Neutrophil extracellular traps in breast cancer and beyond: current perspectives on NET stimuli, thrombosis and metastasis, and clinical utility for diagnosis and treatment

Figure 2 shows the integration of multiple NET stimuli and the downstream effects of NET release including enhanced VTE and metastasis. NETs have been identified as a prognostic indicator of VTE and are at least partially responsible for the hypercoagulable state observed in cancer patients. It is estimated that women with breast cancer are three to four times more likely to develop VTE compared to age-matched women without breast cancer [97]. VTE occurring in breast cancer patients has also been linked to reduced patient survival and tumor recurrence. Mauracher et al. [26] recently observed that high plasma levels of NET marker, citH3, were predictive of an increased risk of VTE for 2 years post diagnosis or relapse in a cohort of nearly a thousand cancer patients; interestingly, levels of circulating DNA were only predictive of increased VTE risk during the first 3 to 6 months. Of the tumor sites examined, brain, lung, and breast showed the highest frequency of patients whose sera contained elevated citH3 levels. For these patients, the 2-year risk of VTE was 14.5%, as opposed to 8.5% for patients lacking elevated citH3. An increase in citH3 of only 100 ng/mL was found to translate to a 13% higher risk of VTE, suggesting that even mild NETosis may severely impact prognosis. The hypercoagulability characteristic of cancer patients has been shown to be largely NETosis dependent in breast cancer models [21]. Demers et al. [19] showed that G-CSF seems to further drive the prothrombotic state by priming neutrophils for NETosis in a 4T1 mammary carcinoma mouse model. Tumor-bearing mice experienced a significant decrease in both platelet and neutrophil counts consistent with thrombus formation and had reduced tail bleed times. In tumor-burdened mice, the highest levels of citH3 were present in the later stages of disease.

It has also been suggested that NETs promote the formation of the pre-metastatic niche, at least partly by entrapping circulating tumor cells in their web-like structure, providing a site from which these cells may then extravasate. Cools-Lartigue et al. [10] showed via intravital microscopy that NETs were able to arrest circulating tumor cells in the pulmonary and hepatic microvasculature. A septic state was induced in C57BL/6 J mice prior to intrasplenic injection of H59 Lewis lung carcinoma cells. Micrometastases were observed within 48 h, with both NE inhibitor and DNase I treatment abolishing this effect; non-septic mice showed few micrometastases, suggesting that pro-NETotic stimuli are required to enhance NET-mediated metastasis. Park et al.’s [27] recent study shows that 4T1 breast cancer cells injected into the tail vein of LysM-EGFP mice were found within lungs and caused NET formation; immunofluorescence staining of lung tissue sections showed via DNA and NE fluorescence that tumor cells were sufficient to provoke increased rates of lytic NETosis for up to 4 days post injection. DNase I-coated nanoparticle treatment reduced tumor invasion in vitro and the number and size of lung metastases in vivo.

Interestingly, surgical resection aimed at decreasing tumor burden can actually promote metastasis through neutrophil activation and subsequent NET formation. Increased NETosis in patients undergoing liver resection due to metastatic colorectal cancer correlates to markedly lower cancer-free survival [14]. This effect of increased metastasis following surgical stress was reproduced in mouse models and was abolished by DNase I treatment or inhibiting PAD4 to dissociate NETs or prevent their release, respectively. Neutrophil HMGB1 release occurred concurrently with pro-NETotic stimulation, and HMGB1 was also associated with NETs. This represents a means by which NETs may directly activate platelets and other neutrophils, eventually forming a thrombus. Circulating tumor cells (the presence of which may be increased by surgical disruption of the primary tumor) could then be captured due to partial vessel blockage and the coagulating microenvironment around the NET. Simultaneously, the capacity of NETs to damage endothelial cells likely enables arrested tumor cells to adhere to the activated endothelium, eventually extravasating and establishing a new metastatic site. Interestingly, NETs have also been shown to activate dormant single breast tumor cells in mouse lungs, which can then lead to metastasis development. Cancer cell activation from dormancy is thought to occur via the remodeling of extracellular matrix due to NET-associated NE and is further facilitated by G-CSF [13].

