The association between adherence to elective nodal volume guidelines and reduced distant metastasis in esophageal and gastroesophageal junction cancer: a retrospective analysis of elective nodal irradiation

Abstract

Background

In 2015, expert guidelines on esophageal/gastroesophageal junction (GEJ) cancer contouring for intensity-modulated radiation therapy (IMRT) were published in International Journal of Radiation Oncology, Biology and Physics (IJROBP) delineating recommended elective nodal basins (celiac, para-aortic, gastrohepatic ligament, supraclavicular) to be irradiated depending on the primary tumor location. The aim of this study seeks to determine if chemoradiotherapy (CRT) that is compliant with the recommendations from these guidelines affects disease free survival and distant failure rates. We hypothesize that incomplete coverage of these areas increases the risk of developing distant failures.

Methods

Patients treated for non-metastatic esophageal or GEJ cancer with CRT pre-operatively or definitively from 2012 to 2021 were retrospectively identified from a single institution database. Radiation plans of eligible patients were then analyzed by tumor location. Plans were deemed guideline-compliant if radiation dose coverage, greater than 41.4 Gy, encompassed nodal basins recommended by the 2015 guidelines. The primary endpoint of this study was the overall rate of distant metastatic disease. Summary and descriptive statistics were used to characterize key cohort features.

Results

With a median follow-up of 22.3 months, 38 patients, with a median age at diagnosis of 66 years, were included in the study. Most patients were male (94.7%) with cT3 (52.6%), cN0 (44.7%), moderately differentiated (44.7%) and poorly differentiated (44.7%) adenocarcinoma (73.7%) located at the GEJ (57.9%). The median radiation dose used was 50.4 Gy, with most patients receiving concurrent carboplatin and paclitaxel (86.8%). Four patients received induction chemotherapy and 20 (52.6%) underwent esophagectomy. When examining guideline compliance, 17 (44.7%) radiation plans demonstrated adequate elective nodal irradiation (ENI). The most common improperly covered nodal basin was para-aortic (66.7%), followed by gastrohepatic (23.8%). One patient with sufficient ENI coverage (1/17) developed distant failure compared to 38.1% (8/21) with insufficient coverage (P=0.02). There were inappreciable differences in locoregional or local failure rates between those with and without complete ENI.

Conclusions

These results suggest that proper coverage of nodal basins could improve distant metastasis. ENI analysis of previous prospective CRT studies for esophageal cancer could validate these findings.

Highlight box.

Key findings

• This retrospective single-institution study evaluated adherence to 2015 expert contouring guidelines for elective nodal irradiation (ENI) in patients with non-metastatic esophageal and gastroesophageal junction (GEJ) cancers treated with chemoradiotherapy. Less than half of radiation plans were guideline-compliant. Patients who received adequate ENI had a significantly lower rate of distant metastatic failure compared with those who had incomplete nodal coverage (5.9% vs. 38.1%, P=0.02). Locoregional and local control rates were similar regardless of ENI compliance, suggesting the observed benefit was specific to distant disease control.

What is known and what is new?

• This study provides early clinical evidence linking guideline-based elective nodal coverage to improved distant metastasis outcomes in esophageal and GEJ cancers. While prior guidelines were largely consensus-based and have already provided a framework for adequate nodal basin irradiation, this analysis supports their clinical relevance and highlights the oncologic consequences of incomplete nodal irradiation. It addresses a critical knowledge gap regarding how radiation field design influences systemic disease patterns.

What is the implication, and what should change now?

• These findings underscore the importance of meticulous radiation contouring and adherence to established nodal irradiation guidelines in esophageal cancer treatment. They suggest that inadequate ENI may contribute to distant failure, potentially by allowing occult nodal disease to persist. The study supports re-evaluation of radiation plans in both clinical practice and prior prospective trials and provides a rationale for future prospective validation. If confirmed, strict ENI compliance could represent a modifiable treatment factor to reduce distant metastasis and improve outcomes in this high-risk population.

