Real-world applications of postneoadjuvant stage I of the staging system for oesophageal squamous cell cancer: what is beyond a pathological complete response?

Abstract

OBJECTIVES

The 8th edition of the American Joint Committee on Cancer classification system defines the postneoadjuvant pathological T0-2N0M0 (ypT0-2N0M0) stage as ypI for oesophageal cancer. This study was designed to explore the possible differences in survival among patients with yp stage I.

METHODS

The records of patients with oesophageal squamous cell cancer staged ypI who underwent a radical oesophagectomy between July 2009 and July 2022 at the Sun Yat-sen University Cancer Center were reviewed. The survival differences were compared between ypT0, ypTis, ypT1a, ypT1b and ypT2.

RESULTS

A total of 355 patients (65 deaths and 34 recurrences) were reviewed. Overall survival (OS) and disease-free survival (DFS) differed significantly between patients in the total population staged as ypT0-1a and ypT1b-2 [5-year OS (5-OS): 81.5% vs 70.2%, P = 0.043; 5-year DFS (5-DFS): 80.8% vs 71.9%, P = 0.016) in the neoadjuvant chemoradiotherapy (NCRT) group (5-OS: 82.0% vs 67.5%, P = 0.032; 5-DFS: 84.6% vs 71.2%, P = 0.027), but not in the neoadjuvant chemotherapy±I group (5-OS: 80.3% vs 72.9%, P = 0.34; 5-DFS: 66.5% vs 72.4%, P = 0.41). The OS and DFS of patients with ypT0 stage (pathological complete response) disease were similar to those of patients with ypTis-1a disease in the total population (P = 0.16 and 0.94) and the NCRT group (P = 0.40 and 0.57). Cox regression analysis revealed that ypT1b-2 was an independent prognostic factor for survival in the total population (hazard ratio = 1.760, 95% confidence interval = 1.056–2.933; P = 0.030).

CONCLUSIONS

Patients with ypIa-new (ypT0-1aN0M0) disease had significantly better survival than those with ypIb-new (ypT1b-2N0M0) disease among patients who received NCRT. Patients with ypIa-new (pathological complete response and ypTis-1aN0M0) disease had similar survival data. Adjuvant immunotherapy might not be required for patients with ypTis-1aN0M0 stage disease.

In light of the latest global cancer statistics, the incidence and mortality of oesophageal cancer are still high [1].

GRAPHICAL ABSTRACT

INTRODUCTION

In light of the latest global cancer statistics, the incidence and mortality of oesophageal cancer are still high [1]. Neoadjuvant chemoradiotherapy (NCRT) is recommended for locally advanced oesophageal cancer based on strong evidence from the NEOCRTEC5010 (Neoadjuvant Chemoradiotherapy Followed by Surgery versus Surgery Alone for Locally Advanced Oesophageal Squamous Cell Carcinoma) trial [2] and the CROSS (Chemoradiotherapy for Oesophageal Cancer Followed by Surgery Study) trial [3]. Immunotherapy is gradually gaining importance as a neoadjuvant therapy, although the available data are immature [4–6]. The pathological stage after neoadjuvant therapy (ypTNM) has completely different prognostic significance and was independently classified according to the 8th edition of the American Joint Committee on Cancer/Union for International Cancer Control/Union for International Cancer Control (AJCC/UICC) classification system [7].

Recently, researchers have reported the application of the ypTNM stage with proposals and revisions for the modification of new staging systems, mostly concerning the poor prognosis [8–10]. Because the 8th edition staging system of the AJCC/UICC defines all ypT0-2N0M0 tumours as yp stage I, it seems inappropriate that pathological complete response (PCR) was also included in the yp stage I. A subsequent analysis of the NEOCRTEC-5010 trial showed that PCR after NCRT was an important prognostic indicator of overall survival (OS) and disease-free survival (DFS) in patients with oesophageal squamous cell cancer. In addition, the CROSS trial performed by van Hagen reported similar results in patients with oesophageal adenocarcinoma [11]. After neoadjuvant chemoradiotherapy followed by an operation, non-PCR patients with oesophageal cancer can obtain improved survival from maintaining immunotherapy regardless of whether they have squamous carcinoma or adenocarcinoma, as reported by the CheckMate-577 trial (NCT02743494) [12] and afterwards recommended by the National Comprehensive Cancer Network guidelines [13]. In clinical practice, physicians and surgeons tend to choose adjuvant strategies based on both tumour staging before treatment and postoperative pathological results. In particular, the situation is more complicated in patients with yp stage I disease, which means that the risk–benefit balance should be carefully evaluated.

