Florent Stasiak¹, Loic Reppel^2,3, Mickael Schaeffer⁴, Guillaume Gauchotte^5,6, Arthur Streit¹, Joelle Siat¹, Lucie Schnedecker⁵, Joseph Seitlinger¹, Stephane Renaud^1,6*

¹Department of Thoracic Surgery, Nancy Regional University Hospital, 54500 Nancy, France

²Department of Cell Therapy and Tissue Bank Unit, Nancy Regional University Hospital, 54500 Nancy, France

³Research Unit UMR 7365, IMOPA, Lorraine University, 54500 Nancy, France

⁴Statistical analyses, consulting, modelling, ALSTATS Company, 67750 Scherwiller, France

⁵Department of Pathology and Molecular Biology, Nancy Regional University Hospital, 54500 Nancy, France

⁶Research Unit INSERM U1256, NGERE Unit, Lorraine University, 54500 Nancy, France

^*Corresponding Author: Stephane Renaud, Department of Thoracic Surgery, Nancy Regional University Hospital, 54500 Nancy, France

Received: 03 February 2026; Accepted: 09 February 2026; Published: 20 February 2026

Article Information

Citation: Florent Stasiak, Loic Reppel, Mickael Schaeffer, Guillaume Gauchotte, Arthur Streit, Joelle Siat, Lucie Schnedecker, Joseph Seitlinger, Stephane Renaud. Sentinel Lymph Node in Non- Small Cell Lung Cancer: Longterm Outcomes and Immunopathological Consequences. Journal of Surgery and Research. 9 (2026): 98-109.

DOI: 10.26502/jsr.10020493

Abstract

Keywords

Lung cancer, Sentinel lymph node, Indocyanine green, Micrometastases, Anti-tumor immunity, Immunoprofilling, Immunotherapy

Article Details

1. Introduction

Lung cancer is a major public health concern and the leading cause of cancer-related mortality worldwide [1]. Surgery remains the cornerstone of treatment across several stages of disease, often in combination with peri-operative therapies [2]. Accurate staging of lymph-node involvement is critical, as it informs prognosis and subsequent therapeutic decisions [3]. Traditionally, systematic lymph-node dissection (SLND) is undertaken at the time of surgical resection [4]. Although this approach improves staging accuracy, it may inadvertently compromise anti-tumour immunity by removing immunocompetent lymphoid tissue. The advent of immunotherapy has underscored the central role of the immune system in disease control and long-term survival [5-7]. Re-examining the role of lymph-node dissection in the immunotherapy era is therefore essential. The sentinel lymph node (SLN), the first node draining the tumour, is a surgical concept routinely applied in several cancers [8,9]. In breast cancer and melanoma, for example, SLN biopsy has replaced SLND in selected cases, reducing morbidity without compromising oncological outcomes [10]. By enabling enhanced detection of occult micrometastases often missed by standard pathological assessment through finer sectioning and time-consuming, dedicated immunohistochemistry (IHC) techniques, the SLN approach can yield crucial information on both metastatic status and immune contexture [11,12]. Adaptation of this strategy to thoracic surgery has been explored over many years, but has been hampered by unsatisfactory SLN detection rates, notable complications, and organisational challenges between hospital departments [13,14]. More recently, the use of indocyanine green (ICG) with near-infrared (NIR) fluorescence imaging has renewed interest in SLN mapping in thoracic surgery [15]. Several groups, including ours, have already demonstrated the safety and feasibility of this technique, with an SLN identification rate of 75% and no adverse events attributable to ICG [16-18]. We also observed strong pathological concordance between the SLN and the remainder of the lymph-node dissection [19]. When the SLN showed no neoplastic involvement, the other resected lymph nodes likewise appeared free of cancer. Finally, IHC analyses performed on the SLN in our series led to upstaging in nearly 10% of patients, demonstrating the value of the SLN for improving staging. However, the SLN technique is not currently approved for the surgical management of non-small-cell lung cancer (NSCLC). Its role in anti-tumour immunity and in predicting immunotherapy response remains under investigation. Understanding its immune composition could clarify whether SLN evaluation may serve as a reliable alternative to complete lymphadenectomy while preserving immune function. This clinical and translational study aimed, first, to assess the long-term outcomes associated with the SLN technique in NSCLC surgery and, secondly, to explore the immune environment of the SLN compared with the remainder of the lymph-node dissection in patients with NSCLC.

2. Materials and methods

2.1. Clinical outcomes

This retrospective, observational, single-centre study was conducted in the Thoracic Surgery Department of Nancy Regional University Hospital (France). The study protocol received ethics approval (NCT05136014), and written informed consent was obtained from each participant.

2.1.1. Study outcomes

The primary endpoint was five-year disease-free survival (DFS) and overall survival (OS) among patients who underwent major lung resection and SLND with the SLN technique for lung cancer in our department. DFS was defined as the interval from surgery to radiological and/or pathological confirmation of either local or distant recurrence, or to the date of last follow-up imaging. OS was defined as the interval from enrolment until death from any cause.

