The incidence of adenocarcinoma of esophagogastric junction (AEG) has increased globally (1,2). Although surgery is the mainstay of treatment for locally advanced AEG, prognosis remains poor even after complete resection (3). Therefore, multimodality approaches, such as preoperative (neoadjuvant) and postoperative (adjuvant) therapies, including chemotherapy or chemoradiation, have been developed to improve survival (4-8). Combined modality therapy is currently incorporated in treatment guidelines (9,10), but the recommended strategies differ across countries and even centers, and the best approach has not been clearly established. In Japan, the standard treatment for resectable AEG, based on the results of clinical trials for stage II/III gastric cancer, is a primary resection followed by adjuvant chemotherapy (11-13); however, the therapeutic outcomes are unsatisfactory. A more intensive strategy to improve prognosis is required for this disease.

Neoadjuvant chemotherapy (NAC) represents an advantage over adjuvant chemotherapy because it is administered before surgery, whereas the intensive adjuvant setting is limited by the poor compliance of post-gastrectomy patients to chemotherapy during the postoperative recovery phase (14,15). Some clinical trials and experimental studies have demonstrated the survival benefits of preoperative chemotherapy for AEG (16-18). Infusional 5-FU-based regimens such as cisplatin and fluorouracil (CS), epirubicin and cisplatin plus either fluorouracil or capecitabine (ECF/EOX), and fluorouracil, leucovorin, oxaliplatin, and docetaxel (FLOT) are conventionally used as preoperative and perioperative chemotherapy for AEG in Western countries (4-6), while oral fluoropyrimidine such as S-1 or capecitabine-based regimens for unresectable or recurrent gastric adenocarcinoma are generally adopted for neoadjuvant settings in Japan (18-20). However, few studies have investigated the efficacy of NAC using an oral fluoropyrimidine regimen for resectable AEG.

In this retrospective study, we aimed to compare the therapeutic outcomes of patients with locally advanced AEG between patients who had surgery first and patients who had surgery after NAC using an oral fluoropyrimidine-platinum regimen and to evaluate the potential benefit of NAC on prognosis.

Patients. A total of 35 patients with locally advanced AGE underwent surgery with curative intent at the Kobe University Hospital from January 2007 to December 2018. Of these, 21 patients underwent surgery first (surgery first group) and 14 patients received NAC before surgery (NAC group). We analyzed the patients’ clinicopathological and treatment data extracted from medical records. The clinical stage was assessed by esophagogastroduodenoscopy (EGD) and a thoracoabdominal computed tomography (CT) scan and was diagnosed according to the Japanese classification of gastric carcinoma (21). The eligibility criteria for NAC were as follows: cT2 or deeper tumor or cN+ stage. During the NAC sessions, six patients received the SOX regimen, two received the CS regimen, and six received the XP plus trastuzumab regimen. The SOX regimen consisted of three cycles of oral S-1 (40 mg/m² twice daily on days 1-14, followed by a one-week rest period) and an intravenous infusion of oxaliplatin (130 mg/m² on day 1). The CS regimen consisted of two courses of oral S-1 (40 mg/m² twice daily on days 1-21, followed by a one-week rest period) and an intravenous infusion of cisplatin (60 mg/m² on day 8). For marginally resectable HER2-positive tumors with invasion to adjacent organs or metastases to extensive lymph nodes, XP plus trastuzumab was administered via two cycles of oral capecitabine (2,000 mg/m² twice daily on days 1-14) and intravenous cisplatin (60 mg/m² on day 1) plus trastuzumab (8 mg/kg during the first cycle and 6 mg/kg during the second cycle on day 1). For the patients who underwent NAC, EGD and a CT scan were repeated once each cycle of chemotherapy to evaluate the clinical response. The pathological response was evaluated based on the Japanese classification of gastric carcinoma: Grade 0, no evidence of effect; Grade 1a, viable tumor cells occupy more than 2/3 of the tumorous area; Grade 1b, viable tumor cells remain in more than 1/3 but less than 2/3 of the tumorous area; Grade 2, viable tumor cells remain in less than 1/3 of the tumorous area; Grade 3, no viable tumor cells remain (21). Adverse events associated with chemotherapy were accessed by the Common Terminology Criteria for Adverse Events (version 5.0).

