Cancer occurrence in Ardabil: Results of a population-based Cancer Registry from Iran



The provincial health authority reported a high mortality rate from upper GI cancer in the newly established Ardabil Province of northwest Iran. A comprehensive search was undertaken to survey and register all cases of cancer during a 4-year (1996–1999) period among the indigenous population of Ardabil Province, including subjects seeking care in the cities of Tabriz and Tehran. Diagnosis of cancer was based on histopathology in 71.4%, clinical or radiologic findings in 25% and death certificate in 3.6% of cases. A total of 3,455 cancers (mean age 57.1 ± 17.3 years) was found during the study. Of these, 60% (2,072) were in males. ASRs for all cancers in males and females were 132.0 and 96.3, respectively. The top 5 cancers in males (excluding skin cancer) according to the calculated ASR were stomach (25.4), esophagus (15.4), lung and bronchus (7.9), colon and rectum (7.9) and bladder (7.6); in females, these were stomach (25.42), esophagus (14.4), breast (7.6), colon and rectum (5. 9) and lung and bronchus (3.6). Compared to rates obtained 30 years ago, the incidence of upper GI cancer in this region has increased about 100%, and there is a striking increase in the incidence of gastric cancer with a decline in the esophageal cancer rate. ASRs for gastric cancer in Ardabil were 49.1 for males and 25.4 for females, while for esophageal cancer these were 15.4 and 14.4, respectively. The ASR for cervical cancer was the lowest (0.4) recorded in the world before. Gastric cancer alone constitutes one-third of all cancers in Ardabil, the ASR of which is the highest reported from Iran up to now and one of the highest in the world. © 2003 Wiley-Liss, Inc.

For more than 40 years, several groups have tried to map out cancer incidence rates in different areas of Iran.1, 2, 3, 4, 5 Among these, only the Caspian Cancer Registry in the city of Babol, which was established in 1969 by joint collaboration of the IPHR of Tehran University and the IARC, has provided a reliable source of data on cancer incidence in the Caspian littoral of Iran.6 Although the study area was limited to the Caspian coastline [2 provinces of Mazandaran and Gilan and the district of Ardabil (now a province)], data published by this registry alarmed the local and national health authorities, by documenting one of the highest incidence rates of esophageal cancer in the world to be in the northeast of the Caspian littoral, the Turkman plain.7, 8 This led to an international effort to search for the risk factors of this disease in this region.9, 10 However, these efforts were discontinued due to the sociopolitical events of the 1980s in Iran. According to recent statistics issued by the Ministry of Health, cancer is the third most common known cause of death in Iran, after cardiovascular diseases and accidents.11 In recent years, there has been an increasing demand for reliable cancer incidence data by national health officials and researchers in the field of cancer.12, 13

We selected Ardabil Province for 2 main reasons: first, Ardabil was reported to have the highest incidence of upper GI cancer in Iran11 and, second, the cancer registry data of this region from the 1970s were available for comparison. In addition, Ardabil has a homogeneous Azeri ethnic population with a minimum immigration rate, which provides an excellent set-up for field studies, and established cancer registries. We also aimed to obtain cancer incidence data from the whole province, to select the target populations and hot spots for our future upper GI cancer field studies. The project began with active surveillance for all cancer cases and the establishment of a population-based cancer registry in this province.


ASR, age-standardized rate; DCO, death certificate only; DDRC, Digestive Disease Research Center; GI, gastrointestinal; ICD-O, International Classification of Disease for Oncology; IPHR, Institute of Public Health Research; LE, life expectancy; MV, microscopic verification; TUMS, Tehran University of Medical Sciences.