We also reported the SLN identification rate using NIR fluorescence imaging after ICG injection, the nodal upstaging rate, and pathological concordance between the SLN and the remaining lymph nodes removed during dissection. Nodal upstaging was defined as the presence of malignancy in one or more lymph nodes on histopathological analysis despite a negative pre-operative assessment.

2.1.2. Study population

Adult patients (≥ 18 years) with confirmed or suspected surgically resectable cT1a–cT2b N0 (clinical stage IA to IIA) NSCLC and no evidence of lymph-node involvement on pre-operative 18F-fluoro-deoxy-glucose positron emission tomography (18F-FDG PET) were enrolled between December 2020 and September 2025. Pre-operative thoracic computed tomography (CT), 18F-FDG PET, and brain magnetic resonance imaging (MRI) were reviewed by both a radiologist and a nuclear radiologist specialising in thoracic oncology. Pulmonary function testing and staging were performed in accordance with European Respiratory Society/European Society of Thoracic Surgeons (ERS/ESTS) guidelines [20]. Pathological staging was assigned according to the eighth edition of the TNM lung cancer classification system.

All patients received a peritumoural injection of ICG, administered either transpleurally or via electromagnetic navigational bronchoscopy (ENB), for SLN identification, following the protocol described by Phillips et al. and by our team [15,17].

2.2. Lymph-node analysis

This experimental study was undertaken in the Thoracic Surgery Department, Pathology Department, and the Cell Therapy and Tissue Bank Unit of Nancy Regional University Hospital (France). Written consent was obtained from each participant. The study was approved by our institutional review board (IRB approval NCT05136014).

2.2.1. Study outcomes and design

The principal objective was to describe and compare the immune populations of the SLN with those of the other lymph nodes removed during NSCLC surgery, to evaluate the potential immunological impact of lymph-node dissection on anti-tumour immunity and immunotherapy response in NSCLC.

We conducted a single-center study at CHRU Nancy with both retrospective and prospective components. Retrospective immunohistochemistry (IHC) was performed on archived lymph-node samples from resected NSCLC patients between January and December 2024, while prospective flow cytometry (FCM) analysis was performed on fresh lymph nodes collected during surgical resections between June and August 2025.

2.2.2. IHC – Study population and protocol

Two groups were defined: a “SLN” group and a “pN+” group. The “SLN” group included adult patients who underwent major lung resection with lymph-node dissection for NSCLC and who benefited from the department’s SLN

protocol. Exclusion criteria were: atypical resection, multiple SLNs, histology other than NSCLC, lymph nodes with major fibrohyaline involvement, and lymph-node metastasis. To form a control group, adult patients who underwent major lung resection with lymph-node dissection for NSCLC without the SLN protocol but with a single lymph-node metastasis were selected to constitute the “pN+” group. Patients with a single lymph-node metastasis were chosen because this node may correspond to the tumour-draining lymph node and can therefore be considered equivalent to the SLN. Exclusion criteria were: atypical resection, multiple metastatic nodes, histology other than NSCLC, and lymph nodes with major fibrohyaline involvement.

To compare the immune populations of the SLN or the pN+ node to the other lymph nodes, the pN0 SLN and 3 other pN0 lymph nodes from 3 di6erent stations were chosen by patient in the “SLN” group, while the single pN+ lymph node and 3 other pN0 lymph nodes from 3 di6erent stations were chosen by patient in the “pN+” group. Para6in-embedded lymph-node tissues were cut using a microtome to generate six microscope slides. One slide was used for HES staining (Coverstainer automated system, Agilent®, Santa Clara, United States) to confirm or rule out metastatic invasion. Five slides were dedicated to characterising the immune environment by IHC (DAKO Omnis automated system, Agilent®) with five antibodies: CD3 (T lymphocytes), CD4 (CD4 T lymphocytes), CD8 (CD8 T lymphocytes), CD20 (B lymphocytes), and CD56 (natural killer [NK] cells).

The detailed protocol is provided in the supplementary data.

2.2.3. FCM – Study population and protocol

As for the IHC component, we aimed to create a “SLN” group and a “pN+” group. The “SLN” group comprised patients with histologically proven or imaging-suspected NSCLC at localised stage cT1a–T2b N0 (stage IA–IIA), treated by anatomical lung resection with lymph-node dissection and the SLN protocol. The “pN+” group comprised patients with histologically proven or imaging-suspected NSCLC with a single histologically proven or suspected lymph-node involvement at staging, treated by anatomical lung resection with lymph-node dissection.

The SLN was detected as described above, and the pN+ node was identified visually during surgery. Per patient, the pN0 SLN or single pN+ node and three other pN0 lymph nodes from three di6erent stations were sampled using an 18-gauge needle by manual aspiration; specimens were stored in cell-culture fluid (RPMI 1640, Eurobio ScientificR) at 4 °C until immunophenotyping and FCM acquisition (MACSQuant10, MiltenyiR). We established three antibody panels to identify immune-cell subsets: a “basal” panel to characterise the global immune landscape; a “T lymphocyte” panel to define T-cell subpopulations; and a “memory” panel to assess memory subsets.

The composition of each panel and the immunophenotyping protocol are detailed in the supplementary data.

