Latest insights into the global epidemiological features, screening, early diagnosis and prognosis prediction of esophageal squamous cell carcinoma

Abstract

As a highly invasive carcinoma, esophageal cancer (EC) was the eighth most prevalent malignancy and the sixth leading cause of cancer-related death worldwide in 2020. Esophageal squamous cell carcinoma (ESCC) is the major histological subtype of EC, and its incidence and mortality rates are decreasing globally. Due to the lack of specific early symptoms, ESCC patients are usually diagnosed with advanced-stage disease with a poor prognosis, and the incidence and mortality rates are still high in many countries, especially in China. Therefore, enormous challenges still exist in the management of ESCC, and novel strategies are urgently needed to further decrease the incidence and mortality rates of ESCC. Although the key molecular mechanisms underlying ESCC pathogenesis have not been fully elucidated, certain promising biomarkers are being investigated to facilitate clinical decision-making. With the advent and advancement of high-throughput technologies, such as genomics, proteomics and metabolomics, valuable biomarkers with high sensitivity, specificity and stability could be identified for ESCC. Herein, we aimed to determine the epidemiological features of ESCC in different regions of the world, especially in China, and focused on novel molecular biomarkers associated with ESCC screening, early diagnosis and prognosis prediction.

Core Tip: Esophageal squamous cell carcinoma (ESCC) is a common malignancy with decreasing incidence and mortality rates worldwide; however, it was the eighth most prevalent cancer and the sixth leading cause of cancer-related death in 2020. Various molecular mechanisms have been suggested, and a better understanding of the epidemiology and characteristic biomarkers may facilitate ESCC screening, early diagnosis and prognosis prediction. This review aimed to discuss the epidemiological features, screening, early diagnosis and prognosis prediction of ESCC, with a focus on promising molecular biomarkers, to provide a guide for the development of novel strategies for ESCC patients.

INTRODUCTION

With age-standardized incidence and mortality rates of 6.3/100000 and 5.6/100000, respectively, in 2020, esophageal cancer (EC) was the eighth most prevalent malignancy and the sixth leading cause of cancer-related death according to the Global Cancer Statistics (GLOBOCAN) 2020 database (https://gco.iarc.fr/)[1] (Figure 1). EC incidence and mortality rates are diverse among different geographic areas, sexes, races, etc.[1]; for instance, EC is the sixth most prevalent malignancy and the fourth leading cause of cancer-related death in China[1-3]. Notably, approximately 90% of EC patients have esophageal squamous cell carcinoma (ESCC), which is the predominant histological subtype of EC in China and worldwide[4-7]. Due to the lack of specific early symptoms, only 20% of ESCC patients are diagnosed in the early stage; for patients diagnosed early, the 5-year survival rate is 85%[8]. However, ESCC patients are usually diagnosed at an advanced stage by gastrointestinal endoscopy or enhanced thoracic computerized tomography, and the 5-year survival rate is approximately 20%[9-12]. Due to advances in preventative, screening, early diagnostic and therapeutic strategies, the incidence and mortality rates of ESCC have decreased, even in many high-risk areas; however, there were still 512500 new cases of ESCC in 2020 worldwide[1,13]. Therefore, ESCC remains a major global health issue with a large clinical burden, and certain challenges still exist in the management of ESCC[1,5].

Figure 1.

The proportions of incidences and mortalities of all cancer types in 2020 worldwide. A: The proportions of incidences of all cancer types in 2020 worldwide; B: The proportions of mortalities of all cancer types in 2020 worldwide. Bar plots show the proportions of incidences or mortalities of all cancer types in both sexes, males and females, respectively. The data source for constructions were from website GLOBOCAN 2020 (https://gco.iarc.fr/today).

Recently, the molecular mechanisms correlated with ESCC have drawn much attention, and potential biomarkers are being investigated to facilitate clinical decision-making. However, the key mechanisms underlying ESCC pathogenesis have not been fully elucidated, and promising biomarkers for screening, early diagnosis, and evaluation of progression and prognosis have not yet been identified. This review aimed to discuss the epidemiological features of ESCC in different regions of the world, especially in China, and focused on novel molecular biomarkers associated with screening, early diagnosis and prognosis prediction in ESCC patients, thereby providing information for clinical decision-making.

GLOBAL EPIDEMIOLOGICAL FEATURES OF ESCC

Epidemiological features of ESCC worldwide

According to GLOBOCAN 2020 estimates, there were approximately 19292789 new cancer cases in 2020; EC accounted for 604100 of these cases, ranking eighth in incidence among all cancer types in 185 countries[1] (Figure 1A). The age-standardized incidence rate (ASIR) is highest in Asia (8.5/100000), followed by Africa, Europe, Oceania, and North America, and it is lowest in Latin America and the Caribbean (Figure 2A), suggesting that ASIRs in different geographic areas across the world can differ by more than 3-fold[1]. The three regions with the highest ASIRs in Asia are Mongolia (17.1/100000), Bangladesh (14.8/100000) and China (13.8/100000) (Figure 3A). Moreover, among all cancer types, EC ranks seventh in males (ASIR 9.3/100000) and ninth in females (ASIR 3.6/100000) in terms of incidence, and approximately 70% of the EC patients are males, suggesting a distinct sex distribution with more than a 2-fold difference[1] (Figures 1A and 2A). In addition, detailed information on the sex distribution of EC in Asia is shown in Figures 2A, 3B and 3C. The two most common histologic subtypes of EC, ESCC, and esophageal adenocarcinoma (EAC), have different etiologies[1]. There were nearly 512500 ESCC patients, 85700 EAC patients, and approximately 6000 patients with other subtypes of EC in 2020[1,13]. Since the majority of EC cases are ESCC[4,10,14], the sex distribution of ESCC is similar to that of EC in males and females, with incidence rates of 7.8/100000 in males and 3.2/100000 in females[13]. Moreover, regardless of the age range, both ESCC and EAC are more common in males than in females[1,15]. Although ESCC represents the most common subtype of EC in many areas worldwide, higher EAC incidence rates may be observed in some developed areas[4,10,14,16]. Specifically, southeastern and central Asia have the highest incidence rates of ESCC, which accounts for approximately 90% of all EC cases; only approximately 10% of EC cases are EAC. However, EAC usually appears in Western Europe, North America and Oceania[1,6]. ESCC and EAC not only have notable geographic differences but could also be affected by different risk factors. ESCC is usually affected by certain biological/genetic factors (such as race, age, and sex) and various environmental factors, including chemical carcinogen exposure, cigarette smoking and alcohol consumption, hot drinks, a diet with low fruits and vegetables, high consumption of pickled vegetables or processed meat, and low socioeconomic status[1,17-19] (Figure 4). Nevertheless, increased excessive body weight, gastroesophageal reflux disease and Barrett’s esophagus are the key risk factors for EAC[1,17,20,21].

