Formosan Journal of Surgery (2015) 48, 185e197

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.e-f js .com

REVIEW ARTICLE

Therapeutic strategies for esophagogastricjunction cancer

Pei-Ming Huang, Chiung-Nien Chen*

Department of Surgery, National Taiwan University Hospital and National Taiwan University College ofMedicine, Taipei, Taiwan

Received 5 May 2015; received in revised form 14 July 2015; accepted 15 August 2015Available online 25 December 2015

KEYWORDScarcinoma of thedistal esophagus;

carcinoma of theesophagogastricjunction;

cardia carcinoma;multimodal therapy;resection

Conflicts of interest: All contributi* Corresponding author. Division of G

South Road, Taipei 10023, Taiwan.E-mail address: cnchen@ntu.edu.t

http://dx.doi.org/10.1016/j.fjs.2015.1682-606X/Copyright ª 2015, Taiwan

Summary The incidence and prevalence of esophagogastric junction (EGJ) cancer haveincreased in Western and Asian countries in recent decades. However, previous studies onthe classification, surgical treatment, and multimodal therapy in relation to these tumors haveindicated multiple inconsistencies. Despite their single anatomical location and stage, studieson EGJ cancer-related therapy are scarce, because patient cases with EGJ cancer are typicallyexamined in esophageal and gastric cancer treatment trials. Despite improvements in surgicaland radiotherapy techniques and refinements to chemotherapeutic regimens, the long-termsurvival of patients with EGJ cancer remains poor. Surgical resection is the only standardapproach for the early stage of the disease (Stage I). The primary goal of EGJ cancer operationis the complete removal of the primary tumor and its lymphatic drainage. Treatment ap-proaches for localized, resectable EGJ cancer are based on the location of primary tumors,their histology, and patient comorbidities; however, because surgery is the fundamentalapproach for the treatment of Stage IIeIII resectable EGJ cancer, multimodality therapy withchemotherapy or chemoradiotherapy remains uncertain. This review focuses on the surgicalapproaches, and provides an overview of the evidence for perioperative treatment and therole of targeted therapies and ongoing clinical trials regarding EGJ cancer.Copyright ª 2015, Taiwan Surgical Association. Published by Elsevier Taiwan LLC. All rightsreserved.

1. Introduction

In 2013, 29,590 patientswere diagnosedwith esophageal andgastric cancer, of whom>90% are predicted to die from their

ng authors declare no conflicts ofeneral Surgery, Department of Su

w (C.-N. Chen).

08.003Surgical Association. Published by

condition.1 Gastric cancer is the third most lethal canceroverall, and esophageal cancer has the most rapidlyincreasing incidence inWestern countries.2 With a combinedmortality of >1.1 million people, esophageal and gastric

interest.rgery, National Taiwan University Hospital, Number 7, Chung-Shan

Elsevier Taiwan LLC. All rights reserved.

186 P.-M. Huang, C.-N. Chen

cancers have become the second most common causes ofcancer death, the first being lung cancer. Moreover, theincidence and prevalence of esophagogastric junction (EGJ)cancer with the clinicopathologic characteristics of bothesophageal cancer and gastric cancer have increased inWestern and Asian countries in recent decades.3,4 EGJ can-cer features squamous-cell carcinoma (SCC) of the distalesophagus, adenocarcinoma of the distal esophagus, andcardiac adenocarcinoma. These carcinomas differ ingeographic incidence, etiology, and biology. Previous studieson the classification, surgical treatment, and multimodaltherapy for these tumors reported multiple inconsistencies,primarily because of the borderline location of the tumorsbetween the esophagus and the stomach. Siewert classifiedEGJ cancer based on the tumor location as follows: adeno-carcinomas located in the distal esophagus as Type I, ade-nocarcinomas centeredwithin 1 cm of the EGJ as Type II, andcarcinomas located 1e5 cmdistal to the EGJ as Type III.5 Thisclinical classification was based on the results of esoph-agography, endoscopy with a retroflexed view of the EGJ,computed tomography (CT), and intraoperative observa-tions. This precise definition and classification is intended tohelp with describing and comparing these tumor types. Thetumor-nodes-metastasis (TNM) staging system in the Unionfor International Cancer Control/American Joint Committeeon Cancer (UICC/AJCC) Cancer Staging Manual (7th edition)defines EGJ cancer as esophageal cancer, including the SCCof the esophagus and adenocarcinoma of the esophagus,having its center between 5 cm proximal and 5 cm distal tothe EGJ.6,7 The UICC/AJCC categorized all types of cardiaccancer with no EGJ invasion as gastric cancer, irrespective oftumor location.

The molecular biology and tumor location of the SCC andadenocarcinoma of the esophagus vary, and therefore, bothdiseases must be considered individually and treated opti-mally with their respective management strategies, such asseparate surgical approaches and a distinct pattern oflymphadenectomy.8,9 Consequently, selecting a particularmanagement strategy for tumors located at the EGJ be-comes a controversial issue. The available randomized datasuggest that various surgical approaches are appropriate,and neoadjuvant therapy improves the survival of patientswith the locally advanced disease. However, the currentlyrandomized therapeutic data on EGJ cancer are based onthe results of multiple trials with a focus on esophageal andgastric cancers with various tumor locations combined. Thestaging system, surgical approaches, and systemic thera-pies differ greatly for esophageal SCC or adenocarcinoma,EGJ cancer, and gastric cancer. However, in any study, it isessential to consider the patient population along with eachhistological subtype, anatomical location, and geographywhen adopting an evidence-based approach, because thismay significantly influence the findings. This review focuseson updating disease-related knowledge and identifyingoptimal systemic therapy targeted at a particular cancertype located around the EGJ.

2. Classification of EGJ cancer

This review investigated every tumor centered within 5 cmproximal and 5 cm distal to the EGJ, particularly after

considering its histological type and distinguishing it basedon its relevant location to the EGJ. All such tumors werecategorized into five groups, namely, E (Sq), E (Ade), EGJ(Sq), EGJ (Ade), and C (Ade), for optimizing therapeuticstrategies and improving the outcome (Figure 1). In thisreview, E (Sq or Ade) tumors were defined as SCC oradenocarcinoma of the lower esophagus without EGJinvolvement and Siewert Type I tumors. Several researchershave indicated that Siewert Type II and Type III tumorsshould be considered a part of gastric cancer, irrespectiveof EGJ involvement, according to their clinicopathologicfeatures and prognosis.10,11 Therefore, EGJ (Sq or Ade) le-sions were considered Siewert Type II and Type III tumorsand subsites of cancer that extend into the EGJ, regardlessof tumor location (central within EGJ or cardiac). Lesionsthat arise within 5 cm of the EGJ but do not extend to theesophagus were classified as C (Ade) tumors. This definitionis objective and applicable for analyzing currently availableoptimal therapies, but varies from the current Siewertclassification. Classifying tumors based on this classificationdepends solely on the morphological appearance of atumor, and is according to the combined results of thecontrast radiogram, endoscopy with a view of the EGJ, CT,and pathological observations. Several reports have justi-fied differentiating epidemiological and morphologicalfeatures between each subtype of EGJ cancer. E (Sq) or E(Ade) tumors were observed mostly in men.12 Patients withE (Ade) tumors are most likely to exhibit a hiatal hernia anda long history of gastroesophageal reflux disease. Anotherstudy identified Barrett’s esophagus (intestinal epithelialmetaplasia in the distal esophagus) with progressive severedysplasia as the main precursor Type E (Ade) tumors.13 TypeC and EGJ tumors are associated with short segments ofintestinal metaplasia at or below the EGJ, although theyare uncommon.14 A Helicobacter pylori infection reducesgastric acidity and the subsequent toxicity exposure levelof the esophagus toward the gastric contents.15 The distinctepidemiology for esophageal, EGJ, and gastric adenocar-cinomas may imply the roles of biological differencesassociated with the location. Furthermore, SCC is associ-ated with smoking and alcohol abuse.16

The major lymphatic pathways of the lower esophagusadvance upward into the mediastinum and downward alongthe celiac axis, whereas those advancing particularly fromthe gastric cardia pass toward the celiac axis, splenic hilum,and paraaortic lymph nodes.17 This explains the distinctpattern of the lymphatic spread in EGJ cancer.18 Cervicallymph node metastases are observed only in E (Sq) tumors.However, E (Ade) tumors exhibit a nodalmetastatic spreadingpattern similar to that of EGJ (Ade) tumors,19,20 with a 26%mediastinal and 47% abdominal node positivity. Moreover,patients with E (Ade) tumors exhibit a higher incidence ofnodal metastasis at the mediastinal lymph node.18

3. Molecular targets in EGJ cancer

This brief review also presents a discussion on the geneticpathways of cancer, which can facilitate the development ofpotential future therapies. The genetic pathogenesis ofesophagogastric cancer involves multiple pathways. Theprevalence and pattern of genetic abnormalities detected

Figure 1 Tumor location and classification of esophagogastric junction (EGJ) cancer. We categorized EGJ cancer into five groupsas E (Sq), E (Ade), EGJ (Sq), EGJ (Ade), and C (Ade) tumors based on histological type and tumor location. E (Sq or Ade) tumors aredefined as squamous cell carcinoma or adenocarcinoma of the lower esophagus without EGJ involvement and Siewert Type I tu-mors. EGJ (Sq or Ade) lesions are considered Siewert Type II and Type III as well as cancer subsites that extend into the EGJ. Lesionsthat arise within 5 cm of the EGJ but do not extend to the esophagus are classified as C (Ade) tumors.

