The impact of extended lymphadenectomy on survival in esophageal adenocarcinoma with complete pathologic response: a retrospective study

Introduction

Esophageal adenocarcinoma (EAC) remains a morbid disease despite advancements in neoadjuvant and immunotherapy regimens (1). Following the CROSS trial in 2012, the mainstay of treatment is a trimodal approach with neoadjuvant chemoradiation followed by esophagectomy (2,3). Despite this standardization and the knowledge that nodal involvement remains the most impactful driver in prognosis, there is currently debate surrounding lymphadenectomy for EAC (4,5). While current National Comprehensive Cancer Network (NCCN) guidelines advocate for a 15-node harvest, recent data have identified a survival benefit with more extended lymphadenectomy (6-10). The application of this knowledge to those with node-negative or less invasive disease remains unclear and is an area of active conversation (11-13).

In the era of advancing neoadjuvant regimens, an increasingly substantial population of patients achieves complete pathologic response (CPR) at resection (14,15). As these patients are thought to have total resolution of disease at the microscopic level, it is logical to infer a limited benefit derived from lymphadenectomy. To our knowledge, there remains minimal data evaluating this population. As such, we sought to investigate the impact of lymph node harvest on survival in patients with EAC receiving guideline-concordant therapy that achieve CPR at esophagectomy. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-172/rc).

Methods

The National Cancer Database (NCDB) was queried from 2006–2021 for patients with non-metastatic EAC that received neoadjuvant chemoradiotherapy. Inclusion criteria required patients to be >18 years of age, with complete clinical and pathologic staging, survival, and lymph node harvest data. Those with staging data coded under the 7^th edition American Joint Committee on Cancer (AJCC) guidelines were restaged to the latest 8^th edition using appropriate Tumor, Node, Metastasis (TNM) criteria. Three lymph node groups were created: <10 nodes representing a limited nodal harvest, 10–20 nodes mirroring guideline concordance, and >20 nodes offering a more extensive lymphadenectomy. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Statistical analysis

Demographic and clinical characteristics were compared using Chi-squared and two-sided t-tests where appropriate. Multivariable logistic regression models were used to assess the predictive value of lymph node harvest when considering survival. Univariate analysis of overall survival was conducted using the Kaplan-Meier method. All analyses were two-sided with significance considered at P<0.05. Analyses were performed using SAS v9.4 (SAS Institute; Cary, NC, USA) in concert with a biostatistician at the Medical College of Wisconsin. This study was submitted to our institutional review board and deemed exempt as the NCDB is a publicly available database.

Results

Patient demographics

Of 17,292 patients with non-metastatic EAC receiving trimodal therapy identified in the NCDB, 3,547 (20.51%) were identified as having CPR. Of those, 6.82% (n=242) were clinical stage (cSTAGE) I, 21.23% (n=753) were cSTAGE II, and 71.95% (n=2,552) were cSTAGE III at diagnosis. There were no demographic or baseline clinical differences identified between these groups. There was a difference in lymph node harvest by cSTAGE with higher stage disease portending a higher rate of extended lymph node harvest (Table 1).

Lymph node harvest as an independent prognostic factor

In a multivariable analysis of all patients achieving CPR, lymph node harvest of 10–20 nodes [hazard ratio (HR) 0.847, 95% confidence interval (CI): 0.727–0.987, P=0.03] and >20 nodes (HR 0.768, 95% CI: 0.639–0.924, P=0.01) was a positive independent prognostic factor for survival (Table 2). Subset analyses by cSTAGE were completed. In cSTAGE I disease with CPR, lymph node harvest was not a predictor of survival at 10–20 (P=0.29) or >20 (P=0.52) nodes (Table 3). Similarly, in cSTAGE II disease with CPR, lymph node harvest did not independently impact survival in 10–20 nodes (P=0.60) or >20 nodes (P=0.34) (Table 4). In those with cSTAGE III disease with CPR, however, lymph node harvest of both 10–20 nodes (HR 0.811, 95% CI: 0.675–0.975, P=0.03) and >20 nodes (HR 0.770, 95% CI: 0.621–0.955, P=0.02) served as an independent prognostic factor for improved overall survival (Table 5).