Thus, our study was designed to explore the possible difference in survival according to yp stage I among patients with oesophageal squamous cell cancer who received neoadjuvant therapy followed by an operation. This study investigated the separation of better survival in yp stage I to provide preliminary evidence for revision of the 8th edition AJCC/UICC staging system.

METHODS

Patients

Data between July 2009 and July 2022 were acquired from an oesophageal cancer-specific database within the big-data intelligence framework at the Sun Yat-sen University Cancer Center (Supplementary Material, Fig. S1); the data of 355 consecutive patients with pathologically staged ypI oesophageal cancer who underwent radical oesophagectomy were reviewed. All patients met the following inclusion criteria: (i) diagnosed histologically with oesophageal squamous cell cancer; (ii) underwent radical oesophagectomy with R0 resection; (iii) had a pathological stage of ypI according to the 8th edition staging of the AJCC/UICC classification system; and (iv) had complete and retrievable clinical data. Patients with ypT0N0 (PCR), ypT1aN0, ypT1bN0 and ypT2N0 were compared to investigate the beginning of survival difference. This study was approved by the institutional review board of the Sun Yat-sen University Cancer Center. The requirement for patient consent was waived due to the retrospective design. The registration number in the Chinese Clinical Trial Registry is ChiCTR2400082409.

Pretreatment workup

The pretreatment workup included a physical examination, routine laboratory tests, X-ray examination, plain and enhanced chest and abdominal computed tomography, pulmonary function tests, electrocardiogram, oesophagogastroduodenoscopy with biopsies, endoscopic ultrasonography and cervical ultrasonography. If indicated, electronic bronchoscopy was performed to exclude tumour infiltration into the trachea or the bronchial tree. Positron emission tomography–computed tomography was performed to rule out distant metastasis, if indicated. Clinical tumour staging was classified according to the 8th edition staging of the AJCC/UICC classification system.

Surgery

All patients underwent open or minimally invasive oesophagectomy with a 2-field or a 3-field lymphadenectomy. At least 15 lymph nodes were harvested during the operation for accurate pathological staging. Planned surgery, salvage surgery or conversion surgery were all included in this study.

Pathology

All paraffin-processed specimens were reviewed according to the 8th edition of AJCC/UICC staging classification [7]. The final pathological diagnosis was confirmed independently by 2 senior pathologists who specialized in oesophageal cancer to ensure diagnostic accuracy and consistency. Each slide underwent a comprehensive evaluation including histological examination and immunohistochemical staining.

Follow-up

Patients were followed up every 3 months during the first year, every 6 months during the next 2 years and annually thereafter. The follow-up assessments performed at each visit included a physical examination, routine laboratory tests, X-ray examination, plain and enhanced chest and abdominal computed tomography, cervical and abdominal ultrasonography and oesophagogastroduodenoscopy.

Statistical analyses

The cut-off date for censored data analysis was 1 April 2024. The OS was calculated as the time from the date of the operation to the date of death, with patients who were still alive being censored at the time of the last follow-up. The DFS was calculated as the time from the date of the operation to the date of recurrence or metastasis or death, with patients who were still alive or who had recurrence or metastasis being censored at the time of the last follow-up. Categorical variables were compared using the χ² test when expected cell frequencies were ≥5; otherwise, Fisher’s exact test was applied. Continuous variables were assessed for normality using the Shapiro–Wilk test. Parametric data (normally distributed) were compared using Student’s t-test, whereas nonparametric data were analysed with the Mann–Whitney U-test. The Kaplan–Meier method was used to calculate overall survival, which was compared using the log-rank test. Univariable analyses with the Cox regression model were used to assess the effect of different factors on survival. The cut-off point of age was selected based on the median age of the study population. Variables (P < 0.2) in univariable analyses were selected for multivariable Cox regression. All the statistical analyses were performed using SPSS 25.0 software (SPSS, Chicago, IL, USA), and a 2-sided P value less than 0.05 was used to indicate statistical significance. The flow chart is shown in Fig. 1.

