Clinico-Pathological Findings and Spatial Distributions of Esophageal Cancer in Arsi Zone, Central Ethiopia

Background: Arsi Zone is one of the esophageal cancer endemic areas and reported to have high incidence of esophageal cancer in Ethiopia. This study assessed clinico-pathological patterns and spatial distributions of esophageal cancer in Arsi Zone, Oromia, Central Ethiopia. Methods: A cross sectional study design was carried out by abstracting data from 403 patients` records that were treated from January 2015 to January 2019. Data were collected by using structured data collection form and Geographical Positioning System software. The collected data were summarized in the form of tables, gures, means and standard deviations. Statistical data analysis was done using SPSS software version 21.0 while geospatial analysis was conducted using the Arc-GIS 10.1 software. Results: The disease was prominent among individuals aged 50 to 59 years. The mean age of females and males was 52.2(SD+10.41) and 56.94(SD +12.27) years respectively. The vast majority (98.3%) of the cases had squamous cell carcinoma. Dysphagia with weight loss accounted for 87.6 percent of the presenting symptoms. The tumor was located at the lower thorax of 42.9% of the patients. Majority (62%) of the patients were from the eastern and north-eastern parts of the Arsi zone. Overwhelming majorities of the patients were from the Robe, Jaju, Disksis and Sude districts. Conclusion: Dysphagia with weight loss was the major symptom during rst visit. Squamous cell carcinoma was the predominant histopathological type. Eastern and the northeastern parts of the Zone are the most affected regions. Future studies should focus on the determinants of EC and precise areas with high incidence by using prospective studies through the population and/or facility-based cancer registry systems


Introduction
Esophageal cancer (EC) is the malignant tumor that develops inside the esophagus as the result of alteration in epithelial linings(1) It is 7th cancer in incidences and the 6th causes of cancer mortality in Projection of cancer in Africa anticipated staggering increases in cancer corresponding to 1.27 million new cases and 0.97 million deaths in 2030 (11). The incidences of EC were rare until the 1930s in East, Central, and Southern Africa (12). With gradual increases over time, the eastern and southern Africa became the epicenter of esophageal carcinoma (5,(13)(14)(15). Concentrations of EC cases were notably observed in rural and Eastern Cape provinces of South Africa (16), western parts of Kenya (17,18), among Darfurian tribes in Sudan (19) and populations living in Kilimanjaro areas in Tanzania (20). Arsi Zone was one of the EC endemic areas in Africa`s esophageal cancer belt for years (21) and a constantly reported high incidence area of EC in Ethiopia (22)(23)(24).
Identifying a cluster of cases of EC within the endemic area highlights high-risk populations and helps in generating hypothesis about the potential risk factors associated with that disease [23][24][25]. Yet, except few observational studies at a country and regional levels (23,28,29), no study has ever been conducted to assess the clinico-pathological patterns and spatial distributions of esophageal cancer in esophageal cancer endemic areas in Ethiopia.
Therefore, this study was conducted to determine the clinico-pathological patterns and spatial distributions of esophageal cancer in Arsi Zone, Oromia, Central Ethiopia.

Study setting and period
Arsi Zone is found in the central part of the Oromia Regional State in central Ethiopia. Asella is the capital city located 175 Kilo meters South East of Addis Ababa, the capital city of Ethiopia. A cross sectional study design was carried out by retrospectively abstracting data from the patients` records treated at multiple health facilities over ve years period. Data were collected from seven referral hospitals and four clinics which render diagnostic services, consultations, and cancer treatments in Asella, Adama, and Addis Ababa (additional le1). All EC patients from Arsi Zone who visited the selected health facilities during the study period were considered as source population. The study populations constituted radiographically diagnosed and pathologically con rmed EC patients from Arsi Zone. Endoscopy and pathologically con rmed esophageal cancer patients with complete data were included. Records of eight ve patients were excluded from the study because of incomplete demographic and clinical data. The data collection period covers from January 2015 to January 2019.

Data collection tool
Nine trained BSc Nurses have participated in the data collection process. The questionnaires (checklists) were developed by the research team and tested to make the required adjustments and familiarize data collectors with data collection tools. The data were retrieved from patients` registration books and charts from the surgical, oncology, endoscopy, and pathology clinics. Data were collected on demographic characteristics (age, sex and residence), clinical, histopathological type of cancer, tumor location, degree of tumor differentiation, cancer stages and the primary residences of the patientsThe severity of dysphagia was graded as follows: grade 1: normal swallowing; grade 2: di culty in swallowing solids; grade 3: di culty in swallowing semisolids; grade 4: di culty in swallowing liquids; grade 5: di culty in swallowing own saliva. The clinical staging was performed using TNM (American Joint Committee on Cancer (AJCC) cancer staging manual); this is a staging system that is an expression of the anatomical extent of the disease based on the extent of the primary tumor (T), absence or presence of and extent of regional lymph node metastasis (N) and absence or presence of distant metastasis (M) (30). The administrative units, geographical locations, and agro-climatic divisions were gathered from the Arsi Zone nance and economic development o ce while the topologies of the study area were collected by using Geographical Information System (GIS) obtained from the agriculture departments of Arsi Zone (additional le1 2).

