Variable sensitivity despite high detection rates: is sentinel node navigation surgery feasible following non-curative gastric ESD?

Dear Editor, We read with great interest the recent study by Park et al, “Feasibility of sentinel basin mapping after non-curative endoscopic resection for early gastric cancer: a prospective multicenter cohort study (SENORITA 2 trial)”[1]. Non-curative endoscopic submucosal dissection (ESD) for early gastric cancer (EGC) necessitates additional radical gastrectomy with lymphadenectomy. However, only 5–10% of such cases exhibit lymph node metastasis[2], underscoring the need for sentinel node navigation surgery (SNNS) to guide tailored intervention. Although evidence is accumulating, clinical feasibility remains controversial due to heterogeneity design and limited sample sizes. We suggest that further discussion and investigation are warranted to strengthen the reliability and clinical applicability of these conclusions. Therefore, we conducted a comprehensive search of MEDLINE, EMBASE, Web of Science and the Cochrane central register of controlled trials databases up until January 2026 to identify relevant articles (Table 1) in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses framework (Figure 1). No regional or language restriction was imposed. The search algorithms consisted of Medical Subject Headings and free text terms including as follows: “stomach neoplasm,” “gastric cancer,” “sentinel node navigation surgery,” “sentinel basin mapping,” “lymphatic drainage,” “endoscopic submucosal dissection.” Additional eligible studies were identified by manually checking the reference lists of retrieved literature. Abstracts retrieved from the initial search were screened for relevance to the study question. Studies involving SNNS or mapping after non-curative ESD, followed by additional systematic lymph node dissection, were also included. Studies without additional systematic lymph node dissection, but with reported detection rates of sentinel node mapping, may also be included. Potentially eligible full-text articles were obtained and evaluated in detail. Two investigators (Z.W. and Z.Z.) independently performed the literature search and subsequently assessed the eligibility of all retrieved studies. Any disagreements regarding inclusion were resolved through consensus. A standardized data extraction form, adapted from the Cochrane Consumers and Communication Review Group template was used to collect information and the Newcastle-Ottawa Scale (star-based scoring system with a maximum score of 9) was used to assess risk of bias (Supplemental Digital Content 1, available at: https://links.lww.com/JS9/H299). To synthesize the available high-quality evidence comprehensively, studies with diverse clinical characteristics were included. However, heterogeneity in interventions across trials precluded a quantitative meta-analysis. Consequently, the evidence was integrated through a narrative synthesis. Figure 1.: Identification of studies via databases and registers. Table 1 - Sentinel node navigation surgery following non-curative ESD. Study design Year Country Design Author Patients (n) SLNs HM TD (mm) LDM Tracers Additional systematic lymph node dissection after SNNS 2013 Japan Single center Arigami 16 3 HE + IHC 22.0 Pick-up method 16 RI 2014 Japan Single center Mayanagi 40 5 NR 18.3 Lymphatic basin dissection 40 Dual 2020 Japan Multicenter Mayanagi 132 5 NR 19.0 Lymphatic basin dissection 68 Dual 49 Dye 15 RI 2020 Korea Single center Roh 98 NR HE 21.4 Lymphadenectomy at each lymph node station 98 ICG 2025 Korea Multicenter Park 238 NR HE 20.0 Lymphatic basin dissection 216 Dual 22 ICG No additional systematic lymph node dissection was performed after SNNS 2005 Japan Single center Nobutsugu 5 15 HE 33.0 Laparoscopic lymph node dissection (llnd) 5 ICG 2024 Japan Single center Takahashi 33 8 HE + IHC NR Lymphatic basin dissection 33 ICG Total - - - - 522 - - - - 284 Dual 49 Dye 31 RI 158 ICG Concentration IM Injection time SEN (%) FNR (%) NPV (%) Detection rate (%) 4 mCi 99 m Tc tin colloid SM One day before surgery 100 (2/2) 0 (0/2) 100 (14/14) 100 (16/16) 4 mCi 99 m Tc tin colloid and 10 mg/ml isosulfan blue or ICG SM RI 1 day before surgery 100 (1/1) 0 (0/1) 100 (39/39) 100 (40/40) Dye on the day of surgery 4 mCi 99 m Tc tin colloid and 5 mg/ml ICG patent blue indigo carmine SM RI 1 day before surgery 88.9 (8/9) 11.1 (1/9) 99.2 (123/124) 98.5 (130/132) Dye on the day of surgery 0.625 mg/ml ICG SM One day before surgery 100 (9/9) 0 (0/9) 100 (89/89) 100 (98/98) 0.2 mCi 99mTc-phytate and 5 mg/ml ICG SM On the day of surgery Total 69.6 (16/23) Total 30.4 (7/23) 96.8 (215/222) Total 97.1 (231/238) Dual 70% (14/20) Dual 30.4 (6/20) Dual 97.7 (211/216) 0.25 mg/ml ICG ICG 66.7% (2/3) ICG 33.3 (1/3) ICG 90.9 (20/22) 5 mg/ml ICG SM On the day of surgery - - - 100 (5/5) 5 mg/ml ICG SM On the day of surgery - - - 97.0 (32/33) - - - 81.4 (35/43) 18.6 (8/43) 98.4 (479/487) 98.1 (512/522) Dual, dual-tracer method; FNR, false negative rate; HE, hematoxylineosin; HM, histopathologic method; IHC, immunohistochemistry; IHC, immunohistochemistry; ICG, indocyanine green; LDM, lymph node dissection method; NPV, negative predictive value; NR, not report; RI, radioisotope; SLN, sentinel lymph node; SNNS, sentinel lymph node navigation surgery; SM, submucosal; TD, tumor diameter; xSEN, sensibility. Among seven included studies, two studies involving the dual-tracer method had an overlapping cohort of 40 patients. After excluding these duplicate patients to ensure independent