Figure 2.

Estimated age-standardized incidence and mortality rates of esophageal cancer in 2020 worldwide. A: Estimated age-standardized incidence rates of esophageal cancer in 2020 worldwide; B: Estimated age-standardized mortality rates of esophageal cancer in 2020 worldwide. Bar plots show the estimated age-standardized incidence rates and estimated age-standardized mortality rates of esophageal cancer in both sexes, males and females, respectively. The data source for constructions were from website GLOBOCAN 2020 (https://gco.iarc.fr/today).

Figure 3.

Estimated age-standardized incidence and mortality rates of esophageal cancer in 2020 in Asia. A: Estimated age-standardized incidence rates of esophageal cancer in 2020 in Asia; B: Estimated age-standardized incidence rates of esophageal cancer in males in 2020 in Asia; C: Estimated age-standardized incidence rates of esophageal cancer in females in 2020 in Asia; D: Estimated age-standardized mortality rates of esophageal cancer in 2020 in Asia; E: Estimated age-standardized mortality rates of esophageal cancer in males in 2020 in Asia; F: Estimated age-standardized mortality rates of esophageal cancer in females in 2020 in Asia. A-F: They were constructed by six separate figures published on the website GLOBOCAN 2020. The data source is from GLOBOCAN 2020 (https://gco.iarc.fr/), and the map production is from IARC/WHO (https://gco.iarc.fr/today). IARC exercises copyright over its materials, and all rights are reserved. A request for permission to reproduce IARC copyrighted material was sent and a copyright permission to reprint and reproduce the IARC/WHO copyrighted materials in this paper is authorized by IARC/WHO. And a legal agreement between authors for this paper and IARC/WHO is acquired, granting authors a license to use the Licensed Materials subject to this paper herein (Supplementary material).

Figure 4.

Summary of risk factors associated with esophageal squamous cell carcinoma.

According to the GLOBOCAN 2020 database, EC was the sixth most common cause of cancer-related death, with 544076 related deaths in 2020[1] (Figure 1B). The International Agency for Research on Cancer demonstrated that one in every 18 cancer-related deaths was due to EC in 2020[1]. The age-standardized mortality rate (ASMR) of EC is highest in Asia (7.6/100000) (Figure 2B), and the three regions with the highest ASMRs in Asia are Mongolia (16.2/100000), Bangladesh (13.9/100000) and China (12.7/100000) (Figure 3D); additionally, EC is the leading cause of cancer-related death in Bangladeshi people and Malawian males[1]. Consistent with the findings concerning EC incidence, the mortality rate of EC is the sixth highest in males and the ninth highest in females among all cancer types, and there is also a more than 2-fold difference in the ASMR between males (8.3/100000) and females (3.2/100000)[1] (Figures 1B and 2B). In addition, detailed information about the sex distribution of EC in Asia is shown in Figures 2B, 3E and 3F. Due to the differences in pathological mechanisms, histological subtypes and respective management strategies, the survival rates differ between patients with ESCC and patients with EAC[22-25]. For instance, a comparative study of 185302 EC patients enrolled from 2010 to 2014 in seven countries, including Australia, Canada, Denmark, Ireland, New Zealand, Norway and the United Kingdom, and revealed that the 1-year survival rate of patients with EAC was greater than that of patients with ESCC in Norway (53.2% vs 40.0%), Denmark (50.7% vs 39.6%) and the United Kingdom (50.6% vs 43.4%)[24]. In addition, another comparative study recruited 93167 EC patients from 1973 to 2009 in the United States and concluded that EAC patients may have better long-term survival outcomes than ESCC patients[22].

In recent years, the incidence of EC has gradually declined in many areas, especially in Asian countries except for Japan[26]; however, it has increased in Africa[27]. While EAC incidence rates are increasing in some developed countries, ESCC is still more common, but its incidence rates are decreasing along with decreases in overall EC incidence[1,6,26,28]. It is speculated that the main cause of the decreasing trend in the incidence of ESCC is the significant decrease in cigarette smoking and alcohol consumption[6,28,29]. With the introduction of more effective preventative strategies according to risk factors in high-risk group patients, the incidence rate of EC might further decrease. Notably, if the incidence and mortality rates of EC remain unchanged, 957000 new cases (806000 with ESCC and 141300 with EAC) and 880000 deaths from EC are expected to occur in 2040 worldwide[1,13].