Geographic and demographic description

Ardabil was previously a district of East Azerbaijan Province but became a separate province 5 years ago. It is located in northwest Iran, an area 50 km inland from the western Caspian coastline (Fig. 1), with an area of about 17,953 km2 (1.09% of the total area of Iran). The capital city of Ardabil is at 38°-14′-32″ latitude N and 48°-17′-53″ longitude E. The province is mountainous with an altitude ranging from 4 to 4,811 m above sea level. Ardabil city is near the silent volcano of Sabalan, with an altitude of 4,811 m. The weather is moderate, with cold winters and mild summers. The total population of Ardabil Province, according to the 1999 census, is 1,128,864 (1.94% of Iran's total population), with 46% living in 9 cities and 54% in rural areas. The population is homogeneous, 95% being from the Azeri ethnic background, which is of Aryan Caucasoid ancestry. They speak in Azeri, which is a variant of the Turkish language, and are Shi'a Muslims. Consanguinity is common in both urban and rural areas. Sixty-three percent of the population is younger than 25 (mean age 23, median 18 years) (Fig. 2). LE at birth is 60 years for men and 61 for women. Sixty-six percent of men are farmers, and 90% of women are housewives, who occasionally work on the farms as well. The rest of the population are workers in local industries, in business or government employees. Based on the 1999 health statistics, the physician-to-patient ratio is 1/2,830; there are 10 hospitals with 974 beds, 14 outpatient clinics, 230 private physician offices, 49 clinical diagnostic laboratories, 5 pathology laboratories and 18 diagnostic radiology clinics.14 During the period of study, 1 oncologist, 1 gastroenterologist, 3 urologists, 20 gynecologists and 25 internists had the facilities and ability to precisely diagnose cancers and to provide appropriate chemotherapy. Most types of surgery were performed in Ardabil city, but there was no facility for performing radiotherapy; therefore, some patients were referred to Tabriz and Tehran for radiotherapy or for other speciality cares.14 This report describes the data derived from the Ardabil Cancer Registry, the first population-based cancer registry in the Islamic Republic of Iran.

Figure 1.

Map of Iran and Ardabil Province.

Figure 2.

Population cascade of Ardabil Province.


The Ardabil Cancer Registry office is located in Ardabil, the capital city of Ardabil Province, with one full-time physician. This registry is run by Ardabil University of Medical Sciences and supervised by the DDRC, TUMS. The Medical Ethics Committee of the DDRC approved the survey methods. The registry unit at the DDRC consists of 1 gastroenterologist, 2 epidemiologists and 2 pathologists; the survey team at Ardabil consists of 5 general practitioners and 10 medical students. The survey team had been trained to go to hospitals, pathology laboratories, diagnostic radiology clinics, outpatient public and private clinics and check records for cancer cases and, whenever possible, to make a copy of the documents, according to which a diagnosis of cancer was made. These were then sent to the registry office in Ardabil and later to the registry unit in the DDRC.

During a period of 6 months (January–July 2001), the survey team actively collected and compiled data for a period of 4 years (1996–1999) from 10/10 hospitals, 10/14 outpatient clinics, 30/150 private offices, 5/5 pathology laboratories, 10/18 radiology clinics and the central death registry office in Ardabil city. All pathology laboratories of the province (all located in Ardabil city) were included, their data collected and a copy of the pathology report made and sent to the registry office.

Most private radiology centers recorded the identifying data of subjects and only occasionally kept the radiologic or sonographic reports, while the public centers, usually located in the hospitals, kept radiology reports for the majority of patients. All private and public hospitals and clinics have a filing system for endoscopy reports.

Mortality data were evaluated in 2 ways: death certificate and annual health census. Physicians in all 9 districts issued death certificates in duplicate, one for themselves and one to be kept by graveyard authorities. The physicians sent their copies to the Deputy of Drug and Treatment in Ardabil city to be compiled later. Death certificates from urban areas (for all 9 cities) were obtained from the office of medical documents in the Deputy of Drug and Treatment.

The Deputy of Health, covering the rural area of the whole province, conducted annual health censuses, using behvarzes (auxiliary health-care workers in health houses who are responsible for public health care and the health census of each village) at the beginning of each year through direct, home-to-home data collection, including data on the number of new births and deaths and causes of death. The results of these censuses, including information on the cause of death in rural areas, are compiled in the Deputy of Health, where the final annual death statistics are produced. Data about rural cancer cases that appear to be more comprehensive than the data in urban areas were obtained from the Census Office in the Deputy of Health. The survey team enjoyed the close collaboration of all health authorities and physicians in Ardabil, who provided all necessary data and documents for the study.