2.3. Statistical analysis

Di6erences between continuous variables were compared using Student’s t-test, while categorical variables were compared using the χ2 test. Analyses were performed using StataMP® statistical software (Stata/MP version 13.0). A p-value < 0.05 was considered statistically significant.

2.3.1. Clinical outcomes

Log-rank tests and hazard ratios were used to assess DFS and OS. All statistical tests and visualisations were conducted using GraphPad Prism v10.2.2 (Boston, USA).

2.3.2. Lymph-node analysis

To compare immune populations between lymph nodes, we used a mixed linear model to account for intra-subject correlation, thereby comparing each subject’s data with their own control data. Results are presented as the mean di6erence estimated by the model. An equivalence analysis was conducted using the TOST (Two One-Sided Tests) method. An equivalence margin of 10% for the most abundant cells and 20% for the least abundant cells was deemed clinically relevant and prespecified. Equivalence was concluded when p < 0.05. These equivalence analyses were performed using R software (R Foundation for Statistical Computing, Vienna, Austria).

3. Results

3.1. Clinical outcomes

3.1.1. Patient characteristics

In total, 266 patients consented to undergo SLN identification. The characteristics of the study population are presented in table 1.

3.1.2. SLN – key results

An SLN was identified in 213/266 patients, corresponding to an identification rate of 80.1%. We then focused on patients in whom an SLN was identified and analysed and for whom pathology confirmed a primary lung malignancy. Consequently, 37 patients (17.8%) with a benign lesion and 9 patients (4.2%) with a metastatic lesion were excluded

from subsequent analyses.

Among the remaining 167 patients with primary lung cancer and an analysed SLN, there were 11 cases of SLN invasion by micrometastases detected by IHC, corresponding to a 6.6% nodal upstaging rate. In pN0 SLN, 4/156 (2.6%) false-negative cases were observed, i.e., metastatic invasion was found in a non-SLN.

SD, Standard Deviation ; VATS, Video-Assisted Thoracoscopic Surgery ; FEV1, Forced Expiratory Volume in one second ; DLCO, DiCusing Capacity of the Lung for Carbon Monoxide

Table 1: Demographic characteristics of the clinical part population

3.1.3. Survival analysis

The median follow-up for this cohort was 19 months (IQR = 32).

For OS, median survival was 27 months in the SLN pN+ group, whereas it was not reached in the SLN pN0 group (HR = 9.7, 95% CI 2.61–36.22, p < 0.001). OS is illustrated in figure 1.

pN0, pathologically non-invaded lymph node ; pN+, pathologically invaded lymph node ; SLN, Sentinel Lymph Node ; HR, Hazard Ratio

Figure 1: Five-year overall survival in patients with pN0 SLN and pN+ SLN.

For DFS, median survival was 12 months in the SLN pN+ group, whereas it was not reached in the SLN pN0 group (HR = 10.1, 95% CI 3.68–27.83, p < 0.001). DFS is illustrated in figure 2.

pN0, pathologically non-invaded lymph node ; pN+, pathologically invaded lymph node ; SLN, Sentinel Lymph Node ; HR, Hazard Ratio

Figure 2: Five-year disease-free survival in patients with pN0 SLN and pN+ SLN

3.2. Lymph-node analysis

3.2.1. IHC

3.2.1.1. Patient characteristics

A total of 20 patients were included, divided into two groups: 10 in the “pN+” group and 10 in the “SLN” group. The two populations were statistically comparable for age, smoking status, and ASA score. However, there was a significant difference in respiratory function in favour of the SLN group, as well as in the number of segmentectomies performed. 

Demographic characteristics are summarised in table 2.

SD, Standard Deviation ; ASA, American Society of Anesthesiologists ; FEV1, Forced Expiratory Volume in one second ; DLCO, DiCusing Capacity of the Lung for Carbon Monoxide ; VO2max, maximum oxygen consumption

Table 2: Demographic characteristics of the population studied with IHC.

3.2.1.2. IHC slides

In the “SLN” group, a standard distribution of B lymphocytes (CD20) was observed in the SLN and in other uninvolved lymph nodes removed during dissection, predominantly within follicles in the cortical region. These follicles were numerous and varied in size, indicating potential activation.

T lymphocytes (CD3) were mainly present in the paracortical sinuses. CD4 cells were sometimes arranged in a ring around the follicles, and a few CD8 cells were present in the germinal centers, suggesting potential B–T cell interactions. In the “pN+” group, the parenchyma of the metastatic lymph node was markedly altered by tumour infiltration. Nonetheless, various immune cells remained detectable, with some CD8 cells identified within the cancerous tissue. The other pN0 lymph nodes removed showed a cellular distribution similar to that observed in the “SLN” group.

NK-cell (CD56) staining appeared negative in both groups.

To compare lymph nodes, we performed a CD8 T-cell count. B cells were too numerous to count, and CD4 T cells were weakly labelled and confounded by CD4-positive histiocytes. CD8 T-cell counts were performed on four 0.25 mm² microscopic fields to obtain an average per 1 mm².

