Introduction

Cancer is the main source of mortality in the age group of 40 to 70 years and the subsequent driving reason for death in all age groups, representing a fourth of all passings worldwide (Siegel et al. 2016). As per worldwide reports, the quantity of death because of cancer in 2012 was 14.1 million, and the quantity of deaths from cancer is projected to be 22.2 million by 2030 (Bray et al. 2012). According to a nationwide cancer registry in Iran, cancer is the third leading cause of death, with more than 30,000 people dying of cancer each year. Also, according to 2010 surveys, the highest incidence of cancers in Iran was reported in gastric, esophageal, breast, prostate, and colon cancers, respectively (Kolahdoozan et al. 2010; Saadat et al. 2015). In light of a public cancer registry, cancer is the third driving reason for death in Iran, and over 30,000 individuals die of cancer every year. Likewise, gastric, esophageal, breast, prostate, and colon cancer were the most well-known cancer types in 2010 (Khlifi et al. 2013). Gastric cancer is the most common cancer in males and third in females (Fauci 2011). Then again, colorectal cancer is the third and fourth most normal cancer in people internationally and has, as of late, been perceived as a significant wellbeing concern around the world. Colorectal cancer accounts for a significant share of cancer mortality and disability (Parkin et al. 2005; Siegel et al. 2013). The incidence of gastrointestinal cancers has been steadily increasing in recent years. Gastrointestinal cancer's risk factors are diet, socioeconomic conditions, toxic substances, and environmental contaminants through food chains.

Heavy metals are among the most serious pollutants in the environment that play an essential role in human diseases. Increased amounts of heavy metals as a result of natural processes (weathering and biodegradation in nature) and human-made (industrial activities, agriculture, vehicles) in the environment and cause various clinical effects in human and animals (Mansouri et al. 2012a, 2012b; Rezaei et al. 2021; Farnia et al. 2021). Heavy metals in two groups of essential metals (including iron (Fe), zinc (Zn), and copper (Cu)) and non-essential metals (such as cadmium (Cd) and lead (Pb)) for biological activities can have human health effects (Norouzi et al. 2012; Mansouri et al. 2012c; Majnoni et al. 2013; Maleki et al. 2015). Metals such as Cd and Pb as potential carcinogens are associated with some diseases, including cardiovascular diseases, kidney, nervous system, hypothyroidism, and thyroid cancer (Solenkova et al. 2014; Rezaei et al. 2019, 2021). Metals like Fe, Cu, and Zn are essential minerals that play an important role in the body's metabolic and physiological activities. Still, their concentrations should not exceed the permissible limit, otherwise damaging the target tissues. When Fe and nickel (Ni) exceed the body’s amount, it reacts with peroxides to produce free radicals. The free radicals can cause DNA, lipids, proteins, and tissue carbohydrates damage that eventually lead to oxidative stress (Al Faris and Ahmad 2011; Nanda and Agrawal 2016). Sohrabi et al. (2018) stated that heavy metal concentrations as environmental pollutants play an important role in developing gastrointestinal cancers; they also found that the concentration of metals in patients with gastrointestinal cancer was significantly different from control samples. Ionescu et al. (2006) also found significant Ni concentrations in breast cancer tissue compared to healthy samples.

Some studies have been carried out on heavy metals’ effect on gastrointestinal cancer samples (Roy et al. 2004; Pasha et al. 2010; Rinaldi et al. 2015). Still, insufficient attention has been paid to existing research on the amount and impact of heavy metals on Iran’s gastrointestinal cancer incidence. Also, given the increasing trend of gastrointestinal cancers in Iran and especially in the eastern part of the country in recent decades, such studies are needed. Therefore, in this study, we conducted a case-control study to assess the association between heavy metals including cadmium, chromium, cobalt, copper, iron, lead, nickel, selenium, zinc (Cd, Cr, Co, Cu, Fe, Pb, Ni, Se, and Zn) and arsenic (As) in cancerous and non-cancerous tissues with the risk of gastrointestinal cancers in Eastern Iran.

Materials and methods

Study population

The present study is a case-control study. The data collected in this study include the measurement of heavy metals in gastric cancer biopsy specimens in South Khorasan province, eastern Iran. Biopsy specimens of stomach (n = 20), esophageal (n = 19), and colorectal (n = 24) cancers along with controls (n = 63) were collected from patients recruited to Birjand Imam Reza Hospital from January to December 2019.