Epidemiological features of EC in China

In China, EC was the sixth most prevalent malignancy and the fourth leading cause of cancer-related death in 2020[1-3], and ESCC is the predominant histological subtype, accounting for nearly 90% of all EC cases[4,10,14]. Although China has prominently high incidence and mortality rates of EC worldwide, a downward trend in incidence and mortality rates was found from 1990 to 2019 based on data collected from the Global Burden of Disease Study 2019 Results database (https://vizhub.healthdata.org/gbd-results/). Specifically, the ASIR of EC decreased from 20.97/100000 in 1990 to 13.90/100000 in 2019, and the ASMR decreased from 22.08/100000 in 1990 to 13.15/100000 in 2019. In addition, with the same decreasing trend from 1990 to 2019, the number of new EC males was approximately 207923 (ASIR 21.94/100000), and the number of new EC females was approximately 70197 (ASIR 6.83/100000) in China in 2019 (Figure 5). Moreover, the GLOBOCAN 2020 database stated that the ASIR of EC in China was 13.8/100000 in 2020, and new cases in China accounted for more than 50% of all new EC cases worldwide. In addition, the ASMR of EC in China was 12.7/100000 in 2020, accounting for more than 50% of all EC-related deaths globally[1]. As in terms of the sex distribution of EC incidence and mortality worldwide, Chinese males had a greater ASIR and ASMR than females (Figure 5). Regarding geographic differences, it was demonstrated that the incidence of EC varied in different regions of China in 2016[30]. Specifically, the ASIR and ASMR of EC were the highest in Central China (ASIR 14.3/100000, ASMR 10.5/100000), followed by the eastern (ASIR 13.5/100000, ASMR 10.1/100000) and southwestern (ASIR 13.5/100000, ASMR 9.7/100000) regions of China, and Northeast China had the lowest ASIR of 4.6/100000 and ASMR of 3.9/100000[30]. Generally, the ASIR and ASMR of EC in rural areas are approximately twice those in urban areas (ASIR 15.0/100000 vs 8.2/100000, ASMR 11.0/100000 vs 6.2/100000)[30]. Undoubtedly, environmental factors potentially affect the occurrence and progression of EC[30]. The inequitable distribution of medical resources may also contribute to such regional differences[3]. Accordingly, China has implemented policies to ensure that the entire population can access certain basic health care depending on an equitable, efficient, affordable and effective health system[31,32]. Moreover, socioeconomic status is another important factor contributing to such differences among distinct geographic areas. Specifically, socioeconomic status can be evaluated according to education level, household size, appliance ownership, etc.[33]. Tran et al[34] reported that the risk of ESCC decreased as education level increased in the population living in Linxian, China, and people who completed middle school or higher education had a lower risk of ESCC (57%) than people who completed only primary school (78%). Several other studies also demonstrated that the risk of ESCC is inversely related to education level[35,36]. In addition, people who have more family members, perform high amounts of physical labor or own certain appliances, such as private cars, motors, washing machines and refrigerators, also have decreased ESCC risk[35]. A study involving 7763 EC patients in Taiwan, China, indicated that occupation and family income are inversely associated with ESCC incidence, and that diagnosis may be delayed in EC patients with lower socioeconomic status[37]. Conversely, higher socioeconomic status may be associated with longer survival[38]. However, it is difficult to reduce the morbidity and mortality rates through short-term overall economic improvements, and people’s awareness of and ability to prevent ESCC can improve quickly after appropriate education[39]. Although a general decreasing trend in the incidence and mortality rates of EC in China was observed from 1990 to 2020, the total numbers of new EC patients and EC-related deaths increased each year, mainly due to the increasing and aging population[40,41]. Furthermore, the estimated number of new EC patients in China would be 324422 for 2030 and 529621 for 2040, and the estimated number of EC-related deaths would be 416509 in 2030 or even 525362 in 2040, according to data calculations based on the GLOBOCAN 2020 database and expected future population data[1,40].

Figure 5.

Trends in age-standardized incidence and mortality rates of esophageal cancer in China from 1990 to 2019, according to the Global Burden of Disease Study 2019 database. A: Trends in age-standardized incidence rates of esophageal cancer in China from 1990 to 2019, according to the Global Burden of Disease Study 2019 (GBD 2019); B: Trends in age-standardized mortality rates of esophageal cancer in China from 1990 to 2019, according to the GBD 2019. Line charts show the trends in age-standardized incidence and mortality rates of esophageal cancer in China from 1990 to 2019 in both sexes, males and females, respectively. Source: Institute for Health Metrics and Evaluation. Global Burden of Disease Collaborative Network, Global Burden of Disease Study 2019 Results, Seattle, United States: Institute for Health Metrics and Evaluation, 2020. Available from https://vizhub.healthdata.org/gbd-results/.

SCREENING AND EARLY DIAGNOSIS OF ESCC

Due to the lack of specific early symptoms, 70% of ESCC patients are already in intermediate or advanced stages at the time of initial diagnosis[42]. The life expectancy and prognosis of advanced-stage ESCC patients are largely dependent on treatment strategies, local recurrence, distant metastasis and other factors[43,44]. Since the 1970s, the Chinese government has been screening for EC in high-risk areas, which has decreased EC mortality rates in these areas[45,46]. Additionally, since 1972, 17 large-scale screening initiatives in the rural areas of Henan, Hebei and northern Jiangsu provinces in China have screened 3000 participants with a high risk of upper gastrointestinal carcinomas from more than 160000 asymptomatic participants, and 757 patients were ultimately diagnosed with superficial ESCC; among these 757 ESCC patients, 420 underwent surgical treatment, and the survival rates increased[8]. Similarly, several other studies have demonstrated that endoscopic screening could significantly reduce the risk of EC-related death and prolong survival for early diagnosis and timely treatment[47,48]. Therefore, population-based cancer screening is the most effective approach, and accepted screening and early diagnostic strategies for improving the prognosis and survival of patients with EC are urgently needed. Notably, various factors are correlated with ESCC occurrence and progression, and the pathogenesis of ESCC is complex. Therefore, determining the underlying mechanisms is pivotal, and effective diagnostic strategies could be formulated to identify patients at increased risk of ESCC so that preventive measures can be implemented in a timely manner. Since early diagnosis could provide the optimal treatment window, more attention has been given to the clinical screening and early diagnosis of ESCC.

Endoscopy, cytology, histopathology and imaging are utilized for clinical screening and early diagnosis of ESCC, and upper gastrointestinal endoscopy with biopsy is recognized as the gold standard for early diagnosis of cancerous lesions. However, potential molecular biomarkers are available only for assisting in the screening and early diagnosis of ESCC. Cytokeratin-19-fragment (CYFRA21-1), carcinoembryonic antigen (CEA) and SCC antigen are traditional ESCC biomarkers and are considered to have better specificity than sensitivity[49,50]. Moreover, Xu et al[51] indicated the sensitivity of a panel of autoantibodies against six tumor-associated antigens, including p53, NY-ESO-1, matrix metalloproteinase-7 (MMP-7), heat shock protein 70 (Hsp70), peroxiredoxin VI and the BMI1 polycomb ring finger oncogene, for the early detection of ESCC was only 45% [95% confidence interval (CI): 32%-59%], with a specificity of 95% (95%CI: 89%-98%), and similar results could be obtained in the validation cohort. However, Ju et al[52] showed that the sensitivities of CEA and CYFRA21-1 for diagnosing ESCC were 80% and 88.89%, respectively, and the specificities were 53% and 58.5%, respectively. At present, certain genes, proteins, metabolites, circulating tumor cells (CTCs), microRNAs (miRNAs) and DNA methylation have been identified and considered new biomarkers with potential auxiliary diagnostic value[53,54]. Herein, we summarize the current genomics, proteomics and metabolomics studies related to ESCC, hoping to identify valuable molecular biomarkers of ESCC with high sensitivity, specificity and stability.

