Segment-specific patterns of regional lymph-node metastasis in primary small-cell carcinoma of the esophagus: A systematic review and meta-analysis

Abstract

BACKGROUND

Primary small-cell carcinoma of the esophagus (SCCE) is a rare, biologically aggressive malignancy with a marked propensity for lymphatic dissemination. We estimated the prevalence of pathologically confirmed lymph-node metastasis in surgically staged SCCE and mapped the available segment-specific evidence at the nodal-station level.

AIM

To estimate the patient-level prevalence of pathologic lymph-node metastasis in surgically staged SCCE and to map the available segment-specific evidence at the station level.

METHODS

We conducted a PRISMA-informed exploratory systematic review and meta-analysis of a deduplicated evidence set comprising 99 SCCE-related records. Records were classified as direct nodal-station evidence, cohort-level N-status or prognostic evidence, or contextual case-based/background evidence. Surgically staged cohorts reporting SCCE-specific pathologic nodal positivity, including the CHiSCEC multicenter preprint, were pooled using logit-transformed proportions and a random-effects model. Because station denominators, nodal maps, and reporting definitions were insufficiently compatible for station-by-segment pooling, station-level findings were summarized as an evidence map.

RESULTS

Four surgical cohorts with compatible pathologic nodal denominators included 579 patients, of whom 352 were node-positive. The pooled prevalence of pathologic lymph-node metastasis was 61.1% (95%CI: 51.4% to 70.0%; I² = 76.4%; τ² = 0.113), indicating frequent nodal involvement and substantial between-study heterogeneity.

CONCLUSION

SCCE is associated with a high regional nodal burden, with a pooled prevalence of pathologic nodal metastasis of approximately 61%. Available station-level evidence suggests predominant recurrent laryngeal nerve and upper mediastinal involvement in upper thoracic SCCE, bidirectional spread in middle thoracic SCCE, and predominant perigastric and abdominal involvement in lower thoracic SCCE.

Key Words: Small-cell carcinoma of the esophagus; Lymph-node metastasis; Nodal stations; Systematic review

Core Tip: In small-cell carcinoma of the esophagus (SCCE), lymph-node metastasis is common and anatomically widespread rather than confined to peritumoral stations. Across four surgical cohorts, the pooled prevalence of pathologic nodal metastasis was 61.1%, with a consistent node-level metastatic ratio of about 11%-13%. Station-level evidence shows segment-specific but bidirectional spread involving cervical, mediastinal, and abdominal compartments, with frequent recurrent laryngeal nerve and perigastric node involvement and non-negligible skip metastasis. These findings highlight that SCCE should be understood as a highly lymphotropic disease and support broader, segment-adapted nodal assessment and clinical target-volume design rather than limited local field assumptions.

INTRODUCTION

Primary small-cell carcinoma of the esophagus (SCCE) is rare, yet its clinical significance is disproportionate to its incidence. Early pathologic and clinicopathologic series have established SCCE as a distinct, high-grade esophageal malignancy marked by rapid progression and poor survival[1-3]. More recent genomic and multi-omics studies have identified molecular features consistent with small-cell biology, including disruption of RB1-related pathways and other high-grade neuroendocrine programs[4,5]. These features may explain why SCCE exhibits the longitudinal lymphatic spread of esophageal cancer and the early lymphatic and systemic dissemination characteristic of small-cell carcinoma.

For locoregional management, the clinically relevant question is not merely whether nodal metastasis is present but where it occurs. Large esophageal nodal mapping studies have observed that tumor location influences both the distribution of lymph nodes and the potential therapeutic value of lymph node dissection, underscoring the importance of station-level anatomy even before SCCE-specific evidence is considered[6,7]. In SCCE, metastasis to the right recurrent laryngeal nerve, subcarinal, lower paraesophageal, perigastric, and celiac nodes has different implications for operative exposure, radiotherapy field design, toxicity tradeoffs, and recurrence surveillance. Broad categories such as limited stage, extensive stage, or clinical N status may therefore obscure the anatomic detail needed for multidisciplinary planning.

