Neutrophil extracellular traps predict sensitivity to neoadjuvant chemotherapy combined with immunotherapy in locally advanced gastric cancer

Abstract

BACKGROUND

Neutrophil extracellular traps (NETs) are associated with an immunosuppressive tumor microenvironment and may influence the efficacy of immune-based therapies. However, their role in neoadjuvant chemotherapy combined with immunotherapy (NACI) for locally advanced gastric cancer (LAGC) remains unclear.

AIM

To investigate the prognostic and predictive value of NET density in LAGC patients undergoing NACI.

METHODS

We enrolled 31 LAGC patients treated with NACI. NET density was assessed through dual immunofluorescence staining of citrullinated histone H3 and myeloperoxidase in pretreatment biopsy and post-treatment surgical specimens. Patients were stratified into high and low pre-NACI NET groups based on median NET density. Pathological complete response (pCR) and overall response rates were evaluated in relation to NET density. Logistic regression analyses were performed to identify independent predictors of treatment outcomes. Dynamic changes in NET density during NACI were also analyzed.

RESULTS

Patients with low pre-NACI NET density demonstrated significantly higher rates of pCR (40% vs 6%, P = 0.037) and overall response (53% vs 12%, P = 0.023) compared to those with high NET density. Low pre-NACI NET density and higher programmed death protein ligand 1 expression were identified as independent protective factors for achieving pCR and better response rates. NACI increased NET density; however, this increase was primarily observed in non-pCR and nonresponder groups. Patients in the pCR and responder groups showed stable NET density before and after treatment. Higher post-NACI NET density was associated with poorer respond to NACI. High post-NACI NET density was associated with increased infiltration of immunosuppressive FOXP3⁺ T regulatory cells (P = 0.025) and CD68⁺ macrophages (P = 0.038).

CONCLUSION

Pre-NACI NET density serves as a prognostic and predictive biomarker for NACI efficacy in LAGC patients. Low pretreatment NET density is associated with favorable outcomes, while increased post-treatment NET density correlates with poorer response. Targeting NET formation may represent a novel therapeutic strategy to enhance NACI efficacy in LAGC.

Key Words: Neutrophil extracellular traps; Locally advanced gastric cancer; Neoadjuvant chemotherapy; Neoadjuvant immunotherapy; Treatment sensitivity

Core Tip: This study identified neutrophil extracellular traps (NETs) as a novel prognostic and predictive biomarker for neoadjuvant chemotherapy combined with immunotherapy (NACI) in locally advanced gastric cancer (LAGC). Low pretreatment NET density correlated with higher pathological complete response and overall response rates, while increased post-treatment NET density was associated with poorer outcomes. These findings suggest that targeting NET formation could enhance NACI efficacy, offering a potential therapeutic strategy for LAGC patients.

INTRODUCTION

Gastric cancer (GC) remains a significant global health burden and ranks as one of the leading causes of cancer-related mortality[1]. Despite advances in diagnosis and treatment, the prognosis of locally advanced GC (LAGC) remains poor, primarily because of the high rates of recurrence and metastasis following surgery[2]. Neoadjuvant chemotherapy (NAC) has been established as a standard strategy to downstage tumors and improve resectability, with the additional benefit of eliminating micrometastatic disease[2,3]. More recently, the advent of immunotherapy, particularly immune checkpoint inhibitors (ICIs), has revolutionized cancer treatment and shown promise in enhancing the efficacy of traditional therapies. NAC combined with immunotherapy (NACI) is emerging as a promising treatment modality for LAGC, to improve long-term survival outcomes[4]. However, the heterogeneity of treatment responses underscores the need for robust biomarkers to predict efficacy and guide personalized therapeutic strategies[5,6].

