1 INTRODUCTION

Small intestine (SI) cancers are recognized as the rare types of worldwide malignancies. The four main histological subtypes of SI cancers include adenocarcinomas, neuroendocrine tumors, sarcomas, and lymphomas.1 It is estimated that the incidence rate of SI cancer would experience a considerable growth in high-income countries such as the United States, the United Kingdom, and France. SI consists of three different areas including the duodenum, jejunum, and ileum. One of the most common sites for tumor growth is the duodenum where a large amount of digestive enzymes of a human body is released. In previous studies, it is reported that from 55% to 88% of SI cancers occur in this smallest area of SI while jejunum figures 11%–25% of all these kinds of cancer and between 7% and 17% of the cases belongs to the ileum.2-4

According to recent epidemiological studies, the incidence rate of SI cancer was 2.3 per 100 000 person-years in the United States in 2018. In addition, SI cancer was placed at 23rd in the ranking of most common cancers in this country.1 Researchers demonstrated that the incidence rate of SI among men was about 2.6 while this rate was 2.0 among women. Although the incidence rate of this neoplasm has gradually increased, the mortality rate has remained unchanged. Despite the fact that, deaths from SI cancer have increased in the United Kingdom, its rate of increase is less than the rate of increase in the incidence of this cancer, indicating an increase in survivals.1, 5

The 5-year survival of SI cancer varies among various countries across the world. Although this rate was 67.6% in the United States, the average 5-year survival of SI patients in European countries was considerably lower than that of the United States. The histology type of tumor has a pivotal role in the 5-year survival of SI neoplasm. As the results of recent researches indicated, neuroendocrine cancers and carcinomas had the highest and lowest survival rate respectively. Several risk factors have been identified for SI cancer, including race and ethnicity, age, sex, stage, subsite, diet, alcohol, and obesity.1, 6

A few studies have been conducted to evaluate the incidence, risk factors, and survival analysis of this fatal worldwide malignancy. However, to the best of the authors' knowledge, there is no comprehensive study investigating the epidemiology of SI cancer in Iran. The aim of the present study was to conduct a population-based study to determine the incidence pattern of SI cancer in Iran to enable health authorities, clinical experts, and policymakers to provide more targeted and earlier diagnosis and better care.

2 METHODS 2.1 Study population

For the current study data on patients with SI neoplasms were analyzed, which had been reported to the Iran National Cancer Registry (INCR), between March 20, 2005 and March 20, 2015. This is a national, mandatory registration program to collect information on clinical procedures, pathological examinations, and causes of death on cancer cases in the population. All health centers including hospitals, clinics, and pathology laboratories were required to report patient information to the INCR. Being a time interval between cancer data collection and recording these data in patient's medical records made it is not possible for us to use the updated data during the course of the study. For that reason, the latest available national data were used in this study.

The data used potentially faced with some problems; problems such as multiple recorded data for some patients. In addition, incorrect morphology may have been recorded for a topography related to SI cancer. Moreover, the diagnostic method of SI cancer may have been incorrect or invalid. Furthermore, the patients' age and date of birth did not match in some cases. Therefore, the quality of data needs to be evaluated.

The extraction and clean-up of the analyzed dataset by the INCR was done in several steps. In the first step, the consistency of the topographical and morphological information of the patients were evaluated. If there was any inconsistency, the data would be re-examined and if the data was inaccurate, the subject would be omitted from the study. Subsequently, the type of patients' tumor was assessed to see whether it was diagnosed based on the correct procedure used for SI or not. In the next step, the accuracy of patients' birthdays and the date of cancer diagnosis were checked; the incorrect information was modified or removed from the dataset. The pathology (or cytology), clinical outcome, and death certificate only were applied to identify the cancer diagnosis. Duplicate cases were identified through checking their first name and surname, sex, and father's name. Patients with the exactly matched records were assigned to duplicate records and were removed automatically.

