INTRODUCTION

Colombia, in 2012, presented the 15th highest cancer age-standardized mortality rate (85.0 deaths per 100,000 people) among the 20 Latin American countries. However, in relation to leukemia mortality, this country presented the sixth highest rate (4.1/100,000), similar to Uruguay (4.0/100,000), and overcoming countries like Brazil (3.5/100,000), Argentina (3.5/100,000), and Chile (3.1/100,000)¹. This scenario has remained, according to the most recent estimates. In 2018, Colombia presented the 16th highest age-standardized mortality rate for cancer (79.2/100,000) and the sixth highest mortality rate for leukemia (4.1/100,000)².

According to cancer mortality estimates from the Cancerology National Institute of Colombia, during 2007-2011, the cancer locations responsible for the highest number of cancer deaths among men were stomach, prostate, lung, colorectal and anus, and due to leukemia. In women, the main locations were breast, cervix, stomach, lung, and colorectal and anus. For children (0-14 years old), the highest number of cancer deaths were due to leukemia. There were 869 (4.2/100,000) leukemia deaths per year in men, 741 (3.3/100,000) in women, and 256 in children (2.2/100,000 boys and 1.7/100,000 girls)³.

In developed countries, the leukemia age-adjusted mortality rate trend showed a steady declining pattern since the 1970s. In the European Union, Malvezzi et al.⁴ observed a continuing fall in mortality rates from 1970 to 2012, for both sexes (Average Annual Percent Change - AAPC = -0.7 for men and AAPC = -1.0 for women, p < 0.05). In the USA, Chatenoud et al.⁵ found a steady decrease in leukemia mortality since the early 1980s. However, this continuing fall was not observed in the developing countries of Latin America, which presented limited declines in Argentina, Chile, Costa Rica, Cuba, Puerto Rico, and Venezuela; and increases in Brazil and Ecuador. In Colombia, mortality rate was stable for both sexes from 1984 to 2009 (AAPC = -0.2 for men and AAPC = 0.2 for women, p > 0.05). Moreover, even for developed countries, the mortality rates declining differ according to age groups, being more consistent at a young age⁶. Curado et al.⁷, when analyzing the leukemia mortality trend in children, adolescents, and young adults, found stability for Colombia between 1984 and 2004 for the age groups 0-14 and 15-24 years, in both sexes. However, there are no published studies on the leukemia mortality trend in Colombia and in the country's capital, Bogotá, according to age groups and leukemia type.

Therefore, the aim of this study was to analyze the leukemia mortality trend in children and adults living in Colombia and Bogotá, between 1985 and 2012, by sex, age, and leukemia type, identifying the similarities and differences between the AAPC in these two locations.

METHODS

RESULTS

In Colombia, between 1985 and 2012, there was a predominance of leukemia deaths in males (19,545 deaths, Age Standardized Rate - ASR 3.93/100,000) in relation to females (16,951 deaths, ASR 3.19/100,000). In only 235 (0.64%) of these death records, age was not discriminated. In Bogotá, for this same period, age was not discriminated in 0.07% (seven) of the records and mortality rates were 7.54/100,000 men (5,265 deaths) and 5.59/100,000 women (4,625 deaths).

Both in Colombia (Table 1) and in Bogotá (Table 2), the highest leukemia mortality rates, in all periods analyzed, occurred in the population ≥ 70 years old and this pattern repeats itself for lymphoid and myeloid leukemia.

The mortality rates for leukemia presented a decrease pattern, during the study period, between 0-14 years old in Colombia (AAPC = -0.9) and in Bogotá (AAPC = -2.0). In Bogotá, when both sexes were considered, this decrease pattern repeats itself for other age groups, except for the ≥ 70 years old range, which presented a mean increase in mortality rates by 0.9% per year, and this increase was higher in women (AAPC = 1.5). A statistically significant increase in leukemia mortality rates was also observed in Colombia in the age range of 60-69 (AAPC = 1.1), ≥ 70 years (AAPC = 2.0), and all ages (AAPC = 0.3) (Table 3).

