Incidence of leukemia in Eritrea:  11-year Laboratory -based retrospective analysis of nationally representative data

doi:10.21203/rs.3.rs-4917105/v1

Download PDF

Research Article

Incidence of leukemia in Eritrea: 11-year Laboratory -based retrospective analysis of nationally representative data

https://doi.org/10.21203/rs.3.rs-4917105/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Introduction:

Little or no research has been conducted on the epidemiology of leukemias in Eritrea. In this retrospective study, we evaluated the burden and trends of acute lymphoblastic leukemia (ALL), Acute myelogenous leukemia (AML), chronic myelogenous leukemia (CML), chronic lymphocytic leukemia (CLL) and overall leukemia in Eritrea.

Methods: An audit of leukemia cases recorded in laboratory logbooks at the National Health Laboratory (NHL) and Orotta Referral and Teaching Hospital (ORTH) between January 2010 and December 2021 was performed. Aside from leukemia sub-types, additional variables that were retrieved included age, sex, years of incidence, residency. Relevant estimates assessed included crude incidence rates (CIR), age-standardised rates (ASIR) and estimated annual percentage change (EAPC).

Results: In total, 372 confirmed cases of leukemia were recorded between, 2010-2020. The median [interquartile range (IQR)] age, maximum – minimum age, and male/female ratio were as follows: 48 years (24.5 – 60 years), 2 - 91 years, and 210/161 (1.3: 1), respectively. Estimated all-age CIR and ASIR over the study period was 9.22 per 100 000 and 30.1 per 100 000 respectively. Analysis of cumulative (2010 - 2020) CIR per 100 000 (ASIR per 100 000) for ALL, AML, CLL, and CML were as follows: 2.01(3.87); 0.94(2.38); 2.94(15.37) and 3.61(24.03). Additionally, median (IQR) age differed significantly across different subtypes of leukemia – ALL (23.0 years, IQR: 10.0 – 39.0); AML (30 years, IQR: 20 – 56 years), CLL (59.0 years, IQR: 40.75 – 66.75 years), and CML (49 years, IQR: 39.25 – 60 years), p value (Kruskal Wallis), < 0.05). No sex specific differences were observed in median (IQR) for different types of leukemia. Unlike other leukemia sub-types evaluation of EAPC demonstrated that the incidence of leukemia has increased overtime, 21.9 (95 CI: 3.1-44.1), p-value = 0.025.

Conclusions: The burden of leukemia was relatively stable. However, due to underreporting and underdiagnosis, it’s our belief that the true burden of leukemia is likely higher. Further, an upward trend in the burden of ALL was uncovered. Lastly, expansion of diagnostic services to other sub-zones, establishment of a national cancer registry and research remains a priority in Eritrea

Epidemiology

Oncology

Leukemia

hematological malignancy

Incidence

Trend

Eritrea

Leukemia is a compound term for a group of hematological malignancies characterised by clonal expansion of leucocytes (in the blood and/or bone marrow) subsequent to the malignant transformation of one or more normal hematopoietic progenitor cells [1]. The disease is subdivided by the tempo of disease (acute vs. chronic). Based on this subdivision, four main types of leukemia, which vary in pathogenesis, origin and prognosis, are recognized – acute (acute lymphoblastic leukemia (ALL) and Acute myelogenous leukemia (AML)) and chronic (chronic myeloid leukemia (CML)) and chronic lymphocytic leukemia (CLL) [2]. Additionally, a number of rarely diagnosed leukemia exists - Hairy cell leukemia. Temporal trends in incidence and prevalence of Leukemias are dynamic and multifactorial [3]. For instance, Global burden of disease (GBD) estimates suggest that global crude incidence rates (CIR) of leukemia in 2019 was 350 000 (Uncertainty interval (UI): 308 000–390 000) − 351 000 (UI: 308 000–390 000) in males vs 293 000 (UI: 263 000–322 000) [4]. Corresponding age-standardised rates (ASIR) was 8.2 (UI: 7.5–8.9) per 100 000 - (9.4 (UI: 8.3–10.5) per 100 000 in males vs 7.2 (UI: 6.5-8.0) per 100 000 in females. In terms of deaths and age standardised mortality rates (ASMR) these estimates indicates that the number of deaths recorded in 2019 was 335 000 (307 000-360 000) − 188 000 (165 0000-208 0000) in males vs 146 000 (132 000-158 000) in females. Global and location-specific annualized rates of change in CIR, ASIR, ASMR, disability adjusted life years (DALY) from 2010 to 2019 exhibited a highly mixed picture with some metrics demonstrating relative improvements and others demonstrating the converse [4]. Analysis of these metrics in leukemia subtypes demonstrating a similar trend [5].

Despite recent efforts devoted to cancer epidemiology, data from low-income and middle-income countries (LMICs) in sub-Saharan Africa (SSA) is sparse and, at times of questionable quality. However, even with this problematic data situation, data on the scope and trends in the burden of leukemia have become clearer [6]. One recurring theme in emerging estimates is that some sub-regions in the WHO Africa regions have some of the lowest ASIRs for leukemia globally − 3.9 per 100,000 persons in Western SSA and 3.89 per 100,000 persons in Central SSA [4, 6]. It also had the regions with the largest declines in ASIR in 2019: EAPC = − 1.11 in Eastern SSA. At the moment, patterning of leukemia ASIR and EAPC by country level of development or other epidemiological dynamics is difficult to explain. First, poor diagnostic infrastructure, poor access to formal health-care system, low awareness on the part of patients, low number high-quality population-based cancer registries (PBCRs) in most countries suggests that incidence of leukemias are under-researched and underreported in some LMICs. Extensive data analysis also indicates that even though cancer mortality rates have surpassed those of infectious diseases such as tuberculosis (TB), HIV/AIDS, and malaria combined; cancer has been given a low priority in the research field, and in healthcare services [7].