The SCCE literature comprises several therapeutic and prognostic reports but few datasets with pathologic nodal information at the station level. The most directly relevant evidence comes from resected or station-oriented SCCE cohorts that quantify nodal positivity, recurrence, or segment-specific nodal patterns[8-10]. In addition, limited-stage Chinese series, risk-factor analyses, and the CHiSCEC (Multicenter Retrospective Trial from China on Small-cell Carcinoma of the Esophagus) multicenter preprint have provided complementary evidence regarding nodal risk and implications for radiotherapy target-volume delineation[11-13]. Accordingly, this review treated patient-level nodal positivity and station-level distribution as separate analytic questions: The former can be pooled across compatible surgical cohorts, whereas the latter should be summarized as an evidence map until station denominators and nodal definitions are harmonized.

This systematic review addressed two related questions: (1) What is the pooled prevalence of pathologic lymph node metastasis among surgically staged patients with SCCE? and (2) What station-level patterns are reported for upper, middle, and lower thoracic SCCE? The study aimed to provide a clinically useful synthesis of appropriately matched evidence while avoiding unsupported claims about station-specific survival or efficacy indices.

MATERIALS AND METHODS

Design and objectives

A PRISMA-informed exploratory systematic review and meta-analysis of regional lymph node metastasis in SCCE was conducted. The primary quantitative outcome was the patient-level prevalence of pathologic nodal positivity among surgically staged patients, and the secondary objective was to map segment-specific involvement by nodal station or compartment.

Although the evidence base was extensive, the main-text bibliography was intentionally limited to 55 core references. Records were retained if they directly informed SCCE nodal or station-level evidence, provided extractable pathologic nodal denominators, contributed clinically relevant therapeutic or prognostic context, furnished molecular or biomarker information, or clarified esophageal nodal anatomy (Figure 1). The review protocol was registered in PROSPERO (CRD420261382730) (Supplementary Table 1).

Figure 1 PRISMA flow diagram for literature search and study selection. SCCE: Small cell carcinoma of the esophagus.

Search concepts and eligibility

Search terms encompassed primary SCCE, small-cell esophageal carcinoma, esophageal small-cell neuroendocrine carcinoma, lymph-node metastasis, nodal recurrence, recurrent laryngeal nerve lymph nodes, perigastric, mediastinal, abdominal, nodal station, and skip metastasis. The searches were last updated on April 27, 2026.

Eligible records described SCCE or an equivalent historical term and provided clinical, pathologic, nodal, treatment, survival, molecular, or contextual information relevant to SCCE. Studies that combined SCCE with other esophageal neuroendocrine carcinomas were retained only if SCCE-specific data were extractable or the report provided essential contextual information. Single-case reports were cited only for historical or pathobiologic context.

Data extraction and definitions

Study design, country or center, SCCE case definition, sample size, treatment setting, pathologic vs clinical nodal ascertainment, total number of dissected nodes, number of metastatic nodes, tumor segment, station or compartment definitions, recurrence-site reporting, and survival endpoints were extracted. Tumor location was harmonized to upper, middle, or lower thoracic esophagus whenever possible. Nodal stations were grouped into cervical/neck, upper mediastinal, middle mediastinal, lower mediastinal, recurrent laryngeal nerve, perigastric, and abdominal compartments.

Risk of bias was assessed across domains informed by Joanna Briggs Institute and Risk of Bias in Non-randomized Studies of Interventions principles for retrospective surgical oncology studies: Patient selection, SCCE case definition, pathologic nodal ascertainment, completeness of station reporting, treatment heterogeneity, control of confounding, and outcome reporting. Certainty was not formally graded at the station level as the available evidence was retrospective, heterogeneous, and incompletely documented.

Statistical analysis

Patient-level pathologic nodal positivity was pooled as logit-transformed proportions using an inverse-variance random-effects model, with between-study variance estimated by the DerSimonian-Laird method and 95%CIs back-transformed to the proportion scale. Between-study heterogeneity was quantified with Cochran’s Q, τ², and I². Furthermore, the unweighted aggregate proportion was reported for transparency. Sensitivity analyses excluded the CHiSCEC multicenter preprint and, separately, Xu et al[8] to assess robustness to non-peer-reviewed evidence and potential source-population overlap.

The node-level metastatic ratio, defined as the number of metastatic nodes divided by the number of dissected nodes, was summarized descriptively. As nodes are clustered within patients, individual nodes cannot be treated as independent biological observations. Therefore, a formal station-by-segment meta-analysis was not conducted owing to differences in station denominators, nodal maps, and reporting definitions across studies.

RESULTS

Evidence set and analytical hierarchy

Six retained records directly informed SCCE nodal distribution, station involvement, and metastatic patterns. Four studies provided compatible patient-level pathologic denominators and were included in the random-effects meta-analysis. Contemporary treatment and prognostic context were derived from perioperative, risk-stratification, and survival-model cohorts[14-16]; 15-year and multicenter treatment analyses[17-19]; and a multimodality surgical cohort[20].