Figure 1:

Flow chart of the study.

RESULTS

Baseline characteristics

Between July 2009 and July 2022, a total of 355 patients (65 deaths and 34 recurrences) with oesophageal squamous cell cancer staged ypI (ypT0-2N0M0) were enrolled in this study; these patients included 190 patients who received NCRT, 102 patients who received neoadjuvant chemotherapy (NCT) and 63 patients who received neoadjuvant immunochemotherapy (NICT). The median follow-up time was 35 months. Because of the complexity of a real-world study conducted over 13 years, we classified the chemotherapy regimen into 3 categories for analysis. We compared the baseline characteristics between ypT0-1a and ypT1b-2 in 3 subgroups (Table 1) and only found significant differences in the chemotherapy regimen among patients who received NCT (P < 0.05). There were 5 patients with stage IVb disease in the NCRT group, 9 patients with stage IVb disease in the NCT group and 19 patients with stage IVb disease in the NICT group. All these patients, who were postoperatively staged at ypI, had supraclavicular lymph node metastases and received 3-field lymph node dissection. The PCR rates were 45.3% in the NCRT group and 29.1% in the NCT + NICT group.

Table 1:

Baseline characteristics

Characteristics	NCRT		P-value	NCT		P-value	NICT		P-value
	ypT0-1a, No. (%) (n = 105)	ypT1b-2, No. (%) (n = 85)		ypT0-1a, No. (%) (n = 46)	ypT1b-2, No. (%) (n = 56)		ypT0-1a, No. (%) (n = 37)	ypT1b-2, No. (%) (n = 26)
Agea (range)	58 (40–73)	59 (42–70)	0.962	59 (44–72)	59 (33–86)	0.699	59 (29–72)	59 (44–69)	0.978
Sex
Male	86 (81.9)	70 (82.3)	0.936	32 (69.6)	46 (82.1)	0.136	31 (83.8)	17 (65.4)	0.091
Female	19 (18.1)	15 (17.6)		14 (30.4)	10 (17.9)		6 (16.2)	9 (34.6)
Tumour location
Upper	8 (7.6)	13 (15.3)	0.234	3 (6.5)	9 (16.1)	0.322b	0	2 (7.7)	0.094b
Middle	68 (64.8)	52 (61.2)		32 (69.6)	34 (60.7)		22 (59.5)	18 (69.2)
Lower	29 (27.6)	20 (23.5)		11 (23.9)	13 (23.2)		15 (40.5)	6 (23.1)
Histologic grade
G1	2 (1.9)	2 (2.4)	0.175b	0	3 (5.4)	0.389b	1 (2.7)	0	0.881b
G2	73 (69.5)	48 (56.5)		27 (58.7)	31 (55.4)		21 (56.8)	14 (53.8)
G3	30 (28.6)	35 (41.2)		19 (41.3)	22 (39.3)		15 (40.5)	12 (46.2)
Conduits of reconstruction
Gastric tube	104 (99.0)	85 (100)	0.503b	46 (100)	55 (98.2)	1.000b	37 (100)	26 (100)	-
Colon	1 (1.0)	0		0	1 (1.8)		0	0
Surgical approaches
McKeown style	93 (88.6)	78 (91.8)	0.530b	37 (80.4)	34 (60.7)	0.051b	29 (78.4)	23 (88.5)	0.390b
Ivor Lewis style	6 (5.7)	2 (2.4)		7 (15.2)	12 (21.4)		4 (10.8)	0
Sweet style	3 (2.9)	4 (4.7)		2 (4.3)	10 (17.9)		3 (8.1)	2 (7.7)
Transhiatal	3 (2.9)	1 (1.2)		0	0		1 (2.7)	1 (3.8)
ECOG score
0	1 (1.0)	0	1.000b	1 (2.2)	0	0.451b	37 (100.0)	26 (100.0)	-
1	104 (99.0)	85 (100.0)		45 (97.8)	56 (100)		0	0
Hypertension
No	89 (84.8)	75 (88.2)	0.489	35 (76.1)	45 (80.4)	0.602	29 (78.4)	19 (73.1)	0.627
Yes	16 (15.2)	10 (11.8)		11 (23.9)	11 (19.6)		8 (21.6)	7 (26.9)
Heart disease
No	102 (97.1)	85 (100)	0.254b	44 (95.7)	55 (98.2)	0.587b	37 (100)	26 (100)	-
Yes	3 (2.9)	0		2 (4.3)	1 (1.8)		0	0
Diabetes
No	97 (92.4)	77 (90.6)	0.840	41 (89.1)	52 (92.9)	0.728b	32 (86.5)	23 (88.5)	1.000b
Yes	8 (7.6)	8 (9.4)		5 (10.9)	4 (7.1)		5 (13.5)	3 (11.5)
Smoking history
No	45 (42.9)	37 (43.5)	0.926	23 (50.0)	31 (55.4)	0.590	20 (54.1)	16 (61.5)	0.555
Yes	60 (57.1)	48 (56.5)		23 (50.0)	25 (44.6)		17 (45.9)	10 (38.5)
Drinking history
No	54 (51.4)	55 (64.7)	0.066	33 (71.7)	36 (64.3)	0.423	23 (62.2)	16 (61.5)	0.960
Yes	51 (48.6)	30 (35.3)		13 (28.3)	20 (35.7)		14 (37.8)	10 (38.5)
Clinical stages
I	2 (1.9)	2 (2.4)	0.934b	4 (8.7)	5 (8.9)	0.804b	5 (13.5)	4 (15.4)	0.913b
II	21 (20.0)	14 (16.5)		12 (26.1)	15 (26.8)		11 (29.7)	5 (19.2)
III	58 (55.2)	48 (56.5)		17 (37.0)	25 (44.6)		15 (40.5)	13 (50.0)
IVa	22 (21.0)	18 (21.2)		9 (19.6)	6 (10.7)		3 (8.1)	2 (7.7)
IVb	2 (1.9)	3 (3.5)		4 (8.7)	5 (8.9)		3 (8.1)	2 (7.7)
Chemotherapy regimen
Antimicrotubule+platinum	86 (81.9)	64 (75.3)	0.266	22 (47.8)	36 (64.3)	0.000b	14 (37.8)	13 (50.0)	0.509
Antimetabolites+platinum	19 (18.1)	21 (24.7)		11 (23.9)	19 (33.9)		15 (40.5)	7 (26.9)
Antimicrotubule+ antimetabolites	0	0		13 (28.3)	1 (1.8)		8 (21.6)	6 (23.1)
Adjuvant therapy
None	94 (89.5)	73 (85.9)	0.146b	37 (80.4)	39 (69.6)	0.448b	28 (75.7)	20 (76.9)	1.000b
Chemotherapy	4 (3.8)	5 (5.9)		6 (13.0)	13 (23.2)		1 (2.7)	0
Immunochemotherapy	2 (1.9)	6 (7.1)		3 (6.5)	3 (5.4)		8 (21.6)	6 (23.1)
Radiotherapy	5 (4.8)	1 (1.2)		0	0		0	0