With few exceptions, the situation highlighted in the foregoing paragraph, describes the situation in Eritrea. The country does not have a cancer registry and the International Agency for Research in Cancer (IARC), GLOBOCAN estimates rely mostly on low quality data or extrapolations [8]. Additionally, access to cancer diagnostic is extremely limited due to the limited number of laboratories that can support morphological profiling involved in leukemia diagnosis - immunohistochemical, cytogenetic techniques or other advanced techniques are seldom available. Importantly, the country has a highly centralized laboratory services and a limited number of pathologists. This creates a significant rural-urban disparity in access to cancer diagnostics. Overall, there are many unanswered questions regarding the basic epidemiology of multiple cancers in Eritrea. Therefore, this study was designed to determine the CIR, ASIR, EAPC of ALL, AML, CML, and CLL using data from the two laboratories that provide diagnostic services for leukemias in Eritrea. Tracking the burden and epidemiological dynamics could provide relevant data for better policy making.

Study Design and Settings

An audit of Leukemia cases presenting to the National Health Laboratory (NHL) and Orotta Referral and Teaching Hospital (ORTH) between January 2010 and December 2020 was performed. At present, the two facilities are the only centers in Eritrea with the ability to diagnose Leukemia. NHL serves as a referral laboratory facility for all hospitals the country. On the other hand, Orotta hospital serves as a national referral hospital. Therefore, all leukemia suspected specimens in the country are processed at these two facilities.

Study population

Patients suspected of having leukemia in facilities across Eritrea and referred to NHL or ORTH for care of subsequent confirmation of diagnosis. All Eritrean patients recorded in the registry with diagnosis of leukemia, both genders, in different age groups between the periods of 1st January 2010 to 31st December 2020 were included in this study. On the other hand, Non-Eritrean patients with leukemia and registered cases missing principle data leukemia status or subtypes were excluded in this study. Overall, a total of 5600 samples were processed at the two facilities from 2010 to 2020 (2000 in NHL vs 3600 in ORTH).

Data Collection procedure

Data captured for the study consisted of information gathered from laboratory records in the two institutions. Age at diagnosis, sex, region of residence, and referring institution were included as key variables. Although the type of leukemia diagnosed was included; information on cancer stage and eventual outcomes were omitted. To maintain confidentiality and minimize duplications, codes were used. The inclusion criteria were all male and female patients of all ages who were diagnosed with leukemia in Eritrea. For comprehensiveness and minimization of missing information; data collected was cross-checked by three different enumerators.

Operational definition

Diagnosis of leukemia: The International Classification of Diseases (ICD) definition of disparate leukemia sub-types developed by the WHO was adopted in this study. Relevant codes include: AML (C92.0 - C92.02, C92.3 - C92.62, C93.0 - C93.02, C94.0 - C94.02, C94.2 - C94.22), ALL (C91.0 - C91.02), CML (C92.1 - C92.12), CLL (C91.1 - C91.12) [9]. Diagnosis was as per established cytological protocols [10].

Main outcome

The primary outcomes included age standardized incidence rate and crude incidences ALL, AML, CLL and CML. Other variables included hematological profiles associated with specific leukemia’s.

Data analysis

Data were reviewed for incompleteness or missing entries and then entered into the computer package using IBM SPSS Software (Statistical Package for the Social Sciences, version 20, Chicago, Ill, USA). Where applicable, mean ± standard deviation (± SD) or median + interquartile range (IQR) was used to summarize the data. Categorical variables such as age, gender, types of leukemia were presented as frequencies or percentages. A two-sided Chi-square test (χ²) was subsequently used to evaluate relevant associations. CIR, ASIR and estimated average percentage change (EAPC) were calculated using. Briefly, annualized CIR, was estimated by dividing the total number of cases by the corresponding population at risk, expressed per 100 000 person-years. The direct method was used to estimate ASIR. In this analysis, a 10-year age band (< 40–49, 50–59, 60–69 and > 70) was employed. ASIR conversions were based on world population net 2014. The following formula was used to compute ASIR or ASR:

Where (where α_k = distribution of the selected reference standard population in the k age groups, and β_k = age-specific rate). For annualized computation of EAPC, generalized linear model with Gaussian distribution was employed. A regression line is fitted to natural logarithm of annual ASR rates (i.e. y = α + βx + ε, where x = year and y = In (rate), from 95% CI were obtained; EAPC was computed as 100 x (e^β-1) [11].

Ethical consideration

The study was approved by the Health Research Ethics and Protocol Review Committee of the Eritrean Ministry of Health (MOH). Consent to participate was not obtained from the patients. This waiver was premised on the consideration that the study was based on anonymized patient’s records.