In the “pN+” group, the estimated di6erence in CD8 cells between the metastatic lymph node and healthy lymph nodes was −9.175 [−26.074; 7.724], p = 0.07, with an equivalence margin set at 10%.

In the “SLN” group, the estimated di6erence in CD8 cells between SLN and non-SLN values was −4.383 [−24.832; 16.066], p = 0.04, with an equivalence margin set at 10%.

3.2.2. FCM

3.2.2.1. Patient characteristics

Eighteen patients were initially included in the “SLN” group. However, two had intraparenchymal SLNs that could not be retrieved despite very distal dissection and were excluded; two patients initially classified as cN0 were found to have micrometastatic invasion of the SLN on final pathology (diagnosed by IHC with anti-cytokeratin AE1/AE3) without other nodal involvement and were therefore reclassified into the “pN+” group. In one patient, the SLN could not be identified (no lymph node took up ICG after exploration of the various stations) and was excluded. Thus, analyses were performed in 13 patients in the pN0 “SLN” group.

In the “pN+” group, only five patients could be included (including the two reclassified from the “SLN” group) owing to the small number of patients with a single lymph-node metastasis on pre-operative assessment during the inclusion period. One patient was excluded because FCM results were uninterpretable. The final “pN+” group therefore comprised four patients.

The two groups were statistically comparable for age, ASA score, and respiratory function. However, there were significant di6erences in the number of segmentectomies performed and in localised stages (clinical staging) in favour of the “SLN” group. Excluding the two reclassified patients, the labelling technique in the “SLN” group was performed

by ENB in 46% (6/13) and by transpleural injection in 54% (7/13). Regarding histology, the “pN+” group included 4/4 (100%) adenocarcinomas (1 stage IIB and 3 stage IIIA). In the “SLN” group, there were 7/13 (53.8%) adenocarcinomas (1 stage IA1, 3 stage IA2, and 3 stage IA3), 1/13 (7.7%) squamous-cell carcinoma (stage IB), 1/13

(7.7%) neuroendocrine carcinoma (stage IB), 1/13 (7.7%) metastasis of endometrial adenocarcinoma, and 3/13 (23.1%) benign lesions (one bronchiectasis and two giant-cell epithelioid granulomas with necrosis). Patients with histology other than NSCLC were retained for immune-cell analyses, the aim being to compare populations between SLN and other nodes removed during dissection, but will, of course, be excluded from any future survival analysis. Demographic characteristics are summarised in table 3.

SD, Standard Deviation ; ASA, American Society of Anesthesiologists ; FEV1, Forced Expiratory Volume in one second ; DLCO, DiBusing Capacity of the Lung for Carbon Monoxide

Table 3: Demographic characteristics of the population studied with FCM.

3.2.2.2. “SLN” group

3.2.2.2.1. Basal panel

With a 10% equivalence margin, the SLN was statistically comparable to other lymph nodes for CD45 and CD3 cells; for CD56 cells, equivalence was demonstrated with a 20% margin.

3.2.2.2.2. T-lymphocyte panel

With a 10% equivalence margin, the SLN was statistically comparable to other lymph nodes for CD4 and CD8 cells; for Treg cells, equivalence was demonstrated with a 20% margin.

3.2.2.2.3. Memory panel

Four memory T-cell subpopulations were identified:

• • T stem-cell memory (Tscm): CD45RA+ CCR7+ CD95+
• • T central memory (Tcm): CD45RA− CCR7+
• • T e6ector memory (Tem): CD45RA− CCR7−
• • T e6ector memory re-expressing CD45RA (Temra): CD45RA+ CCR7−

With a 20% equivalence margin, the SLN was comparable to other nodes for CD3 Tscm,

CD3 Tcm, CD4 Tscm, CD4 Tcm, and CD8 Tcm.

3.2.2.3. “pN+” group

3.2.2.3.1. Basal panel

With a 10% equivalence margin, pN+ lymph nodes were statistically comparable to pN0 lymph nodes for CD45 cells.

3.2.2.3.2. T-lymphocyte panel

With a 10% equivalence margin, pN+ lymph nodes were statistically comparable to pN0 lymph nodes for CD4 cells.

3.2.2.3.3. Memory panel

The same memory subpopulations were assessed. With a 20% equivalence margin, the pN+ lymph node was statistically comparable to pN0 lymph nodes for CD4 Tscm.

4. Discussion

This clinical and translational study sought to evaluate, first, five-year survival among patients undergoing major lung resection for lung cancer with an SLN protocol at Nancy University Hospital, and secondly to analyse lymph nodes at the cellular level to understand how lymph-node dissection might alter anti-tumour immunity and the capacity to respond to immunotherapy. The overarching aim was to consider whether SLN assessment could serve as an alternative to SLND in a context where the rise of immunotherapy is reshaping oncological priorities and encourages, as far as possible, the preservation of functional immune compartments.

