Deciphering Mechanisms of Resistance to Macrophage-Targeted Therapies Daniela F

Published OnlineFirst November 28, 2016; DOI: 10.1158/1078-0432.CCR-16-0133

Molecular Pathways Clinical Cancer Research Molecular Pathways: Deciphering Mechanisms of Resistance to Macrophage-Targeted Therapies Daniela F. Quail1,2 and Johanna A. Joyce3,4

Abstract

Tumor-associated macrophages (TAMs) are a major cellular clinical strategies to target macrophages in cancer and discuss component of numerous tumor types. TAM-targeted therapies potential explanations for why some strategies are effective include depletion strategies, inhibiting their effector functions while other approaches have shown limited success. Clin Cancer or reprogramming toward an antitumorigenic phenotype, with Res; 23(4); 1–9. 2016 AACR. varying degrees of efﬁcacy. Here, we review preclinical and

Background recruited in response to cytokines and chemokines that are released as a result of tissue imbalance or injury. In cancer, Tumor-associated macrophages (TAMs) are a major immune macrophages can be hijacked by the tumor, even though many component of many types of cancer and can account for up to of their fundamental biological pathways are maintained. Within approximately 30% of cells within the tumor microenvironment the tumor microenvironment, macrophages can be "educated" by (1, 2). In adult tissues, macrophages can accumulate in tumors the tumor to promote cancer development, progression, and through mobilization and recruitment from systemic reservoirs, metastasis. Therefore, understanding the mechanisms of mobili- differentiation from monocyte precursors, and in some cases, zation, differentiation, and activation in normal macrophage proliferation within local microenvironments. In addition to their biology is critical for designing effective targeted strategies against abundance, the activation status and phenotypes of TAMs are also TAMs in cancer. important considerations for tumor biology. TAMs are highly Macrophages arise from differentiation of precursor cells plastic cells and can adopt either pro- or anti-inflammatory through at least two distinct ontogenetic processes, as determined activation states in response to cytokine exposure. In turn, TAMs through mouse genetic studies. Yolk sac–derived macrophages produce a variety of factors, including growth factors, cytokines, (F4/80hi) arise during early developmental hematopoiesis and and proteases, which can contribute significantly to regulating seed tissues prior to birth. Bone marrow–derived macrophages disease progression. As such, they are attractive targets for recali- (F4/80lo) arise from monocyte precursors and are recruited to brating immune responses within the tumor microenvironment. tissues in response to inflammation. Another source of macro- Not surprisingly, drugs that aim to either deplete or reprogram phages within adult tissues is through local proliferation in macrophages are meriting considerable attention. Here, we will response to specificinflammatory stimuli (3–6). review key components of macrophage biology, discuss thera- Macrophage differentiation is largely driven by colony-stim- peutic approaches that are currently being employed to target ulating factor-1 (CSF-1 or M-CSF) and IL34 signaling via their macrophages in both clinical and preclinical settings, highlight cognate receptor, CSF-1R, which is located on the plasma mem- mechanisms by which TAMs can develop resistance, and delineate brane (7). These same factors are also involved in mobilization strategies to overcome such resistance. of monocytes from systemic reservoirs and their recruitment to tissues when required. In CSF-1–null mouse models (e.g., Key facets of macrophage biology osteopetrotic op/op mice), monocytes and macrophages are In normal tissues, macrophages play important roles during depleted in several tissues, and there is a complete deficiency of tissue homeostasis. Tissue-resident macrophages, including liver bone macrophages (osteoclasts), resulting in aberrant bone Kupffer cells, brain microglia, skin Langerhans cells, etc., are remodeling (8, 9). Therefore, inhibiting CSF-1R signaling is a important for maintaining steady-state homeostasis within a major focus of current macrophage-targeted therapies, which given organ, whereas peripherally derived macrophages are will be discussed here in detail. Macrophages can be activated by a variety of different cytokines within the microenvironment. Perhaps the most classical groups are 1Cancer Biology and Genetics Program, Memorial Sloan Kettering Cancer Center, Th1 and Th2 cytokines, which give rise to pro-inflammatory and New York, New York. 2Goodman Cancer Research Centre, Department of anti-inflammatory states, respectively. Th1-activated macrophages Physiology, McGill University, Montréal, Quebec, Canada. 3Ludwig Institute for (classical activation/M1-like) are associated with antitumorigenic 4 Cancer Research, University of Lausanne, Lausanne, Switzerland. Department functions, whereas Th2-activated macrophages (alternative activa- of Oncology, University of Lausanne, Lausanne, Switzerland. tion/M2-like) tend to be associated with protumorigenic pheno- Corresponding Author: Johanna A. Joyce, University of Lausanne, Chemin des types (10, 11). In recent years, this polarization model has been Boveresses 155, Lausanne, CH 1066, Switzerland. Phone: 412-1692-5937; Fax: regarded as oversimplistic, as studies have demonstrated that 412-1692-5995; E-mail: [email protected] macrophages can adopt overlapping M1-like and M2-like gene doi: 10.1158/1078-0432.CCR-16-0133 expression programs (12). In light of these observations, it is now 2016 American Association for Cancer Research. appreciated that macrophages adopt a more spectrum-based