2.2 Data variables

In the current study, patients' information regarding the topography and histology of SI was classified based on the third edition (first revision) of the International Classification of Diseases for Oncology (ICD-O3). All types of the SI cancers (ICD-O topography codes C17-0-C17.9), with malignant behavior (ICD-O behavior code/3) were presented in four various categories including carcinomas, neuroendocrine cancers, sarcomas, and lymphomas.7 Other information such as sex, age, birth date, city of residence, phone number, and date of diagnosis were also registered during the data collection used in this study. Through national census data performed by the Statistical Center of Iran, the incidence rate of this cancer in different age groups was calculated while using Iran's total population in 2006, 2011, and 2016. Not having access to information regarding the population of other years caused these figures to be estimated through growth rate between two consecutive censuses. Therefore, each year's population was calculated by multiplying the population of the previous year in the growth rate.

Twenty percent of the patients whose medical record were registered in national census data were randomly selected from all over the country and were interviewed by telephone. In this telephone interview, the type of cancer recorded in the data was checked and verified. Besides, their survival status, death status due to SIs, and patient's death date were questioned and recorded.

2.3 Statistical analysis Crude incidence rates (per 100 000 person-years) and 95% confidence intervals (CIs) were calculated for total cases and each age group (age-specific incidence rate) as follows: in which is a standardized normal deviate, r refers to the numbered of cases occurred in the ith age class and n refers to the person-years of observation in the age class during the same period of time as cases were counted. The age-standardized incidence rates (ASIRs) per 100 000 and 95% CI were computed based on new World Health Organization (WHO) standard population (2000–2025) in which these estimations are regarded as the weights in the standardization method. These formulas can be expressed as8, 9: in which and where Zα/2 is a standardized normal deviate, ai is the crude incidence rate in each age category (age-specific incidence rate), and wi is the new WHO standard population. By applying a direct method, the 95% CI standardized rate ratio () were obtained10 as follows: in which

The average annual percent change (AAPC) and 95% CI were calculated by applying Joinpoint software to summarize the ASIR trend in the study.11 The ASIRs, SRRs, and AAPCs were assessed according to patients' histology information and various subsites such as duodenum (C17.0), jejunum (C17.1), ileum (C17.2), Meckel's diverticulum (C17.3), overlapping sites (C17.8), and unspecified site (C17.9).

Survival information of patients with malignant behavior included all samples recorded by telephone interview, from 2005 to 2015 and patients were followed up till 2020. The 1- and 5-year survival rate were computed through employing the Kaplan–Meier approach by sex, age, calendar year, and cancer's subsites. The log-rank test was used to compare survival distribution among variable levels.12 The overall survival rate with 95% CI was obtained by Kaplan–Meier survival curve.13 The cause-specific survival method as an alternative to relative-survival methods was applied for estimating cancer survival, since the suitable life-tables were not available.14, 15 The current study was approved by the Ethics Committee of Shahid Beheshti University of Medical Sciences (IR.SBMU.CRC.REC.1399.004).

3 RESULTS

In total, 4928 SI cancers were diagnosed during 2005–2015, after removing patients with wrong histology information (n = 80) and duplicated data (n = 128) (Table S1). From March 20, 2005 to March 20, 2010 and from March 21, 2010 to March 20, 2015, 1826 and 3102 cases of SI cancer had been reported, respectively. The highest frequency of SI cancers was observed in the age group of 50–80. The overall age-specific incidence rate was higher than 1.0 per 100 000 person-years in patients over 50 and the highest age-specific incidence rate was observed in age group of 80–84 years with a value of 7.21 (6.47–7.95) per 100 000 person-years. In the current study, the overall ASIR (95% CI) was 0.90 (0.85–0.90) per 100 000 person-years, while the ASIR in males was 1.03 (0.99–1.07) and in females was 0.71 (0.68–0.74) per 100 000 person-years (Table 1 and Table S2, Figure 1).