Impact of lymph node harvest on survival

When evaluating survival in patients of all cSTAGEs who achieved CPR, extended lymph node harvest of >20 nodes appeared to offer a benefit in both five-year overall survival (5-y OS) (58.54%) and median survival (81.15 months, 95% CI: 72.38–103.26, P<0.01) compared to <10 nodes (5-y OS 49.84%, median survival 59.70 months, 95% CI: 52.37–77.86). The advantage observed in the >20 node group was not significantly more than the 10–20 node group (5-y OS 58.15%, median survival 91.70 months, 95% CI: 80.10–108.94) (Figure 1).

Figure 1 Survival in all EAC patients with CPR by lymph node harvest. CPR, complete pathologic response; EAC, esophageal adenocarcinoma.

A subset analysis was completed by cSTAGE. In those with cSTAGE I disease, there was no significant difference in survival with increasing extent of lymphadenectomy. Patients with <10 nodes harvested had a 5-y OS of 69.85% and median survival of 143.57 months (95% CI: 110.68–NE) while a 10–20 node harvest demonstrated a similar 5-y OS of 64.30% and median survival of 137.86 months (95% CI: 72.97–NE). Extended lymphadenectomy of >20 nodes conferred a 5-y OS of 76.15% and median survival of 133.68 months (95% CI: 90.81–NE) (P=0.25) (Table 6, Figure 2).

Figure 2 Survival in cSTAGE I EAC with CPR by lymph node harvest. CPR, complete pathologic response; cSTAGE, clinical stage; EAC, esophageal adenocarcinoma.

Those with cSTAGE II disease also did not demonstrate a difference in median or overall survival by lymph node harvest. The 5-y OS was 50.76% in the <10 node group, 61.20% in the 10–20 node group, and 59.98% in the >20 node group. Median survival was variable at 65.28 months (95% CI: 49.91–95.59) in the <10 node group, 104.97 months (95% CI: 76.94–124.98) in the 10–20 node group, and 106.05 months (95% CI: 60.81–NE) in the >20 node group, however not statistically different (P=0.12) (Table 6, Figure 3).

Figure 3 Survival in cSTAGE II EAC with CPR by lymph node harvest. CPR, complete pathologic response; cSTAGE, clinical stage; EAC, esophageal adenocarcinoma.

A significant difference in survival was noted with increased lymph node harvest in patients with cSTAGE III disease with CPR. The 5-y OS in those with <10 nodes harvested was 46.82%, compared to 56.52% in the 10–20 node group and 56.18% in the >20 node group. Median survival demonstrated similar trends at 53.32 months (95% CI: 42.12–66.14) in the <10 node group, 84.21 months (95% CI: 71.70–99.71) in the 10–20 node group, and 74.64 months (95% CI: 64.20–92.22) in the >20 node group (P<0.01) (Table 6, Figure 4). Interestingly, in those with cSTAGE III disease, this model also revealed prognostic value in age (HR 1.563, 95% CI: 1.335–1.830, P<0.001), female sex (HR 0.723, 95% CI: 0.563–0.928, P=0.01) and higher Charlson-Deyo score (HR 1.687, 95% CI: 1.047–2.718, P=0.03) (Table 5).

Figure 4 Survival in cSTAGE III EAC with CPR by lymph node harvest. CPR, complete pathologic response; cSTAGE, clinical stage; EAC, esophageal adenocarcinoma.

The NCDB utilizes a 30-day and 90-day mortality variable to identify early postoperative deaths. These patients are often captured by the unique 90-day variable and not identified through Kaplan-Meier modeling. The 90-day survival was comparable across cSTAGEs I, II, and III (P=0.28) (Table 1) as well as across lymph node harvest groups (Table 6). Kaplan-Meier modeling thus begins at the three-month mark to better demonstrate these trends unique to the NCDB EAC registry (Figures 1-4).

Discussion

There remains much conversation regarding adequate lymphadenectomy in EAC. While current guidelines advocate for a 15-node harvest at minimum, recent data suggest that extended lymphadenectomy may offer a survival advantage in those with residual disease following neoadjuvant therapy (6-9,16). In the era of advancing neoadjuvant regimens, there exists a growing subset of patients with remarkable treatment response for whom the utility and role of extended lymphadenectomy remains less clear. This ambiguity is well summarized by Yeung et al. in their review (17). They acknowledge the dual role of nodal harvest in both establishing accurate staging and positively influencing survival. It is this duality that remains conflicting in those with node-negative and early-stage disease, and largely unaddressed in those with CPR.

