Upstaging Risk and Prognosis of Clinical T2-3N0M0 Esophageal Squamous Cell Carcinoma Patients after a Curative Esophagectomy: A Propensity Score-Matched Analysis

Abstract

Background

Neoadjuvant chemoradiation therapy (nCRT) has survival benefits for patients with cT1N1 or cT2–3N0-1 esophageal squamous cell carcinoma (ESCC), but heterogeneous disease behavior has been observed in node-negative tumors. This study aimed to investigate the risk of upstaging and the efficacy of upfront esophagectomy compared with nCRT followed by surgery in patients with cT2–3N0 ESCC.

Methods

ESCC patients treated at Taipei Veterans General Hospital, Taiwan, between January 2010 and December 2022 were retrospectively reviewed. Patients who underwent upfront surgery (n=72) and those who received neoadjuvant therapy (n=36) were included in this analysis on an intention-to-treat basis after propensity score matching. Comparisons of TNM (tumor, node, and metastasis) classification shifts between patients who underwent upfront surgery and those who received neoadjuvant therapy, as well as primary outcomes including overall survival (OS) and recurrence-free survival (RFS), were analyzed.

Results

Among patients who underwent upfront surgery, 6 (16.7%) in the cT2 group and 10 (27.8%) in the cT3 group had occult lymph node (LN) metastasis. Most patients (77.4%) who received neoadjuvant therapy experienced downstaging; however, there were no significant differences in OS (5-year OS 53.8% vs. 51.6%, p=0.666) or RFS (5-year RFS 48.6% vs. 44.8%, p=0.864) compared with upfront surgery. No significant differences were observed before propensity score matching for either OS (p=0.809) or RFS (p=0.856).

Conclusion

Upfront esophagectomy in patients with cT2–3N0 ESCC provides comparable survival outcomes, despite the higher risk of upstaging and occult LN metastasis.

Introduction

Esophageal cancer is an aggressive gastrointestinal malignancy with a poor prognosis despite advances in treatment modalities, and its predominant subtype, esophageal squamous cell carcinoma (ESCC) [1], has the highest regional incidence in Eastern Asia. Surgical resection remains the standard therapy for patients with locally advanced esophageal cancer, but neoadjuvant chemoradiation therapy (nCRT) has shown survival benefits in patients with cT1N1 or cT2–3N0-1 ESCC according to the Chemoradiotherapy for Oesophageal Cancer Followed by Surgery Study (CROSS) [2,3]. For locally advanced ESCC, except for early-stage cT1N0 disease, current ESMO Clinical Practice Guidelines recommend nCRT followed by surgery [4]. However, heterogeneous disease behavior has been observed, including among node-negative cT2–T3 tumors. Therefore, evidence remains insufficient to establish firm recommendations for preoperative treatment in these cases.

A previous study reported that patients with esophageal cancer staged lower than pT2N0 have a better prognosis with upfront surgery, as this approach avoids unnecessary chemoradiation-related morbidity or mortality and potential surgical complications [5]. Thus, identifying higher-risk patients who would benefit from induction therapy is expected to yield better outcomes. The application of nCRT may be advantageous within the range of diagnostic uncertainty described in the literature. Semenkovich et al. [6] conducted a decision analysis demonstrating that the therapeutic benefit of nCRT depends on whether the probability of upstaging exceeds 48.1%. Consequently, accurate preoperative staging in potentially resectable esophageal cancer is essential for risk-benefit assessment and to minimize inappropriate treatment selection.