Therapeutic strategies for EGJ cancer 187

through comparative genomic hybridization vary accordingto the different location of EGJ adenocarcinoma.21 Theabnormal expression of proto-oncogenes, such as Src, Ras,myc, Sis, and Myk, leads to abnormal cellular proliferationand differentiation. The phosphoinositide 3-kinase (PI3K)/PTEN pathway plays a major role in carcinogenesis. The lossof PTEN expression is a potential biomarker of resistance inantiepidermal growth factor receptor (anti-EGFR) therapyfor advanced-stage gastric cancer (Figure 2). It has been re-ported in >50% of cases in a retrospective study and is asso-ciated with a poor outcome.22 The deletion or inactivation oftumor suppressor genes such as Rb and p53 (on chromosome17p) inhibits cell growth, resulting in programmedcell death.Mutations in p53 typically occur in esophageal cancer, andare indicative of progression toward SCC and adenocarci-noma.23 The expression of the CXCL12 chemokine and itssingle-nucleotide polymorphism rs1801157 GA/AA is signifi-cantly associated with distant metastasis (p Z 0.026).24

Because CXCL12 polymorphisms mediate tumor cell dissem-ination in EGJ cancer, theymay represent amarker indicatingan advanced stage of the disease. However, CXCL12 expres-sion is nonsignificantly associated with the depth, lymphnode metastasis, and grading of the tumor. Researchinvolving the identification of biomarkers in the blood ortumor tissue is crucial for enhancing the treatment efficacy inpatients with advanced-stage EGJ cancer.

EGFR is a transmembrane glycoprotein. Its over-expression can lead to increased tumor invasiveness,angiogenesis, and apoptosis inhibition. Human epidermalgrowth factor receptor-2 (HER-2) is a member of the EGFRfamily. HER-2/NEU amplification and expression areapparent in gastric and EGJ cancers. Approximately 71% ofresected esophageal SCC and 88% of lymph node metastasesexhibit a higher frequency of EGFR overexpression.25 EGFR

overexpression is noted in 33e50% of resectable esophagealadenocarcinomas, and is negatively associated withdisease-free survival and overall survival (OS).26 The over-expression of EGFR and possibly that of HER-2 implies apoor prognosis as potentially vital therapeutic targets forselected patients. The inclusion of trastuzumab in cytotoxicchemotherapy was found to yield a survival advantage inpatients with HER-2/NEU-positive gastric cancer. However,the prognosis of advanced EGJ (Ade) tumor remains poorafter therapy including these new cytotoxic drugs and tar-geted agents.27 Cetuximab and panitumumab, two mono-clonal antibodies against EGFR, and the dual EGFR and HER-2 tyrosine kinase inhibitor (TKI), lapatinib, are currentlyunder Phase III clinical trials in EGJ cancer28 (Figure 2).

Several predictive markers are available for identifyingpatients who may benefit more from systemic chemo-therapy in advanced EGJ cancers. A higher tumor vesseldensity is associated with a greater risk of tumor metas-tasis. The disruption of the vascular endothelial growthfactor (VEGF) pathway inhibits tumor growth by reducingthe blood supply to the tumor. The new invasion and stem-cell markers VEGF receptor-3 (VEGFR-3) and CXCR4 wereexamined in patients with aggressive EGJ cancer, and inchemotherapies including oxaliplatin/leucovorin/5-fluorouracil (5-FU) (FLO) versus cisplatin/leucovorin/5-FU(FLP). The CXCR4-positive patients exhibited an improvedOS after FLP, whereas FLO was more effective in CXCR4-and VEGFR-3-negative patients compared with FLP.29

However, a retrospective comparative study on two meth-ylenetetrahydrofolate reductase (MTHFR) polymorphisms inesophageal cancer, esophagogastric cancer, and gastriccancer (i.e., MTHFR A1298C polymorphisms) were found toplay an independent prognostic factor only in patients withneoadjuvantly treated gastric adenocarcinomas, but not in

Figure 2 General mechanisms of the molecular targets of epidermal growth factor receptor (EGFR) and human epidermal growthfactor receptor (HER) dimer pathways in esophagogastric junction (EGJ) cancer. The dimer binding stabilizes and downregulatesthe PI3K/AKT and RAS/RAF/MAPK signaling cascade, inhibiting apoptosis. Genomic aberrations in the PI3K pathway are a commonevent in various cancer types. Monoclonal antibodies recognize and attach to the extracellular domain of EGFR (cetuximab andpantitumumab) or HER-2 (trastuzumab and pertuzumab). These antibodies are currently undergoing clinical trials in advancedesophagogastric cancer. Inhibitors targeting the intracellular tyrosine kinase domain of EGFR include erlotinib, lapatinib, andgefitinib. Lapatinib (a tyrosine kinase inhibitor of the HER-2 and EGFR pathways) and everolimus (inhibitor of intracellulardownstream signaling protein mTOR) are currently undergoing Phase III evaluation in advanced-stage gastric cancer. AKT Z proteinkinase B; MAPK Z mitogen-activated protein kinase; MEK Z MAP kinase kinase; mTOR Z mammalian target of rapamycin;PI3K Z phosphatidylinositol 3 kinase; RAF Z GTPase Raf; RAS Z GTPase Ras.

188 P.-M. Huang, C.-N. Chen

other cancer types nor in primarily resected patients,indicating that this enzyme affects cancer-associatedoutcomes.30

4. Diagnosis of EGJ cancer

Extensive preoperative staging is essential for the correctselection of a suitable therapeutic strategy against EGJ can-cer. Endoscopic ultrasonography (EUS) is the most accuratemethod for assessing the locoregional extent of esophagusand EGJ carcinomas. The EUS accuracy rates for evaluatingthe T, N, and M stages are 73%, 80%, and 78%, respectively.31

EUS has thus become mandatory in the preoperative assess-ment of tumors located at the EGJ. However, EUS accuracyfor staging EGJ cancer is lower in tumors spanning >5 cm.Predicting the pTcategory and local resectability using EUS ispossible in 85% of patients. Furthermore, magnifying narrow-band imaging (NBI) of endoscopy can enable contrasting thesurface structure and vascular architecture of the mucosa,thus facilitating the evaluation of mucosal features thatcorrespond to their histology. The morphology of the intra-papillary capillary loop in EGJ cancer, as observed throughNBImagnification, varies according to the tumor invasion depth.Furthermore, NBI magnification can provide valuable infor-mation for distinguishing EGJ neoplasia fromnon-neoplasia aswell as identifying the extent of neoplasia.32 The assessmentof peritoneal and abdominal metastases conducted usingnoninvasive imaging yields inaccurate results. Diagnosticlaparoscopy is the most accurate tool for identifying theperitoneal tumor spread or livermetastases in 25%of patientswith locally advanced EGJ tumors.33

Patients with locally advanced EGJ tumors who respondto induction chemotherapy or chemoradiotherapy (CRT)exhibit significantly improved survival compared with pa-tients with no response. A systematic review evaluatedlong-term survival outcomes in patients with advancedesophageal and EGJ cancers using the predictive value of18F-fluorodeoxyglucose positron emission tomography(18FDG-PET). 18FDG-PET efficiently predicts long-term out-comes compared with histopathological tumor responses.34

Early PET permits the identification of nonresponsive pa-tients for earlier operation and the discontinuation ofineffective preoperative therapy, without a survival detri-ment. The use of early PET during preoperative chemo-therapy is undergoing evaluation in a CALGB/Alliance trial(ClinicalTrials.gov identifier: NCT01333033).

5. Treatment strategies for EGJ cancer

The management of EGJ cancer is controversial because ofseveral factors. The first is that the definition of EGJ cancerhas not been standardized. Thus, clinicopathologic charac-teristics and treatment policies have not been integrated.Moreover, most reports on EGJ cancer have included mixedresults regarding EGJ cancer and gastric carcinoma.