Introduction

As the 8^th most common cancer worldwide, 22,370 new cases of esophageal cancer arise every year in the United States (1). For those with locally advanced esophageal or gastroesophageal junction (GEJ) cancer, pre-operative or definitive chemoradiotherapy (CRT) remains part of the treatment paradigm. Most relapses occur due to distant metastatic spread (2,3). In 2015, U.S. gastrointestinal radiation oncologists created guidelines for recommended radiation targets and volumes when designing intensity-modulated radiation therapy (IMRT) for esophageal and GEJ cancer (4). These guidelines outlined elective nodal basins (celiac, para-aortic, gastrohepatic ligament, and supraclavicular lymph nodes) that should be irradiated depending on the location of the primary disease. However, the impact of adherence to these guidelines when designing elective nodal irradiation (ENI) has not been examined.

We hypothesize that incomplete coverage of these areas increases the risk of developing distant failures in patients with esophageal and GEJ cancers who are receiving CRT. In this study, we retrospectively compared the rates of distant failures of these patients who did or did not receive adequate ENI by the 2015 guideline recommendations. We present this article in accordance with the STROBE reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-609/rc).

Methods

This study included adults with non-metastatic gastroesophageal (GE) and esophageal cancer patients who completed CRT pre-operatively or definitively from 2012 to 2021 from a single institution. Non-metastatic disease state was assessed by positron emission tomography (PET) imaging (92.1%) or computed tomography (CT) imaging alone (7.9%). All patients fulfilling these inclusion criteria were selected to participate in the study. All patients completed either an exhale/inhale breath hold CT or four dimensional (4D)-CT during simulation to account for tumor motion. Standard 3–4 cm proximal and distant expansions from the primary disease were applied per consensus guidelines. Radial border expansion varied depending on the degree of confidence of motion management with a minimum of 1 cm to account for at-risk adjacent paraesophageal nodes but was cropped out of normal uninvolved structures (lung, heart, liver) if 4D-CT was obtained. Finally, expansions of 0.5 to 1 cm to generate the planning target volume were used. Because this study included definitive and pre-operative cases, doses administered ranged from 41.4 to 50.4 Gy. Radiation plans were then analyzed by tumor location by the senior investigator, who is a board-certified radiation oncologist. Plans were deemed guideline-compliant if radiation dose coverage of at least 41.4 Gy encompassed nodal basins recommended by the 2015 guideline by the senior investigator (4). Table 1 outlines recommended ENI by primary tumor location.

Table 1. Recommended ENI by primary tumor location.

Primary tumor location	Recommended nodal coverage
Cervical esophagus	Supraclavicular, high echelon cervical
Proximal 1/3 esophagus	Paraesophageal, supraclavicular
Middle 1/3 esophagus	Paraesophageal
Distal 1/3 esophagus	Paraesophageal, lesser curvature, splenic (provider discretion), celiac axis
GEJ	Gastrohepatic, para-aortic

ENI, elective nodal irradiation; GEJ, gastroesophageal junction.

The primary endpoint of this study was the rate of distant metastasis defined as disease beyond the primary and regional lymph nodes. Other endpoints examined include locoregional failure, overall survival (OS), and distant metastasis-free survival (DMFS). Locoregional failure was defined as failure within the regional lymph nodes of the primary tumor, as well as within the gross tumor volume. OS rates were calculated from the time a patient completed radiation to their recorded date of death or last recorded oncology appointment as appropriate. DMFS was calculated from the time a patient completed radiation to their recorded date of distant failure or any cause of death. Failure rates were calculated from the time a patient finished chemoradiation and/or surgery to the date of confirmed failure. Follow-up data was gathered from the patient’s chart as the last time the patient was recorded meeting with an oncologist. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of University of California Irvine (IRB No. 2021-6535) and due to the retrospective and non-experimental nature of the study, no informed consent was obtained and all patient data remains anonymous.