Although the recent CROSS trial demonstrated superior outcomes with nCRT, the absolute benefit for cN0 patients remains uncertain. For example, patients with cN0 esophageal adenocarcinoma who received nCRT in addition to surgery did not exhibit a significant overall survival (OS) advantage compared with surgery alone [7]. A retrospective study examining the optimal treatment strategy for clinical T2–T3N0M0 esophageal cancer found that neoadjuvant chemotherapy did not improve prognosis in either the cT2 or cT3 subgroups individually and that most cT2–T3N0M0 patients (60%) had occult lymph node involvement despite induction therapy [8]. Another nationwide analysis based on the National Cancer Database (NCDB) reported that induction therapy did not improve survival in patients with clinical T2N0 (cT2N0) esophageal cancer, even though the upstaging rate was 41.6% [9]. In contrast, a recent nationwide study from Taiwan [10] demonstrated that applying nCRT before esophagectomy significantly improved OS and recurrence-free survival (RFS) in patients with cT3N0 ESCC [10]. Nevertheless, few studies have specifically addressed the oncologic prognosis of clinically staged T2–3 node-negative ESCC. Therefore, this study aimed to assess the risk of upstaging and evaluate the efficacy of upfront esophagectomy versus nCRT followed by surgery in patients with ESCC categorized as cT2N0 or cT3N0.

Methods

Study design and eligibility criteria

This single-center study retrospectively reviewed patients from a prospectively maintained database of the Taipei Veterans General Hospital multidisciplinary tumor board between January 2010 and December 2022. The principles of multidisciplinary management have been previously described [11]. Patients were considered eligible if they had squamous cell carcinoma of the esophagus clinically staged as cT2–T3N0M0 and were scheduled for esophagectomy, with or without neoadjuvant therapy before surgery. Patients were excluded if they lacked a pre-treatment positron emission tomography–computed tomography (PET/CT) scan for cancer staging, underwent salvage-intended treatment, experienced surgery-related mortality, or were lost to follow-up (within 2 years) without recurrence events or death.

Preoperative staging included a physical examination, laboratory testing, tumor marker assays, esophagogastroduodenoscopy, and flexible bronchoscopy (for upper and middle third tumors). Additional imaging studies consisted of CT scans from the neck to the upper abdomen and radionuclide bone scans. Clinical nodal status (cN0) was primarily assessed using PET/CT scans, based on the background standardized uptake of lymph nodes. Clinical tumor staging (cT2–3) was determined using endoscopic ultrasound (EUS) and chest CT findings, with any discordant results resolved through multidisciplinary team consensus.

The study protocol was approved by the Institutional Review Board of Taipei Veterans General Hospital (approval no., 2024-12-006AC), and informed consent was waived due to the retrospective study design. The study was conducted in accordance with institutional guidelines and the principles of the Declaration of Helsinki.

Neoadjuvant therapy and surgical treatment

Upfront surgical resection was primarily performed using the McKeown tri-incisional esophagectomy approach, which included mediastinal, abdominal pericardial, and celiac axis lymph node dissection, as well as regional cervical lymph node dissection near the esophagogastric anastomosis. At our institution, neoadjuvant therapy followed by esophagectomy is recommended for patients with locally advanced ESCC (cT2–4N0) or positive clinical nodal status (cN+). However, in this study, treatment selection for patients with cT2–3N0 disease depended on the surgeon’s expertise and preference. The neoadjuvant regimen consisted of 2 courses of chemotherapy, with intravenous cisplatin followed by continuous intravenous infusion of 5-fluorouracil. Chemotherapy was administered concurrently with 41.4–50.4 Gy of external-beam radiation delivered in 1.8–2 Gy fractions targeting the primary tumor. The clinical target volume encompassed the gross tumor volume delineated on CT and other diagnostic images, extended by 3–5 cm cephalad and at least 5 cm caudally. The final target volume was adjusted according to the relative risk of nodal metastasis at specific nodal locations based on the tumor’s site of origin. The interval between the completion of radiotherapy and curative esophagectomy was approximately 6–10 weeks. Pathological staging followed the 8th edition of the American Joint Committee on Cancer TNM (tumor, node, and metastasis) classification. Occult nodal metastasis was defined as the presence of pathologically positive lymph nodes in the absence of preoperative imaging evidence of nodal disease. Post-neoadjuvant tumor response was graded using the College of American Pathologists protocol for esophageal carcinoma as follows: tumor regression grade (TRG) 0 (pathologic complete response), no residual cancer cells; TRG 1 (moderate response), minimal residual cancer; TRG 2 (minimal response), partial tumor regression; and TRG 3 (poor response), no definite histologic regression. After curative esophagectomy, postoperative adjuvant therapy was recommended by multidisciplinary consensus for patients with close margins or positive lymph nodes, while it was considered optional for squamous cell carcinoma with T3 disease.