Data put forth by Rice et al. showed a need for a far more substantial lymphadenectomy in patients with well-differentiated, smaller tumors to determine nodal positivity (up to 60 nodes in tumors <2.5 cm), with limited survival benefit offered by such extensive nodal harvest in those with lower-stage disease (11). Their findings align with the retrospective work of Altorki that highlighted a survival benefit only with lymphadenectomy of greater than 40 nodes in those with pathologic node-negative disease (13). These studies, along with others, offer the consensus that in early-stage and node-negative disease, a far more substantial lymphadenectomy is necessary to confer limited staging and survival benefits (17).

The subgroup of patients found to have CPR at esophagectomy offers a unique opportunity for investigation. As CPR implies disease resolution even at the microscopic level, the benefit of lymphadenectomy, particularly of an extended nature, is thought to be limited. Lutfi et al. attempted to investigate this theory in their NCDB study spanning 2004 to 2014 with data showing a survival benefit with increasing nodal harvest up to 25 nodes (18). Their findings are certainly interesting as they contradict the hypothesis that lymphadenectomy may not offer benefit in those with CPR, however generalizability is limited by a heterogeneous cohort (nearly 30% of patients were classified as esophageal squamous cell carcinoma) and notable variations in neoadjuvant regimens with a lack of stratification by cSTAGE.

Amid such active conversation regarding the role of lymph node harvest in EAC, the goal of our study was to identify the role of extended lymphadenectomy in those with CPR. The greatest limitation to studies in those with CPR is the inability to predict CPR with post-neoadjuvant imaging studies. To address this, we stratified our analysis by cSTAGE to allow for application of our findings to clinical data available in the preoperative period. We elected to create three groups: those with <10 nodes representing a more limited harvest, 10–20 nodes within the region of guideline concordance, and >20 nodes as a surrogate for extended lymphadenectomy. These groups were intended to encompass a less extensive, guideline concordant, and more extensive lymphadenectomy, respectively. Additionally, as we evaluated the dataset, we identified a trimodal distribution of nodes harvested. This also favored the creation of our groups, with a peak seen around 8, 15, and 25 nodes.

Within our study cohort, patients with EAC who received trimodal therapy in the modern era demonstrated a CPR rate of 20.51%, consistent with current trends in the literature (15,19). The majority of these patients were diagnosed with cSTAGE III disease, mirroring persistent tendencies of advanced stage at diagnosis (1). Our initial analysis evaluating the utility of lymph node harvest as an independent predictor of overall survival found that in patients of all cSTAGEs achieving CPR, a lymph node harvest of 10–20 nodes was independently associated with increased survival compared to <10 nodes (HR 0.847, P=0.03). However, there was no additional survival associated with a more extensive lymphadenectomy of >20 nodes. On subgroup analysis by cSTAGE, we determined that the survival advantage initially observed in all patients was largely driven by the cSTAGE III group. In both cSTAGE I and II patients, lymphadenectomy was not independently associated with a survival advantage (P=0.52 and P=0.60, respectively) at any level. Additional analyses of both median and 5-y OS also failed to demonstrate a survival advantage in these groups. However, in the cSTAGE III group, a survival benefit of approximately 10 months was shown in patients with 10–20 nodes and >20 nodes harvested when compared to patients with <10 nodes (P<0.01). No additional survival benefit was observed when comparing between 10–20 and >20 nodes harvested (P=0.84). It is important to note that guideline-concordant or extended lymphadenectomy did not serve as a negative prognostic indicator in any group.

These trends merit further consideration, particularly in the context of recent discussion regarding the utility of extended lymphadenectomy in EAC. As CPR implies the absence of tumor cells within a specimen, it was our hypothesis that lymphadenectomy at any level, irrespective of cSTAGE, would not offer a survival advantage. The lack of improvement in survival observed in those with cSTAGE I and II disease supports this theory, however the notable survival benefit observed in cSTAGE III patients with at least 10 nodes harvested is intriguing and calls into question the dual role of lymphadenectomy in EAC. It is reasonable to conclude from our dataset that the survival benefit observed in cSTAGE III patients with CPR may be attributable to positive nodal disease present outside the captured lymphadenectomy. This aligns with the previously mentioned findings of Altorki and Rice, demonstrating a need for an extensive lymphadenectomy of over 40 nodes in those with notable treatment response, an exceedingly rare occurrence given the nature of lymphadenectomy in esophagectomy (11,13,17). As such, our data advocate for a traditional approach to resection and lymphadenectomy for those with cSTAGE III disease with CPR, even in the setting of apparent remarkable treatment response with suspicion for CPR. This is further supported by our multivariate model identifying at risk subpopulations of this cohort consistent with previously published data (15,20). Contrastingly, in those with cSTAGE I and II disease, the benefit of lymphadenectomy, particularly of an extended nature, appears to be quite limited. As such, the benefits of an extensive dissection with goal of increased lymph node harvest likely do not outweigh the risks and should be considered in operative planning.

