Comparative genomic analysis of esophageal squamous cell carcinoma and adenocarcinoma: New opportunities towards molecularly targeted therapy

Abstract

Esophageal cancer is one of the most lethal cancers worldwide because of its rapid progression and poor prognosis. Esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC) are two major subtypes of esophageal cancer. ESCC predominantly affects African and Asian populations, which is closely related to chronic smoking and alcohol consumption. EAC typically arises in Barrett's esophagus with a predilection for Western countries. While surgical operation and chemoradiotherapy have been applied to combat this deadly cancer, molecularly targeted therapy is still at the early stages. With the development of large-scale next-generation sequencing, various genomic alterations in ESCC and EAC have been revealed and their potential roles in the initiation and progression of esophageal cancer have been studied. Potential therapeutic targets have been identified and novel approaches have been developed to combat esophageal cancer. In this review, we comprehensively analyze the genomic alterations in EAC and ESCC and summarize the potential role of the genetic alterations in the development of esophageal cancer. Progresses in the therapeutics based on the different tissue types and molecular signatures have also been reviewed and discussed.

Graphical abstract

Esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC) are two subtypes of esophageal cancer with diverse epidemiology and pathogenesis as well as different molecular profiles. This review analyzes the genomic alterations in ESCC and EAC and summarizes the potential therapeutic targets of ESCC and EAC.

1. Introduction

Carcinoma of the esophagus is the seventh most common cancer and the sixth leading cause of cancer-related mortality in the world¹. Esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC) are two major subtypes of esophageal cancer. ESCC represents the majority of esophageal cancer worldwide, especially in Asian and African populations and is thought to be associated with diet habit and exposure to carcinogens². In contrast, EAC which develops from intestinal metaplasia of the esophageal epithelium due to chronic gastroesophageal reflux disease is the predominant subtype of esophageal cancer in Western countries³. Esophageal cancer is known for its rapid progression and poor outcomes. As esophageal cancer is often asymptomatic in its early stages, the disease is usually at an advanced stage at the time of diagnosis and the prognosis of esophageal cancer is poor with the overall five-year survival rate from 15% to 25%⁴. However, the options for the treatment of esophageal cancer are limited at present. Endoscopic or surgical treatment is applied to early-stage patients, while radiotherapy or/and chemotherapy is predominant for patients with advanced or metastatic cancer⁵. Therefore, it is urgent to develop new therapies targeting the essential factors involving in the initiation and progression of esophageal carcinoma.

Although emerging evidence shows that ESCC and EAC are two different diseases, molecular mechanism of tumorigenesis for each subtype remains unclear. Massively parallel, high-throughput next-generation DNA sequencing (NGS) has enabled comprehensive characterization of somatic mutations and copy number variations (CNVs) in a large scale of samples, which provides new clues and targets for the precise therapy of ESCC and EAC.

In this review, we summarize the recent progress in genomic and molecular characterization of ESCC and EAC. We also propose the opportunities to target the functional alterations for the therapy of ESCC and EAC.

2. Genetic alterations in ESCC and EAC

In light of the molecular events in the tumorigenesis, molecularly targeted therapies have been successfully employed for the treatment of tumors with identified driver genes, including non-small cell lung cancer (EGFR, ALK), melanoma (BRAF), renal cancer (VEGFR), hepatocellular cancer (PDGFRB), breast cancer (HER2, CDK4, CDK6), chronic lymphocytic leukemia (BTK), ovarian cancer (PARP) and gastrointestinal cancer (VEGFR, HER2)⁶. Esophageal cancer has been benefitted little from the targeted therapies due to the unclear pathogenesis. Therefore, defining genomic landscape of esophageal cancer would gain insights into the initiation and development of esophageal cancer and facilitate to develop molecularly targeted therapies. NGS has revolutionized cancer genomic research by providing a comprehensive profiling of cancer genome, including point mutations, insertions, deletions and copy number variations, making it a powerful tool to identify potential driver genes or pathways of human cancer⁷. In recent years, several NGS studies have been performed in esophageal cancer patients across the world. The cBioPortal for cancer genomics (https://www.cbioportal.org) is an open platform integrating cancer genomics data from multiple resources with high credibility. We obtained the genomic data of ESCC and EAC samples from cBioPortal (Supporting Information Tables S1–S4) and compared their genomic profile in an effort to identify potential therapeutic targets. The top 15 cancer-related genes with highest rate in mutation or copy number variation in ESCC or EAC were presented in Fig. 1. The genes in red have been reported to be associated with esophageal cancer and are or potentially are therapeutic targets for esophageal cancer. As shown in Fig. 1, TP53, PIK3CA and KMT2D are commonly mutated (Fig. 1A), and copy number variations in CCND1, FGF3, FGF4, FGF19, CDKN2A and CDKN2B are frequently found (Fig. 1B) in both subtypes of esophageal cancer, while distinct genetic alterations were also found in ESCC and EAC. Here, we summarize these cancer-related genes and outlined their potential functions in ESCC and EAC.

Figure 1.

Frequent genetic alterations in human esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC). The data of genetic alterations in human ESCC and EAC were collected from cBioPortal, www.cbioportal.org, accessed in April 2021 (Tables S1–S4). The top 15 cancer-related genes with highest rate in mutation (A) or copy number variation (B) in ESCC and EAC were presented as a Venn diagram. Genes in red have been reported to be associated with esophageal cancer and are or potentially are therapeutic targets. ∗EGFR has been verified as a promising target to treat ESCC though it is not among the top 15 altered genes.

2.1. Genetic alterations with similar frequency in ESCC and EAC

By comparing the frequency of genetic alterations between ESCC and EAC, we found that TP53, CDKN2A, MYC and KMT2C are commonly altered with similar frequency in both subtypes of esophageal cancer (Fig. 2), reflecting deregulation in genomic stability and cell cycle progression in both ESCC and EAC (Fig. 3).