Follow-up protocol

Patients were followed every 3 months during the first 2 years and every 6–12 months thereafter. Postoperative evaluations included serum tumor marker testing, chest radiography, and CT scans from the neck to the upper abdomen. Endoscopy, abdominal sonography, brain magnetic resonance imaging, radionuclide bone scans, and PET/CT scans were performed when clinically indicated. Follow-up data regarding survival and recurrence were collected through February 2025.

Statistical analyses

Categorical variables were compared using the Pearson chi-square test or Fisher’s exact test, while continuous variables were analyzed using the independent t-test. The primary outcomes were OS and RFS. OS was defined as the period from surgical resection (in the upfront surgery group) or initiation of neoadjuvant therapy (in the neoadjuvant setting group) to death or the last follow-up. RFS was defined as the period from surgical resection (upfront surgery group) or initiation of neoadjuvant therapy (neoadjuvant setting group) to the first recurrence, death, or the last follow-up if no recurrence occurred. Survival curves were generated using the Kaplan-Meier method and compared using the log-rank test. A p-value <0.05 was considered statistically significant. Propensity score matching was performed using logistic regression and optimal 1:2 matching based on the following covariates: age, sex, clinical T stage, tumor length, preoperative EUS assessment, and performance status. Standardized mean differences were used to evaluate covariate balance between groups. Statistical analyses were conducted using IBM SPSS ver. 25.0 (IBM Corp.), and visualization was performed with the Survival, ggplot2, MatchIt, and survminer packages in R ver. 4.5.1 (The R Foundation for Statistical Computing).

Results

Clinical characteristics

Data from 149 consecutive patients with ESCC clinically staged as cT2–3N0 and scheduled for curative esophagectomy were collected. Among patients who underwent upfront surgery (n=107), 3 lacked a preoperative PET/CT scan, 3 were lost to follow-up, and 1 patient with a cervical tumor location was excluded. Of the 42 patients who received neoadjuvant treatment, 1 lacked a preoperative PET/CT scan, 2 were lost to follow-up, and 3 initially declined surgery due to tumor shrinkage but later underwent salvage esophagectomy after tumor relapse and were therefore excluded. Ultimately, 108 clinically T2–3N0 patients were included in the analysis on an intention-to-treat basis after propensity score matching (Fig. 1). A PET/CT scan was obtained for all patients, and an EUS was performed in 81 patients (75%). The cohort was divided into 2 groups: upfront surgery (n=72) and neoadjuvant therapy (n=36). After propensity score matching, there were no significant differences between the 2 groups in sex, age, tumor length, tumor location, clinical T stage, PET/CT SUVmax value, performance status, or comorbidities. In the neoadjuvant therapy group, 5 patients did not proceed to esophagectomy: 2 due to poor tolerance from chemoradiotherapy-related adverse effects, 2 owing to disease progression, and 1 who died during neoadjuvant treatment. The median observation time was 51.5 months. Patient demographics and clinical characteristics are summarized in Table 1.

Fig. 1.

Flow chart of study design. ESCC, esophageal squamous cell carcinoma; PET/CT, positron emission tomography–computed tomography.

Table 1.

Characteristics of patients with cT2-3N0M0 squamous cell carcinoma (n=136)