This study’s protocol was approved by the Ethics Committee in Birjand University of Medical Sciences (IR.BUMS.REC.1397.072). The patients for this study were aged between 30 and 70 with confirmed gastric cancer. All cases were newly diagnosed and previously untreated. Clinical characteristics, including basic medical data, were obtained from medical records. Patients in the case group were diagnosed by specialist physicians based on inclusion criteria. Then, the consent form was given to the patients in the research project. If included in the study, the demographic checklist (age, gender, type of cancer, smoking, and anthropometric measurements) of the subjects was completed by the questionnaire. Cases and controls had no distinction between geographic or cultural groups since they were the native aborigines in South Khorasan province. Both cases and controls had no occupational exposure to heavy metals. The inclusion criteria for the gastrointestinal cancer patients group were history, physical exam, and an endoscopic study indicating gastrointestinal cancers. In the control group, inclusion criteria were not indicating gastrointestinal cancer or intestinal diseases. The exclusion criteria were patients’ reluctance, history of non-gastrointestinal cancers, malignancies, metabolic or nutritional disorders, vitamin supplementation use, hormones, chelators, and receiving chemotherapy or radiation.

Biopsy samples

The biopsy specimens were then put away in formalin solution with the endorsement of a gastroenterologist and pathologist. Specimens were thought to be cancerous in the control group, yet these tissues were resolved healthy after pathological evaluation. These samples were kept in formalin solution at Imam Reza hospital in Birjand city and selected as a control sample like cancerous tissues.

Determination of heavy metals

We applied the nitric-perchloric acid digestion method to prepare samples for measuring cadmium. Approximately 50 mg of each sample was mixed with 2 mL of ultrapure (HNO3, 65%, Merck, Germany) and 1 mL of perchloric acid (HClO4, 70%, Merck, Germany), and the samples were left overnight to be slowly digested. Then, the mixture was placed in a water bath (Bain Marie; TW12, Julabo GmbH, Germany) at 100°C for approximately 6 h or until the solutions were clear. After cooling, the mixture was made up to 25 mL by adding double-distilled water (18.2 MΩ-cm at 25 °C, Fistreem, WSC044, UK). The prepared solution was measured for Cd, Cr, Co, Cu, Fe, Pb, Ni, Se, Zn, and As using Inductively Coupled Plasma Mass Spectrometry (ICP-MS; Agilent 7900). The operation parameters for ICP-MS were as follows: Radiofrequency power: 1.5 kW, Plasma gas flow rate: 15 L/min., Carrier gas flow: 1.01 L/min., Make up gas: 0.15 L/min., Sample uptake rate: 1.7, Sample depth: 10 mm, Detector mode: auto, Scan type: peak hopping with three sweeps per reading and three readings per replicate, and scan number: 3.

Statistical analysis

Results were presented as mean ± SD for numerical variables and number (percentage) for categorical variables. A t-test or Chi-squared test was used to evaluate the differences between subjects' characteristics. Since the metal concentrations were highly skewed, to make the analysis more robust against extreme values, the concentrations were standardized using the mean and interquartile range of natural log-transformed concentrations. If the normality assumption was violated, a t-test or Wilcoxon rank-sum test was used to compare the concentrations between cancer and healthy group.

The logistic regression model was used to assess the effect of single and multiple metal exposures of heavy metals on cancer risk using unadjusted and adjusted risk differences (RDs) and respective 95% confidence intervals. The outcome was a binary indicator of participants in the case-control group. RDs were adjusted for age, Body mass index (BMI), smoking status, education concentration, and family history of cancer in first-degree relatives. Risk differences were obtained using single metal exposure (unadjusted), using single metal and covariates (adjusted), and using covariates and multiple metals (adjusted). Furthermore, to deal with the perfect separation issue in the logistic model, a Bayesian model with non-informative prior assumptions proposed by Gelman et al. (2008) was used. Moreover, generalized additive models (GAM) analysis was performed under the logistic regression model to investigate if the effect of the trace elements is present, it is a nonlinear effect or not. As we have mentioned in the Results section, such an effect was not confirmed by GAM analysis. All statistical analyses were performed using R, the open-source language, and computing software.

