope of this report. However, our estimates for all cancer sites combined at global and regional levels in 2050 are probably underestimated. Indeed, countries in transition are seeing the incidence of cancers common in the oldest population, naming breast, colorectal and prostate cancers, rising. In the same time, the decrease of infection-related cancers, notably in gastric and cervical cancers, are decreasing in high-resourced countries. 50 Finally, the definition of oldest-old might vary according to the life expectancy of populations, and it is possible that setting the cut-off at 80 years might be too high for less developed regions of the world.

| CONCLUSION
The substantial rise in cancer cases among the oldest adults in the next decades represents a considerable challenge for healthcare systems across all world regions. Our study highlights the need for prioritising cancer prevention over the entire lifespan and the inclusion of the oldest adults in cancer control programs through the creation of age-friendly healthcare systems and of multidisciplinary teams with geriatric expertise and training which can provide high-quality care for this growing population of patients with cancer.