It is reasonable to conclude that tumor-driven NETosis alone, even without surgical stress or major infection, can also serve to drive metastasis. Pro-NETotic factors are known to be overexpressed by many tumor lines, and multiple murine breast cancer models have been shown to promote NETosis. However, much remains unclear about the specific mechanism in which cancer promotes metastasis through NET formation. It is currently unknown whether NETs predominantly contribute to metastatic establishment via endothelial damage or direct sequestration of tumor cells. Additionally, little is known about the timeline of NET generation. It is possible that the primary tumor site must reach enough development to elicit NETosis, which then promotes the establishment of metastases. Alternatively, pioneering tumor cells may secrete pro-NETotic factors which then provoke a NET-induced inflammatory state from surrounding neutrophils, favoring tumor cell invasion and further sequestration of circulating tumor cells.

The ability to detect NETs would likely be of significant prognostic use in differentiating patients at higher risk of metastatic progression or VTE, thereby enabling clinicians to better personalize treatment regimens. To develop a clinical screening tool for NETs, a standardized definition of “normal” levels of NETosis would need to be established and has not yet been presented in the literature. The simplest means of in vivo NET detection involves measurement of NET-associated products in the blood such as circulating cell-free DNA, citH3, NE, and MPO. For example, free circulating DNA has been quantified in both colorectal and breast cancer patient serum samples via a simple nucleic acid staining assay [98, 99]. However, even though circulating DNA is known to correlate with breast tumor size and malignancy [100], it lacks specificity in measuring NETosis. An increased amount of DNA in cancer patient serum can also be attributable to other factors such as apoptotic and necrotic cells. Measuring circulating MPO/DNA conjugates is more specific for NET formation than evaluation of cell-free DNA alone [101]. Citrullinated histone H3 (citH3) is formed as a result of PAD4-mediated citrullination during NET formation and represents the most specific biomarker for circulating NETs [26]. In addition, citH3 may be of prognostic significance, as Thålin et al. [102] observed that high plasma content of citH3 was a significant indicator of short-term mortality in late-stage cancer patients, even when compared to severely ill patients without cancer. Additionally, IL-8 levels were found to correlate with levels of citH3. Since higher levels of IL-8 would result in increased neutrophil recruitment, it would be reasonable to conclude that this higher density of neutrophils would subsequently lead to increased NETosis. Despite this, other markers associated with NETs including NE and MPO were not found to differ significantly between severely ill patients with and without malignancy; however, these neutrophil-derived enzymes can be independently released during neutrophil degranulation in the absence of NET formation, and therefore may not be reliable NET-specific biomarkers. Indeed, citH3 seems to be the most consistent indicator of NETosis. While levels of other markers may provide useful insight into neutrophil behavior, citH3 is highly specific to NETosis and thus would be valuable in understanding variances between other NET-associated biomarkers. CitH3 levels are also predictive of VTE risk in newly diagnosed patients, further supporting its diagnostic utility [26].

The development of clinical therapies specifically targeting NETs in cancer is in its infancy. Inhibition of NETosis has been achieved through several means, though these vary in their potential for clinical therapies. For instance, DNase I treatment degrades NETs and results in a loss of the web-like structure and a reduction in the capacity to promote metastasis in several studies [10, 14, 21, 34]. In addition, DNase I has been shown to decrease tumor volume in rats when injected intramuscularly or intraperitoneally in conjunction with other proteases (papain, trypsin, and chymotrypsin) [103]; however, it is not known whether these effects are due primarily to NET inhibition. Currently, DNase I is used clinically in the treatment of cystic fibrosis, as it decreases the NETosis-mediated buildup of mucous viscosity, resulting in improved lung function [6]. However, in this context, DNase I is delivered via nebulizer, which would likely be ineffectual in most cancer treatments, though it would be fascinating to observe whether nebulized DNase I would have a preventative effect on lung metastasis. Additionally, DNase I injection may have off-target effects, including compromising the immunoprotective function of NETs.