Results

Characteristic of the study population

Of 126 participants with 63 cases (24 esophagus, 19 colon, 20 stomach) and 63 controls (17 esophagus, 19 colon, 27 stomach), 64 (51%) were men and 62 (49%) were women. The mean age of participants was 59.95 ± 17.41 (18 to 92 years). Moreover, only eight percent of participants had an academic degree, whereas 43% were uneducated, 42% less than five years, and 16% completed high school. Compared to controls, cases were significantly older, less likely to be overweight (57% with BMI < 25), more likely to have a lower level of education (Table 1). Regarding gender, smoking, and family history of cancer among first-degree relatives, cases and controls were comparable. The comparison of participants’ characteristics revealed no statistically significant difference in the distribution of gender (Chi-squared test, P = 0.856), BMI (Chi-squared test, P = 0.364), the likelihood to be ever smoker (Chi-squared test, P = 0.611), and the family history of cancer (Chi-squared test, P = 0.676) between the cases and controls (Table 1). However, cases were significantly older than controls (t-test, P = 0.002).

Table 1 Characteristic of the study population
Full size table

Age effects on gastrointestinal cancers

For each cancer type, Table 2 shows the mean age as well as the distribution of education levels for participants in each studied group. The age differences were not statistically significant between cases and controls suffering from the esophagus and stomach cancer. For colon cancer, cases (with a mean age of 64.71 years) were markedly older than controls (with a mean age of 51.65 years) (t-test, P = 0.013). Comparison between education levels showed a significant association between education levels and studied groups for colon cancer (Fisher test, P = 0.0007) and esophagus cancer (Fisher test, P = 0.0006).

Table 2 Comparison of age and education of participants between three types of cancer
Full size table

A comparison of element concentration levels was made between cases and controls to investigate the effect of age (Table 3). There was no difference between Fe, Se, and Cd concentrations in cancerous and control tissues, but compared to controls, elevated levels of Cr, Co, Ni, Zn, and Pb were observed in cancerous patients. The levels of Cu and As were found to be higher in controls (Table 4).

Table 3 Comparison of trace element levels (μg kg−1) between colon cancer cases and their controls using the Mann-Whitney test
Full size table
Table 4 Concentrations of trace element levels (μg kg−1) among cases and controls for three types of cancer
Full size table

Mann-Whitney test showed that the concentrations of Cr (p < 0.001), Fe (P = 0.021), Co (P = 0.039), Ni (P = 0.0004), Zn (P = 0.034), Se (P = 0.003), and Pb (P = 0.004) were significantly higher among esophagus cancer cases than controls (Table 3). Compared to controls, the concentration concentrations of Cr (P = 0.001), Co (P = 0.015), Ni (p < 0.001), Zn (P = 0.002), and Pb (P = 0.003) were markedly high among cases with colon cancer, whereas they had lower concentrations of Cu (P = 0.022) and As (P = 0.007). Similarly, cases with stomach cancer showed higher concentrations of Cr (P = 0.003), Co (P = 0.030), Ni (p < 0.001), Zn (P = 0.002), and Se (P = 0.013) and lower concentration concentrations of Cu (p < 0.001) when compared to their controls (Table 3).

Correlation among trace elements

The correlation between metals with each other is evaluated using spearman’s correlation coefficient. The results are presented in Fig. 1. As it can be seen, correlation coefficients between element pairs were mostly positive and low to moderate (less than 0.50). A few exceptions with correlation coefficient over 0.50 and highly significant at 0.001 level were Co-Cd pair (esophagus r = 0.86, colon r = 0.80, stomach r = 0.79), Co-Pb (esophagus r = 0.72, colon r = 0.82, stomach r = 0.66), Cd-Pb (esophagus r = 0.66, colon r = 0.68, stomach r = 0.63), Cr-Ni (esophagus r = 0.70, colon r = 0.64, stomach r = 0.66). Moreover, Cu was negatively correlated with other elements for both colon and stomach cancer types (Fig. 1).