Statistical analysis

Summary and descriptive statistics were used to characterize key cohort features. Tests of differences utilized Chi-squared and Fisher’s exact test for categorical variables and two-sided t-tests for continuous measures. Kaplan-Meier survival plots were also created to visualize disease free and distant metastasis survival. To assess sensitivity of the results to death as a competing risk, Fine-Gray estimates of cumulative incidence of distant metastasis were also calculated and are reported in Table S1. The association of various characteristics including tumor location, grade, and the risk of failure for each endpoint of distant, locoregional, and local failures was assessed using univariate Cox proportional hazards (PH) regression with a statistical significance threshold of P<0.05.

Results

A total of 38 patients were included in the study. The median age at diagnosis was 66 years with most patients being male (94.7%). Patients commonly presented with T3 (52.6%) N0 (44.7%) moderately (44.7%) or poorly differentiated (44.7%) adenocarcinoma (73.7%) located at the GEJ (57.9%). During radiation therapy, patients received either carboplatin + paclitaxel (86.8%), folinic acid, fluorouracil, oxaliplatin (FOLFOX) (10.5%), or a combination of treatment regimens (2.6%) for concurrent chemotherapy. A total of 4 patients received induction chemotherapy with either FOLFOX (3 patients) or carboplatin/paclitaxel (1 patient). The median radiation dose delivered to the primary tumor was 50.4 Gy (range, 41.4–50.4 Gy). Twenty patients underwent an esophagectomy. A complete list of patient characteristics can be found in Table 2.

Table 2. Patient characteristics stratified by guideline adherence.

Variable	Overall	Non-guideline adherent treatment group	Guideline adherent treatment group	P value
Median age at diagnosis (years)	66	61	67	0.20
Sex, n				0.50
Male	36	19	17
Female	2	2	0
Race, n				0.60
White	23	14	9
Hispanic	8	3	5
African American	6	3	3
Unstated	1	1	0
Staging, n
T				0.80
T1	7	5	2
T2	8	4	4
T3	20	10	10
T4	3	2	1
N				0.02
N0	17	9	8
N1	14	11	3
N2	7	1	6
Grade, n				0.009
Well differentiated	3	3	0
Moderately differentiated	17	5	12
Poorly differentiated	17	12	5
Not initially stated	1	1	0
Histologic subtype, n				0.70
Adenocarcinoma	28	16	12
Squamous cell carcinoma	10	5	5
Chemotherapy, n				0.20
Carboplatin + paclitaxel	33	20	13
FOLFOX	4	1	3
More than 1 regimen	1	0	1
Induction chemotherapy, n				0.08
FOLFOX	3	0	3
Carboplatin + paclitaxel	1	1	0
Primary tumor location, n				0.70
Proximal esophagus	2	1	1
Middle esophagus	2	2	0
Distal esophagus	13	6	7
GEJ	21	12	9
Surgical treatment, n				0.70
Ivor-Lewis	18	11	7
Transhiatal	1	1	0
Non-specified	1	0	1
Received adjuvant with nivolumab, n	4	1	3	0.30

FOLFOX, folinic acid, fluorouracil, oxaliplatin; GEJ, gastroesophageal junction; N, node; T, tumor.

A total of 17 patients received irradiation that properly covered recommended nodal basins as specified by the IMRT treatment guidelines. Only 1 of these patients demonstrated evidence of distant failures. In contrast, a total of 21 patients were not treated according to the guidelines and did not have adequate ENI coverage. The most common improperly covered nodal basin was para-aortic (66.7%), followed by gastrohepatic (aka lesser curvature) (23.8%) and celiac (23.8%), and supraclavicular (14.3%). Complete data can be found in Table 3. Examples of treatment plans displaying inadequate coverage can be seen in Figures 1,2.

Table 3. Patient results of current retrospective study.