Figure 2.

Comparison of genetic alterations between ESCC and EAC. The frequency of alteration of the genes in red presented in Fig. 1 was compared between ESCC and EAC according to the data from cBioPortal. SF: similar frequency. If the frequency of alteration in a gene is greater than 2-fold (≥2) in ESCC compared to that in EAC, the gene is assigned to ‘ESCC > EAC’. If the frequency of alteration in a gene is greater than 2-fold (≥2) in EAC compared to that in ESCC, the gene is assigned to ‘ESCC < EAC’. If the difference in the frequency of alteration in a gene is less than 2-fold, the gene was considered as ‘similar frequency’ in both ESCC and EAC. (A) Frequency of mutations. (B) Frequency of copy number variations.

Figure 3.

Significantly dysregulated pathways in human ESCC (A) and EAC (B). The most altered genes were classified into three signaling pathways. The frequency of genetic alterations in ESCC and EAC were retrieved from cBioPortal (www.cbioportal.org, accessed in April 2021.). (a) RTK–MAPK–PI3K signaling. (b) Cell cycle regulation. (c) Epigenetic modulation.

2.1.1. TP53

The TP53 gene, localizing on chromosome 17p13.1, encodes a tumor suppressor protein responding to diverse cellular stresses to ensure cell and tissue homeostasis⁸. Somatic mutations of TP53 are one of the most common alterations in human cancer, and about 80% of esophageal cancer patients carry TP53 mutations (Figure 2, Figure 3). Most of them are hotspot mutations in DNA-binding domain⁹ (codons 175, 245, 248, 273, and 282), and truncated mutation at R342X in P53 tetramerization motif. These mutations abrogate the tumor suppressive function of P53, resulting in unstable genome. A recent study indicated that G245C and R273H mutants played important roles in tumorigenesis of ESCC, which might be exploited for diagnosis and therapy of patients carrying the mutants¹⁰. In a multivariable analysis in 161 patients with resectable ESCC, researchers revealed that the non-disruptive mutation in TP53 DNA-binding domain is a potential independent prognostic factor indicating prolonged survival and patients with elevated P53 expression showed a better outcome¹¹. A genomic analysis of Barrett's esophagus (BE) patients found significantly higher numbers of TP53 mutations in BE patients with progression to EAC¹². Taken together, these results suggest TP53 mutations are closely linked to the pathogenesis of esophageal cancer, which could be used as potential prognosis biomarker dependent on disease contexts.

2.1.2. CDKN2A

CDKN2A encodes a cyclin-dependent kinase inhibitor p16 (Ink4a), a tumor suppressor that regulates the transition of cell cycle from G1 to S phase. The CDKN2A gene is located within the frequently deleted region of chromosomal 9¹³. Loss of p16 is common in variant cancers. Inactive or decreased p16 fails to bind to cyclin D, which assists cyclin-dependent kinase 4/6 (CDK4/6) in phosphorylating Rb protein, and results in loss of brake in cell cycle progression. CDKN2A deletion is found in about 63% ESCC (Figure 2, Figure 3A). A comprehensive analysis of whole-genome sequencing (WGS) of 31 ESCC tumors and paired normal samples found that CDKN2A was deleted in 29 out of 31 ESCCs¹⁴. Homozygous deletion and promotor methylation of CDKN2A were also found to occur in 86% of a cohort of 21 cases of ESCC¹⁵. In addition, loss-of-function mutations of CDKN2A are detected in about 17% EAC patients (Figure 2, Figure 3B). A whole exome sequencing from 25 pairs of EAC and BE revealed that the progression of BE often emerged with the inactivation of CDKN2A¹⁶. Furthermore, researchers have verified that methylation of the CDKN2A promotor is the predominant mechanism of p16 inactivation, which is a very common event in EAC and occurs as early as metaplasia¹⁷.

2.1.3. MYC

The MYC gene, located on chromosome 8q24.21, encodes a nuclear phosphoprotein that plays important roles in cell cycle progression, apoptosis and cellular transformation. The amplification of MYC occurs frequently in a variety of human cancers, contributing to tumor progression and indicating poor outcome¹⁸. According to the data from cBioPortal, MYC amplification occurs in nearly 20% esophageal cancer patients (Figure 2, Figure 3). MYC amplification was detected in the normal mucosa adjacent to the tumor in 23% of a group of esophageal cancer patients from high-risk region of China¹⁹. High expression level of MYC was significantly correlated with poor prognosis and also associated with invasion and lymph node metastasis in ESCC patients²⁰. The amplification of MYC is also frequently detected in EAC, which is likely to be an early event during EAC carcinogenesis²¹.

2.1.4. KMT2C

KMT2C, mapped on chromosome 7q36.1, also known as myeloid/lymphoid or mixed-lineage leukemia protein 3 (MLL3), encodes a nuclear protein which possesses histone methylation activity and participates in transcriptional coactivation. The lysine methyltransferase 2 (KMT2) family members play essential roles in regulating cell growth and development, and mutations in these genes are frequently found in blood and solid cancers²². KMT2C has been reported to be frequently mutated in a variety of tumors. There are about 5% esophageal cancer patients carrying mutations in KMT2C (Figure 2, Figure 3), which have also been identified in other studies^23,24. However, the function of KMT2C in esophageal cancer and the relationship between KMT2C mutations and clinicopathological characteristics of esophageal cancer remain poorly understood.

2.2. Differential genetic alterations in ESCC and EAC

Though ESCC and EAC share a few common genetic alterations, they exhibit distinct molecular profile as two subtypes of esophageal cancer (Fig. 2). In terms of genetic mutations, significantly higher mutational rates in NFE2L2, KMT2D, NOTCH1 and PIK3CA are found in ESCC, whereas more frequent mutations in CDKN2A and ARID1A are observed in EAC (Fig. 2A). There are also distinctions in CNVs between ESCC and EAC. HER2, VEGFA and KRAS are significantly more amplified in EAC, while ESCC exhibits higher CNVs of CDKN2A, CDKN2B, CCND1, FGFs, PIK3CA and EGFR (Fig. 2B). These molecular differences suggest ESCC and EAC should not be considered as one disease and the therapeutic targets might be different.