For lymphoid leukemia, mortality rate trends behave differently in Colombia and Bogotá. In Colombia, there was a tendency of an increase in mortality rates for all ages (AAPC = 1.0) and in different age groups, except for ages 0-14 years (AAPC = 0.2). In Bogotá, there is a decrease by -0.8% per year in mortality rate and this behavior occurs mainly in the age groups 0-14 years (AAPC = -1.4) and 60-69 years (AAPC = -1.3) (Table 3).

For myeloid leukemia in Colombia, the AAPC shows a statistically significant increase in mortality rates across all age groups. However, this trend is only constant for the age groups 45-59 years (AAPC = 1.4) and ≥ 70 years (AAPC = 4.2) (Table 3). For the other age groups, the general trend is divided into different periods, since the final model has one or two joinpoints. Thus, it can be observed that between 1985 and 2000 there was a statistically significant upward trend, but in the period of 2000 to 2012 there was stability in mortality rates (Table 4).

In Bogotá, myeloid leukemia trends were also not constant for age groups 0-14, 15-44, and 60-69 years. There was a first period of a statistically significant increase and then a period of stability for the ages 0-14 years (APC = -7.7; between 2003-2012) or a period of decrease for the ages 15-44 years (APC = -4.0; 95% CI -6.3, -1.7; between 1999-2012) and 60-69 years (APC = -4.1; 95% CI -7.6, -0.4; between 2000-2012) (Table 4). However, in Bogotá, for seniors ≥ 70 years old, there was an increase in myeloid leukemia mortality rates, throughout all the studied period (AAPC = 3.0; 95% CI 1.8, 4.2) (Table 3).

DISCUSSION

Leukemia is a neoplasm characterized by imbalance in the white blood cells development, produced by genetic factors and chemical or environmental exposures^12,13. The existing types of leukemia affect the age groups differently, being the findings of this article consistent with the epidemiological profile reported in the literature. For children less than 15 years old, acute lymphoblastic leukemia (ALL) represents 78% of cases, while the myeloid leukemias corresponds to 17% of cases (16% for acute myeloid leukemia and 1% for chronic myeloid leukemia). From the second decade of life onwards, the proportion of myeloid leukemias increases, with a corresponding percentage decrease in acute lymphoblastic leukemia¹⁴. Thus, the analysis of the temporal evolution of leukemia mortality rates is important not only for their magnitude but also for the complexity of the determinants.

The study of leukemia mortality profile is strongly influenced by the diagnostic capacity, the mechanisms used to consolidate information, and the population's access to specialized medical centers. Therefore, the analysis of mortality patterns measures the effort undertaken to control the disease, as well as the possible inequalities in the health care of the studied localities^15,16. In Colombia, in 2012, the mortality rates for leukemia found in men (3.81/100,000) were lower than the estimate for the European Union in 2012 (4.3/100,000). However, the rate for men in Bogotá (7.04/100,000) exceeded this estimate, just as the rates for women in Colombia (3.28/100,000) and in Bogotá (4.98/100,000) were higher than the estimate for women in European Union (2,59/100,000)¹⁷, suggesting that, despite advances, there is still a need for improvements in the treatment of this neoplasm for Colombians.