5.1. Surgical approach

Surgery is an effective and mainstay therapy for resectableEGJ cancer. Surgery alone may be appropriate for severalpatients with early disease (intramucosal and Stage IA), with

Therapeutic strategies for EGJ cancer 189

a 5-year OS for 80e100% of those affected,35 a poor per-formance status, or increased risk of CRT. Surgical consid-erations specific to EGJ cancer include variousesophagectomies and gastric resections, the choice of sur-gical approaches, and the optimal extent of lymph nodedissection. The complete resection of a tumor with negativemargins (R0 resection) and its entire lymphatic drainage isthe objective of surgical resection for improving the survivaloutcome compared with limited surgery.5 We assessedvarious surgical approaches for EGJ cancer, including en blocesophagogastrectomy, subtotal esophagectomy with theresection of the proximal stomach, total gastrectomy withthe transhiatal resection of the distal esophagus, and alimited resection of the EGJ. The survival outcome of thecomplete resection is similar to that of the traditionalapproach of esophagectomy, and selecting an appropriatemethod is based on surgical expertise.36 Tumors near theEGJ are treated primarily with transhiatal or thor-acoabdominal esophagectomy. The left thoracoabdominalapproach for surgery does not improve the survival outcome,and leads to increased morbidity37 compared with anabdominaletranshiatal approach for EGJ cancer.38

Moreover, EGJ (Sq or Ade) cancer carries the risk of boththoracic and abdominal metastases. Hence, a limitedresection without mediastinal lymphadenectomy is selectedfor limited benefit, but this method should not be chosen forEGJ (Sq or Ade) tumors.18,39,40 Esophagectomy with medias-tinal and abdominal lymphadenectomy is adequate for E (Sq)tumors, and total gastrectomy with lower mediastinal andabdominal lymphadenectomy is suitable for E (Ade) tu-mors.41 Subtotal esophagectomy and extended total gas-trectomy are recommended for all subtypes of EGJ cancer.42

When EGJ cancer is beyond its early stage, esophagectomyand regional lymph node resection combined with proximalgastrectomy or total gastrectomy are recommended. Forpreoperative patients with Stage T1 tumors and with no ev-idence of lymph node metastasis, a limited resection of theproximal stomach, cardia, and distal esophagus permits acomplete EGJ (Sq or Ade) tumor resection, adequate lym-phadenectomy, and excellent functional results.

The risk of postoperative complications and quality oflife are crucial. Surgery alone can no longer be consideredthe standard of care with little change in a poor prognosisassociated with carcinoma of the distal esophagus and EGJ.This leads to a debate on the optimal surgical approach andextent of regional lymphadenectomy. The rate of locore-gional and distant recurrence after surgery is high. There-fore, multimodality therapy should be selected for thesepatients, unless indicated otherwise. The multimodalitytherapy with radiotherapy, chemotherapy, and surgeryconfers the greatest opportunity for margin-negativeresection, improved locoregional control, and a cure inpatients with locally advanced esophagogastric carcinomas.Neoadjuvant therapy shrinks the primary tumor, lowers theT and N categories, and eradicates micrometastases.

5.2. Roles of lymphadenectomy in EGJ cancer

Tumors limited to the T1a (mucosa) and pT1b (submucosa)of E (Ade) and EGJ (Ade) practically never present withlymph node metastases.12 The new technology of the

endoscopic mucosal resection offers an even more limitedapproach to early EGJ tumors. Because lymphadenectomyis not possible with this technique, a minimally invasiveendoscopic mucosal resection can be recommended onlyfor patients with tumors limited to the mucosa or superfi-cial submucosa.43 Once an invasion is beyond approxi-mately the first third of the submucosa,44 the risk of lymphnode involvement becomes significant, and esophagealresection is recommended for staging. Endoscopic resectionmay typically be preferable to photodynamic therapy,because the depth of invasion and the margins can bepathologically confirmed. Furthermore, the inaccuracy ofcurrent preoperative staging modalities, including endo-scopic ultrasound for differentiating mucosal from submu-cosal tumors, currently limits the broad clinical applicationof these already-limited procedures.

Several similar encouraging data with limited resectionin patients with early tumor stages have been reported,with a limited risk of lymph node metastasis. Transhiatalesophagectomy with partial gastrectomy for EGJ (Sq orAde) tumors allows the use of gastric remnants for recon-struction, and avoids the need for a thoracotomy. Underthis condition, the mediastinal lymph node dissection isrecommended.41 Resection with thoracic approaches al-lows an extensive resection of periesophageal tissuecompared with transhiatal resection in multi-institutionalrandomized trials.45 Regardless, the pattern of recurrenceremains similar, and perioperative morbidity is substan-tially less in transhiatal resection, with similar mortality.The advantages of this approach include less discomfortand a lower risk of respiratory complications. However, nosignificant difference in long-term survival between trans-thoracic and transhiatal esophagectomy was observed inpatients with E (Ade) tumors.12 Because cervical or medi-astinal lymph node metastasis has not been noted in pa-tients with C (Ade) tumors, subtotal esophagectomy for C(Ade) tumors is not recommended,19 provided that a com-plete tumor resection is achieved. However, this surgicallymanaged population exhibits improved survival for distalesophagus and EGJ locations compared with gastric cardia,perhaps reflecting the tendency toward detection at amore advanced stage during diagnosis for cardiaclocations.46,47

5.3. Multimodal therapy

The advantages of neoadjuvant therapy include the po-tential for tumor downstaging, rendering tumors resect-able, and eradicating metastatic lesions. Although apathologic complete response (pCR) can increase the like-lihood of R0 resection, a complete response is not manda-tory to achieve a negative margin; tumor downstaging maybe sufficient, particularly in adenocarcinoma, where peri-operative chemotherapy results in significant downstagingand increased R0 resections, despite relatively few com-plete tumor regressions. Patients with locally advancedtumors, in whom preoperative staging implies that thelikelihood of complete tumor removal by primary resectionis questionable, should be evaluated for multimodaltreatment. However, in localized disease, less than half ofpatients are cured through surgery, even when combined

190 P.-M. Huang, C.-N. Chen

with perioperative chemotherapy or CRT.14,48 Advancedesophageal cancer is a heterogeneous group of diseases, forwhich not all patients benefit in a similar manner from themultimodality approach. Only 40e50% of patients respondto neoadjuvant CRT.49 Data obtained from randomizedstudies using a purely neoadjuvant therapy are inconsis-tent, and further research is required to delineate the roleof multimodal therapy in EGJ cancer.

5.4. Neoadjuvant chemotherapy in EGJ cancer

Reim et al50 retrospectively analyzed 394 patients whounderwent preoperative platin-based chemotherapy fol-lowed by surgery for locally advanced EGJ (Ade) tumors in asingle-center study. Among them, 100 patients (25%)exhibited a histopathological response (<10% of residualtumor), which was significantly associated with OS(p < 0.0001). A lower baseline cT stage and baseline cN0stage were significantly associated with the histopatholog-ical response (p Z 0.015 and p Z 0.002, respectively). AFederation Nationale des Centres de Lutte Contre le Cancerand Federation Francophone de Cancerologie Digestive(FFCD) multicenter Phase III trial found that perioperativechemotherapy including 5-FU plus cisplatin for resectablegastroesophageal adenocarcinoma significantly improvedthe curative resection rate (84% vs. 73%, p Z 0.04), 5-yeardisease-free survival rate [34% vs. 19%, hazard ratio(HR) Z 0.65, p Z 0.003], and 5-year OS (38% vs. 24%,HR Z 0.69, p Z 0.02) compared with surgery alone (Table1).51 Another meta-analysis including eight randomizedtrials that involved a comparison of neoadjuvant chemo-therapy with surgery alone for resectable esophagogastriccancer demonstrated a survival benefit for neoadjuvantchemotherapy [HR Z 0.90 (0.81e1.00), p Z 0.05]. How-ever, this benefit was compelled by that for adenocarci-noma [HR Z 0.78 (0.64e0.95), p Z 0.014], because thedifference was nonsignificant for E (Sq) tumors [HR Z 0.88(0.75e1.03), p Z 0.12].52 Neoadjuvant chemotherapy is nolonger considered a standard approach for E (Sq) tumors.Neoadjuvant chemotherapy with a cisplatin-based regimenfollowed by surgical resection considerably improves sur-vival in patients with locally advanced EGJ (Ade) tumors.12

However, certain patients fail to achieve a measurableresponse to chemotherapy, and exhibit chemotherapy-related toxic effects. The identification of such non-responders may potentially improve the treatment efficacy,and spare patients from serious adverse effects. Moreover,the rate of locoregional tumor recurrence is high, indicatingthat a combination of preoperative chemotherapy withradiotherapy should be considered in future studies.