Registry-based studies complemented institutional cohorts by describing population-level outcomes in SEER, the National Cancer Database, and related datasets[21-23]. Reviews of esophageal neuroendocrine carcinoma and population-based analyses of neuroendocrine carcinoma were retained solely for contextual interpretation and were not used for station-level inference in SCCE[24-26].

Patient-level pathologic nodal positivity

Four surgical cohorts met the pooling criteria. Wu et al[11] reported nodal positivity in 15 of 21 surgically treated patients with limited-stage disease, Xu et al[8] in 60 of 125 resected patients, Liu et al[9] in 100 of 147 patients with R0-resected thoracic SCCE, and CHiSCEC in 177 of 286 surgically treated patients[8,9,11,13].

Collectively, the four cohorts comprised 579 surgically staged patients and 352 node-positive events. The unweighted aggregate prevalence was 60.8% (352/579), and the random-effects pooled prevalence of pathologic lymph-node metastasis was 61.1% (95%CI: 51.4%-70.0%). Heterogeneity was substantial (I² = 76.4%; τ² = 0.113; Cochran’s Q = 12.71; df = 3), indicating that differences in case mix, treatment sequence, and surgical or pathologic assessment materially influenced the estimates. The pooled result should therefore be interpreted as a descriptive summary of available surgical series rather than a precise population-level estimate (Figure 2 and Table 1).

Figure 2 Forest plot of the patient-level prevalence of pathologic lymph-node metastasis in surgically staged small-cell carcinoma of the esophagus. Logit-transformed random-effects model; Q = 12.71, df = 3, τ² = 0.113, I² = 76.4%. The CHiSCEC study is a non-peer-reviewed preprint and should therefore be interpreted with caution.

Table 1 Surgical cohorts included in the patient-level pathologic lymph-node metastasis prevalence meta-analysis.

Ref.	Setting	Patients	Node-positive patients	Crude prevalence	Role in synthesis
Wu et al[11]	Limited-stage Chinese surgical series	21	15	71.4%	Small surgical denominator; directionally useful
Xu et al[8]	Retrospective resected SCCE cohort	125	60	48.0%	Patient-level and dissected-node denominator
Liu et al[9]	Detailed thoracic SCCE station-level cohort	147	100	68.0%	Main peer-reviewed station-level anchor
CHiSCEC[13]	Multicenter preprint with surgical denominator	286	177	61.9%	Largest denominator; interpreted cautiously
Random-effects synthesis	Logit-transformed proportions	579	352	61.1% pooled	95%CI: 51.4% to 70.0%; I² = 76.4%

SCCE: Small cell carcinoma of the esophagus.

In a sensitivity analysis that excluded the CHiSCEC multicenter preprint, the remaining three cohorts comprised 293 patients and 175 node-positive events. Using the same random-effects model, the pooled prevalence was 61.6% (95%CI: 45.0%-75.9%; I² = 83.8%; τ² = 0.275; Cochran’s Q = 12.33; df = 2). The crude aggregate proportion was 59.7% (175/293) and should not be interpreted as the random-effects estimate. In a second sensitivity analysis excluding Xu et al[8] because of potential source-population overlap with CHiSCEC[13], the remaining three cohorts comprised 454 patients and 292 node-positive events; the pooled prevalence was 64.3% (95%CI: 59.6%-68.8%; I² = 3.5%; τ² = 0.001; Cochran’s Q = 2.07; df = 2). Both sensitivity analyses were directionally consistent with the primary estimate.

Node-level metastatic burden

Node-level metastatic ratios were similar across studies reporting dissected-node denominators. Xu et al[8] documented 141 metastatic nodes among 1061 dissected nodes (13.3%), Liu et al[9] reported 401 of 3560 (11.3%), and CHiSCEC reported 597 of 4850 (12.3%)[8,9,13]. A smaller limited-stage surgical series recorded a comparable ratio of 12.0%[11].

Station-level and segment-specific patterns

Station-level evidence consistently argued against a narrow peritumoral model of spread. In the most detailed peer-reviewed cohort of thoracic SCCE, Liu et al[9] found involvement of cervical/neck (8.8%), upper mediastinal (27.9%), middle mediastinal (23.1%), lower mediastinal (15.6%), and abdominal (35.4%) nodes. CHiSCEC provided directionally concordant multicenter evidence, identifying high-risk stations that varied by tumor segment and with potential implications for clinical target-volume design in radiotherapy[13] (Table 2).