Fig. 1
figure 1

Correlation plot between trace elements in studied cancer types. The number represents Spearman's correlation coefficients. a: colon; b: esophagus, c: stomach

Full size image

Relationships between elements and gastrointestinal cancers

The results of Cr, Ni, Zn, Se, and Pb metals showed significant risk differences on esophagus cancer when considered as the single predictor unadjusted for other metals and covariates RD (95% CI) – Cr: −0.274 (−0.463, −0.086), Ni: −0.288 (−0.457, −0.118), Zn: −0.171 (−0.463, −0.086), Se: −0.243 (−0.434, −0.051), and Pb: −0.094 (−0.143, −0.045) (Table 5). The metal significance remained unchanged when adjustment was applied for confounders and other metals. The negative sign of the RD denotes excessive risk of cases when metal concentrations increased. For instance, RD = −0.274 for Cr shows that when the concentrations of Cr increased, on average, the excessive risk of esophagus cancer was 27.4 % and it was significant (CI does not include zero). Similarly, the unadjusted RD of Cr, Ni, Cu, Zn, and As metals were significant for colon and Cr, Ni, Cu, Zn, Se, and Pb metals for stomach cancer. We also assessed the nonlinear relationship of metals using GAM analysis and found no evidence of nonlinear associations.

Table 5 Logistic regression analysis of trace elements among cases and controls for three types of cancer
Full size table

Discussion

While researchers keep on finding the primary cause, the etiology of a few cancers is as yet indistinct. Numerous studies reported the association of heavy metals like Cd, Cu, Ni, Pb, Zn, and As in the etiology of many cancers, including breast and prostate cancers (Marouf 2018). The International Agency for Research on Cancer introduced As, Cd, Cr, and Ni as category 1 heavy metals (IARC 1980). These components are found in nature, and exposure to them can inhibit tumor suppressor genes, damage repair processes, and enzymatic activities concerned with metabolism via oxidative damage (Bánfalvi 2011; Kim et al. 2015). We found that the concentrations of Co, Cr, Fe, Ni, Pb, Se, and Zn were significantly higher among esophagus cancer cases than controls. Additionally, the Co, Cr, Ni, Pb, and Zn concentrations were markedly higher among colon cancer cases than controls.

Essential trace elements

In our study, the high concentration of copper (Cu) in cancerous tissues of the stomach and esophagus is consistent with findings from previous studies (Sun et al. 2011; Lin et al. 2020). Also, unlike stomach and esophagus cancers, Cu concentrations in cancerous tissues of colon cancer and the control group were not significantly different (p > 0.05). Cu is an essential component of various biochemical functions. Its concentration is increased in cancer tissues to promote cancer development through angiogenesis processes (Wang et al. 2010). High concentrations of Cu were detected in the liver and spleen of patients with the respiratory system, urinary tract, genital tract, breast, esophagus, bronchus, and gastrointestinal cancers (Magalhães et al. 2010; Szewczyk et al. 2013). On the contrary, Reddy et al. (2003) reported lower Cu concentrations in kidney cancer patients than non-cancerous tissues. Zinc (Zn) is a mineral element involved in many homeostatic mechanisms of the body, including specific immunity, inflammation, and oxidative stress (John et al. 2010). Cancer cells have upregulated zinc importers and frequently increased zinc concentrations, which allow them to survive. Previous in vivo and in vitro studies have shown Zn’s role in carcinogenesis (Morton 2016). However, studies that evaluated the role of zinc in malignant diseases have a long history of contradiction. According to this study's results, Zn’s mean concentrations in cancerous tissue of patients with gastric, esophageal, and colon cancer tissues were significantly higher than the control group (p < 0.05). Klimczak et al. (2016) showed similar results in breast and colon cancers. Oppositely, Yaman et al. (2007) demonstrated higher Zn concentrations in normal tissues than the cancerous gastric tissues (P = 0.005).

Selenium (Se) has various health and disease roles. It can function as a toxin, a carcinogen, and a cancer chemopreventive agent in mammals (Vinceti et al. 2018). Early studies provided evidence that Se has a role in cancer prevention. Decades of studies specifically prove that Se compounds inhibit the development of malignant cells in different laboratory model systems. However, growth-modulating and cytotoxic pathways are complex and far from clear (Wallenberg et al. 2014). However, none of the Se compounds have been scientifically accepted as anticancer drugs, partially because researchers have achieved contradictory findings over the years (Tan et al. 2018). Data on possible associations between selenium and esophageal and gastric cancer are sparse, and even less is known about the subtypes of these cancers. Some studies reported that high Se levels might be inversely associated with gastric cancer (GC) risk and GC mortality (Gong 2016; Charalabopoulos et al. 2009). Our findings indicated that the concentrations of Se were significantly higher among esophagus cancer cases than controls (p < 0.05). Similarly, cases with stomach cancer showed higher Se levels compared to their controls (p < 0.05).