Neutrophil extracellular traps (NETs) are web-like structures composed of chromatin and granular proteins released by activated neutrophils[7]. Originally identified as a host defense mechanism against pathogens, NETs have since been implicated in cancer progression and metastasis[8]. Within the tumor microenvironment (TME), NETs promote tumor growth, facilitate immune evasion, and contribute to the establishment of an immunosuppressive milieu[9]. Increasing evidence suggests that NETs may interfere with the efficacy of immune-based therapies, including ICIs, by suppressing antitumor immune responses and fostering an immunosuppressive TME[9]. However, the role of NETs in modulating the response to NACI in LAGC remains largely unexplored. NETs are not merely passive scaffolds but active regulators of the TME. They facilitate tumor progression through multiple mechanisms: (1) Physical barrier formation: NETs-derived DNA-histone complexes physically shield tumor cells from cytotoxic T lymphocytes and natural killer cells; (2) Immunosuppressive signaling: Proteases embedded in NETs, such as NE and myeloperoxidase (MPO), cleave chemokines critical for T-cell recruitment (e.g., CXCL9/10), thereby impairing antitumor immunity[10]; and (3) Checkpoint ligand presentation: Recent studies have demonstrated that NETs carry immune checkpoint molecules like PD-L1, directly inhibiting T-cell activation[11]. For instance, some researchers revealed that tumor-associated platelet particles (TRAPs) induce PD-L1-decorated NETs, which suppress CD8⁺ T-cell function via PD-1/PD-L1 axis activation[11]. These findings underscore the dual role of NETs as both structural and functional mediators of immune evasion.

This study investigated the predictive significance of NETs in LAGC patients undergoing NACI. By assessing the density of NETs in pre-NACI biopsies and post-NACI surgical specimens, we aimed to elucidate their relationship with treatment response and pathological complete response (pCR). Additionally, we explored dynamic changes in NET density during NACI and their potential impact on treatment efficacy.

MATERIALS AND METHODS

Patient cohort

This retrospective study was conducted on gastric adenocarcinoma patients who underwent radical resection at the 7^th Medical Center of the Chinese PLA General Hospital. Clinical stage II or III (T3-4 any N or T ≤ 2 N+) tumors according to the 8^th edition of the American Joint Committee on Cancer (AJCC) TNM classification for GC were defined as LAGC[12]. Enhanced computed tomography of the abdomen and magnetic resonance imaging of the gastric region were used to confirm the TNM stage.

The inclusion criteria were: (1) Diagnosis of GC by preoperative biopsy; (2) Surgical cohort had postoperative pathological staging of LAGC, and the neoadjuvant therapy cohort was diagnosed with LAGC by preoperative imaging data and biopsy; (3) Complete medical and follow-up records; and (4) Controlled comorbidities. The exclusion criteria were: (1) Incomplete neoadjuvant therapy; (2) Comorbidities with severe infections, autoimmune diseases, and other life-threatening conditions; (3) Death within 30 days after surgery; (4) Gastric tumors from other origins; and (5) History of other cancers. Using these criteria, 31 LAGC patients who received NACI from December 2023 to December 2024 were enrolled. This study was conducted under the approval of the Institutional Ethics Committee.

NACI regimen

The chemotherapy regimen included SOX [S-1 (tegafur) and oxaliplatin] or paclitaxel. Immunotherapy options included nivolumab, camrelizumab, penpulimab, toripalimab, or zimberelimab. After two treatment cycles, imaging evaluation (enhanced abdominal CT) was performed to assess tumor response. Responses were categorized as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD). For patients with CR, PR, or SD, the existing regimen was continued for an additional two cycles, followed by a final imaging evaluation. For patients with PD, the treatment regimen was discontinued.

Assessment of tumor regression grade

Tumor regression grade (TRG) was evaluated in biopsy samples following the AJCC criteria, using a four-tier grading scale: 0, complete regression with no detectable viable tumor cells; 1, near-complete regression with only isolated cells or small clusters of tumor cells; 2, partial regression characterized by residual tumor being outweighed by fibrosis; and 3, minimal or absent regression with extensive residual malignancy and little to no fibrotic response. Patients were divided into two groups: Responders (TRG 0 or 1) and nonresponders (TRG 2 or 3). TRG 0 was specifically defined as pCR, while TRG 1-3 were grouped under incomplete pathological remission (i-pCR).