2.2.1. Genetic alterations with higher frequency in ESCC

2.2.1.1. FGFs

The family of fibroblast growth factors (FGFs) regulates a wide range of biological functions, including cellular proliferation, survival, migration and differentiation²⁵. There is mounting evidence for the importance of FGF signaling in the pathogenesis of various tumor types, and clinical inhibitors targeting FGFs or their receptors are being rapidly developed²⁶. FGF3/4/19, located at 11q13-14, are found amplified in more than 50% ESCC patients (Figure 2, Figure 3A), which often coincided with CCND1 amplification. Amplification of FGF3/4 was detected by DNA microarrays in 9 of 20 surgically resected primary ESCC tumors²⁷. In a cohort of 3342 gastroesophageal cancers, amplification of FGF3, FGF4, FGF19 and FGFR1 were significantly more frequent in ESCC compared with EAC²⁸, which indicated the preferential activation of FGF pathway in ESCC. Fibroblast growth factor receptor 1 (FGFR1), FGFR4 and fibroblast growth factor receptor like 1 (FGFRL1) was implicated in the progression of ESCC²⁹, and genetic depletion of FGFRL1 was able to attenuate the motility and invasion in ESCC cells³⁰. Moreover, several studies have demonstrated that FGFR1 amplification was an independent factor of poor prognosis in ESCC patients^31,32. Given that the defined roles of abnormal FGF/FGFR signaling in tumorigenesis, agents targeting FGF/FGFR may be a potential opportunity for ESCC treatment.

2.2.1.2. EGFR

EGFR, located on chromosome 7p11.2, encodes a transmembrane glycoprotein that belongs to the ERBB family of receptor tyrosine kinases (RTK). Binding of epidermal growth factor receptor (EGFR) to epidermal growth factors triggers its autophosphorylation and subsequent activation of signal transduction pathways leading to cell proliferation, differentiation and survival³³. Abnormal activation of EGFR is often found in epithelial tumors and related to poor prognosis and decreased survival³⁴. Amplification of EGFR occurs in about 16% ESCC patients (Figure 2, Figure 3A), which was significantly associated with overexpression of EGFR³⁵. The survival rate of patients with high EGFR expression was significantly lower than that of patients with lower expression in a study consisting of 441 ESCC patients, suggesting that expression of EGFR could be a prognostic predictor for ESCC³⁶. Furthermore, a molecular prognostic model have showcased that comprised expression of EGFR, p-Sp1 and Fascin proteins was significantly associated with poor clinical outcome of ESCC patients³⁷.

2.2.1.3. NOTCH1

The NOTCH1 gene, located on chromosome 9q34.3, encodes a member of the NOTCH family proteins. NOTCH signaling is an evolutionarily conserved intercellular signaling pathway that regulates interactions between physically adjacent cells through binding of NOTCH family receptors to their cognate ligands. There is emerging evidence that mutations in NOTCH genes and dysregulated NOTCH signaling are important factors in tumor initiation and development38, 39, 40. NOTCH1 mutation is found in 14% ESCC according to the data from cBioPortal (Figure 2, Figure 3A). The mutations of NOTCH1 in ESCC tended to cluster in the EGF-like repeats and resulted in loss of function⁴¹. Loss of NOTCH1 was found to predispose the esophagus to precancer and squamous cell carcinoma partially via accelerated telomere erosion⁴². Constitutively active NOTCH signaling pathway, regardless of transient or stable expressions of NOTCH1, was able to inhibit the proliferation of ESCC EC9706 cells via arresting cells at G1 phase⁴³. Analysis of NGS data from 104 ESCCs from China found that patients with loss of function NOTCH1 mutations failed to respond to chemotherapy and had shorter survival times than patients without mutations in this gene⁴⁴. Therefore, NOTCH seems to act as a tumor suppressor in the development of ESCC.

2.2.1.4. PIK3CA

The PIK3CA gene, mapped on chromosome 3q26.32, encodes the catalytic subunit of phosphoinositide 3-kinase (PI3K), p110α. Class Ⅰ PI3K is a lipid kinase composed of an 85 kDa regulatory subunit and a 110 kDa catalytic subunit, which is responsible for phosphorylating phosphatidylinositol-4,5-bisphosphate (PIP₂) to generate the second messenger phosphatidylinositol-3,4,5-trisphosphate (PIP₃). The PI3K signaling pathway regulates various cellular processes, such as proliferation, cell cycle, apoptosis and cytoskeletal rearrangement⁴⁵. PI3K pathway is frequently disturbed in many human cancers, mostly PIK3CA alterations⁴⁶. The mutation and amplification of PIK3CA occur in 13% and 30% ESCC patients respectively (Figure 2, Figure 3A), which have been verified in a number of studies^23,24,47. A comprehensive analysis of PI3Kα expression in ESCC patients indicated that overexpression of PI3Kα was associated with lymph node metastasis and PI3Kα may play a crucial role in the development of ESCC⁴⁸. PIK3CA amplification was also associated with shorter survival in an investigation of 534 curatively resected ESCC patients, suggesting its role as a poor prognostic factor in resected ESCC⁴⁹. The alteration of PIK3CA is common in ESCC, making it a promising target for ESCC treatment and potential biomarkers for individualized molecularly targeted therapy for ESCC⁵⁰.