5.5. Radiotherapy in EGJ cancer

Radiotherapy may reduce the risk of local recurrence, andmay be particularly beneficial for patients with E (Sq) andEGJ (Sq) tumors because of their highly radiosensitive na-ture. Crucial concerns regarding radiotherapy include un-certainty for whether radiation synergizes the benefit ofneoadjuvant chemotherapy and the appropriate dose. Foradministering radiotherapy in EGJ cancer, the treatment ofceliac nodes is recommended. Moreover, a radial margin of

1.5 cm around the periesophageal lymphatics is suggestedbased on endoscopic, CT, and PET observations. The dosesused for preoperative treatment range from 4140 cGy to5040 cGy, in fractionated doses of 180e200 cGy/d. Higherdoses did not improve the outcome in a trial involving pa-tients with adenocarcinoma and SCC. In a randomizedPhase III trial in patients with EGJ (Ade) tumors, preoper-ative radiation with 40 Gy improved the survival ratessubstantially compared with primary resection.53 Radio-therapy in combination with chemotherapy may play amajor role in neoadjuvant or adjuvant therapy targetinglocally advanced EGJ cancer; however, this remainscontroversial because certain trials have demonstrated thebenefits of chemotherapy alone. Trials using radiotherapyhave found a tendency toward improved local control andsurvival outcome. Therefore, radiotherapy is a rationalapproach in multiple modality therapy when treatmentvolumes are safe, with no other substantial contraindica-tions. A Cochrane meta-analysis reported the superiority ofa combined approach over radiotherapy alone. Based onthese data, the use of radiotherapy alone cannot be rec-ommended for EGJ cancer. However, a poor performancestatus or contraindications to CRT may force patients to optfor radiotherapy.

5.6. Neoadjuvant CRT in EGJ cancer

Neoadjuvant therapy delivers both locoregional and sys-temic treatments, and is better tolerated compared withadjuvant therapy. In a report involving 2920 surgicallytreated patients, 5-year and 10-year survival rates for N1(1e2 nodes) disease were 34% and 24%, respectively, and asignificantly worse outcome was observed in patients whounderwent surgery alone and those with more nodes; thismay encourage the use of combined modality therapies forimproved outcomes.54 A combination of chemotherapy andradiotherapy yields more satisfactory results, although thisresult has yet to be confirmed in EGJ cancer. PreoperativeCRT is a common treatment option for EGJ carcinoma in theUnited States. A meta-analysis including 10 randomizedstudies comparing neoadjuvant CRT with surgery alone inpatients with E (Sq) tumors confirmed a 13% survival dif-ference at 2 years.52 Neoadjuvant CRT is superior to neo-adjuvant chemotherapy alone for improving resectabilityand maintaining locoregional control. A randomized trialreported a higher survival benefit with neoadjuvant CRTcompared with primary resection in patients with poten-tially resectable E (Ade) and EGJ (Ade) tumors.55 The largerChemoradiotherapy for Esophageal Cancer followed bySurgery Study (CROSS) trial and a study by the University ofMichigan reported significant improvements in survival, anda reduced incidence of local or regional failure.14,55,56 The3-year survival rate was 32% versus 6e16% in both adeno-carcinoma and SCC. In the CROSS trial, 49% of SCC patientstreated with CRT achieved a complete response comparedwith 23% of patients with adenocarcinoma (p Z 0.008).Neither regimen was superior to 5-FU-containing regimensfor improving the outcome regarding this disease. Sudoet al57 reported that, among 518 patients with E (Ade) tu-mors who completed trimodality therapy, 188 patients(36%) exhibited distant metastases, with locoregional

Table 1 Results of prospective randomized trials in esophagogastric junction cancer since the year 2000.

Tumor type Treatment type Author Year Regimens 3-yearsurvival

5-yearsurvival

% localrelapse

% distantrelapse

p value ofsurvival

E (Ade) Neoadjuvantchemoradiotherapy

Urba et al.14 2001 2 cycles of CF þvinblastin and 45 Gy

30%

E (Ade) Surgery alone Urba et al. 2001 16.0% 0.15E (Sq) /EGJ (Sq) Neoadjuvant

chemoradiotherapyBurmeister et al.65 2005 1 cycle of CF and 35 Gy

E (Sq) /EGJ (Sq) Surgery alone Burmeister et al. 2005 HR0.69 0.16E (Ade)/ EGJ (Ade) Neoadjuvant

chemoradiotherapyBurmeister et al. 2005 1 cycle of CF and 35 Gy

E (Ade)/ EGJ (Ade) Surgery alone Burmeister et al. 2005 HR1.04 0.81E (Ade)/ EGJ (Ade) Perioperative

chemotherapyCunningham et al./Magic48 2006 3 cycles of pre-op CF

and epirubicin and3 cycles of postop CFand epirubicin

36.3% 14.40% 24.40%

E (Ade)/ EGJ (Ade) Surgery alone Cunningham et al./Magic 2006 23.0% 20.60% 36.80% 0.009E (Sq) Neoadjuvant

chemoradiotherapyNatsugoe et al.86 2006 2 cycles of CF and 40 Gy 57.0% 5% 45.50%

E (Sq) Surgery alone Natsugoe et al. 2006 41.0% 4% 56.50% 0.58E (Sq) /EGJ (Sq) Neoadjuvant

chemotherapyKelsen et al./ROTG8911,USA Intergroup 11384

2007 3 cycles of CF 23% 19% 49%

E (Ade)/ EGJ (Ade) Neoadjuvantchemotherapy

Kelsen et al./ROTG8911,USA Intergroup 113

2007 3 cycles of CF 26% 21% 51%

Surgery alone Kelsen et al./ROTG8911,USA Intergroup 113

2007 HR 0.78 0.003

E (Ade)/ EGJ (Ade) Neoadjuvantchemotherapy

Stahl et al.60 2009 2.5 cycles of PLF 27.70% 23.70% 22%

E (Ade)/ EGJ (Ade) Neoadjuvantchemoradiotherapy

Stahl et al. 2009 2 cycles of PLF then3 weeks of CRT

47.40% 15% 16.70% 0.07

E (Ade)/ EGJ (Ade) Neoadjuvantchemotherapy

Allum et al./OEO262 2009 2 cycles of CF 22.6%

E (Ade)/ EGJ (Ade) Surgery alone Allum et al./OEO2 2009 17.6%E (Sq) /EGJ (Sq) Neoadjuvant

chemotherapyAllum et al./OEO2 2009 2 cycles of CF 25.5%

E (Sq) /EGJ (Sq) Surgery alone Allum et al./OEO2 2009 17.0%E(Sq/ Ade) /EGJ

(Sq/ Ade)Allum et al./OEO2 2009 HR0.84 23% vs. 17.1% 0.03

E (Ade)/ EGJ (Ade) Perioperativechemotherapy

Ychou et al./FNCLCC-FFCD51

2011 2e3 cycles of pre-opCF and 3e4 cycles ofpostop CF

38.0% 12% 30%

E (Ade)/ EGJ (Ade) Surgery alone Ychou et al./FNCLCC-FFCD

2011 24.0% 8% 38% 0.02

CF Z cisplatin and 5-fluorouracil; CRT Z chemoradiotherapy; HR Z hazard ratio; PLF Z cisplatin, fluorouracil, and leucovorin.

Therapeutic

strategie

sforEGJca

nce

r191

192 P.-M. Huang, C.-N. Chen

failure noted in 27 patients (5%). Most locoregional failures(89%) occurred within 36 months of surgery. The median OSof 27 patients with locoregional failure was 17 months, and10 patients (37%) survived for longer than 2 years. Only 2%of all 518 patients benefited from surveillance or salvagestrategies. Locoregional failures are rare after trimodalitytherapy; however, salvage strategies are not considerablybeneficial. An improved local control of the tumor may notoffer an advantage for survival in patients with gastro-esophageal adenocarcinoma who do not respond to initialchemotherapy and they are instead treated with salvageradiotherapy.

Although R0 resection rates are improved after the addi-tion of radiotherapy, the survival yields no improvementbecause of the high occurrence rate of systemic metastaticdisease.58 Most patients with EGJ (Ade) tumors are morelikely to experience a relapse with distant disease, andtherefore, the systemic approach with chemotherapy islikely to bemore beneficial for these patients comparedwitha localized treatment strategy.18,47,59 The German Esopha-geal Study Group 39 conducted a multi-institutional ran-domized trial of Siewert Type IeIII adenocarcinoma by usingneoadjuvant platinum and 5-FU chemotherapy with orwithout radiotherapy.60 The trial was terminated earlybecause of insufficient power. No definitive conclusion wasdrawn because of the modest sample size. However, animprovement in the 3-year survival rate and local progressionwas noted in the overallmeta-analysis. Postoperative deathsincreased when radiotherapy was used (4% vs. 10%,p Z 0.26), although the lengths of intensive care unit stayand hospitalization were similar. The incidence of localfailure was nonsignificantly lower with neoadjuvantchemotherapy alone in the US intergroup [Radiation TherapyOncology Group (RTOG) 8911]61 and the MRC OEO213,62 trials(Table 1). Neoadjuvant chemotherapy-related deaths werereported in 2% of patients, with a nonsignificant increase inpostoperative mortality (6% in both arms) or morbidity (26%vs. 31% in the chemotherapy arm).