Demographic characteristics of patients

In this study, 372 confirmed cases of leukemia identified in existing records. The median age (interquartile range (IQR)) at diagnosis was, 48 years (IQR: 24.5 – 60 years), minimum and maximum (Min-Max) age was 2-91 years and the male-to- female ratio of 210/161 or 56.6% vs 43.4%. The age-related distribution of all leukemia was 69 (18.6%) in for young patients (1–20 years), 84 (22.6%) for young adults (21 - 40 years), 122 (32.8%) for adults (41–60 years), and 160 (43%) for older adults (61 - 90 years) (Table 1). In addition, most of the patients came from the OPD, 167 (49.1%). Other patients were from the medical ward, 84 (24.7%) and the emergency ward 51 cases (15.0%).

Table 1. Sociodemographic characteristics of Leukemic patients diagnosed in Asmara, Eritrea: 2010-2020

Characteristic		No of patients	Frequency (%)
Gender
	Male	209	56.6
	Female	160	43.4
Total		369	100
Age group
	0-10	30	8.1
	11-20	39	10.5
	21-30	40	10.8
	31-40	44	11.8
	41-50	59	15.9
	51-60	63	16.9
	61-70	49	13.2
	71-80	37	9.9
	81-90	11	3.0
Total		372	100
Patients portal
	OPD	167	49.1
	Medical ward	84	24.7
	Emergency Ward	51	15.0
	Podiatric	36	10.7
Total		338	100

Abbreviation: OPD – Outpatient Department

Frequency of different types of leukemia recorded from 2010 to 2020

Table 2 presents data on the frequency of disparate types of leukemia in this cohort. Overall, 132 (35.4%) vs 240 (64.5%) had acute and chronic leukemia, respectively. The frequency of ALL, AML, undifferentiated acute leukemia, was 75 (20.1%), 41(11.0%) and 16 (4.3%), respectively. Similarly, frequency of CLL and CML was 108 (29.0%) and 132 (35.4%), respectively.

Table 2. Proportion of the different Leukaemia varieties diagnosed in Asmara, Eritrea: 2010-2020

No of cases
Type of leukemia	N	Relative(%)	Group (%)
Total	372	100	100
Acute leukemia	132	35.4	100
ALL	75	20.1	56.8
AML	41	11.0	31.0
Undifferentiated	16	4.3	12.1
Chronic leukemia	240	64.5	100
CLL	108	29.0	45
CML	132	35.4	55

Abbreviation’s: ALL: Acute lymphocytic leukemia, AML: Acute myelogenous leukemia, CLL: Chronic lymphocytic leukemia, CML: Chronic myeloid leukemia

Age distribution of different types of leukemia

The overall age distribution of leukemia cases in this population of bimodal. In decreasing frequency, the estimates were in the order: 50-59 years, 63(16.9%); 40-49 years, 59 (15.8%); 60.9 years 49(13.1%); > 70 years, 49 (13.1%); 30-39 years, 44 (11.8%); 20 – 29 years, 40 (10.7); 10 – 19 years, 39 (10.5%); <10 years, 30 (8.0%). Additionally, median (IQR) age differed significantly across different subtypes of leukemia – ALL (23.0 years, IQR: 10.0 – 39.0); AML (30 years, IQR: 20 – 56 years), CLL (59.0 years, IQR: 40.75 – 66.75 years), and CML (49 years, IQR: 39.25 – 60 years), p value (Kruskal Wallis), < 0.05). No sex specific differences were observed in median (IQR) for different types of leukemia. See Figure 2 for additional information.

Hematological profiles of patients at presentation

To evaluate presentation at diagnosis, complete blood count (CBC) for selected group of patients, 102 out of 369 patients was undertaken. Due to missing data, this analysis was restricted to the following timeframe: 2016, 2017, 2019 and 2020. The median (IQR) value of the total WBC, MCV, platelet count (PLTs), and hemoglobin (Hb) at diagnosis was 50.26 x 10⁹/L (IQR: 20 x 10⁹ /L – 134.7 x 10⁹ /L); 90.6 fL (86.1 – 97.8 fL); 97.0 x10⁹/L (27.5 x 10⁹ - 270.25 x 10⁹) and 10.0 g/dL (7.8 - 11.0 g/dL), respectively. Overall, 62% of the patients were anaemic at diagnosis (Hb < 10g/dL), while 51.6% were thrombocytopenic (platelet count < 100 x 10⁹/L). As seen in Fig 3, significant differences (p-value < 0.05) in median (IQR) values were observed in disparate subtypes of leukemia. On analysis of symptoms and signs, it was observed that fever was the commonest symptom, followed by weakness, bleeding and bone pain. The incidence of hepatomegaly and splenomegaly was common mainly in CML and CLL.

Incidence of Leukemia per gender diagnosed in Asmara, Eritrea: 2010-2020

Table 3. Incidence of Leukaemia per gender diagnosed in Asmara, Eritrea: 2010-2020

Year	Male			Female			Male: Female	Both Total
Year	Count	CIR	ASIR	Count	CIR	ASIR	Male: Female	Count	CIR	ASIR
2010	21	1.16	1.54	4	0.22	0.28	5.3:1	25	0.69	1.82
2011	18	0.10	1.33	15	0.83	1.2	1.2:1	33	0.91	2.53
2012	17	0.94	1.77	25	1.38	1.86	1:1.5	42	1.16	3.63
2013	24	1.33	2.12	12	0.67	1.01	2:1	36	0.10	3.13
2014	17	0.67	1.59	11	0.61	0.92	1.6:1	28	0.64	2.51
2015	15	0.83	1.28	13	0.72	0.88	1.2:1	28	0.78	2.16
2016	20	1.11	1.75	19	1.05	1.37	1.1:1	39	1.08	3.12
2017	24	1.33	2.09	22	1.22	1.48	1:1.1	46	1.26	3.57
2018	10	0.55	0.82	11	0.61	0.92	1:1.1	21	0.58	1.74
2019	12	0.67	1.01	17	0.94	1.22	1:1.4	29	0.80	2.23
2020	32	1.77	2.73	12	0.67	0.92	2.7:1	44	1.22	3.65
AAPC	2.1			-3.2				-0.1
95 (CI)	-5.7-10.6			-10-4.1				-6.1-6.7
*p-value*	0.56			0.35				0.9
Total	210	10.46	18.03	161	8.92	12.06	1.3:1	371	9.22	30.1