Immunohistochemistry

Immunohistochemistry was conducted on 4-μm-thick formalin-fixed paraffin-embedded sections that were dewaxed, hydrated, and underwent antigen retrieval in EDTA, using anti-CD68 macrophages (Abcam, ab213363), and anti-FOXP3 T regulatory (Treg) cells (Abcam, ab20034). After blocking with 5% bovine serum albumin for 30 minutes, the sections were incubated overnight at 4 °C with primary antibody. On the next day, the sections were incubated with secondary antibody for 1 hour at room temperature and diaminobenzidine for 3 minutes followed by hematoxylin staining. All sections were dehydrated and scanned using Leica CS2 and analyzed using Aperio Image Scope.

Quantification of Treg Density Using QuPath: Treg cell density in tumor regions was evaluated. CD68 expression was graded based on staining intensity (negative, weak, moderate, or strong) and the percentage of positive cells was determined. The scoring system categorized samples as follows: Low expression (0 or 1): ≤ 20% positive cells (moderate/strong intensity) or ≤ 100% positive cells (weak intensity)[13]. High expression (2 or 3): ≥ 21% positive cells with moderate/strong intensity. Samples were classified into low (0 or 1) and high (2 or 3) expression groups to analyze the correlation with tumor characteristics.

Immunofluorescence staining

Tumor samples were fixed in formalin, embedded in paraffin, and sectioned at a thickness of 4 μm under the supervision of a pathologist. The tissue slides underwent deparaffinization, rehydration, and antigen retrieval using EDTA buffer. Primary antibodies targeting citrullinated histone H3 (CitH3, 1:1000, ab5103, Abcam) and myeloperoxidase (MPO, 1:1000, AF3667, Abcam) were applied, followed by overnight incubation at 4 °C. Subsequently, fluorophore-conjugated secondary antibodies were added and incubated for 1 h. After washing, nuclei were counterstained, and the slides were visualized using an Olympus Fluoview FV3000 confocal microscope.

Identification and quantification of NETs

Tumor areas were first determined by hematoxylin and eosin staining. Immunofluorescent imaging of these regions was performed with a confocal microscope (Olympus Fluoview FV3000). The images were analyzed using QuPath (v0.4.1) and ImageJ (v1.52c). NETs were recognized based on the colocalization of MPO and CitH3 signals. To quantify NETs, five random fields per slide were selected, and average NET density was calculated. The density was expressed as the percentage of NETs per DAPI-stained nuclei using the following formula: (NET count/DAPI count) × 100.

Statistical analysis

R version 4.2.1 was used in this study. The χ² test and Fisher's exact test were used to analyze categorical variables, and the Mann-Whitney U test was used to analyze continuous variables. We used Kaplan-Meier analysis, univariate and multivariate analysis to examine risk factors for disease-free survival. We used logistic multivariate analysis to identify variables associated with TRG. All statistical methods used two-sided analysis, and P < 0.05 was considered statistically significant.

RESULTS

Patient characteristics

We enrolled 31 patients with LAGC who underwent NACI (Table 1). The cohort included 13 patients aged ≥ 65 years and 18 aged < 65 years. There were 11 female and 20 male patients. Clinical staging revealed seven patients classified as cT3, 7 as cT4a, and 17 as cT4b. For lymph node involvement, one patient was classified as cN1, 11 as cN2, and 19 as cN3. Nine tumors were in the lower stomach, 15 in the middle, and seven in the upper part. Preoperative biopsy grading indicated one patient with a G1 tumor, 10 with G2 tumors, and 20 with G3 tumors. The median PD-L1 score was 7 (interquartile range: 2.5-9.5), and the neutrophil-to-lymphocyte ratio (NLR) was 3 ± 1.38. In terms of immunotherapy regimens, 15 patients received nivolumab, 12 received tislelizumab, and three each received sepranolimab, camrelizumab, toripalimab, or penpulimab. For chemotherapy, 29 patients received the SOX regimen, while two were switched to paclitaxel because of adverse events. Twenty-four patients underwent surgery, while seven did not because of tumor progression.

Table 1 Baseline characteristics of locally advanced gastric cancer patients before neoadjuvant chemotherapy combined with immunotherapy treatment, n (%).