2.2.1.5. CDKN2B

The gene of CDKN2B encodes a cyclin-dependent kinase inhibitor p15, which interacts with CDK4 or CDK6, and prevents the activation of CDKs. Thus, the function of p15 is to regulate cell cycle progression by restraining G1/S transition. CDKN2B is located on chromosome 9p21.3, a region often implicated in the pathogenesis of multiple cancers. Genetic variants at 9p21.3 may modulate the expression of p15 and contribute to ESCC susceptibility^51,52, highlighting the importance of 9p21.3 genetic variants in tumorigenesis. Among the genetic variants found in this region, deletion of CDKN2B is the most common alteration in ESCC, which occurs in 56% ESCC patients (Figure 2, Figure 3A). In a genetic analysis of 33 human esophageal carcinomas and their adjacent normal tissues, deletion of CDKN2B was observed in a large proportion of ESCC samples⁵³. Additionally, the homozygous deletion of CDKN2B was notably associated with lymph node metastasis in ESCC⁵⁴. Lately, a study revealed CDKN2A/2B loss could be a biomarker predictive of CDK4/6 inhibitor efficacy in both patient-derived cell and patient-derived xenograft models⁵⁵, which is consistent with disorder of cell cycle in ESCC due to CDKN2A/2B deletion.

2.2.1.6. CCND1

CCND1, located on chromosome 11q13.3, encodes a highly conserved cyclin family member cyclin D1, which forms a complex with CDK4/6 and participates in cell cycle G1/S transition. Mutation, amplification and overexpression of this gene are able to release a cell from its controlled cell cycle and cause transformation to a malignant phenotype⁵⁶. The amplification of CCND1 is detected in 54% ESCC patients (Figure 2, Figure 3A). In a whole exome sequencing analysis of tissues collected from 144 Japanese ESCC patients, cell cycle regulators constituted the most frequently disrupted category, including TP53 mutations (93.1%), CCND1 amplification (46.5%) and CDKN2A deletion (53.5%)⁵⁷. Another study with 55 ESCC patients also found that ratios of CCND1 amplification or cyclin D1 overexpression were 42% or 58%, respectively⁵⁸. High expression of cyclin D1 was able to increase distant metastasis, decrease overall survival and distant metastasis-free survival in resectable ESCC, which may help to identify patients with high risk of postoperative metastases⁵⁹. Collectively, the alterations of CCND1 exist in almost half of ESCC patients, manifesting its critical character in the development and progression of ESCC.

2.2.1.7. KMT2D

KMT2D, localized to chromosome 12q13.12, encodes a histone methyltransferase that is responsible for methylating histone 3 at lysine 4. KMT2D plays critical roles in regulation of development, differentiation, metabolism and tumor suppression⁶⁰. Frequent mutations in genes involving in histone modification including KMT2D have been found in ESCC. KMT2D mutations occur in 18% of ESCC (Figure 2, Figure 3A), the majority of which are truncated mutations resulting in proteins lacking the key methyltransferase domain, indicating a tumor suppressor role of KMT2D in ESCC^{23,24,28,47,57}. Notably, KMT2D was mutated in 60% of the metastatic ESCC and in only 15.3% of the primary ESCC, proposing an important role of KMT2D in ESCC metastasis⁶¹. However, the clinical significance and prognostic value of KMT2D mutations in ESCC have not been fully elucidated, which deserves further investigation.

2.2.1.8. NFE2L2

The NFE2L2 gene, mapped on chromosome 2q31.3, encodes a transcription factor which is a member of small family of basic leucine zipper proteins. NFE2L2 is the master regulator of cellular antioxidant responses. Mutations of NFE2L2 frequently occur in human cancers, which drive continuous NFE2L2 activation and correlate with poor prognosis⁶². NFE2L2 mutations is found in 18% ESCC according to the data from cBioPortal (Fig. 2A). An analysis of 1145 tumor samples detected that NFE2L2 mutations existed in 11.4% samples, accompanied with increased NFE2L2 expression in the nuclei⁶³. Recently, a WGS of 508 Chinese ESCC patients identified that NFE2L2 mutations were significantly associated with worse prognosis of ESCC⁶⁴. Mutant NFE2L2 conferred attachment-independent cell survival, which was correlated with lymph node metastasis and tumor progression⁶⁵. Down-regulation of mutant NFE2L2 by short hairpin RNA increased the sensitivity of ESCC cells to chemotherapy⁶⁵. These results indicate the oncogenic function of NFE2L2 mutation in ESCC and strategies are proposed to target the NFE2L2 for future therapy⁶⁶.

2.2.2. Genetic alterations with higher frequency in EAC

2.2.2.1. HER2

HER2, located on chromosome 17q12, encodes a member of EGFR family of RTKs. Although HER2 possesses an active tyrosine kinase domain, no direct ligand has been identified. HER2 exists in the extended active conformation, rendering HER2 constitutively available for dimerization^67,68. HER2 is often amplified in human cancers such as breast cancer, gastric and esophageal cancer69, 70, 71. According to the data from cBioPortal, HER2 amplification occurs in 28% of EAC (Figure 2, Figure 3B). The HER2 expression in EAC patients is highly concordant with its copy number⁷². The overexpression of HER2 was also a typical feature in an animal model of BE-related EAC⁷³, indicating a crucial role of HER2 in EAC pathogenesis. HER2 expression is likely to indicate a positive prognosis in EAC. Several studies have reported that EAC patients with high HER2 expression showed a superior overall survival compared to HER2-negative patients^72,74,75. In consideration of the vital role in EAC, HER2 has become a therapeutic target for EAC patients.