A minority of Ardabil Province residents seek medical care outside the province, mainly in Tabriz and Tehran. A thorough search was done for the records of Ardabil residents with a diagnosis of cancer in 5 major hospitals and 3 major pathology laboratories in Tabriz. The hospital-based cancer registry of the Cancer Institute, Tehran, currently provides the most reliable and comprehensive data in this large city. All incident cancer cases from Ardabil, recorded in Tabriz or the Cancer Institute of Tehran, were also included for the same time period (1996–1999).

The data were summarized in a data sheet and coded using the ICD-O.15 After data collection was complete, all data were alphabetically organized and duplicate cases with the same name, sex, age and place of residence were eliminated by manual and computerized linkage. Each alphabetical group was assessed manually by 2 individuals on 2 different occasions.

The data were computerized using SPSS (Chicago, IL) software, version 10.0, and MS EXCEL (Microsoft, Redmond, WA) software with Persian fonts. We calculated person-years of the population at risk using each year method. The results are presented as incidence rates of cases by site (ICD-O), sex, age, crude, age-specific rates (Figs. 3 and 4) and ASRs per 100,000 person-years, using the direct method of standardization to the world population.16

Figure 3.

Age-specific incidence of the major cancers in males.

Figure 4.

Age-specific incidence of the major cancers in females.


Over a 4-year period (1996–1999), 3,455 new cancer cases were registered in Ardabil Province, 2,072 men (60%), 1,309 women (37.9%) and 74 with undefined gender in registry documents (2.1%). The mean (±SD) age at the time of first diagnosis was 57.1 ± 17.3 years (59.1 ± 16.7 for males and 53.0 ± 17.8 for females).

Tables I and II show the principal cancer sites, 10-year age-specific incidence rates, total number of cancer cases, relative frequencies as well as crude rates and ASRs for both males and females. Diagnosis of cancer was based on histopathology in 71.4% (MV percentage shown in Table III), clinical or radiology in 25% and DCO in 3.6% of cases. The top 5 cancers in males according to the calculated ASRs were stomach (49.1), esophagus (15.4), lung and bronchus (7.9), colon and rectum (7.9) and bladder (7.6); in women, these were stomach (25.4), esophagus (14.4), breast (7.6), colon and rectum (5.9) and lung and bronchus (3.6). The ASR for all cancers in males was 132, and that for females was 96.3. The ASR for cervical cancer (0.4) was very low. The age-specific incidence of the major cancers in male and female are depicted in Figures 3 and 4.

Table I. Age-Specific Incidence Rates, Average Annual Crude Incidence Rates and ASR in Males in Ardabil, 1996–1999
Site0–1415–2425–3435–4445–5455–6465All ages% TotalCrude rateASR world
Oral cavity00.
Colon and rectum00.
Gallbladder, etc.0000.
Bronchus and lung00.41.31.511.729.856.21215.95.17.9
Connective tissue00.20.60.500090.20.20.2
Skin melanoma000001.81.840.20.20.3
Other skin0.
Other male genital00000.80.9020.10.10.2
Kidney, etc.0.300.
Nervous system11.
Hodgkin's disease011.30.501.81.8140.70.60.7
Non-Hodgkin's lymphoma00.20.327.511.4831.81.62.6
Lymphoid leukemia00.20.311.73.51.8120.50.40.7
Myeloid leukemia0.
Other leukemia00.20.601.71.72.7100.50.40.6
All sites4.212.53159.4116.6528.1769.82,07210085.7132
Table II. Age-Specific Incidence Rates, Average Annual Crude Incidence Rates and ASR in Females in Ardabil, 1996–1999
Site0–1415–2425–3435–4445–5455–6465All ages% TotalCrude rateASR world
Oral cavity0.2001.
Colon and rectum000.9314.323.924.5745.73.35.9
Gallbladder, etc.
Bronchus and lung0002.15.412.424.6443.423.6
Connective tissue0.30.20.3020080.60.30.4
Skin melanoma020.3002.1040.30.20.2
Other skin0.
Female genital, other000002.
Kidney, etc.00.301.725.25.3171.30.81.3
Nervous system1.
Hodgkin's disease0.
Non-Hodgkin's lymphoma0.
Lymphoid leukemia0.10.20.332.741.1211.60.91.4
Myeloid leukemia0.
Other leukemia0.
All sites5.412.427.575.8177.2328.5470.71,30910058.196.3
Table III. Percentage of Cases Diagnosed with MV
SiteMV (%)TotalDCO
Colon and rectum63.51802.2
Bronchus and lung35.11624.2
Brain, nervous system60.41281.6
Hodgkin's disease92.327
Non-Hodgkin's lymphoma90.161


