Cancer Incidence in Egypt: Results of the National Population-Based Cancer Registry Program

Background. This paper aims to present cancer incidence rates at national and regional level of Egypt, based upon results of National Cancer Registry Program (NCRP). Methods. NCRP stratified Egypt into 3 geographical strata: lower, middle, and upper. One governorate represented each region. Abstractors collected data from medical records of cancer centers, national tertiary care institutions, Health Insurance Organization, Government-Subsidized Treatment Program, and death records. Data entry was online. Incidence rates were calculated at a regional and a national level. Future projection up to 2050 was also calculated. Results. Age-standardized incidence rates per 100,000 were 166.6 (both sexes), 175.9 (males), and 157.0 (females). Commonest sites were liver (23.8%), breast (15.4%), and bladder (6.9%) (both sexes): liver (33.6%) and bladder (10.7%) among men, and breast (32.0%) and liver (13.5%) among women. By 2050, a 3-fold increase in incident cancer relative to 2013 was estimated. Conclusion. These data are the only available cancer rates at national and regional levels of Egypt. The pattern of cancer indicated the increased burden of liver cancer. Breast cancer occupied the second rank. Study of rates of individual sites of cancer might help in giving clues for preventive programs.


Introduction
Egypt was completely lacking incidence rates at national level until the results given in the current report were obtained. Available statistics were proportions derived from single or multicenter hospital registries that could not be used for calculation of incidence rates [1][2][3][4][5][6][7]. The only published incidence rates are those from a cancer registry in one district in Nile delta (Gharbiah governorate). The last internal reports of this registry are for 2002 [8,9]. Incidence rates up to 2007 were published in Volumes IX and X of Cancer Incidence in Five Continents [10,11], date of end of registry activities due to failure of sustainability. The published crude and age-standardized incidence rates from that registry are 96.5 and 132.6/100,000 males and 97.3 and 122.1/100,000 females. The commonest sites of cancer in males are liver (18.7%), bladder (12.7%), non-Hodgkin's lymphoma (11.0%) and trachea, bronchus, and lung (8.2%). The 4 sites represent 50.6% of all cancer in males. The commonest sites in females are breast (38.8%), non-Hodgkin's lymphoma (8.5%), liver (4.6%), and ovary (4.5%); all together represent 56.4% of cancer in females. There is no mention of rates of both sexes together. Studies are published using these Gharbiah data up to 2007 and are limited to specific sites of cancer mainly breast [12][13][14][15], gastrointestinal [16][17][18], hematopoietic [19], bladder [20], and gynecological cancers [21]. None of these geographically-limited studies and published rates could be considered as representative of Egypt, being based on results of one registry in a single delta governorate and do not have an impact on understanding the current situation of cancer at the national level.
The National Cancer Registry Program (NCRP) was established in 2008 and became the only source for cancer incidence in the country [22]. The main objective of the current publication is to present the incidence rates of cancer in Egypt in 2008-2011 based upon data of the National Cancer Registry Program of Egypt with estimated incidence of the disease up to 2050.