TABLE 1. The 5-year annual frequency (age-specific incidence rate per 100 000 person-year) of small intestine cancer, Iran, 2005–2015 Age group 2005–2010 2010–2015 2005–2015 Total Male Female Total Male Female Total Male Female 0–4 17 (0.06) 12 (0.08) 5 (0.04) 40 (0.13) 24 (0.15) 16 (0.10) 57 (0.09) 36 (0.12) 21 (0.07) 5–9 25 (0.09) 18 (0.13) 7 (0.05) 17 (0.06) 13 (0.09) 4 (0.03) 42 (0.07) 31 (0.11) 11 (0.04) 10–14 15 (0.05) 10 (0.06) 5 (0.03) 13 (0.05) 11 (0.08) 2 (0.01) 28 (0.05) 21 (0.07) 7 (0.02) 0 15–19 27 (0.07) 19 (0.09) 8 (0.04) 23 (0.07) 15 (0.09) 8 (0.05) 50 (0.07) 34 (0.09) 16 (0.04) 20–24 38 (0.09) 27 (0.12) 11 (0.05) 54 (0.14) 39 (0.20) 15 (0.08) 92 (0.11) 66 (0.16) 26 (0.06) 25–29 65 (0.17) 32 (0.17) 33 (0.18) 52 (0.12) 28 (0.13) 24 (0.11) 117 (0.15) 60 (0.15) 57 (0.14) 30–34 71 (0.24) 44 (0.30) 27 (0.19) 91 (0.25) 62 (0.34) 29 (0.16) 162 (0.25) 106 (0.32) 56 (0.17) 35–39 77 (0.31) 47 (0.36) 30 (0.24) 96 (0.33) 58 (0.39) 38 (0.26) 173 (0.32) 105 (0.38) 68 (0.25) 40–44 97 (0.46) 58 (0.54) 39 (0.37) 164 (0.65) 100 (0.79) 64 (0.52) 261 (0.56) 158 (0.67) 103 (0.45) 45–49 145 (0.80) 84 (0.91) 61 (0.68) 223 (1.06) 127 (1.20) 96 (0.92) 368 (0.94) 211 (1.07) 157 (0.81) 50–54 191 (1.32) 122 (1.67) 69 (0.96) 316 (1.76) 176 (1.96) 140 (1.57) 507 (1.56) 298 (1.83) 209 (1.29) 55–59 165 (1.62) 95 (1.90) 70 (1.35) 297 (2.13) 182 (2.63) 115 (1.63) 462 (1.91) 277 (2.32) 185 (1.51) 60–64 202 (2.62) 100 (2.65) 102 (2.60) 329 (3.30) 193 (4.04) 136 (2.62) 531 (3.00) 293 (3.42) 238 (2.61) 65–69 163 (2.66) 99 (3.16) 64 (2.14) 300 (4.22) 158 (4.65) 142 (3.84) 463 (3.50) 257 (3.93) 206 (3.08) 70–74 209 (3.73) 140 (4.74) 69 (2.61) 340 (6.00) 213 (7.54) 127 (4.47) 549 (4.87) 353 (6.11) 196 (3.57) 75–79 175 (4.76) 111 (5.66) 64 (3.73) 337 (7.51) 203 (8.70) 134 (6.22) 512 (6.27) 314 (7.31) 198 (5.12) 80–84 108 (5.09) 73 (6.69) 35 (3.40) 259 (8.72) 153 (10.14 106 (7.26) 367 (7.21) 226 (8.69) 141 (5.66) 85+ 36 (2.57) 17 (2.34) 19 (2.82) 151 (7.80) 89 (8.93) 62 (6.60) 187 (5.61) 106 (6.16) 81 (5.02) Total 1826 (0.70, 0.67–0.73)a 1108 (0.83, 0.78–0.88) 718 (0.57, 0.52–0.61) 3102 (1.01, 0.98–1.05) 1844 (1.20, 1.14–1.25) 1258 (0.82, 0.78–0.87) 4928 (0.87, 0.85–0.90) 2952 (1.03 0.99–1.07) 1976 (0.71, 0.68–0.74) a ASIRs, 95% CI and ASIRs are age-standardized incidence rates to the new WHO standard population (per 100 000 person-year).