samples, dual-tracer (n = 284), radioisotope alone (n = 31), dye (n = 49), and indocyanine green (ICG) fluorescence (n = 158) methods all demonstrated comparably high detection rates (range 90.9–100%). These results are consistent with those reported for primary gastric cancer patients without prior ESD (range 93.7–97.5%)[3], indicating that prior non-curative ESD does not compromise mapping. Sensitivity outcomes, however, varied. Four studies (n = 246) reported favorable sensitivity (89.9–100%). In contrast, the SENORITA 2 trial (n = 238) reported an overall sensitivity of only 69.6% (70% for dual-tracer, n = 216; 66.7% for ICG, n = 22). It is premature to dismiss SNNS after non-curative ESD based exclusively on this false-negative rate without a comprehensive critical appraisal of all available evidence. This discrepancy cannot be explained solely by the limited number of node-positive cases in earlier studies, as SENORITA 2 itself included a similar nodal-positive cohort (n =23 vs. n =20 in all prior studies). The primary concern was whether ESD alters lymphatic drainage. Fibrosis resulting from ESD may modify lymphatic flow and impede tracer. In a porcine survival model (n = 12), ICG drainage mapping remained unchanged post-ESD in 10 of 12 lesions (83.3%). In the two cases with altered patterns, the identified nodal stations were either consistent with or more extensive than preoperative findings[4]. Given these results and the high detection rates consistently reported in prior studies, precise lymphatic mapping and sentinel node detection do not appear to be a major obstacle following ESD. Notably, in all seven false-negative cases from the SENORITA 2 trial, missed metastatic nodes were located distant from the ESD sites, with two exhibiting skip metastases beyond the perigastric first-tier nodes, a pattern not commonly reported elsewhere. Although skip metastasis is a recognized cause of false negatives, its incidence remains low (2.7–4.8%)[5] and even rarer in EGC (0.8%)[6]. The adopted sentinel lymph node basin concept has effectively mitigated the challenges posed by the complex anatomy of gastric lymphatic drainage and atypical metastatic pathways[7], leading to a low overall false-negative rate in primary tumors[8]. These contradictions indicate that atypical metastatic patterns are not the predominant factor underlying false-negative events. Together, these observations suggest that atypical metastatic patterns may not be the primary contributor to false-negative outcomes. As highlighted in the SENORITA 2 trial, false-negative cases were clustered in participating centers and during the early enrollment phase, exhibiting a clear correlation. Notably, the lead center, which enrolled 75.3% of patients, reported no false-negative cases after 2019. As the only long-term, multicenter study of its kind, its feasibility demonstrated a high dependence on technical proficiency. This highlights the critical importance of standardized protocols and rigorous quality control. The significant association reported between detection failure and surgeon inexperience or a low yield of identified sentinel nodes further underscores the demanding requirements for institutional expertise[9], procedural quality control, and multidisciplinary collaboration[10]. Multivariate analysis confirmed that surgeon inexperience (≤30 cases) was an independent risk factor for failed sentinel lymph node detection (OR = 11.870, P = 0.001)[9]. The learning curve for gastric cancer sentinel node mapping by a single surgeon showed that approximately 26 cases were required to achieve a 95% sentinel lymph node detection rate[11]. For institutions with experience in SNNS for EGC, quality control for the standardization of laparoscopic sentinel basin dissection could be attained after an average of 15 cases[12]. Moreover, the technique’s reliance on precise lymph station harvesting may also affect outcomes. These findings collectively indicate that variable technical proficiency and the absence of standardized protocols are the primary reasons for SNNS failure following ESD. Given the technical complexity of this procedure and the characteristics of the patient cohort, further validation in high-volume centers under stringent quality control protocols is imperative. Despite the lymph station-based evidence presented in the high-volume, single-center study by Roh et al. (2020), well-designed validation studies of lymphatic basin dissection using ICG-based fluorescence imaging remain lacking. Therefore, the clinical adoption of this technique should be approached with caution and selectivity. Among the 522 patients in this cohort, only 8 patients were false-negative cases (1.5%). The remaining 514 patients (98.5%) successfully underwent function-preserving surgery without compromising tumor safety. This highlights the success of the non-curative ESD followed by SNNS strategy in preserving function for the vast majority of patients. However, even such a low false-negative rate may carry significant oncologic consequences. Disease progression in this cohort should be effectively controlled through meticulous surveillance and timely intervention. Meanwhile, long-term disease-free survival data are essential for risk assessment. Ultimately, similar to the expansion of indications for ESD, a well-designed randomized controlled trial comparing SNNS versus standard gastrectomy in patients with non-curative ESD focusing on disease-free survival and quality-of-life outcomes is necessary to provide definitive evidence for advancing this field.