In our cohort, survival analyses underscore the prognostic importance of lymph-node status in patients undergoing surgery for early-stage NSCLC. The survival curves show clear separation between SLN pN0 and SLN pN+ groups for both OS and DFS. Median OS was not reached in the SLN pN0 group, compared with 27 months in the SLN pN+ group (p < 0.001), indicating an almost tenfold higher risk of death in the presence of nodal invasion. Similarly, median DFS was 12 months in the SLN pN+ group, while it was not reached in the SLN pN0 group (p < 0.001). These results confirm that micrometastatic disease in the SLN remains a major prognostic factor, even when pre-operative imaging

shows no lymph-node involvement [11,21,22]. They also support the relevance of SLN assessment as a staging tool to identify a subgroup of patients whose prognosis might be improved with tailored adjuvant strategies, such as post-operative immunotherapy.

Anti-tumour immunity relies on the coordinated action of multiple immune populations [23]. Cytotoxic CD8 T lymphocytes are principal e6ectors, capable of recognising tumour antigens and directly destroying cancer cells, with their action reinforced by CD4 T lymphocytes that secrete IFN-γ and IL-2, stimulating both CD8 cells and macrophages. NK cells provide rapid elimination of tumour cells that evade adaptive surveillance, while dendritic cells capture and present antigens, playing a key role in initiating immune responses. B lymphocytes, through antibody production and antigen presentation, complete this defensive network. Using IHC and the FCM “basal” panel, we showed that SLNs and other dissected lymph nodes contain all these essential immune populations. In our samples, leukocytes (CD45) represented on average 96% (± 3.79) of isolated cells, underscoring the profoundly immune nature of lymph nodes. Among these, CD4 and CD8 T lymphocytes and B lymphocytes (CD19/CD20) constituted predominant subpopulations. Furthermore, in patients with benign histology, these cells were also present despite the absence of tumour cells and can therefore be mobilised if needed. These findings emphasise that lymph nodes are not merely anatomical relays for tumour dissemination, but genuine immunological sites. The functional interaction  between the tumour-draining node and the tumour microenvironment, as demonstrated in prior studies [24], supports the hypothesis that SLND could deplete the cellular reservoir required for e6ective anti-tumour responses, potentially compromising the e6icacy of adjuvant immunotherapies.

These immune cells are central to the principles of immunotherapy and condition its e6ectiveness. After neoadjuvant immunotherapy, patients with complete pathological responses have been shown to exhibit significantly higher levels of CD8 T lymphocytes and dendritic cells in both the tumour microenvironment and lymph nodes compared with those without complete responses [25]. Honigsberg et al. demonstrated that the tumour-draining lymph node actively orchestrates anti-tumour immune responses by producing two types of tumour-specific lymphocyte clones: local CD8 T lymphocytes with potent intranodal activity, and circulating CD8 T lymphocytes with migratory, distal surveillance capacity [7]. Immunotherapy also increases the diversity and cytotoxic activity of these two clonal pools. Removing lymph nodes during surgery may therefore excise the very site of T-cell activation and impair systemic responses. In line with this, a murine model showed that recurrence-free survival was significantly better when lung-cancer surgery was performed without lymph-node dissection than with dissection [26]. Although these data derive from animals, Deng et al. reported in humans that extensive lymph-node dissection (resection of more than 16 lymph nodes) was associated with decreased anti-tumour immunity and reduced response to immunotherapy at recurrence after surgical resection [27].

Moreover, among trials of adjuvant immunotherapy, some have yielded negative results, whereas IMpower010—evaluating one year of atezolizumab (a monoclonal antibody targeting PD-L1) after surgical resection and adjuvant chemotherapy—demonstrated a benefit in recurrence-free survival [28]. One explanation may be that patients in that study did not undergo complete SLND but rather dissection of stations 7 and 4 on the right and stations 7, 5, and 6 on the left, or even simple sampling. Partial preservation of the lymphatic immune system may have allowed a more e6ective response to immunotherapy, which could account for these outcomes. Collectively, this raises the possibility of an immune-preserving nodal strategy i.e., a non-extended lymphadenectomy that provides high-quality staging while avoiding immune system impairment—within which the SLN technique could be particularly valuable.

We showed by IHC that the SLN harbours a CD8 T-cell population comparable to that of other nodes in the dissection, whereas the metastatic node appears depleted in CD8 cells relative to healthy lymph nodes. FCM analysis confirmed equivalence in CD8 cells within the “SLN” group and absence of equivalence in the “pN+” group, although in the latter the difference favoured the metastatic node, which contained more CD8 cells than the other nodes (positive di6erence of 3.014). This apparent discrepancy may reflect that the pN+ nodes studied by IHC were all invaded by metastases > 2 mm, likely creating a highly immunosuppressive environment, whereas in FCM, 50% (2/4) of pN+ nodes contained micrometastases of 0.2–2 mm, potentially triggering an anti-tumour immune response. That CD8 T lymphocytes are present in similar proportions in SLN and non-SLN nodes suggests that, when the SLN is pN0, omitting SLND could preserve an immunocompetent nodal compartment and thereby maintain the patient’s immune capacity. Conversely, the relative paucity of CD8 cells in metastatic nodes strongly indicates locally impaired immunity, reinforcing the notion of persistent immune activity in nodes more distant from the tumour, as previously described (29). The “memory” panel enabled exploration of immune-memory subpopulations within lymph nodes. We demonstrated the presence of stem-like memory T cells (Tscm) and central memory T cells (Tcm) in SLN and in other nodes, with good concordance. Memory T cells underpin durable anti-tumour immunity through long-term persistence and rapid reactivation upon antigen re-encounter. Tscm (CD45RA+ CCR7+ CD95+) combine robust self-renewal with the capacity to di6erentiate into all other memory and e6ector T subsets (30). Their longevity and proliferative potential make them attractive targets in cancer immunotherapy, notably as reservoirs capable of sustaining long-term responses,