The age-specific incidence rate (per 100 000 person-year) of small intestine cancer, Iran, 2005–2015

The percentage of carcinomas (n = 2835, ASIR = 0.51), unspecified malignancies (n = 934, ASIR = 0.17), lymphomas (n = 704, IR = 0.11), sarcomas (n = 228, ASIR = 0.04), and neuroendocrine cancers (n = 214, ASIR = 0.04) accounted for approximately 57.5%, 19.0%, 14.3%, 4.6%, and 4.3% in all cases of SI cancers, respectively. Eighty-one percentage of all kind of carcinomas were adenocarcinomas, not otherwise specified (NOS) (n = 2306, ASIR = 0.42). According to this study results, 52% of neuroendocrine cancers were carcinoid, NOS (n = 111, ASIR = 0.02); 67% of sarcomas were gastrointestinal stromal tumors (GIST) (n = 153, ASIR = 0.03). Besides, among all lymphoma cases, 37% were non-Hodgkin lymphoma (NHL) and lymphoma-NOS (n = 260, ASIR = 0.04) (Table 2 and Table S3).

TABLE 2. The frequency, ASIR (per 100 000 person-year), SRR, and AAPC of patients with small intestine cancers based on the main groups of the ICD-O-3, Iran, 2005–2015 ICD-O-3 group No. of patients (ASIR) SRR (95% CI) ASIR (95% CI) SRR (95% CI) AAPC (95% CI) Total Male Female Male to female 2005–2010 2010–2015 2010–2015 to 2005–2010 Total 4928 (0.87) 2952 (1.03) 1976 (0.71) 1.45 (1.37–1.54) 0.70 (0.67–0.73) 1.01 (0.98–1.05) 1.44 (1.36–1.53) 9.6d (5.7–13.7) Carcinomas 2835 (0.51) 1619 (0.58) 1149 (0.42) 1.37 (1.27–1.48) 0.47 (0.45–0.50) 0.55 (0.52–0.57) 1.15 (1.07–1.24) 7.2d (1.9–12.8) Neuroendocrine cancers 214 (0.04) 111 (0.04) 102 (0.04) 1.11 (0.85–1.46) 0.02 (0.02–0.03) 0.05 (0.04–0.06) 2.19 (1.65–2.92) 17.8d (5.8–31.2) Sarcomas 228 (0.04) 111 (0.04) 114 (0.04) 0.94 (0.72–1.23) 0.03 (0.03–0.04) 0.04 (0.04–0.05) 1.25 (0.96–1.63) 3.6 (−11.9–21.9) Lymphomas 704 (0.11) 440 (0.14) 238 (0.07) 1.85 (1.57–2.17) 0.12 (0.11–0.13) 0.10 (0.09–0.11) 0.81 (0.70–0.95) −1.4 (−6.5–4.0) Other specified malignancies 13 (<0.01) 14 (<0.01) 6 (<0.01) 2.56 (0.73–8.95) <0.01 <0.01 — — Unspecified malignancies 934 (0.17) 557d (0.20) 370 (0.14) 1.47 (1.29–1.68) 0.04 (0.04–0.05) 0.27 (0.25–0.29) 6.09 (5.17–7.17) 60.9d (27.5–103.1) a Age-standardized incidence rate to the new WHO standard population. b Standardized rate ratio. c Changes in trends among each line segment and overall lines were calculated using annual percent change (APC) and average annual percent change (AAPC), However, only AAPC was reported. d AAPC is significantly different from zero at the level of 0.05.

