Abstract
Background: This study aimed to identify prognostic factors for adenocarcinoma of the esophagogastric junction (AEG) in Siewert type II and characterize the population whose prognosis is expected to improve. Patients and Methods: We retrospectively reviewed a database of 68 AEG type II patients who had undergone surgical curative resection without preoperative treatment. Results: Although patients with pathological N0 (pN0) showed favorable 3-year disease-free survival (91%), patients with pN1-3 had poor outcomes (42%, 23%, and 10%). Multivariate analyses showed that the only independent prognostic factor was lymph node metastasis, and that tumor depth (cT3-4) and tumor size (≥4 cm) were correlated closely with lymph node metastasis. Conclusion: Surgical curative resection without preoperative treatment is insufficient to treat AEG type II with lymph node metastasis. Among AEG patients, those with the high-risk factors of preoperative tumor depth (T3-4) or/and size (≥4 cm) might need intensive multimodal treatment, including perioperative adjuvant chemotherapy.
Adenocarcinoma of the esophagogastric junction (AEG) is defined as a tumor with an epicenter within 5 cm of the esophagogastric junction (EGJ) per the eighth edition of the tumor-node-metastasis (TNM) classification of the Union for International Cancer Control (UICC) and has been classified into three subtypes by Siewert (1, 2). The worldwide incidence of AEG Siewert type II, which is defined as a tumor that invades the EGJ with the epicenter located between 1 cm above and 2 cm below the EGJ, has increased rapidly (3, 4). Because AEG type II is located along the border between the mediastinum and abdomen, the standard treatment strategy for AEG type II, including surgical procedure, perioperative adjuvant therapy, remains controversial (5-7). Surgery with lymph node dissection is the main therapeutic modality for AEG; however, with this modality alone, the 5-year survival rate (<30%) is poor (8, 9). These prognoses are the same or worse than those of gastric and esophageal cancers (10). Additionally, because outcomes in AEG type II patients treated with surgery alone remain unsatisfactory, multimodal treatment involving neoadjuvant therapy for locally advanced AEG type II is the current strategy in Western countries (11, 12). However, criteria for identifying AEG type II patients who are candidates for perioperative treatment must be established, and there are some uncertainties regarding the prognostic factors of AEG type II.
This study aimed to investigate the prognostic factors for AEG type II patients who undergo curative surgery without preoperative treatment and identify the population whose poor prognosis is expected to improve.
Patients and Methods
Patients. The clinicopathological and follow-up data of 68 patients who had undergone curative resection without preoperative treatment for AEG type II from April 2000 to December 2016 at Kobe University Hospital were retrospectively analyzed. AEG type II was diagnosed by performing biopsy by esophagogastroscopy with total-body computed tomography (CT) scanning. For evaluation of clinical lymph node metastasis, a size of 8 mm or more was considered positive with reference to Japan Clinical Oncology Group (JCOG) 1302A study (13). The inclusion criteria were gastric and esophageal resection for AEG type II, ≥D1 grade radical lymphadenectomy, and macroscopic curative resection. Patients who received chemotherapy or radiotherapy before surgery and those who underwent D0 grade lymphadenectomy or palliative resection were excluded. Histological and clinicopathological evaluations were performed in accordance with the TNM Classification of Carcinomas of the Esophagogastric Junction by UICC 8th (1). All patients were followed up by total-body CT and blood collection. This study was approved by the Ethics Committee of Kobe University (No. B200072). All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964 and later versions.
Surgical procedures. The standard surgical procedure was total gastrectomy and lower esophagectomy with lower mediastinal and abdominal lymphadenectomy via a transhiatal approach. A subtotal esophagectomy by transthoracic approach was included if the length of esophageal invasion was ≥3 cm. Proximal gastrectomy was performed in patients with T1/T2 tumors. Total gastrectomy was performed in patients with T3/T4 tumors.
Statistical analyses. The associations between clinicopathological factors were estimated using the χ2 test. Disease-free survival (DFS) and overall survival (OS) curves were constructed using the Kaplan-Meier method and compared using the log-rank test. Parameters that were significantly associated with DFS rates in univariate analyses using log-rank tests were further analyzed by multivariate analysis using a Cox proportional hazard regression model. The odds ratio represented the odds of a tumor occurring in patients with lymph node metastasis relative to that in those without lymph node metastasis. A p-value of <0.05 was considered statistically significant. Covariates found to be significant in the univariate analysis at p<0.05 were included in the multivariate model. Receiver operating characteristic (ROC) curve analysis for the optimal tumor size of interest for predicting lymph node metastasis was performed. Statistical analyses were performed using JMP ver. 14 software (SAS, Cary, NC, USA).
