Abstract
Background/Aim: To determine the clinical significance of pre-treatment circumferential tumor location within the esophageal wall in patients with locally advanced esophageal squamous cell carcinoma who underwent curative resection after neoadjuvant chemotherapy. Patients and Methods: Patients (n=96) with cStage IB-III esophageal squamous cell carcinoma who underwent curative resection after neoadjuvant chemotherapy were categorized into two groups based on the circumferential tumor location within the esophageal wall on diagnostic endoscopy: left or anterior wall group (L/A patients, n=49); right or posterior wall group (R/P patients, n=47). Analyses were conducted to examine the relationship between circumferential tumor location, clinicopathological characteristics, 3-year overall survival (3Y-OS), and 3-year recurrence-free survival (3Y-RFS). Results: The lymph node status and recurrence rates of mediastinal lymph node metastases were significantly higher in patients with L/A than in patients with R/P. Furthermore, patients with L/A had significantly poorer 3Y-OS and 3Y-RFS than those with R/P. Tumor location within the esophageal wall was identified in multivariate analysis as an independent risk factor for 3Y-RFS (hazard ratio=2.92, 95% confidence interval=1.35-6.32, p=0.0064). Conclusion: Pre-treatment of circumferential tumor located within the esophageal wall may be a useful prognostic factor in patients with cStageIB-III esophageal squamous cell carcinoma who underwent curative resection after neoadjuvant chemotherapy.
Esophageal cancer (EC) is the eighth most common carcinoma worldwide, with 604,100 new cases in 2020, and the sixth leading cause of cancer-related deaths, with 544,076 deaths in 2020 (1). Curative resection after neoadjuvant chemotherapy (NAC) or neoadjuvant chemoradiation is the standard treatment for locally advanced resectable EC (2-9). However, the prognosis of patients remains unsatisfactory (10). Patients with locally advanced EC had a 5-year relative survival rate of 47% between 2010 and 2016 in the United States and 52.1% in 2013 in Japan (3). Treatment decisions are based on the disease stage, patient performance status, and preferences. However, patients with the same disease stage treated with the same chemotherapy regimen followed by curative surgery can have different prognoses. Therefore, useful prognostic biomarkers are very important, and several have been reported (11-15). Based on prognostic biomarkers, individualized treatments such as switching to a different NAC regimen, extended lymph node dissection surgery, and enhanced postoperative follow-up are expected to improve patient survival.
The prognostic significance of circumferential tumor location within the esophageal wall in patients with esophageal squamous cell cancer (ESCC) treated with only surgery has only been reported in a study by Mine et al. (16). Their study showed that in contrast to patients with tumors located on the left or anterior part of the esophageal wall (L/A patients), those with tumors located on the right or posterior part (R/P patients) had a lower pathological lymph node stage (pN) but shorter disease-free survival (DFS). Furthermore, circumferential tumor location within the esophageal wall proved to be an independent prognostic factor in multivariate analysis of patients with pT1b or pT2 ESCC.
To the best of our knowledge, no study has investigated the relationship between circumferential tumor location within the esophageal wall and survival in patients with locally advanced ESCC treated with NAC followed by curative resection, and the results of the study by Mine et al. (16) have not been validated. This study investigated the relationship between tumor location within the esophageal wall before treatment and clinicopathological features or survival in patients with locally advanced ESCC treated with NAC followed by curative resection as the standard of care.
Patients and Methods
Patients. This study was approved by the Institutional Review Board of Kanagawa Cancer Center (KCC), Yokohama, Japan (2018 Epidemiology-174). Patients were retrospectively selected from the medical database of the Department of Gastrointestinal Surgery, KCC, Yokohama, Japan, according to the following criteria: 1) histologically proven ESCC according to the Japanese classification of Esophageal Cancer, 11th Edition (17), 2) clinical stage IB, II, or III diseases according to the 8th edition of the UICC TNM Classification of Esophageal Cancer (18) and 3) treatment with curative resection (R0 resection), including lymph node dissection in three areas (neck, chest, and abdomen) after two courses of 5-Fluorouracil (5-FU)+cisplatin (CDDP) as NAC at KCC between June 2010 and December 2015. Patients diagnosed with a tumor occupying the entire circumference on pre-treatment endoscopy were excluded.
