Neutrophil extracellular traps in relationship to efficacy of systemic therapy for metastatic renal cell carcinoma

Abstract Background The efficacy of systemic therapy regimens, such as immune checkpoint inhibitors and tyrosine kinase inhibitors (IO‐TKI) and targeted therapy, for metastatic clear cell renal cell carcinoma (ccRCC) remains unpredictable due to the lack of effective biomarkers. Neutrophil extracellular trap (NET) plays an important role in promoting ccRCC. This study explores the NET predictive value of the efficacy in metastatic ccRCC. Methods In this retrospective study, patients with metastatic ccRCC who received targeted drugs and IO‐TKI were included. Immunofluorescence staining was utilized to quantify the levels of tissue NETs through cell counts of H3Cit(+) and MPO(+) cells. Results A total of 183 patients with metastatic ccRCC were enrolled, including 150 patients who received TKIs and 33 patients who received IO‐TKI. The levels of NETs in tumor tissue were significantly higher than in para‐tumor tissue (p < 0.001). In terms of predicting drug efficacy, a correlation between NET levels and progression‐free survival (PFS) was observed in the TKI with metachronous metastasis group (HR 1.73 [95% CI 1.02–2.91], log‐rank p = 0.037), while no correlation was observed in the TKI with synchronous metastasis group and IO‐TKI group. Regarding overall survival (OS), activated NET levels were associated with poor OS in both TKI (HR 1.60 [95% CI 1.05–2.43], log‐rank p  = 0.017) and IO‐TKI group (HR 4.35 [95% CI 1.06–17.82], log‐rank p =0.047). IMDC score (HR 1.462 [95% CI 1.030–2.075], p = 0.033) and tumor tissue NET levels (HR 1.733 [95% CI 1.165–2.579], p = 0.007) were independent prognostic risk factors for OS in patients with metastatic ccRCC.NET level was associated with poor OS in both TKI (HR 1.60 [95% CI 1.05–2.43], log‐rank p  = 0.017). Conclusions The active NET levels in tumor tissue can predict drug efficacy in patients with metastatic ccRCC who received systemic therapy. Elevated levels of NETs in tumor tissue were also associated with poor efficacy in OS.


| INTRODUCTION
Neutrophils, as the most enriched subtype of white blood cells in humans, serve an important role in the adaptive immune and tumorigenesis. 1 Neutrophil lysis releases network structures composed of nuclear chromatin and antimicrobial granules, known as neutrophil extracellular trap (NET). 2 NETs contain DNA-coated molecules such as neutrophil elastase (NE), myeloperoxidase (MPO), histones, etc. 3,4 Some specific proteins are used as quantifying biomarkers in human tissue and peripheral blood: Citrullinated Histone H3 (H3Cit), MPO, MPO-DNA complex, and NE-DNA complex. 5,6The tumor-promoting or anti-tumor role of NET can be altered, depending on the interaction of the immune system and tumor microenvironment.8][9] By immunofluorescence staining, Park et al. discovered NET presence in both primary lesions and metastasis of breast cancer. 10or patients with colorectal cancer liver metastasis, the enrichment of tumor-associated neutrophils (TAN) and NET was higher than that of normal liver tissue as well. 11urthermore, serum baseline level of MPO-DNA was significantly higher in tumor patients than in healthy population controls and was correlated with poor disease-free survival (DFS) and overall survival (OS). 11ince neither tumor mutation burden (TMB) nor PD-L1 correlated with immune checkpoint inhibitor (ICI) response, 12 the effective biomarker in tissue for efficacy prediction is needed.Neutrophils have a great impact on the progression of kidney cancer because they reflect the intrinsic immune status.For the efficacy of immune checkpoint blockade, increased neutrophil-lymphocyte ratio (NLR) was associated with worsened progression-free survival (PFS). 13NLR is also an independent risk factor for poor prognosis in kidney cancer. 14,15The International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) proposed a scoring criteria to stratify the prognostic risk of metastatic renal cell carcinoma (RCC), which annotated elevated neutrophil count as a risk factor. 16IMDC risk group is used to screen population with favorable efficacy in clinical trials of immune checkpoint inhibitors with tyrosine kinase inhibitor (IO-TKI) and targeted agents. 17,18Neutrophils exert tumor-promoting effects in a variety of mechanisms, including NET.The fibrin network and DNA framework released by NET arrest circulating tumor cells (CTCs) in kidney cancer, then help them escape from attack by the immune system and survive. 19In previous studies, we constructed gene sets based on NET and found that activated NET in pan-cancer (lung adenocarcinoma, triple-negative breast cancer, colon adenocarcinoma, and kidney renal clear cell carcinoma) indicated poor clinical outcomes. 20To date, the NET predictive value for the efficacy of systemic therapy in metastatic ccRCC remains unclear.In the present study, we collected tumor tissue specimens from patients with metastatic ccRCC who received targeted agents and IO-TKI, then quantified NET in tumor tissue.As NET plays an important role in promoting ccRCC, this study explores the NET predictive value of the efficacy in metastatic ccRCC.