Comparisons were performed using the χ² test or Student’s t-test, except

^ausing Mann–Whitney U-test;

^busing Fisher’s exact test.

ECOG: Eastern Cooperative Oncology Group; NCRT: neoadjuvant chemoradiotherapy; NCT: neoadjuvant chemotherapy; NICT: neoadjuvant immunochemotherapy.

Survival analyses

The Kaplan–Meier curves and log-rank tests were conducted for all yp stage I (Supplementary Material, Fig. S2). The curves indicated an obvious decline starting at the ypT1b stage. Thus, ypT0-1a stages were grouped together and compared to the ypT1b-2 stages. In the total population, significant differences were found in OS and DFS between the ypT0-1a and ypT1b-2 stages (Fig. 2; P = 0.043 for OS and P = 0.016 for DFS). For patients classified as ypT0-1a, the 5-OS was 81.5% (95% CI: 90.0–73.8%), whereas the 5-DFS was 80.8% (95% CI: 89.8–72.7%). Conversely, for those with the ypT1b-2 stage, the 5-OS was 70.2% (95% CI: 79.5–61.9%), and the 5-DFS was 71.9% (95% CI: 80.8–64.0%). In subgroup analyses, significant differences were found in OS and DFS between the ypT0-1a and ypT1b-2 stages in the NCRT group (Fig. 3; P = 0.032 and 0.027) but not in the NCT±I group (P = 0.34 and 0.41). Further analyses indicated that patients with the ypT0 stage (PCR) disease had similar OS and DFS outcomes compared to those with ypTis-1a disease (Fig. 4; P > 0.05).