Abbreviation: ASIR: Age standardized Incidence Rate; CIR: Crude incidence rate

The cumulative CIR, ASIR and 10 year - EAPC for the population under study were 9.22 per 100 000 persons, 30.1 per 100 000 persons and – 1 (-6.1 – 6.7), p-value 0.9. Disaggregation of CIR, ASIR and AAPC data by sex was undertaken. In this analysis, estimates indicated a CIR of 10.46 per 100 000 persons in males and 8.92 per 100 000 persons in females. Associated ASIR were 18.3 per 100 000 (2010 – 2020 AAPC 95 CI: - 2.1 (-5.7 – 10.6), p value 0.56) in males vs 12.06 per 100 000 persons (2010 – 2021 EAPC 95% CI: - 3.2 (-10 – 4.1), p-value 0.035) in females.

Crude incidence rate and age-standardized incidence rates for subtypes of leukaemia diagnosed in Eritrea: 2010-2020

The CIR, ASIR and EAPC over the study period (2010 – 2021) were as follows: 2.01 per 100 000 and 3.87 per 100 000, EAPC, 21.9%(95 CI: 3.1-44.1%), p value = 0.025 for ALL; 0.94 per 100 000 and 2.38 per 100 000, EAPC, 14.9%, p value = 0.57, for AML; 2.94 per 100 000 and 15.37 per 100 0000, EAPC, -2.1%, p value = 0.65, for CLL; 3.61 per 100 000 and 24.03 per 100 000, EAPC, -0.2%, p value = 0.95 for CML; and 9.68 per 100 000 and 30.5 per 100 000. See Table 4 for additional details.

Table 4.Crude incidence rate and age-standardized incidence rates for leukaemia for the types of leukaemia diagnosed in Asmara, Eritrea: 2010-2020

Year of diagnosis	ALL		AML		CLL		CML		Leukemia
	Cases (CIR)	ASIR	Cases (CIR)	ASIR	Cases (CIR)	ASIR	Cases (CIR)	ASIR	Cases (CIR)	ASIR
2010	7(0.19)	0.06	4(0.11)	0.22	6(0.16)	0.83	8(0.22)	2.33	25(0.69	3.44
2011	2(0.05)	0.24	0(0)	0.00	13(0.36)	1.31	15(0.41)	1.23	30(0.83	2.78
2012	6(0.16)	0.30	7(0.19)	0.34	10(0.27)	1.49	17(0.47)	2.26	40(1.11	4.39
2013	8(0.22)	0.50	4(0.11)	0.06	14(0.38)	2.35	11(0.30)	2.17	37(1.02	5.08
2014	6(0.16)	0.26	2(0.05)	0.19	8(0.22)	1.11	12(0.33)	3.24	28(0.77	4.80
2015	4(0.11)	0.05	2(0.05)	0.19	13(0.36)	1.60	9(0.24)	2.15	28(0.77	3.99
2016	5(0.13)	0.39	2(0.05)	0.03	12(0.33)	1.73	15(0.41)	2.53	34(0.94	4.68
2017	9(0.24)	0.66	3(0.08)	0.04	14(0.38)	2.31	14(0.38)	2.03	40(1.11	5.04
2018	9(0.24)	0.68	0(0)	0.00	6(0.16)	0.96	6(0.16)	0.84	21(0.58	2.48
2019	5(0.13)	0.25	7(0.19)	0.74	7(0.19	0.94	8(0.22)	0.48	27(0.75	2.41
2020	14(0.38)	0.48	4(0.11)	0.57	5(0.13)	0.74	17(0.47)	4.77	40(1.11	6.56
EAPC		21.9		14.9		-2.1		-0.2
95 (CI)		3.1-44.1
*P-value*		0.025		0.57		0.65		0.95
Total	75 (2.01)	3.87	35(0.94)	2.38	108(2.94)	15.37	116(3.61)	24.03	350 (9.68)

Abbreviation: ASIR: Age standardized Incidence Rate; CIR: Crude incidence rate, EAPC: Average annual percentage change; ALL: acute

Crude incidence rate and age-standardized incidence rates for Leukemia across 10-year age bands: 2010-2021

The cumulative CIR (ASIR) in the disparate age-bands was as follows: < 10 years, 0.51 per 100 000(1.02 per 100 000); 10 – 19 years, 1.01 per 100 000 (1.54 per 100 000); 30 – 39 years, 0.42 per 100 000 (2.63 per 100 0000); 40 – 49 years, 000 per 100 0000 (5.08 per 100 000); 50 – 59 years, 3.88 per 100 000 (7.16 per 100 0000); 60 – 69 years, 0.79 per 100 000 (3.66 per 100 000); 70 – 79 years, 5.05 per 100 000 (3.66 per 100 000) and > 80 years 3.96 per 100 000 and 1.45 per 100 000. See table 6 for additional details.

