Resected lymph nodes and survival of patients with esophageal squamous cell carcinoma: an observational study

Background: The incidence and mortality of esophageal cancer are high. Therefore, the authors aimed to investigate how the number of dissected lymph nodes (LNs) during esophagectomy for esophageal squamous cell carcinoma impacts overall survival (OS), particularly that of patients with positive LNs. Materials and methods: Data from 2010 to 2017 were obtained from the Sichuan Cancer Hospital and Institute Esophageal Cancer Case Management Database. Participants were divided into two groups: patients with negative lymph nodes (N0) and patients with positive lymph nodes (N+). The median number of resected LNs during surgery was 24; therefore, patients with 15–23 and those with 24 or more resected LNs were assigned to subgroups A and B, respectively. Results: After a median follow-up of 60.33 months, 1624 patients who underwent esophagectomy were evaluated; 60.53 and 39.47% had a pathological diagnosis of N+ or N0, respectively. The median OS was 33.9 months for the N+ group; however, the N0 group did not achieve the median OS. The mean OS was 84.9 months. In the N+ group, the median OS times of subgroups A and B were 31.2 and 37.1 months, respectively. The OS rates at 1, 3, and 5 years were 82, 43, and 34%, respectively, for subgroup A of the N+ group; they were 86, 51, and 38%, respectively, for subgroup B of the N+ group. Subgroups A and B of the N0 group exhibited no statistically significant differences. Conclusion: Increasing the number of LNs harvested during surgery to 24 or more could improve the OS of patients with positive LNs but not that of patients with negative LNs.


Introduction
According to the Systematic Analysis for the Global Burden of Disease Study, the numbers of new cases and deaths of esophageal cancer in 2019 were 535 000 and 498 000, respectively, and its incidence and mortality were ranked high among the 29 tumor categories investigated [1] . Esophageal cancer is the fourth primary cause of cancer-related death and has the sixth highest incidence in China [2] . Esophageal squamous cell carcinoma (ESCC) is the main subtype in China, accounting for~90% of cases [3] . Radical resection of esophageal cancer is recommended for all resectable cases. Surgery with radiotherapy and chemotherapy comprise the cornerstone treatment for localized ESCC [4] .
In recent years, the CROSS and CheckMate-577 trials have enabled remarkable achievements. The CROSS trial indicated that the median disease-free survival of patients undergoing surgery alone was 11.6 months; however, that of patients with squamous cell carcinoma treated with neoadjuvant chemoradiotherapy was 74.7 months. CheckMate-577 revealed that the HIGHLIGHTS • Esophageal squamous cell carcinoma is the main occurring subtype in China. • Overall survival is associated with an increased number of resected lymph nodes. • An esophagectomy was mainly performed through a right transthoracic procedure. • The pathological N stage is an independent prognostic factor for overall survival. median disease-free survival of patients treated with neoadjuvant chemoradiotherapy was 11 months for the squamous cell carcinoma subgroup; however, it was 22.4 months for the immunomaintenance therapy subgroup [5,6] . After the randomized, controlled CROSS trial, the CROSS study group launched the TIGER study, which demonstrated that the establishment of an optimal surgical strategy for esophageal cancer patients has been increasingly emphasized by Western surgeons [7,8] . Several studies have confirmed that lymph node (LN) metastasis (LNM) is an important factor impacting overall survival (OS) associated with esophageal cancer [9][10][11] . In recent years, Chinese researchers have conducted studies of LN dissection; however, the sample sizes were small [12][13][14] .
In this study, we evaluated the impact of the number of resected LNs (RLNs) on the OS of patients with thoracic ESCC (TESCC) who underwent esophagectomy. In addition, we determined the effects of different numbers of RLN segments during surgery on the OS of patients with positive LNs (N + ) and those with negative LNs (N0).