Studies on cancer epidemiology in developing countries such as Iran are limited by the dearth of tools for disease control and surveillance. The first report on cancer incidence in Iran dates back to the 1970s, a study of cancer in the Caspian littoral region between 1968 and 1972.1 This study showed that the ASRs of all cancers in Ardabil for men and women were 71.9 and 45.0, respectively, and the most common cancer was that of the esophagus, with decreasing incidence from the east to the west of the littoral. Comparing the rates reported 30 years ago,1, 7, 8 the incidence of upper GI cancers has increased about 100% in this region. The most plausible explanation for this increase might be the underreporting of the 1970–1971 study.

Ardabil, a newly established small province with a homogeneous (98%) Azeri ethnic population, has a very low rate of migration in and out of the province, very cooperative and enthusiastic health authorities and a comprehensive rural health infrastructure, which made it a suitable choice for screening and prevention trails.

The well-trained survey team had access to all patient records, with close monitoring of expert pathologists and epidemiologists at the registry branch of the DDRC in Tehran. These data are the most complete and accurate estimate of cancer occurrence that have been reported up to now from Iran. The survey team obtained a copy of pathologic, imaging and endoscopic documents in the majority of cases; and the epidemiologist and pathologists were able to recheck the information, specifically to exclude benign or suspicious cases and to reclassify the cancers as necessary.

One of the shortcomings of our study was its retrospective nature in the setting of poor-quality medical records kept in private and, to some extent, in public hospitals and other medical centers. Also, not using the ICD coding system in some of the medical centers at the time of diagnosis made it impossible to classify all cancer cases from the exact topographic point of view. The alphabetical organization of data we used to eliminate cases with similar names may have resulted in 5–10% of doubles. Despite these shortcomings, we did our best to find all new cancer patients, but we do not know exactly how successful we were; most likely, we have underestimated some tumors. Therefore, our results should be taken as the minimum incidence rate of cancer in Ardabil.

Our results demonstrate that upper GI cancers alone account for >42% of all cancers in this region and that the GI tract as a whole constitutes >50% of all cancers in Ardabil, with gastric cancer being the most common malignancy in the province (31%). This represents the highest reported incidence of gastric cancer in the country, which puts Ardabil among one of the highest incidence areas of gastric cancer in the world. Gastric and esophageal cancers constituted about 30% and 12%, respectively, of all cancers in our study. These relative frequencies are strikingly different from those reported 30 years ago,7 which were 8% and 56% for the stomach and esophagus, respectively. These figures are also very different from the relative frequencies reported from south Iran 30 years ago,17 which were 5.5% and 2.3% for stomach and esophageal cancer, respectively.

The striking difference from cancer registry data obtained 30 years ago is in part due to the differences between survey techniques and the size of the population studied; e.g., during the 1970s registry, the diagnosis of esophageal cancer was confirmed histologically in about 27% of cases,1, 7 while during our study diagnosis was based on histologic reports in 75.2% of cases (Table III). It is likely, therefore, that the results obtained then reflected an overestimation of the true diagnosis of esophageal cancer. Improved diagnostic facilities, e.g., the availability of endoscopy for diagnosis and exact localization of upper GI cancer specifically distinguishing cardia cancer from distal esophageal cancer, may be another reason for the decrease in esophageal cancer and the rise in gastric cancer found in our study. Another explanation may be change in lifestyle, shifting from the rural toward the urban, which has taken place over the past 30 years.