and they are now considered strategic targets for adoptive immunotherapies (e.g., CAR-T cells, tumour-infiltrating lymphocyte transfer) [31]. Tcm (CD45RA− CCR7+) predominantly reside in lymph nodes, have high proliferative capacity, and provide an essential link between Tscm and e6ector memory T cells (Tem) [32]. Unlike Tem, which circulate in peripheral tissues and rapidly exert cytotoxic function, Tcm generate powerful secondary e6ectors after antigenic stimulation while persisting over the long term. Several studies have associated the presence of Tcm and Tscm in tumour-draining nodes with greater e6icacy of immunotherapies [27]. Connolly et al. showed that stem-like CD8 T cells persist in draining lymph nodes and continuously fuel intratumoural anti-tumour responses [33], highlighting lymph nodes as active immune reservoirs.

The evolving role of the SLN in thoracic surgery mirrors experience in breast cancer, where practice has shifted from systematic axillary dissection to a more selective, conservative strategy that combines reliable staging, reduced morbidity, and immune preservation. Historically, treatment relied on systematic axillary lymph-node dissection [34], which was associated with substantial morbidity [35]. Introduction of the SLN technique profoundly changed practice: randomised trials demonstrated that SLN analysis alone provides reliable staging while significantly reducing post-operative complications [36]. Moreover, there is no di6erence in OS, DFS, or local control between women undergoing

additional axillary dissection and those who do not when the SLN is pN0 [10]. Axillary dissection has therefore become reserved for patients with proven or extensive nodal disease, whereas in most cases SLN excision and analysis su6ice to guide therapy [37]. Beyond staging, several studies have identified the SLN as a key site of immune interaction with the tumour [38], reinforcing the view that it is not simply an anatomical relay but a central immunological player. This evolution in breast cancer paves the way for similar considerations in lung cancer, where the SLN technique may allow high-quality staging alongside preservation of anti-tumour immunity.

This study has limitations: the ENB technique entails significant equipment costs and a learning curve; the sample sizes for both IHC and FCM analyses were small; patients in the “SLN” group were highly selected (stage I), reducing representativeness; and follow-up in the translational component remains insufficient to fully assess prognostic implications. It also has strengths: to our knowledge, this is the largest SLN cohort in lung cancer evaluating long-term outcomes, and the first to explore lymph-node immune populations in lung-cancer patients in such an exhaustive manner. Unlike most previous studies, it relies on identification of the true SLN using a specific technique, rather than analysis of a proximal drainage node, ensuring more accurate staging and improved prognostic assessment. Integration of retrospective and prospective elements using two complementary analytical methods strengthens the validity of our findings.

5. Conclusion

This study supports the proposition that the SLN technique may allow limitation of the extent of lymph-node dissection while ensuring reliable and potentially more accurate staging, thereby contributing to more conservative surgery that preserves the patient’s immune system. Although these findings require confirmation in larger cohorts with longer follow-up, they open the way to changes in surgical practice in thoracic oncology, integrating accurate staging with immune preservation to optimise responses to systemic treatments.