Materials and Methods
For registration purposes, Egypt was stratified into 3 geographical strata (regions), namely, Lower Egypt (north of Cairo), Middle Egypt (south of Cairo), and Upper Egypt (further south, reaching the southern frontier of the country). The current report covered three districts (governorate), each representing one of the 3 regions, namely, Damietta (Nile delta), Minya (Middle Egypt), and Aswan (Upper Egypt) ( Figure 1). A population-based registry was established in each of the 3 governorates, located in the Ministry of Health Cancer Center.
Trained medical doctors in the 3 registries abstracted records from their cancer centers and regularly visited establishments that dealt with cancer within the governorate for active data collection from medical records. Other sources of data were major tertiary centers on the national level as the National Cancer Institute of Cairo University, Pediatric Oncology Hospital in Cairo, and South Egypt Cancer Institute of Assiut University. Data managers in these institutions reported incident cancer cases among residents of the 3 governorates that were diagnosed/treated in these institutions. Health Insurance Organization and the Government-Subsidized Treatment Program periodically supplied their data. Death registers in local health directorates were regularly checked for cancer deaths from the 3 governorates. A web-based software was developed for online data entry, validity checks, and data analysis. The database was centralized in the Ministry of Communication and Information Technology server ( Figure 2) with backups in the Ministry of Health and the National Cancer Institute. Computer checks were achieved using DEPedits Conversion and Check Programs for Cancer Registries software [23]. Duplicates were eliminated using the National Identification Number and a clean database was achieved ready for analysis. Registration covered all invasive cancers (behavior code/3), in situ breast cancer (topography code C50. and behavior code/2), in situ urinary bladder cancer (topography code C67. and behavior code/2), and borderline tumors of the brain (topography code C71. and behavior code/1) [24].
For the current report, we used data from Aswan (2008), Minya (2009), and Damietta (2009-2011) to represent the 3 geographical strata of Egypt (Table 1). Crude, age-specific, and world population age-standardized incidence rates were Journal of Cancer Epidemiology 3 Table 1: Population parameters of the 3 regions of Egypt and corresponding registries used in the current report to calculate regional and national incidence rates.   5-167.8) calculated and expressed/100,000 population. Statistics were published at the level of the 3-character ICD-10 codes [10]. The format of IARC publication: "Cancer Incidence in Five Continents, Vol. X" [11], was used for grouping of ICD codes. Statistics for all sites of cancer were expressed twice, with and without nonmelanoma skin cancer (C44). Confidence intervals were calculated following SEER methodology [25].
We developed a model to use incidence data of the population-based registries of the 3 regions (Lower, Middle, and Upper Egypt) to get incidence rates for the entire country. The first step was to apply the age-specific incidence rates of each registry to the population of the stratum it represented to get the number per gender of incident cases in different age groups and for all ages together in this region using data of Egypt last census [26]. Then, we used these statistics to estimate the crude and age specific incidence rates (ASIR) of cancer in the 3 regions together, considered to be representative of the entire country. These rates were applied to the total population of Egypt to get number of incident cancers and crude rate and ASR (world) at national. This procedure was applied for each cancer site and for all sites together per gender. For projections up to 2050, the ASIRs were applied to the projected population of the corresponding year to get the number of patients/age group. Crude and ASR (world) could then be obtained. Steps of calculations are detailed in the Appendix.

National Incidence Rates and Proportions.
The crude incidence rates on the national level for all sites excluding nonmelanoma skin cancer (C44) were 113.1/100,000 (both sexes), 115.7/100,000 (males), and 110.3/100,000 (females). The age-standardized rates (world) were 166.6/100,000 (both sexes), 175.9/100,000 (males), and 157.0/100,000 (females) as shown in Table 2. Proportions, crude, age standardized incidence rates and detailed age-specific rates of cancer sites according to ICD-10 format are shown in Tables 3 and 4 for              Table 5 depicts the proportions and rates of the most frequent cancer sites by gender. There was predominance of liver, breast, and bladder cancer that represented approximately 46% of all cancers. Liver and bladder cancers represented approximately 44% of cancer in males. In females, breast and liver cancer occupied the top ranks accounting for around 45% of all cancers.

Frequencies and Incidence
Rates/Geographical Strata. The frequencies of individual sites of cancer and their incidence rates by geographical stratum and sex are detailed in Tables 6  and 7. Table 8    Among males in the 3 regions, liver and bladder cancer occupied the top 2 ranks. The proportion and ASR of liver cancer were highest in Lower Egypt (41.7% and 81.0/100,000) and lowest in Upper Egypt (11.8% and 17.5/100,000). Cancer of the lung occupied the third or fourth ranks representing 5-7% of cancers and Non-Hodgkin lymphoma was among the 5 most common cancers in Lower Egypt only having a proportion of 6.0% and ASR 10.3/100,000.
Among females, the pattern in the 3 regions was dominated by the high frequency of breast cancer and liver cancer. Proportion of liver cancer was highest in Lower Egypt (16.4%), less in Middle Egypt, and lowest in Upper Egypt (8.9% and 5.1% resp.).

Estimated Number of Incident Cancer Cases 2013-2050.
During the period 2013-2050, population of Egypt is expected to increase to approximately 160% the 2013 population size. Applying the current age-specific incidence rates to successive populations would lead to a progressive increase in number of incident cases from 114,985 in 2013 to 331,169 in 2050, approximately 290% of 2013 incidence (Table 9 and Figure 7). This increase reflected both population growth and demographic change mainly due to ageing of population.   Population growth alone would increase the number of incident cases by 55.2% in 2015. This fraction progressively decreased to become 32.8% in 2050. The fraction due to ageing gradually increased to reach 67.2% in 2050 (Figure 8).