According to a survey carried out by the Ministry of Health,14 at least 50% of the people in this region are overweight and at least 70% now have access to fresh meat, fruits and vegetables. Approximately 50% of gastric cancers in Ardabil are adenocarcinomas of the cardia,18 and Helicobacter pylori infection is present in close to 90% of the adult population; this infection appears to occur early in life, with >50% of children infected before age 15.19 The high frequency of endoscopic reflux esophagitis (36.7%) among the asymptomatic Ardabil endoscopic screening participants20 correlated with 60% of them being overweight, with a body mass index >25 kg/m;2;2 that endoscopic survey also revealed that cardia inflammation (carditis) existed in 75% of asymptomatic volunteers.20

Ardabil Province has some geographic similarities to other high-incidence areas of gastric cancer, so-called volcanic countries like Costa Rica, Chile and Japan,21 by being located near the silent volcano of Sabalan. Therefore, a similar environmental factor like nitrous volcanic soils may contribute to the higher incidence of upper GI cancer in this area.

Although prostate cancer is one of the most common cancers among males in Western countries,22 it ranks ninth in Ardabil; this may be partly due to the fact that in the rural areas people older than 65 years do not receive a full diagnostic workup when they are sick, and a significant number of deaths in older people are being attributed to old age in the rural health census. A very interesting finding in this survey is the surprisingly low incidence of uterine cervical cancer, a rate lower than any reported in the world. We rechecked this very low rate of cervical cancer several times, and there is no doubt that despite the presence of 20 gynecologists (19 female and 1 male) and regular periodic Pap smear screening tests by both rural and urban health centers, this type of cancer is very uncommon. The reason for this finding is not clear but may be partially the fact that Ardabil is one of the most religious cities in Iran, with >99% of residents being Shi'a Muslims, a very strong dependence on family-based traditions and almost no extramarital sexual relations, very rare sexually transmitted diseases among women and a very uncommon practice of cigarette smoking in females.23 The incidence of other gynecologic malignancies, not related to sexual behavior, is also very low in Ardabil; and the possibility of underdiagnosis or underreporting should be considered and addressed in the prospective cancer registry study of Ardabil in the future.

With regard to the top 10 cancers, there is a remarkable difference between the results of our study and the estimated cancer incidence for Iran by GLOBOCAN 2000 (Table IV).22, 24 The GLOBOCAN 2000 database uses a large amount of data, derived from cancer registries from different national populations or subsamples from selected regions. Therefore, the difference is probably due to the fact that the GLOBOCAN 2000 estimate is based on a hypothetical incidence for all cancers calculated from other registries in the region and the relative frequency of cancers in the hospital-based survey reported from other low-incidence areas of Iran5, 12 with imperfect data collection. Further investigation in other provinces is necessary for mapping out the real cancer incidence in Iran.

Table IV. Top 10 Cancers in Ardabil Compared to Estimated Cancer Incidence for Iran by GLOBOCAN 2000
Ardabil Cancer RegistryGLOBOCAN 2000
Colorectal5.047.86Colorectal3.335.89Colorectal16.419.1Lung and bronchus11.211.1
Nervous system3.144.44Nervous system2.253.13Esophagus9.110.8Ovary, etc.6.46.5
Oral cavity2.503.75Liver1.302.28Bladder8.510.0Corpus uteri6.36.4
Non-Hodgkin's lymphoma1.572.60Oral cavity0.991.71Non-Hodgkin's lymphoma5.56.1Esophagus4.44.5
Bone1.441.69Lymph leukemia0.911.43Larynx4.75.5Non-Hodgkin's lymphoma4.04.0


We thank Dr. D.M. Parkin (IARC, Lyon, France) for reading the manuscript and offering very helpful comments. We are also grateful to Dr. Gooya from the Center for Disease Control in the Ministry of Health, medical students on the survey team and colleagues who contributed to this survey, particularly all local health officials, behvarzes at health houses and health staff in the treatment and diagnostic centers in Ardabil Province.