www.aacrjournals.org OF1

Quail and Joyce

activation state (10, 11, 13–15), whereby different cytokines can models can have variable effects on outcome. For example, in integrate to induce a broad and dynamic range of gene expression proneural glioblastoma multiforme, Pyonteck and colleagues patterns and biological functions. showed that early intervention with a CSF-1R inhibitor, BLZ945, prevented tumor initiation and that late intervention caused a Targeting the CSF-1R pathway in cancer: Successes and rapid and robust tumor debulking after just 1 week of treatment limitations (16). In this study, microglia were depleted in the normal brain as Given its pivotal role in regulating multiple aspects of macro- expected, but not within the tumor region due to the presence of phage biology, inhibiting CSF-1R has been the focus of many glioma-enriched survival factors. Surviving TAMs were instead preclinical cancer studies (Fig. 1A and B). It has been shown by reeducated by CSF-1R inhibition to downregulate protumori- several groups that blocking CSF-1R signaling in mouse tumor genic/M2-like activation programs (16).

Disrupting chemokine axes

Enhancing immune responses D Neutralizing Ab TC C TCR T cell MHCII Chemokine gradient (e.g., CXCL12, CCL2)

Recruitment FGK45 CD40L CD40 Targeting TEMs agonist E Neutralizing Ab Other checkpoint inhibitors? TIE-2 B Neutralizing Ab Macrophage (e.g., RG7115) CSF-1R Ang1/2 Proangiogenic factors (e.g., VEGF, bFGF) CSF-1 BLZ945 PLX3397 GW2580 A Reeducation Targeting CSF-1R M1-like M2-like

Il12, Il23, Tlr4, Arg1, Mrc1, Chil3, Th1 cytokines MHCII, Cd86 F13a1, Retnla Th2 cytokines (e.g., IL1, TNFα, IFNγ) (e.g., IL4, IL13, IL10)

Macrophage reprogramming

Figure 1. Diversity of macrophage-targeting strategies in cancer. A, Rather than depleting macrophages, therapies aimed at reprogramming or reeducating M2-like macrophages to adopt antitumor M1-like phenotypes or enhanced antigen-presenting capacity have demonstrated survival benefits in preclinical models (16, 19, 31, 41,67). B, Most macrophage-targeted therapies are currently focused on CSF-1R inhibitors, which either deplete or reprogram macrophages depending on the context. For example, in brain tumors, CSF-1R inhibitors have been shown to reprogram macrophages toward an antitumor phenotype (16, 19), whereas in breast cancer, CSF-1R inhibitors have been shown to reduce macrophage numbers and thereby improve responses to cytotoxic therapies (17, 23–25). C, Macrophages can be targeted to reprogram the immunosuppressive tumor microenvironment and consequently enhance antitumor T cell responses. For example, in models of pancreatic ductal adenocarcinoma, CD40 agonists have been shown to enhance antigen presentation through MHCII expression on myeloid cells, leading to enhanced antitumor outcomes (39), and CSF-1R inhibitors have yielded similar results (41). D, Another strategy to target macrophages is to prevent their systemic mobilization and recruitment toward specific tissues through blockade of chemokine gradients. Two examples of well-characterized gradients of this nature are SDF1/CXCR4 (45, 46) and CCL2/CCR2 (50–54). E, Myeloid cells can be targeted by virtue of specific behaviors, such as angiogenesis. For example, TEMs can be targeted in different cancer models via neutralizing TIE-2 or angiopoietins, to reduce vascularization of tumors (55–60). TC, tumor cell; TCR, T cell receptor; TEM, TIE-2–expressing monocyte/macrophage.