Characteristic	Before propensity score matching				After propensity score matching
	Upfront surgery (n=100)	Neoadjuvant+surgery (n=36)	p-value		Upfront surgery (n=72)	Neoadjuvant setting (n=36)	p-value	SMD
Age (yr)	61 (53.3–69)	62.5 (53–67.8)	0.960		61.5 (54.3–69.8)	62.5 (53–67.8)	0.797	0.0102
Sex			0.438				0.584	0.1356
Male	86 (86)	29 (80.6)			61 (84.7)	29 (80.6)
Female	14 (14)	7 (19.4)			11 (15.3)	7 (19.4)
Tumor location			0.833				0.753
Upper	21 (21)	7 (19.4)			10 (13.9)	7 (19.4)
Middle	47 (47)	19 (52.8)			40 (55.6)	19 (52.8)
Lower/gastroesophageal	32 (32)	10 (27.8)			22 (30.5)	10 (27.8)
Clinical T stage			0.024				0.172	0.4493
cT2	58 (58)	13 (36.1)			36 (50)	13 (36.1)
cT3	42 (42)	23 (63.9)			36 (50)	23 (63.9)
Surgical approach			0.172				0.299
Open	5 (5)	0			4 (5.6)	0
MIE	95 (95)	36 (100)			68 (94.4)	36 (100)
Operative method			0.236				0.109
McKeown	98 (98)	34 (94.4)			72 (100)	34 (94.4)
Ivor-Lewis	1 (1)	2 (5.6)			0	2 (5.6)
TPLE	1 (1)	0			0	0
Route of reconstruction			0.363				0.606
Retrosternal	76 (76)	30 (83.3)			57 (79.2)	30 (83.3)
Posterior mediastinal	24 (24)	6 (16.7)			15 (20.8)	6 (16.7)
Tumor length (cm)	3.5 (2.5–4.8)	3.0 (1.8–4.0)	0.053		3.05 (2.5–4.5)	3.0 (1.78–4.0)	0.318	–0.3382
PET/CT SUVmax	12.3 (8.35–16.1)	12.6 (8.89–17.48)	0.407		12.6 (8.18–16.1)	12.6 (8.89–17.48)	0.336
Preoperative EUS	54 (54)	30 (83.3)	0.002		51 (70.8)	30 (83.3)	0.157	0.7871
Charlson comorbidity index			0.918				0.892
≥1	51 (51)	18 (50)			37 (51.4)	18 (50)
0	49 (49)	18 (50)			35 (48.6)	18 (50)
ECOG performance status			0.573				0.639	0.0990
1	8 (8)	4 (11.1)			6 (8.3)	4 (11.1)
0	92 (92)	32 (88.9)			66 (91.7)	32 (88.9)

Values are presented as median (interquartile range) or number (%) unless otherwise stated.

SMD, standardized mean difference; MIE, minimally invasive esophagectomy; TPLE, total pharyngolaryngoesophagectomy; PET/CT, positron emission tomography/computed tomography; SUV, standardized uptake value; EUS, endoscopic ultrasonography; ECOG, Eastern Cooperative Oncology Group.

Pathological findings and TNM classification shifts

Pathological findings are summarized in Table 2. Compared with the neoadjuvant group, the upfront surgery group showed significantly more advanced pathological stage (p<0.001), pathological T stage (p<0.001), and tumor grade (p<0.001), as well as a higher proportion of patients who received postoperative treatment (p<0.001). The upfront surgery group also exhibited more frequent lymphovascular invasion (p=0.006), perineural invasion (p=0.017), and a greater number of harvested lymph nodes (p<0.001). There were no statistically significant differences in pathological N stage (p=0.144) or the number of metastatic lymph nodes between the 2 groups. Comparisons of TNM classification shifts between the upfront surgery group (n=72) and the neoadjuvant therapy group (n=36) are presented in Table 3. Among patients who underwent upfront surgery in the cT2 group (n=36), 6 (16.7%) had occult lymph node metastasis (N+), and 12 (33.3%) showed upstaging in either T or N classification. In the cT3 upfront surgery group (n=36), 10 patients (27.8%) had occult lymph node metastasis and 11 (30.6%) experienced T or N upstaging. In contrast, the overall pathologic complete response (ypT0N0) rate in the neoadjuvant therapy group was 41.7%. Most patients experienced downstaging after neoadjuvant therapy—91.7% in the cT2 group and 73.7% in the cT3 group. Only 2 patients (10.5%) in the cT3 group exhibited postoperative regional lymph node metastasis. The overall upstaging risk in the total cohort was 24.1%, with 31.9% in the upfront surgery group and 8.3% in patients who received nCRT.