Fundings source

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References

1. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 71 (2021): 209-249.
2. Jeon H, Wang S, Song J, et al. Update 2025: Management of Non-Small- Cell Lung Cancer. Lung 203 (2025): 53.
3. Watanabe S. Lymph node dissection for lung cancer: past, present, and future. Gen Thorac Cardiovasc Surg 62 (2014): 407-414.
4. Lardinois D, De Leyn P, Van Schil P, et al. ESTS guidelines for intraoperative lymph node staging in non-small cell lung cancer. Eur J Cardiothorac Surg 30 (2006): 787-792.
5. Tosello J, Borcoman E, Sedlik C, et al. Le role des ganglions lymphatiques drainant la tumeur a l’ere des immunotherapies. Bulletin de l’Academie Nationale de Medecine 206 (2025): 485-495.
6. Jones D, Villena-Vargas J. The Evolving Role of Lymph Node Resection in the Era of Immunotherapy for Non-Small Cell Lung Cancer. Semin Thorac Cardiovasc Surg 14 (2025).
7. Honigsberg R, Cruz T, Yoffe L, et al. Tumor-specific draining lymph node CD8 T cells orchestrate an anti-tumor response to neoadjuvant PD-1 immune checkpoint blockade. bioRxiv (2025).
8. Cheng TW, Hartsough E, Giubellino A. Sentinel lymph node assessment in melanoma: current state and future directions. Histopathology 83 (2023): 669-684.
9. Giammarile F, Vidal-Sicart S, Paez D, et al. Sentinel Lymph Node Methods in Breast Cancer. Semin Nucl Med 52 (2022): 551-560.
10. Krag DN, Anderson SJ, Julian TB, et al. Sentinellymph- node resection compared with conventional axillary-lymph-node dissection in clinically node-negative patients with breast cancer: overall survival findings from the NSABP B-32 randomised phase 3 trial. Lancet Oncol 11 (2010): 927-933.
11. Rusch VW, Hawes D, Decker PA, et al. Occult metastases in lymph nodes predict survival in resectable non-small-cell lung cancer: report of the ACOSOG Z0040 trial. J Clin Oncol 29 (2011): 4313-4319.
12. Caso R, Jones GD, Jones DR. Sentinel lymph node mapping in lung cancer: a step forward? J Thorac Dis 10 (2018): S3254-S3256.
13. Little AG, DeHoyos A, Kirgan DM, et al. Intraoperative lymphatic mapping for non-small cell lung cancer: the sentinel node technique. J Thorac Cardiovasc Surg 117 (1998): 220-224.
14. Gilmore DM, Khullar OV, Colson YL. Developing intrathoracic sentinel lymph node mapping with near-infrared fluorescent imaging in non–small cell lung cancer. The Journal of Thoracic and Cardiovascular Surgery 144 (2012): S80-84.
15. Phillips WW, Weiss KD, Digesu CS, et al. Finding the ‘True’ N0 Cohort: Technical Aspects of Near-infrared Sentinel Lymph Node Mapping in Non-small Cell Lung Cancer. Ann Surg 272 (2020): 583-588.
16. Seitlinger J, Stasiak F, Piccoli J, et al. What is the appropriate ‘first lymph node’ in the era of segmentectomy for non-small cell lung cancer? Front Oncol 12 (2022): 1078606.
17. Stasiak F, Seitlinger J, Streit A, et al. Sentinel Lymph Node in Non-Small Cell Lung Cancer: Assessment of Feasibility and Safety by Near-Infrared Fluorescence Imaging and Clinical Consequences. J Pers Med 13 (2022): 90.
18. Wollbrett C, Seitlinger J, Stasiak F, et al. Clinicopathological factors associated with sentinel lymph node detection in non-small-cell lung cancer. J Cardiothorac Surg 19 (2024): 145.
19. Stasiak F, Seitlinger J, Walsh LC, et al. Sentinel lymph node detection for lung cancer surgery: a possible pathological surrogate of overall lymph node dissection. Front Oncol 15 (2025): 1474887.
20. Brunelli A, Charloux A, Bolliger CT, et al. ERS/ESTS clinical guidelines on fitness for radical therapy in lung cancer patients (surgery and chemo-radiotherapy). European Respiratory Journal 12 (2009): 17-41.
21. Carretta A. Clinical value of nodal micrometastases in patients with non-small cell lung cancer: time for reconsideration? Journal of Thoracic Disease 8 (2016): 12.
22. Huyuk M, Fiocco M, Postmus PE, et al. Systematic review and meta-analysis of the prognostic impact of lymph node micrometastasis and isolated tumour cells in patients with stage I–IIIA non-small cell lung cancer. Histopathology 82 (2023): 650-663.
23. Immunotherapie des cancers Inserm, La science pour la sante. Inserm (2025).
24. Nagasaki J, Inozume T, Sax N, et al. PD-1 blockade therapy promotes infiltration of tumor-attacking exhausted T cell clonotypes. Cell Reports 38 (2022): 12.
25. Bai Z, Cheng X, Ma T, et al. CD8+ T cells infiltrating into tumors were controlled by immune status of pulmonary lymph nodes and correlated with non-small cell lung cancer (NSCLC) patients’ prognosis treated with chemoimmunotherapy. Lung Cancer 197 (2024): 107991.
26. Fear VS, Forbes CA, Neeve SA, et al. Tumour draining lymph node-generated CD8 T cells play a role in controlling lung metastases after a primary tumour is removed but not when adjuvant immunotherapy is used. Cancer Immunol Immunother 70 (2021): 3249-3258.
27. Deng H, Zhou J, Chen H, et al. Impact of lymphadenectomy extent on immunotherapy e6icacy in postresectional recurred non-small cell lung cancer: a multi-institutional retrospective cohort study. Int J Surg 110 (2024): 238-252.
28. Felip E, Altorki N, Zhou C, et al. Five-Year Survival Outcomes With Atezolizumab After Chemotherapy in Resected Stage IB-IIIA Non–Small Cell Lung Cancer (IMpower010): An Open-Label, Randomized, Phase III Trial. JCO (22025): 2343-2349.
29. Cochran AJ, Huang RR, Lee J, et al. Tumour-induced immune modulation of sentinel lymph nodes. Nat Rev Immunol 6 (2006): 659-670.
30. Gattinoni L, Speiser DE, Lichterfeld M, et al. T memory stem cells in health and disease. Nat Med 23 (2017): 18-27.
31. Rosenberg SA, Restifo NP. Adoptive cell transfer as personalized immunotherapy for human cancer. Science 348 (2015): 62-68.
32. Liu Q, Sun Z, Chen L. Memory T cells: strategies for optimizing tumor immunotherapy. Protein Cell 11 (2020): 549-564.
33. Connolly KA, Kuchroo M, Venkat A, et al. A reservoir of stem-like CD8+ T cells in the tumor-draining lymph node preserves the ongoing antitumor immune response. Sci Immunol 6 (2021): 7836.
34. Beck AC, Morrow M. Axillary lymph node dissection: Dead or still alive? Breast (2023): 469-475.
35. Rockson SG. Lymphedema after Breast Cancer Treatment. N Engl J Med 379 (2018): 1937-1944.
36. Lucci A, McCall LM, Beitsch PD, et al. Surgical complications associated with sentinel lymph node dissection (SLND) plus axillary lymph node dissection compared with SLND alone in the American College of Surgeons Oncology Group Trial Z0011. J Clin Oncol 25 (2007): 3657-3663.
37. Costaz H, Rouffiac M, Boulle D, et al. Strategies in case of metastatic sentinel lymph node in breast cancer. Bull Cancer 107 (2020): 672-685.
38. Ribeiro JM, Mendes J, Gante I, et al. Two Di6erent Immune Profiles Are Identified in Sentinel Lymph Nodes of Early-Stage Breast Cancer. Cancers (Basel) 16 (2024): 2881.