Genomics

In the human body, many proto-oncogenes play crucial roles in regulating cellular physiological processes; however, these proto-oncogenes can be activated to form oncogenes under certain circumstances, which may promote cancer occurrence and progression[55]. With the advent of high-throughput technologies such as next-generation sequencing, tremendous advances have been made in cancer genomics, which can reveal specific genes as characteristic molecular biomarkers to provide insights into the underlying pathogenesis of ESCC and the development of novel management strategies for ESCC[9,56].

Song et al[57] conducted a comprehensive genomic analysis of 158 ESCC patients; whole-genome sequencing was performed for 17 patients, whole-exome sequencing was performed for 71 patients, and array comparative genomic hybridization analysis was conducted for 123 patients of which 53 were selected from patients performed with whole-exome sequencing. Eight gene mutations were found to be significantly associated with ESCC, including six acknowledged tumor-associated genes [tumor suppressor gene (TP53), retinoblastoma susceptibility gene, cyclin-dependent kinase inhibitor 2A (CDKN2A), phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA), Notch homolog 1 (NOTCH1) and nuclear factor, erythroid 2, like 2 (NFE2L2)] and two genes [a disintegrin and metalloprotease domain 29 and family with sequence similarity 135 member B (FAM135B)] that have not been described in ESCC previously[57]. In another whole-exome sequencing analysis of cancerous and noncancerous esophageal tissues sampled from 144 ESCC patients in Japan, Sawada et al[58] reported a high proportion of gene mutations in these cancerous samples, with C to T substitutions in CpG dinucleotides (called the CpG signature) or C to G/T substitutions with a flanking 5’ thymine (called the APOBEC signature). These mutations may occur in genes that regulate the cell cycle [TP53, CyclinD1 (CCND1), CDKN2A, and F-Box and WD Repeat Domain Containing 7], epigenetic processes [mixed lineage leukemia 2 (MLL2), E1A-binding protein p300 (EP300), CREB binding protein, and ten-eleven translocation methylcytosine dioxygenase 2], or the NOTCH (NOTCH1 and NOTCH3), WNT [FAT atypical cadherin 1 (FAT1), Yes-associated protein 1, and ajuba LIM protein] and receptor-tyrosine kinase-phosphoinositide 3-kinase signaling pathways [PIK3CA, epidermal growth factor receptor (EGFR), and epidermal growth factor receptor-2][58]. Some of these genes, such as TP53, CUB and Sushi multiple domains 3 (CSMD3), nuclear envelope protein 1, PIK3CA, and NOTCH1, also exhibited frequent mutations in an Indian ESCC cohort[59]. Moreover, a Chinese study enrolling 225 ESCC patients indicated that TP53, CCND1, fibroblast growth factor 4 (FGF4), FGF19, FGF3, CDKN2A, PIK3CA, NOTCH1, histone-lysine N-methyltransferase 2D, FAT1, and low-density lipoprotein receptor-related protein 1B were the most commonly mutated genes in ESCC patients[60]. Another Chinese study revealed significant differences in the mutation frequencies of NOTCH1 and NOTCH2 and in the frequency of cancer-related Hippo, Wnt and nuclear factor E2-related factor 2 signaling pathways in ESCC tissues compared with paired normal esophageal mucosa[61]. Furthermore, it was reported that triple knockout of Trp53, Cdkn2a, and Notch1 could induce the neoplastic and immune evasion features of ESCC via CCL2 in murine esophageal organoids[62]. Therefore, the genes mentioned above, such as PIK3CA and NOTCH1, have become hotspots in ESCC-related studies.

In 2009, Akagi et al[63] conducted a genetic analysis of PIK3CA mutations in ESCC cells from patients who underwent surgery, and PIK3CA mutation and overexpression were used as detection indices for ESCC. Analysis of mutations in exons 9 and 20 of PIK3CA and increased copy numbers, mRNA and protein expression levels in ESCC samples indicated that PIK3CA could be a promising biomarker of ESCC and may also predict lymph node metastasis[63]. Another clinical study reported that PIK3CA mutations in exon 9 were independently associated with poor survival, and PIK3CA overexpression was correlated with recurrence and worse survival in these female ESCC patients[64]. However, certain opposite views exist. For instance, PIK3CA was shown to promote the proliferation and motility of Eca109 and EC9706 cells but was not correlated with lymph node metastasis or prognosis in ESCC patients[65]. In addition, PIK3CA mutations were reported to be associated with longer survival in ESCC patients with PIK3CA mutations[66].

Interestingly, many studies have shown that the Notch signaling pathway plays dual roles as a tumor-suppressive pathway and an oncogenic pathway in many cancer types, including ESCC[67-69]. Previous data demonstrated that Notch1 could inhibit proliferation and induce apoptosis through the Wnt signaling pathway in EC9706 cells[70], and that the loss of Notch signaling may lead to ESCC[71,72]. In addition, mutations in NOTCH1 and the Notch pathway are common in the preclinical and early stages of ESCC[73]. In contrast, several different opinions have indicated that Notch signaling may have a carcinogenic effect. For example, Notch1 promoted the growth of xenograft tumors and the initiation of ESCC in concert with transforming growth factor-β[74], and upregulated Notch4/Hey1 signaling may also induce proliferation and promote epithelial mesenchymal transition in ESCC cells[75]. Therefore, it remains to be elucidated whether the activation or inhibition of NOTCH signaling prevents ESCC occurrence and progression.