Table 2.

Pathological findings of patients with cT2-3N0M0 squamous cell carcinoma after matching (n=108)

Variable	Upfront surgery (n=72)	Neoadjuvant setting (n=36)	p-value
Pathological stage			<0.001
0 (Pathologic complete response)	0 (0.0)	15 (41.7)
IA	2 (2.8)	3 (8.3)
IB	12 (16.7)	6 (16.7)
IIA	22 (30.5)	2 (5.6)
IIB	21 (29.2)	3 (8.3)
IIIA	1 (1.4)	0 (0.0)
IIIB	14 (19.4)	2 (5.6)
Without surgery	0 (0.0)	5 (13.9)
Pathological T stage			<0.001
0 (Pathologic complete response)	0 (0.0)	15 (41.7)
1a	2 (2.8)	3 (8.3)
1b	14 (19.4)	6 (16.7)
2	16 (22.2)	1 (2.8)
3	39 (54.2)	6 (16.7)
4a	1 (1.4)	0 (0.0)
Without surgery	0 (0.0)	5 (13.9)
Pathological N stage			0.144
N0	56 (77.8)	29 (80.6)
N1	14 (19.4)	2 (5.6)
N2	2 (2.8)	0
Without surgery	0 (0.0)	5 (13.9)
Tumor grade			<0.001
Gx (undifferentiated)	0 (0.0)	15 (41.7)
G1 (well differentiated)	1 (2.8)	0 (0.0)
G2 (moderately differentiated)	55 (76.4)	14 (38.9)
G3 (poorly differentiated)	16 (22.2)	2 (5.6)
Without surgery	0 (0.0)	5 (13.9)
Lymphatic vascular invasion	23 (31.9)	2 (5.6)	0.006
Perineural invasion	23 (31.9)	3 (8.3)	0.017
Postoperative treatment			<0.001
Adjuvant therapy	28 (38.9)	2 (5.6)
None	44 (61.1)	34 (94.4)
No. of LNs harvested	21.5 (16.0–31.8)	10.0 (14.0–19.0)	<0.001
No. of metastatic LNs			0.087
Cervical	2 (2.8)	0 (0.0)
Mediastinum	12 (16.7)	1 (2.8)
Abdominal	4 (5.6)	2 (5.6)
30-day surgical related mortality	4 (5.6)	0 (0.0)	0.299
R1 resection	5 (6.9)	0 (0.0)	0.167

Values are presented as number (%) or median (interquartile range) unless otherwise stated.

LN, lymph node.

Table 3.

Comparison of TNM classification shifts in patients with upfront surgery or neoadjuvant therapy after matching

Variable	pN or ypN	No. (%)	pT or ypT	No. (%)
Upfront surgery	pN		pT
cT2N0 (n=36)	pN0	30 (83.3)	pT1a-b	13 (36.1)
			pT2	11 (30.5)
			pT3	6 (16.7)
	pN+	6 (16.7)	pT1bN1	2 (5.5)
			pT2N1	1 (2.8)
			pT3N1	2 (5.5)
			pT3N2	1 (2.8)
cT3N0 (n=36)	pN0	26 (72.2)	pT1b	1 (2.8)
			pT2	3 (8.3)
			pT3	21 (58.3)
			pT4a	1 (2.8)
	pN+	10 (27.8)	pT2N1	1 (2.8)
			pT3N1	8 (22.2)
			pT3N2	1 (2.8)
Neoadjuvant therapy	ypN		ypT
cT2N0 (n=12)	ypN0	12 (100.0))	ypT0	5 (41.7)
			ypT1a-b (TRG1: 2, TRG2: 4)	6 (50.0)
			ypT3 (TRG3)	1 (8.3)
cT3N0 (n=19)	ypN0	17 (89.5)	ypT0	10 (52.6)
			ypT1a-b (TRG1)	3 (15.8)
			ypT2 (TRG1: 1)	1 (5.3)
			ypT3 (TRG2)	3 (15.8)
	ypN+	2 (10.5)	ypT3N1 (TRG1: 1, TRG2: 1)	2 (10.5)
Without surgery (n=5)

TRG, tumor regression grade.