According to the findings of the present study (1985-2012), in the 0-14 year age group, mortality from all leukemias revealed a statistically significant declining trend (APC = -0.9; 95% CI -1.2, -0.6), confirming the findings of the previous study by Piñeros et al.¹⁸, in which an average annual percentage change of -1%, for the same age group, was found in an analysis of child cancer mortality in Colombia, from 1985 to 2008. The decrease evidenced in the 1990's coincide, in the political sphere, with the implementation of Law 100/1993, which regulated the General System of Social Security in Health, establishing health care for the entire Colombian population in areas of prevention, diagnosis, and treatment¹⁹. It is also important to highlight other more recent resolutions, aimed at improving child cancer survival. The Law 1388/2010 proposes to significantly reduce the death rate from cancer in children and people under 18 years old, through the guarantee of all the services required for early detection and comprehensive treatment²⁰. In addition, the Public Health Surveillance Protocol, published in 2008, proposes a sentinel surveillance strategy to accelerate pediatric acute leukemia diagnosis and treatment initiation²¹. However, the decrease revealed in Colombia was lower and later in comparison to that observed, in the same age group, in developed countries like Japan (APC = -4.77, 1976-2006, for boys; and APC = -4.53, 1970-2006, for girls); United States (APC = -4.95, 1970-1984, and APC = -3.39, 1984-2005, for boys; and APC = -6.09, 1970-1980, and APC = -3.14, 1980-2005, for girls)²², and European Union countries between 1970-2006 (Denmark: APC = -4.91 and -4.16; Italy: APC = -3.67 and -4.33; United Kingdom: APC = -3.83 and -3.82; Portugal: APC = -3.67 and -2.53, for boys and girls, respectively)²³. However, it could coincide with the limited reduction observed in some Latin American countries⁷.

The analysis, by type of leukemia, in the 0-14 years old group shows, however, a statistically significant increase for myeloid leukemia (AAPC = 2.4; 95% CI 0.8, 3.9), mainly in males (AAPC = 2.7; 95% CI 0.4, 5.2). For lymphoid leukemia, there was a stable behavior (AAPC = 0.2; 95% CI -0.1, 0.4). In Bogotá, there was a stable behavior for myeloid leukemia (AAPC -0.2, 95% CI -4.0, 3.7) and a decrease for lymphoid leukemia (AAPC = -1.4; 95% CI -2.2, -0.6). Less favorable results for myeloid leukemia compared to lymphoid leukemia were expected given the higher probability of survival in pediatric patients with lymphoid leukemia^24,25,26, that varies from 50% to more than 90% in Puerto Rico, Canada, USA, and many European countries. In Colombia, the estimated probability of survival for children with acute lymphoblastic leukemia is 68.9%²⁷. The higher rate reductions in Bogotá, compared to Colombia, make it possible to highlight the differences in access to health services and, therefore, the existence of several situations that may interfere in the flow of events related to the health service use. Among these situations, the individual's perception of personal or family health needs stand out, giving rise to the desire to obtain care, which could imply the decision to search outpatient or hospital services to obtain care²⁸. One possible measure to minimize these differences is to consolidate the role of health professionals in the primary care level in identifying warning signs at all stages of life and knowledge about the health impact of exposure to harmful environmental and residential agents, to achieve social mobilization in search of healthy environments and promote self-care actions in the use of household chemicals and pesticides.

The decrease pattern for lymphoid leukemia observed in the 0-14 years old group, in Bogotá, was not observed in the 15-44 (APC = -0.4; 95% CI -1.0, 0.3) and 45-59 years old groups (APC = -0.6; 95% CI -1.7, 0.5). It is known that the prognosis of lymphoid leukemia is less favorable for adolescents and adults compared to children¹⁴. However, even for these age groups some improvements have been verified in developed countries, with a better relapse control using allogenic hematopoietic stem cell transplantation, monitoring of minimal residual diseases and improving the management of treatment complications¹⁷.

The ≥ 70 years old group showed the highest mortality rates, in Colombia and in Bogotá, for all types and for lymphoid and myeloid leukemias. Moreover, there were increases in mortality rates for this age group, for men, women, and both sexes, in all analyses, except for women and both sexes for lymphoid leukemia in Bogotá (AAPC = -0.2; 95% CI -1.4, 1.1). The 60-69 years old group had also an increase in mortality rates for all types (AAPC = 1.1; 95% CI 0.3, 2.0) and for lymphoid (AAPC = 1.2; 95% CI 0.6, 1.9) and myeloid (AAPC = 3.3; 95% CI 2.1, 4.4) leukemias in Colombia. Studies regarding the evolution of mortality rates in the elderly are much more scarce compared to studies in other age groups, especially in children. However, it is possible to notice that, even in developed countries, older people have worse outcomes than younger people. In the United States, where there has been a decline in rates among the youngest, since 1960, there has been a notable increase in rates from 65 years old at least up to the year 2000²⁹. Bertuccio et al.¹⁷, when comparing leukemia mortality rates for the European Union, in 1977 and 2007, showed a decrease in all age groups, except for the ≥ 70 years range, which showed a percentage increase of 5.58 and 1.69, for men and women, respectively.