Materials and methods
The data were obtained from the database at our institution. We performed a retrospective analysis of patients with esophageal cancer from January 2010 to December 2017. The study was approved by the Ethics Committee for Medical Research and New Medical Technology of our hospital. The research has been reported in line with the Strengthening the Reporting of Cohort Studies in Surgery (STROCSS) guidelines [15] , Supplemental Digital Content 1, http://links.lww.com/JS9/A571.
The retrieved data included demographic and pathological information, such as sex, age, T stage, N stage, tumornode-metastasis (TNM) stage, tumor location (upper, middle, or lower thoracic), tumor grade, nerve invasion, central LNM, and radical resection. Esophagectomy was mainly performed using a right transthoracic procedure with two-field or three-field LN dissection; the surgical approaches were dependent on the patient's characteristics and the surgeon's discretion. The disease stage was classified according to the American Joint Committee on Cancer eighth edition of the TNM system. Patients were followed up once every 3 months for the first 2 years; thereafter, they were followed up once every 6 months for 3 to 5 years. A total of 2957 patients with TESCC treated with esophagectomy were identified from January 2010 to December 2017. There were three inclusion criteria: esophagectomy was performed; the tumor was located in the thoracic esophagus; and the pathology results confirmed squamous cell carcinoma. The exclusion criteria were as follows: less than 15 RLNs were resected during surgery; other malignant tumors were present; pathological T stage (Tis/T1a/T4/M1); preoperative treatment was performed; and required data were missing. The OS was calculated from the month and year of surgery to death or the last follow-up evaluation in March 2021. For survival outcomes, we used the median as far as possible; when the medians were not achieved, means were used.
Patients were divided into two groups. Patients with N0 according to the pathological results were assigned to the N0 group. Patients whose pathological results showed N + were assigned to the N + group. Because the median number of RLNs during surgery was 24, patients with 15-23 RLNs were assigned to subgroup A, and those with 24 or more RLNs were assigned to subgroup B. The TNM stages were used to compare clinical outcomes and survival data.

Statistical analysis
Class variables are expressed as percentages. We calculated the results using χ 2 or Fisher precision tests. Independent OS-related risk factors were identified by single-variable and multivariable logistic regression analyses; hazard ratios and 95% CIs were calculated. The impact of all baseline covariates on the results was assessed using the Cox proportional disaster hazards regression model. The observation system was evaluated using Kaplan-Meier curves, and the results of the logarithmic grade tests were compared to describe the median at specific time points as the 95% CI. P < 0.05 was considered statistically significant. All analyses were performed using SPSS version 23.0 (Chicago).

Patient characteristics
From January 2010 to December 2017, the data of 1624 patients were retrieved and retrospectively analyzed; of these, 641 (39.5%) and 983 (60.5%) had pathological N0 and N + , respectively. Males comprised 83.0% (1348/1,624), and females comprised 17.0% (276/1624) of the patients. More than half of them (60.5%; 983/1,624) exceeded pathological stage III. There were 1473 patients who underwent two-field LN dissection, 88 patients who underwent three-field LN dissection, and 63 patients who underwent two-field LN dissection combined with unilateral neck dissection. Among these patient groups, the median numbers of dissected LNs were 23, 39, and 36, respectively. A total of 3683 (8.6%) supraclavicular LNs were dissected. Moreover, 259 of 3683 (7.0%) LNs were positive. There were 23 608 (55.3%) mediastinal (thoracic) LNs, and 2030 of 23 608 (8.5%) LNs were positive. There were 15 397 (36.1%) abdominal LNs. Moreover, 1343 of 15 379 (8.7%) LNs were positive. During this study, 641 participants comprised the N0 group, and 983 comprised the N + group (Fig. 1). The clinicopathological and pathological characteristics of the N0 and N + groups are presented in Table 1. Subgroup B of the N + group had more patients with a poor pathological TNM stage than subgroup A of the N + group (

Discussion
This study clarified the value of RLNs during lymphadenectomy in terms of long-term survival benefits. Compared with the N0 group, the N + group exhibited significantly improved OS as the number of RLNs increased. In the N + group, the OS of subgroup B was better than that of subgroup A, and there were no statistically significant differences in the demographic characteristics between groups, except for the TNM stage and surgical techniques for lymphadenectomy. However, in the N + group, the number of cases worse than TNM stage IV was higher in subgroup B (19.2%) than in subgroup A (11.7%) (P = 0.006).
Although the surgical techniques for esophageal carcinoma were not the same in terms of LN harvesting, surgical techniques for lymphadenectomy had no statistical significance in the singlefactor analysis ( Table 2). Moreover, the three groups of patients were very close in the Kaplan-Meier curves, with a nonsignificant P-value in total patients and the N + group (Fig. 3). Sex, tumor grade, T stage, and N stage significantly influenced the OS of patients in this study (Table 3).
These findings suggest that the treatment strategy should depend on the presence of LNM. For patients without LNM, there was no significant difference in OS between patients with