Inhibition of components integral to NETosis, such as NE or PAD4, would likely have similar off-target effects due to their involvement in other key pathways, potentially disrupting normal neutrophil function. Small molecule inhibitors of PAD4 for NET inhibition are under active investigation and include Cl-amidine and F-amidine, irreversible inhibitors that inactivate calcium-bound PAD4 [104]. However, these lack specificity and interact with other PAD-family enzymes. Recently, Lewis et al. [105] synthesized two reversible inhibitors which overcome this hurdle, GSK199 and GSK484, both of which exhibit high specificity for PAD4 and inhibit NETosis in both mouse and human neutrophils. GSK484 was recently shown to prevent tumor-associated renal dysfunction in mice, which was determined to be NET-mediated; the inhibitory effects of GSK484 were as effective as DNase I [106]. Additionally, a recent study by Yazdani et al. [107] indicates that PAD4-knockout mice challenged with subcutaneous tumor injection of colorectal and hepatocellular carcinoma tumor cells experienced slower tumor growth and smaller metastases similar to mice treated with daily DNase I injection. NETs were not observable in excised tumor tissue in PAD4-knockout mice. Finally, the authors showed that NETosis at the primary tumor site may contribute to tumor cell survival through enhanced mitochondrial biogenesis. This data further supports the need to develop NET-targeting treatments, as these would be of great therapeutic benefit in both the context of the primary tumor site and the pre-metastatic niche.

Efforts targeting cell adhesive molecules, such as P-selectin, could also prove problematic. Though successful results of a stage II clinical trial for the use of the P-selectin inhibitor crizanlizumab in sickle cell anemia to prevent vaso-occlusion were recently published [108], it would be reasonable to conclude that such a therapy may interfere with leukocyte function. Though P-selectin and PSGL-1 antibodies have been shown to inhibit NETosis in mice [41], the disruption of leukocyte adhesion molecule binding capacity could decrease neutrophil recruitment in response to infection in cancer patients already suffering from an immunocompromised state. Off-target effects could potentially be mitigated via the development of new, more specific delivery vehicles, such as functionalized, targeted nanoparticles.

Alternatively, the adaptation of FDA-approved drugs could facilitate the development of effective anti-NET treatments. For instance, the inhibitory effect of aspirin on NETs has yielded some promising results in animal models. Lapponi et al. [109] showed that aspirin prevented NET-induced injury of the lung endothelium by inhibiting platelet activation and subsequent NET formation in mice. The inhibitory effect of aspirin on NF-κB, an inflammatory transcriptional regulator that plays a role in some pathways promoting NETosis, was also demonstrated. The authors found that aspirin treatment effectively inhibited NETs in human neutrophils in vitro and resulted in higher bacteria counts in infection-burdened mice in vivo, suggesting a loss of normal NET functionality. There is evidence to support the use of aspirin in clinical treatment. In one meta-analysis, patients using aspirin daily had significantly reduced mortality and risk of distant metastases for adenocarcinomas. Interestingly, this effect did not appear to be dose dependent [110]. Aspirin has also been shown to be effective in reducing metastasis in patients suffering from breast cancer specifically [111].

Another FDA-approved drug, hydroxychloroquine, originally used to treat malaria, has been shown to inhibit NETosis [17, 54, 112]. While the mechanism behind NET inhibition by hydroxychloroquine is unclear, it may be related to autophagy inhibition [113]. However, a phase II clinical study on patients with advanced pancreatic cancer produced little clinical effect. The authors do suggest, however, that combination therapy may prove more effective [114]. Furthermore, use of hydroxychloroquine as a neoadjuvant treatment in earlier stage disease holds significant promise [115]. Remarkably, and perhaps not coincidentally, hydroxychloroquine also inhibits leukocyte phagocytosis [116]. Thus, it may be possible that hydroxychloroquine could inhibit neutrophil uptake of tumor-derived EVs, thus reducing NETosis. However, the precise mechanism by which this uptake occurs is unknown, as are the mechanisms behind tumor-derived EV stimulated NETosis. Due to the associated complications of NETs including increased VTE risk and metastasis, which are both negatively associated with breast cancer patient outcome, it is crucial for future research efforts to focus on further investigation of new specific targets to prevent NET formation.