Results
Clinicopathological findings and surgical procedure. Clinical characteristics of the patients are shown in Table I. The median patient age was 69 years (interquartile range=61-75 years), and 57 (84%) were men. Fifty-eight (85%) patients had differentiated adenocarcinoma. The median tumor size was 3.8 cm (interquartile range=2.5-5.5), and 11 (16%) patients had a tumor invading the esophagus by >3 cm. More than half (59%) of the patients presented with pT1 or pT2 tumors. Pathological lymph node metastasis was observed in 32 (47%) patients. One patient had M1 status with liver metastasis. All patients underwent abdominal lymph node dissection and half of the patients underwent mediastinal lymph node dissection (34/64 cases). The patients underwent curative resection with hepatectomy. Fifty-four (79%) patients were treated via the transhiatal approach with lower esophagectomy. Thirty-two (47%) patients underwent total gastrectomy, and 36 (52%) patients underwent proximal gastrectomy. The median number of harvested lymph nodes was 27 (interquartile range=18-46). The median postoperative follow-up period was 4.0 years.
Patient characteristics and surgical procedure.
Recurrence rates and patterns based on N status. Patients with pathological N0 status showed favorable outcomes (3-year DFS rate of 91%, Figure 1A). However, patients with N1, N2, or N3 showed poor outcomes (3-year DFS rate of 42%, 23%, and 10%, respectively). The median DFS from time of surgery was significantly shorter in patients with N1, N2, or N3 status than in patients with N0 status (log-rank test, 2.7 years, 1.4 years, and 0.7 years vs. undefined, p<0.001). The 5-year OS rate was 83% in patients with pathological N0 status. Patients with N1-3 condition had poor OS rate (14%, 46%, and 10%, respectively, Figure 1B).
Survival analysis of all 68 patients with adenocarcinoma of the esophagogastric junction (AEG) in Siewert type II. (A) Disease-free survival and (B) overall survival for patients with AEG type II according to pathological N status.
Univariate analysis revealed that tumor depth (pT3-4), tumor size (≥4 cm), esophageal invasion, and lymph node metastasis were significantly associated with DFS (Table II). In the multivariate analysis, only lymph node metastasis remained a significant prognostic factor [hazard ratio (HR)=4.2; 95% confidence interval (CI)=1.68-10.6; p<0.01].
Univariate and multivariate Cox regression analyses for disease-free survival.
Recurrence in patients with pathological N0 status was observed in one (3%) case. The patient had recurrence at the anastomotic site, but not at other sites (Table III). Recurrence in patients with N1-3 status was observed in five (56%), nine (69%), and eight (80%) cases, respectively. Notably, two (40%) out of five relapsed patients with N1 status had multiple sites of recurrence at the time of the first recurrence, including lymphatic recurrence and hematogenous metastases and dissemination. This pattern of recurrence was similarly prevalent in patients with advanced N status [N2, 5/9 (56%); N3, 4/8(50%)].
Comparison of recurrence sites according to N status.
Preoperative predictors of lymph node metastasis. Table IV shows the cross table for clinical and pathological diagnosis of lymph node metastasis. Among 21 patients who were diagnosed with clinical lymph node metastasis, 18 had histologically positive nodes (positive predictive value 86%), whereas 33 of 47 patients who were clinically node negative did not have histological nodal metastasis (negative predictive value 70%). Preoperative prediction of lymph node metastasis using the CT criteria had high specificity (92%) but low sensitivity (56%).
Clinical and pathological diagnosis of lymph node metastasis.
ROC curve analysis showed that the optimal tumor cutoff size that significantly correlated with pathological lymph node metastasis was 4.1 cm (AUC, 0.796; sensitivity, 0.719; specificity, 0.824; Figure 2A). Tumor depth (T3-4) and tumor size (≥4 cm) were significantly associated with pathological lymph node metastasis in the multivariate analyses, but tumor location, histology, and esophageal invasion were not (p<0.001, Table V). The DFS curves of patients based on the tumor depth and size are presented in Figure 2B. The 3-year DFS rate was 20% in the cT3-4 plus tumor size ≥4 cm status.
Correlation between tumor size and overall survival of patients with adenocarcinoma of the esophagogastric junction (AEG) in Siewert type II. (A) ROC analysis of tumor size for predicting pathological lymphatic metastasis. The ROC curve demonstrates how the use of a tumor size cutoff of 4.1 cm helps differentiate N0 status from N1-N3. (B) Kaplan-Meier analysis of disease-free survival for patients with adenocarcinoma of the esophagogastric junction of Siewert type II according to tumor depth and tumor size.
Odds of tumor factors in patients with pathological lymph node metastasis.
Discussion
This study showed that lymph node metastasis was an independent prognostic factor for AEG type II patients, and the data from surgical curative resection without preoperative chemotherapy showed that AEG type II patients had significantly poorer survival if there was at least one lymph node metastasis than that of patients without. In addition, tumor depth more than T3 and tumor size ≥4 cm were strongly correlated with lymph node metastasis.