Discussion
When the National Cancer Registry Program of Egypt was designed, there had been a number of challenges that were addressed. One of the most important challenges was design of a sample that could be representative of such a big country. Other challenges included complete capture of incident cancer cases among residents of selected governorates and best guarantee of quality of data and sustainability of the program to avoid going through survey-like data collection that need to be repeated, a methodology that proved to be inappropriate [27].
Dealing with population-based registration, the word national might be confusing. Actually, and with the exception of very few examples mainly old registries like that of Denmark and registries in small countries like Singapore, complete national coverage is not accomplished [27,28]. The US is an example of a huge country with 18 population-based registries spread over the states without a central national  registry [29]. Recently, national incidence rates of Turkey were published based on results of 8 geographically spread peripheral registries [30]. With this in mind, Egypt was geographically stratified for registration purposes into 3 regions: Lower Egypt, to the north of Cairo, and Middle, and Upper Egypt to the south. Three governorates were selected from the 3 regions representing 20.9% of the total population of Egypt. The greater Cairo, including the capital city, and Alexandria, the second capital (approximately 18 million population), were not included in the program due to difficulty in data collection and coverage of a population that daily migrates in and out of these 2 regions. Other regions that were not covered were frontier governorates; namely Matrooh, New Valley, Red Sea, and Sinai, together representing 3.3% of total population [26]. This exclusion was due to logistic reasons being mainly inhabited by nomads or internal migrants working in the tourist industry.
To ensure complete coverage of incident cancer patients in the selected governorates, data were actively collected from medical records of oncology centers within the governorates and from national referral centers like National Cancer Institute in Cairo. Registration covered health-insured patients and patients treated on government-subsidized treatment program would minimize under registration considered a point of strength of the registry. Also, regular check of death registers helped to decrease the possibility of under registration. Death certificate only cases (DCO) accounted for almost 8% of cases which could be considered an acceptable level for a newly established registry [28].
One of the elements that contributed to the National Cancer Registry Program of Egypt was the use of a unique national identification number for duplicate elimination to prevent over registration, which is a serious threat to population-based registries [7,28]. A major concern during the development of a cancer registry especially populationbased, is its sustainability. An alternative that was applied in Europe and the US was to conduct successive cancer surveys as with the US Health Interview and Examination surveys. Evaluation of these surveys was negative and this method was not recommended [28]. The main reason for lack of sustainability is unavailability of needed funding and lack of collaboration of treating physicians [28]. From the very beginning, the national cancer registry was planned to be a national program and not a project with start and end dates. During all stages of its development it was gradually incorporated within the infrastructure of the health system to gradually become part of everyday routine work [31]. Quality of data was assured through computer validation and regular manual checks on the peripheral and central levels.
We also developed a mathematical model to apply the regional age-specific incidence rates to the corresponding population structure to get national estimates of all cancers and for individual cancer sites for any specified year. Mathematical modeling is a common practice used to estimate national rates from statistics of regional registries. Recently, a model was developed for china to get national rates based upon regional registries [32]. The advantage of our Egyptian model was use of Egyptian national population-based data without importing data of adjacent countries [33].
Nonetheless, the program has its points of weakness. The data should be carefully interpreted in view of the short initial phase of registration that would be of better accuracy on successive years [28]. Furthermore, some private patients that are not covered by the Government-Subsidized Treatment Program might escape registration. This would be assumed to be a small proportion. With the rising cost of cancer treatment, most patients would seek government financial support which is a constriction right.
The results given in the current report are the first ever published incidence rates on a national and regional level for Egypt derived from a population-based cancer registry program. These incidence rates would replace the proportions that have always been derived from hospital-based results [1][2][3][4][5]. The results were those of Damietta (Lower Egypt), Minya (Middle Egypt), and Aswan (Upper Egypt). Damietta was selected to represent the Nile delta instead of Gharbiah registry that stopped its activities before establishment of the NCRP and needed some time to be restructured to join the program. Results of these governorates were used to compute regional rates that were used to get a national estimate based upon age-specific incidence rates and population structure.