Definition of circumferential ESCC location within the esophageal wall by endoscopy before NAC. The location of the tumor and percentage of the circumferential tumor size in esophageal wall were determined by analyzing pictures taken during esophagogastroduodenoscopy (EGD) before NAC as follows: the location where the left main bronchus indented the esophagus was defined as the “anterior” wall. The center of the tumor was used to determine its location (Figure 1). Tumor diameter was routinely measured during endoscopy before NAC. Patients were classified into two groups based on the circumferential tumor location in the esophageal wall: L/A patients and R/P patients.
Definition of circumferential tumor location in the esophageal wall by endoscopy before neoadjuvant chemotherapy. The anterior esophageal wall was defined as the area where the left main bronchus was pushed into the esophageal lumen from outside the wall. If the tumor was ulcerated, the location of the deepest part of the ulcer was defined as the tumor center; otherwise, the center of the esophageal carcinoma was defined as the tumor center. The circumferential distribution of the tumor within the esophageal wall was divided into two groups: anterior or left wall (L/A patients) and posterior or right wall (R/P patients).
NAC. Patients with clinical TNM stage IB, II, or III ESCC received NAC with 5-FU and CDDP before surgery according to a standard protocol (19). NAC was administered in two cycles every 3 weeks. 5-FU was administered at a dose of 800 mg/m2 by continuous infusion from days 1 to 5, and CDDP was administered at a dose of 80 mg/m2 by intravenous drip infusion for 2 h on day 1.
Surgery. Surgery was performed 4 to 6 weeks after the final NAC. All patients underwent open thoracic subtotal esophagectomy in three fields (cervical, thoracic, and abdominal) and D2 lymph node dissection following the Japanese Esophageal Cancer Treatment Guidelines (20).
Pathological diagnosis after NAC+surgery, postoperative complications, and survival. The pathological depth of invasion of the primary tumor (ypT), lymph node status (ypN), and pathological stage were determined according to the UICC TNM 8th edition for esophageal cancer and the esophago-gastric junction, respectively (21). In addition, pathological primary tumor regression was evaluated according to the Japanese classification of Esophageal Cancer, 11th Edition, as follows: Grade 0, no recognizable cytological or histological therapeutic effect; Grade 1: viable cancer cells account for 1/3 or more of the tumor tissue, with some evidence of cancer tissue or cell degeneration; Grade 2: viable cancer cells account for less than 1/3 of the tumor tissue with evidence of severely degenerated or necrotic cancer cells; Grade 3: no viable cancer cells. Postoperative surgical complications were defined as Clavien-Dindo classification grade ≥2, occurring within 30 days after surgery (22).
Overall survival (OS) was defined as the time from the first day of NAC to the date of the patient’s death from any cause or date of the last follow-up. Relapse-free survival (RFS) was defined as the time from the first day of NAC to the date of tumor recurrence or the last follow-up. After discharge, all patients underwent physical examinations, blood tests, chest abdominal computed tomography every 3 months for at least 3 years, and annual upper gastrointestinal endoscopy for 5 years.
Statistical analyses. Fisher’s exact test was used to compare the patients in the L/A and R/P groups. The 3-year OS and RFS rates were calculated using the Kaplan-Meier method, and a log-rank test was performed to compare the groups. Univariate and multivariate Cox proportional hazards models were used to analyze the hazard ratios (HRs) for OS and RFS. Variables with p-values <0.05 in univariate analysis were included in the multivariate analysis. For all tests, a two-sided p-value <0.05 was considered significant. All statistical analyses were performed using the SPSS Statistics software program, ver. 25 (IBM Corp., Armonk, NY, USA).
Results
Patient cohort. A total of 225 patients with EC underwent esophagectomy between June 2010 and December 2015 at the KCC Hospital, in Yokohama, Japan. Among these patients, 129 were excluded from the current study due to the following reasons: 89 patients were diagnosed with clinical stage IA, stage IV, or did not receive NAC; 12 were diagnosed with histologically special type tumors (e.g., not ESCC); six patients had a non-curative (R1 or R2) resection; and 22 were diagnosed with tumors occupying the whole circumference at pre-treatment endoscopy. Eventually, 96 patients were included in the final analyses (Figure 2).