Siewert type I tumors were resected by transthoracic esophagectomy with two-field lymph node dissection and gastric conduit reconstruction. Siewert type II/III tumors were resected by a transhiatal abdominothoracic lower esophagectomy and total or proximal gastrectomy with intrathoracic reconstruction. Postoperative morbidities higher than grade II on the Clavien-Dindo classification were defined as morbidities (22). Morbidities higher than grade IIIa were defined as severe. Adjuvant chemotherapy was adopted for pathological stage II/III disease in both groups. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Declaration of Helsinki. This study was approved by the Institutional Review Board of the Graduate School of Medicine, Kobe University (approval number: B200093).

Statistical analyses. Statistical comparisons between the two groups were performed with the Mann-Whitney U-test for continuous variables and with the Chi-square test and Fisher’s exact test for categorical variables, as appropriate. Overall survival (OS) was defined as the period from the date of surgery to death from any cause or the last follow-up date. Relapse-free survival (RFS) was defined as the period from the date of surgery to recurrence or death. The cumulative survival rate was estimated using the Kaplan-Meier method, and were compared using the log-rank test. p-Values <0.05 were considered statistically significant. All statistical analyses were performed using JMP statistical software, ver. 14 (SAS, Cary, NC, USA).

Clinicopathological characteristics did not differ significantly between the two groups in terms of gender, American Society of Anesthesiologists Physical Status (ASA-PS), and tumor location, although the surgery first group tended to have a higher median age (p=0.071, Table I). The NAC group had a more advanced cT category than the primary surgery group (p=0.007). Likewise, the cN category tended to be more advanced in the NAC group (p=0.100). Table II shows the adverse events of NAC in the 14 patients. Grade 3 or 4 adverse events recorded during NAC included leukopenia (14%), neutropenia (14%), thrombocytopenia (14%), and febrile neutropenia (7%) in four total patients; no chemotherapy-related deaths occurred. A total of 71% of patients completed the planned courses of NAC. The therapeutic outcomes are presented in Table III. There was no significant difference in the operative procedure. R0 resection was achieved in all 35 patients. The (y)pT and (y)pStage categories in the NAC group were significantly less advanced than those in the surgery first group (p=0.014 and p=0.019, respectively). A clinical downstage effect of NAC was observed in seven patients (50%), and no patients exhibited disease progression. The pathological response of the primary tumor was distributed as follows: grade 1a in seven patients, grade 1b in one patient, grade 2 in two patients, and grade 3 in four patients; thus, the response rate was 50% (7 of 14 patients). Of the six patients who received additional trastuzumab, three (50%) downstaging and two (33%) pathological grade 3 responses were obtained. Adjuvant chemotherapy was continued for six patients in each group. The NAC group had a significantly lower rate of recurrence than the surgery first group (21% vs. 62%, p=0.019). Postoperative morbidities are summarized in Table IV. Grade II or higher morbidities were observed in 10 patients (48%) in surgery first group and 8 patients (57%) in NAC group (p=0.58). Grade III or higher morbidities were observed in 8 patients (38%) in surgery first group and 3 patients (21%) in NAC group (p=0.25). The median follow-up period was 39.5 months. NAC was significantly associated with superior three-year relapse-free survival (RFS) compared to surgery first [78% (95% confidence interval (CI)=0.39-0.94) vs. 22% (95% CI=0.03-0.40), p=0.011, Figure 1a]. The overall survival (OS) was significantly longer in the NAC group than in the surgery first group [median survival, not reached (95% CI=0.45-1.00) vs. 29 months (95% CI=0.21-0.68), p=0.032, Figure 1b].

The present study demonstrated that the OS and RFS were significantly superior in patients who received NAC with an oral fluoropyrimidine-platinum regimen compared to those who underwent primary resection for locally advanced AEG.