15-23
LNs and those with 24 or more LNs dissected during surgery. Therefore, the current Chinese expert consensus and guidelines, which indicate that the required number of resected LNs should exceed 15, are reasonable [4] . However, for patients with LNM, the surgical requirements described in the Chinese Society of Clinical Oncology diagnosis and treatment guidelines for malignant LNs in 2021 require further consideration. In our study, patients with LNM who underwent the removal of at least 24 LNs exhibited considerably better OS than patients who underwent the removal of 15-23 LNs. This suggests that the surgical strategy for patients with suspected LNM should include the removal of at least 24 LNs during esophagectomy. The OS of patients with N + can be considerably improved by increasing the number of LNs removed during surgery to 24 or more. Therefore, this is a favorable surgical treatment strategy. Moreover, the clinical stage of patients should be carefully distinguished before esophagectomy, especially when there is suspicion of LNM indicated by computed tomography, ultrasonography, or positron emission tomography [16,17] , and the surgical treatment strategy should be adjusted appropriately [18,19] . The Memorial Sloan-Kettering Cancer Center proposed that 5-year OS is dependent on the T classification: for pT1, the resection of 10 LNs is adequate; for pT2, at least 20 LNs should be removed; and for pT3/T4, 30 or more LNs should be removed [20] . Similarly, different pN stages require the removal of different LNs. A study performed at the Cleveland Clinic suggested that the removal of at least 25 LNs was optimal for patients receiving neoadjuvant therapy for esophageal adenocarcinoma [20] , and our study showed no obvious difference in the OS of subgroups A and B. This indicates that the OS of patients did not change much after the removal of a sufficient number of LNs. However, LNM is an independent risk factor for OS; therefore, we individually evaluated the LNM of each patient. In this study, the number of LNs removed during lymphadenectomy was associated with significantly improved OS for TESCC patients with LNM. We hope to provide more evidence regarding the characteristics and surgical treatment of TESCC for medical professionals in China. Concerning the clinical efficacy of LN removal at each station for TESCC patients, further studies including more samples and clinical data from more centers are required to verify these findings and improve the OS of patients with esophageal cancer.
In recent years, various studies, such as CROSS, NEOCRTEC5010, and CheckMate-577, of the treatment of esophageal cancer have attracted considerable attention, and a comprehensive treatment modality based on esophagectomy was formulated. The OS of patients was significantly different in the NEOCRTEC5010 and CROSS trials; however, LN dissection might have affected these results, especially those of patients with suspected LNM [20][21][22][23][24][25] . Metastases of vestigial LNs could negatively impact OS and reduce the effects of subsequent treatment. We deduced from our results that the removal of at least 24 LNs is crucial to the surgical treatment of patients with N + .
Unfortunately, the relevant lymphadenectomy guidelines and consensus for patients with suspected LNM do not adequately apply to all patients. When patients have suspiciously large LNs, it is sufficient to remove 15 or more LNs; however, for patients with N0, some stations do not require consideration in accordance with lymphadenectomy standards. Conversely, a careful and systematic evaluation of each LN station is vital for an accurate and appropriate lymphadenectomy. Therefore, these challenges must be urgently addressed.
There are some limitations to our study. This study did not consider real-world confounding factors that could have influenced the results. Furthermore, 12 groups at our center performed esophagectomy from January 2010 to December 2017. McKeown esophagectomy and Ivor-Lewis esophagectomy were the main surgical types, two-field or three-field lymphadenectomy was performed, and careful systematic lymphadenectomy was not performed at each station. Therefore, certain subjective selection bias exists in the results. The clinical value and efficacy index of different lymph stations were different. Our research of the efficacy index will be presented in the near future. This study included only single-center data that were retrospectively collected; therefore, their generalizability requires consideration. Moreover, the results of postoperative complications were lacking in this study, and only the OS outcomes were evaluated. Multicenter cooperation among hospitals in China is required to obtain larger cohorts to collect compelling evidence to enhance the guidelines for lymphadenectomy for esophageal cancer. By combining the results of multicenter data analyses, specific and detailed treatment options can be developed.

Conclusions
Increasing the number of LNs harvested during surgery to 24 or more could improve the OS of patients with N + . Patients without LNM achieve better OS than those with LNM. The guidelines recommend dissecting at least 15 LNs for patients without LNM, and no further OS benefit was observed when the number of harvested LNs was increased for these patients. Patients with different stages of disease should be provided with different LN dissection strategies.

Ethical approval
All procedures performed in this study were in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Ethics Committee (EC) for Medical Research and New Medical Technology of Sichuan Cancer Hospital (SCCHEC-02-2022-050). Consent was waived by the Ethics Committee (EC) due to the retrospective nature of the study.