The Incidence of Esophageal Cancer in Iran: A Systematic Review and Meta-analysis

Background: Esophageal Cancer (EC) is the eighth most common cancer in terms of incidence. Despite various studies, there is no general estimate of EC incidence in Iran. Therefore, the present study is conducted to evaluate the incidence rates of EC in Iran. 
Method: A systematic search was conducted on all published studies of EC incidence using Medline/PubMed, Scopus, Web of sciences, Google scholar, and four Iranian databases (Scientific Information Database, MagIran, IranMedex, and IranDoc) until November 2017. This systematic review was done according to the preferred reporting items for systematic reviews and meta-analyses (PRISMA). 
Result: The database, grey literature searches, and hand searching yielded 346 potentially relevant studies. A total of 22 studies were included. The results of the random effect model were demonstrated the age-standardized rate (ASR) of EC was 25.05, 95% CI (20.84 to 29.26) among males and 22.93 95 % CI (18.97-26.88) among females. 
Conclusion: In comparison to other geographical locations, the incidence of EC is higher in Iran. However, organized system for collecting data of cancer is required to specify the incidence and trend of EC in Iran.


(a) Search strategy of systematic reviews
A literature search of published studies was conducted using international databases Medline/PubMed, Scopus, Web of sciences, and Google Scholar for English papers and Iranian databases Scientific Information Database (www.sid.ir), MagIran (www.magiran.com), IranMedex (www.iranmedex.com), and Irandoc (www.irandoc.ac.ir), for Persian papers.

(b) Inclusion and exclusion criteria
All studies with results of ASR of EC and reports of Iranian populations were included in this review. Furthermore, studies with following criteria were not considered in this review; studies which reported prevalence rate based on pathological data, studies with inadequate sample size, and research articles (all type of conference abstracts, poster papers, letters, comments, and editorial).

(c) Quality assessment
In order to assess the quality of the articles, a checklist prepared by The Joanna Briggs Institute (JBI) was used [21]. The purpose of this appraisal is to assess the methodological quality of a study and to determine the extent to which a study has addressed the possibility of bias in its design, conduct, and analysis.

(d) Risk of bias across studies
Random effect model was used for minimizing the risk of bias across the studies [22,23].
(e) Statistical analysis STATA version 12.0 software (Stata Corp LP, College Station, TX, USA) was used to perform all analysis. Statistical heterogeneity between the results of obtained studies was assessed using Cochran's Q statistic (with a significance level of p≤0.1) combined with I 2 statistic (with a significance level of>50%). The Meta-analysis was conducted with a random effect model (with inverse variance method) in the studies with significant heterogeneity (p≤0.1 and I 2 ≥50%). Additionally, in the absence of heterogeneity (p>0.1 and I 2 <50%), the fixed effect model was used.

(a) Description of literature search
The database, grey literature searches, and hand searching yielded 346 potentially relevant studies. In total, 274 unique studies were reviewed, and 91 studies were entered into the second stage of evaluation. Overall, our review included 22 unique studies. Study retrieval and selection has been outlined in Figure 1. Some studies were excluded from the review due to not being relevant to the topic (n=190), incorrect study population (n=35), duplicate study (n=6), and inadequate data (n=21). The flowchart of the included studies in this review has been shown in Figure 1.

(c) The results of individual studies
The results of the study showed the sex ASR ratio of male to female is 1.09. The highest ASR was reported from Golestan province between 1995 and 1997 (144.09 per 100,000) for men [38], and 1968 (174.1 per 100000) for women [43]. The lowest ASR was reported from Fars province between 1990 to 2005 (1.05 for males and 0.87 for females per 100,000) [28].    Figure 2 and Figure 3 for males and females, respectively. All measurements in the forest plot were multiplied by 10 5 because of the small values of ASRs.

(e) Publication Bias
Publication bias was assessed using Egger's tests [45]. Results of Egger's tests showed a lack of publication bias (p=0.197 for males and P= 0.442 for females).

Discussion
EC is the eighth most common cancer worldwide (3.8% of all cancers) and is the sixth most common cause of cancer-related deaths (5.4% of all cancer-related deaths). More than 80 percent of all EC deaths occur in developing countries [1,2]. Cancer is the third most common cause of death in Iran [46]. So far, few studies have been conducted on the epidemiology of cancer in developing countries such as Iran [47][48][49][50]. The first study on cancer incidence in Iran dates back to the 1970s. That study investigated cancer incidence in the Caspian littoral region between 1968 and 1972 [43].