Table 6:- Crude incidence rate and age-standardized incidence rates for Leukemia across 10-year age bands: 2010-2020

Year	<10	10-19	20-29	30-39	40-49	50-59	60-69	70-79	>80
Year	CIR/ASIR	CIR/ASIR	CIR/ASIR	CIR/ASIR	CIR/ASIR	CIR/ASIR	CIR/ASIR	CIR/ASIR	CIR/AISR
2010	0.51/0.15	1.01/0.31	0/0	0.42/0.12	0/0	3.89/0.68	0.79/0.12	5.05/0.3	3.96/0.14
2011	0.61/0.18	0.23/0.08	1.04/0.22	1.05/0.18	4.56/0.7	3.33/0.44	6.34/0.52	2.52/0.21	0/0
2012	0.4/0.07	0.56/0.12	0.52/0.12	3.15/0.49	6.32/0.79	7.77/1.24	3.97/0.32	3.79/0.19	7.93/0.29
2013	0/0	0.11/0.04	1.57/0.39	1.05/0.18	3.51/0.72	3.89/0.68	9.52/0.85	6.32/0.27	0/0
2014	0.31/0.08	0.33/0.08	0/0	1.05/0.25	1.40/0.27	4.44/0.8	10.31/0.84	3.79/0.19	0/0
2015	0.31/0.08	0.68/0.12	1.04/0.22	1.47/0.31	1.75/0.36	1.66/0.57	3.96/0.2	1.26/0.3	0/0
2016	0.41/0.11	0.90/0.2	1.22/0.23	1.47/0.25	1.40/0.36	3.331/0.44	8.72/0.87	13.9/0.66	0/0
2017	0.71/0.18	1.12/0.24	0.87/0.17	2.73/0.43	2.45/0.54	5.00/0.91	2.38/0.22	12.63/0.67	15.87/0.2
2018	0/0	0.34/0.08	1.22/0.28	0.21/0.06	1.40/0.27	2.78/0.32	0.79/0.12	7.58/0.26	19.8/0.35
2019	0.41/0.07	0.56/0.12	1.74/0.39	1.05/0.18	1.40/0.17	4.99/0.53	3.17/0.33	3.79/0.3	3.97/0.21
2020	0.51/0.18	0.45/0.15	0.7/0.23	0.63/0.18	3.51/0.9	2.78/0.55	6.35/0.9	3.79/0.31	7.93/0.25

Abbreviations: CIR: crude incidence rates, ASIR: Age adjusted incidence rates

The results of this retrospective analysis demonstrate a substantial burden of leukemia in Eritrea. Accordingly, the main finding is our study is that the cumulative (2010 – 2021) CIR and ASIR of leukemia in this population was 9.22 per 100 000 and 30.1 per 100 000 respectively, male/female (M/F) ratio = 1.3:1. The incidence of leukemia was higher than previous estimates by IARC-GLOBOCAN estimates for Congo, Brazzaville (2009 - 2013), 2.2 per 100 000; Ethiopia (2013 - 12) [12]. Although the ASIR estimates are comparable (3.2 per 100 000 in 2020) to those in a recent report; they are lower than estimates from Ethiopia, 9.5 per 100 000 (CI: 5.3 - 13.4 per 100 000) [13]. The higher incidence of leukemia in males was remarkably consistent with previous reports - 2017 GBD estimates were 197 270/150310 (1.3/1) [14].

Disaggregation of data by age showed a higher incidence of acute leukemias (ALL and AML) in younger patients and a higher incidence of chronic leukemias (CML and CLL) in older patients. Importantly, past reports have indicated that the incidence of leukemias is higher in individuals aged >50 years. These estimates are broadly consistent with previous reports. The median age of patients with CML (49 years, IQR: 39.25 – 60 years) was strikingly similar to those reported in a clinical trial of imatinib (a tyrosine kinase inhibitor) from high income countries (HICs) [15]. This age distribution may reflect the age structure of the population in Eritrea or other biases.

Further, assessment of the frequency of the disparate subtypes of leukemia in this population uncovered some remarkable findings. In terms of proportions, we found that the overall all-age distributions were as follows: ALL, AML, CLL and CML were as follows: 75 (20.1%), 41(11.0%), 108 (29.0%) and 132 (35.4%), respectively. In contrast, a difference distribution was observed globally and in the region. In North West Ethiopia, the following distribution was observed: CLL - 61.1% (58/95); ALL - 23.2% (22/95); CML - 6.3% (6/95); AML - 4.2% (4/95), and other hematological malignancies - 5.3% (5/95) [16]. According to Dong et al., CLL and AML were the predominant subtypes in high-SDI countries located in Europe and North America [17]. In contrast, CML was rarely diagnosed. Relative to these reports, our result is highly unusually considering the disproportionate diagnosis of CML. These notwithstanding, we have to emphasize the fact that proportions of disparate leukemia subtypes is highly heterogeneous in between-country comparisons and that the underlying reasons behind these disparities is poorly understood.