A global standard regimen for AEG consists of a fluoropyrimidine plus a platinum compound. In addition, the ToGA trial demonstrated that the addition of trastuzumab to a combination of cisplatin plus capecitabine or infusional 5-FU improved both the response rate and the OS for HER 2-positive tumors (23). The 5-FU continuous intravenous infusion regimen for AEG is commonly administered in Western countries. Differing chemotherapy regimens might also contribute to divergent results (24). In a recent network meta-analysis of clinical trials for AEG, patients who received three fluoropyrimidines such as 5-FU, capecitabine, and S-1 had similar overall and progression-free survivals (25). In the present study, which included six patients (43%) who were administered trastuzumab, chemotherapy using oral fluoropyrimidine S-1 or capecitabine had a 50% histological response rate of grade 1b or higher and achieved a pCR rate of 28.6%, which was rather higher than the previous reports of 0-17.2% in 5-FU-based regimens (26). Oral fluoropyrimidines are generally more convenient than infusional 5-FU in clinical practice. In particular, S-1-based regimens had a lower incidence of adverse events, such as less febrile neutropenia and toxicity in Western patients, compared to regimens that included 5-FU (25). Grade 3 or higher leukopenia (21% and 27%), neutropenia (39% and 51%) and thrombopenia (3% and 2%) were recorded for the ECF/ECX and FLOT arms in the FLOT-AOI trial, respectively (6). In this study cohort, grade 3 or higher leukopenia, neutropenia, and thrombopenia each occurred in 14% of patients, and no severe non-hematological toxicity was observed. Thus, an oral fluoropyrimidine-platinum regimen seems to have a manageable and acceptable safety profile for clinical use in NAC settings.

Based on previous results including the MAGIC trial and the FLOT4-AIO trial in Western countries, perioperative chemotherapy is now a standard treatment for patients with resectable AEG (4,6). Although perioperative chemotherapy may be more intensive compared with preoperative chemotherapy theoretically, it is unclear whether improved survival results from NAC or adjuvant chemotherapy. In the MAGIC trial, only 55% of patients in the perioperative chemotherapy group were administered postoperative chemotherapy. This may imply that the preoperative portion mainly influenced the efficacy of the perioperative treatment. Similarly, postoperative chemotherapy was administered to 55-60% of the patients in the FLOT-AIO trial, implicating a larger role of preoperative chemotherapy. Chemoradiotherapy is reported to be superior to chemotherapy in terms of the local control rate in Western countries (7). However, the local recurrence rate was very low in AEG in Japan and Korea, likely because of D2 lymph node dissection and ensuring resection margins by intraoperative frozen sections (27,28). The ARIST trial, which was conducted in Korea, failed to prove the additional effect of postoperative radiotherapy to adjuvant chemotherapy for resected gastric cancer. Furthermore, the survival benefit of additional radiotherapy to chemotherapy remains a matter of debate (26).

Accurately diagnosing the AEG stage is crucial for decision-making for therapeutic strategies, but the diagnostic accuracy of the T and N categories by endoscopy and imaging is limited (29). The assessment of nodal metastasis is particularly difficult, and the clinical stage of AEG tends to be underdiagnosed, as has been previously reported (30). For example, the negative predictive value (pathological node negative/clinical node negative) of gastric cancer was reported to be only 47.8% (31). We could accurately diagnose only 44.4% (4/9) of nodes as negative in the surgery first group. Lymph node metastasis is a strong predictor of survival in AEG (2). In light of these findings, we adopt NAC for node-positive tumors or cT2 or deeper tumors that have a high possibility of nodal metastasis. This aggressive implementation of NAC is one plausible explanation why the pCR rate was relatively high and the therapeutic outcomes of patients who received NAC were favorable in this study cohort.

Our study has several important limitations. This study was based on retrospective data collected at a single center. The NAC regimen was not standardized, few patients were included, and the median follow-up period was short. Although some limitations were present, oral fluoropyrimidine-platinum regimen had low toxicity and convenience for AEG patients. A larger and well-designed prospective clinical trial is warranted to confirm our results.

In conclusion, NAC using an oral fluoropyrimidine-platinum regimen potentially provides survival benefits over a surgery-first approach in patients with locally advanced AEG. This approach for AEG should be considered from the perspective of therapeutic efficacy and clinical convenience.

The Authors declare no conflicts of interest.

Study conception and design: SS and YK. Acquisition of data: GT, HH, YM, KY, and TM. Analysis and interpretation of data: SK, NU, TO and TN. Drafting of manuscript: SS and YK. Final approval of the version to be published: YK.