A Retrospective Approach to Predict Metastatic Recurrence Risk After Endoscopic Resection for Esophageal Squamous Cell Carcinoma

Abstract

Recent progress in early detection of esophageal squamous cell carcinoma (ESCC) has led to the widespread acceptance of endoscopic submucosal dissection (ESD) as a minimally invasive treatment. Advances in ESD techniques and the implementation of preventive measures for post-ESD strictures—such as local triamcinolone injections—have yielded favorable outcomes [1], contributing to the expansion of ESD indications to larger and more advanced lesions. We are in the process of expanding complete circumferential resection to larger ESCC [2]. Treatment strategies are typically determined based on preoperative assessment of invasion depth. When preoperative diagnosis is cT1a-epithelium (EP)/lamina propria mucosae (LPM), endoscopic resection (ER) is generally indicated [1]. Even in cases diagnosed as cT1a-mascularis mucosae (MM) or cT1b-submucosa (SM) 1, observation following ER is feasible in a substantial proportion of cases [3], leading to the broad adoption of ESD, which facilitates reliable en-bloc resection for lesions requiring precise histological evaluation. Limitation of preoperative invasion diagnosis is also the reason to facilitate the use of ESD. The diagnostic accuracy of ME-NBI using B2 vessels for cT1a–MM and cT1b–SM1 lesions remains modest (55.7%) [1], emphasizing the importance of pathological evaluation after ER.

When postoperative pathological assessment reveals pT1a-EP/LPM/MM without lymphovascular invasion (LVI), favorable clinical outcomes have been reported without additional therapy [1, 3-7]. However, in cases with submucosal invasion or LVI, the risk of metastasis or recurrence is considered high, and additional treatment is generally recommended. Consequently, esophagectomy or chemoradiation therapy (CRT) is commonly performed, both of which have demonstrated favorable clinical outcomes [8]. Nevertheless, accurate data regarding the rate of metastasis or recurrence in patients who do not receive additional treatment remain lacking.

Treatment strategies have previously been guided by lymph node (LN) metastasis rates based on the depth of invasion diagnosed in surgical pathology specimens and the evaluation of metastasis in dissected LNs during esophagectomy. However, surgical specimens are typically sectioned at 5-mm intervals, whereas ER specimens are processed at 2-mm intervals. This discrepancy in tissue processing leads to a potential underestimation of invasion depth in surgical cases, resulting in an overestimation of metastasis risk. For instance, the reported LN metastasis rate for pT1a-MM lesions in surgical cases was 14.6% (95% CI: 10.0–20.3), whereas ER cases show a considerably lower rate of 5.6% (95% CI: 2.9–9.5) [1]. These differences have gradually become more apparent, and the metastasis rate based on ER specimens has emerged as a necessary metric for clinical decision-making.

What is the expected metastatic risk following ER in cases of pT1b-SM1/SM2 or pT1a-MM with LVI? These cases are frequently treated with additional therapy, and therefore, data on metastasis rates under observation alone are extremely limited. Patients who do not undergo further treatment often have restrictions due to advanced age or comorbidities, making long-term cumulative metastasis rates difficult to assess. Single-center studies tend to have insufficient case numbers, and multi-center studies face challenges in consistent data aggregation. Nevertheless, when encountering patients considering observation without additional treatment, estimating the metastatic risk becomes a crucial issue. Clarifying this risk also enables evaluation of the additional effect of CRT or esophagectomy. Several retrospective studies [5-8] have reported on cases that did not undergo additional treatment, and while each is highly informative, the metastatic risk across different invasion depths and LVI status remains insufficiently elucidated.

In this issue of Digestive Endoscopy, Ishihara et al. [9] report on the metastatic rates in patients who underwent additional esophagectomy following ESD for ESCC with pathological findings of submucosal invasion or LVI. Notably, survival status at 5 years post-surgery was confirmed in 92.3% of cases. This analytical approach is particularly robust, as depth of invasion and presence of LVI were assessed based on ER specimens sliced at 2-mm intervals. As a result, both the frequency of LN metastasis at the time of surgical resection and metastasis rates after surgery can be evaluated. In this report, metastatic rates were calculated according to depth of invasion, presence of LVI, and status of the vertical margin (VM). As anticipated, the rates were significantly lower than those reported in previous upfront surgical cases, likely reflecting differences in specimen sectioning methods. For example, the metastatic rate for pT1b lesions was reported as 16.3% in this study, markedly lower than the previously published rates of 25.3% (95% CI: 19.0–32.5) for pSM1 and 25.0% (95% CI: 19.1–31.7) for pSM2 [1]. Consistent with prior findings, LVI remained a strong predictive factor for metastasis; however, the difference in metastatic frequency between SM1 and SM2 was minimal. Specifically, the metastatic rates for SM1 and SM2 lesions without LVI were 8.0% and 9.4%, respectively, while those with LVI were reported as 25.7% for MM, 32.3% for SM1, and 27.7% for SM2. The notably high metastatic rate of 35.1% and local residual rate of 19.2% observed in cases with VM1 or VMX were also striking findings.

At present, the metastatic recurrence rates reported in this study are considered the most reliable indicators for predicting outcomes in patients with ESCC who exhibit submucosal invasion or LVI following ER and are managed with observation. These data are especially valuable in cases where additional esophagectomy or CRT is not pursued due to advanced age, comorbidities, or patient preference. Notably, treatment-related mortality rates for esophagectomy and CRT have been reported as 1.9% and 1.7%, respectively [10], underscoring the importance of individualized risk–benefit assessment. Integrating these pathological risk factors with patient-specific prognostic indicators—such as age, sex, Charlson comorbidity index, and prognostic nutritional index—enables a more practical evaluation of therapeutic options. They offer critical support for evaluating the risks and benefits of additional therapy. By leveraging these findings, clinicians can provide informed explanations to patients, facilitating shared decision-making and the selection of optimal treatment strategies.

The author declares no conflicts of interest.

Risk of Metastasis and Local Residual Cancer After Non-Curative Endoscopic Submucosal Dissection for Esophageal Cancer, http://doi.org/10.1111/den.15082.