Figure 2:

Overall survival and disease-free survival in patients with staged ypT0-1a and ypT1b-2 disease.

Figure 3:

Subgroup analysis in overall survival and disease-free survival. (A) Patients who received neoadjuvant chemoradiotherapy; (B) patients who received neoadjuvant chemotherapy ± I).

Figure 4:

Overall survival and disease-free survival in patients with staged ypT0 and ypTis-1a disease. (A) Total population. (B) Patients who received neoadjuvant chemoradiotherapy.

Cox regression analyses

Univariable Cox regression showed that only ypT1b-2 was a prognostic factor for OS. The value of the hazard ratio (HR) starts to be greater than 1 from ypT1b. Variables (P < 0.2) in univariable analyses, including pathological stages, age and drinking history, were selected for multivariable Cox regression (Table 2). The results showed that ypT1b-2 was an independent prognostic factor for OS (HR = 1.760, 95% CI = 1.056–2.933; P = 0.030).

Table 2:

Cox regression analyses of overall survival

Variables	Univariable analyses			Multivariable analyses
	HR	95% CI	P-value	HR	95% CI	P-value
Sex (female vs male)	1.239	0.647–2.375	0.518
Age (≤60 vs >60)	1.507	0.919–2.471	0.104	1.533	0.934–2.517	0.091
Location
Upper	Ref
Middle	0.642	0.311–1.327	0.232
Lower	0.716	0.316–1.624	0.424
Histologic grade
G1	Ref
G2	1.765	0.241–12.92	0.576
G3	2.803	0.382–20.57	0.311
Hypertension	1.095	0.556–2.157	0.792
Diabetes	1.114	0.481–2.582	0.801
Smoking history	0.860	0.529–1.400	0.544
Drinking history	1.488	0.914–2.421	0.110	1.538	0.941–2.515	0.086
Clinical stage (I & II & III vs Iva & IVb)	1.168	0.655–2.084	0.598
Neoadjuvant therapy
NCRT	Ref
NCT	1.001	0.583–1.717	0.998
NCT+I	0.930	0.387–2.234	0.871
Chemotherapy regimen
Antimicrotubule+platinum	Ref
Antimetabolites+platinum	1.222	0.705–2.120	0.475
Antimicrotubule+antimetabolites	0.701	0.168–2.929	0.626
Adjuvant therapy
Chemotherapy	1.405	0.764–2.584	0.274
Immunotherapy	1.784	0.634–5.023	0.273
Radiotherapy	2.540	0.619–10.418	0.196	2.917	0.694–12.258	0.144
Pathological stages
ypT0	Ref
ypTis	0.249	0.033–1.861	0.176
ypT1a	0.709	0.211–2.389	0.579
ypT1b	1.195	0.603–2.369	0.609
ypT2	1.578	0.883–2.820	0.124
(ypT0-1a vs ypT1b-2)	1.674	1.010–2.773	0.046	1.760	1.056–2.933	0.030
(ypT0 vs ypTis-1a)	0.472	0.161–1.386	0.172

HR: hazard ratio; Ref: reference.

DISCUSSION

Several years have passed since the 8th edition of the AJCC/UICC classification system first became widely applied in the staging of oesophageal cancer with reference to a worldwide database [14]. Trimodal therapy for locally advanced oesophageal cancer has become a standard treatment. The ypTNM staging system provides a more accurate prognostic prediction than the pTNM staging system [15]. However, because neoadjuvant therapy is generally applied, many issues in the ypTNM staging classification system have been revealed from real-world studies [8–10, 16]. It seems inappropriate to classify both PCR and ypTis-2N0M0 as yp stage I according to the 8th edition staging of the AJCC/UICC classification system [7, 17]. To date, many studies have explored the efficacy of immunotherapy in combination with neoadjuvant therapy [4, 5, 18]. Given the need to adapt the ypTNM staging classification system to the progress of neoadjuvant therapy, this study was conducted with the objective of obtaining preliminary evidence for revision of the yp stage I.