Overall survival and recurrence-free survival

The 5-year OS and RFS rates for the total cohort were 52.4% and 46.0%, respectively. There were no significant differences between the upfront surgery and neoadjuvant therapy groups (5-year OS: 52.8% vs. 53.8%, p=0.809; 5-year RFS: 43.8% vs. 48.6%, p=0.856) (Fig. 2A, B). After propensity score matching, survival outcomes remained comparable (5-year OS: 51.6% vs. 53.8%, p=0.666; 5-year RFS: 44.8% vs. 48.6%, p=0.864) (Fig. 2C, D). No significant differences were observed between the 2 groups regarding recurrence events or recurrence patterns (Table 4). Patients who developed recurrence were treated with definitive chemotherapy or chemoradiotherapy following relapse.

Fig. 2.

Overall survival and recurrence-free survival curve of patients with cT2–3 esophageal squamous cell carcinoma who underwent upfront surgery or neoadjuvant therapy followed by surgery (A, B) and patients who underwent upfront surgery or neoadjuvant therapy followed by surgery after propensity score matching (C, D). (Continued on next page.)

Table 4.

Recurrence events and patterns in patients with upfront surgery or neoadjuvant therapy after matching (n=108)

Variable	Upfront surgery (n=72)	Neoadjuvant setting (n=36)	p-value
Recurrence event	27 (37.5)	12 (33.3)	0.671
Recurrence pattern			0.381
No recurrence or metastasis	45 (62.5)	24 (66.7)
Locoregional recurrence	17 (23.6)	6 (16.7)
Distal metastasis	7 (9.7)	6 (16.7)
Locoregional and distal recurrence	3 (4.2)	0
With locoregional recurrence	20 (27.8)	6 (16.7)	0.203
With distal metastasis	10 (13.9)	6 (16.7)	0.702

Values are presented as number (%), unless otherwise stated.

Discussion

The Neoadjuvant Chemoradiotherapy Followed by Surgery Versus Surgery Alone for Locally Advanced Squamous Cell Carcinoma of the Esophagus (NEOCRTEC5010) trial demonstrated the benefit of nCRT among patients with locally advanced ESCC (cT1–4N1 or cT4N0), with more than 80% of patients presenting with clinically staged N1 disease [12,13]. Because patients with esophageal cancer and nodal involvement benefit from neoadjuvant chemoradiation, identifying underestimated nodal metastasis preoperatively is critical [2,7]. To address the therapeutic dilemma arising from inaccurate staging, multidisciplinary assessment and treatment planning for preoperative tumors are essential [4]. Therefore, this study assessed the clinical value and potential benefits of these approaches in patients with cT2– 3N0 ESCC.

All patients in our cohort were evaluated by PET/CT scans as part of routine clinical staging, and most also underwent both EUS and PET/CT examinations preoperatively. A total of 26 patients (24.1%) were pathologically upstaged, including 23 patients (31.9%) in the upfront esophagectomy group who showed T or N classification upstaging, and 16 patients (22.2%) with nodal involvement. Most patients achieved downstaging after nCRT with a favorable pathologic complete response rate, likely attributable to the squamous histology. Stiles et al. [8] retrospectively analyzed 102 patients clinically staged as cT2–3N0, of whom approximately 92% had undergone PET/CT or EUS. Survival outcomes were comparable between cT2 and cT3 patients treated with or without neoadjuvant therapy. Pathologic lymph node positivity was the strongest predictor of overall survival, highlighting the prognostic impact of nodal metastasis. Based on EUS staging accuracy, patients with cT2N0 disease may derive a survival advantage from nCRT, which should be considered during treatment planning if the probability of upstaging exceeds the 48.1% threshold [6].