The worst results for myeloid leukemias in the present analysis by leukemia type are in line with the scientific literature. Data from 18 high quality population-based cancer registries, available on the National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) Program, for cases diagnosed from 2007 to 2013, demonstrate that survival in older adults ages (75-84 years old) is poor, particularly for acute myeloid leukemia (relative survival = 2.7%,; 95% CI 2.0-3.6)³⁰. The structural and functional changes of the aging process lead to a decrease in the vital capacity of the organs, making the induction therapy less tolerated, with a higher risk of infection and with an increase in treatment-related mortality^31,32.

The results of this study present some limitations to be highlighted. Firstly, the use of data from the DANE website brings the inherent problems of secondary data, such as inability to complete missing information or correct any data inconsistencies. In addition, the fact that, during the period considered in the study, two different ICD revisions were used, and this divergence was addressed using a code equivalence guide³³. Nevertheless, the comparison of the trend in the Colombian mortality rates in relation to those observed in the capital shows inequities in health and access conditions, representing a decrease in the effectiveness of the treatment among the Colombian population. Thus, the findings suggest the need for studies that address the distribution and hierarchization of the health network in the country, identifying the need for a greater supply of quality specialized services according to the population's health demands, respecting the balance between the demand and the provision of services. The location of specialized health entities in oncology should agree with the cancer epidemiological behavior. When financial, administrative, and security conditions do not allow the establishment of these entities, it is necessary to provide the creation and maintenance of parents and patients organizations, so that they support families and users during the treatment period, avoiding the inequality referenced in some national studies^34,35.

We acknowledge that the proportion of unspecified leukemia in the death certificates can change over the time, putting uncertainty in the interpretation of the mortality rates trends. However, according to the World Health Organization, the Colombian mortality statistics were considered of medium quality, it means that completeness vary between 70% and 90%, and the proportion of deaths certified as signs, symptoms, and ill-defined conditions was between 10% and 20%³⁶. Nevertheless, there is evidence indicating that 92.8% of the cancer deaths were correctly assigned by the mortality system⁹. In addition to that, according to Bravo et al.³⁷, the proportion in unspecified leukemia in children decreased over the time, achieving less than 1% in the period 2007-2011, likely reflecting improvement in the quality of ill-defined cause of deaths in the mortality system.

On the other hand, in the last two decades, the Colombian health system has presented substantial progress in improving equity in the access to health care facilities, moving toward the complete implementation for the universal health coverage³⁸. The public health system provides programs of primary and secondary cancer prevention, early detection, and treatment and palliative care. However, the main weakness and challenges consists to delivery effective high quality oncology care for all population³⁹.

The findings of this article are very important to understand the behavior of leukemia in the country and in the capital. Moreover, the results can be considered as a reference point to identify the impact of the implementation of Law 1388/2010, which regulated state actions, to ensure the right to life for children with cancer throughout the country²⁰, and to evaluate the epidemiological behavior of leukemias in other age groups.

CONCLUSION

In the 0-14 years old group, a declining trend in mortality rates for all types of leukemia was identified in Colombia, most pronounced in the capital. There was a decrease in historical series at ages 15 and 69 years in Bogotá, and a positive tendency for all types of leukemia deaths in age of ≥ 70 years old in the two studied areas. The analysis, by type of leukemia, also shows a decreasing trend for lymphoid leukemia in boys and girls from 0-14 years old in Bogotá. Therefore, the epidemiological mortality pattern is consistent with that described in the international literature. The results found in Bogotá are possibly due to the availability of specialized health centers in pediatric oncology and access to medical treatment.