The results from the trials using neoadjuvant chemo-therapy or CRT followed by surgery are encouraging; how-ever, they have demonstrated that only patients with majorpCR benefit from neoadjuvant therapy. Moreover, thesetherapies are expensive. When a combination of cisplatinand 5-FU with radiotherapy (45 Gy) in 25 fractionated doseswas administered to patients with Stage I or II esophagealcancer, no improvement in survival was observed with theaddition of neoadjuvant CRT to surgery.63 This indicates alack of benefit from multimodality management in the earlydisease stage. Regardless of these advances, the selectionof the most beneficial treatment strategy for patients withEGJ cancer remains challenging, and is influenced by de-ficiencies in predictive and prognostic markers. Blanket al64 examined preoperative therapy targeting esoph-agogastric cancer, and reported that nonresponding EGJ(Ade) patients do not have a disadvantage compared withprimary resection patients, but nonresponders with E (Sq)tumors had a worse prognosis. This finding warrants acritical patient selection for this tumor entity to be treatedsurgically. This indicates a lack of benefit from multi-modality management in the early disease stage. Moreover,an insufficient dosage of chemotherapy or radiation con-tributes to lower discrimination.65

5.7. Definitive CRT in EGJ cancer

Definitive CRT has become the frequent choice of treatmentfor patients with E (Sq) tumors because of its superiority toradiotherapy alone and noninferiority to neoadjuvant CRTfollowed by surgery. Definitive CRTwithout surgery can cureSCC patients, with an improved 5-year OS for 26% of patients,as demonstrated in the RTOG 85-01 trial including 121 pa-tients. The trial results revealed that only 7% of enrolledpatients had adenocarcinoma. For EGJ (Ade) tumors, evi-dence supporting the benefits of definitive CRT withoutsurgery is scant. However, this approach must be reeval-uated in patients who are not surgical candidates. The trialresults showed no benefit in the 2-year OS, with similarlocoregional failure rates between high and low doses ofradiotherapy (31% of 64.8-Gy CRT vs. 40% of 50.4-Gy CRT). Arandomized study involving the use of definitive CRT or sur-gical approach in 80 patients with SCC indicated that theperioperative mortality after salvage surgery for persistentor recurrent disease was 16.7% versus 6.8%, respectively, inthe surgery-only group, with no significant difference in themedian disease-free survival or 2-year OS (58.3% vs. 54.5%,respectively, with surgery).66 The FFCD 9102 trial involvedtreatment for 444 patients with localized (T3N0e1M0)esophageal cancer using conventional CRT, and 259 patients(58.3%) who responded to the initial CRTwere randomized toreceive either surgery or definitive CRT. Two-year OS wasfound for 40% of patients treated with definitive CRTcompared with 34% in the trimodality arm (p Z 0.03). Thisstudy confirmed definitive CRT as a standard treatment op-tion for patients who respond to initial CRT.67

5.8. Targeted therapy in EGJ cancer

Because of the high incidence and lethal nature of this dis-ease, novel therapies, including EGFR-inhibiting monoclonalantibodies (cetuximab, panitumumab, and matuzumab) andHER-2-inhibiting monoclonal antibodies (trastuzumab andpertuzumab), bevacizumab, lapatinib, and everolimus, areundergoing Phase III testing for patients with esoph-agogastric cancer (Figure 2). Several potential extracellularand intracellular antitumor mechanisms of the blockade ofHER signaling have been identified, including the activationof antibody-dependent cellular cytotoxicity, reduction ofangiogenesis, and decreased DNA repair. In an advanced E(Sq) tumor setting, a randomized Phase II study evaluatingthe addition of cetuximab to standard doses of cisplatin plus5-FU chemotherapy found manageable toxic effects andsimilar response rates, but prolonged progression-free sur-vival (5.9 months vs. 3.6 months) and OS (9.5 months vs.5.5 months) after the addition of cetuximab compared withchemotherapy alone.68 However, the study was not poweredsufficiently to determine a survival difference. Anotherstudy evaluated cetuximab in combination with cisplatin,irinotecan, and CRT in locally advanced E (Sq or Ade) tumors.A combination of cetuximab and radiotherapy results in aleast comparable pCR rate compared with chemotherapyand radiotherapy.69 The pCR ratewas 67% for SCC and 28% foradenocarcinoma. An earlier disease stage was associatedwith increased pCR. Becerra et al69 conducted a Phase II,open-label, single-arm, multicenter study with cetuximab

Therapeutic strategies for EGJ cancer 193

and radiotherapy in patients with surgically resectableesophageal and EGJ cancer. Thirty-nine patients completedthe study, most of whom (78%) had a histology of adenocar-cinoma. The results showed that tumors were located in theesophagus (63%). Cetuximab and radiotherapy result in asimilar pCR rate, with superior tolerance compared withpreoperative chemotherapy and radiotherapy in patientswith resectable esophageal cancer. However, the results ofongoing larger randomized studies are required to determinewhether cetuximab plays a role in this regard.

Trastuzumab is a licensed drug indicated for EGJ (Ade)tumors and gastric cancer in the European Union. In aprevious study, patients with an HER-2 immunohistochem-istry (IHC) score of 3 or more, or an IHC score of 2 plus geneamplification (detected by fluorescence in situ hybridiza-tion) had the greatest benefit from trastuzumab.70 Themedian OS was 16 months [HR Z 0.65, 95% confidence in-terval (CI) Z 0.51e0.83]. The sensitivity of EGFR-TKIs de-pends on the primary-tumor location. Patients with gastriccancer exhibit no activity with the therapy, whereas pa-tients with the EGJ (Ade) tumors exhibit some activity afterthe failure of surgery and radiotherapy.71,72 The efficacy oftherapy in these studies is not impressive, and research onEGFR-TKI monotherapy in esophagogastric cancer is scant.The only clinically approved targeted therapy is trastuzu-mab, which targets the HER-2/NEU oncogene. However,only a small group of patients with esophageal cancerexhibit amplified HER-2. With the identification of otheressential targets and pathways, several other therapies areanticipated for approval in the future.

5.9. Predictive molecular markers of neoadjuvanttherapy in EGJ cancer

Patients with EGJ cancer are often treated with similartherapies; however, they exhibit different outcomes. Theprognoses of patients who do not respond to trimodalitytherapy strategies are inferior to those of patients who un-dergo surgery. Moreover, trimodality therapy strategies canresult in significant undesirable consequences. The devel-opment of validated predictive molecular markers can aid inidentifying patients who are more likely to be responders,and in the selection of more efficient strategies involving atailored therapy. Predictive molecular markers for prevent-ing expensive, noneffective, and potentially harmful thera-pies must be investigated. Predictive markers are evaluatedthrough the clinical response to specific therapy, duration ofprogression, and associated toxicity. Molecular markers canbe protein targets, such as growth factor receptors, angio-genesis enzymes, cell-cycle regulators, gene expressionlevels, and genetic variations in the tumor or host. However,disease progression is greatly influenced by complex path-ways, and the analysis of a single marker is insufficient forpredicting disease progression with reproducibility.

The predictive value of EGFR and HER-2/NEU messengerRNA (mRNA) expression in E (Ade) tumors after neoadjuvantCRT with cisplatin and 5-FU followed by transthoracicesophagectomy is not associated with the degree of histo-pathological response; the low intratumoral expressionlevels of HER-2/NEU are significantly associated with pCR.73

The roles of p53 and survivin mRNA as predictive markers of

pCR in patients with locally advanced EGJ cancer whoreceived trimodality therapy are minor and inconsistent.74

Cyclooxygenase-2 (COX-2) is involved in the regulation oftumor progression, angiogenesis, and resistance tochemotherapy. High COX-2 expression is less sensitive toneoadjuvant CRT, and has demonstrated an extremely poorprognosis.75

The excision repair cross-complementing 1 (ERCC1) geneencodes an enzyme required in the nucleotide excisionrepair pathway. A more efficient DNA repair system maylead to increased clinical resistance to a chemotherapeuticagent. The overexpression of ERCC1 mRNA may indicate apoor response to neoadjuvant chemotherapy.76 Geneexpression microarray profiling may be a powerful tool tosystematically search for the molecular markers of cancerclassification and outcome prediction. This is a potentiallyvaluable technology for separating patients with EGJ can-cer sensitive to trimodality treatment regimens from thosewho are resistant.77 The hypermethylation of the CpG is-land methylator phenotype promoter is responsible forsilencing and inactivating the tumor suppressor gene. Thenumber of methylated genes per patient is significantlylower in patients who experienced pCR compared withthose who did not.78

5.10. Radical surgery following inductionneoadjuvant therapy for Stage IV EGJ cancer

With the introduction of new chemotherapeutic regimensand targeted agents, improved survival can be attained,even for metastatic cases. Only a few studies havecompared tumor resection in Stage IV EGJ cancer withpalliative chemotherapy. Most studies have reported asurvival benefit from resection; however, this can be due toa selection bias. Reevaluating the combination of chemo-therapy with radical surgery in a large patient populationmay lead to median survival times of 14.6e25.4 months,with a remarkably low postoperative mortality.70,79 How-ever, a complete tumor resection and histopathologicaltumor regression remain two independent prognostic fac-tors.80 Patients responding to chemotherapy have a greaterlikelihood of R0 resection, which explains the use ofaggressive chemotherapy agents in palliative treatment.

