A critical review on the relevance, essentiality, and analytical techniques of trace elements in human cancer

Abstract

Trace elements (TEs) are indispensable nutritional elements, playing a pivotal role in maintaining human health and serving as essential cofactors for numerous enzymes that facilitate crucial biological processes. The dysregulation (excess or deficiency) of TEs can affect the proper functioning of various organs and lead to diseases like cancer. However, the current research findings remain contentious, and the association between TE variations and cancer remains elusive. This article reviews the recent advances in the quantitative detection of TEs in tumor research to fully understand the important role of TEs in disease diagnosis and prognosis. The changes in the levels of various elements (such as Cu, Zn, Fe, Se, Ca, etc.) are analyzed and summarized from five systems of the human body, including the digestive system, urinary system, reproductive system, endocrine system, and respiratory system. By analyzing the relevant findings in diverse biological samples, we systematically investigate the disruption of TEs homeostasis in cancer patients, thereby underscoring the potential of TEs as cancer biomarkers. We also present novel analytical techniques such as isotope ratio determination and bioimaging, along with advanced auxiliary tools like machine learning, for the detection of TEs in disease research. This review aims to provide a comprehensive overview of TEs variations in the main cancer types of different systems, which addresses the knowledge gap in TEs on human health, and provides proposals for future research.

Graphical Abstract

Open in new tabDownload slide

Introduction

Trace elements (TEs) refer to elements that constitute less than 0.01% of the total mass of the human body. These elements can be categorized into three groups: essential elements, including iodine (I), zinc (Zn), selenium (Se), copper (Cu), molybdenum (Mo), chromium (Cr), cobalt (Co), and iron (Fe); potentially essential elements, including manganese (Mn), silicon (Si), nickel (Ni), boron (B), and vanadium (V), and elements that exhibit potential toxicity but have functional roles in the body at low doses, such as fluorine (F), lead (Pb), cadmium (Cd), mercury (Hg), arsenic (As), aluminum (Al), lithium (Li), and tin (Sn) [1, 2]. Although TEs are extremely low in abundance, they are necessary to maintain the normal metabolism of organisms. As the key cofactors of metalloproteins, certain TEs are involved in a variety of basic biological processes, such as enzyme catalysis, cell signal transduction, redox regulation, and immune response [3]. For instance, Cu/Zn superoxide dismutase (SOD) assembled by Zn and Cu enables the decomposition of superoxide radicals into H₂O₂, which plays an important role in antioxidant defense against oxidative stress. Selenoproteins carry Se in the form of selenocysteine (Sec), which can eliminate detrimental peroxides in tissues and safeguard biofilms and biomacromolecular structures against oxidative damage as an important antioxidant enzyme. In general, TEs have an optimal concentration range in the body; either insufficient or excessive could result in physiological abnormalities or diseases, and even cancer. For example, Fe is essential for oxygen utilization and DNA synthesis, while excessive Fe reacts with peroxides to produce free radicals that damage DNA, lipids, proteins, and carbohydrates in tissues, leading to oxidative stress [4]. Zinc deficiency can compromise the antioxidant capacity and affect the occurrence and development of tumors, as it serves as a crucial component in numerous enzymes and coenzymes in organisms [5].

Cancer is one of the leading cause of death and a significant impediment to enhancing life expectancy, which has become a major public health problem. According to GLOBOCAN data, there were approximately 19.3 million new cancer cases and nearly 10.0 million cancer deaths worldwide in 2020 [6]. It is estimated that 40% of cancer risk factors in China or other developed countries can be attributed to environmental and lifestyle factors, with smoking accounting for about 24.5% of cancers in men and chronic infections accounting for about 17% of cancers [7]. The burden of cancer-related diseases not only occupies more medical resources but also undoubtedly affects the functional status, mental health and quality of life of cancer patients and their relatives. Compared with the United States and the United Kingdom, the low early detection rate of cancer in different regions of China leads to higher mortality rates of cancer, which are 30% and 40% higher respectively in 2018 [7]. Prompt initiation of treatment necessitates the early detection and diagnosis of cancer.

Currently, liquid biopsy, a noninvasive tool, is widely applied for the diagnosis and conformation of cancers through the analysis of biological fluids to detect cancer-related biomarkers such as prostate-specific antigen (PSA) and carcinoma antigen-125. However, many existing liquid biopsy strategies for early detection of cancer often come with the drawbacks of complex procedures and limited sensitivity. Compared with biomarkers of proteins and metabolites, TEs are more prevalent in organisms and exhibit greater stability, making them easier to detect [8]. Schwartz previously summarized the pivotal role of TEs, such as Cu, Se, and Zn, in cancer development, while also discussing their potential utility as diagnostic markers [9]. Furthermore, the imbalances of TEs have also been observed to be associated with various types of cancer in humans. For instance, abnormally elevated Cu concentration can stimulate tumor cell angiogenesis and induce DNA damage via free radicals, thereby facilitating tumor proliferation [10]. Reportedly, elevated serum Cu may be a dangerous factor for various cancers, such as oral cancer (OC) [11], gastrointestinal cancer [12], prostate cancer [13], breast cancer [14], thyroid cancer [15], and lung cancer [16]. In addition, excess Fe can induce heightened oxidative stress in cells, thereby accelerating cell and DNA damage, which is related to the development of hepatocellular carcinoma [17]. Conversely, Se and Zn are often considered as antitumor TEs due to their antioxidant capacity [18, 19]. The deficiency of these two TEs may indicate an elevated susceptibility to liver cancer (LC) [20].

However, there is still controversy among many current studies, that is, the change of a certain TE does not mean the occurrence of disease, which should be a comprehensive result of multiple elements. Currently, there is a lack of consensus regarding variation of TEs in cancer patients; therefore, the present review focuses on recently published case-control studies of the most common cancer types. We aim to extend the current view by systematically reviewing and analyzing the relationship between the occurrence of specific cancers and TEs disorders in different human systems. Furthermore, the recent advances in analytical techniques for the detection of TEs in different biological samples, including traditional methods, isotope ratio determination, bioimaging techniques, and machine learning (ML)-assisted cancer detection are also summarized. Finally, we outline the challenges and prospects of research on the relationship between TEs and cancer, with the aim of providing a more comprehensive understanding of the role of TEs in different cancers and to better develop appropriate treatment strategies.

Trace elements in human cancerous tissues

At present, cancers occur in different sites of the body's various systems, accompanied by different TEs variations. In this study, we provide a comprehensive analysis of the relationship between changes of TEs in different biological samples and the main types of cancer in five key systems of the human body, namely the digestive system (Fig. 1), urinary system (Fig. 2), reproductive system (male and female) (Fig. 2), endocrine system, and respiratory system (Fig. 3). Table 1 presents a summary of the alterations in TEs observed across various major cancer types within different systems, as reported in recent literatures (additional references can be found in Table S1).