Figure 3. Forest plot of the random-effect meta-analysis for ASRs of Esophageal cancer in females in the Iran
Iran is one of the countries with very high incidence of EC [8,26]. Some parts of Iran, which have common borders with Turkmenistan and Afghanistan countries, experience much higher incidence rates of this cancer. These regions include Mazandaran, Golestan, and Khorasan provinces [14,35].
Results of the present study indicate that the incidence rate of EC among Iranian men and women (ASR=25.05 and 22.93 per 100,000 in men and women, respectively) is higher compared to other countries in the region. However, some Asian countries such as Turkmenistan (ASR=24 for men and 16.4 for women per 100,000), Mongolia (21.2 for men and 14.9 for women per 100,000), and Tajikistan (19.8 for men and 10.7 for women per 100,000) have ASR. Also, countries such as United Arab Emirates (1.8 per 100 thousand), Qatar (.1.8 per 100 thousand) and Nepal (3.6 per 100 thousand) in men and South Korea (0.4 per 100 thousand), Vietnam (0.7 per 100 Thousand) and Thailand (0.8 per 100,000) in women have the lowest ASR [51].
Globally, the highest age standardized incidence rate of EC is observed in Eastern Asia (ASR=11 per 100,000), and Western Pacific (ASR=10.2 per 100,000). Also, Central America (ASR=1.1 per 100,000), and Western Africa (ASR=0.6 per 100,000) has the lowest incidence rates of EC [1].
According to the study that examined the correlation between the human development index (HDI) and the ASR of EC, a significant reverse relationship between HDI and the standardized incidence rate of EC was identified. Thus, countries with higher levels of HDI reported lower incidencerates of EC [1].The difference in incidence rates of EC among different countries can be due to the development, distribution, and prevalence of risk factors, as well as better application of cancer registration techniques and more accurate diagnosis of the disease [4,[52][53][54].
According to the results of this study, the highest ASR of EC in Iranian men and women is observed in Golestan province (ASR=144 per 100 thousand in men during 1995-7 and 174.1 per 100,000 in women in 1968). The high incidence of cancer in different regions is due to the difference in the distribution of the risk factors and its associated exposures in one region compared to another. Based on studies, risk factors such as drinking hot tea, burned opium consumption, gastric Helicobacter pylori infection, family history of esophageal cancer, drinking contaminated water, inappropriate diet, low physical activity, and smoking are associated with EC incidence [19,55]. The genetic role is also a major contributor to the high incidence of EC in Golestan province, having one of the highest worldwide rates of this type of cancer incidence [56][57][58].
The results of this study indicated that the lowest ASR of EC in Iran, in both sexes, is observed in Fars province (1.05 for men and 0.87 for women). The low incidence rate of EC in this province can be attributed to the demographic characteristics of people living in this area, differences in lifestyle, and the presence of other types of risk factors and diseases. In Fars province, other types of cancer such as breast, colorectal, and gastric cancers in women and bladder, prostate and gastric cancers in men have higher incidence rates [30,59,60].
Finally, it should be noted that the prevalence of cancer-related risk factors in Iran is high and is rising with an upward trend. The number of new cancer cases in Iran is expected to rise in future due to the epidemiological transition, increasing life expectancy, and aging the population.

Conclusion
In comparison to other geographical locations, the incidence of EC is higher in Iran. However, organized system for collecting data of cancer is required to specify the incidence and trend of EC in Iran.

Open Access
This article is distributed under the terms of the Creative Commons Attribution License (CCBY4.0) which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.

List of abbreviations
ASR: age-standardised rate; EC: Esophageal Cancer; NCR: National Cancer Registry System; PRISMA: Preferred reporting items for systematic reviews and meta-analyses 9. Ethics approval and consent to participate Not to be applied

Competing interests
The authors declare that they have no conflicts of interest.

Funding
The present study was financially supported by Guilan University of Medical Sciences, Rasht, Iran.

Authors' contributions
All authors contributed to the design of the research, HS, FMG, SM, FJ, EA, MF extracted the data and summarized it. HS, SH AND MAZ analyzed the data. All authors drafted the first version.

Acknowledgments
The present study was financially supported by Guilan University of Medical Sciences, Rasht, Iran.