Characteristics	Total (n = 31)	High-NETs (n = 16)	Low-NETs (n = 15)	P value
Sex				0.537
Female	11 (35)	7 (44)	4 (27)
Male	20 (65)	9 (56)	11 (73)
Age				0.378
< 65	18 (58)	11 (69)	7 (47)
≥ 65	13 (42)	5 (31)	8 (53)
Location				0.426
Lower third	9 (29)	5 (31)	4 (27)
Middle third	15 (48)	9 (56)	6 (40)
Upper third	7 (23)	2 (12)	5 (33)
cT				0.505
cT3	7 (23)	4 (25)	3 (20)
cT4a	7 (23)	5 (31)	2 (13)
cT4b	17 (55)	7 (44)	10 (67)
cN				0.014
cN1	1 (3)	1 (6)	0 (0)
cN2	11 (35)	2 (12)	9 (60)
cN3	19 (61)	13 (81)	6 (40)
Tumor grade				0.34
G1	1 (3)	0 (0)	1 (7)
G2	10 (32)	4 (25)	6 (40)
G3	20 (65)	12 (75)	8 (53)
PD-L1 scores, median (Q1, Q3)	7 (2.5, 9.5)	3.5 (1, 7)	9 (5, 16)	0.004
NLR, mean ± SD	3 ± 1.38	3.16 ± 1.53	2.84 ± 1.24	0.532
PD1				0.131
Nivolumab	15 (48)	5 (31)	10 (67)
Tislelizumab	12 (39)	8 (50)	4 (27)
Camrelizumab	1 (3)	1 (6)	0 (0)
Penpulimab	1 (3)	1 (6)	0 (0)
Toripalimab	1 (3)	1 (6)	0 (0)
Zimberelimab	1 (3)	0 (0)	1 (7)
Chemotherapy				0.484
Paclitaxel	2 (6)	2 (12)	0 (0)
SOX	29 (94)	14 (88)	15 (100)

NET: Neutrophil extracellular traps; NLR: Neutrophil-to-lymphocyte ratio; SOX: S-1 (tegafur) and oxaliplatin.

Impact of pretreatment NET density on NACI response

To investigate the relationship between NET density and NACI sensitivity in LAGC, we performed dual immunofluorescence analysis of CitH3 and MPO on pre-NACI biopsy tissues from all 31 patients and post-NACI surgical specimens from 24 patients. Representative images of NETs are shown in Figure 1. Patients were stratified into high and low pre-NACI NETs groups based on median density. Table 1 summarizes the clinical features of these groups. The high pre-NACI NETs group exhibited higher cN stages and lower PD-L1 scores than the low pre-NACI NETs group. However, no significant differences were observed in other clinical features, including age, gender, tumor location, cT stage, tumor grade, NLR, and treatment regimens.

Figure 1 Representative images of immunofluorescence staining of high neutrophil extracellular trap density (upper panel) and low neutrophil extracellular trap density (lower panel) in locally advanced gastric cancer samples without neoadjuvant therapy. Magnification 40 ×. DAPI, Blue; MPO, Green; CitH3, Red. NET: Neutrophil extracellular trap; MPO: Myeloperoxidase.

We next assessed the impact of pre-NACI NET density on pCR and response to NACI. The low pre-NACI NETs group demonstrated a significantly higher pCR rate (40% vs 6%, P = 0.037) and response rate (53% vs 12%, P = 0.023) compared to the high pre-NACI NETs group (Table 2). Post-NACI NET density was also evaluated. The pCR group had lower post-NACI NET density than the no-pCR group (0.450 vs 1.290, P = 0.0388) (Figure 2A). Similarly, the response group showed lower post-NACI NETs density than the no-response group (0.460 vs 1.330, P = 0.0338) (Figure 2A).

Figure 2 Evaluation of pre-neoadjuvant chemotherapy combined with immunotherapy and post-neoadjuvant chemotherapy combined with immunotherapy neutrophil extracellular traps in association with response to neoadjuvant chemotherapy combined with immunotherapy in locally advanced gastric cancer patients. A and B: Pre-neoadjuvant chemotherapy combined with immunotherapy (NACI) neutrophil extracellular trap (NETs) and post-NACI NETs according to response to NETs in locally advanced gastric cancer patients treated with NACI; C-G: Differences between pre-NACI NETs and post-NACI NETs between NET-treated responder, nonresponder, pathological complete response (pCR) and non-pCR groups. The plotted values encompass the range from minimum to maximum, with the inclusion of the median. NET: Neutrophil extracellular trap; NACI: Neoadjuvant chemotherapy combined with immunotherapy; pCR: Pathological complete response.