Table 2 Segment-specific station-level interpretation for multidisciplinary planning.

Tumor segment	Dominant signal	Supporting evidence	Practical interpretation
Upper thoracic SCCE	Upper mediastinal and recurrent laryngeal nerve involvement; CHiSCEC high-risk stations No. 105, No. 106recR, and No. 106recL	Liu et al[9]; CHiSCEC[13]; Jiao et al[10]	upper mediastinal nodal assessment should be explicit; lower mediastinal assessment remains warranted when imaging, tumor length, or pathology suggests risk
Middle thoracic SCCE	Most diffuse pattern, including upper and middle mediastinal, recurrent nerve, perigastric, and abdominal compartments; CHiSCEC high-risk stations No.107, No. 108, and abdominal No. 7	Liu et al[9]; CHiSCEC[13]; Wu et al[11]	Bidirectional nodal evaluation is more defensible than a narrow peritumoral field
Lower thoracic SCCE	Dominant lower mediastinal, perigastric, and abdominal signal; CHiSCEC high-risk stations No. 107, No. 110, abdominal No. 2, and abdominal No. 7	Liu et al[9]; CHiSCEC[13]	Upper abdominal and perigastric risk should be weighed against operative and radiotherapy morbidity; upper mediastinal skip risk remains possible
Cross-segment issue	Skip metastasis and named station risk are clinically relevant but not yet independently linked to survival	Liu et al[9]; Xu et al[8]; CHiSCEC[13]	Standardized station reporting is needed before classical efficacy-index or station-specific survival claims can be made

SCCE: Small cell carcinoma of the esophagus.

Upper thoracic SCCE showed the strongest signal in the upper mediastinum. Liu et al[9] reported prominent involvement of the upper mediastinum and the recurrent laryngeal nerve, and CHiSCEC provided exploratory evidence of relatively frequent involvement of stations No. 105, No. 106recR, and No. 106recL[9,13]. Earlier Chinese station-pattern studies yielded directionally consistent results, supporting upper mediastinal predominance in limited-stage upper thoracic disease[10,11].

Middle thoracic SCCE exhibited the broadest and most bidirectional pattern of spread. Liu et al[9] reported involvement of the upper, middle, and abdominal compartments, including the recurrent laryngeal nerve and perigastric nodes. CHiSCEC provided exploratory evidence of relatively frequent involvement of stations No. 107, No. 108, and abdominal station No. 7 in middle thoracic tumors[13].

Lower thoracic SCCE was characterized predominantly by lower mediastinal, perigastric, and abdominal involvement, although upper mediastinal spread remained possible. Liu et al[9] identified the abdominal compartment as the most commonly involved region in lower thoracic tumors, and CHiSCEC provided exploratory evidence of relatively frequent involvement of stations No. 107, No. 110, abdominal No. 2, and abdominal No. 7[9,13].

Skip metastasis, clinicopathologic correlates, and prognosis

Skip metastasis was not uncommon in the detailed station-level cohort, particularly among middle and lower thoracic tumors, and several skip events involved the abdominal compartment[9]. Available evidence does not establish skip status as an independent predictor of survival after adjusting for stage and treatment. The current significance of skip metastasis is therefore anatomic rather than prognostic: Negative peritumoral nodes do not exclude more distant regional nodal disease.

Depth of invasion, tumor length, lymphovascular invasion, and advanced stage were the most consistently reported correlates of nodal metastasis or burden. Xu et al[8] associated the depth of invasion and tumor length with lymph node metastasis; Liu et al[9] identified lymphovascular invasion as a strong correlate; and CHiSCEC reported age, T category, and tumor length as factors associated with nodal metastasis[8,9,13].

The largest station-level dataset evaluated cancer-specific and disease-free survival as well as nodal distribution. Advanced TNM stage, fewer than four chemotherapy cycles, lymphovascular invasion, advanced T category, and advanced N category were the strongest adverse factors; the M category was independently linked to disease-free survival[9]. Broader treatment cohorts likewise support the clinical relevance of nodal disease, although the independent effect of N status varies as survival is also influenced by systemic therapy, distant relapse, and small-cell biology[18-20] (Figure 3).

Figure 3  Evidence map of lymph node station-specific metastasis patterns in small cell carcinoma of the esophagus.