| Patients selection and study design
The study retrospectively enrolled patients with metastatic clear cell renal cell carcinoma (ccRCC) who received systemic therapy from January 1, 2006 to June 1, 2022.Systemic therapy regimens referred to vascular endothelial growth factor-tyrosine kinase inhibitor (VEGF-TKI) and PD-1 immune checkpoint blockade with axitinib (IO-TKI).We collected the clinicopathological data and specimen wax of the enrolled patients.The pathologic TNM stage of all patients was confirmed by pathologists according to the 2017 American Joint Committee on Cancer (AJCC) 8th Edition. 21his retrospective study was approved by the local Ethics Committee of the National Cancer Center and waived informed consent (Ethics Approval Number: NCC2020C-122).Among the patients with metastatic ccRCC who received systemic therapy, the VEGF-TKIs included sunitinib, pazopanib, and axitinib.PD-1 inhibitors included pembrolizumab and tislelizumab.For the 33 patients who received second-line IO-TKI or TKI treatment, the prior treatments were regarding to sunitinib or pazopanib.The inclusion criteria for the cohort were as follows: 1. Patients enrolled in blinded clinical trials 2. Patients with pathological initial diagnosis of ccRCC, but the report is suspicious 3. Other malignant tumors during drug treatment 4. History of hematology-related diseases 5. Low-quality slice staining 6. Co-infection or short-term use of antibiotics 7. Non-recommended VEGF-TKI in NCCN guidelines As shown in Figure 1, a total of 949 patients received TKIs and 179 patients received IO-TKI.Through the above screening criteria, 117 cases received first-line sunitinib, 18 cases with first-line pazopanib, 15 cases with second-line axitinib, 15 cases with first-line IO-TKI, and 18 cases with second-line IO-TKI.A total of 183 metastatic ccRCC patients treated with systemic therapy were enrolled and analyzed as tissue microarray (TMA).For patients enrolled in the study, we assessed clinicopathologic data including sex, age at initial diagnosis, tumor diameter, TNM stage, histologic grade, presence or absence of renal venous tumor thrombus, PFS, OS, time from diagnosis to treatment, Karnofsky performance status (KPS), pretreatment NLR, and IMDC risk group.3][24] PFS and OS are defined as the date on which the patients start systemic therapy until disease progression or death/last F I G U R E 1 Workflow of patient screening for metastatic ccRCC.
follow-up respectively.For ccRCC patients treated with IO-TKI, the objective response rate (ORR) was also evaluated by RECIST v1.1. 25

| TMA design and immunofluorescence staining
A blank wax block of 30*25*10 mm was selected for designing a microarray of 12*8 sites.For each patient, three tumor tissue sites and one para-tumor tissue site were collected.A single TMA wax block consists of tissue from 18 to 22 patients.The constructed tissue array wax blocks were placed in a 55°C incubator for 10 min, then cooled at room temperature, and stored in a 4°C freezer.The tissue wax block was precooled at −10°C and fixed on a microtome for 4 μm continuous slices.Hematoxylin-eosin (HE) staining was performed to confirm specimens.TMAs were heated at 60°C for 1 h, dewaxed at room temperature, and washed for 30 min.2.5% donkey serum (abs935, Absin Biotechnology) was used for incubation.For primary antibody incubation, 1:100 dilution of anti-myeloperoxidase antibody (MAB3174, R&D Systems) and 1:100 dilution of anti-histone H3 (citrulline R2 + R8 + R17) antibody (ab5103, Abcam) were used and incubated with TMA overnight at 4°C.For second antibody incubation, 1:200 dilution of Alexa Fluor 488 (711-545-152, Jackson ImmunoResearch) and 1:200 dilution of Alexa Fluor 594 (715-585-150, Jackson ImmunoResearch) were used.DAPI was used for nuclear staining (Boster Biological Technology), and a laser scanning microscope (KF-PRO-020-HI, China) was used to observe and scan the TMA.Random 5 fields of view were captured for each site in K-Viewer software.The co-expression cell count of H3Cit(+) and MPO(+) was used to quantify tissue NET level.Two independent researchers measured, calculated, and averaged cell counts.Any discrepancies were discussed and solved by a senior researcher (J.Shou).