OF2 Clin Cancer Res; 23(4) February 15, 2017 Clinical Cancer Research

Macrophages and Cancer Therapy

Using the same compound, Strachan and colleagues showed 6-month follow-up period (20). Patients in this study varied in that CSF-1R inhibition delayed the growth of orthotopically terms of macrophage depletion post-treatment, with several gli- transplanted mammary cancer cell lines derived from the omas showing sustained abundance. In light of the high frequen- MMTV-PyMT mouse model (17). In this study, macrophages cy of PI3K and phosphatase and tensin homolog (PTEN) altera- were depleted within the tumor in response to CSF-1R inhibition, tions in human glioblastoma multiforme (22), these results raise leading to a shift in the immune profile of the breast tumor the critical question of whether inherent resistance to CSF-1R þ microenvironment, characterized by high numbers of CD8 inhibitors in glioblastoma multiforme could be mechanistically cytotoxic T (Tc) cells. This observation of macrophage depletion similar to acquired resistance through PI3K activation (19), as þ concomitant with CD8 Tc cell infiltration extends to the K14- discussed above. If so, this suggests that dual inhibition of PI3K HPV-16 primary cervical tumor model, where BLZ945 treatment and CSF-1R may be important upfront in the subset of patients is also efficacious (17). These studies highlight key examples of with PTEN/PI3K pathway mutations. For a list of CSF-1R inhibi- CSF-1R blockade efficacy via completely distinct mechanisms: tors currently in clinical trials, and available published trial Pyonteck and colleagues found that macrophages were main- results, see Table 1. tained and reprogrammed to become antitumorigenic, whereas Although these studies were the first to identify mechanisms Strachan and colleagues showed that anticancer efficacy was of acquired resistance to CSF-1R inhibitors in glioblastoma dependent on macrophage depletion. As a result of these studies multiforme (19), there were earlier reports of inherent resis- and others, CSF-1R inhibitors are being tested in clinical trials for tance to CSF-1R blockade in other tumor types. DeNardo and several cancer types, including but not limited to breast cancer, colleagues showed that PLX3397 caused no change in primary glioma, ovarian cancer, melanoma, and lung cancer (18). tumor growth of an orthotopic PyMT mammary tumor trans- In a more recent preclinical study on glioblastoma multiforme, plantation model, despite effective macrophage depletion (23). it was shown that acquired resistance to CSF-1R inhibition even- CSF-1R inhibition in this case still improved disease outcome, tually emerges within a subset of mice (19). Resistance was found as macrophage depletion led to enhanced response to chemo- þ to be driven by PI3K hyperactivation after long-term exposure to therapy and allowed for infiltration of antitumorigenic CD8 Tc the CSF-1R inhibitor BLZ945. This PI3K activation phenotype was cells, consistent with other reports (17, 23–25). Strachan and driven by insulin-like growth factor 1 (IGF-1) supplied by macro- colleagues also showed that spontaneous metastasis of breast phages and IGF-1R upregulated on tumor cells, resulting in cancer to lung was not affected by BLZ945 monotherapy in canonical PI3K signaling in BLZ945-resistant tumors. Of note, PyMT genetic models. In this case, BLZ945 did not cause IGF-1 production by macrophages appeared to be triggered in part macrophage depletion in the lungs, which was required for þ by Tc cells in the microenvironment, again demonstrating a link antitumorigenic CD8 Tc cell infiltration (17). Given the role of þ between Tc cells, macrophages, and CSF-1R inhibition in cancer. CD8 Tc cells in the efficacy reported in several of these studies, þ Interestingly, in recurrent tumors, combined inhibition of either and the findings from Quail and colleagues that CD8 Tc cells IGF-1R or PI3K with CSF-1R blockade resulted in a pronounced can in part be responsible for engaging the IGF-1/PI3K axis in improvement in overall survival. However, none of these inhi- the context of acquired CSF-1R resistance in glioblastoma bitors were effective as monotherapy in recurrent disease (19). multiforme (19), these data together emphasize that the inter- These findings thus demonstrate that sustained inhibition of section between Tc cells and macrophages ought to be evalu- CSF-1R is necessary to expose the signaling dependency on ated as a putative predictor of drug response, and in cases of IGF-1R/PI3K and thereby render the tumors susceptible to sec- resistance, may serve as an additional therapeutic target. ondary pathway inhibition; one of several examples that will be reviewed here of how CSF-1R inhibition, even in cases where Combining CSF-1R inhibitors with other RTK inhibitors seemingly ineffective alone, can nonetheless significantly As discussed above, at least one mechanism of CSF-1R improve response to additional therapies. inhibitor resistance has been attributed to aberrant activation In clinical studies in glioblastoma multiforme [NCT01790503, of PI3K signaling through IGF-1R in glioblastoma multiforme NCT01349036 (20), and newly initiated NCT02829723], CSF-1R (19). Indeed, several studies have reported enhanced treat- inhibitors have shown limited efficacy to date. Butowsky and ment efficacy when CSF-1R inhibitors are used in combination colleagues recently reported findings from a phase II trial with a with additional receptor tyrosine kinase (RTK) inhibitors different inhibitor of CSF-1R, PLX3397 (which had shown some (26–28). This suggests that cross-talk between different RTK efficacy in a preclinical glioma model; ref.21), in 37 patients with pathways can compensate for one another, causing reactiva- recurrent aggressive glioblastoma multiforme. Overall, treated tion of alternative survival pathways that then require addi- patients showed no change in progression-free survival over a tional drug intervention.