Event analyzed	Number of patients
Number of patients receiving
Guideline adherent treatment	17
Guideline non-adherent treatment	21
Occurrences of improperly irradiated nodal basins
Gastrohepatic	5
Para-aortic	14
Celiac	5
Supraclavicular	3
Rates of distant failures
Guideline adherent treatment	1/17
Guideline non-adherent treatment	8/21
Median follow-up in weeks
Guideline adherent group	107
Guideline non-adherent group	96

Figure 1.

Patient with gastroesophageal cancer without adequate gastrohepatic coverage of minimum 41.40 Gy as indicated by the arrow.

Figure 2.

Patient with gastroesophageal cancer with incomplete coverage of para-aortic lymph nodes posteriorly as indicated by the arrow.

Pattern of failure by ENI coverage

The median follow-up of guideline-compliant and non-compliant group was similar (107 vs. 96 weeks, P=0.30). Out of the 21 patients with suboptimal ENI, eight patients showed evidence of distant failure within three years following treatment cessation. A total of seven patients with distant metastases also experienced local nodal failures. Overall, only one patient with sufficient ENI coverage (1/17) developed distant failure compared to eight patients (8/21) with insufficient coverage (P=0.02). There were also appreciable differences in locoregional failure rates, with patients in the non-guideline adherent group showing locoregional (1/21) and locoregional plus distant (7/21) metastases and one patient in the guideline adherent group showing evidence of local recurrent disease between those with (0/17) and without complete ENI (P=0.009). When examining other tumor characteristics [histologic subtype, location, human epidermal growth factor 2 (HER2) status, esophagectomy rate] of patients with and without complete ENI, no further differences were noted. The Cox PH univariate analysis data for distant relapse can be found in Table 4.

Table 4. Univariate Cox PH analysis for distant failure.

Patient and tumor characteristic	HR	LCL	UCL	P value
Esophageal location of tumor
Upper	0.00	0.00	Infinity	>0.99
Middle	0.00	0.00	Infinity	>0.99
Lower	1.73	0.44	6.83	0.44
Induction chemotherapy
FOLFOX	0.00	0.00	Infinity	>0.99
Carbotaxol	0.00	0.00	Infinity	>0.99
Adjuvant chemotherapy	9.40	1.92	46.03	0.01
T-stage
2	2.42	0.22	26.95	0.47
3	1.97	0.24	16.43	0.53
4	0.00	0.00	Infinity	>0.99
N-stage
1	5.68	1.18	27.37	0.03
2	0.00	0.00	Infinity	>0.99
Tumor grade
Moderate	0.00	0.00	0.10	0.001
Poor	0.04	0.00	0.47	0.01
Tumor histology
Adenocarcinoma	0.88	0.18	4.28	0.87
Female	2.63	0.31	22.67	0.38
Age at diagnosis	0.94	0.89	1.00	0.03
Surgery performed
Ivor-Lewis	1.46	0.29	7.44	0.65
Mckeown	–	–	–	–
Transhiatal	–	–	–	–

FOLFOX, folinic acid, fluorouracil, oxaliplatin; HR, hazard ratio; LCL, lower confidence interval limit; N, node; T, tumor; PH, proportional hazards; UCL, upper confidence interval limit.

DMFS

Median DMFS was significantly higher among the 17 patients who completed guidelines as compared to DMFS among the 21 non-guideline compliant patients (median not reached vs. 30 months, P=0.01), as depicted in Figure 3. Only one patient in the guideline-adherent group developed distant failure. On univariate analysis, receiving adjuvant immunotherapy was found to correlate with metastatic disease with a hazard ratio (HR) of 9.4 [95% confidence interval (CI): 1.92–46.03]. Patients with N1 disease were also more likely to experience distant metastatic failure with a HR of 5.68 (95% CI: 1.18–27.37) when compared to patients with N0 disease. To further elucidate this point, we ran a Cox PH analysis which found that the effect of guideline-compliance remained protectively associated with DMFS (HR =0.17) even when adjusting for node (N)-stage. The adjusted exploratory Cox PH analysis demonstrated better overall fit when compared to the univariate N-stage model (likelihood ratio P value: 0.047). This Cox PH univariate analysis data for DMFS can be found in Table 5. When examining death as a competing factor, a total of 16 deaths were recorded. Among the 16 deaths, 9 patients were from the non-guideline compliant group, and 7 patients were from the guideline compliant group. There does not appear to be a difference in the risk of death when comparing guideline compliance groups (P>0.9). Further analysis can be found in Tables S1-S3.