Data processing and analysis
Questionnaires were checked daily for accuracy, consistency, and completeness by supervisors. Furthermore, the supervisor and the principal investigator gave feedback and correction on daily bases. Patients were categorized into highland, midland, and lowland agro-climatic divisions (Supplementary le 3). The information related to administrative units (Districts) was extracted from the map of Arsi Zone with the scale ratio of 1:1200. The extracted demographic information on spatial locations was geotagged to the database containing administrative districts from where patients came.
The categorical variables of the cases were analyzed by descriptive statistical methods and presented in the form of tables and gures. Continuous variables were summarized using means and standard deviations. The association between categorical variables was determined using the Chi-square (χ2) test. P-values <0.05 were declared as signi cant. Statistical data analysis was done using SPSS software version 21.0 (SPSS, Inc, Chicago, IL, USA).
Categorical attributes of the number of males and females were identi ed and the coordinates of the point were taken using Geographical positioning System (GPS) at the district level. The number of EC patients per district was used for delineating and mapping the distributions of EC per district in the Arsi Zone. In order to analyze the data spatially, CSI, 2013 Ethio-GIS administrative data were used as input data to automatically mark out and map the burden of cancer by using the Arc-GIS 10.1 software. Finally, different types of colors were used to indicate the burden of EC cancer per district based on the following category: Red color for the districts which had 31 and above patients, Pink color for the districts which had 21 to 30 patients yellow color for the districts which had 11 to 20 patients and Green color for the districts which have 1 to 10 patients.

Ethics approval and consent to participate
Ethical clearance was obtained from the Institutions Research Board (IRB) of Jimma University and approval of the research was obtained from Research Ethical Committee (REC) and administrations of the selected health facilities. Personally identi able information was not taken from patients records to keep anonymity.

Patients` characteristics
Out of 403 patients` whose cards reviewed, 388(96.3%) of them were rural residents. Two hundred nine (51.8%) of the patients were females. The male to female ratio was 0.93:1.0 with no statistically signi cant difference. The majority of patients (72.2% females vs. 50.5% males) were below the age of 60 years. The overall mean age was 54.5(SD ± 11.6) years. The mean age of female patients was 52.2(SD ± 10.41) years which ranging from 25 to 85 years. The corresponding mean age for males was 56.94(SD ± 12.27) years with the age ranging from 19 to 88 years (Table 1).  Tijo) contributed one to 10 patients (Fig. 1).

Spatial distribution of oesophageal cancer
In this study, 141 (35%) of the patients were from the eastern parts of the zone. The northeast, southeast, and northern parts of the zone had a total of 108 (26.8%), 47(11.7%) and 31(7.7%) of patients, respectively. Additionally, the zonal capital and surrounding districts (Asella &Tiyo) contributed to a total of 30(7.4%) patients while 27(6.7%) and 17(4.2%) of the cases were found to be the residents of the western and southwest parts, in that order. Out of the total cases, 178(44.2%), 137(34.0%) and 88(21.8%) of the patients were from the districts located in the highland, midland and lowland agro-climatic divisions, respectively (Fig. 2).