Table 1. Compounds currently in clinical trials for cancer, with targets relevant to macrophage biology Compound Target Cancer trials Trial results BLZ945 CSF-1R antagonist Phase I/II In progress PLX3397 CSF-1R/Kit/Flt3 antagonist Phase II In progress RG7155 CSF-1R mAb Phase I In progress CP 870893 CD40 mAb agonist Phase I 39, 69 AMD3100 CXCR4 antagonist Phase III 70–78 MLN1202 CCR2 mAb Phase I/II In progress PF-04136309 CCR2 antagonist Phase I/II 79 CNTO888 CCL2 mAb Phase I/II 80 CEP-11981 TIE-2/VEGFR/FGFR antagonist Phase I 81 AMG386 Ang1/2 antagonist Phase III 82–94

www.aacrjournals.org Clin Cancer Res; 23(4) February 15, 2017 OF3

Quail and Joyce

In one study, Patwardhan and colleagues showed that PLX3397 Myeloid cells have also been shown to play an immunosup- treatment caused a delay in tumor outgrowth of malignant pressive role in some cancers; this is particularly relevant in the peripheral nerve sheath tumors (26). However, consistent with context of melanoma, where immunotherapy is proving effective the findings from Quail and colleagues, combination treatment of in a growing patient population. The successes with immuno- PLX3397 and rapamycin [a mechanistic target of rapamycin therapy in melanoma indicate that there is a large immune (mTOR) inhibitor] further improved this response, whereas component to these tumors, and studies have suggested a role treatment of tumors with either of these drugs as monotherapy for macrophages during immunosuppression and regulating was not as effective by comparison (19, 26). Although Pataward- checkpoint expression (32–34). In preclinical melanoma models han and colleagues reported a decrease in macrophages in tissue where CSF-1 is secreted by tumors to cause recruitment of myeloid samples in response to PLX3397 treatment, these cells were still cells, Mok and colleagues demonstrated that inhibition of CSF-1R clearly abundant in the reported immunohistochemical images can improve efficacy of vemurafenib [a B-Raf proto-oncogene, of the general macrophage marker, ionized calcium-binding serine/threonine kinase (BRAF) inhibitor] by reducing infiltra- adapter molecule 1 (Iba1). This observation is consistent with tion of immunosuppressive myeloid cells (28). However, the those from Quail and colleagues, where macrophage number was authors did not find a myeloid-based mechanism of resistance to somewhat reduced in BLZ945-resistant glioblastoma multi- vemurafenib; rather, depletion of myeloid cells allowed for þ forme tumors compared with vehicle controls; however, TAMs increased antitumorigenic CD8 Tc cell infiltration, providing were not completely depleted. It would therefore be of interest an improved "immunological space" for the BRAF inhibitor to be þ to determine whether there is a similar heterotypic signaling effective. Consistently, depletion of CD8 Tc cells in the context loop between TAM-derived IGF-1 and the nerve sheath tumor of dual targeting of CSF-1R and BRAF abrogated the benefits of cells to cause activation of PI3K signaling specifically in the this combination therapy (28). Similar to other studies discussed context of CSF-1R inhibition, and thus explain the observations here, inhibition of CSF-1R alone had minimal benefit for disease of enhanced sensitivity to rapamycin treatment. outcome and was instead most effective at enhancing efficacy of An intersection between myeloid cells and PI3K signaling has the BRAF inhibitor. also been shown in the context of immune regulation. Schmid and colleagues reported that G-protein–coupled receptors, RTKs, Alternative strategies for targeting macrophages in cancer and Toll-like receptors are capable of activating PI3Kg in biology þ þ CD11b Gr1 myeloid cells, causing their infiltration into tumors In addition to targeting CSF-1R pathways to either deplete or in response to various chemoattractants, including CSF-1 (29). reprogram macrophage populations within tumors (Fig. 1A Kaneda and