Figure 3.

Distant metastasis free survival Kaplan-Meier curve comparing patients who did and did not receive guideline-adherent therapy.

Table 5. Univariate Cox PH analysis for distant metastasis-free survival.

Tumor and patient characteristics	HR	LCL	UCL	P value
Esophageal location of tumor
Upper	0.00	0.00	Infinity	>0.99
Middle	0.00	0.00	Infinity	>0.99
Lower	1.73	0.44	6.83	0.44
Induction chemotherapy
FOLFOX	0.00	0.00	Infinity	>0.99
Carbotaxol	0.00	0.00	Infinity	>0.99
Adjuvant chemotherapy	9.40	1.92	46.03	0.01
T-stage
2	3.29	0.28	38.99	0.35
3	2.80	0.33	23.38	0.34
4	0.00	0.00	Infinity	>0.99
N-stage
1	5.68	1.18	27.37	0.03
2	0.00	0.00	Infinity	>0.99
Tumor grade
Moderate	0.01	0.00	0.14	0.001
Poor	0.04	0.00	0.43	0.01
Tumor histology
Adenocarcinoma	0.88	0.18	4.28	0.87
Female	2.63	0.31	22.67	0.38
Age at diagnosis	0.96	0.91	1.01	0.11
Surgery performed
Ivor-Lewis	1.05	0.26	4.35	0.94
Mckeown	–	–	–	–
Transhiatal	–	–	–	–

FOLFOX, folinic acid, fluorouracil, oxaliplatin; HR, hazard ratio; LCL, lower confidence interval limit; N, node; T, tumor; PH, proportional hazards; UCL, upper confidence interval limit.

OS

The median OS in the non-guideline compliant group was 30 months but was not reached in the guideline compliant group, as illustrated in Figure 4 (median not reached vs. 30 months, P=0.03). On univariate analysis, having N1 disease was correlated with lower OS, with a HR of 7.47 (95% CI: 2.06–27.08) when compared to patients with N0 disease. The Cox PH univariate analysis data for OS can be found in Table 6.

Figure 4.

OS Kaplan-Meier curve comparing patients who did and did not receive guideline-adherent therapy. OS, overall survival.

Table 6. Univariate Cox PH analysis for overall survival.

Patient and tumor characteristic	HR	LCL	UCL	P value
Esophageal location of tumor
Upper	1.59	0.19	13.00	0.66
Middle	24.00	3.10	185.85	0.001
Lower	1.37	0.46	4.08	0.57
Induction chemotherapy
FOLFOX	0.68	0.09	5.17	0.71
Carbotaxol	0.00	0.00	Infinity	>0.99
Adjuvant chemotherapy	2.57	0.96	6.87	0.06
T-stage
2	3.63	0.38	34.93	0.26
3	5.01	0.65	38.90	0.12
4	3.19	0.20	50.99	0.41
N-stage
1	7.47	2.06	27.08	0.001
2	1.65	0.27	9.86	0.59
Tumor grade
Moderate	0.24	0.05	1.25	0.09
Poor	0.40	0.08	1.91	0.25
Tumor histology
Adenocarcinoma	0.51	0.19	1.42	0.20
Female	1.06	0.14	8.02	0.96
Age at diagnosis	0.99	0.95	1.03	0.56
Surgery performed
Ivor-Lewis	0.55	0.19	1.54	0.25
Mckeown	–	–	–	–
Transhiatal	27.33	1.67	448.02	0.02

FOLFOX, folinic acid, fluorouracil, oxaliplatin; HR, hazard ratio; LCL, lower confidence interval limit; N, node; T, tumor; PH, proportional hazards; UCL, upper confidence interval limit.