In the case of carcinomas and lymphomas, the ASIR in males were 0.58 and 0.14 respectively, per 100 000 person-years, being 0.16 and 0.07 higher than females, respectively. However, the ASIR in neuroendocrine cancers and sarcomas were equal in males and females. In total, the ASIR experienced a significant increase during 2010–2015 in comparison to that of 2005–2010 (SRR = 1.44, 95% CI: 1.36–1.53). Over the course of this study, this increasing trend was observed at 9.6% (95% CI: 5.70–13.70) per year from 2005 to 2015. The upward trend of carcinomas and neuroendocrine cancers was illustrated at 7.20% and 17.80% per year during the course of the study, respectively. On the flip side, this measure in sarcomas and lymphomas similar to other major histological subtypes did not fluctuate significantly (Table 2). The ASIR of SI cancer had a rising trend despite some fluctuations (Figure 2).

The ASIR (per 100 000 person-year) of small intestine cancer, Iran, 2005–2015. ASIR, age-standardized incidence rates

Carcinomas' frequency was about seven times higher in the duodenum than in the jejunum and ileum. Moreover, the incidence of lymphoma and neuroendocrine cancers was higher in the duodenum than in the jejunum. The ASIR of carcinomas and lymphomas in the jejunum was significantly lower than the ASIR of this histology subtype in the duodenum. Similar results were observed for carcinomas and neuroendocrine cancers in the ileum compared to the duodenum. The ASIR of neuroendocrine cancer and sarcomas in the duodenum, jejunum, and ileum were not higher than 0.02 per 100 000 person-years (Table 3).

TABLE 3. ASIR (per 100 000 person-year) and SRR of small intestinal cancer according to subsite, Iran, 2005–2015 Subsite Carcinomas Neuroendocrine cancers Sarcomas Lymphomas Index No. ASIR SRR (95% CI) No. ASIR SRR (95% CI) No. ASIR SRR (95% CI) No. ASIR SRR (95% CI) Duodenum 1221 0.23 (Reference) 67 0.01 (Reference) 21 <0.01 (Reference) 105 0.02 (Reference) Jejunum 137 0.02 0.10 (0.09–0.12) 12 <0.01 — 31 0.01 — 49 0.01 0.44 (0.31–0.61) Ileum 194 0.03 0.15 (0.13–0.17) 46 0.01 0.66 (0.45–0.96) 22 <0.01 — 122 0.02 1.12 (0.86–1.46) Meckel diverticulum 4 <0.01 — 2 <0.01 — 1 <0.01 — 0 — — Overlapping lesion of SI 18 <0.01 — 0 — — 1 <0.01 — 4 <0.01 — Small intestine, NOS 1261 0.23 1.00 (0.93–1.09) 87 0.02 1.31 (0.95–1.81) 152 0.03 — 424 0.06 3.77 (3.09–4.61) Abbreviations: NOS, not otherwise specified; SI, small intestine. a Age-standardized incidence rate to the new WHO standard population. b Standardized rate ratio.

The overall 1- and 5-year survival rate for all SI cancer were 62.4% and 35.3%, respectively. Among histological subtypes, the survival rates of patients with lymphomas, sarcomas, and neuroendocrine were moderate (more than 50.0%); however, the survival rate of carcinomas was the lowest (55.6 at 1 year and 24.1 at 5 years). Overall survival rate was significantly affected by gender (log-rank test: p-value <.05), while carcinomas patients had lower survival in males compared to females (log-rank p-values <.05). But other histology subtypes did not have significant different survival rates among males and females. The survival rate of carcinomas, sarcomas, and lymphomas patients who were younger than 60 was markedly higher than older patients. The 5-year survival rate was significantly improved in lymphomas over the second 5 years of the study period compared to the first one (log-rank test: p-value <.05). The 5-year survival rate of SI cancer located in the duodenum had the smallest percentage among carcinomas (21.0%) and sarcomas (25.0%). Moreover, the lowest survival rates of SI cancer occurred in the jejunum (35.0%) and ileum (39.5%) were carcinomas (Table 4). The Kaplan–Meier survival curve showed cumulative survival of patients about 0.3 after 11 years from the start of the study and remained stable until the end of the study (Figure 3A). In addition, the cumulative survival of patients revealed significant differences between different histology types (Figure 3B).