| Statistics
The optimal cutoff values of NET level were determined using X-tile 3.6.1. 26The statistical analyses were performed in SPSS 23.0 and GraphPad Prism 8.0.2, and the immunofluorescence was quantified in K-Viewer 1.7.0.Counts and percentages were used for the features of the patient's clinicopathological data.The Mann-Whitney U test was used to compare the expression of NET in tumor tissue and para-tumor tissue.PFS and OS survival were analyzed with the Kaplan-Meier method and log-rank test, and the difference between groups was shown as hazard ratio (HR) and 95% confidence interval (CI).Statistical analysis was identified as significant when p < 0.05.
By immunofluorescence staining, we analyzed NET levels in metastatic ccRCC.Notably, MPO expression was mainly located on the cell membrane, whereas H3Cit was located in the nucleus (Figure 2D). Figure 2E shows H3Cit and MPO levels in matched tumor tissue and para-tumor tissue of the same patient.Among all patients, NET was more activated in tumor tissue than in adjacent normal tissue (p < 0.001).By calculating NET(+) cell counts, we analyzed the correlation between NET level in tumor tissue and clinicopathological features (Figure 3).The results suggested that a higher level of NET was significantly correlated with tumor N stage (p = 0.02), and NLR (p = 0.02), but there was no significant correlation with age, ISUP grade, tumor T stage, tumor M stage, venous tumor thrombus, and IMDC risk (p > 0.05).The full results of clinicopathological features are presented in Figure S2.

| NET predicts systemic therapy efficacy in patients with metastatic ccRCC
The median PFS was 14.0 months and the median OS was 32.6 months in the TKI group.The median PFS was 9.0 months, and the median OS was not reached in the IO-TKI group.For the targeted therapy group with synchronous metastasis (Figure 4A), no correlation between the NET level and PFS was observed (HR 1.26 [95% CI 0.76-2.07],log-rank p = 0.35).Notably, in the TKI group with metachronous metastasis (Figure 4B), the lower NET level was associated with better PFS (HR 1.73 [95% CI 1.02-2.91],log-rank p = 0.037).Overall,

| DISCUSSION
Consistent with neutrophil properties, NET can induce cancer-promoting or cancer-suppressing effects in tumors.This study mainly explored the cancer-promoting effect of NET in metastatic ccRCC and revealed the relationship between NET and clinicopathological features.By analyzing the biomarker level of NET in tumor and para-tumor tissue, we found NET could be a predictor of drug efficacy and prognosis.H3Cit and MPO are specific markers of NET and are used to quantify NET in tissue and peripheral blood.Previous studies in pan-cancer indicated that H3Cit and MPO were effective prognostic markers. 20,279][30] Compared with using H3Cit or MPO as biomarker alone, we quantified NET with cell counts co-expressed with H3Cit and MPO simultaneously, with better reliability.Immunofluorescence staining showed that NET in tumor tissue was more active than adjacent normal tissue.Neutrophil infiltration in kidney cancer has been considered a signature of poor prognosis. 31As one of the mechanisms by which neutrophils function, NET activity is consistent with neutrophil infiltration.Notably, tissue NET level was associated with tumor N stage at initial diagnosis.Secondary lymph node organs serve as sites for the colonization and proliferation of immune cells, providing a microenvironment for the interaction of innate and adaptive immune systems.
In the peripheral circulation, neutrophils can be recruited in tumor-draining lymph nodes and NET can be induced when stimulated by chemokines such as IL-17. 32,33To date, the underlying mechanism of lymph node metastasis in kidney cancer remains unclear, and NET might be a potential trigger.In addition, there was no significant correlation between NET and initial M stage (synchronous metastasis vs metachronous metastasis), indicating that the tissue NET activity of primary tumor might not be a key inducer in metastasis.However, NET in peripheral blood promoted hematogenous metastasis by enveloping CTCs and assisting them in survival and migration. 19he recommended first-line treatment for metastatic ccRCC is IO-TKI, but there are no effective predictors of drug efficacy.Labriola et al analyzed tumor specimens of metastatic ccRCC treated with IO and found no correlation between therapy response and traditional efficacy markers such as TMB and PD-L1 expression. 12As one of the parameters of IMDC risk, neutrophil counts are of great significance for patient prognosis stratification and guiding treatment.We assumed that NET might be a potential predictor of drug efficacy.Chen et al. constructed a risk model of six NET-related genes in head and neck squamous cell carcinoma, and the low-risk group received more benefits from anti-PD-1 therapy than the high-risk patients. 34Interestingly, tissue NET level was also a negative predictor of OS with both TKI and IO-TKI for metastatic ccRCC.However, no association between PFS and ORR was observed with NET in the IO-TKI group.In patients who received IO+TKI therapy, no difference in PFS was observed between the NETs-high and NETs-low groups, while the OS in the NETs-low group was significantly longer compared to the NETs-high group.The proportion of patients who received first-line and second-line treatment in the IO-TKI group was close, which might affect the difference in PFS.Another possible explanation is that NET has an effect on prognosis in RCC beyond the medical treatment. 20,27Further exploration with long-time follow-up in large-scale cohorts is needed.A total of 33 patients were enrolled in the IO-TKI group, with a small sample size and ratio of first-line versus second-line at 5:6, and the results of this study may be altered by enlarging the sample size and optimizing the sequence of treatment.Santoni et al. revealed that host immune status was associated with VEGF-TKI efficacy in metastatic kidney cancer that occurred more than 5 years after radical nephrectomy, with a median PFS of 25.1 months (95% CI 11.7-NA) for NLR <3 patients and 15.8 months (95% CI 7.1-31.3)for NLR ≥ 3. 35 Consistently, elevated NET level with metachronous metastasis was associated with poor PFS in the TKI group.Therefore, NET levels in primary tumor tissue may reflect the host immune response to tumor, while patients with active local immune status respond better to TKIs.In the clinical dissection of tumor, tumor tissue NET can be assessed to guide the potential dose of targeted agents.The IMDC risk score stratifies patients according to median OS and guided treatment.Active NET in tumor tissue was associated with poor OS in metastatic ccRCC.Multivariate COX analysis showed that NET and IMDC had a greater impact on prognosis than risk factors such as primary tumor stage, histological grade, and venous tumor thrombus.After progression to distant metastasis, T A B L E 3 COX analysis of overall survival in patients with metastatic ccRCC (*p<0.05).In this study, NET is more predictable for PFS and OS in the TKI group than in the IO-TKI group.Immune checkpoint blockade of PD-1 and PD-L1 can restore the function and survival of exhausted CD8+ T cells, and then activate the cytotoxicity toward tumor cells. 36NET can interact with infiltrating CD8+ T cells to promote an immunosuppressive tumor microenvironment.Neutrophils isolated from humans could form NET and induce exhaustion and dysfunction of CD8+ T cells. 37For colorectal cancer, DNase I inhibits NET formation in vivo and reverses anti-PD-1 blockade resistance through increasing CD8+ T cell infiltration and cytotoxicity. 38To date, it remains unclear if NET is associated with drug resistance of VEGF-TKI, though NET plays a role in promoting angiogenisis by interacting with angiogenic factors (VEGF-A and angiopoietins). 39,40Therefore, NET effects on immune checkpoint blockade with anti-angiogenisis might be more complex and less predictable than using ICI or TKI alone.