In addition, several other genes may participate in the occurrence, progression and prognosis of ESCC. CSMD3 mutations may correlate with a better prognosis in Asian ESCC patients, while mutations in TP53, EP300, and NFE2L2 are more likely to promote ESCC[76]. Moreover, the long noncoding RNA (lncRNA) SNHG1 was found to be significantly upregulated in ESCC tissues, and SNHG1 overexpression was strongly correlated with invasion depth, lymph node metastasis and tumor-node-metastasis (TNM) stage, as well as reduced survival time in ESCC patients[77]. In conclusion, the elucidation of cancer-related gene mutations could provide new perspectives for the screening and early diagnosis of ESCC, and even lead to the development of treatments targeting the verified characteristic gene mutations.

Proteomics

Proteins are the executors of functional gene transcription within cells, and proteomic analysis could reveal valuable biomarkers that are mostly relevant to functional cellular states[78]. As an important novel technique, proteomics can be used to identify differentially expressed proteins in ESCC tissues compared with paired normal esophageal tissues, to identify effective protein biomarkers of ESCC[79]. Proteomics was performed to explore the expression profiles of serum samples from ESCC patients; the concentrations of serum Hsp70 autoantibodies were found to be increased in ESCC patients compared with controls[80]. As another member of the HSP family, Hsp27 has been proven to be highly expressed in ESCC tissues[81,82]. Moreover, Hsp27 expression in ESCC tissues is positively correlated with lymphatic metastasis and a poor prognosis[81,82]. Furthermore, marked upregulation of Hsp90 and its target proteins in cancerous lesions has also been demonstrated in an immune-competent mouse model of ESCC[83].

Recently, Liu et al[84] demonstrated that ENO1, TPI1, PGAM1, SAA1 and S100A8/A9 are potential biomarkers of ESCC with high diagnostic sensitivity and specificity. Among these proteins, ENO1, TPI1 and PGAM1 are key regulators of glycolysis, and SAA1 and S100A8/A9 are widely believed to play a role in inflammation[84]. Glycolytic pathways have been extensively studied and are generally recognized to correlate with ESCC cell proliferation and tumor occurrence[85,86]. In addition, glycolysis-related proteins, such as PKG1, have also been found to be overexpressed during ESCC progression[87]. Therefore, the joint monitoring of such glycolysis-related proteins may be beneficial for the screening and diagnosis of ESCC.

Moreover, abnormal glycosylation has been demonstrated to be associated with the occurrence, progression and metastasis of cancers, and glycoproteomics has been applied to various cancers, such as breast cancer, prostate cancer and ESCC[88,89]. An ESCC proteomic study revealed increased expression of alpha-2-HS-glycoprotein and leucine-rich alpha-2-glycoprotein in the plasma of ESCC patients, and these proteins may be used as potential biomarkers for the early diagnosis of ESCC[90]. N-linked glycoproteomic profiling analysis indicated that the expression of total procathepsin D and high-mannose procathepsin D was upregulated, while the expression of total haptoglobin, high-mannose clusterin and the GlcNAc/sialic acid-containing fraction of 14-3-3ζ was downregulated in ESCC tissues[91]. In addition, Shu et al[92] reported that the proportion of ESCC patients with GlcNAc or Galβ1-4GlcNAc-containing N-glycans was increased. Similarly, the expression of glycosylated proteins was distinct between the saliva of ESCC patients and that of healthy controls, and screening for ESCC biomarkers in saliva is noninvasive and may be a more acceptable approach[92]. In addition to screening and early diagnosis, glycosylated proteins may also be associated with the metastasis and prognosis of ESCC. For instance, by conducting a dynamic profiling analysis of glycoprotein changes during lymph node metastasis progression, Gao et al[93] showed that integrin beta-1 and CD276 may contribute to the occurrence and progression of ESCC. In conclusion, proteomic analysis not only is helpful for screening and early diagnosis of ESCC, but is also valuable for predicting ESCC progression and prognosis. In addition, different types of proteomic analysis, such as plasma proteomics, glycosylated proteomics and phosphoproteomic analysis, may provide additional comprehensive information from another perspective for ESCC management.

Metabolomics

With the clinical application of metabolomics analysis, it is possible to search for certain metabolic features that are valuable for assessing ESCC pathological states. In such studies, we can identify characteristic metabolites, such as amino acids (AAs), lipids, carbohydrates, nucleotides, etc., as well as metabolic pathways that are associated with ESCC by analyzing changes in metabolites in the human body. Several metabolomics studies have shown that ESCC occurrence is correlated with the abnormal metabolism of AAs and fatty acids[94-97]. In general, AA metabolism is upregulated in ESCC patients. For instance, Zhang et al[94] demonstrated that the levels of eight serum AAs were greater in ESCC patients than in healthy controls, and that the levels of four serum AAs were greater in ESCC patients than in subjects with esophageal squamous dysplasia (ESD). Specifically, the panel of propanoic acid, l-leucine and hydroxyproline for the screening of ESD revealed the area under the curve (AUC), sensitivity and specificity values of 0.819, 0.76 and 0.72, respectively; while, the combination of hypoxanthine, 2-ketoisocaproic acid, l-glutamate and l-aspartate had similar AUC, sensitivity and specificity values of 0.818, 0.83, and 0.74, respectively, in distinguishing ESCC patients from ESD patients[94]. In addition, Zhao et al[96] selected a panel of serum metabolites, including tryptophan (Trp), citrulline, L-carnitine, lysine and acetylcarnitine, and verified that this panel has great accuracy in distinguishing early-stage ESCC patients from controls. However, the changes in fatty acid metabolism were not consistent with the changes in AA metabolism. Zang et al[95] reported that the levels of seventeen free fatty acids were obviously increased in ESCC tissues, while two other studies reported decreased fatty acid levels in the serum of ESCC patients compared to those in the controls[96]. Thus, AA metabolites may be better biomarkers for ESCC than fatty acids. Furthermore, the metabolism of phospholipids, especially glycerophospholipids, has been demonstrated to be associated with cancer occurrence and progression[98,99]. Focusing on metabolic profiling in plasma, Liu et al[100] showed that six phospholipids, including phosphatidylserine, phosphatidic acid, phosphatidyl choline, phosphatidylinositol, phosphatidyl ethanolamine and sphinganine 1-phosphate, were upregulated in ESCC; and Yang et al[101] reported that phosphatidylserine synthase 1 and lysophosphatidylcholine acyltransferase 1 levels in tissues may have good diagnostic value for ESCC.