The estimated incidence rates showed differences between the 3 regions that were mainly in the incidence of liver and bladder cancers. Among males, the proportion of incident liver cancer was highest in Lower Egypt (41.7%) and next in Middle Egypt (20.4%) and lowest in Upper Egypt (11.8%). Bladder cancer ranked next to liver cancer in Lower Egypt (8.8%). In Middle Egypt, the proportion was 14.2%, still lower than liver cancer. In Upper Egypt, liver cancer was still the most common cancer, with a small difference from bladder cancer (12.6%). These differences could be attributed   to the high prevalence of hepatitis C viral infection (HCV), which is one of the highest prevalence rates worldwide [34,35]. The distribution of liver cancer in the 3 regions followed the distribution of HCV, which is more frequent in Nile delta with decreasing prevalence going south [34]. The main risk factor for bladder cancer in Egypt was urinary Schistosomiasis which was more frequent in Upper Egypt and its prevalence decreased when going north [34]. Despite control of Schistosomiasis, its effect on bladder cancer needs time to disappear. Another difference between the 3 regions was the proportion of non-Hodgkin lymphoma that was distributed in a pattern similar to that of liver cancer [36,37]. Among females, the pattern in Lower, Middle, and Upper Egypt was dominated by the high frequency of breast cancer (33.8%, 26.8% and 38.7% resp.) and liver cancer (16.4%, 8.9% and 5.1% resp.). This pattern of liver cancer was similar to that of males with similar relation to the prevalence of HCV. Other differences in site distribution between the 3 regions will be detailed in a separate publication [38].
The national age standardized incidence rates for all cancers in Egypt, excluding nonmelanoma skin cancer, were 175.9/100,000 for males, 157.0/100,000 for females, and 166.6/100,000 for both sexes. The age-standardized rates were intermediate between the rates of more and less developed countries [33,39,40]. For both sexes, the rates for all cancers excluding nonmelanoma skin cancer were 268.3/100,000 for more developed countries and 147.7/100,000 for less developed countries compared to 166.6/100,000 in Egypt. For males, the age-standardized rates were 308.7/100,000 for more developed countries and 163.0/100,000 for less developed countries compared to 175.9/100,000 in Egypt. The rates of females were 240.6/100,000 in more developed countries and 135.8/100,000 in less developed countries compared to 157.0/100,000 for Egypt.     The model used in our study revealed the seriousness of the liver cancer that ranked first among cancers in males (33.6%) and next to breast cancer. Among females the proportion of breast cancer was 32.0% followed by liver cancer (13.5%). The high prevalence of HCV especially genotype IV would explain this high incidence [34,35]. The agestandardized rates for liver cancer for both genders were 5.4/100,000 in more developed countries, 12.0/100,000 in less developed countries compared to 43.6/100,000 in Egypt. Among males the rates were 8.6/100,000 in more developed countries, 17.8/100,000 in less developed countries compared to 61.8/100,000 in Egypt. Among females, the rates were 2.7/100,000 (more developed), 6.6/100,000 (less developed), and 24.4/100,000 (Egypt).
Incidence rate of bladder cancer that ranked next to liver cancer in males was also high. The age-standardized rates are 16.9/100,000 (more developed) and 5.3/100,000 (less developed) compared to 21.1/100,000 in Egypt. Breast cancer was the most frequent cancer among females. The age-standardized rates are 74.1/100,000 (more developed), 31.3/100,000 (less developed), and 48.8/100,000 (Egypt). Analysis of individual sites of cancers will be detailed in a separate publication [38].
Applying the model over successive years for the projected populations showed an increase in number of incident cases from approximately 115,000 patients in 2013 to more than 331,000 in 2050, almost 3-fold increase. The fraction of increase due to population growth gradually decreased over the years with a corresponding increase in the fraction due to demographic transition with ageing of the population. The cancer problem in Egypt would thus be expected to continue simply due to the inevitable ageing of the population with better standards of health care (Table 10).
Rates estimated by the model developed for the study were based on certain assumptions that should be considered in interpretation of results. These assumptions were: (a) constant age-specific rates during the study period; (b) populations of Greater Cairo and Alexandria would be considered a mosaic of the 3 registration regions and would affect the number of incident cases, and not rates; and (c) exclusion of frontier governorates (3.3% of total population of Egypt) that need special studies to get reliable cancer statistics.

Conclusion
These results are the first ever published incidence rates for Egypt on a national and regional level and clearly demonstrated the seriousness of the cancer problem of Egypt with age-adjusted incidence rates approaching those of the more developed countries. Liver cancer is a serious if not the most serious cancer problem in Egypt. Nonetheless, these rates should be carefully interpreted being based on a short initial phase of registration and a mathematical model that used regional incidence rates.