5.11. Adjuvant therapy in EGJ cancer

Adjuvant chemotherapy in patients after esophagogas-trectomy remains challenging. Common reasons for notstarting postoperative chemotherapy are disease progres-sion or early death, patient choice, postoperative compli-cations, and previous chemotherapy-related toxicity, asreported in the MAGIC trial on a small patient populationcompleting adjuvant treatment.48 Surgery with post-operative therapy presents potential advantages. Theremoval of the primary tumor may help patients who didnot respond well to neoadjuvant therapy, or those with alocally advanced tumor. In addition, adjuvant therapy canadd therapeutic effects until full surgical recovery isensured following neoadjuvant therapy and surgery.Although a universally accepted standard treatment is un-available for advanced-stage EGJ cancer, these treatments

194 P.-M. Huang, C.-N. Chen

alone have been found to have a median survival of9e12 months in randomized controlled trials.81 No signifi-cant benefit has been found in 5-year OS after the additionof adjuvant chemotherapy82 or adjuvant radiotherapy83 toR0 resection; however, the studies were limited because oftheir small samples.

The data necessary for the recommended use of adju-vant chemotherapy for resected EGJ adenocarcinoma areinsufficient. Several studies have indicated that adjuvantchemotherapy is poorly tolerated, with only 50% of patientscompleting the planned postoperative therapy. Anotherstudy reported that postoperative complications and mor-tality rates were similar across all arms, and that the 5-yearsurvival rates improved significantly, from 23% to 36%.48

Data on the proportion of patients with EGJ (Ade) neces-sary to measure the relevance to postoperative CRT arelacking. An adjuvant CRT group and surgery alone or be-tween histologic types exhibited similar outcomes. Inanother study, the 5-year survival rate was similar across allarms, and was influenced only by achieving the R0 resection(32% vs. 5%, respectively, for lesser resections).84 TheIntergroup 0116 study investigated adjuvant CRT in 556randomized patients with resected adenocarcinoma of thestomach and EGJ (20%)85; 64% of patients completed theplanned treatment, and the 3-year OS was 50% in the CRTgroup versus 41% in the surgery-only group (HR Z 1.35, 95%CI Z 1.09e1.66, p Z 0.005). These data suggest thatadjuvant CRT should be considered for palliative treatmentor when surgical risks are high, whereas surgery alone re-mains a potentially effective method that can cure patientswho cannot tolerate standard trimodality therapy.

5.12. Further directions in the management of EGJcancer

Although diagnostic tools (e.g., endoscopic ultrasound) andsurgical skills (e.g., thoracoscopic and laparoscopic pro-cedures) have been improved, the survival of patients withEGJ cancer remains low. Type E (Sq or Ade) and EGJ (Sq orAde) tumors in this study were categorized as esophagealcancer, in accordance with UICC/AJCC criteria; patientswith these tumors exhibited differences in clinicopatho-logic characteristics. With an emphasis on EGJ tumors as awell-defined entity, future studies are likely to provideprecise classification and optimal management. Currentchemotherapy regimens, although real, lead to minor im-provements. Other agents are being investigated worldwidefor metastatic esophageal cancer and EGJ cancer, andpromising agents should be evaluated along with standardneoadjuvant CRT. Recent research on the pathogenesis ofEGJ cancer, a brief review of the commonly known geneticpathways in EGJ cancer, and the identification of newgenes with DNA microarray analysis will guide the devel-opment of future therapies. The identification of molecularmarker subsets that can predict responses to multimodalitytreatment strategies, the assessment of disease aggres-siveness using prognostic markers, and the likelihood ofrecurrence after surgery will assist clinicians in personal-izing the combined modality treatment by optimizing drugefficacy and minimizing adverse effects. The variables ofgenetic polymorphisms in drug targets, metabolizing

enzymes, transporters, and influential receptors are crucialfor analyzing drug efficacy.

The traditionally used CRT with platinum plus 5-FU maybe replaced with new regimens, including taxanes and iri-notecan, which enable ease of prescription. Further data ona comparison between chemotherapy and CRT are availablein the MAGIC versus CROSS trial (ClinicalTrials.gov identifier:NCT01726452). HER-2 is currently the most well-establishedtarget, based on the results of the ToGA trial,70 whichshowed that the addition of trastuzumab to chemotherapyimproves OS in patients with advanced gastric or EGJ cancer.The use of trastuzumab or other anti-HER-2 therapies in aneoadjuvant or adjuvant setting is currently being evaluatedin various clinical trials (ClinicalTrials.gov identifiers:NCT01748773, NCT01196390, and NCT00450203).

6. Conclusion

A detailed preoperative staging and assessment of locore-gional resectability and the exclusion of hepatic metastasesor peritoneal tumor spread are conducted using CT, PET,and endoscopy in the staging of malignancies at the EGJ.Early stage disease with no evidence of nodal spread andinvasion confined to the submucosa was found to have agood prognosis, with complete local resection to ensurenegative margins, including lymph node dissection. Patientswith lymph node metastasis are required to confirm thebenefit of combined modality therapy, particularly becauseceliac axis nodes are considered regional nodes, rather thanM1a metastasis. Extended lymphadenectomy in the lowerposterior mediastinum and the celiac area may improve theprognosis of patients with a limited number of positivelymph nodes. Certain long-term survival data for EGJ can-cer have been summarized in this review, indicating thenecessity to improve the outcome using aggressive com-bined modality therapy, particularly for patients withadvanced-stage cancer. Neoadjuvant chemotherapy is nolonger considered a standard approach for esophagogastricSCC, and has been superseded by neoadjuvant or definitiveCRT. Neoadjuvant chemotherapy is an appropriate treat-ment option for locally advanced esophagogastric adeno-carcinoma, and evidence implies that CRT yields optimalresults. Definitive CRT is a valid treatment option forresectable esophagogastric SCC, avoiding potential surgicalmorbidity and mortality, with salvage surgery reserved forthose with persistent cancer or those noted with a tumorrelapse. However, neoadjuvant CRT followed by surgery is avalid option for patients at lower surgical risk with nocomorbidities. Several surgical approaches are available,however, additional data are required to select the mostappropriate approach for patients in R0 resection to lowerthe risk of complications and improve the quality of life.

References

1. American Cancer Society Facts and Figures 2013. Atlanta, GA:American Cancer Society; 2013.

2. World Health Organization. Cancer Surveillance Database.Available from: http://www.dep.iarc.fr. [Accessed 22 May2014].

Therapeutic strategies for EGJ cancer 195

3. Deans C, Yeo MS, Soe MY, Shabbir A, Ti TK, So JB. Cancer of thegastric cardia is rising in incidence in an Asian population and isassociated with adverse outcome. World J Surg. 2011;35:617e624.

4. Pohl H, Welch HG. The role of overdiagnosis and reclassifica-tion in the marked increase of esophageal adenocarcinomaincidence. J Natl Cancer Inst. 2005;97:142e146.

5. Siewert JR, Stein HJ, Feith M. Adenocarcinoma of the esoph-ago-gastric junction. Scand J Surg. 2006;95:260e269.

6. Edge SB, Compton CC. The American Joint Committee onCancer: the 7th edition of the AJCC cancer staging manual andthe future of TNM. Ann Surg Oncol. 2010;17:1471e1474.

7. Sobin LH, Gospodarowicz MK, Wittekind C. TNM Classification ofMalignant Tumors. 7th ed. Oxford, UK: Wiley-Blackwell; 2010.

8. D’Amico TA, Harpole Jr DH. Molecular biology of esophagealcancer. Chest Surg Clin N Am. 2000;10:451e469.

9. Metzger R, Schneider PM, Warnecke-Eberz U, Brabender J,Holscher AH. Molecular biology of esophageal cancer. Onkolo-gie. 2004;27:200e206.

10. Suh YS, Han DS, Kong SH, et al. Should adenocarcinoma of theesophagogastric junction be classified as esophageal cancer? Acomparative analysis according to the seventh AJCC TNMclassification. Ann Surg. 2012;255:908e915.

11. Hasegawa S, Yoshikawa T, Aoyama T, et al. Esophagus or stom-ach? The seventh TNM classification for Siewert type II/IIIjunctional adenocarcinoma. Ann Surg Oncol. 2013;20:773e779.

12. Stein HJ, Feith M, Siewert JR. Cancer of the esophagogastricjunction. Surg Oncol. 2000;9:35e41.

13. Medical Research Council Oesophageal Cancer Working Group.Surgical resection with or without preoperative chemotherapyin oesophageal cancer: a randomised controlled trial. Lancet.2002;359:1727e1733.