Patient flow diagram. A total of 225 patients with esophageal cancer underwent surgery between June 2010 and December 2015 at Kanagawa Cancer Center Hospital. Among these patients, 89 diagnosed with clinical stage IA or IV or did not receive neoadjuvant chemotherapy, 12 with histological types other than squamous cell cancer, six underwent non-curative resection (R1 or R2 resection), and 22 with whole circumferential tumor diagnosed at pre-treatment endoscopy were excluded. Eventually, 96 patients were included in this study.
Association between circumferential tumor location within the esophageal wall and clinicopathological factors. Table I shows the patient characteristics and association between circumferential tumor location within the esophageal wall and clinicopathological features. A marked difference was observed between the L/A and R/P groups regarding the ypN.
Patient characteristics and association between circumferential tumor location within the esophageal wall and clinicopathological features.
Survival analyses. The Kaplan-Meier curves for OS are shown in Figure 3A. The 3-year OS rate after surgery was significantly lower in L/A patients compared to R/P patients (74.8% and 91.2%, respectively; p=0.0071 using the log-rank test). On univariate analyses, pathological tumor regression grade, ypT, percentage of circumference occupied by the tumor of the esophageal wall, tumor diameter, and tumor location were considerably associated with OS. On multivariate analysis, circumferential tumor location was identified as an independent predictive factor for OS [HR=3.05, 95% confidence interval (CI)=1.10-8.39, p=0.03; Table II].
Kaplan-Meier curves for overall survival (OS) and relapse-free survival (RFS). (A) The 3-year OS after the time of first NAC in patients with left or anterior esophageal wall located tumor (L/A patients) is poorer than that in patients with right or posterior esophageal wall located tumor (R/P patients) (74.8% and 88.8%, respectively, HR=3.6, 95%CI=1.32-9.77, p=0.0071). (B) The 3-year RFS rate after surgery in the L/A patients are significantly poorer than that in the R/P patients (56.4% and 82.6%, respectively, HR=2.48, 95%CI=1.21-5.08, p=0.0096).
Results of univariate and multivariate Cox proportional hazards analyses for overall survival (OS).
Kaplan-Meier curves for RFS are shown in Figure 3B. The 3-year RFS rate after surgery was significantly lower in L/A patients than in R/P patients (56.4% and 82.6%, respectively; p=0.0053). On univariate analyses for RFS, circumferential tumor location within the esophageal wall, pathological tumor regression grade, ypT factor, ypN factor, percentage of diameter of the esophageal wall occupied by the tumor, and tumor diameter were markedly associated with RFS. On multivariate analysis, circumferential tumor location within the esophageal wall and ypT and ypN were identified as independent predictive factors for RFS (HR=2.92, 95%CI=1.35-6.32, p=0.0064; Table III).
Univariate and multivariate Cox proportional hazards analyses for relapse-free survival (RFS).
Sites of first recurrence after curative esophagectomy after NAC. Overall, the most common site of recurrence in both groups was the mediastinal lymph nodes. Recurrence in the mediastinal lymph nodes was notably more common in the L/A group than in the R/P group (Table IV). In addition, a detailed examination of the mediastinal lymph node metastases revealed that the No. 106recL lymph node had the highest number of metastases, all of which were recurrences in the L/A patients (Table V).
Comparison of the initial site of recurrence between left and anterior wall group (L/A) and right and posterior wall group (R/P) patients.
Details of metastatic lymph nodes in left and anterior wall group (L/A) (n=14) and right and posterior wall group (R/P) (n=5) patients with mediastinal lymph node recurrence.
Discussion
This study investigated the association between pre-treatment circumferential tumor location within the esophageal wall, clinicopathological characteristics, and survival in patients with cTNM stage IB-III ESCC who underwent curative resection after standard NAC. The results showed that L/A patients had a higher frequency of lymph node metastasis than R/P patients. In addition, the OS and RFS of patients with L/A were considerably shorter than those with R/P. Multivariate analysis showed that circumferential tumor location within the esophageal wall could predict OS and RFS, independent of known prognostic factors.