DISCUSSION

This review presents a deliberately conservative synthesis of a rare and clinically challenging malignancy. Across four compatible surgical cohorts, approximately three-fifths of surgically staged patients with SCCE exhibited pathologic lymph node metastasis[8,9,11,13]. Although station-level evidence was insufficiently standardized for formal pooling, available studies consistently indicated segment-dependent rather than uniform nodal spread[9,10,13].

The central methodological decision was to avoid combining incompatible evidence. Patient-level pN positivity could be pooled as events and denominators were interpretable across surgical cohorts[8,9,11,13]. By contrast, station-specific involvement could not be reliably pooled as station maps, lymphadenectomy fields, and station-specific denominators varied across reports[9,10,13]. Possible source-population overlap between Xu et al[8] and CHiSCEC[13] was therefore examined in a separate sensitivity analysis.

The findings suggested that SCCE followed the lymphatic anatomy of the esophagus but exhibited a more aggressive small-cell phenotype. Large-scale nodal mapping studies of esophageal squamous cell carcinoma illustrated the significance of tumor location and nodal station for regional treatment planning[6,7]. In SCCE, upper thoracic tumors showed recurrent laryngeal nerve and upper mediastinal involvement; middle thoracic tumors displayed bidirectional dissemination; and lower thoracic tumors demonstrated predominant lower mediastinal, perigastric, and abdominal involvement[9,11,13]. These patterns support deliberate, segment-informed nodal assessment rather than indiscriminate expansion of treatment fields.

For surgeons and pathologists, the clearest practical implication is the need for station-specific reporting. Xu et al[8] linked nodal metastasis and recurrence to post-resection outcomes, whereas Liu et al[9] revealed that station and compartment data materially affect the interpretation of nodal burden. Total and positive-node counts alone are therefore insufficient for characterizing recurrence risk or informing future target-volume design. At a minimum, operative and pathology reports should distinguish among the recurrent laryngeal; upper, middle, and lower mediastinal; perigastric; and abdominal nodal groups.

For radiation oncologists, these findings provide an anatomic framework for individualized target-volume design rather than support for a single standardized elective template. CHiSCEC offered exploratory evidence linking tumor segment to patterns of nodal involvement and potential target-delineation considerations[13]. Perioperative and multimodality cohorts, however, emphasized that local therapy must be interpreted in the context of combined treatment strategies[14,20]. For medical oncologists, the high nodal burden reinforces the importance of systemic therapy; however, treatment-response assumptions cannot substitute for anatomic staging or station-specific reporting.

The broader treatment literature supports this cautious interpretation. Surgical and radiotherapy series suggested that local treatment may benefit selected patients, but most were retrospective and lacked station-specific nodal denominators[27-29]. Additional limited-stage surgical cohorts and prognostic analyses supported multimodality management, yet these cohorts remained vulnerable to selection bias related to stage, operability, and treatment sequence[30-32]. Earlier surgical and multidisciplinary studies similarly implied benefit from local therapy in selected patients but could not define segment-specific nodal fields[33-35]. International and tertiary-center series further established the rarity and heterogeneity of SCCE[36,37].

Population-level and institutional datasets improved generalizability but not anatomic resolution. Nationwide, cross-population, and institutional analyses confirmed that outcomes vary by disease extent, treatment modality, and cohort composition[38-40]. Patient-data meta-analyses, single-institution reviews, and clinicopathologic studies remained useful for historical and clinical context but should not be used to infer station-specific probabilities[41-43]. Recent narrative reviews summarized therapeutic progress while underscoring the rarity of reporting SCCE-specific station denominators[44].

Analyses of radiotherapy, chemoradiotherapy, and adjuvant treatment provided complementary evidence that locoregional therapy may be clinically relevant, but they did not identify a universally appropriate elective nodal template[45-49]. The studies addressed immune infiltration, RB1/SOX2 biology, chemoimmunotherapy, serum markers, and contemporary prognostic or treatment-strategy analyses[50-55]. These data reinforce the need for a systemic perspective but do not obviate the need for standardized anatomic staging and station-level reporting.

Available studies report station frequencies but generally do not provide survival curves for station-positive patients, station-specific hazard ratios, or station-specific recurrence denominators. Therefore, a conventional efficacy index cannot be estimated defensibly for SCCE. Multiplying station frequency by whole-cohort survival yields only a frequency-weighted ranking, not an estimate of station-specific treatment benefit. This distinction is important because the strongest survival signals in contemporary SCCE cohorts are typically stage, treatment sequence, systemic therapy, and adverse tumor biology, rather than isolated nodal station status[15-17].