A further difference that was observed in this population was the age at diagnosis of ALL (23.0 years vs 5.0 years in Egypt) [18]; CLL (59.0 years vs 59 in Nigeria and 70 years in the USA) [19]; and CML [49 years vs. 34 years in Rwanda]. Of note, the low median age for CML and its predominance in this population is unusual. In general, the dominant risk factors for CML are aging and exposure to radiation [4]. Indeed, past reports suggest that incidence of CML correlates strongly with advancing age peaking at 10.3 cases per 100,000 among individuals 80 - 84 years of age with frequent age at diagnosis of 65 and 74 years of age [4]. More recently, Hoglund et al. argued that the geographic disparities in ASIRs of CML could partly be attributed to the methodological factors such as the overrepresentation of referral patients in some registries [20]. Others have argued that the advent of tyrosine kinases inhibitors (TKIs) has transformed CML from a fatal disease to a disorder that is compatible with a normal lifespan and that this has increased its prevalence [21]. We don’t believe that this is the case in LMICs. However, this argument together with existing demographic trend in HICs may explain the high burden of CML in these regions. Altogether, we have to concede that the unusual (for the region) nature of the distribution of leukemia subtypes or age at diagnosis is hard to explain and warrant further investigations.

Importantly, we noted that the incidence of AML, CLL and CML remained stable over the years. This is contrary to recent reports which has uncovered increases in CLL (0.46% per years from 1990 – 2017); decreases in CML in most regions. In another unusually finding, we noted that the EAPC associated will ALL has increased significantly over the years. In contrast, a recent report noted that the ASIR associated with ALL experienced a subtle decrease from 1990 – 2017, EAPC = −0.08 (95% CI −0.15, −0.02) with the greatest decrease detected in Ghana, EAPC= -5.18, 95% (CI: - 6.23, - 4) [17]. Altogether, the increasing ASIR in ALL may be caused by environmental factors and requires further research.

In the past, the observed pattern in the disparities in cancer burden or patterns of distribution and trends has been explained by invocation of specific risk factor dynamics, underdiagnosis, under registration, misdiagnosis or changes in classification protocols [22]. Underdiagnosis is mostly a by-product of the well documented deficiencies in laboratory systems and health delivery components such as management of health information [23]. Emphasizing this point, authors of a recent GBD report noted that < 10% of the population in Africa and other LMICs in Asia and South America are covered by high-quality cancer registries [22]. According to Lam et al., some types of cancers are poorly represented in data from some regions [24]. According to this opinion underreporting (deficit between actual and reported incidences) is more pronounced in malignancies such as acute leukemias, neuroblastomas, and bone tumors due to the requirement of advanced diagnostic techniques (Flow Cytometry Markers - immunophenotyping, immunohistopathology and advanced imaging technologies such as positron emission tomography (PET) scans, among others) [24]. Altogether, the problem of underreporting and lack of PBCRs means that estimates of cancer burden rely heavily on covariate selection in models and regional patterns [22]. Much of what is described here is true for Eritrea. As previously noted, the concentration of oncologic and diagnostic infrastructure in Asmara, Zoba-Maekel; limits access to cancer care in the country. This problem is potentially compounded by knowledge deficit among health workers and poor referrals to specialized/reference hospitals.

A detailed analysis of the etiologies behind the observed variation in leukemia incidence across countries in SSA goes beyond the scope of this analysis. However, multiple factors from studies in HICs and non-Africa populations have been invoked. For example, the diverse incidence of leukemias across GBD-super regions is driven by between-country disparities in occupational exposures to putative risk factors such as aging and environmental hazards – benzene, formaldehyde, and among other persistent organic pollutants (POPs) [26, 27, 28]; therapeutic exposure (alkylating agents and topoisomerase II inhibitors); pre-natal and post-natal exposure to ionizing radiation (particularly X rays) [29], herbicides/or pesticides, obesity (Body mass index (BMI) ≥ 30 Kg/m²; smoking or tobacco use; birth weight [18, 30]. The role that these factors are currently playing in the observed increase in ALL is unknown. However, in prevalence and ASIRs between males and females is generally attributed to differential exposures to risk factors such as smoking (a predominantly male pass time) and obesity [31]. Interestingly, a recent reported noted that in 2019, ALL and AML were the leading causes of leukemia attributable to occupational risk-related DALYs and death rates [32]. Further, the development of leukemia has been partly ascribed to a number of syndromes whose epidemiology in SSA is poorly understood - Fanconi anemia, Bloom syndrome, Schwachman Diamond syndrome, Diamond Blackfan anemia, among others. Others factors which have a role in leukemogenesis include acquired of inherited genetic and epigenetic mutations – CLL (chromosome 13q (del[13q]), Hyperdiploidy [18]; CML, 9:22 translocation (Philadelphia chromosome +) (q34; q11.2) or BCR-ABL fusion, AML, t(15;17), t(8;21), inv(16), or t(16;16) among others. Obviously, research on the connection between this panoply of risk factors and leukemia in SSA is inadequate. Therefore it follows, and an overview of literature supports the fact that very little is known about the etiological factors driving leukemia incidence or molecular epidemiology of leukemias in most countries in SSA. The need for such investigations in Eritrea is therefore apparent.