Non-essential trace elements

In our study, arsenic concentration was notably elevated in cancerous tissue patients with gastric, esophageal, and colon cancer, consistent with other reports (Yang et al. 2008; Lee et al. 2016; Safarzad et al. 2019). The inorganic form of arsenic (As) is highly toxic and considered group 1 (Straif et al. 2009). The exact association between AS and cancer is still undiscovered. Still, it has been shown that AS can interfere with several biological processes, such as chromosomal damage, DNA methylation, and DNA repair, further producing ROS (Raju et al. 2017). We also showed a high concentration of cadmium (Cd) in cancer tissues. Sayadi et al. (2020) illustrated that the Cd and Cr concentrations in herbs in rural residents living in Eastern Iran were relatively high. Risk assessments revealed that consuming these herbs may affect the health of both children and adults. Cigarette smoking is one of the most important sources of exposure to Cd (Ashraf 2012). Prolonged exposure to Cd leads to cardiovascular disease and cancer. Cd’s concentration in cancerous tissue of patients with gastric, esophageal, and colon cancers was significantly higher than the control group in the present study. Kellen et al. (2007) reported a higher concentration of Cd in the blood of patients with bladder cancer than the control group. Strumylaite et al. (2011) evaluated Cd’s concentration in breast tissue, urine, and blood of 57 breast cancer and 51 benign tumor patients. The results showed a higher Cd concentration in breast tumors and urine of cancer patients. They supported a possible relationship between Cd and breast cancer.

Meta-analyses and original studies have proposed that there may be excess risks for lung and stomach cancers among industrial workers with lead (Pb) exposed for a long time. However, stomach cancer was associated with organic Pb (Rousseau et al. 2007). In our study, Pb concentration in cancerous tissue of patients with esophageal, gastric, and colon cancers was significantly higher than normal tissues in the control group (p < 0.05). Some evidence showed that the level of Pb and Cr contamination of drinking water in Birjand flood plain was higher than international guidelines given by the USEPA and WHO, and that may explain the higher level of both elements in cancerous tissues in our study (Mansouri et al. 2012d; Sinkakarimi et al. 2020). Moreover, the findings of Rezaei et al. (2019) reported that toxic metals such as Pb, Cd, and Cr could increase the risk of developing hypothyroidism and thyroid cancer in patients with thyroid diseases living in Birjand City, east of Iran.

Chromium (Cr) is the seventh most common element on Earth. Cr can produce reactive oxygen species in successive oxidation states. As a result, it will be able to damage DNA, causing mutations and eventually cellular changes. Chromium is known to be carcinogenic (Sohrabi et al. 2018). According to the present study results, the mean concentrations of Cr in cancerous tissue of patients with gastric, esophageal, and colon cancers significantly increased compared to healthy tissues (p < 0.05). Sohrabi et al. (2018) earned the same result. In their study, the median of Cr in colorectal cancer (CRC) tissues was significantly higher than in healthy tissues. This finding is also in agreement with the meta-analysis results performed by Deng et al. (2019). They showed that male workers exposed to Cr were at risk of prostate cancer, oral cancer, and to a greater extent, stomach cancer. Another meta-analysis performed by Welling et al. (2015) suggested Cr is a stomach carcinogen in humans. Iron (Fe) is an important element in cell proliferation and oxidative activities. While existing research on the carcinogenic function of iron deficiency and over-sufficiency is mixed, there is general support for these possibilities. Both iron deficiency and iron over-sufficiency may raise the risk of gastrointestinal tract cancers such as stomach and CRC cancers (Hossein Davoodi et al. 2016). This increased risk can be explained by biochemical modification and inflammatory status. The difference in the metabolism of iron occurs as an initiator (Verma and Cherayil 2017). In the present study, iron concentrations in cancerous tissues of patients with stomach and esophageal cancers were significantly lower than non-cancerous tissues of healthy controls (p < 0.05), which is consistent with the findings in similar studies (Nelson et al. 1994; Pusatcioglu et al. 2014). On the contrary, in a study conducted by Yaman et al. (2007), the iron concentrations in tissue samples of patients with gastric cancer were significantly higher than non-cancerous tissues.