To facilitate the analyses of the substantial differences in survival associated with patients in the yp stage I, we defined ypT0-1aN0M0 as the ypIa-new stage and ypT1b-2N0M0 as the ypIb-new stage. The OS and DFS between patients with ypIa-new and ypIb-new disease were significantly different in the total population (P = 0.043 and 0.016) and among those who received NCRT (P = 0.032 and 0.027), but not in the NCT±I group (P = 0.34 and 0.41). Regarding the ypIa-new stage, patients with ypT0N0M0/PCR stage disease had OS and DFS values similar to those of patients with ypTis-1aN0M0 stage disease (NCRT group: P = 0.40 and 0.57; total population: P = 0.16 and 0.94).

The discrepancy between NCRT and NCT±I groups may stem from fundamental differences in therapeutic mechanisms. NCRT combines radiotherapy—which enhances local tumour control and amplifies the prognostic value of residual tumour depth (ypT)—with chemotherapy, whereas NCT±I primarily exerts systemic effects that may decouple survival outcomes from yp T stage among yp stage I. Notably, despite these mechanistic distinctions, the results comparing OS and DFS of different neoadjuvant therapies showed no significant difference for yp stage I (Supplementary Material, Fig. S3).

Our study results indicated that patients with ypIa-new stage disease could be considered oncologically equivalent to PCR in status. Patients who achieved PCR had better oncological outcomes than non-PCR patients did, as was proven in the NEOCRTEC5010 and CROSS trials [11, 19]. PCR is an essential prognostic factor for evaluating the tumour response to neoadjuvant therapy in clinical trials and has even been reported as a primary outcome [18]. Accurate diagnosis via PCR is crucial for identifying adjuvant therapy strategies but has led to difficulties in pathological evaluation [20].

For NCRT, the ypIa-new and ypIb-new staging classifications showed better prognostic value than the ypI staging classification. The ypIa-new stage, which includes PCR and the ypTis-1aN0M0 stage, provides additional clinical value for evaluating tumour response. The application of immunotherapy in combination with neoadjuvant therapy has shown positive results in the NEOCRTEC1901 [4], NICE (Neoadjuvant Immunochemotherapy for Oesophageal Squamous Cell Carcinoma) [21], TD-NICE [22] and other clinical trials [23]. However, the question of whether immunotherapy is effective as an adjuvant therapy for patients who have been operated on requires additional evidence. Under these circumstances, a new ypTNM classification system adapted for the progression of neoadjuvant therapy is warranted.

After neoadjuvant therapy, residual tumours mostly occur in the mucosal and/or submucosal layers [24, 25]. The most widely recognized tumour regression grading systems published by Chirieac [26] and the National Comprehensive Cancer Network [13] both categorize PCR as distinct on their own. However, patients with tiny subtle residual tumours in the mucosa layer are occasionally observed in clinical practice. In some situations, it is difficult to distinguish between ypTis and ypT1a because of severe inflammation after neoadjuvant therapy. Residual tumour cells may exhibit therapy-induced nuclear pleomorphism and increased cytoplasmic eosinophilia, vacuolation and degenerative changes [27]. Additionally, squamous metaplasia of the oesophageal submucosal glands or ducts with atypical nuclei may be observed, which mimics residual carcinoma [20]. These changes increase the difficulty of staging tumours. Thus, the ypIa-new staging system, which includes both PCR and ypTis-1aN0M0, has more practical value in this situation.

For non-PCR patients, evidence for adjuvant chemotherapy is insufficient, but adjuvant immunotherapy is recommended [12, 28]. In the CheckMate-577 trial, 13% of the non-PCR patients in whom grade 3 or 4 adverse events were reported had received nivolumab adjuvant therapy, and 9% of the patients discontinued immunotherapy [12]. In our study, only 1 patient with stage ypTis-1aN0M0 disease received adjuvant immunotherapy. The results showed that patients with ypT0N0M0 stage disease had survival rates similar to those with ypTis-1aN0M0 stage disease, which indicated that patients with ypTis-1aN0M0 stage disease might be exempted from adjuvant therapy to avoid high costs and adverse effects.