Given the limitations of current clinical staging practices in EAC, the question then becomes how to best implement these findings into practice. Any meaningful change in surgical management of patients with EAC proceeding to esophagectomy remains contingent on advancements in clinical staging techniques. Within our current staging and treatment paradigm, we believe our work offers two key findings: first, there is an urgent need to improve staging accuracy, particularly for patients with earlier-stage disease. Our data suggests that, if able to accurately predict CPR in this group, patients may benefit from a more conservative intraoperative approach by reducing the risk of an unnecessary extended lymphadenectomy. Additionally, there are instances in which <15 lymph nodes are harvested, and staging is rendered inadequate by current guidelines; by our data, this may in fact be sufficient for reasonable prognostication. Second, for patients with more advanced stage disease at diagnosis (cSTAGE III), even achieving CPR does not fully mitigate the need for extensive resection of nodal tissue, even when pathologically negative. Accordingly, clinical restaging is of less significance in this group as their intraoperative approach remains unchanged. Though we remain hindered by clinical staging modalities, as CPR rates continue to increase and staging concordance becomes a more realistic goal, analysis of this group remains of utmost importance.

Our study has limitations, the most notable of which is a lack of understanding of the requisite number of nodes harvested to ensure accurate pathologic staging in those with local resolution of disease. This is fundamentally important work that is underway within our group. There are also limitations to our study inherent to all large database endeavors. Despite the NCDB serving as a representative national cohort, there remains potential for selection bias and variations in coding practices. The NCDB also lacks detail regarding specific neoadjuvant regimens, extent of regional lymph node dissection, and type of esophagectomy as well as data regarding disease-free survival and disease recurrence. Additionally, as is characteristic of database studies, variation in pathologic analysis from center to center may limit external validity, though previous works have demonstrated comparable rates of CPR identification by treatment center type (15). Similarly, missingness of data results in a higher burden of censoring in analyses. This study, while the largest current cohort to our knowledge, has limited power in smaller subset analyses which may reduce the generalizability of more significant trends.

Conclusions

The utility of extended lymphadenectomy in esophageal cancer remains a topic of ongoing discussion. While the consensus has shifted to emphasize the benefit of greater lymph node harvest in most groups, debate persists surrounding its efficacy in select patient subsets, particularly those with early-stage disease and a remarkable treatment response. Our data demonstrate that among patients with early clinical-stage disease who achieve a complete pathologic response (CPR), greater lymph node harvest does not confer a survival advantage. In contrast, for patients presenting with cSTAGE III disease, lymph node harvest of at least 10 nodes remains associated with improved survival, although extending lymphadenectomy beyond current guideline recommendations (represented by our >20 lymph node group) does not offer additional benefit. As neoadjuvant advances continue to improve CPR rates in EAC, understanding these distinct patient populations will be critical to refining surgical strategies and improving outcomes.

Acknowledgments

The abstract of this article was presented at the International Thoracic Surgical Oncology Summit on September 28, 2024 in Times Square, New York, NY, USA.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-172/rc

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-172/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-172/coif). The authors have no conflicts of interest to declare.