14. Urba SG, Orringer MB, Turrisi A, Iannettoni M, Forastiere A,Strawderman M. Randomized trial of preoperative chemo-radiation versus surgery alone in patients with locoregionalesophageal carcinoma. J Clin Oncol. 2001;19:305e313.

15. Whiteman DC, Parmar P, Fahey P, et al. Association of Heli-cobacter pylori infection with reduced risk for esophagealcancer is independent of environmental and genetic modifiers.Gastroenterology. 2010;139:73e83.

16. Freedman ND, Abnet CC, Leitzmann MF, et al. A prospectivestudy of tobacco, alcohol, and the risk of esophageal andgastric cancer subtypes. Am J Epidemiol. 2007;165:1424e1433.

17. Kodama I, Kofuji K, Yano S, et al. Lymph node metastasis andlymphadenectomy for carcinoma in the gastric cardia: clinicalexperience. Int Surg. 1998;83:205e209.

18. Hosokawa Y, Kinoshita T, Konishi M, et al. Recurrence patternsof esophagogastric junction adenocarcinoma according toSiewert’s classification after radical resection. Anticancer Res.2014;34:4391e4397.

19. Ito H, Inoue H, Odaka N, et al. Clinicopathological character-istics and optimal management for esophagogastric junctionalcancer; a single center retrospective cohort study. J Exp ClinCancer Res. 2013;32:2e13.

20. Leers JM, DeMeester SR, Chan N, et al. Clinical characteristics,biologic behavior, and survival after esophagectomy are similarfor adenocarcinoma of the gastroesophageal junction and thedistal esophagus. J Thorac Cardiovasc Surg. 2009;138:594e602.

21. van Dekken H, Geelen E, Dinjens WN, et al. Comparativegenomic hybridization of cancer of the gastroesophagealjunction: deletion of 14Q31-32.1 discriminates betweenesophageal (Barrett’s) and gastric cardia adenocarcinomas.Cancer Res. 1999;59:748e752.

22. Im SA, Lee KE, Nam E, et al. Potential prognostic significance ofp185(HER2) overexpression with loss of PTEN expression ingastric carcinomas. Tumori. 2005;91:513e521.

23. Bennett WP, Hollstein MC, Metcalf RA, et al. p53 mutation andprotein accumulation during multistage human esophagealcarcinogenesis. Cancer Res. 1992;52:6092e6097.

24. Schimanski CC, Jordan M, Schlaegel F, et al. SNP rs1801157significantly correlates with distant metastasis in CXCL12expressing esophagogastric cancer. Int J Oncol. 2011;39:515e520.

25. Itakura Y, Sasano H, Shiga C, et al. Epidermal growth factorreceptor overexpression in esophageal carcinoma. An immu-nohistochemical study correlated with clinicopathologic find-ings and DNA amplification. Cancer. 1994;74:795e804.

26. Wang KL, Wu TT, Choi IS, et al. Expression of epidermal growthfactor receptor in esophageal and esophagogastric junctionadenocarcinomas: association with poor outcome. Cancer.2007;109:658e667.

27. Jiang Y. Impact of genetic targets on cancer therapy inesophagogastric cancer. Adv Exp Med Biol. 2013;779:55e65.

28. Okines A, Cunningham D, Chau I. Targeting the human EGFRfamily in esophagogastric cancer. Nat Rev Clin Oncol. 2011;8:492e503.

29. Thomaidis T, Maderer A, Al-Batran SE, et al. VEGFR-3 andCXCR4 as predictive markers for treatment with fluorouracil,leucovorin plus either oxaliplatin or cisplatin in patients withadvanced esophagogastric cancer: a comparative study of theArbeitsgemeinschaft Internistische Onkologie (AIO). BMC Can-cer. 2014;14:476e484.

30. Blank S, Rachakonda S, Keller G, et al. A retrospectivecomparative exploratory study on two methylentetrahy-drofolate reductase (MTHFR) polymorphisms in esoph-agogastric cancer: the A1298C MTHFR polymorphism is anindependent prognostic factor only in neoadjuvantly treatedgastric cancer patients. BMC Cancer. 2014;14:58e72.

31. Heeren PA, van Westreenen HL, Geersing GJ, vanDullemen HM, Plukker JT. Influence of tumor characteristics onthe accuracy of endoscopic ultrasonography in staging cancerof the esophagus and esophagogastric junction. Endoscopy.2004;36:966e971.

32. Uedo N, Fujishiro M, Goda K, et al. Role of narrow band imagingfor diagnosis of early-stage esophagogastric cancer: currentconsensus of experienced endoscopists in Asia-Pacific region.Dig Endosc. 2011;23:58e71.

33. Stein HJ, Kraemer SJ, Feussner H, Fink U, Siewert JR. Clinicalvalue of diagnostic laparoscopy with laparoscopic ultrasound inpatients with cancer of the esophagus or cardia. J GastrointestSurg. 1997;1:167e172.

34. Kwee RM. Prediction of tumor response to neoadjuvant therapyin patients with esophageal cancer with use of 18F FDG PET: asystematic review. Radiology. 2010;254:707e717.

35. Lerut T, Nafteux P, Moons J, et al. Three-field lymphadenec-tomy for carcinoma of the esophagus and gastroesophagealjunction in 174 R0 resections: impact on staging, disease-freesurvival, and outcome: a plea for adaptation of TNM classifi-cation in upper-half esophageal carcinoma. Ann Surg. 2004;240:962e972.

36. Dantoc M, Cox MR, Eslick GD. Evidence to support the use ofminimally invasive esophagectomy for esophageal cancer: ameta-analysis. Arch Surg. 2012;147:768e776.

37. Sasako M, Sano T, Yamamoto S, et al. Left thoracoabdominalapproach versus abdominal-transhiatal approach for gastriccancer of the cardia or subcardia: a randomised controlledtrial. Lancet Oncol. 2006;7:644e651.

38. Carboni F, Lorusso R, Santoro R, et al. Adenocarcinoma of theesophagogastric junction: the role of abdominal-transhiatalresection. Ann Surg Oncol. 2009;16:304e310.

39. Gronnier C, Piessen G, Mariette C. Diagnosis and treatmentof non-metastatic esophagogastric junction adenocarci-noma: what are the current options? J Visc Surg. 2012;149:e23ee33.

196 P.-M. Huang, C.-N. Chen

40. Hosokawa Y, Kinoshita T, Konishi M, et al. Clinicopathologicalfeatures and prognostic factors of adenocarcinoma of theesophagogastric junction according to Siewert classification:experiences at a single institution in Japan. Ann Surg Oncol.2012;19:677e683.

41. Kakeji Y, Yamamoto M, Ito S, et al. Lymph node metastasisfrom cancer of the esophagogastric junction, and determina-tion of the appropriate nodal dissection. Surg Today. 2012;42:351e358.

42. Schiesser M, Schneider PM. Surgical strategies for adenocarci-noma of the esophagogastric junction. Recent Results CancerRes. 2010;182:93e106.

43. Stein HJ, Feith M, Bruecher BL, Naehrig J, Sarbia M, Siewert JR.Early esophageal cancer: pattern of lymphatic spread andprognostic factors for long-term survival after surgical resec-tion. Ann Surg. 2005;242:566e573.

44. Raja S, Rice TW, Goldblum JR, et al. Esophageal submucosa:the watershed for esophageal cancer. J Thorac CardiovascSurg. 2011;142, 1403e1411.e1.

45. Hulscher JB, van Lanschot JJ. Individualised surgical treatmentof patients with an adenocarcinoma of the distal oesophagus orgastro-oesophageal junction. Dig Surg. 2005;22:130e134.

46. Kleinberg L. Therapy for locally advanced adenocarcinoma ofthe gastroesophageal junction: optimizing outcome. SeminRadiat Oncol. 2013;23:38e50.

47. Curtis NJ, Noble F, Bailey IS, et al. The relevance of the Sie-wert classification in the era of multimodal therapy foradenocarcinoma of the gastro-oesophageal junction. J SurgOncol. 2014;109:202e207.

48. Cunningham D, Allum WH, Stenning SP, et al. Perioperativechemotherapy versus surgery alone for resectable gastro-esophageal cancer. N Engl J Med. 2006;355:11e20.

49. Krause BJ, Herrmann K, Wieder H, zum Buschenfelde CM. 18F-FDG PET and 18F-FDG PET/CT for assessing response to therapyin esophageal cancer. J Nucl Med. 2009;50:89Se96S.

50. Reim D, Gertler R, Novotny A, et al. Adenocarcinomas of theesophagogastric junction are more likely to respond to pre-operative chemotherapy than distal gastric cancer. Ann SurgOncol. 2012;19:2108e2118.

51. Ychou M, Boige V, Pignon JP, et al. Perioperative chemotherapycompared with surgery alone for resectable gastroesophagealadenocarcinoma: an FNCLCC and FFCD multicenter phase IIItrial. J Clin Oncol. 2011;29:1715e1721.