Variation
Neutrophil infiltration is an indication of poor prognosis for ccRCC, and this study revealed NET as a potential role of neutrophils affecting the tumor microenvironment.With long-time follow-up of patient cohorts, NET can be a biomarker of systemic therapy efficacy.Some limitations of the research should be stated.First, the sample size of IO-TKI cohort was small, and the ratio of first-line and second-line was 5:6, which might not represent the real-world experience.Second, due to tissue necrosis and low staining quality after treatment, matching sample analysis before and after treatment was missing.Third, p-value of 0.048 in the PFS analysis of all targeted therapy patients may alter conclusions when adjusting the sample size.

| CONCLUSIONS
For metastatic ccRCC, NET was more active in tumor tissue than in para-tumor tissue.The lower level of tumor tissue NET predicts better drug efficacy in patients treated with TKI and IO-TKI.NET could be the mechanism of neutrophil-promoting tumor, and NET formation in ccRCC tumor tissue was associated with poor clinical outcomes.Quantifying NET may optimize the treatment strategy for kidney cancer.

1 .
Histologically confirmed as ccRCC by pathologists 2. Previous first-line/second-line system therapy (VEGF-TKI and IO-TKI) 3. Available specimen wax archived in the Department of Pathology 4. Complete medical records, laboratory tests, imaging examinations, and pathological data Patient exclusion criteria were as follows:

F I G U R E 3
Association of NET with clinicopathological features in ccRCC.(A) N stage (B) Neutrophil-to-lymphocyte ratio (NLR) (C) M stage (D) IMDC risk I. (*p < 0.05, ***p < 0.001, ns: not significant).F I G U R E 4 Association between NET and efficacy of targeted therapy/IO-TKI in metastatic ccRCC.(A) Progression-free survival (PFS) in the TKI group with synchronous metastasis (B) PFS in the TKI group with metachronous metastasis (C) PFS in all patients of the TKI group (D) PFS in the IO-TKI group (E) Objective response rate (ORR) in the IO-TKI group (ns: not significant).

2
COX analysis of targeted therapy group with metachronous metastasis (*p<0.05).F I G U R E 5 Association between NET and survival prognosis in ccRCC.(A) Overall survival (OS) in the TKI group with synchronous metastasis (B) OS in the TKI group with metachronous metastasis (C) OS in the TKI group (D) OS in the IO-TKI group.