Digestive system

Oral cavity

OC is a general term for malignant tumors that manifest in the oral cavity, representing one of the most prevalent malignant tumors in the head and neck. Multiple etiological factors contribute to OC development, including prolonged tobacco and alcohol addiction, poor oral hygiene, betel nut chewing, chronic foreign body stimulation, malnutrition, mucosal leukoplakia, or erythema, as well as recurrent oral ulcers. TEs play a vital role in the body's enzyme function, and alterations in the levels of these biomarkers in serum seem to be extremely critical in the pathogenesis of OC. Undoubtedly, early detection of precancerous lesions based on TE levels has the potential to significantly reduce cancer incidence [21].

During the past decade, many epidemiological studies have investigated the relationship between serum levels of TEs such as Cu, Zn, Fe, Se and the risk of OC. In a hospital-based case-control study conducted in Fujian, China, essential elements (Cu, Se, Zn, Sr, and Cr) and nonessential elements (As, Li, Th, Ce, Ti, and Sc) exhibited significant associations with OC. Elevated serum Zn and V levels adversely affect OC, while Sr, Se, and Cu have significant positive effects [22]. Although a study from Turkey reported reductions in serum Cu and Zn levels in OC patients, it is important to note that these findings were based on a small sample size [23]. Because of the sampling limitation, most studies have been conducted on smaller scales, thus failing to provide solid evidence for establishing a correlation between serum TE levels and OC. In order to accurately assess the situation of OC patients in China, Chen et al. conducted a large-scale study involving 344 OC cases and 1122 matched healthy controls, which preliminarily suggested that abnormal serum levels of Cu or Zn levels (excess or deficiency) may serve as risk factors for OC [24].

While environmental factors, such as geographical differences, may be associated with OC, two other studies from India also observed increased serum Cu levels in OC patients [25, 26]. The long-term consumption of betel nut is the primary etiological factor for OC in India. Coincidentally, the average Cu content in betel nut reaches as high as 302 nmol/g. Considering the high prevalence of serum Cu abnormalities among OC patients, it may be deemed a cost-effective and noninvasive alternative approach for screening, diagnosis, and monitoring of OC.

Recently, a study investigated the alterations in TEs levels in serum and saliva of patients with OC. Compared with healthy controls, serum Cu levels increased and Zn, Se, and Mo levels decreased, whereas saliva Cu and Zn increased significantly [11]. Although changes in saliva TEs may indicate disease progression, the short-term secretion of saliva and the local oral environment are more susceptible to fluctuations compared to serum, potentially influenced by daily dietary or other effects. Therefore, further exploration is needed regarding the role of salivary TEs in early diagnosis and prognosis of OC, as well as detection of TEs in alternative biological samples such as tissue. In brief, abnormally elevated serum Cu level may indicate an increased risk of OC, warranting attention and further comprehensive investigation.

Stomach

Cancer of the stomach, also known as gastric cancer (GC), is characterized by the development of malignant cells in the gastric epithelial tissue [27]. The precise etiology of GC remains unclear; however, factors such as Helicobacter pylori infection, tobacco and alcohol consumption, nitrite intake, occupational exposure to various dusts, and other variables have been identified as potential risk factors for this disease [28].

To date, a large number of studies have been conducted to monitor and quantify TEs levels in different biological samples, including serum, hair, and tissue, as indicators for assessing nutritional status and early diagnosis of GC [29–31]. For example, an earlier study reported a statistically significant increase of 46% in the average concentration of Cu in malignant tissues compared to normal tissues [32]. Recently, Sohrabi et al. reported increased levels of Zn, Cr, and Sn in gastric tissues [29]. Elevated Sn levels may be attributed to excessive consumption of canned foods. These differences in TEs might reflect mechanisms in cellular changes during tumorigenesis, but larger cohort studies are needed to confirm them. It is essential to consider complex factors such as gender (work exposure, hormonal influences, variations in defense mechanisms) and intake risks (dietary habits, environmental pollution).

In addition to the investigation of cancer tissue samples, an imbalance in serum TEs has also been associated with GC. A study conducted in Fujian, China demonstrated a positive correlation between serum Cu levels and the overall risk of GC, as well as an elevated Cu/Zn ratio. Conversely, Se, Ca, and Mo appear to mitigate this risk [12]. Further in vivo animal studies are still needed to provide clear evidence. Turkdogan et al. found that the levels of Cd, Ni, Fe, and Mn in the serum of Turkish GC patients decreased [30]. The decline in serum Fe and Zn levels could be attributed to excessive nutrient consumption during cancer development and patient malnutrition, while Cu levels may increase abnormally due to its requirement for tumor growth promotion [33].

A comprehensive study with a large sample size revealed significant positive correlations between Ca, Zn, Fe, Al, Cr, Pb, Se, and V levels in hair and fingernails of patients with GC. Conversely, Mg, Mn, Sr, and As exhibited an inverse relationship [31]. The accumulation of Cd and Cr in hair may be attributed to exogenous factors like smoking and alcohol consumption. Although nails and hair are more susceptible to exogenous contamination compared to biofluids, they provide noninvasive and convenient sampling methods for detecting TEs during precancer screening owing to their long-term accumulation properties.

Colorectum

Colorectal cancer (CRC) pertains to malignancies occurring in the colon (large intestine) or rectum, which constitute the distal part of the gastrointestinal tract. Over the past decade in China, due to inadequate early screening and treatment approaches compared to those employed in the United States, both incidence and mortality rates of CRC have exhibited an upward trend [34]. Therefore, there is an urgent need for novel biomarkers associated with CRC that can facilitate early disease detection.

Studies conducted in China and the United States have demonstrated a significant association between serum levels of Cu and Zn with CRC [35, 36]. Recent studies on CRC patients in Malaysia and Turkey have reported conflicting findings regarding alterations in serum Zn and Cu levels [37, 38]. Factors like gender, age, and region can influence these results. Large-scale and multicenter longitudinal studies are needed to resolve these discrepancies.

Changes in TEs not only have a correlation with cancer occurrence but also provide insights into different stages of the disease. A study in Guangdong, China reported that CRC patients had lower levels of TEs like Fe, Zn, and Se in their whole blood compared to healthy controls, while Cu and Ni levels were higher [39]. Interestingly, the levels of Mg, Fe, Zn, Se, and Ba in the whole blood gradually decrease with the development of tumor node metastasis classification. This may be attributed to the excessive demand for these TEs during malignant tumor progression. These TEs changes in whole blood could potentially conduce to staging diagnosis of CRC and improve clinical treatment strategies.