Discussion
In this study, we attempted to review the clinico-pathological patterns and spatial distributions of esophageal cancer in Arsi Zone. Accordingly, vast proportions of the patients were rural residents similar to the preceding studies that reported predominance of EC in rural areas (16,28,31,32). A high proportion of EC among rural residents may be attributed to exposure to various environmental and occupational carcinogens (33)(34)(35). Furthermore, Patel K. et.al. [2013] argued that the predominance of EC among rural populations is linked to infection from fungi and bacteria that replicate in grass-thatched houses (35). Moreover, inadequate nutrient intake and lower socioeconomic status could be another risk factors associated with increased incidences of EC among the rural populations (36,37).
In the present study, 7.1% of patients were ≤ 39 years of age and the youngest male and female patients were diagnosed at the age of 19 and 25 years, respectively. This is contradictory to the pathogenesis of EC that usually increases with age and attaining a climax after 70 years (1). Some studies in Africa also found increasing in the proportion of young patients as early as 12 years of age (38,39). The proportion of youngest patients in this study is greater than the ndings in Ethiopia (29) and Kenya (39). Observing EC among younger populations may indicate early childhood exposure to the carcinogens that may alter the natural pathogenesis of the disease.
Dysphagia with weight loss accounted for 87.6 percent of the presenting symptoms. Besides, the majority of the patients were at advanced stages of cancer. Dysphagia and late presentations became the cardinal presentations of EC patients in Ethiopia (23,28,29), Africa (16,33,35,(40)(41)(42) and elsewhere (43)(44)(45)(46). The reason could be because of the fact that esophagus has a remarkable capability to stretch until signi cantly blocked by the tumor (47).In addition, presenting at a late stage of cancer with dysphagia could be linked to failure to understand the early symptoms associated with esophageal carcinoma.
Squamous cell carcinoma was the dominant (98.3%) histo-pathological type similar to previous studies in Ethiopia (23,28,29,48). The nding is very conceivable since squamous cell carcinoma is the leading histo-pathological type of EC in developing countries (1,49,50).
The lower thorax was the commonest anatomical site for a tumor followed by mid thorax and GEJ with no signi cant differences between male and female patients. The nding is similar with the studies that found lower thorax as the commonest anatomical sites of esophageal cancer (40,48) but contrary to the studies that identi ed middle thorax as major cancer location in Ethiopia (23,28), Uganda (40) The majority of the patients in this study had a moderately differentiated tumor differently to studies that found well-differentiated tumors among EC patients [23,38]. The disparities might be because of lower inconclusive results in this study (11.7%) than the rate of undetermined pathological results (62.7%) in the aforementioned studies.
The geographical distribution of EC in the Arsi Zone showed asymmetrical patterns. About 62% of the patients were from the eastern and north-eastern parts. The majority of the patients were from the Robe district followed by Jaju, Disksis and Sude districts. Additionally, Doddota, Zuway-dugda,Lode-hetosa, and Shirka were succeeding districts in an increased number of cases. Besides, 44.2% of the cases were from the highland agro-climatic zone. The nding is in agreement with the study that reported high altitude areas as the main location of EC in Arsi Zone (42). The predominance of EC in those areas seems to be related to the population sizes because the eastern, north-eastern, and highland agro-ecological zone are the most populated areas in Arsi Zone (53,54). In strati ed analysis though, the effect of population sizes was not observed. For instance, Robe district which had the highest number of cases encompasses equivalent population size (6.2% of the zonal population) to Munessa district that had only 4.2% of the total cases. Whereas, Jaju and Diksis districts have lower populations than Munessa district. Conversely, the 6th populated Digalu-tijo district contributed to a single esophageal cancer patient.
Remarkably, greater numbers of patients were from contiguous districts and located farther from the main referral hospital in the Zone. The clustering of EC at demarcated areas and the disparities between and within geographical locations are the peculiar characteristics of EC in Africa, Asia, and elsewhere (10,21,55,56). High prevalence of EC in geographically adjacent areas may signal the existence of potentially harmful factors at high burden areas than areas with lower prevalence of esophageal cancer. Nevertheless, there are no established evidences whether the routinely reported "hot wheat porridge consumption" is widely practiced in higher burden districts of Arsi Zone. Hence, similar to multiple studies (57-62), the discrepancies in the geographical distribution of EC in the Arsi Zone may be primarily linked to differences in dietary practices, lifestyles, genetics and exposure to environmental carcinogens.
In the wake of increasing chronic diseases in developing countries like Ethiopia(63), the identi ed clinicpathological ndings and high-risk geographical areas can be used as the baseline for exploring the potential risk factors and designing preventive strategies.

Limitations
Being the retrospective study and absence of proper cancer registry systems in uenced retrieving of important socio-demographic and clinical information which might provide better perspectives to our study. Besides, identi cation of the patients s exact agro-climatic areas due to existing overlapping in agro-ecological zones within the districts.

Conclusion
Dysphagia with weight loss was the major symptom during rst visit. Squamous cell carcinoma was the predominant histopathological type. Eastern and the northeastern parts of the Zone are the most affected regions. Education on early symptoms associated with EC and bene t of timely seeking health care should be provided using various education channels. Future studies should focus on the determinants of EC and precise areas with high incidence by using prospective studies through the population and/or facility-based cancer registry systems.

Declarations
Ethics approval and consent to participate Ethical permission was obtained from the Institutions Research Board (IRB) of Jimma University and approval of the research activity was sought from each health facility administration o ces before the study was carried out. Con dentiality of the data was maintained by avoiding names and other personal identi ers known or written in the questionnaires.

Consent for publication
Not applicable Availability of data and material All data generated or analysed during this study are included in this published article [and its supplementary information les]. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.  Figure 1 Total cases of esophageal cancer per districts in Arsi Zone, central Ethiopia