2.2.2.2. VEGFA

VEGFA, located on chromosome 6p21.1, encodes a heparin-binding protein, which exists as a disulfide-linked homodimer. This growth factor stimulated proliferation and migration of vascular endothelial cells, which is essential for angiogenesis under both physiological and pathological circumstances. As angiogenesis is important to provide nutrients for tumor growth, vascular endothelial growth factor A (VEGFA) is highly expressed in cancer tissues and associated with more aggressive features. Data from cBioPortal shows that VEGFA amplification occurs in 14% of EAC (Figure 2, Figure 3B). In consistency with the high frequency of amplification, significant increase in VEGFA expression was found in tumor and peritumoral adipose tissue during the progression of normal squamous epithelium to EAC^76,77. Elevated expression of VEGFA was also observed in lymph node metastases of EAC, indicating an important role in tumor metastasis⁷⁸. Furthermore, VEGFA expression was significantly associated with poor prognosis, which supports the targeted-VEGFR therapy for EAC patients⁷⁹.

2.2.2.3. KRAS

The KRAS gene located on chromosome 12p12.1 encodes the small GTPase KRAS. KRAS acts as a molecular switch, regulating diverse signal transduction pathways⁸⁰. As one of the most well-known proto-oncogenes, KRAS has been intensively studied in the past years. The activating mutations of KRAS occur in approximately 30% of all human cancers, indicating its pivotal role in driving cancer progression⁸¹. Nevertheless, the amplification of KRAS is more common in EAC. In a WGS analysis of 22 fresh-frozen EAC samples, the amplification of KRAS by breakage-fusion-bridge was found in 6 samples, proposing a potential role in fueling EAC progression through KRAS amplification⁸². Additionally, KRAS amplification significantly related with nodal positive patients and poorer survival in the subgroup without neoadjuvant treatment. Co-existing of amplification in KRAS, HER2 as well as TP53 mutations has been reported to indicate an unfavorable prognosis in EAC⁸³.

2.2.2.4. ARID1A

ARID1A, mapped on chromosome 1p36.11, encodes a member of SWI/SNF family, which is thought to regulate transcription of specific genes by modulating the chromatin structure. Inactivating mutations in ARID1A have been identified in a broad spectrum of cancers, indicating its action as a tumor suppressor⁸⁴. There are about 12% of EAC harboring ARID1A mutations (Figure 2, Figure 3B). In a WGS study of endoscopic biopsies, ARID1A mutations were detected in 15% of high-grade dysplasia or EAC and loss of ARID1A was also identified in BE/high-grade dysplasia/EAC tissues. ARID1A-deficient EAC was reported to be accompanied with high microsatellite instability⁸⁵, suggesting its role in genomic stability. Knock down of ARID1A in EAC OE33 cells was able to enhance cell growth, proliferation and invasion, suggesting its suppressive activity in EAC pathogenesis⁸⁶. The overall survival time of patients with loss of ARID1A is apparently shorter than those with wild-type ARID1A (26.2 months vs. 60.1 months)⁸⁷. Collectively, ARID1A acts as a tumor suppressor in the initiation and progression of EAC.

3. Emerging therapeutic targets in esophageal cancer

With the development of molecularly targeted drugs and revelation of genetic alterations in esophageal cancer, molecularly targeted therapy has emerged as a promising opportunity for the treatment of esophageal cancer. Here, we summarized the recent progress in the treatment of ESCC and EAC by exploiting their distinct genetic alterations.

3.1. Common therapeutic targets in esophageal cancer

3.1.1. c-MET

c-MET is a transmembrane receptor tyrosine kinase, which is activated by hepatocyte growth factor (HGF) via autophosphorylation at the activation loop of the kinase domain. The signal mediated by c-MET is involved in cell proliferation, motility, invasion and survival. The amplification and overexpression of c-MET is implicated in a wide variety of human cancers including esophageal cancer^88,89. MET amplification was an independent prognosis factor in surgical ESCC and negatively affected the survival of ESCC patients⁹⁰. The esophagogastric adenocarcinoma patients with MET amplification showed a substantially shorter median survival time and more sensitive to the c-MET inhibitor crizotinib⁹¹. These findings suggest c-MET may be a potential therapeutic target for esophageal cancer and crizotinib is being tested in clinical trials (Fig. 4 and Table 1). Anti-c-MET antibody telisotuzumab showed acceptable safety profile and clinical activity in esophageal cancer patients carrying MET amplification. Among four gastroesophageal cancer patients, three achieved a partial response and one had progressive disease as best response⁹². The completed phase Ⅰ and Ⅱ clinical studies with c-MET inhibitor AMG337 found that EAC patients with MET amplification responded well to AMG337 with manageable toxicities, and an objective response rate of 18% was observed in gastric/gastroesophageal junction/EAC patients with MET amplification^93,94. Collectively, c-MET is a promising target for esophageal cancer patients with MET amplification.

Figure 4.

Potential therapeutic targets and drugs for the treatment of ESCC and EAC. Drugs in red have been approved by FDA, drugs in black are at different stages of clinical trials, while drugs in blue are in preclinical studies.

Table 1.

Ongoing clinical trials of molecularly targeted drugs in esophageal cancer.

NCT No.	Drug	Target	Disease	Phase
NCT02465060	Crizotinib	c-MET	Esophageal cancer	Ⅱ
NCT04491942	Elimusertib + chemotherapy	ATR	Esophageal cancer	Ⅰ
NCT04460937	Adavosertib + radiotherapy	WEE1	Esophageal cancer	Ⅰ
NCT04171700	Rucaparib	PARP	Esophageal cancer	Ⅱ
NCT01366144	Veliparib + chemotherapy	PARP	Esophageal cancer	Ⅰ
NCT03233724	Decitabine	DNMT	Esophageal cancer	Ⅰ/Ⅱ
NCT03544736	Nivolumab + radiotherapy	PD-1	Esophageal cancer	Ⅰ/Ⅱ
NCT03691090	Camrelizumab + chemotherapy	PD-1	Esophageal cancer	Ⅲ
NCT02639065	Durvalumab	PD-L1	Esophageal cancer	Ⅱ
NCT03940976	Afatinib	EGFR	ESCC	Ⅱ
NCT02409186	Nimotuzumab + chemoradiotherapy	EGFR	ESCC	Ⅲ
NCT03770988	Poziotinib	EGFR	ESCC	Ⅱ
NCT03126708	Cetuximab	EGFR	ESCC	Ⅱ
NCT02699606	Erdafitinib	FGFR	Esophageal cancer	Ⅱ
NCT01795768	AZD4547	FGFR	Esophageal cancer	Ⅱ
NCT03292250	Alpelisib	PI3Kα	ESCC	Ⅱ
NCT03544905	CYH33	PI3Kα	ESCC	Ⅰ
NCT01351103	LGK974	Wnt	ESCC	Ⅰ
NCT04430738	Tucatinib	HER2	EAC	Ⅰ/Ⅱ
NCT03783936	Trastuzumab + avelumab	HER2	EAC	Ⅱ
NCT04499924	Tucatinib	HER2	EAC	Ⅱ/Ⅲ
NCT01359397	Bevacizumab + chemotherapy	VEGFR	EAC	Ⅱ
NCT02898077	Apatinib + chemotherapy	VEGFR	Esophageal cancer	Ⅱ