In addition to screening and early diagnosis, metabolomics may also play a guiding role in determining the association between lymph node metastasis and the prognosis of ESCC. By comparing the metabolomic differences between ESCC patients with or without lymph node metastasis, Jin et al[102] reported that ESCC patients with lymph node metastasis had multiple metabolic abnormalities, and three serum metabolites, valine, γ-aminobutyric acid and pyrrole-2-carboxylic acid, may have greater diagnostic value for ESCC metastasis. In addition, another clinical study[103] demonstrated that the ratios of the concentrations of serum kynurenine, 5-hydroxytryptophan, 5-hydroxyindole-3-acetic acid and 5-hydroxytryptamine to the concentration of their precursor Trp can be used not only as screening biomarkers to facilitate early ESCC diagnosis, but also as metabolite biomarkers to distinguish ESCC patients with metastasis from those without metastasis. Moreover, Zhang et al[94] reported a significant accumulation of kynurenine in ESCC tissues, and indicated that numerous other metabolites, including glutamate, oleic acid, lysoPC(15:0), uracil, inosine choline and serine, may also be related to tumor invasion, lymph node metastasis and survival. In addition, the combination of plasma kynurenine and LPC(14:0)sn-1 could also be used to assess the risk of lymphatic metastasis in ESCC patients[104]. Among these metabolites, kynurenine, a downstream metabolite of Trp, was shown to have a satisfactory predictive accuracy of 0.84 as a diagnostic biomarker for ESCC in tissues[105], and high plasma kynurenine levels may be correlated with a poorer prognosis in patients with ESCC[104]. The occurrence and progression of cancers are related to changes in various metabolites in the serum, plasma, other fluids or tissues. Metabolomic analysis might reveal molecular biomarkers that are closely associated with cancer development and could be used to predict cancer occurrence, progression and prognosis.

Other potential biomarkers

After shedding from the original or metastatic tumor, CTCs circulate in the peripheral blood and may further disseminate to form metastases[106]. The presence of CTCs in ESCC patients may indicate tumor differentiation, metastasis or recurrence, and usually predicts a poor prognosis; however, CTCs are of little significance for ESCC screening[107,108]. miRNAs have also been proven to be potential biomarkers for early diagnosis and prognosis evaluation in ESCC patients[109,110]. A meta-analysis of data from 995 ESCC patients and 733 healthy controls aimed to assess the potential diagnostic value of certain miRNAs, including miR-10a, miR-22, miR-100, miR-148b, miR-223, miR-133a, miR-127-3p, miR-21, miR-375, miR-31, miR-18a, miR-1246, miR-144, miR-10, miR-451, and miR-1322, and demonstrated pooled AUC, sensitivity and specificity values of 0.91 (95%CI: 0.88-0.93), 0.81 (95%CI: 0.76-0.85) and 0.83 (95%CI: 0.76-0.88), respectively, suggesting the excellent diagnostic value of miRNAs for ESCC[111]. However, miRNAs have certain limitations, such as instability and rapid degradation[112]. In addition, only a few studies have aimed to distinguish the characteristic expression of miRNAs in ESCC compared with that in other gastrointestinal cancers, so the available data are not sufficient to reach any definite conclusions.

Notably, aberrant DNA methylation is another potential diagnostic and prognostic biomarker for various malignancies, and usually plays an important role in the occurrence and progression of many cancers including ESCC[113-115]. Generally, the global genomic DNA hypomethylation and promoter region hypermethylation are the major alterations associated with aberrant DNA methylation[114,116,117]. For instance, long interspersed transposable element-1 (LINE-1), which constitutes a substantial portion (approximately 17%) of human genome[118], is considered as valuable indicators of genome-wide methylation levels[119-122]. LINE-1 hypomethylation is co-related with chromosomal instability in ESCC tissues, and was also found to be positively associated with lymphatic and venous invasion, as well as lymph node metastasis[119]. In addition, LINE-1 hypomethylation was detected in the noncancerous esophageal mucosa of ESCC patients with a history of smoking[123]. The expression of lysyl oxidase (LOX) was upregulated in ESCC along with LINE-1 hypomethylation, and patients with high LOX expression may have a significantly shorter overall survival[124]. Moreover, abnormally hypermethylated promoters usually appear in many genes involved in the cell cycle, DNA damage repair and cancer-related signaling pathways[116,125]. The increased methylation frequencies of human mismatch repair genes, including O6-methylguanine-DNA methyltransferase, mut-L homolog 1, and mut-S homolog 2, may correlate with ESCC development[126,127], and certain cell cycle-related genes, such as CDKN2A and ras association domain family 1 isoform, also have significantly hypermethylated sites in ESCC[122,126,128]. In addition, SRY-box containing gene 17 (SOX17) may be silenced by frequent methylation in EC cell lines, esophageal dysplasia and ESCC[129], and SOX17 methylation combined with other methylation biomarkers could be used to predict the prognosis of ESCC or the response of ESCC patients to chemoradiotherapy[130,131]. Furthermore, some hypomethylated promoters may also affect ESCC progression. Lu et al[132] reported that the hypomethylation of CpG sites upstream of the miR-10b-3p gene may upregulate the expression of miR-10b-3p in ESCC tissues, and miR-10b-3p was shown to contribute to ESCC progression. Although many studies have focused on the screening and prognostic implications of methylation in ESCC, only a few of these markers have been verified as markers for routine clinical practice[133]. Moreover, due to the lack of standardization in methylation studies, certain inconsistencies exist and cannot be ignored[134].