Table 2 Pre-neoadjuvant chemotherapy combined with immunotherapy neutrophil extracellular traps density correlates with treatment response in locally advanced gastric cancer patients receiving neoadjuvant chemotherapy combined with immunotherapy, n (%).

Characteristics	High NETs (n = 16)	Low NETs (n = 15)	P value
pCR			0.037
no-pCR	15 (94)	9 (60)
pCR	1 (6)	6 (40)
Respond			0.023
No	14 (88)	7 (47)
Yes	2 (12)	8 (53)

NET: Neutrophil extracellular traps; pCR: Pathological complete response.

Univariate logistic regression analysis identified low pre-NACI NET density [hazard ratio (HR) = 0.10, 95% confidence interval (CI): 0.01-0.97, P = 0.047] and higher PD-L1 scores (HR = 0.87, 95%CI: 0.77-0.99, P = 0.028) as significant factors for achieving pCR with NACI (Table 3). Multivariate analysis confirmed these findings, showing that low pre-NACI NET density (HR = 0.07, 95%CI: 0.0-0.96, P = 0.047) and higher PD-L1 scores (HR = 0.85, 95%CI: 0.73-0.99, P = 0.041) were independent protective factors for pCR (Table 3). Furthermore, both univariate and multivariate logistic regression analyses consistently demonstrated that low pre-NACI NET density (HR = 0.13, 95%CI: 0.02-0.75, P = 0.023) was an independent protective factor associated with better treatment response to NACI (Supplementary Table 1).

Table 3 Logistic regression analysis for predictors of pathological complete response in locally advanced gastric cancer patients receiving neoadjuvant chemotherapy combined with immunotherapy.

Variables	Univariate analyses		Multivariate analyses
	HR (95%CI)	P value	HR (95%CI)	P value
Sex
Female	1 [Reference]	NA
Male	1.50 (0.27-8.38)	0.644
Age
< 65	1 [Reference]	NA
≥ 65	2.12 (0.34-13.13)	0.421
Tumor location
Lower third	1 [Reference]	NA
Middle third	5.20 (0.71-37.90)	0.104
Upper third	4.80 (0.40-58.01)	0.217
cT stage
cT3	1 [Reference]	NA
cT4a	1.00 (0.05-19.96)	1.000
cT4b	0.40 (0.04-4.24)	0.447
cN stage
cN1	1 [Reference]	NA
cN2	0.00 (0.00-Inf)	0.995
cN3	0.00 (0.00-Inf)	0.995
PD-L1 scores	0.87 (0.77-0.99)	0.028	0.85 (0.73-0.99)	0.041
Tumor grade
G1-2	1 [Reference]	NA
G3	1.50 (0.27-8.38)	0.644
Immunotherapy
Nivolumab	1 [Reference]	NA
Tislelizumab	1.82 (0.27-12.17)	0.538
Others	1.09 (0.09-13.78)	0.946
NLR
High	1 [Reference]	NA
Low	0.33 (0.03-3.26)	0.345
Pre-NACI NET
High	1 [Reference]	NA	1 [Reference]	NA
Low	0.10 (0.01-0.97)	0.047	0.07 (0.00-0.96)	0.047

NACI: Neoadjuvant chemotherapy combined with immunotherapy; HR: Hazard ratio; CI: Confidence interval; NET: Neutrophil extracellular trap; NLR: Neutrophil-to-lymphocyte ratio.

Association between post-treatment NET density and NACI efficacy

We evaluated the relationship between post-NACI NET density and sensitivity after NACI. The pCR group had significantly lower post-NACI NET density compared to the no-pCR group (0.510 vs 4.610, P = 0.0008) (Figure 2B). Similarly, the response group had a lower post-NACI NETs density compared to the no-response group (0.600 vs 4.765, P = 0.0161) (Figure 2B).