3.1.2. CDK4/6

The cell cycle-related genes including TP53, CDKN2A, CDKN2B and CCND1 frequently altered in esophageal cancer (Fig. 3), indicating a highly disordered cell cycle progression. In recent years, therapies targeting cell cycle progression has made great progress. Three CDK4/6 inhibitors, abemaciclib, palbociclib and ribociclib, have emerged for the therapy of breast cancer. In light of the dysregulation of cell cycle, targeting CDK4/6 might be a promising approach for treatment of esophageal cancer. Abemaciclib was able to induce apoptosis and inhibit proliferation in EAC cells and significantly inhibited the tumor growth in vivo⁹⁵. Clinical trials to test the efficacy of CDK4/6 inhibitors in esophageal cancer are warranted.

3.1.3. Genes involving DNA damage response

Esophageal cancer is characterized by a high mutation in TP53, which plays pivotal roles in the DNA damage response⁹⁶. In response to DNA damages, activation of P53 initiates pathways involved in cell cycle arrest, DNA repair, and apoptosis in order to avoid propagation of damaged cells. Mutations in TP53 cause defective and weakened G1 and G2 checkpoints, rendering cells highly dependent on activated WEE1 to achieve cell cycle arrest in response to DNA damage. WEE1 was identified as a synthetic lethal target in TP53-mutated cancer. WEE1 inhibitor AZD1775 in combination with chemotherapy showed superior response rate in TP53 mutated patients (21%) compared with those with wild-type TP53 (12%)⁹⁷. Combination DNA damage response inhibitors such as inhibitors targeting ATR, CHK1/2 with chemotherapy/radiotherapy also displayed potential for the treatment of esophageal cancer⁹⁸. The combination of an ATR inhibitor, elimusertib, or a WEE1 inhibitor, adavosertib, with chemotherapy/radiotherapy for the treatment of advanced solid tumors including esophageal carcinoma is being tested in clinical trials (Fig. 4 and Table 1). PARP inhibitors alone or in combination with paclitaxel are being evaluated in esophageal cancer patients carrying deleterious mutation in homologous recombination repair genes (Fig. 4 and Table 1).

3.1.4. Epigenetic modulators

Epigenetic alterations such as KMT2C, KMT2D and ARID1A are common in esophageal cancer (Fig. 3). Though the role of these alterations has not been clearly elucidated, drug candidates targeting epigenetic modulators have displayed potential activity against esophageal cancer. Pretreatment of the DNA methyltransferase inhibitor azacitidine enhanced the efficacy of chemotherapy in advanced EAC patients. All 12 patients achieved R0 section after treatment of azacitidine plus chemotherapy. 3 patients (25%) had a complete response, 5 patients (42%) had a partial response, and 4 (33%) had stable disease⁹⁹. Combination of a histone deacetylase inhibitor MS-275 and azacytidine was efficient to inhibit esophageal cancer cells, indicating a potential targeted therapy for the esophageal cancer¹⁰⁰. Moreover, the DNA methyltransferase (DNMT) inhibitor decitabine is being tested in clinical trials (Fig. 4 and Table 1). As the role of epigenetic alterations in esophageal cancer is gradually revealed, epigenetic-targeted therapy might be a possible option for esophageal cancer patients.

3.1.5. Immune checkpoint

For the past few years, remarkable breakthrough of immunotherapy has been made for the treatment of solid tumors. Immune checkpoint inhibitors such as anti-CTLA4 or anti-PD-1/PD-L1 antibodies have been approved by the U.S. Food and Drug Administration (FDA) for the treatment of multiple solid tumors. PD-1 inhibitor pembrolizumab exhibits desired activity in esophageal cancer patients, which has been used to treat ESCC with high expression of PD-L1 (Fig. 4 and Table 2)¹⁰¹. Immune checkpoint inhibitors alone or combined with chemoradiotherapy are being assessed in clinical trials (Fig. 4 and Table 1) and encouraging results have been obtained in esophageal cancer. In a single-arm phase Ⅱ trial, PD-1 inhibitor camrelizumab combined with apatinib and chemotherapy demonstrated significant efficacy and manageable safety in advanced ESCC patients. The objective response rate and disease control rate were 80% and 96.7% respectively. All adverse events were handled with appropriate medical care, which suggested a new combination strategy for ESCC treatment¹⁰². The level of serum lactate dehydrogenase may serve as a potential marker for the response of camrelizumab in ESCC patients¹⁰³. Recently, the efficacy and safety of pembrolizumab in combination with chemoradiotherapy are being evaluated in both ESCC and EAC patients, which may help define the role of immunotherapy as a first-line treatment option for patients with esophageal carcinoma¹⁰⁴.

Table 2.

Approved molecularly targeted drugs for esophageal cancer.