Although promising, current clinical studies evaluating biomarkers for ESCC screening and diagnosis have not verified any definite sensitive or specific biomarkers that could be applied in clinical practice, and large-scale multicenter omics analysis is still needed and should be conducted in various geographic areas, including Eastern and Southeastern Africa, Northeastern Iran, South America and Central Asia, to provide complete information on structural features, nuclear mutation signatures and other important information[135-137]. Moreover, more effort should be made to identify specific driver biomarkers for ESCC diagnosis, disease stratification, prognosis evaluation, and to access strategies related to certain therapeutic targets, such as human epidermal growth factor receptor 2 in breast cancer, v-kit Hardy-Zuckerman 4 feline sarcoma viral oncogene homolog and platelet-derived growth subunit A in gastrointestinal stromal tumors, and EGFR, echinoderm microtubule-associated protein-like 4 gene-ALK variant and reactive oxygen species in lung cancer[138]. Moreover, the combination of various omics studies may facilitate personalized risk-stratification and therapeutic strategy development for each patient[133,139,140]. For instance, Huang et al[141] found that nicotinate and nicotinamide metabolism were dysregulated in ESCC patients with lymph node metastasis by combining single-cell RNA sequencing data with metabolomics in ESCC tissues and plasma samples; and these authors also found that N-methyltransferase, which regulates nicotinamide metabolism, may promote the epithelial-mesenchymal transition (EMT) and metastasis of ESCC in vitro and in vivo. Undoubtedly, multiomics analysis could be applied to systematically characterize the full landscape of molecular alterations involved in disease occurrence, development and progression, and could play a vital role in precision medicine in the future[87,142].

PROGNOSIS OF ESCC

According to the GLOBOCAN 2020 database, EC accounted for 3.1% of new cancer patients and 5.5% of all cancer-related deaths in 2020[1]. The prognosis of EC is usually affected by various factors, among which the clinical TNM stage at initial diagnosis and therapeutic strategies are the most pivotal factors[143,144]. For ESCC patients diagnosed at stage I or stage II, surgical resection is usually considered the standard treatment and is associated with a better prognosis[145]. However, according to the depth of tumor invasion and lymph node metastasis, adjuvant radiotherapy or chemotherapy may be adopted to improve therapeutic efficacy[146]. For patients with advanced ESCC (stage III or IV), acknowledged therapeutic strategies usually include surgery with radiotherapy, chemotherapy or a combination of both[147]. Notably, for ESCC patients with T1a or superficial T1b tumors less than 3 cm in diameter and without clear lymphovascular invasion or poorly differentiated histology, endoscopic resection could be introduced, with chemotherapy serving as an adjunctive therapy to surgery[148]. In recent decades, targeted therapies have been confirmed to play an essential role in the treatment of ESCC[149,150], and may benefit difficult-to-treat patients by extending their life expectancy[151]. For instance, by multiomics analysis, Zhang et al[152] reported the promising effect of neoadjuvant therapy with toripalimab, nab-paclitaxel and S-1 in ESCC patients, and indicated that the integration of CD4⁺ T cells with plasma FN-γ, Gal.1 or LAMP3 could facilitate the discrimination of responders from nonresponders. Notably, the establishment of patient-origin esophageal 3D organoids may facilitate the discovery and validation of novel translational applications in ESCC[153,154]. For instance, Kijima et al[155] found that ESCC cells with high CD44 expression undergoing autophagy may be resistant to 5-fluorouracil treatment in 3D culture organoid models. Despite the current evidence, personalized treatment of ESCC is still a major challenge[156]. Furthermore, immunotherapies, including immune checkpoint inhibitors (ICIs), therapeutic vaccines, monoclonal antibodies and adoptive cellular immunotherapy, have shown promising results in treating advanced ESCC[147,157]. For instance, programmed cell death ligand 1 (PD-L1) is an immune checkpoint molecule that can suppress the immune response after binding to programmed death receptor 1 (PD-1), thereby allowing tumors to evade immune attack[158,159]. High PD-L1 expression is generally associated with a poor prognosis in ESCC patients[160], and a targeted next-generation sequencing study revealed that approximately 36.7% of 139 Chinese patients with ESCC expressed high levels of PD-L1[161]. ICIs targeting PD-L1 are already widely used to treat numerous cancers, including ESCC, to improve patient prognosis[162]. A recent randomized, double-blind phase 3 clinical trial demonstrated that the progression-free survival and overall survival of previously untreated, PD-L1-positive advanced ESCC patients could be significantly improved by treatment with serplulimab plus PD-1 blockade administered every 2 wk[163]. Similarly, another randomized phase 2 clinical study concluded that neoadjuvant socazolimab (a PD-L1 inhibitor) combined with chemotherapy may be associated with a promising major pathological response and pathological complete response rates and significant T downstaging in locally advanced ESCC without increasing surgical complication rates[164]. In addition, it was reported that peptide vaccines could stimulate tumor-specific immune responses[165]. The combination of a multiple-peptide vaccine and chemoradiation could prolong the survival of patients with HLA-A*2402-positive unresectable ESCC[166]; and three peptide vaccines, RYCNLEGPPI, VYGIRLEHF and KTVNEQNL, may decrease the recurrence rate significantly and improve the survival of ESCC patients[167]. Furthermore, various other important factors affecting the prognosis of ESCC include the degree of tumor differentiation, the presence of comorbidities, and the patient response to formulated treatments[168].