Similarly, skip metastasis should be viewed as an anatomic warning rather than a validated prognostic marker. Negative peritumoral nodes do not rule out more distant regional nodal disease, particularly in middle and lower thoracic tumors[9,13]. No available SCCE dataset demonstrates that skip status identifies a subgroup in which treating a specific elective nodal station improves survival.

This review has several limitations. Most evidence was retrospective, and the surgical cohorts were susceptible to selection bias, differences in treatment sequencing, and variation in lymphadenectomy fields[8-20]. One large dataset was available only as a preprint and was therefore interpreted cautiously[13]. Potential overlap among Chinese cohorts, including possible source-population overlap between Xu et al[8] and CHiSCEC[13] and overlap among Chinese multicenter treatment cohorts, could not be ruled out based on publicly available information[18-20] (Table 3). The overlap-focused sensitivity analysis excluding Xu et al[8] yielded a pooled prevalence of 64.3%, supporting the robustness of the overall direction while not eliminating the underlying uncertainty.

Table 3 Main limitations and safeguards applied in the review.

Issue	Risk to inference	Safeguard used here
Retrospective surgical cohorts	Selection bias and variable treatment sequence	Pooled only compatible pathologic surgical denominators; interpreted estimates as surgical-series probabilities
Heterogeneous station maps	Invalid formal station-by-segment pooling	Summarized station evidence as an evidence map
Preprint evidence	Non-peer-reviewed data and possible cohort-overlap uncertainty	Included CHiSCEC for completeness, flagged its preprint status and possible overlap with Xu et al[8], and performed an overlap sensitivity analysis[8,13]
Node-level clustering	Individual nodes are not independent observations	Reported node-level metastatic ratios descriptively only
Unavailable station-positive survival endpoints	Station-specific survival benefit and efficacy index cannot be estimated	Avoided claims that treating any specific station improves survival

Station-level inference was dominated by a single detailed, peer-reviewed surgical cohort and a large multicenter preprint[9,13]. Earlier Chinese studies provided directionally consistent support but had smaller sample sizes and less harmonized station definitions[10-12]. In addition, the evidence set was PRISMA-informed but assembled from multiple curated sources. As source-level retrieval counts before deduplication were unavailable, a conventional database-specific flow diagram could not be constructed.

Future SCCE studies should adopt standardized nodal reporting templates. Minimum data elements should include tumor segment, tumor length, invasion depth or T category, pure vs mixed SCCE histology, lymphovascular invasion, neoadjuvant and adjuvant treatment, surgical approach, lymphadenectomy field, total nodes examined, station-specific nodes examined and positive, recurrent laryngeal nerve status, perigastric or celiac status, radiotherapy nodal volumes, station of nodal recurrence, site of distant recurrence, disease-free survival, cancer-specific survival, and overall survival[9,13].

Multicenter collaboration is essential as SCCE is rare. The priority is not merely a larger cohort but a larger cohort with harmonized station maps, transparent documentation of cohort independence, and station-specific recurrence and survival endpoints.

All four surgical cohorts included in the quantitative synthesis were from China, limiting the generalizability of the findings to other populations. Differences in epidemiology, healthcare systems, diagnostic pathways, and treatment strategies may influence disease presentation and the observed patterns of nodal metastasis. Variations in surgical practice, the extent of lymphadenectomy, and pathologic reporting standards between Eastern and Western centers may further limit direct extrapolation of station-level distributions. The present findings should therefore be interpreted primarily as reflecting Chinese surgical cohorts and require validation in multiregional datasets, including those from Western centers.

As only four studies were included in the quantitative synthesis, publication bias was not assessed using funnel plots or Egger’s regression, which have limited power and are unreliable when < 10 studies are available. Therefore, publication bias cannot be excluded.

CONCLUSION

SCCE is associated with a high regional nodal burden. Four surgically staged cohorts included 579 patients and 352 node-positive events, yielding a pooled prevalence of pathologic lymph node metastasis of 61.1%. Available station-level evidence suggests predominant recurrent laryngeal nerve and upper mediastinal involvement in upper thoracic SCCE, bidirectional spread in middle thoracic SCCE, and predominant lower mediastinal, perigastric, and abdominal involvement in lower thoracic SCCE.