colleagues recently provided further insights into and B), multiple alternative strategies have been explored in the roles of macrophages in immune regulation by showing that animal models that involve abrogating macrophage recruit- the PI3Kg isoform in macrophages regulates a critical switch ment and/or functions. between immune stimulation and suppression in cancer (30). Accordingly, macrophage-specific inactivation of PI3Kg restored Creating an imbalance between tissue-resident and recruited mac- þ CD8 Tc cell activation, leading to enhanced efficacy of immune rophage populations. Different approaches to target specific mac- checkpoint inhibition. These findings are intriguing as tumor cells rophage subpopulations in several animal models have also been often hijack or co-opt existing biological processes; in this case, the investigated. This is largely based on the hypothesis that depleting relationship between myeloid cells and PI3Kg may lend insight one type of macrophage may result in disruption of the monocyte into how tumor cells mimic these immune pathways to support immune ecosystem and consequently result in replenishment of more aggressive disease. new macrophage populations from the periphery. For instance, It has been shown that macrophages play an important role MacDonald and colleagues demonstrated that a neutralization during tumor angiogenesis in a variety of cancer types. As such, antibody specifically targeting CSF-1R depleted tissue-resident targeting macrophages together with proangiogenic pathways has monocytes across a variety of different healthy peripheral tissues been investigated in preclinical models. In one study, Priceman and tumors (35, 36). The authors hypothesized that reducing and colleagues demonstrated that treatment of Lewis lung carci- tissue-resident macrophages in this manner would create a noma (LLC) with the CSF-1R inhibitor GW2580 prevented requirement for monocyte replenishment specifically under monocytic myeloid cell infiltration into tumors and reduced inflammatory stress conditions. When this hypothesis was tested angiogenesis but provided no overall benefit in blocking tumor using different inflammatory models, depletion of the resident growth (27). By comparison, treatment of LLC tumors with the monocyte pool did not result in an overall loss of monocytes; VEGFR inhibitor DC101 caused a substantial reduction in tumor rather, local depletion was countered by enhanced recruitment of volume, attributed to a robust devascularization and significant pro-inflammatory monocytes from the bone marrow (36). These increase in tumor necrosis. However, combining both treatment data suggest that antibody-mediated neutralization of CSF-1R regimens had a synergistic effect, whereby tumor volume and may specifically target monocytes that are committed to produc- angiogenesis were further reduced (27). These data suggest that ing tissue-resident macrophages, but not inflammatory mono- CSF-1R inhibition causes an increased dependency of LLC tumors cytes, despite both of these populations expressing CSF-1R on on effective angiogenesis, whereby dual targeting of CSF-1R and their cell surface. In other words, CSF-1 may not be necessary for VEGFR would result in better outcomes than VEGFR inhibition monocyte differentiation in response to inflammatory stimuli. In alone. Thus, even in cases where there seems to be inherent the studies discussed up to this point, it is unclear whether tissue- resistance to CSF-1R blockade as a monotherapy, these inhibitors resident or recruited macrophages are most affected by pharma- may nonetheless prove useful to improve response in combina- cologic blockade of CSF-1R, or how efficacy of these agents might tion with alternative therapies that are currently under clinical or be altered by inflammatory monocyte recruitment; these will be preclinical investigation (19, 23, 25, 28, 31). important questions to address in future studies.