Discussion

In this study, we found an association between non-compliance to ENI guidelines and increased risk of distant metastasis. The most common areas of inadequate radiation coverage were the para-aortic (66.7%) or gastrohepatic/celiac nodal basins (23.8%). Out of the 38 patients, only 17 received radiation plans with adequate ENI. Only one patient in this group developed metastatic disease compared to eight patients with non-compliant plans who developed distant failures. This is one of the first studies that demonstrates the potential impact of incomplete coverage of elective nodal stations in esophageal cancer treated with CRT.

Prior studies examining the benefit of ENI for esophageal squamous cell carcinoma have not demonstrated improved outcomes. A study from Shandong Cancer Hospital examined 126 patients with T2-4N0M0 thoracic esophageal squamous cell carcinoma who completed definitive chemoradiation did not appreciate a difference in OS at 1-year, 3-year, or 5-year between those who did not and did complete ENI, respectively (5). While the ENI group had a numerically superior median progression-free survival (mPFS) of 30 months, this was not statistically different (5). All patients analyzed in this study had esophageal squamous cell carcinoma only. A meta-analysis of six randomized controlled trials and 17 retrospective studies selecting for those who completed definitive chemoradiation investigated OS and PFS with and without ENI. Of the 41,420 patients, 1,841 received ENI and only 2.4% demonstrated non-squamous cell histology. There was found to be a higher OS among those who did not receive ENI. Out of the 23 total studies in this meta-analysis, 18 of them only analyzed patients with esophageal squamous cell carcinoma. From the five remaining studies in the meta-analysis, patients with esophageal adenocarcinoma comprised <10% of the total patient population studied. The investigators also appreciated lower rates of high-grade adverse treatment events including esophagitis (6).

These studies suggest that ENI should not impact distant metastasis, contradicting our findings. However, significant differences exist between our study and the ones previously discussed. The primary difference being that approximately 80% of patients in our study presented with adenocarcinoma rather than squamous cell carcinoma, which present with different patterns of spread and higher risk of progression after definitive therapy (3). Additionally, esophageal adenocarcinoma is a more radioresistant disease as evidenced by lower rates of pathologic response after surgery compared to their squamous counterparts (3).

This study highlights a potential dissemination mechanism of esophageal cancer. Traditionally, metastases are thought to spread from hematogenous dissemination of the primary tumor (7). However, ENI addresses occult lymphatic disease. Therefore, our findings suggest that impairing the spread of micro-metastatic nodal disease through comprehensive ENI impacts the rate of distant metastasis. This could be explained by cancer’s potential ability to suppress immune activation once in the lymph nodes and subsequent travel through the vascular supply of involved lymph nodes to other organs (8). This mechanism of spread is supported by previous studies, which noted that esophageal squamous cell carcinoma with high mRNA expression of vascular endothelial growth factor (VEGF)-C is correlated with increased risk of lymph node metastasis and overall poor prognosis (9). Several VEGF signals have been found to be associated with metastasis formation in esophageal cancer (10). Because VEGF promotes lymphangiogenesis and angiogenesis, metastatic seeding between occult metastatic lymph nodes and solid organs could occur. By ensuring proper ENI coverage during radiation treatment planning, this process could be interrupted.