Limitations of the study

The unique feature of this study is that we provided a comprehensive overview of the burden and trends of leukemias in Eritrea from 2010 – 2021. However, the study has a number of limitations. First, retrospective studies have a data quality problem – missing information and variables, questionable robustness or data collection or data quality evaluation among others. Secondly, it can be argued that the diagnostic and classification criteria for leukemia subtypes have changed overtime and that these may have an impact on the comparability of our findings. In Eritrea, diagnosis of leukemia is exclusively – further subtype subclassification using cytogenetics, molecular genetics or immunophenotyping is not undertaken. Third, and most importantly, it should be noted that although the ASIR is a useful measure for comparing disease burden across populations with disparate age structures and sizes; in countries with relatively small populations like Eritrea, the reliability of ASIR may be compromised due to small number of cases and the resultant outsized effect of random variation. Due to this problem, uncovering trends can be problematic. This notwithstanding, and for reasons detailed elsewhere in this study, it’s our belief that the approach adopted in this study is superior to alternatives which rely exclusively on modelling and extrapolations from neighboring countries.

The burden of leukemia was relatively stable. However, due to underreporting and underdiagnosis, it’s our belief that the true burden of leukemia is likely higher. Therefore, our result suggests the need for increased cancer prevention and control efforts as well as the need to establish a national cancer registry. Towards this end, experts have emphasized the need for effective national cancer control plans (NCCPs). However, establishments of NCCPs can be hindered by poor data on local cancer epidemiology. Therefore, additional research covering other aspects of leukemia epidemiology in Eritrea is required. In the last two decades, advances in laboratory techniques and technologies (next-generation sequencing – cancer genomics, immunophenotyping or flow cytometry, cytogenetics, among others) have greatly improved the diagnosis and management of leukemias. Although a daunting or perhaps impossible task in LMICs, the adoption, at scale, of some of these technologies in laboratories in Eritrea should be explored.

Funding: This research received no external funding.

Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki and approved by Ethical Committee of the Eritrean Ministry of Health. Written consent was not obtained from patients. Instead, the data was de-identified and all ethical and professional considerations undergirding confidentiality/or others protections of patient information were rigorously implemented.

Data Availability Statement: The data are not publicly available due to privacy and confidentiality. Restrictions apply to the availability of data but are available for the authors with the permission of the organization.

Author’s contribution: Conceptualization: DMA and EG. Data curation: DMA, EG, EG, GEZ, RK, AT, SN, TD, and SM. Formal analysis: STM, and OOA. Investigation: DMA, EG, STM, and OOA. Methodology: DMA, EG, EG, GEZ, RK, AT, SN, TD, and SM. Project administration: DMA and EG. Writing—original draft: DMA, EG, and OOA. Writing—review and editing: DMA, EG, STM, and OOA. All authors read and approved the fnal manuscript

Conflicts of Interest: The authors declare no conflict of interest.

