Endoscopic submucosal dissection in early esophageal neoplasia: Celebrating efficacy, confronting heterogeneity, and refining surveillance for high-risk patients

Abstract

Esophageal cancer poses a severe global healthcare burden, with a dismal prognosis primarily due to late-stage diagnosis - only 10%-30% of patients survive 5 years when symptoms trigger medical attention. Endoscopic submucosal dissection (ESD) has emerged as a transformative minimally invasive therapy for early esophageal neoplastic lesions, offering curative potential while preserving organ function. However, the clinical landscape of early esophageal neoplasia is highly heterogeneous, with low-grade intraepithelial neoplasia, high-grade intraepithelial neoplasia, early-stage carcinoma, and superficial carcinoma differing significantly in biological behavior, endoscopic features, and treatment response. This article examines the recent retrospective study published by Zhang et al, which analyzed 245 patients with 264 lesions treated with ESD between 2014 and 2022. The study fills a critical gap in understanding this heterogeneity by systematically linking lesion stage to clinicopathological characteristics, ESD efficacy, and long-term prognosis. It not only validates ESD’s role in early disease but also raises urgent questions about refining stratified management and addressing unmet needs in high-risk populations. This article discusses the implications of the study’s findings, contextualizes them within current clinical practice, and outlines directions for future research to advance care for patients with early esophageal neoplasia.

Key Words: Early esophageal neoplasia; Endoscopic submucosal dissection; Lesion heterogeneity; Surveillance; Esophageal stricture

Core Tip: This article explores early esophageal neoplastic lesions (low/high-grade intraepithelial neoplasia, early/superficial carcinoma) treated with endoscopic submucosal dissection, revealing heterogeneity in endoscopic/pathological features, curative resection rate, and prognosis. Key factors include non-curative resection, esophageal stricture, age, and prior cancer history. Biomarkers (Ki-67, p53) differ by lesion grade, correlating with invasion and treatment response, informing stratified management, adjuvant therapy, and personalized surveillance.

TO THE EDITOR

Esophageal cancer remains a major global healthcare challenge, with its poor prognosis largely attributed to late diagnosis. When symptoms finally lead patients to seek medical care, only 10%-30% of them survive for 5 years[1]. In this context, endoscopic submucosal dissection (ESD) has emerged as a revolutionary minimally invasive treatment for early esophageal neoplastic lesions. It not only holds the potential for cure but also preserves the normal function of the esophagus. Nevertheless, the clinical manifestations of early esophageal neoplasia are far from uniform. Low-grade intraepithelial neoplasia (LGIN), high-grade intraepithelial neoplasia (HGIN), early-stage carcinoma, and superficial carcinoma differ greatly in their biological behaviors, endoscopic features, and responses to treatment. Physiologically, LGIN is characterized by mild epithelial dysplasia with preserved cell polarity and minimal invasion risk, while HGIN shows disrupted cell architecture, increased mitotic activity, and heightened progression risk; early-stage carcinoma is confined to the mucosal layer (without submucosal involvement) and has low metastatic potential, whereas superficial carcinoma invades the submucosal layer and is associated with higher lymphovascular invasion rates. Biomarker-wise, LGIN rarely exhibits abnormal p53 expression or elevated Ki-67 (a proliferation marker), but HGIN and carcinomas often present with p53 mutation (detectable via immunohistochemistry) and Ki-67 index > 20% - these biomarkers help distinguish lesion aggressiveness and guide treatment decisions[2].

The recently published study by Zhang et al[3] is a retrospective analysis involving 245 patients with 264 lesions who underwent ESD between 2014 and 2022. This study fills a crucial gap in our understanding of the heterogeneity of early esophageal neoplasia. By systematically establishing the connections between lesion stage and clinicopathological characteristics, ESD efficacy, and long-term prognosis, it not only confirms the important role of ESD in the treatment of early esophageal neoplasia but also raises urgent questions about how to improve stratified management strategies and address the unmet medical needs of high-risk patient groups.

BEYOND “ONE-SIZE-FITS-ALL”: THE VALUE OF UNRAVELING LESION HETEROGENEITY

A prominent strength of Zhang et al’s study[3] lies in its detailed analysis of how endoscopic and pathological features vary across different lesion stages. These findings are not merely descriptive but also have significant clinical relevance. For endoscopists, the differentiation between LGIN and more advanced lesions (such as early carcinoma or superficial carcinoma) no longer relies solely on histological examination. Instead, there are clear endoscopic indicators to aid in their assessment. For example, LGIN rarely presents with surface vesiculation (only 8.5%) or submucosal vascular blurring (1.4%), while early carcinoma shows these features in 33.8% and 13.8% of cases, respectively.