Although the yp staging system primarily reflects a pathological response to neoadjuvant therapy, our findings also suggested that its prognostic implications are not entirely independent of the initial clinical stage. For instance, patients with clinical stage IVA and ypT0-1a disease had numerically poorer survival outcomes compared to those with clinical stage II and ypT1b-2 disease (Supplementary Material, Fig. S4). Although this difference did not reach statistical significance due to the limited sample size, this observation underscores the potential influence of baseline tumour biology on long-term outcomes, even among patients with favourable pathological responses. Minor metastases might already occur in cases with stage IVA/IVB disease. In other cancers, such as lung cancer, the circulating tumour DNA has shown significant value in detecting recurrence. However, the prognostic biomarkers for oesophageal cancer remain unclear. To address how to truly screen patients who actually need adjuvant therapy, we are now conducting a clinical trial to detect minimal residual disease by individual circulating tumour DNA analysis. Moving forward, our goal is to re-define adjuvant treatment strategies by integrating multi-omics biomarkers to complement conventional TNM staging.

The patients included in our study included 190 patients who received NCRT, 102 patients who received NICT and 63 patients who received NCT; the data of these patients came primarily from clinical trials, such as the NEOCRTEC5010 trial [2], the NEOCRTEC1601 trial [29], the NEOCRTEC2001 trial (ongoing) and the NEOCRTEC308 trial (ongoing). Thus, our data represent the Chinese population well because of the objectivity of the study. Given that NCRT is the standard treatment for locally advanced oesophageal cancer, several possible reasons for patients not receiving NCRT in our study were as follows: (i) borderline resectable oesophageal cancer; (ii) contraindications to radiotherapy, such as oesophageal perforation; (iii) insufficient function of the heart or respiratory system that improved preoperatively; (iv) advanced oesophageal cancer; and (v) refusal to receive radiotherapy.

This study has several limitations. First, as a single-centre retrospective analysis, our findings may be influenced by institutional selection bias and treatment protocols. Second, the sample size was small, implying that the results need further external validation. The conclusions of this study need to be verified by analysis of other databases. Third, the median follow-up of 35 months may be insufficient to capture long-term recurrence patterns, especially for early-stage disease. Although our data highlight the prognostic heterogeneity within yp stage I, these results should be interpreted as hypothesis-generating, requiring validation in larger, prospective cohorts.

In conclusion, the current 8th edition AJCC yp stage I criteria for oesophageal squamous cell cancer might need potential refinement. Patients with ypIa-new (ypT0-1aN0M0) disease had different survival results than those with ypIb-new (ypT1b-2N0M0) disease among patients who received NCRT, while patients with ypIa-new (PCR and ypTis-1aN0M0) disease had similar survival results. The results implied that patients with ypTis-1aN0M0 stage disease might be exempted from receiving adjuvant immunotherapy, providing a potential landscape shifter for trimodal treatment of oesophageal cancer. However, further larger multicentre studies are warranted.

Glossary

ABBREVIATIONS

AJCC/UICC

American Joint Committee on Cancer/Union for International Cancer Control

DFS

Disease-free survival

NCRT

Neoadjuvant chemoradiotherapy

NCT

Neoadjuvant chemotherapy

NICT

Neoadjuvant immunochemotherapy

OS

Overall survival

PCR

Pathological complete response

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

FUNDING

This study was supported in part by grants from the National Natural Science Foundation of China [grant number 82272881].

Conflict of interest: The authors have no relevant financial or nonfinancial interests to disclose.

DATA AVAILABILITY

The data were accessed from the database at the Sun Yat-Sen University Cancer Center. The derived data generated in this study are available from the corresponding author upon reasonable request.

Author contributions

Yan Huang: Conceptualization; Data curation; Investigation; Methodology; Writing—original draft; Writing—review & editing. Zhichao Li: Data curation; Investigation. Jiadi Wu: Investigation; Methodology. Sheng Huang: Data curation; Investigation. Caiyan Fang: Data curation; Investigation. Kaiwen Li: Data curation; Investigation. Jiyang Chen: Investigation. Huilin Su: Data curation. Hong Yang: Conceptualization; Methodology; Writing—review & editing. Kongjia Luo: Conceptualization; Investigation; Methodology; Project administration; Writing—original draft; Writing—review & editing. Jianhua Fu: Methodology; Project administration; Resources; Supervision; Writing—review & editing

Reviewer information

The European Journal of Cardio-Thoracic Surgery thanks Hasan Fevzi Batirel, Frank A. Baciewicz Jr. and the other, anonymous reviewer(s) for their contributions to the peer review process of this article.