We considered two potential reasons for our study’s differences in survival between L/A and R/P patients. First, there is a difference in the perineural invasion of esophageal cancer cells between L/A and R/P patients. Although this study did not determine the relationship between pre-treatment circumferential tumor location within the esophageal wall and perineural invasion of esophageal cancer, in a previous study Doubilet et al. examined 32 cadavers and reported more nerve fiber contact in the anterior wall of the esophagus than in the posterior wall (23). Furthermore, perineural invasion of cancer cells has been reported to predict OS and DFS (24-26). However, it should be noted that the present study could not demonstrate a relationship between the circumferential nature of esophageal cancer and perineural invasion. Second, there is a difference in the difficulty of lymph node dissection in the right and left mediastinum in radical esophageal cancer surgery. For example, dissection of the left recurrent nerve lymph node (No. 106recL) and the left tracheobronchial lymph node (No. 106tbL) is more difficult than the dissection of other lymph nodes, as these lymph nodes are located in very narrow areas adjacent to the trachea, esophagus, and right aortic arch. Furthermore, the left recurrent laryngeal nerve, which must be preserved, runs through the fatty tissue to be dissected. In our study, L/A patients had considerably more mediastinal lymph node recurrences than R/P patients. The mediastinal lymph nodes with the highest recurrences were No. 106recL and No. 106tbL, all of which exclusively occurred in L/A patients. In other words, we hypothesized that the difficulty in completely dissecting these lymph nodes might lead to residual micrometastases in L/A patients, which may be related to recurrence and decreased survival.
To date, the only report examining the relationship between the circumferential distribution of tumors within the esophageal wall and clinicopathologic characteristics and survival was the study by Mine et al. in patients with pTNM stage IB-II ESCC who underwent only surgical resection without NAC. Their study also reported a higher frequency of pathologic lymph node metastases in L/A patients and tumor circumference distribution as an independent prognostic factor (p=0.003) in multivariate analysis (16). However, the relationship between tumor circumference distribution and survival contradicted our findings. This difference could be attributed to differences in the treatment strategies [surgery alone (16) versus NAC+surgery (current study)], the stage of the disease within the two cohorts, or differences in lymph node dissection between the studies. In addition, the current study included patients with stage IB-III disease, whereas Mine et al. included only patients with stage IB-II disease. This could explain the increased frequency of higher ypN stages in the L/A patient group in the current study.
Interestingly, in the study by Mine et al. (16), in which NAC was not performed, R/P patients had more frequent hematogenous metastatic recurrences than L/A patients. It was reported that this might be related to differences in esophageal lymphatic drainage and that cancer cells from tumors located in the R/P are more likely to enter the systemic circulation via the thoracic duct (27). In contrast, in the present study, in which patients were treated with NAC before surgery, there was no significant difference in hematogenous recurrence between the L/A and R/P patients. We hypothesize that this might be because NAC controls hematogenous micrometastases, so differences in lymphatic drainage within the esophageal wall may become irrelevant concerning hematogenous tumor spread.
This study has a few limitations. First, this was a retrospective, single-center cohort study, and the number of patients enrolled was relatively small (n=96). The results of the current study need to be validated in large multicenter prospective studies using, ideally, a standardized endoscopic assessment, surgical approach, and workup of the resection specimen in the pathology department. Second, circumferential tumor location was classified by two independent observers. If the tumor appeared to be located between 1 and 2 o’clock or between 7 and 8 o’clock, the classification proposed by Kuge et al. (4) was followed. Still, in some cases, the two observers disagreed, and a third observer had to make a final decision. In some cases, the site of indentation of the esophageal wall by the left bronchus, which was used as a landmark on the anterior wall, was unclear in the still images. In these cases, the decision was made based on several other images.
In conclusion, our study demonstrated that pre-treatment circumferential tumor location within the esophageal wall may be a useful prognostic factor in patients with cStage IB-III ESCC who underwent curative resection after NAC. If these results are validated in large multicenter prospective studies, outcomes of patients with locally advanced ESCC may improve if individualized treatments, such as intensive neoadjuvant or perioperative chemotherapy, surgery with extended lymphadenectomy, and close postoperative follow-up are performed for L/A patients.
Acknowledgements
This work was supported by JSPS KAKENHI Grant Number 20K08997. The Authors would like to thank the patients, families, and facility staff who participated in this study.
Footnotes
Authors’ Contributions
SN and HG designed the study. Data collection and literature search were performed by SN and TO. Data analysis and interpretation were performed by SN, and TO. Data interpretation was performed by all researchers. The paper and figures were drafted by SN, TO, and HG. Finally, the paper was revised and approved by all researchers. Thus, all Authors actively participated in this study.
Conflicts of Interest
The Authors have no actual or potential conflicts of interest to declare in relation to this study.
- Received March 28, 2023.
- Revision received April 6, 2023.
- Accepted April 7, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.