IMRT guidelines published by U.S. gastrointestinal radiation oncologists in 2015 recommended ENI of anatomical zones (celiac, gastrohepatic, para-aortic) based on tumor location to reduce risk of lymph node recurrence. However, this differs compared to seminal prospective trials examining the role of chemoradiation outside of the U.S. The CROSS regimen specifies the radiation volume expansion necessary to treat subclinical disease by distance based on the location of the primary disease (3). On the Neo-AEGIS trial, the subclinical nodal disease was addressed by treating the adjacent esophageal fat pad of the primary tumor (11). Elective nodal basins were covered only if involved. The recently released ESOPEC trial also utilized the CROSS design for radiation volumes (12). Because these protocols likely provide partial coverage of the guideline-recommended ENI basins, an investigation into the patterns of failure similar to this study examining guideline-compliant ENI could verify our findings and encourage integration of US consensus guidelines into ENI design for future trials involving chemoradiation.

Various limitations exist within this study including its retrospective approach and limited number of patients. Both pre-operative and definitive chemoradiation patients were included in this study which increases patient population heterogeneity and therefore could confound the results of this study. Another potential confounding factor could be the presence of more patients in the guideline non-compliant group having N1 disease compared to the guideline compliant group. However, the adjusted Cox PH analysis suggests guideline compliance may still contribute to the understanding of DMFS beyond what is captured by N-staging alone. While there may be more N1 patients in the guideline non-compliant group and more N2 patients in the guideline compliant group, it is difficult to assess their true difference and effect on patient’s outcomes due to the relatively small sample size. It is possible that the patients with higher N-stage disease (N2) tend to have lower rates of distant disease and better outcomes due to these patients being more likely to receive guideline adherent therapy or because these patients are receiving more aggressive CRT. Further studies with an increased sample size would allow for a greater analysis of the role of disease N-stage and grade.

Another potential confounding factor could be the presence of more patients with poor grade disease receiving guideline non-adherent treatment despite having better outcomes. This difference in groups cannot be explained by guideline treatment alone. It is possible that this difference is due to the patients with worse grade disease being more likely to receive more aggressive chemoradiation, nivolumab, or adjuvant therapy. However, because of the relatively small sample size, it is difficult to identify the true effect disease grade has on patient outcomes.

In addition, death was not found to be a competing factor in this study, and future investigators should be cognizant of death as a potential competing risk when compared to distant metastatic free survival. While our competing risks cumulative incidence analysis returned results consistent with those presented in our primary analysis, this does not imply that there is independence between these events. Also, the wide confidence intervals and large p-values illustrated in the result tables, such as the transhiatal surgery results, are attributed to the limited sample size and do not likely illustrate any significant meaning. Larger future studies should be conducted to further verify the validity of the results of this study.

Finally, the radiation dose was selected at a minimum of 41.4 Gy due to recommendations for pre-operative radiation therapy from the CROSS study. However, the absolute dose required to reduce the risk of lymphatic spread is not known. One retrospective study comparing ENI doses of 36 to 40 Gy and higher in 79 patients with inoperable esophageal cancer after chemoradiation did not appreciate any differences in PFS or OS (13).

Conclusions

We appreciated that the most common nodal basins lacking complete dosimetric coverage were the para-aortic and gastrohepatic regions. We noted correlation between compliance to ENI guidelines and fewer distant metastases. This benefit could be relevant for those with lower esophageal and GEJ adenocarcinoma. Improved delineation of patients at higher risk of occult disease using circulating tumor DNA (ctDNA) help to determine who would benefit from ENI. Even though our findings suggest improved outcomes with optimal ENI coverage, the benefit of ENI could change with induction chemotherapy or adjuvant nivolumab. Our findings require further validation, which could be performed by analyzing previously published prospective trials where ENI volumes did not follow U.S. expert guidelines.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of University of California Irvine (IRB No. 2021-6535) and due to the retrospective and non-experimental nature of the study, no informed consent was obtained and all patient data remains anonymous.

Data Sharing Statement

Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-609/dss

jgo-17-01-1-dss.pdf ^{(62.6KB, pdf)}

DOI: 10.21037/jgo-2025-609