Nickel (Ni) is a heavy metal widely distributed in several forms. Epidemiological studies suggest that occupational exposures to nickel compounds are associated with an elevated prevalence of nasal and lung cancers (Lu et al. 2005). Ni carbonyl inhibits cortisone-induced tryptophan pyrrolase and phenothiazine-induced benzopyrene hydroxylase and protein synthesis and DNA synthesis by interfering with DNA-dependent RNA polymerase (Pavela et al. 2017). Khlifi et al. (2013) showed Ni concentrations in tumor head and neck cancer tissues were significantly higher than the non-cancerous tissues in the control group (p < 0.05). Additionally, Pasha et al. (2010) reported a similar result in gastrointestinal cancer. In the present study, the mean concentrations of Ni in the cancerous tissue of patients with gastric, esophageal, and CRC cancers were significantly higher than normal tissues in the control group, which is consistent with the findings in similar studies (p < 0.05). On the contrary, Kohzadi et al. (2017) reported that concentrations in non-cancerous tissue in the control group were significantly higher than cancer tissue in cancer patients (p < 0.05).

Age effects on cancers

Cancer development is a long process that takes several years to complete and can be considered an age-related disease because the incidence of most cancers increases with age, rising more rapidly beginning in midlife (White et al. 2014). According to official reports and studies, more than half of cancers occur in individuals older than 70. Several mechanisms include the role of genomic instability, telomere attrition, epigenetic changes, loss of proteostasis, decreased nutrient sensing and altered metabolism, cellular senescence, and stem cell function can share or diverge aging and cancer development (Aunan et al. 2017). The fundamental reasons for the observed changes are still unknown; however, several aging hypotheses are based on an increase in oxidative stress and concomitant accumulation of oxidized and nitrated proteins, oxidized lipids, and DNA damage. According to in vivo studies, essential trace elements are important during the aging process because they modulate both oxidative stress and immune response by their essential functions, such as enzymatic reactions and signaling pathways. Inadequate trace element consumption is common in the elderlies, resulting in a lower amount, especially serum Se and Zn concentrations (Lossow et al. 2020). However, a low-grade chronic inflammation, common in the elderlies, may play a role in the age-related changes in trace element profiles (Wong et al. 2013). There was no significant association in terms of age between cases and controls suffering from the esophagus and stomach cancer in our study. However, for colon cancer, cases (with a mean age of 64.71 years) were markedly older than controls (with a mean age of 51.65 years) (P = 0.013) (Table 2)

Correlation among trace elements

The Mann-Whitney test was used to compare the element concentration levels between cases and controls to investigate the effect of age (Table 3). The result showed that the concentrations of Cr, Fe, Co, Ni, Zn, Se, and Pb were significantly higher among esophagus cancer cases than controls. Additionally, the concentrations of Cr, Co, Ni, Zn, and Pb were markedly high among cases with colon cancer compared to controls, whereas they had lower concentrations of Cu and As. Similarly, higher concentrations of Cr, Cu, Ni, Zn, and Se and lower concentration concentrations of Cu were detected in cases with stomach cancer compared to controls. In addition, Spearman’s correlation coefficient was used to assess the correlation between metals with each other. The result showed that correlation coefficients between element pairs were mostly positive and low to moderate (less than 0.50).

Conclusion

There is now a considerable amount of epidemiologic evidence concerning trace elements' role in influencing cancer risk. This study aimed to evaluate the association between heavy metals including Cr, Fe, Co, Ni, Cu, Zn, As, Se, Cd, and Pb in cancerous and non-cancerous tissues with the risk of gastrointestinal cancers in Eastern Iran. Based on our result, the concentrations of Cr, Ni, As, Cd, Zn, and Pb in the cancerous tissue of patients with gastric, esophageal, and colorectal cancers were significantly higher than the healthy tissue of non-cancerous patients. Also, we found that the concentration of Fe and Cu in the healthy tissue of non-cancerous individuals was higher than in the cancerous tissues. Although the association between trace element exposure and cancer risk has been examined in some extensive prospective studies, additional studies are needed to elucidate further the mechanisms underlying trace element carcinogenesis.