The results of multivariate analysis in this study suggested that lymph node metastasis was a strong prognostic factor in AEG type II patients. In this study, surgical resection with recent recommended lymphadenectomy for AEG type II was generally performed (7). Previous studies have shown that the number of metastatic lymph nodes (seven or more) could be used to predict the likelihood of systemic disease in patients with AEG or esophageal cancer (14, 15). In our study, the N stage evaluation for 3-year DFS also showed that even patients with N1 stage had a poor prognosis. Our results suggested that there was a high probability of recurrence when there was at least one lymph node metastasis. Postoperative adjuvant chemotherapy might be desirable for AEG type II with pathological lymph node metastasis and requires careful follow-up.
Lymph nodes near the esophagogastric junction, particularly those in the inferior mediastinum, are surrounded by anatomical structures in a narrow space (such as the diaphragm, aorta, and pericardium) and detecting lymph node metastases by contrast CT before surgery is more difficult than lymph nodes around the stomach or colon. Because the preoperative diagnosis rate of lymph node metastasis by CT was low, with a sensitivity of 62.5% and specificity of 65.7% even for gastric cancer, other ancillary factors are required to predict pathological lymph node metastasis for AEG type II (13). In addition, we examined the correlation between clinical and pathological lymph node metastasis. While the pathologically positive rate was high in clinically positive nodes (86%), the detection rate of clinically positive nodes was only 4% of all cases. This fact suggests that predicting metastasis based on lymph node size alone is difficult. We derived from the results of ROC curve analysis that tumor size (≥4.1 cm) could be a simple predictor of lymph node metastasis. We believe that the pathological tumor size is generally smaller than the clinical size considering the effects of formalin fixation, so a preoperative expected tumor size (≥4 cm) would not overestimate the size of a tumor used as a predictor of lymph node metastasis (16). Tumor size (≥4 cm) can be calculated by esophagogastroduodenoscopy, CT, and upper gastrointestinal series performed in the routine preoperative examination for AEG and would be the convenient indicator for predicting pathological lymph node metastasis. Recent reports have described the prediction of lymph node metastasis using a combination of multiple clinicopathological factors. Although calculating tumor volume for estimating lymph node metastasis from CT images has been reported, the measurement method is considered time consuming and cumbersome (17, 18). Zhu et al. also showed a nomogram for predicting the risk of lymph node metastasis in AEG combining tumor size, differentiation, and depth of invasion (19). However, their study included only T1 tumors and was inadequate for predicting the progression of AEG. Therefore, tumor depth and tumor size >4 cm are essential predictors of simple lymph node metastasis as well as prognostic factors.
Several studies have reported the recurrence patterns in AEG type II patients (20, 21). The incidence of postoperative mediastinal and aortic lymph node metastases is known to be relatively high in patients with AEG type II. Wayman et al. showed that the most frequent type of recurrence was hematogenous (54% in AEG type II patients), of which 56% were detected within 1 year after surgery, indicating that hematogenous metastases are also an essential form of recurrence (21). A recent retrospective study also showed that lung recurrence (11%) is more frequent than para-aortic lymph node metastases (9.8%) (22). These findings are consistent with our results, indicating that recurrence after radical resection of AEG occurred at various and multiple sites. Because even patients with N1 disease tended to experience various recurrence patterns, it is crucial to understand the oncogenic nature of AEG to plan surgical therapeutic strategies.
The limitations of the present study include its retrospective design and nonrandomized nature. Moreover, because it was performed in a single center, the number of patients examined was small. Not all patients could avail contrast-enhanced CT for the diagnosis of clinical lymph node metastasis.
In summary, lymph node metastasis was a poor prognostic independent factor for AEG type II, regardless of the number of metastases. Additionally, tumor depth more than T3 and tumor size ≥4 cm were strongly correlated with lymph node metastasis. Considering the difficulty of preoperative diagnosis for lymph node metastasis in AEG type II patients, preoperative tumor depth (T3-4) or tumor size (≥4 cm) might need intensive multimodal treatment, including preoperative chemotherapy.
Acknowledgements
The Authors would like to thank Enago (www.enago.jp) for the English language review.
Footnotes
Authors’ Contributions
Conception and design: Yoshihiro Kakeji, Shingo Kanaji, Satoshi Suzuki, Naoki Urakawa; Acquisition of data: Naoki Urakawa, Hitoshi Harada, Ryuichiro Sawada, Hironobu Goto, Hiroshi Hasegawa, Kimihiro Yamashita, Takeru Matsuda, Taro Oshikiri; Analysis and interpretation of data: Naoki Urakawa, Shingo Kanaji, Satoshi Suzuki; Writing, review, and revision of manuscript: Naoki Urakawa, Shingo Kanaji, Satoshi Suzuki, Yoshihiro Kakeji.
Conflicts of Interest
The Authors have no conflicts of interest associated with this manuscript.
- Received November 18, 2021.
- Revision received January 8, 2022.
- Accepted January 10, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.