A retrospective analysis of 1,559 patients diagnosed with cT2N0 esophageal cancer between 1998 and 2011 in the NCDB revealed that only one-quarter of surgically resected patients had accurate pretreatment staging [9]. Moreover, 41.6% were pathologically upstaged in the upfront surgery group, and 30.2% had node-positive disease. Although patients who received nCRT exhibited lower upstaging and higher downstaging rates, there was no significant difference in long-term survival. A subsequent nationwide NCDB study (2006–2012) focusing on cT2N0 patients who underwent upfront esophagectomy found that 45.7% were pathologically upstaged compared with those in the nCRT group [14]. Among upstaged patients treated with upfront surgery, only 44.2% received adjuvant therapy and experienced poorer survival compared to those treated with nCRT (p<0.001), underscoring the importance of selecting patients at risk of systemic disease for chemotherapy. Similarly, data from the Taiwan Cancer Registry indicated that 43.17% of patients with cT3N0 esophageal cancer (predominantly ESCC) who underwent upfront surgery experienced pathologic upstaging [10]. The application of nCRT before esophagectomy significantly improved survival in cT3N0 ESCC patients; however, the study lacked data on multidisciplinary evaluation before treatment.

Treatment decisions must balance the potential morbidity and quality-of-life effects associated with neoadjuvant therapy. Neoadjuvant chemoradiotherapy for esophageal cancer has been shown to temporarily reduce multiple aspects of health-related quality of life [15]. In clinical practice, some patients may become inoperable during preoperative chemoradiotherapy due to toxicity, clinical deterioration, or disease progression. Indeed, literature reports indicate that up to 22.9% of patients do not proceed to surgical resection, primarily due to disease progression or chemoradiation-related mortality [13,16,17]. A randomized trial (Japan Clinical Oncology Group Study, JCOG 9907) comparing postoperative adjuvant chemotherapy and preoperative chemotherapy for locally advanced ESCC demonstrated superior overall survival with preoperative chemotherapy, without additional serious adverse events [18]. However, subgroup analysis revealed nearly identical outcomes for patients stratified by stage, such as cT3 or cN0 disease, indicating the need for more effective treatment strategies. Carefully selecting postoperative adjuvant therapy based on pathologic findings is advantageous, regardless of the accuracy of preoperative nodal assessment. The JCOG 9204 trial also supported the survival benefit of adjuvant therapy after esophagectomy, although not for patients with pathologic node-negative disease (pN0) [19].

Accurate diagnosis and staging are essential to identify appropriate candidates for chemoradiation and to spare low- risk patients from unnecessary treatment. Beyond routine PET/CT and EUS staging, multidisciplinary tumor boards play a crucial role in improving diagnostic accuracy. Studies have shown that multidisciplinary discussions can alter diagnoses in 18.4%–26.9% of patients with gastrointestinal malignancies [20], offering greater accuracy than individual physician assessment. In our cohort, the relatively low upstaging rate (24.1%) compared with prior studies may be partly attributed to consensus-based multidisciplinary diagnosis in patients with cT2–3N0 ESCC.

Limitations

This study has several limitations. First, it was limited by its retrospective design, despite being based on a prospectively maintained database. The small number of patients with this clinical stage from a single institution may have introduced selection bias, which may persist even after propensity score matching. Second, not all patients with locally advanced ESCC underwent EUS before therapy, leading to variation in clinical staging modalities. This may have resulted in nodal understaging and affected radiation planning accuracy. Additionally, interobserver variability among technicians performing ultrasonography could have influenced tumor depth assessment. Third, surgical techniques and treatment strategies (upfront surgery or nCRT) varied according to the surgeon’s preference, though most procedures were transthoracic esophagectomies (McKeown approach) performed using minimally invasive methods. Finally, this study focused solely on squamous cell carcinoma, excluding adenocarcinoma. Future multicenter studies or randomized controlled trials with larger, more diverse populations are needed to validate these findings and provide further insight into clinically node-negative esophageal cancer.

Conclusions

Compared with neoadjuvant therapy followed by surgery, upfront esophagectomy in patients with cT2–3N0 ESCC yielded comparable survival outcomes and may be considered as an initial treatment for clinically node-negative patients. The addition of chemoradiation before surgical resection was associated with a higher downstaging rate and favorable pathologic response.