An even more compelling finding is the analysis of intraepithelial papillary capillary loops (IPCLs). LGIN is mainly characterized by IPCL-A (accounting for 29.6%), whereas early carcinoma and superficial carcinoma are predominantly IPCL-B2 (accounting for 44.6% and 45.2%, respectively). These observations represent a significant advancement in preoperative assessment. They enable endoscopists to predict the aggressiveness of lesions before performing ESD, helping to select patients who are most likely to benefit from the procedure and avoid over-treatment of indolent LGIN. As noted in the study, approximately 37% of LGIN cases may regress spontaneously.

From a pathological perspective, the study further confirms an obvious fact: As lesions progress from LGIN to superficial carcinoma, the risk of treatment failure increases significantly. For instance, the vertical margin positivity rate of superficial carcinoma is 16.1%, which is in sharp contrast to LGIN (0%) and HGIN (0%). Moreover, superficial carcinoma is much more likely to exceed the absolute indications for ESD (40%, compared to 0% for early carcinoma). This heterogeneity explains why the curative resection (CR) rate drops sharply from 100% in LGIN to 51.6% in superficial carcinoma - it turns out that not all early lesions can be equally “cured” by ESD.

For clinicians, this means abandoning the “one-size-fits-all” approach to ESD and adopting a stage-specific treatment strategy instead. For example, ESD should be prioritized for the treatment of LGIN and HGIN (where the CR rate is nearly 100%), while for superficial carcinoma - especially those with submucosal layer 2 invasion (SM2 invasion, defined as invasion into the middle to deep third of the submucosal layer) or positive margins - clinicians need to carefully weigh the pros and cons of ESD against alternative treatments such as esophagectomy.

ESD’S EFFICACY: A WIN, BUT NOT WITHOUT CAVEATS

The long-term survival data from Zhang et al’s study[3] are encouraging. The 3-year, 5-year, and 8-year disease-free survival rates of the entire cohort are 96.8%, 91.7%, and 86.4%, respectively. These figures are comparable to or even higher than those of traditional surgical methods for early esophageal neoplasia[4]. This further confirms that ESD is a first-line treatment option for eligible patients, especially for the elderly population, in whom surgical treatment is associated with higher morbidity.

However, the study also highlights two critical challenges that cannot be ignored: Non-CR (NCR) and postoperative complications. NCR occurs in 11.4% of patients (30 out of 264), and the management of these patients remains a clinical dilemma[5]. The study reports that 14 of these NCR patients received adjuvant treatment, including 5 who underwent esophagectomy and 9 who received chemoradiotherapy. Unfortunately, 5 of these 14 patients still experienced recurrence[6].

This raises an important question: Is the adjuvant treatment for NCR patients sufficiently tailored to the biological characteristics of the lesions? For example, the study identifies the depth of lesion invasion (SM2) as a strong predictor of recurrent metastasis hazard ratio (HR), with an HR = 30.213[7]. This suggests that NCR patients with SM2 invasion may require more aggressive adjuvant treatment strategies than those with superficial submucosal involvement. The rationale for this recommendation is that SM2 invasion is associated with significantly higher rates of lymphovascular invasion (reported in 35%-50% of SM2 lesions, vs < 10% in superficial submucosal lesions like SM1) and local recurrence (HR = 30.213 for SM2 vs HR ≈ 2-5 for SM1). The deep submucosal layer contains more lymphatic and vascular channels, facilitating tumor cell dissemination; in contrast, superficial submucosal lesions have limited access to these channels, reducing metastatic risk[8]. Regrettably, current clinical guidelines provide limited clarity on this issue. The findings of Zhang et al’s study[3] emphasize the need for prospective clinical trials to compare the efficacy of different adjuvant treatment modalities (such as chemoradiotherapy vs surgery) in different subgroups of NCR patients.

Complications, particularly esophageal stricture, are another major drawback of ESD. The study shows that the incidence of stricture increases with the progression of lesion stage: 4.2% in LGIN, 14.4% in HGIN, 24.6% in early carcinoma, and 35.5% in superficial carcinoma. Female sex [odds ratio (OR) = 2.450], lesion length ≥ 5 cm (OR = 5.833), and circumferential extent ≥ 3/4 (OR = 3.472) are independent risk factors for esophageal stricture[9]. These findings are consistent with previous research but also highlight the lack of effective preventive measures.

Although the study mentions steroid therapy and balloon dilation as methods for preventing and treating esophageal stricture, it also points out that even these interventions fail to completely eliminate strictures in high-risk patients (such as those with circumferential lesions). Future research should focus on exploring new approaches to reduce the morbidity associated with esophageal stricture, such as biodegradable stents or cell-based regenerative therapies, as this remains an unmet clinical need that significantly affects patients’ quality of life.

HIGH-RISK PATIENTS: THE FORGOTTEN CORNERSTONE OF PROGNOSIS

Perhaps the most significant finding of Zhang et al’s study[3] is the identification of two reliable prognostic factors: Patient age and a history of prior esophageal cancer. Each additional year of age increases the risk of poor disease-free survival by 1.8% (HR = 1.018). Patients with a prior diagnosis of esophageal cancer have a 3-fold higher risk of adverse outcomes (HR = 3.050) and an 18-fold higher risk of recurrent metastasis (HR = 18.230).