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. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA Cancer J Clin 2024;74:12-49. [Crossref] [PubMed]
2. Shapiro J, van Lanschot JJB, Hulshof MCCM, et al. Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial. Lancet Oncol 2015;16:1090-8. [Crossref] [PubMed]
3. Ajani JA, D'Amico TA, Bentrem DJ, et al. Esophageal and Esophagogastric Junction Cancers, Version 2.2019, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2019;17:855-83. [Crossref] [PubMed]
4. Schuring N, van Berge Henegouwen MI, Gisbertz SS. History and evidence for state of the art of lymphadenectomy in esophageal cancer surgery. Dis Esophagus 2024;37:doad065. [Crossref] [PubMed]
5. Hagens ERC, van Berge Henegouwen MI, Cuesta MA, et al. The extent of lymphadenectomy in esophageal resection for cancer should be standardized. J Thorac Dis 2017;9:S713-23. [Crossref] [PubMed]
6. Henckens SPG, Hagens ERC, van Berge Henegouwen MI, et al. Impact of increasing lymph node yield on staging, morbidity and survival after esophagectomy for esophageal adenocarcinoma. Eur J Surg Oncol 2023;49:89-96. [Crossref] [PubMed]
7. Sihag S, Nobel T, Hsu M, et al. A More Extensive Lymphadenectomy Enhances Survival After Neoadjuvant Chemoradiotherapy in Locally Advanced Esophageal Adenocarcinoma. Ann Surg 2022;276:312-7. [Crossref] [PubMed]
8. Visser E, Markar SR, Ruurda JP, et al. Prognostic Value of Lymph Node Yield on Overall Survival in Esophageal Cancer Patients: A Systematic Review and Meta-analysis. Ann Surg 2019;269:261-8. [Crossref] [PubMed]
9. Greenstein AJ, Litle VR, Swanson SJ, et al. Effect of the number of lymph nodes sampled on postoperative survival of lymph node-negative esophageal cancer. Cancer 2008;112:1239-46. [Crossref] [PubMed]
10. Bahardoust M, Torabi S, Yarahmadi D, et al. Impact of negative lymph node removal on survival in esophageal cancer: a systematic review and meta-analysis. BMC Surg 2025;25:124. [Crossref] [PubMed]
11. Rice TW, Ishwaran H, Hofstetter WL, et al. Esophageal Cancer: Associations With (pN+) Lymph Node Metastases. Ann Surg 2017;265:122-9. [Crossref] [PubMed]
12. Khoma O, Paredes SR, Park JS, et al. Extensive lymphadenectomy may improve survival in node negative oesophageal cancer. Sci Rep 2024;14:2711. [Crossref] [PubMed]
13. Altorki NK, Zhou XK, Stiles B, et al. Total number of resected lymph nodes predicts survival in esophageal cancer. Ann Surg 2008;248:221-6. [Crossref] [PubMed]
14. Gaber CE, Sarker J, Abdelaziz AI, et al. Pathologic complete response in patients with esophageal cancer receiving neoadjuvant chemotherapy or chemoradiation: A systematic review and meta-analysis. Cancer Med 2024;13:e7076. [Crossref] [PubMed]
15. Donato BB, Campany ME, Brady JT, et al. Complete pathologic response in esophageal adenocarcinoma: does it make a difference? Dis Esophagus 2024;37:doae068. [Crossref] [PubMed]
16. Peyre CG, Hagen JA, DeMeester SR, et al. The number of lymph nodes removed predicts survival in esophageal cancer: an international study on the impact of extent of surgical resection. Ann Surg 2008;248:549-56. [Crossref] [PubMed]
17. Yeung JC, Bains MS, Barbetta A, et al. How Many Nodes Need to be Removed to Make Esophagectomy an Adequate Cancer Operation, and Does the Number Change When a Patient has Chemoradiotherapy Before Surgery? Ann Surg Oncol 2020;27:1227-32. [Crossref] [PubMed]
18. Lutfi W, Martinez-Meehan D, Dhupar R, et al. Higher lymph node harvest in patients with a pathologic complete response after neoadjuvant therapy for esophageal cancer is associated with improved survival. J Surg Oncol 2020;121:654-61. [Crossref] [PubMed]
19. Reynolds JV, Preston SR, O'Neill B, et al. Trimodality therapy versus perioperative chemotherapy in the management of locally advanced adenocarcinoma of the oesophagus and oesophagogastric junction (Neo-AEGIS): an open-label, randomised, phase 3 trial. Lancet Gastroenterol Hepatol 2023;8:1015-27. [Crossref] [PubMed]
20. Donato BB, Campany ME, Brady JT, et al. Lymphovascular Invasion is an Independent Negative Prognostic Factor in Esophageal Adenocarcinoma. Ann Surg Oncol 2024;31:6680-90. [Crossref] [PubMed]