Dynamic changes in NET density during NACI

We explored the dynamic changes in NET density during NACI. NACI significantly increased NET density (mean difference = 2.138, P = 0.0091) (Figure 2C). Subgroup analysis revealed no significant pre- to post-NACI NET changes in the pCR group (mean difference = 0.4986, P = 0.4560) (Figure 2D) or responders (mean difference = 0.1848, P = 0.1848) (Figure 2E). However, significant increases were observed in the no-pCR group (mean difference = 2.814, P = 0.0104) (Figure 2F) and nonresponders (mean difference = 2.473, P = 0.0248) (Figure 2G).

Immunosuppressive cell enrichment in high post-NACI NETs group

We performed immunohistochemistry to quantify the density of immunosuppressive cells, such as FOXP3⁺ Treg cells and CD68⁺ macrophages cells, on tumor tissue (Figure 3A). Our results showed that FOXP3⁺ Treg cells (P = 0.025), and CD68⁺ macrophages (P = 0.038) were significantly enriched in the high post-NACI NETs group compared to the low post-NACI NETs (Figure 3B and C).

Figure 3 Analysis of the correlation between post-neoadjuvant chemotherapy combined with immunotherapy neutrophil extracellular traps and immunosuppressive cells in locally advanced gastric cancer patients. A: Representative images of CD68+ macrophages and FOXP3+ regulatory T cells (Tregs) in the High post-neoadjuvant chemotherapy combined with immunotherapy (NACI) and Low post-NACI groups, respectively; B: Comparison of CD68+ macrophage expression levels between the high post-NACI and low post-NACI groups; C: Differences in FOXP3+ Treg expression between the two groups. In the graphs, the plotted values encompass the range from the minimum to the maximum, with the inclusion of the median. NET: Neutrophil extracellular trap; NACI: Neoadjuvant chemotherapy combined with immunotherapy.

DISCUSSION

We investigated the role of NETs in the context of NACI for LAGC. Pre-NACI NET density was associated with both treatment response and pathological outcomes, and changes in NET density during treatment may provide insights into NACI sensitivity.

Patients with low pre-NACI NET density achieved markedly higher rates of pCR and response compared to those with high pre-NACI NETs density. Logistic regression analysis further identified low pre-NACI NET density as an independent protective factor for achieving pCR and favorable treatment responses. These results align with previous research suggesting that excessive NET formation can contribute to an immunosuppressive TME, which diminishes the efficacy of immune-based therapies[14,15]. Specifically, NETs are implicated in promoting tumor progression through mechanisms such as angiogenesis, immune evasion, and metastasis facilitation[16]. Thus, low pre-NACI NET density may indicate a less immunosuppressive TME, enhancing the antitumor effects of NACI.

NACI treatment increased NET density; however, subgroup analyses revealed this increase was predominantly observed in non-pCR and nonresponder groups. Conversely, NET density remained stable in pCR and responder groups before and after NACI. This suggests that failure to suppress NETs during treatment may reduce therapeutic efficacy. Persistent neutrophil activation and NET formation likely maintain an immunosuppressive TME, undermining the benefits of NACI. Shinde-Jadhav et al[17] have demonstrated that radiotherapy can stimulate NET formation in bladder cancer, with elevated NET levels being closely associated with radiation resistance. Tumor-derived high mobility group box (HMGB)1 released after radiotherapy promotes NET formation through Toll-like receptor 4 activation, underscoring the bidirectional relationship between treatment modalities and NETs[17]. These findings highlight the potential utility of targeting NETs as a strategy to improve NACI efficacy.

Our study also observed that patients in the high pre-NACI NETs group had lower PD-L1 scores, suggesting reduced tumor immunogenicity. Higher PD-L1 expression was independently associated with achieving pCR, consistent with its established role as a predictive biomarker for ICI efficacy[18,19]. These findings point to a complex interplay between NETs and PD-L1 expression, where NETs may influence immune escape mechanisms within the TME. Future investigations should explore the molecular pathways linking NETs to PD-L1 signaling to better understand their combined impact on tumor immunity. Zhou et al[11] revealed that in lung metastasis of breast cancer, TRAPs activate the TLR4/Myd88/ERK/p38 signaling axis by releasing HMGB1, inducing the formation of NETs in human and mouse neutrophils, and PD-L1 carried by TRAPs drives the formation of a premetastatic immunosuppressive microenvironment by inhibiting T-cell function.