Juliusson G, Hough R, Leukemia (2016) Progress tumor Res 43:87–100
Karagianni P, Giannouli S, Voulgarelis M (2021) From the (Epi)genome to metabolism and vice versa; examples from hematologic malignancy. Int J Mol Sci 22:6321
Huang J, Chan SC, Ngai CH, Lok V, Zhang L, Lucero-Prisno DEIII, Xu W, Zheng Z-J, Elcarte E, Withers M, Wong MCS (2022) Disease Burden, Risk Factors, and Trends of Leukaemia: A Global Analysis. Front Oncol 12:904292. 10.3389/fonc.2022.904292
Global Burden of Disease 2019 Cancer Collaboration, Kocarnik JM, Compton K, Dean FE, Fu W, Gaw BL et al (2022) Cancer incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life years for 29 cancer groups from 2010 to 2019: a systematic analysis for the global burden of disease study 2019. JAMA Oncol 8:420–444
Bispo JAB, Pinheiro PS, Kobetz EK (2020) Epidemiology and Etiology of Leukemia and Lymphoma. Cold Spring Harb Perspect Med 10:a034819
Zhang N, Wu J, Liang Y, Li X, Chen G et al (2023) Global burden of hematologic malignancies and evolution patterns over the past 30 years. Blood Cancer J 13:82. https://doi.org/10.1038/s41408-023-00853-3
Hamdi Y, Abdeljaoued-Tej I, Zatchi AA, Abdelhak S, Boubaker S, Brown JS, Benkahla A (2021) Cancer in Africa: The Untold Story. Front Oncol 11:650117. 10.3389/fonc.2021.650117
Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A (2018) Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA J Clin 68(6):394–424
International Agency for Research on Cancer, World Health Organization (2021) International Classification of Diseases for Oncology. ICD-O-3-Online. http://codes.iarc.fr/home. Accessed August 10
Hallek M, Cheson BD, Catovsky D, Caligaris-Cappio F, Dighiero G, Döhner H, Hillmen P, Keating MJ, Montserrat E, Rai KR (2008) Guidelines for the diagnosis and treatment of chronic lymphocytic leukemia: a report from the International Workshop on Chronic Lymphocytic Leukemia updating the National Cancer Institute-Working Group 1996 guidelines. Blood. ;111(12):5446–56
Clegg LX, Hankey BF, Tiwari R, Feuer EJ, Edwards BK (2009) Estimating Average Annual Percent Change in Trend Analysis. Stat Med 28(3):3670–3682. 10.1002/sim.3733
Parkin DM, Ferlay PJ, Jemal A, Borok M, Manraj SS et al Cancer in Sub-Saharan. IARC Scientific Publications – No. 167: https://publications.iarc.fr/Book-And-Report-Series/Iarc-Scientific-Publications/Cancer-In-Sub-Saharan-Africa-2018 accessed on 31.8.23
Awedew AF, Asefa Z, Belay WB (2022) National Burden and Trend of Cancer in Ethiopia, 2010–2019: a systemic analysis for Global burden of disease study. Scientifc Rep 12:12736. https://doi.org/10.1038/s41598-022-17128-9
Zejin OU, Danfeng Y, Yuanhao L, Wenqiao H et al (2020) Analysis of the Global Burden of Disease study highlights the trends in death and disability-adjusted life years of leukemia from 1990 to 2017. Cancer Commun 40:598–610
Hehlmann R, Lauseker M, Saubele S et al (2017) Assessment of imatinib as first-line treatment of chronic myeloid leukemia: 10-year survival results of the randomized CML study IV and impact of non-CML determinants. Leukemia 31:2398–2406
Enawgaw B, Aynalem M, Melku M, Asrie F, Abebe M, Yalew A et al (2021) Hematological malignancies in the Northwest Ethiopia. PLoS ONE 16(12):e0260639. https://doi.org/10.1371/journal. pone.0260639
Dong Y, Shi O, Zeng Q, Lu X, Wang W, Li Y, Wang Q (2020) Leukemia incidence trends at the global, regional, and national level between 1990 and 2017. Exp Hematol Oncol 9:14. https://doi.org/10.1186/s40164-020-00170-6
Adbelmabood S, Fouda AE, Boujetiff F, Mansour A (2020) Treatment outcomes of children with acute lymphoblastic leukemia in a middle-income developing country: high mortalities, early relapses, and poor survival. J Pediatr (Rio J) 96(1):108–116
Madu AJ, Korubo K, Okoye A, Ajuba I et al Presenting features and treatment outcomes of chronic lymphocytic leukaemia in a resource poor Southern Nigeria. Malawi Med J 31 (2): 144–149 June 2019.
Höglund M, Sandin F, Simonsson B (2015) Epidemiology of chronic myeloid leukaemia: an update. Ann Hematol 94(Suppl 2):S241–247. 10.1007/s00277-015-2314-2
Hochhaus A, Larson RA, Guilhot F, Radich JP, Branford S, Hughes TP et al (2017) Long-Term Outcomes of Imatinib Treatment for Chronic Myeloid Leukemia. N Engl J Med 376(10):917–927. 10.1056/NEJMoa1609324
Global Burden of Disease Cancer Collaboration Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived with Disability, and Disability Adjusted Life-years for 32 Cancer Groups, 1990 to 2015: A Systematic Analysis for the Global Burden of Disease Study. JAMA Oncol 2017 April 01; 3(4): 524–548. 10.1001/jamaoncol.2016.5688
Bray F (2016) The evolving scale and profile of cancer worldwide: much ado about everything. Cancer Epidemiol Biomarkers Prev 25(1):3–5. 10.1158/1055-9965.EPI-15-1109
Lam CG, Howard SC, Bouffet E, Pritchard-Jones K (2019) Science and health for all children with cancer. Science 363:1182–1186. 10.1126/science.aaw4892
Morhason-Bello IO, Odedina F, Rebbeck TR, Harford J, Dangou JM, Denny L et al (2013) Challenges and opportunities in cancer control in Africa: a perspective from the African Organisation for Research and Training in Cancer. Lancet Oncol 14:e142–e151. 10.1016/S1470-2045(12)70482-5
Ward MH, Colt JS, Metayer C et al (2009) Residential exposure to polychlorinated biphenyls and organochlorine pesticides and risk of childhood leukemia. Environ Health Perspect 117(6):1007–1013
Ward MH, Colt JS, Deziel NC et al (2014) Residential Levels of Polybrominated Diphenyl Ethers and Risk of Childhood Acute Lymphoblastic Leukemia in California. Environ Health Perspect 122(10):1110–1116
Yi M, Li A, Zhou L, Chu Q et al (2020) The global burden and attributable risk factor analysis of acute myeloid leukemia in 195 countries and territories from 1990 to 2017: estimates based on the global burden of disease study 2017. J Hematol Oncol 1372. https://doi.org/10.1186/s13045-020-00908-z
Nikkilä A, Erme S, Arvela H, Holmgren O, Raitanen J, Lohi O, Auvinen A (2016) Background radiation and childhood leukemia: a nationwide register based case-control study. Int J Cancer 139(9):1975–1982
Andreotti G, Koutros S, Hofmann JN, Sandler DP, Lubin JH, Lynch CF, Lerro CC, De Roos AJ, Parks CG, Alavanja MC et al (2018) Glyphosate Use and Cancer Incidence in the Agricultural Health Study. J Natl Cancer Inst 110(5):509–516
Li X, Wang J, Huang X, Wang H, Li F Global, regional, and national burdens of leukemia from 1990 to 2017: a systematic analysis of the global burden of disease 2017 study. AGING 2021, 13, No. 7
Shi Y, Chen C, Huang Y, Xu Y, Xu D, Shen H, Ye X, Jin J, Tong H, Yu Y, Tang X, Li A, Cui D, Xie W (2022) Global disease burden and trends of leukemia attributable to occupational risk from 1990 to 2019: An observational trend study. Front Public Health 10:1015861. 10.3389/fpubh.2022.1015861

The authors declare no competing interests.

Download PDF

Version 1

posted

You are reading this latest preprint version

Incidence of leukemia in Eritrea: 11-year Laboratory -based retrospective analysis of nationally representative data

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study Design and Settings

Study population

Data Collection procedure

Operational definition

Main outcome

Data analysis

Ethical consideration

Results

Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1