52. Gebski V, Burmeister B, Smithers BM, et al. Survival benefitsfrom neoadjuvant chemoradiotherapy or chemotherapy inoesophageal carcinoma: a meta-analysis. Lancet Oncol. 2007;8:226e234.

53. Zhang ZX, Gu XZ, Yin WB, Huang GJ, Zhang DW, Zhang RG.Randomized clinical trial on the combination of preoperativeirradiation and surgery in the treatment of adenocarcinoma ofgastric cardia (AGC)dreport on 370 patients. Int J RadiatOncol Biol Phys. 1998;42:929e934.

54. Gertler R, Stein HJ, Langer R, et al. Long-term outcome of2920 patients with cancers of the esophagus and esoph-agogastric junction: evaluation of the New Union Inter-nationale Contre le Cancer/American Joint Cancer Committeestaging system. Ann Surg. 2011;253:689e698.

55. Walsh TN, Noonan N, Hollywood D, Kelly A, Keeling N,Hennessy TP. A comparison of multimodal therapy and surgeryfor esophageal adenocarcinoma. N Engl J Med. 1996;335:462e467.

56. van Hagen P, Hulshof MC, van Lanschot JJ, et al. Preoperativechemoradiotherapy for esophageal or junctional cancer. N EnglJ Med. 2012;366:2074e2084.

57. Sudo K, Taketa T, Correa AM, et al. Locoregional failure rateafter preoperative chemoradiation of esophageal adenocarci-noma and the outcomes of salvage strategies. J Clin Oncol.2013;31:4306e4310.

58. zum Buschenfelde CM, Herrmann K, Schuster T, et al. 18F-FDGPET-guided salvage neoadjuvant radiochemotherapy ofadenocarcinoma of the esophagogastric junction: the MUNI-CON II trial. J Nucl Med. 2011;52:1189e1196.

59. Oppedijk V, van der Gaast A, van Lanschot JJ, et al. Patterns ofrecurrence after surgery alone versus preoperative chemo-radiotherapy and surgery in the CROSS trials. J Clin Oncol.2014;32:385e391.

60. Stahl M, Walz MK, Stuschke M, et al. Phase III comparison ofpreoperative chemotherapy comparedwith chemoradiotherapyin patients with locally advanced adenocarcinoma of theesophagogastric junction. J Clin Oncol. 2009;27:851e856.

61. Kelsen DP, Ginsberg R, Pajak TF, et al. Chemotherapy followedby surgery compared with surgery alone for localized esopha-geal cancer. N Engl J Med. 1998;339:1979e1984.

62. Allum WH, Stenning SP, Bancewicz J, Clark PI, Langley RE.Long-term results of a randomized trial of surgery with orwithout preoperative chemotherapy in esophageal cancer. JClin Oncol. 2009;27:5062e5067.

63. Bosset JF, Gignoux M, Triboulet JP, et al. Chemoradiotherapyfollowed by surgery compared with surgery alone in squamous-cell cancer of the esophagus. N Engl J Med. 1997;337:161e167.

64. Blank S, Stange A, Sisic L, et al. Preoperative therapy ofesophagogastric cancer: the problem of nonresponding pa-tients. Langenbecks Arch Surg. 2013;398:211e220.

65. Burmeister BH, Smithers BM, Gebski V, et al. Surgery aloneversus chemoradiotherapy followed by surgery for resectablecancer of the oesophagus: a randomised controlled phase IIItrial. Lancet Oncol. 2005;6:659e668.

66. Chiu PW, Chan AC, Leung SF, et al. Multicenter prospectiverandomized trial comparing standard esophagectomy withchemoradiotherapy for treatment of squamous esophagealcancer: early results from the Chinese University ResearchGroup for Esophageal Cancer (CURE). J Gastrointest Surg.2005;9:794e802.

67. Bedenne L, Michel P, Bouche O, et al. Chemoradiation followedby surgery compared with chemoradiation alone in squamouscancer of the esophagus: FFCD 9102. J Clin Oncol. 2007;25:1160e1168.

68. Lorenzen S, Schuster T, Porschen R, et al. Cetuximab pluscisplatin-5-fluorouracil versus cisplatin-5-fluorouracil alone infirst-line metastatic squamous cell carcinoma of the esoph-agus: a randomized phase II study of the ArbeitsgemeinschaftInternistische Onkologie. Ann Oncol. 2009;20:1667e1673.

69. Becerra CR, Hanna N, McCollum AD, et al. A phase II study withcetuximab and radiation therapy for patients with surgicallyresectable esophageal and GE junction carcinomas: HoosierOncology Group G05-92. J Thorac Oncol. 2013;8:1425e1429.

70. Bang YJ, Van Cutsem E, Feyereislova A, et al. Trastuzumab incombination with chemotherapy versus chemotherapy alonefor treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): a phase 3, open-label,randomised controlled trial. Lancet. 2010;376:687e697.

71. Dragovich T, McCoy S, Fenoglio-Preiser CM, et al. Phase II trialof erlotinib in gastroesophageal junction and gastric adeno-carcinomas: SWOG 0127. J Clin Oncol. 2006;24:4922e4927.

72. Geldart T, Astras G. Isolated CNS relapse in human epidermalgrowth factor receptor 2-positive esophagogastric cancer:effective treatment with trastuzumab after failure of surgeryand radiotherapy. J Clin Oncol. 2011;29:e664ee665.

73. Miyazono F, Metzger R, Warnecke-Eberz U, et al. Quantitativec-erbB-2 but not c-erbB-1 mRNA expression is a promisingmarker to predict minor histopathologic response to neo-adjuvant radiochemotherapy in oesophageal cancer. Br JCancer. 2004;91:666e672.

74. Vallbohmer D, Peters JH, Oh D, et al. Survivin, a potentialbiomarker in the development of Barrett’s adenocarcinoma.Surgery. 2005;138:701e706.

Therapeutic strategies for EGJ cancer 197

75. Xi H, Baldus SE, Warnecke-Eberz U, et al. High cyclooxygenase-2 expression following neoadjuvant radiochemotherapy isassociated with minor histopathologic response and poorprognosis in esophageal cancer. Clin Cancer Res. 2005;11:8341e8347.

76. Warnecke-Eberz U, Metzger R, Miyazono F, et al. High speci-ficity of quantitative excision repair cross-complementing 1messenger RNA expression for prediction of minor histopath-ological response to neoadjuvant radiochemotherapy inesophageal cancer. Clin Cancer Res. 2004;10:3794e3799.

77. Luthra R, Wu TT, Luthra MG, et al. Gene expression profiling oflocalized esophageal carcinomas: association with pathologicresponse to preoperative chemoradiation. J Clin Oncol. 2006;24:259e267.

78. Hamilton JP, Sato F, Greenwald BD, et al. Promoter methyl-ation and response to chemotherapy and radiation in esopha-geal cancer. Clin Gastroenterol Hepatol. 2006;4:701e708.

79. Bader FG, Lordick F, Fink U, et al. Paclitaxel in the neo-adjuvant treatment for adeno carcinoma of the distal esoph-agus (AEG I). A comparison of two phase II trials with long-termfollow-up. Onkologie. 2008;31:366e372.

80. Blank S, Lordick F, Dobritz M, et al. A reliable risk score forstage IV esophagogastric cancer. Eur J Surg Oncol. 2013;39:823e830.

81. Kang YK, Kang WK, Shin DB, et al. Capecitabine/cisplatinversus 5-fluorouracil/cisplatin as first-line therapy in patientswith advanced gastric cancer: a randomised phase III non-inferiority trial. Ann Oncol. 2009;20:666e673.

82. Ando N, Iizuka T, Ide H, et al. Surgery plus chemotherapycompared with surgery alone for localized squamous cell car-cinoma of the thoracic esophagus: a Japan Clinical OncologyGroup StudydJCOG9204. J Clin Oncol. 2003;21:4592e4596.

83. Malthaner RA, Wong RK, Rumble RB, et al. Neoadjuvant oradjuvant therapy for resectable esophageal cancer: a sys-tematic review and meta-analysis. BMC Med. 2004;2:35e51.

84. Kelsen DP, Winter KA, Gunderson LL, et al. Long-term results ofRTOG trial 8911 (USA Intergroup 113): a random assignmenttrial comparison of chemotherapy followed by surgerycompared with surgery alone for esophageal cancer. J ClinOncol. 2007;25:3719e3725.

85. Macdonald JS, Smalley SR, Benedetti J, et al. Chemo-radiotherapy after surgery compared with surgery alone foradenocarcinoma of the stomach or gastroesophageal junction.N Engl J Med. 2001;345:725e730.

86. Natsugoe S, Okumura H, Matsumoto M, et al. Randomizedcontrolled study on preoperative chemoradiotherapy followedby surgery versus surgery alone for esophageal squamous cellcancer in a single institution. Dis Esophagus. 2006;19:468e472.