Supplementary data

Sentinel lymph node in non-small cell lung cancer: longterm outcomes and immunopathological consequences

Florent Stasiak, Loic Reppel, Mickael Schae6er, Guillaume Gauchotte, Arthur Streit, Joelle Siat, Lucie Schnedecker, Joseph Seitlinger, Stephane Renaud

IHC protocol

To prepare the slides, the para6in blocks were cooled in the freezer for 10 to 15 minutes to harden the para6in, making it easier to cut with the microtome. Once cooled, seven slides per block were prepared using the microtome, with a cut thickness of between 3 and 5 micrometres.

HES staining was performed on a Coverstainer automated system (AgilentR, Santa Clara, United States) according to the steps detailed in table 1. For IHC, the blocks were cut to 4 μm, then the slides were dried in an oven at 56°C for 45 minutes, fixed in 10% buffered formalin and then passed through the DAKO Omnis  automated system (AgilentR). The antibody clones used and the protocols employed are shown in table 2.

Table 1. HES staining protocol on Coverstainer automated system

Table 2. Clones and IHC protocols for DAKO Omnis automated systems

CMF protocol

Antibodies and fluorochromes

- Basal panel: CD45 VioBlue (130-110-637) for leukocytes, CD235a (Glycophorin A) FITC (130-117-688) for erythrocytes, CD3 PE (130-113-139) for T lymphocytes, CD19 PerCPVio700 (130-113-648) for B lymphocytes, CD11b PE-Vio770 (130-110-555) for monocytes, CD56 APC (130-113-310) for NK cells, CD66b APC-Vio770 (130-120-060) for granulocytes.

- ‘T lymphocyte’ panel: CD45RO VioBlue (130-119-620) for identification of memory/naive cells, CD4 FITC (130-114-722) for CD4 T lymphocytes, CD8 APC Vio770 (130-110-819) for CD8 T lymphocytes, CD279 PD-1 PE (130-120-388) for PD-1 labelling, CD11c PerCP-Vio700 (130-128-716) for dendritic cells, CD127 PE-Vio770 (130-113-977) and CD25 APC (130-113-846) for Treg lymphocytes.

- Memory panel: CD45 PE (130-110-632) for leukocytes, CD3 FITC (130-113-138) for T lymphocytes, CD4 PeVio770 (130-113-227) for CD4 T lymphocytes, CD8 APC Vio770 (130- 110-681) for CD8 T lymphocytes, CD95 APC (130-113-005), CD45Ra VioGreen (130-113- 369) and CCR7 VioBlue (130-117-353) for memory phenotypes.

Immunophenotyping protocol

- Double filtration using 70μm filters (Pre-Separation Filters, 70μm, 130-095-823) and 30μm (Pre-Separation Filters, 30μm, 130-041-407) in 15mL FalconR tubes to remove tissue fragments and aggregates.

- Centrifugation of tubes at 400G for 10 minutes.

- Removal of the supernatant, then dissociation of the pellets with 300μL of PBS.

- Transfer of 100μL per cytometry tube.

- Preparation of the antibody mix in a cone for each panel: 4μL of each antibody with 4μL of PBS. For the ‘basal’ and ‘T lymphocyte’ panels, addition of 4μL of Viobility 405/520 Fixable Dye (130-130-404) to assess cell  viability. For the ‘memory’ panel, 5μL of 7-AAD Staining Solution (130-111-568) is added 5 minutes before passing through the cytometer as a viability marker. A quantity of 8μL of the antibody mix is added to the corresponding tubes according to the panels.

- Incubation of the tubes for 15 minutes at room temperature and in the dark.

- Washing of the tubes with 1 mL of PBS, then centrifugation for 10 minutes at 400G.

- Removal of the supernatant and addition of 100 μL of PBS to each tube.

- Run through the flow cytometer.