Drug	Target	Company	Indication
Trastuzumab	HER2	Pfizer	EAC
Ramucirumab	VEGFR	Lilly	EAC
Pembrolizumab	PD-1	Merck Sharp & Dohme	ESCC

3.2. Therapeutic targets in ESCC

3.2.1. EGFR

Amplification and overexpression of EGFR occurred frequently in ESCC and significantly correlated with tumor progression and invasion^35,58,105, which indicates that EGFR is a potential molecular target in ESCC. The selective EGFR inhibitor gefitinib showed potent antitumor activity against ESCC in vitro and in vivo and augmented the efficacy of tumor necrosis factor-related death apoptosis inducing ligand in human ESCC cells^106,107. Lapatinib, a dual inhibitor of EGFR and HER2, possessed significant activity in ESCC cells over-expressing EGFR and/or HER2 and synergistically inhibited the growth of ESCC patient-derived xenograft in combination with 5-fluorouracil^108,109. Monoclonal antibodies against EGFR also exhibited favorable efficacy against ESCC. Nimotuzumab was able to promote radiosensitivity in EGFR-overexpressing ESCC cells by upregulating insulin like growth factor binding protein 3 (IGFBP-3)¹¹⁰. Cetuximab significantly inhibited the growth of ESCC cells overexpressing EGFR¹¹¹. Clinical trials of small molecule inhibitors or monoclonal antibodies targeting EGFR for the treatment of ESCC are undergoing (Fig. 4 and Table 1) and clinical efficacy has been observed. Combination of chemoradiation with erlotinib has been shown to be effective and tolerable in ESCC patients in phase Ⅱ studies^112,113. Among the 21 patients treated, 8 (38%) achieved a complete response, 10 (47.6.%) a partial response and 3 (14.4%) stable disease. No patient showed progressive disease¹¹³. In a randomized phase Ⅲ trial for locally advanced ESCC patients, the addition of erlotinib significantly improved overall survival compared with chemoradiotherapy alone (median, 39.4 vs. 27.4 months)¹¹⁴. Icotinib alone or combined with radiotherapy also showed favorable activity in ESCC patients with EGFR overexpression or amplification. The objective response rate and disease control rate was 16.7% or 46.3%, respectively, in a phase Ⅱ trial of icotinib monotherapy in ESCC patients with EGFR overexpression¹¹⁵. In another randomized phase Ⅱ trial with 127 ESCC patients enrolled, icotinib plus radiotherapy improved the median overall survival compared with radiotherapy alone (24.0 months vs. 16.3 months)¹¹⁶. Complete or partial responses were observed in 18 of 20 elderly ESCC patients in a phase Ⅱ trial of combined radiotherapy and gefitinib¹¹⁷. In summary, targeting EGFR is a promising therapeutic option for ESCC patients and EGFR amplification may serve as a biomarker indicating favorable response.

3.2.2. FGFR

FGF/FGFR participate in multiple cellular processes, whereas deregulation of FGF/FGFR signaling pathway is correlated with developmental disorders and cancers¹¹⁸. FGFR inhibitors are developed in recent years and erdafitinib is the first FGFR inhibitor approved by FDA for the therapy of urothelial carcinoma in 2019. Several other FGFR inhibitors also exhibit favorable activity against multiple solid tumors. The study of FGFR inhibitor on ESCC patients is at the early stage. The selective FGFR4 inhibitor H3B-6527 significantly inhibited the growth of ESCC cells in vitro and in vivo by blocking PI3K and mitogen-activated protein kinase (MAPK) signaling pathways¹¹⁹. FGFR inhibitor AZD4245 could enhance the sensitivity of ESCC cells with high level of active FGFR1 to gefitinib, proposing a potential combination for ESCC treatment¹²⁰. Currently, several clinical trials with FGFR inhibitors for esophageal cancer are underway (Fig. 4 and Table 1).

3.2.3. PI3Kα

The mutations and amplification of PIK3CA are frequent and associated with poor prognosis in ESCC, indicating PI3Kα is a promising target for ESCC treatment. As the first PI3Kα-specific inhibitor approved by FDA, alpelisib has been proved to be effective in several solid tumors including head and neck squamous cell carcinoma and ESCC¹²¹. Our previous study has reported that CYH33, a novel PI3Kα-selective inhibitor, displayed significant activity against the proliferation of ESCC cells and the growth of xenografts derived from ESCC cells¹²². CYH33 also sensitized ESCC cells to radiation by abrogating survival signals in tumor cells and tumor microenvironment¹²³. The combination of PI3Kα inhibitors with other targeted drugs also demonstrated potent anti-ESCC activity. Combination of alpelisib with a JNK inhibitor, SP600125, displayed a synergistic activity against ESCC in vitro and in vivo, proposing a new therapeutic strategy for ESCC patients¹²⁴. Concurrent inhibition of PI3Kα and RAS/MEK synergistically inhibited the growth of xenografts harboring HRAS^G12S mutation, which was worthwhile to be considered in the treatment of ESCC patients harboring mutation or amplification in KRAS or HRAS¹²⁵. Alpelisib and CYH33 are currently in clinical trials for the treatment of solid tumors including advanced ESCC (Fig. 4 and Table 1).

3.2.4. WNT signaling pathway

The WNT signal transduction cascade is a main regulator of development and a key driver of most types of tissue stem cells¹²⁶. The alterations in WNT signaling pathway possess a fundamental role in cancer development and drug resistance¹²⁷. WNT signaling pathway was reported to be hyperactivated in ESCC^24,57. The biomarker signature score based on a panel of WNT-related markers (cytoplasmic β-catenin, nuclear c-MYC, nuclear disheveled segment polarity protein 1 and membrane α-catenin) was associated with ESCC recurrence/death¹²⁸. WNT signaling was also found to be involved in radioresistance by promoting DNA damage repair mediated by a chromatin-associated protein trans-activating high-mobility group box 1¹²⁹. These studies indicated that WNT signaling might be a potential therapeutic target for ESCC. LGK974 is a porcupine inhibitor, which inhibits the secretion of WNT and its signaling by blocking the palmitoylation of WNT¹³⁰. LGK974 is currently in clinical trials for the treatment of solid tumors including ESCC (Fig. 4 and Table 1).