Even more striking are the results of the sensitivity analysis. When superficial carcinoma cases (n = 31) were excluded from the analysis, the associations between these two factors and prognosis became even stronger (age: HR = 1.097; prior esophageal cancer history: HR = 10.429). This confirms that these two factors are relevant across all lesion stages of early esophageal neoplasia.

These findings challenge the current follow-up paradigms, which often adopt a uniform schedule (such as endoscopy at 3 months, 6 months, and 12 months after treatment, followed by annual endoscopies) regardless of the patient’s risk level. For elderly patients or those with a prior history of esophageal cancer, this “one-schedule-fits-all” approach may be inadequate. The study shows that 81.8% of recurrences occur within 5 years after treatment. This suggests that high-risk patients need more frequent surveillance during this critical period, such as biannual endoscopies.

In addition, the study notes that 5 out of 10 deaths were caused by second primary tumors, including 3 cases of lung cancer, 1 case of esophageal cancer, and 1 case of gastric cancer. This reminds us that follow-up care for patients with early esophageal neoplasia should not be limited to the esophagus alone. It should also include screening for other malignancies, especially in patients who smoke or have a family history of cancer. Specifically recommended screening tests include: (1) Low-dose computed tomography for lung cancer: Annual low-dose computed tomography reduces lung cancer mortality by 20% in high-risk groups (e.g., smokers with ≥ 30 pack-years or former smokers who quit < 15 years ago)[10]; (2) Gastroscopy for gastric cancer: Biennial to triennial gastroscopy, as esophageal and gastric cancers share risk factors (e.g., Helicobacter pylori infection, chronic atrophic gastritis) and co-occur in 5%-10% of high-risk patients[11]; and (3) Enhanced endoscopic surveillance for esophageal recurrence: Narrow-band imaging endoscopy every 6 months in the first 2 years, as narrow-band imaging improves detection of abnormal IPCLs (a precursor to recurrent lesions) by 30% compared to white-light endoscopy[12].

FUTURE DIRECTIONS: FROM REAL-WORLD DATA TO PRECISION MEDICINE

Zhang et al’s study[3] makes a valuable contribution to the existing literature. However, like all retrospective, single-center studies, it has certain limitations, including a small sample size for superficial carcinoma (n = 31), potential selection bias, and a lack of data on molecular biomarkers. These limitations point to clear priorities for future research.

First, multicenter prospective clinical trials are needed to validate the findings of this study, particularly regarding the role of IPCL staging in preoperative risk stratification and the optimal adjuvant treatment for NCR patients. Second, integrating molecular markers (such as p53, Ki-67, or circulating tumor DNA) with endoscopic and pathological features may enable more accurate prognosis assessment. This could transform the current “stage-based” management approach into a more precise “biology-based” one. Furthermore, leveraging cancer genomics - including custom panel sequencing targeting driver mutations (e.g., TP53, NOTCH1, CDKN2A in esophageal squamous cell carcinoma) and the liquid biopsy paradigm (e.g., detecting circulating tumor DNA or microRNAs in peripheral blood) - can enhance precision. Custom panel sequencing can classify lesions into molecular subtypes (e.g., TP53-mutant vs NOTCH1-mutant) with distinct recurrence risks, guiding adjuvant therapy selection; liquid biopsies (e.g., circulating tumor DNA detection) can identify minimal residual disease 2-6 months before radiological/endoscopic recurrence, enabling early intervention[13,14]. Third, research on the prevention of esophageal stricture should move beyond traditional steroid therapy and balloon dilation to explore innovative treatment methods. For example, autologous cell sheet transplantation or extracellular matrix scaffolds have shown promising results in preclinical models and deserve further clinical investigation.

CONCLUSION

The study published by Zhang et al[3] is more than just a report on the efficacy of ESD. It serves as a call to action for the gastroenterology community to recognize the heterogeneity of early esophageal neoplasia, prioritize the care of high-risk patients, and continuously improve the management of this disease. ESD has already revolutionized the treatment of early esophageal neoplasia, but its full potential can only be realized when we move beyond celebrating its benefits in improving survival rates and instead focus on addressing the unmet clinical needs it has revealed. These include developing tailored adjuvant treatment strategies for NCR patients, improving measures to prevent esophageal stricture, and formulating personalized surveillance plans for elderly patients and those with a prior history of esophageal cancer.

By answering these critical questions, we can ensure that ESD is not only a minimally invasive treatment option but also a precise one - one that optimizes treatment outcomes for every patient, regardless of the stage of their lesion or their risk profile. For clinicians, researchers, and most importantly, patients, this represents the next frontier in the care of early esophageal neoplasia - and Zhang et al’s study[3] has laid a solid foundation for advancing towards this goal.