TE analysis has expanded from serum to tissue samples. A Balkans study found significant differences in Mg, Ca, Cu, Zn, Se, Mn, Cd, Cr, and Hg levels between malignant CRC tissues and adjacent healthy intestinal tissues. In particular, the Cu/Zn ratio was higher in CRC tissues [40], indicating its potential as a predictive marker for CRC invasion and progression. Another cross-sectional study measured eight TEs in cancerous and adjacent noncancerous CRC tissues. Results showed higher median levels of Zn, Cr, Cu, Al, and Pb in cancerous tissues, while Mn, Sn, and Fe were lower in noncancerous tissues [41]. A comprehensive survey and rigorous analysis suggest that gender and smoking history affect specific TEs. A recent study in Eastern Iran has reported significantly elevated concentrations of Co, Cr, Ni, Pb, and Zn in CRC tissues compared to noncancerous tissues [42]. Long-term exposure to heavy metals in Eastern Iran contributes to the high incidence of CRC. In summary, CRC is susceptible to dietary and environmental contaminants. Assessing Cu and Zn levels shows promise for CRC screening, but further research is needed to confirm the Cu/Zn ratio as a reliable marker.

Liver

LC, also known as hepatocellular carcinoma, is a highly prevalent and fatal malignancy primarily affecting the liver. China bears the highest burden of LC cases due to both escalating incidence rates and its large population size. LC usually has no obvious symptoms in the early stage, so nearly 85% of patients are in the middle and late stage at the time of diagnosis [43].

TEs may be implicated in the pathogenesis and progression of LC, as an imbalance in TEs can disrupt the expression of associated proteins. Hepatic dysfunction in LC patients can lead to metabolic disturbances in TEs, leading to abnormal distribution patterns. Precise determination of TEs such as Cu, Zn, and Se levels in LC patients could serve as a fundamental basis for clinical diagnosis, treatment strategies, and prognostic evaluation.

Previous studies have reported on the determination of serum TEs in patients with LC [44, 45]. A study conducted on Mexican cases revealed elevated serum Cu level and decreased serum Zn level in LC patients compared to controls [44]. Another study in Hunan, China also observed these variations, demonstrating increased Fe levels and reduced Se levels [45]. The significant increase in the Cu/Zn ratio observed in patients may provide valuable insights for evaluating suspected hepatocellular malignancies, thereby suggesting an antagonistic relationship between Zn and Cu in LC patients. Studies also show that higher serum Cu levels are associated with poorer LC survival, while tissue concentrations of Zn and Se exhibited opposite effects [46]. The down-regulated expression of MT1G and MT1H genes in liver tissue of LC patients may significantly contribute to the dysregulation of TEs in the body.

A separate study investigated the association between metal concentrations in liver tissue and survival outcomes using two independent cohorts (Peru and France). Both cohorts exhibited higher average Cu concentration in LC tissues compared to noncancerous tissues, while multiple metals including Mn, Fe, Co, Zn, As, Se, Rb, Mo, Cd, Pb, and Sn were found at lower levels in LC tissues than noncancerous tissues [47]. Interestingly, only Se levels in Peru were associated with prolonged survival. Geographical factors should be considered as contributors to cancer, particularly in cases of prolonged exposure to the environment with high levels of heavy metals. A study in northwestern China also reported lower concentrations of As, Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb, Se, and Zn in the tissues of LC patients [20]. The differences in the composition of the study population may be the cause of this contrast. In summary, considering the characteristic alterations in TEs during LC development, focusing on variations specifically related to Cu, Zn, and Se could potentially aid early diagnosis.

Esophagus

Esophageal cancer (EC) is the third most prevalent gastrointestinal tumor, following colon cancer and GC, with poor prognosis and high mortality rates. Various factors including alcohol consumption, tobacco use, environmental exposures, and dietary habits have been implicated in the development of EC [48].

Some essential TEs play a vital role in the prevention of EC. For example, adequate Zn intake has been associated with a reduced risk of EC, while Se ingestion has demonstrated a protective effect against EC development [49]. Numerous clinical and epidemiological studies have explored the correlation between TEs and EC risk over the years [31, 50]. A study of hair samples from EC patients in Pakistan revealed elevated levels of As, Cd, and Ni, whereas the average concentrations of Se and Zn exhibited an inverse pattern [51]. Moreover, the levels of TEs in blood circulation of patients may vary according to tumor progression. A report conducted on Pakistani patients with different clinical stages of EC revealed that serum concentrations of Ca, Pb, Se, and Sb levels were increased in the early stage (stage I), while As, Ag, and Mn showed significant increases in stage II. Ni, Cd, and Co concentrations were higher in stage III, whereas Hg, Sr, Cr, and Cu exhibited their highest levels in the late stage IV [52].

In addition to hair and blood samples, TEs in cancerous tissues also demonstrated alterations when compared to healthy controls. Xie et al. first evaluated the link between tissue element concentration and EC, finding higher concentrations of Mn, Se, Cu, and Ti in tumor tissues compared to noncancerous tissues [50]. Previous studies also showed that elevated levels of Se exposure can effectively reduce cancer risk, particularly in plasma/serum and toenail samples [53]. Notably, Se accumulation was exclusively observed in cancer tissues, while other biological samples like serum, plasma, hair, and toenail displayed a decrease in Se content. Given the heightened oxidative stress in cancer patients, characterized by an upsurge in free radicals and/or reactive oxygen species, Se assumes a crucial role in the development of EC by effectively scavenging excessive oxidant-induced free radicals. Considering the alterations in Se level reported among EC patients as described above, dietary interventions to maintain normal Se levels may help reduce disease risk.

Urinary system

Kidney

Renal or kidney cancer (KC) refers to cancer that originates from kidney cells, with renal cell carcinoma (RCC) being the predominant type. The etiology of RCC is not fully understood but may be associated with factors such as age, smoking, medication use, hormonal influences, coffee consumption, viral infections, radiation exposure, genetics, or TEs.

Several studies have explored the association between TE levels in the human body and KC [54]. However, their relationship remains elusive due to conflicting results across multiple studies. For instance, a study in India reported the accumulation of Zn and the decrease of Fe in renal tissue of RCC patients compared to healthy controls [54]. In contrast, two separate studies on RCC cases from Beijing, China, and Egypt demonstrated elevated tissue Fe concentration alongside reduced tissue Zn level [55, 56]. These differences may stem from the limited sample sizes and regional variations. Despite discrepancies, the indispensability of Zn in RCC remains evident. Studies have indicated a positive correlation between dietary Zn intake and an increased risk of KC, suggesting a potential association between Zn and KC occurrence [57].

A recent study investigated alterations in TEs levels in the blood and urine samples of RCC patients. Compared to controls, significant increases were observed in the blood of As, Cu, Mn, Cd, Pb, and Hg concentrations, while Se levels were decreased. Additionally, elevated levels of urinary Mn and Se were detected [58]. Another study by Zheng et al. reported a reduction in the levels of Mg, Mo, Sn, and V in RCC urine samples, but urine Cd was found to be increased [59]. The elevation of heavy metal TEs in blood or urine may arise from environmental exposure or metabolic abnormalities, contributing to disease initiation. For example, Cd and Pb can generate excessive free radicals leading to oxidative damage that may trigger genetic and epigenetic cascade reactions associated with the development of various cancers. Meyer et al. investigated the interaction between Se and RCC in patients from Berlin, Germany, revealing a significant decrease in blood Se level [60]. This suggests Se could be a potential diagnostic biomarker for RCC. During tumor progression, there is a gradual decline in Cd and K, along with increasing levels of P, Pb, and Zn [61]. The monitoring of TEs changes in biological samples of RCC patients can not only serve as a foundation for supplementary therapy but also reveal the various stages of cancer advancement.