3.3. Therapeutic targets in EAC

3.3.1. HER2

HER2 amplification and overexpression have been found in a number of human cancers, such as breast cancer, gastric cancer and EAC. Currently, HER2-targeted therapy has been applied to the treatment of a few types of cancer overexpressing HER2. The anti-HER2 monoclonal antibody trastuzumab exhibited promising therapeutic effect in HER2-positive EAC patients. In a phase Ⅲ randomized trial for HER2-positive gastric or gastric-esophageal junction adenocarcinoma patients, the median overall survival was 13.8 months in trastuzumab plus chemotherapy group compared with 11.1 months in chemotherapy alone group¹³¹. Considering the favorable clinical efficacy, trastuzumab has been approved by FDA for the treatment of HER2-positive EAC patients in combination of chemotherapy (Fig. 4 and Table 2). Other HER2-targeted drugs are being tested in clinical trials (Fig. 4 and Table 1) and favorable activity against HER2-positive EAC has been observed. Lapatinib, a small-molecule inhibitor of HER-2 and EGFR, has been approved to treat HER2-positive breast cancer. Combined lapatinib and chemotherapy was effective in HER2-positive EAC patients with manageable toxicity¹³². In a phase Ⅱ study of trastuzumab and pertuzumab with chemoradiotherapy in EAC, complete response rate and overall survival was 34% and 71%, respectively. No unexpected adverse events were observed and patients with HER2 overexpression showed better response, indicating that such combination was promising in HER2-positive patients¹³³. Based on the positive results of clinical trials, HER2-targetd drugs displayed great promising in the therapy of HER2-positive EAC.

3.3.2. VEGFR

VEGF plays an essential role in the regulation of angiogenesis and lymphangiogenesis, which is vital for tumor angiogenesis and metastasis. Targeting VEGFR has been proven effective in diverse solid tumors including EAC. Ramucirumab is a monoclonal antibody targeting VEGFR and has displayed efficacy in EAC patients. In a phase Ⅲ trial of ramucirumab monotherapy for gastric or gastric–esophageal junction adenocarcinoma patients, the median overall survival was 5.2 months in ramucirumab group and 3.8 months in placebo group¹³⁴. Besides, ramucirumab in combination with paclitaxel significantly improved overall survival compared with placebo plus paclitaxel in gastric or gastric-esophageal junction adenocarcinoma patients (9.6 months vs. 7.4 months)¹³⁵. Ramucirumab has been approved by FDA for treatment for EAC patients in view of the promising clinical results (Fig. 4 and Table 2). Bevacizumab is also a monoclonal antibody against VEGFR being tested in clinical trials (Fig. 4 and Table 1). In a phase Ⅱ study of bevacizumab combined with chemotherapy in metastatic gastric or gastro–esophageal junction adenocarcinoma patients, the overall response rate was 65% and time to disease progression improved over historical controls by 75%¹³⁶. Small molecule VEGFR inhibitors apatinib also showed favorable antitumor activity in EAC patients and has been approved in China as the third-line treatment for gastro–esophageal junction adenocarcinoma¹³⁷. The clinical trial of apatinib combined with chemotherapy in esophageal cancer patients is now undergoing (Fig. 4 and Table 1).

4. Conclusions and perspectives

Esophageal cancer is one of the most lethal malignancies worldwide, whereas the options of molecularly targeted therapy are very limited. At present, there are only three targeted drugs have been approved for clinical applications (Table 2). With the evolution of DNA sequencing technology, the landscape of esophageal cancer genome is being revealed in recent years. We analyzed the genomic data and found that these recurrently altered genes are enriched in regulators of cell cycle (TP53, CDKN2A, CDKN2B, CCND1 and MYC), members in the RTK-MAPK/PI3K pathway (FGFs, EGFR, HER2, VEGFA, PIK3CA, and KRAS), and epigenetic modulators (KMT2C, KMT2D and ARID1A) (Fig. 3). These alterations could disrupt normal epithelia homeostasis and contribute to tumorigenesis and promotion of esophageal cancer.

Though current sequencing technology have provided bulk genomic data to guide the treatment of esophageal cancer, how to select the genomic data due to the intratumoral heterogeneity is still a considerable challenge in clinical practice. Emerging strategies such as multiregional sequencing, single-cell sequencing and analysis of liquid biopsy samples have been introduced. These strategies may help better understand the tumor heterogeneity and provide more accurate data for precise medicine. However, bulk genomic data are usually utilized in the clinical practice due to the cost and efficiency. With further development of the sequencing technology, more precise genomic data will be available for the molecularly targeted therapy in esophageal cancer.

ESCC and EAC are two subtypes of esophageal cancer with diverse epidemiology and pathogenesis as well as different molecular profiles, which provide distinctive targets for the treatment. Further preclinical studies and clinical trials are in progress and some encouraging clinical efficacy has been observed (Fig. 4). Moreover, the combination of targeted drugs and chemotherapy/radiotherapy displayed great potential for this deadly disease. Combination regimens with improved efficacy and reduced toxicity are being tested in clinical trials (Table 1). With the increasing knowledge of molecular characteristics in ESCC and EAC, more molecularly targeted therapy will be developed to benefit the esophageal cancer patients in the near future.

Author contributions

Linghua Meng conceived the project and revised the manuscript. Xu Zhang and Yuxiang Wang summarized the literature and composed the manuscript. All authors proofread the manuscript and approved to submit the final manuscript.

Conflict of interest

The authors have no conflicts of interest to declare.

Appendix A. Supplementary data

The following are the Supplementary data to this article: