Neutrophils in cancer: prognostic role and therapeutic strategies

Neutrophilic granulocytes (neutrophils) account for 50–70% of all leukocytes and depend on a sequential process of maturation in the bone marrow that provokes the conversion of myeloblasts to segmented neutrophils [7]. Maturation depends on different stimulating factors including the granulocyte–macrophage-colony stimulating factor (GM-CSF) and the granulocyte-colony stimulating factor (G-CSF), two of the most relevant growth factors that control such maturation process. Neutrophil maturation includes: myeloblast, promyelocyte, myelocyte, metamyelocyte, band neutrophil and, finally, segmented neutrophils [7‐9]. Neutrophil lifespan is altered in cancer and it is associated with maturation, extending from 7 h in normal conditions to 17 h in cancer [8, 9]. Of note, the majority of neutrophils remain in the bone marrow, for instance in mice only 1–2% circulate in the peripheral blood [10]. Release of neutrophils from the bone marrow depends on a series of stimulating factors and cytokines including IL-23, IL-17, G-CSF; and CXC chemokine receptors [11, 12]. The generation and maturation of neutrophils have important implications: from the design of therapeutic strategies to the utilization of their expression as a prognostic biomarker.

The role of neutrophils in cancer is multifactorial and not fully understood. Neutrophils reflect a state of host inflammation, which is a hallmark of cancer [13]. They can participate in different stages of the oncogenic process including tumor initiation, growth, proliferation or metastatic spreading [8, 9]. In general neutrophils play a central role in inflammation within the tumor as they are attracted by CXCR2 ligands like CXCL1, CXCL2 and CXCL5, among others [9, 14]. Tumor initiation can be promoted by the release by neutrophils of reactive oxygen species (ROS), reactive nitrogen species (RNS) or proteases, among others [15]. A relevant mechanism is the induction of angiogenesis. Indeed, neutrophil depletion or CXCR2 blocking decrease vessel formation [15]. Some factors that mediate the formation of angiogenesis include the production of vascular endothelial growth factor A (VEGFA), prokineticin 2 (PROK2), or MMP9, among others [16, 17]. Neutrophils can facilitate tumor proliferation by attenuating the immune system. CD8+ T lymphocyte antitumor response can be suppressed by nitric oxide synthase (iNOS), or arginase 1 (ARG1) released by neutrophils under stimulation by TGFβ (Fig. 1a) [18, 19]. They also produce MMP9 that has an important role in tumor initiation. In addition tumor proliferation can be mediated by degradation of the insulin receptor substrate 1 (IRS1), and activation of PI3K signaling due to the transfer of neutrophil elastase to cancer cells [20]. Of note, production of iNOS can also be stimulated in neutrophils by the upregulation of the tyrosine kinase receptor MET [21]. Finally, neutrophils can also motivate the metastatic spreading by inhibiting natural killer function and facilitating the extravasation of tumor cells (Fig. 1a) [22, 23]. As can be seen here, the role of neutrophils in cancer is complex, and can be context and tumor dependent. Indeed, some studies have even shown how neutrophils can antagonize the metastatic spreading, as is the case in lung cancer [24]. It should be mentioned that this difference in function could be linked with the existence of various neutrophil subpopulations [8, 9].

A different population of cells that is generated in the bone marrow from myeloid precursors is the myeloid-derived suppressor cells (MDSC). They migrate to the tumor guided by several stimulating factors, being the chemokines CCL2 and CCL5 the most studied [25‐27]. There are two different type of cells, polymorphonuclear MDSC (PMN-MDSC), that are morphologically similar to neutrophils, and monocytic MDSC (M-MDSC), that are similar to monocytes [27]. Of note, MDSC have a potent suppressor capacity in human cancer [27].

Given the various roles of neutrophils in cancer development and progression, several groups have recently explored the role of neutrophils and other markers of host inflammation on clinical outcomes. Thus, an elevated neutrophil count is an adverse prognostic factor incorporated in a contemporary prognostic score for metastatic renal cell carcinoma (mRCC) treated with targeted therapy [28]. Furthermore, most data are available for the ratio of neutrophils to lymphocytes measured in the peripheral blood, the so-called neutrophil-to-lymphocyte ratio (NLR). An elevated NLR is associated with worse outcomes in many solid tumors, both in early and advanced stage of cancer [3]. Moreover, an elevated NLR is associated with lower response rates in castration-resistant prostate cancer treated with abiraterone or docetaxel [29, 30] and a decline during treatment with cabazitaxel was shown to be associated with longer overall survival [31]. Also, an early decrease of NLR in response to targeted treatment appears to be associated with more favorable outcomes and higher response rates in patients with mRCC, even after adjustment for known prognostic factors including NLR at baseline [5]. In contrast a rising NLR during the first weeks of treatment had the opposite effect. These findings make NLR a biomarker easy to evaluate, and that have potential for the identification of early responders. Table 1 summarizes all the meta-analyses studies performed evaluating the role of NLR expression and outcome in cancer.

Not only elevated numbers of neutrophils in peripheral blood as reflected by NLR are of prognostic relevance, but also their presence in the tumor can be associated with clinical outcome. The expression of neutrophils in the tumor has been linked with detrimental outcome in some indications like in renal cell carcinoma, head and neck cancer or esophageal carcinoma [6, 32, 33]; whereas in other indications it has been associated with better survival [34, 35]. In this context, it should be noted that what mainly impact the worse outcome is the presence of inflammation within the tumor, and the assessment of neutrophils is an indirect measure of this and can vary among tumor types.

To avoid the deleterious effect of neutrophil expression in cancer, strategies intended to reduce its activity have been explored and some have entered clinical evaluation. Table 2 describes characteristics of all ongoing clinical studies. The first approach is to target factors involved in the late stage process of neutrophil maturation. Indeed, some factors can be produced by tumor cells and this may favor the metastatic spreading mediated by neutrophils (Fig. 1b) [36, 37].

Strategies explored to inhibit neutrophils include the inhibition of CXC receptors like CXCR2 that are associated with the migration of neutrophils to tumor areas. CXCR1 and CXCR2 inhibitors are currently in clinical development in cancer [38, 39]. Inhibition of the IL-23 and IL-17 axis is another approach, as IL-17 and IL-23 stimulate expansion of neutrophils mediated by G-CSF (Fig. 1b) [40]. However this approach has not reached yet the oncology field, but drugs targeting these cytokines are approved for the treatment of other medical conditions like psoriasis [41, 42].

Another tactic is to directly inhibit G-CSF and therefore decrease the amount of neutrophils, strategy that has shown efficacy in preclinical models [43]. Agents against this target are currently in its early stage of clinical development in cancer [44]. However, it is unclear if the inhibition of G-CSF and subsequent reduction of neutrophils can have an impact in patient infections, mainly in those under treatment with chemotherapy. Recently, preclinical studies have shown that neutrophil Alox5 inhibition can also decrease metastatic lung dissemination (Fig. 1b) [45].

There are many areas of uncertainty regarding the evaluation of neutrophils as a prognostic marker or in the development of compounds against neutrophils.

Although the NLR is considered as an easy, inexpensive and reproducible biomarker associated with clinical outcome for the majority of tumors some questions remain to be resolved. For instance, the identification of adequate cut-offs, or longitudinal evaluations over a treatment period of time could add more accurate information. Indeed, modifications over time can inform about treatment efficacy. Similarly, comparison of this ratio with the expression of cytokines in blood or the evaluation of neutrophil expression in tumors could help to improve its prognostic or predictive value.

It is also challenging how to optimize therapies against neutrophils. Some studies have suggested an augmented effect when neutrophil targeting agents, CXCR2 inhibitors or anti-Ly6G, were combined with checkpoint inhibitors [46, 47]. Table 2 provides a list of compounds in clinical development. Similarly combinations of antiangiogenic agents with neutrophil targeting agents could be another tactic as resistance to antiangiogenic agents has been linked with neutrophil stimulation [48]. In the case of combination strategies with chemotherapy, data is contradictory with studies supporting the efficacy of the combination and others showing a detrimental effect [49]. Of note clinical studies in combination with chemotherapy are also present. Like with any new therapeutic agent, identification of a biomarker or a specific clinical scenario could undoubtedly help to identify responsive patients. Finally, given the dual role of neutrophils in cancer, the consequences of depleting tumor promoting and anti-tumor neutrophils are unclear, reinforcing the importance for patient identification and biomarker discovery.