Bladder

Bladder cancer (BC) is the most common malignant tumor of the urinary system, and it has the highest incidence rate among all malignant tumors of the urinary system. Tobacco consumption, genetic and environmental exposures, and even alterations in TEs are potential risk factors for BC development. Early diagnosis plays a pivotal role in successful treatment [62].

The imbalance of TEs (Ca, Zn, Cu, Se, etc.) exhibits the potential for the detection, staging and grading of BC [63, 64]. A study in Erlangen, Germany analyzed major elements and TEs in two cohorts (discovery cohort and validation cohort) and observed a significant elevation of serum Ca, Li, K, Ni, and Sr in both cohorts [65]. These findings present a novel and valuable approach for early diagnosis of BC. Recently, the combined characteristics of TEs and their natural isotopes have emerged as promising indicators for identifying BC. For example, Wang et al. developed a ML model based on 2D Cu features, namely the isotope composition and concentration of Cu in plasma and red blood cells [66]. Given that the integration of blood Cu signals with ML enables improved discrimination between BC patients and healthy controls, it holds promise as a versatile tool for cancer research and potential clinical applications.

In addition to common TEs, alterations in As levels were also observed in patients with BC. Significantly elevated As concentration was detected in various biological samples, including hair, serum, and cancer tissue [51, 67]. Excessive exposure to As can promote carcinogenesis by modulating cellular differentiation and proliferation. Apart from environmental contamination by As, occupational hazards and smoking are additional factors that contribute to variation in As levels within the human body. Consequently, monitoring TEs concentrations not only holds potential as an indicator for cancer development but also offers a means of reducing its incidence through risk factor management.

Reproductive system

Male reproductive system

Prostate: Prostate cancer (PCa) is the second most prevalent malignancy and the fifth leading cause of cancer-related mortality in men, making it the most frequently diagnosed cancer among males worldwide. Despite being widely employed for PCa diagnosis guidance, serum PSA quantification has inherent limitations as highlighted by Prensner et al. [68]. Therefore, there is an urgent need to explore novel biomarkers for PCa detection and monitoring. Given the simplicity and noninvasiveness of urine collection, it represents a promising source for the development of new PCa biomarkers. The imbalance of urinary TEs is associated with the etiology of benign prostatic hyperplasia and PCa [69, 70]. Moreover, the consistent reduction of urinary Zn in multiple studies suggests its potential as a superior diagnostic and prognostic indicator for PCa. A mathematical classification model based on urine elemental spectrum has been developed to predict the status of PCa patients with an impressive accuracy of up to 89%, surpassing that of standard PSA detection [69].

Zn deficiency has been observed in various biological samples, including serum, whole blood, cancer tissue [13, 70–72], and prostate secretions of patients with prostatitis [73]. Previous studies have reported variations in the concentrations of Zn, Cu, and Se varying in different regions of prostate tissue in patients with PCa. For instance, individuals with pathological changes in their prostate tissue exhibited an increase in Cu level within the central region but a decrease within the peripheral region [71]. Conversely, the reduction of Zn in PCa tissues has been widely acknowledged [74]. However, elevated Zn levels were observed in toenail samples, suggesting potential poor Zn absorption and emphasizing the significance of maintaining a balanced Zn status in PCa pathogenesis [75]. A recent investigation in South Australia found significantly higher levels of Fe, Cu, Ca, and S in PCa patient plasma, while Se levels were lower [76]. In a small-scale study involving PCa subjects from Iran, hair sample analysis demonstrated a notable increase in Al content alongside reduced Si and P concentrations [77]. Despite multiple studies linking TEs disorders to PCa, large-scale multicenter cohort studies are warranted to elucidate the precise role of TEs in the etiology of PCa due to potential sample effects.

Female reproductive system

Female breast: By 2020, female breast cancer (BCa) surpassed lung cancer as the most prevalent malignancy, with an estimated 2.3 million new cases, accounting for 11.7% of all cancer incidences [6]. In addition to genetic factors, BCa is also influenced by environmental factors such as diet, specific lifestyles, and chemicals that disrupt the homeostasis of TEs in the human body [78]. Several studies have demonstrated a potential association between levels of TEs in various human biological samples, including urine, serum, hair, cancer tissue, nails, pleural fluid, and the risk of BCa [79–84]. A meta-analysis conducted in China using a random effect model found that serum Cu level was significantly elevated in patients with BCa compared to healthy controls, while Zn concentration was found to be statistically lower [5]. These findings suggest a potential association between altered Cu and Zn levels and an increased risk of BCa development. Consistent evidence supporting this observation has also been reported in studies conducted in other regions, including China, Turkey, and Pakistan [14, 85, 86]. Additionally, a prospective multicenter cohort study revealed that an elevated Cu/Zn ratio may be correlated with diminished overall survival rates following BCa diagnosis [87].

Se, as an anticancer element, appears to exhibit promising preventive effects against BCa. Multiple studies have identified reduced serum Se levels in cases, suggesting its involvement in the occurrence and progression of BCa [65, 88]. However, a study in western Iran revealed the presence of Se accumulation in the cancer tissues of BCa subjects, accompanied by an increased Cu level and decreased Zn level [83]. It is widely acknowledged that serum Se level serves as a superior prognostic indicator for BCa (exhibiting a strong negative correlation), while no significant association exists between Se concentration in cancer tissues and BCa. The abnormal elevation of Se concentration in cancer tissues may be attributed to its facile binding to proteins and rapid uptake by proliferating cancer cells.

Furthermore, there appears to be a close relationship between the levels of TEs in hair and BCa incidence [89]. Chanihoon et al. reported a three to four times higher average concentration of Cd in hair compared to the control group, while Zn levels exhibited an opposite trend [90]. A Moscow study found reduced Cu and Mn levels in the hair of cancer patients, which could potentially be associated with BCa development [82]. The abnormal increase of harmful elements such as Cd and Mn in hair may be attributed to exposure through food, drinking water, air, or skin contact, leading to potential disruption of other element metabolism and resulting in a broader range of TEs imbalances. However, due to significant variations among current studies on BCa patients regarding geographical location, sample size, and composition, the results appear inconclusive. Generally, the accumulation of harmful elements like Cu and Cd within cancer tissues along with serum Se depletion may indicate a high risk for BCa.

Cervix uteri: In 2020, cervical cancer (CC) emerged as the fourth most lethal malignancy among females, constituting 6.5% of all cases [6]. Human Papilloma Virus stands out as the primary etiological factor for CC. Concurrently, alterations in TEs levels also play a crucial role in CC pathogenesis [91]. A meta-analysis encompassing 454 CC patients and 880 controls revealed significantly lower serum Fe levels in Chinese CC patients compared to the control group [92]. Further high-quality investigations are warranted to elucidate the precise impact of serum Fe levels on CC risk. A small-scale study in India indicated that elevated plasma Cu and reduced Zn and Se concentrations might serve as potential indicators for CC occurrence [93]. Specifically, plasma Se level was found to be associated with a reduced risk of poor prognosis in CC [94]. Apart from case source variations, the selection of control subjects can also influence the results of the study. Investigating plasma ion changes pre- and post-cisplatin treatment holds significant importance in mitigating adverse reactions to radiotherapy and chemotherapy. It was observed that plasma Pt levels significantly increased while Na, Mg, P, K, Ca, Se, Cu, Zn, Se, Sr, and Ba levels significantly decreased following treatment [95]. Al and Cu ions exhibited a significant correlation with therapeutic efficacy, suggesting their potential association with CC. Essential element supplementation is vital for maintaining ion homeostasis and may prevent or ameliorate certain diseases such as cancer.

Endometrium: Endometrial cancer (ECa) is a group of epithelial malignant tumors that originate in the endometrium and primarily affect perimenopausal and postmenopausal women. The occurrence of ECa often coincides with alterations in serum Cu and Zn levels; however, variations among studies may exist due to regional disparities. For instance, a Turkey study reported lower serum Cu and Zn levels, as well as a reduced Cu/Zn ratio, in patients with ECa compared to controls [96]. Conversely, a Slovakian study observed only a slight decrease in serum Cu level among patients with ECa while noting no significant changes in Zn level [97]. Given their involvement in antioxidant enzymes in the body, Cu and Zn may be influenced by various factors such as inflammation and serum antioxidant levels, potentially leading to disorders.

Higher Cd concentration has been observed in the whole blood of ECa patients [98]. Similarly, it has been suggested that elevated Cd level and decreased Co-concentration are independent risk factors for ECa [99]. Elevated blood Cd concentration appears to be associated with an increased risk of ECa. Considering the significant impact of lifestyle habits like smoking on Cd and Pb concentrations, further studies involving larger populations should consider environmental and lifestyle exposure to heavy metals. Serum Se levels also strongly correlate with ECa risk [100], so comprehensive screening for individuals with low Se levels is recommended to prevent cancer.

Ovary: Ovarian malignant tumor, also known as ovarian cancer (OCa), is prevalent among female reproductive organs. The etiology of OCa remains unclear; however, it is speculated to be associated with environmental factors [101]. A previous study revealed that patients with OCa exhibited significantly elevated average serum Cu levels and Cu/Zn ratios compared to those with benign ovarian lesions [102]. Another study found lower blood Mn and Se in Turkish OCa patients [103]. Consequently, diminished concentrations of Mn and Se along with an elevated blood Cu/Se ratio could potentially serve as favorable indicators for OCa.

Certain TEs act as essential cofactors for antioxidant enzymes involved in cellular redox regulation, such as Mn (Mn-SOD), Cu, Zn (Cu, Zn-SOD), Se (glutathione peroxidase, GPx1). Alterations of TEs levels in serum reflect changes in total antioxidant enzyme activity, leading to an imbalance in the intracellular redox state. An epidemiological study involving African-American women demonstrated that those with a higher dietary intake of Se (>20 mg/day) exhibited a lower risk of developing OCa compared to individuals without Se supplementation, thus highlighting the potential anticancer effects of Se [104]. In OCa patients, the average levels of As, Cd, and Ni in scalp hair were significantly higher than those in standard samples, while Zn and Se levels exhibited an inverse relationship [105]. TEs (such as, As, Cd) that may confer benefits to hair are susceptible to environmental exposure and personal lifestyles. However, Se acts as a constituent of selenoproteins that can mitigate oxidative stress and restrict DNA damage in OCa patients. A meta-analysis and Mendelian randomization study demonstrated reduced circulating Zn levels among individuals with OCa [106]. In summary, further investigation into Zn intervention and Se supplementation for OCa patients hold potential clinical implications.

Endocrine system

Thyroid

Thyroid cancer (TC) is the most prevalent malignant tumor of the endocrine system, accounting for nearly 90% of endocrine cancers [107]. The lack of systematic TC detection poses a current challenge that necessitates identifying appropriate biomarkers to facilitate prompt diagnosis. Many biological samples can be utilized for evaluating TEs, which appear promising as a potential tool for expedited TC detection.

In recent years, several studies have investigated TEs in biological samples from TC patients [108, 109]. A study in Shenzhen, China revealed lower concentrations of Mn, Pb, Fe, Co, Cu, Ba in serum and Fe in urine among TC patients compared to the healthy control group [110]. This study integrated a ML algorithm for TC prediction with high accuracy rates, highlighting the promising potential of metallographic analysis for TC identification. Another study found significantly higher median levels of Mn, Co, Cr, K, Fe, Mg, Pb, Na, and Ni in scalp hair samples from TC patients, while Zn levels were lower [111].

Cd is a prominent endocrine disruptor that plays a crucial role in the pathogenesis of TC. In comparison to healthy tissues, TC tissues exhibited significantly increased levels of Cd and U, while Co, Ni, and Se concentrations were significantly decreased [112]. Notably, the markedly elevated Cd/Se ratio could serve as a distinguishing factor between the two groups under investigation, (i.e. case and control groups). Another meta-analysis has demonstrated a significant correlation between Se element and TC occurrence, suggesting that Se supplementation may prevent and treat TC effectively [113]. For example, supplementation of Se-containing compounds may regulate the disrupted intracellular oxidative stress caused by disease or cancer occurrence, but the dosage and disease stage must be carefully considered. A recent large-scale paired study in China reported lower urine levels of Cr, Mn, Ni, As, Cd, Se, Sb, Tl, and Pb in the case group compared to those in the control group, while Co content exhibited an opposite trend [114]. In general, increased Cu accumulation and decreased Zn or Se levels were predictive factors for TC occurrence [10, 15, 115]. Although thyroglobulin and calcitonin have been utilized as biomarkers for TC diagnosis, monitoring supplementary elements such as Se, Zn, and Cu could potentially enhance diagnostic accuracy.

Pancreas

Pancreatic cancer (PC) is a highly malignant form of human cancer with significant lethality. Risk factors for PC include smoking, chronic pancreatitis, obesity, long-term diabetes, and a strong family history of PC [116]. Previous studies have identified abnormalities in several proteins involved in TE homeostasis in relation to PC [117].

An early Spanish study reported the accumulation of Cd, As and Pb in patients’ toenails, while high concentrations of Se and Ni were found to be negatively correlated with the risk of PC [118]. This finding was consistent with another study in the same area [119], suggesting that TEs may play a significant role in PC development. Pumarega et al. observed a slight increase in Al concentration with longer toenail collection time, whereas Zn and Se concentrations showed a slight decrease [120]. The PC group exhibited significantly lower levels of serum Fe, Cd, Zn, Mn, and Pb compared to the control group; meanwhile, Cu and Ni levels were elevated [121]. Further research is needed to elucidate the underlying physiological mechanisms responsible for these TEs alterations. Compared to healthy controls, urine levels of Ca and Mg were significantly reduced in patients with PC, while Cu and Zn levels were increased [122]. Schilling et al. also found that the preferential excretion of lighter isotopes of Zn (δ⁶⁴Zn) in PC may support the imbalance of metalloproteins in patients [122]. Noninvasive urine analysis for TEs holds great promise as a novel method for precancerous screening, particularly for PCs.

Respiratory system

Lung cancer (LCa), also known as bronchial lung cancer, is a malignant tumor that originates from the bronchial mucosa or glands of the lung. Previous study has demonstrated that long-term heavy smoking plays a significant role as a pathogenic factor for LCa [123]. Given that LCa typically manifests symptoms at an advanced and incurable stage, early detection of disease occurrence assumes paramount importance.

Compared to other types of cancers, the researchers focused on Cu and Zn elements in LCa patients. Elevated blood Cu levels and decreased blood Zn levels are consistent findings across most studies. For example, a study in Poland reported a decrease in whole blood Mn concentrations, as well as serum and whole blood Zn levels, along with an increase in serum Cu level and Cu/Zn ratio [124]. These findings support the hypothesis that alterations in the redox state observed in LCa patients are associated with changes in TEs status (specifically Zn, Mn, and Cu). Notably, several subsequent studies have confirmed these observations [16, 125, 126]. A meta-analysis comprising four European studies and 35 Asian studies demonstrated an association between elevated serum Cu/Zn ratio and increased risk of LCa, with consistent findings observed in both European and Asian populations [126]. Specifically, the reduction in blood Zn concentration may be attributed to heightened demand for Zn by LCa cells, leading to their uptake of additional Zn from the blood.

Zhang et al. [19] revealed that LCa patients exhibited low blood Se levels and high blood Cd levels, suggesting a potential protective effect of higher blood Se concentration against LCa risk. However, a study in Liaoning, China reported inconsistent results regarding urine samples from LCa patients, indicating an increase in Cr, Cd, Cu, Zn, Ni, Se levels alongside a decrease in Mn and Al levels [127]. The different biological samples (blood and urine) employed in the two studies were pivotal determinants for the divergent outcomes, as urinary Se excretion in LCa patients exhibited a significantly higher level compared to that of healthy individuals. Further investigations encompassing both individual and combined analyses of these sample types are imperative to attain consistent findings.

Other sites

In addition to the various cancers in the aforementioned systems, alterations in TEs levels also manifest during the development and progression of other malignancies. The measurement of serum TEs levels (including Fe, Zn, Se, Cu) in patients with hematological tumors can serve as a determinant for assessing myeloma severity, thereby offering potential preventive and therapeutic implications [2, 128]. Lower serum Zn levels in brain cancer patients indicate its potential pivotal role in tumor etiology and prognosis [129]. Compared to adjacent tissues, the content of Mg was increased while that of Cu was decreased in cancerous tissues of patients with brain tumors [130]. Due to the limited number and challenges associated with obtaining tissue samples from individuals with brain cancer, a larger sample size comprising diverse types of cancers is required for validation. A large-scale nested case-control study conducted on American subjects demonstrated an association between elevated levels of Hg in toenails and an increased risk of skin cancer [131].

In conclusion, the association between various TEs and cancer is of significant interest. Apart from exogenous exposure and dietary intake, perturbations in organismal homeostasis can also contribute to alterations in TEs levels. By monitoring TEs variations across different biological samples (e.g. blood, hair, cancer tissue, toenail), it may provide a more robust theoretical foundation for early cancer screening and treatment. This article primarily discusses the disruption of TEs in cancer; however, similar changes have also been observed in other diseases. Table S2 summarizes some of the literature on Cu and Zn imbalances in both cancer and other diseases, highlighting the significance of elemental imbalances to the organism and urging researchers to investigate the underlying mechanisms to reduce disease incidences.

Analytical techniques of trace elements in human cancerous tissues

Canonical methods

The accurate detection of TEs in the human body is a fundamental prerequisite for both theoretical research and clinical applications. Imbalances in TEs are not only associated with common diseases but also serve as indicators of cancer development [36]. Therefore, achieving accurate, rapid, and convenient detection methods for TEs in the human body has emerged as an urgent issue to be addressed.

Since the study of TEs began in the 1970s, the current repertoire of techniques employed for human TEs detection primarily encompasses colorimetry, atomic fluorescence spectrometry, atomic absorption spectrometry (AAS), X-ray fluorescence spectrometry, neutron activation analysis, and the widely used and recognized inductively coupled plasma mass spectrometry (ICP-MS). AAS is a frequently employed method for quantitative analysis of trace and ultra-trace metals; however, it still exhibits certain limitations when compared to ICP-MS, including inadequate sensitivity toward specific elements and incapability to simultaneously determine multiple elements. In recent decades, ICP-MS has emerged as the most promising analytical instrument for inorganic element analysis owing to its rapid mass scanning speed, short operational cycle, minimal interference from ion information provided, and remarkably low detection limit across a wide range of elements. Inductively coupled plasma-optical emission spectroscopy (ICP-OES) is a versatile analytical technique that enables qualitative and quantitative analysis of over 70 metal elements and some nonmetal elements in various sample types, including geological, environmental, chemical, biological, pharmaceutical, food, metallurgical, and agricultural samples. Despite its lower sensitivity compared to ICP-MS, ICP-OES has gained popularity due to its cost-effectiveness [64, 132].

Collision reaction cell and sector field inductively coupled plasma mass spectrometry are emerging techniques that can eliminate multi-atom interferences for improved accuracy and precision [133]. Laser ablation ICP-MS (LA-ICP-MS), an analytical technique offering spatial resolution, enables in-situ and micro-area analysis of elements in biological samples and facilitates elemental imaging [134]. With the increasing demand for isotope detection, multiple collectors inductively coupled plasma mass spectrometry (MC-ICP-MS) has also emerged as a viable option. Furthermore, integration with other instruments, such as ICP-MS/MS, has even been proposed to improve the analysis efficiency [45]. In summary, these commonly used methods have their advantages and disadvantages (Table S3). Some emerging technologies are introduced and discussed below (Fig. 4).

Isotope ratio determination method

Isotope fractionation refers to the phenomenon in which isotopes of an element are distributed in varying proportions among different substances during physical, chemical, biological, and other reaction processes. The stable isotope composition of elements, serving as a robust tracer that provides a distinct dimension from the concentration of TEs, was successfully introduced into multiple research fields in biology during the early 1970s. Recently, employing stable isotopes such as ²H, ¹³C, ¹⁵ N, and ¹⁸O as marker elements for tracing various metabolites (e.g. amino acids and glucose) has become an important approach to investigating metabolic imbalances associated with cancer initiation, growth, and metastasis [135]. Furthermore, from a sample storage perspective, the inorganic components of biological media such as tissue, blood, and urine exhibit greater stability due to their lower susceptibility to organic matter decomposition. The application of TEs in human diseases has been investigated, demonstrating the potential utility of stable isotope analysis as a diagnostic and prognostic tool [136–140].

Various techniques have been developed for high-precision isotope analysis, including thermal ionization mass spectrometry (TIMS), isotope ratio mass spectrometry (IRMS), MC-ICP-MS, and LA combined with MC-ICP-MS (LA-MC-ICP-MS). TIMS offers extremely high measurement precision, but has very low comparative throughput and relatively slow analysis speeds (and therefore cannot be used for clinical testing at present.) IRMS is mainly suited to specific isotope analyses such as C, N, and S, and has limited analytical capability for other elements. Currently, IRMS is employed to investigate the natural variations of stable S isotopes in cancer patients. Balter et al. reported an enrichment of ³²S and a reduction of ³⁴S in the blood samples from hepatocellular carcinoma patients [138]. Moreover, the combination of IRMS with elemental analyzers, gas chromatography, liquid chromatography, and other analytical instruments has been extensively utilized for analyzing specific stable isotopes in complex matrices in various fields such as the food industry and environmental science [141, 142].

Among these techniques, MC-ICP-MS remains the preferred technology due to its ability to combine the advantages of easy high temperature plasma ionization and precise isotope determination using multi-receiver magnetic field MS. It has been successfully applied in determining important TEs such as Fe, Zn, Cu, and Ca in biological samples from patients with cancer and chronic kidney disease [10, 80, 122, 143]. For example, Wang et al. employed MC-ICP-MS to investigate alterations in the concentration of the Cu element and its isotopes in the blood of BC patients [66]. In comparison with healthy controls, BC was found to induce significant modifications in the natural Cu isotope (⁶⁵Cu/⁶³Cu) level in patient blood samples, as evidenced by a decrease in δ⁶⁵Cu values indicating ⁶⁵Cu depletion (enrichment of ⁶³Cu). MC-ICP-MS was also utilized for assessing isotopic compositions (δ⁶⁵Cu, δ⁶³Cu) in serum, urine, and plasma samples from TC patients [10, 110]. The lower δ⁶⁵Cu content observed in patient plasma and higher δ⁶⁵Cu content detected in tumor tissue support the hypothesis that Cu isotopes hold promise as potential diagnostic markers for TC. In addition to cancer, patients with Alzheimer's disease exhibit altered Cu isotope ratios in both serum and whole blood, suggesting a disruption in Cu homeostasis in the body [144].Compared with other isotope analysis methods, MC-ICP-MS exhibits exceptional accuracy. However, solution-based MC-ICP-MS does have certain limitations, such as the requirement for liquid injection during sample analysis and susceptibility to matrix effects. Considering these drawbacks, integrating this technique with other sample introduction systems could be a viable approach to broaden its application scope. For instance, LA-MC-ICP-MS enables direct (imaging) isotope analysis without the need for complex pretreatment of solid samples. Nevertheless, the utilization of this technology for in situ detection of TEs isotopes in cancer tissue samples remains relatively uncommon. In conclusion, while the substantial potential of natural isotopes such as Zn and Cu as diagnostic markers for cancer, when employed through precise stable isotope analysis or in combination with advanced techniques, can lay a solid foundation for future research in cancer therapy, the requirement for expensive instrumentation in isotope determination also imposes certain limitations on their application in cancer research.

Bioimaging technology

Bioimaging technology utilizes optical or electron microscopy to acquire microscopic images of biological cells and tissues, enabling the comprehension of diverse physiological processes. It plays a pivotal role in cancer screening and facilitates the selection of treatment strategies. Generally speaking, bioimaging serves as a preliminary diagnostic tool, necessitating subsequent therapeutic interventions for disease management. Fluorescence imaging (FI), magnetic resonance imaging (MRI), and other advanced technologies have been extensively employed for specific identification imaging of imbalanced elements during disease progression, such as Zn and Hg [145, 146]. These techniques aid in the initial determination of disease occurrence.

Compared to other conventional bioimaging techniques, FI has gained extensive utilization in molecular analysis and disease diagnosis owing to its remarkable sensitivity, exceptional spatial and temporal resolution, as well as minimal invasiveness [147]. Recently, Li et al. harnessed coal humus acid (CHA) as a protective ligand to fabricate environmentally friendly, user-friendly, and biocompatible highly stable fluorescent Cu nanoclusters (CHA-CuNCs) for sequential quantitative monitoring of tryptophan Hg²⁺ [146]. This method has great potential for rapid cancer detection associated with Hg²⁺ dysregulation. In particular, near-infrared FI offers enhanced imaging depth and reduced background interference, facilitating in vivo visualization of deep tissues and real-time monitoring of cancer progression. Recently, multifunctional near infrared-II fluorophores with specific activation have been developed for the analysis and imaging of cancer biomarkers such as mRNA, miRNA, active enzymes, reactive oxygen species, and gases [148]. These advancements are expected to find broader applications in detecting imbalanced TEs in cancer patients. However, challenges persist in FI techniques, particularly in the detection and quantification of Zn²⁺ and Cu²⁺, including the selectivity of the probes for the target cations and their responsiveness at physiologically relevant concentrations.

In contrast to the limitations of light penetration in optical imaging, MRI offers unrestricted depth penetration and high spatial and temporal resolution, enabling whole-body imaging. Although MRI exhibits low sensitivity, the utilization of reactive contrast agents that respond to divalent cations (Zn²⁺, Cu²⁺) in the body allows for real-time visualization of horizontal changes with superior spatial resolution [149]. Trusso Sfrazzetto et al. provided a comprehensive summary on the advancements and constraints associated with synthetic fluorescent probes for intracellular Zn and Cu detection [150]. Furthermore, PCa cancer tissues exhibit a significant decrease in Zn ion level compared to healthy tissues, thereby highlighting promising applications of Zn-responsive MRI probes for noninvasive detection of Zn ion secretion in prostate tissues [151]. A recent study developed a metal-free probe based on 1,2-bis (o-aminophenoxy) ethane-N,N,N’,N’-tetraacetic acid (BAPTA) dye for specific in vivo detection of Zn using ¹⁹F ion exchange saturated transfer MRI, which has the potential to transform clinical diagnosis of PCa [145]. This innovative approach differs from traditional (¹H) MRI techniques. Paranawithana et al. designed a novel gadolinium-based MRI contrast agent that exhibits high selectivity for Cu²⁺ [152]. This technique has demonstrated successful detection of endogenous unstable Cu²⁺ in live mice, offering promising insights into the unique role of Cu in cancer initiation and progression.

Given the distinct advantages and disadvantages associated with each imaging technique, enhancing diagnostic efficiency necessitates not only the integration of imaging techniques (e.g. positron emission tomography (PET)/MRI, PET/computed tomography (CT)), but also the exploration of multimodal imaging as a burgeoning research focus, including targeted imaging and treatment for BCa [153]. In addition to the imaging techniques meticulously described, LA-ICP-MS, which is specifically utilized for in situ bioimaging of TEs and isotopes within tissue sections, is also employed in cancer research [154].