Our observation that high NET density correlated with reduced PD-L1 scores aligns with findings by Au et al[20], where NET-rich tumors exhibited compensatory PD-L1 downregulation due to T-cell exclusion. Additionally, the association between rising NETs and resistance mirrors clinical data from Li et al[21], where serum NET biomarkers predicted immunotherapy failure in GC.

Our findings agree with emerging evidence that NETs dynamically interact with tumor and immune cells. We observed that high post-NACI NET density was significantly associated with increased infiltration of immunosuppressive cells, particularly FOXP3⁺ Treg cells and CD68⁺ macrophages. Several mechanisms may underlie this association. First, NET-derived HMGB1 activates the TLR4/MyD88 pathway in dendritic cells (DCs), impairing their antigen-presenting capacity and subsequent T-cell priming[22,23]. Second, citrullinated histones within NETs directly induce apoptosis of CD8⁺ T cells via caspase-3 activation[20]. Third, NETs serve as reservoirs for proteases (e.g., matrix metalloproteinase-9) that degrade extracellular matrix components, facilitating metastatic niche formation while excluding effector immune cells. This multifaceted immunosuppression likely explains why high pre-NACI NET density correlates with reduced PD-L1 scores in our cohort- a phenomenon potentially reflecting compensatory downregulation of PD-L1 in highly NETs-rich, T-cell-excluded tumors[24].

The findings of this study suggest that NET density could serve as a valuable biomarker for stratifying LAGC patients undergoing NACI. Pretreatment assessment of NET density may help identify patients unlikely to respond to NACI, guiding alternative or combination therapeutic strategies. Additionally, monitoring NET dynamics during treatment could provide real-time insights into therapeutic efficacy. Developing interventions to modulate NETs, such as inhibitors targeting neutrophil activation or NET formation, could enhance treatment outcomes for LAGC patients. Han et al[25] found that exogenous irisin significantly inhibits the formation of NETs through the integrin αVβ5/P38/MAPK signaling axis, which may be a key regulatory target for NET production.

Despite these promising findings, our study had several limitations. First, the small sample size limits generalization of the results. The study included 31 patients who met the inclusion criteria and voluntarily participated. Due to financial constraints and other practical considerations, only this number of patients were enrolled at this stage. While a larger cohort would provide additional statistical power, the current sample size remains sufficient to generate meaningful insights. Larger, multicenter studies are needed to validate the utility of NET density as a predictive biomarker in LAGC. Second, while significant associations were observed, the underlying mechanisms remain speculative. Experimental studies are warranted to establish a causal link between NETs and NACI sensitivity. Third, potential confounders, including comorbidities or variations in neutrophil function, were not accounted for, which may have influenced NETs formation and related outcomes.

The absence of mature survival data currently precludes definitive conclusions about the long-term prognostic value of NET density. Extended follow-up studies are warranted to validate the association between NET dynamics and survival outcomes. Emerging evidence suggests that NETs may recruit or activate immunosuppressive cells such as myeloid-derived suppressor cells (MDSCs) and Treg cells. NET-derived components (e.g., histones and proteases) can activate the TLR9/NF-κB pathway in myeloid cells, promoting MDSC expansion and immunosuppressive functions. Additionally, NET-associated HMGB1 has been shown to enhance MDSC-mediated T-cell suppression via STAT3 signaling[26]. NET-derived DNA and histones may stimulate DCs to secrete transforming growth factor-β and interleukin-10, which are known to drive Treg cell differentiation[27]. In pancreatic cancer, NETs were found to upregulate PD-L1 on DCs, indirectly fostering Treg cell infiltration.

CONCLUSION

This study underscores the prognostic and predictive value of NET density in LAGC patients undergoing NACI. A low pre-NACI NET density was associated with favorable pathological and clinical responses, while an increase in post-NACI NET density was linked to poorer outcomes. These findings highlight the potential of targeting NETs as a novel therapeutic strategy to improve the efficacy of NACI in LAGC.

ACKNOWLEDGEMENTS

The authors acknowledge gratitude to all the staff who participated in this study.