Several biological factors, such as genetic variations, can also significantly influence the prognosis of ESCC[169]. In recent years, researchers have identified various genetic biomarkers that may be used to assess the prognosis of ESCC and facilitate clinical decision-making[170]. For example, TP53, a tumor suppressor gene, is the most commonly mutated gene in ESCC, and is recurrently mutated in precancerous and malignant lesions of ESCC[171-174]; moreover, TP53 mutations are related to a poor prognosis in ESCC patients[171]. Efe et al[175] established a p53-R172H-dependent BRD4-CSF-1 axis that may promote ESCC metastasis, indicating that gain-of-function mutation of p53 could induce invasion and lung metastasis in ESCC patients. Feng et al[176] reported that the mutant p53 G245S (p53-G245S) could promote ESCC progression by activating hnRNPA2B1/AGAP1-mediated exosome formation. Cripto-1 (CR-1) is an epidermal growth factor-related protein that plays an essential role in embryonic development and tumor growth[177,178]. Liu et al[179] reported that increased expression of CR-1 is associated with increased expression of ESCC cell stemness-related genes, such as Sox2, Oct4 and Nanog, and EMT-related genes, including vimentin, snail and MMP-9. Moreover, the expression level of CR-1 in cancerous tissues was positively correlated with TNM stage, tumor invasion depth and lymph node metastasis[179]. Additionally, OV6, an epithelial marker of origin[180], was demonstrated to be a potential marker of metastasis and a valuable indicator of ESCC prognosis, and OV6⁺ ESCC cells may have properties similar to those of stem cells[181]. Specifically, OV6⁺ ESCC cells exhibit increased expression of stem cell-related genes, increased self-renewal ability, increased degrees of tumorigenicity, chemotherapeutic drug resistance, invasion and metastasis[181]. Yang et al[182] reported that leucine zipper-EF-hand containing transmembrane protein 1 (LETM1 protein) was expressed at significantly greater levels in ESCC tissues than in adjacent normal esophageal tissues, and was closely correlated with primary tumor stage, clinical stage, overall survival and disease-free survival. Additionally, LETM1 expression was positively correlated with the expression of markers of cancer stem-like cells, including LSD1, CD44 and OCT4[182]. In addition, LETM1 can bind to kinesin family member 14 (KIF14), regulate the expression of KIF14, and subsequently reduce the proliferation, invasion, migration and angiogenesis of ESCC cells[183]. Dong et al[184] reported that the protein levels of the FAM135B in ESCC tissues were significantly greater than those in adjacent noncancerous tissues, and higher expression of FAM135B was associated with poor clinical prognosis in ESCC patients. Abnormal FAM135B expression may promote the proliferation of ESCC cells in vivo and in vitro, and significantly affect ESCC progression through its interaction with the growth factor granuline, suggesting that FAM135B is a potential therapeutic target and prognostic factor for ESCC[184]. Furthermore, the mutations in PIK3CA discussed above may also be related to survival in ESCC patients and may be prognostic biomarkers[66].

Notably, various ncRNAs, including lncRNAs, circular RNAs and miRNAs, are related to the prognosis of ESCC[185]. Liu et al[186] demonstrated that increased expression of LINC02096 (RIME) in plasma exosomes may be associated with poor prognosis in ESCC patients, indicating that the RIME is a potential prognostic biomarker of ESCC; in addition, the RIME-MLL1-H3K4me3 axis may play an essential role in tumor immunosuppression in ESCC. Moreover, the lncRNA prostate cancer-associated transcript 6 (PCAT6) was shown to be significantly upregulated in ESCC, and high expression of PCAT6 may lead to poor prognosis in ESCC patients[187]. Luo et al[188] also showed that PCAT6 could accelerate the progression of ESCC by activating the JAK/STAT signaling pathway. Li et al[189] showed that a high expression level of circMMP1 is related to a poor prognosis in ESCC patients, and further study has indicated that circMMP1 may enhance ANO1 expression by acting as a miR-671-5p sponge, thereby accelerating ESCC progression. Additionally, circ8199 may significantly inhibit the proliferation of ESCC cells by regulating the JAK2-STAT3 signaling pathway through OGT, and may affect the prognosis of ESCC[190]. Hsa_circ_0026611 was reported to be highly expressed in serum exosomes from ESCC patients, and was positively correlated with lymph node metastasis and poor prognosis in ESCC patients[191]. Further studies revealed that circ_0026611 may inhibit the acetylation and ubiquitination of PROX1 to promote lymphangiogenesis in ESCC[192]. Moreover, miRNAs also participate in the prognosis of ESCC, and Sindhoo et al[193] developed a dynamic database named ESOMIR (miRNA in EC) (https://esomir.dqweilab-sjtu.com), which consists of genes and miRNAs associated with EC. ESOMIR allows users to search for miRNAs, sequences, chromosomal positions, target genes and signaling pathways to acquire EC-related information, and it may be a valuable tool for identifying molecular biomarkers of ESCC[193]. In conclusion, the prognosis of ESCC is affected by various factors, including TNM stage, therapeutic strategy and tumor differentiation. New molecular biomarkers still need to be discovered, and optimal strategies should be investigated to facilitate prognosis prediction and to improve the outcomes of ESCC patients.

CONCLUSION

As an aggressive squamous epithelial malignancy derived from the esophagus, ESCC is the major histological subtype of EC and has a decreasing trend in incidence and mortality rates worldwide; however, it remains a global health issue and was the eighth most prevalent cancer and the sixth leading cause of cancer-related death in 2020. Due to the lack of specific early symptoms, many ESCC patients are diagnosed with advanced-stage disease, which has a poor prognosis; moreover, the incidence and mortality rates of ESCC are still high in many countries, especially in China. ESCC is a complicated and heterogeneous disease, and its initiation involves multiple stages. The histopathological progression is summarized as follows: Normal squamous epithelium, basal cell hyperplasia, mild dysplasia, moderate dysplasia, severe dysplasia, carcinoma in situ and invasive carcinoma[143]. Recent studies have focused on the key molecular mechanisms underlying ESCC pathogenesis, and certain promising biomarkers are being investigated to facilitate ESCC screening, early diagnosis and prognosis prediction. With the advent and advancement of high-throughput technologies, such as genomics, transcriptomics, proteomics and metabolomics, various valuable molecular biomarkers with satisfactory sensitivity, specificity and stability have been identified to be involved in ESCC development. Certain molecular changes, such as gene mutations, transcriptional changes or metabolite alterations, may persist with an increasing trend throughout ESCC development, or may occur at specific points during the occurrence and progression of ESCC[173]. Therefore, we may find specific biomarkers that could guide ESCC screening, early diagnosis and prognosis prediction, and future clinical studies are urgently needed to further verify these promising molecular biomarkers and to reveal novel strategies for ESCC patients.

ACKNOWLEDGEMENTS

We appreciate the works by the International Agency for Research on Cancer/World Health Organization and thank for the permission to re-use related data and re-print related illustrations. We appreciate the works by the Institute for Health Metrics and Evaluation and the Global Burden of Disease study 2019 collaborators, and thank for the permission to re-use related data.