Machine learning assisted cancer detection

Utilizing multidimensional element information based on a substantial number of samples can serve as a potent tool for improved prediction of disease occurrence and progression [155]. However, the manual processing of such extensive data becomes increasingly intricate with its abundance. Recently, ML has emerged as the preferred choice for researchers to alleviate this burden by devising and analyzing algorithms that enable computers to autonomously “learn,” thereby significantly reducing costs. ML algorithms can be categorized into three types: supervised learning, unsupervised learning, and reinforcement learning [156]. Generally, supervised learning is employed in disease risk assessment and identification research where it learns from a training set comprising input and corresponding marker output, subsequently predicting outcomes for new input data based on the acquired training knowledge [156]. In supervised learning, analysts furnish the algorithm with information pertaining to distinct categories of data (e.g. disease or control) to facilitate its “learning” and extrapolation to “invisible” data (i.e. model development and application on a larger sample set) [157].

ML has been widely employed by numerous researchers for disease detection and pattern recognition in various studies [158]. For instance, a previous study utilized Student's t-test and factor analysis for discriminant analysis, followed by cluster analysis and subsequent discussion of results to investigate the potential contribution of element levels in hair and their interrelationships towards BCa diagnosis [127]. Furthermore, random forest (RF), an integrated supervised learning method capable of mitigating excessive decision-making risks, is particularly suitable for small data volume problems. Chen et al. constructed and evaluated the performance of seven ML algorithms for TC prediction based on serum and urine metal spectra from TC patients and control groups [110]. Among these algorithms, the RF model exhibited superior performance with accuracies of 1.000, 0.858, and 0.813 on the training set, five-fold cross-validation set, and test set respectively. The high performance of ML demonstrates the promising potential of elemental analysis in the identification of TC. To classify BCa and non-BCa groups, a RF model was developed using the “leave-one-out” cross-validation test method. Initially, t-SNE dimensionality reduction was applied to the data encompassing four Cu-related variables: plasma Cu concentration, red blood cell Cu concentration, plasma δ⁶⁵Cu value, and red blood cell δ⁶⁵Cu value [66]. Remarkably, this ML model exhibits versatility as it can also be employed for classifying other types of cancers such as hepatocellular carcinoma. Interestingly, a recent study introduced an artificial neural network approach based on principal component analysis, enabling simultaneous determination of the spatial distribution of biological TEs including Mn, Fe, Cu, and Se through micro-X-ray fluorescence imaging techniques [159]. This innovative methodology provides novel insights for developing effective cancer biomarkers.

The rapid advancement of ML technology has also facilitated the emergence of multiomics experiments, as it enables a substantial increase in the volume of data available for analysis. Recently, He et al. proposed a nontargeted metallography technique that utilizes synchrotron radiation X-ray fluorescence along with ML linear partial least squares discriminant analysis and nonlinear support vector machine (SVM) to establish a screening model [160]. They employed five-fold cross-validation to ensure stability and achieved quick and accurate cancer screening results (with an accuracy exceeding 96%). This breakthrough may serve as inspiration for the development of cancer therapeutic drugs. In summary, ML plays an indispensable role in the field of cancer screening and holds the potential as a valuable auxiliary tool for predicting cancer survival rates and making medical decisions.

Conclusions and further perspectives

TEs play a crucial role in the occurrence and progression of cancer, despite their limited presence within the organism. The alteration of Cu and Zn levels appears to be a prevalent issue across numerous studies. It is widely recognized that an elevated Cu level may indicate the presence of cancer, necessitating thorough medical screening and examination for accurate identification. However, the precise mechanisms governing TEs metabolism and homeostasis during disease processes remain incompletely elucidated. Notably, recent advancements in quantitative detection techniques for TEs in biological samples (such as hair, toenails, blood, and tissues) hold promise as potent biomarkers for early diagnosis and prognosis of cancer. Concurrently, TEs as single biomarkers may not provide a comprehensive picture of disease occurrence; therefore, the development of composite and more convincing biomarkers, such as TEs-metabolites and TEs-proteins, may serve as the first line of defense for disease screening and prevention. Furthermore, with innovative analytical methods and the emergence of ML-assisted processing of extensive data sets, precise, and intuitive analysis will gradually unveil the intricate relationship between TEs and cancer. Novel analytical techniques such as isotope ratio determination and bioimaging can be effectively employed in the field of cancer detection related to TEs. Isotope detection not only enables the identification of potential cancer, but also facilitates the analysis of endogenous and exogenous pollutants in the human body by comparing them with elemental fingerprints present in the environment.

In the future, further investigations are anticipated to delve into the intricate relationship between TEs and cancer, elucidating their precise role in disease progression. However, numerous case studies in this field may inevitably exhibit certain limitations, including subject selection bias and environmental confounders. For cancer cases, accuracy of specific types and similarity of pathological examination should be ensured. Regrettably, uncontrollable factors such as case selection bias may arise due to subjects being predominantly from a particular socioeconomic class (particularly those more inclined towards investing in regular health check-ups), thus limiting generalizability. Different lifestyles (e.g. smoking status and dietary habits) as well as environmental factors (e.g. physical activity levels, occupational exposures, residential environment) can influence disease occurrence; hence research designs should encompass consideration of multiple variables. In addition, the consideration of micro-morphology, metal ion valence, multiple measurements, and other factors should be taken into account in the detection of TEs. Furthermore, careful attention should be given to the selection of appropriate biological samples. While blood is a convenient and easily accessible sample, it may be influenced by daily dietary patterns. Noninvasive urine samples only provide insights into short-term changes in TEs among patients; however, hair and toenail samples can be stored for extended periods of time and allow for monitoring long-term variations in the human body.

In summary, the existing studies have not definitively established the association between cancer and TEs (including whether a causal relationship exists), necessitating future comprehensive large-scale cohort studies incorporating in vivo or in vitro experiments to elucidate the underlying biological mechanisms. Drawing from numerous previous investigations, TE-rich nutrition therapy, such as Se supplementation, emerges as one potential anticancer strategy. Furthermore, maintaining appropriate concentrations of abnormal TEs may also contribute to cancer prevention.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (grant no. 21906179 and 22074098), Sichuan Science and Technology Program (2024NSFSC1154 and 2021ZYD0047) and the Fundamental Research Funds for the Central Universities.

Author contributions

Rui Liu (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Writing—Original Draft), Qian Zhang (Conceptualization, Data curation, Writing—Original Draft), Xueting Yan (Conceptualization, Data curation, Funding acquisition, Investigation, Project administration, Supervision, Visualization, Writing—Review & Editing), and Yi Lv (Funding acquisition, Project administration, Supervision, Visualization)

Conflict of interest

None declared.

Funding

None declared.

Data availability

The data underlying this article are available in the article and its online supplementary material.

References