OF4 Clin Cancer Res; 23(4) February 15, 2017 Clinical Cancer Research

Macrophages and Cancer Therapy

A similar study by Varvel and colleagues investigated the effects patients may best respond to T cell–based immunotherapies. This of microglia depletion on macrophage repopulation of brain is consistent with other reports discussed here, where there is an þ tissue, an organ that is protected from inflammation and mono- apparent reciprocal relationship between CD8 Tc cells and cyte recruitment under normal physiologic conditions (37). The macrophages within several tumor types (17, 19, 23, 25, 31, authors showed that depletion of microglia using a transgenic 41). However, it remains unclear whether altering myeloid cell þ CD11b-HSVTK mouse model caused a repopulation of Iba1 function in long-term trials would result in stable management of peripheral monocytes (bone marrow–derived macrophage pre- disease, whether new mechanisms of immunosuppression would cursors) after just 2 weeks. After longer time periods, the distri- develop over time, or whether there would be broader, and þ bution of Iba1 cells within brain tissue from the CD11b-HSVTK perhaps dangerous, consequences for systemic autoimmune com- model was similar to that of microglia in control tissues (37). This plications outside of the context of cancer. study, along with that of MacDonald and colleagues (36), demon- strates that specific depletion of tissue-resident macrophage Disrupting chemokine gradients to prevent macrophages from enter- populations can cause a homeostatic imbalance and compensa- ing tumors. Another strategy to target macrophages is to prevent tory recruitment of peripheral monocytes into tissues, even in their systemic mobilization and recruitment to specific tissue organs such as the brain where inflammation is generally tightly niches through blockade of chemokine gradients (Fig. 1D; refs.42, controlled. Animal models that distinguish between the relative 43). One classic example is the CXCL12 [also known as stromal- contributions of recruited versus resident macrophages to tumor derived factor 1 (SDF1)]/CXCR4 axis (44). Multiphoton intravital initiation and progression are now being developed (38); it will imaging studies by Boimel and colleagues have demonstrated that be important to use these genetically engineered mice to deter- CXCL12 is upregulated by highly metastatic and invasive tumor mine whether inhibiting macrophages broadly, or rather targeting cell lines derived from an MMTV-Neu transgenic mouse model of specific subpopulations to encourage pro-inflammatory cell breast cancer when subsequently implanted in vivo (45). This repopulation, will be most effective in cancer therapy. invasive behavior is thought to be in part the result of enhanced macrophage recruitment, as metastatic phenotypes were reduced Maximizing antitumor immune responses. Another rationale for upon disruption of either CXCL12/CXCR4 or CSF-1/CSF-1R axes targeting macrophages is to reprogram the immunosuppressive (45). In another study, it was shown that autocrine CXCL12 tumor microenvironment and consequently enhance antitumor regulates differentiation of monocytes into macrophages in the T cell responses (Fig. 1C). In one study, Beatty and colleagues perivascular niche of metastatic melanoma, where it is thought to tested whether CD40 activation could overcome the immuno- facilitate CXCR4-mediated metastatic cell recruitment or leuko- suppressive microenvironment in pancreatic ductal adenocarci- cyte infiltration (46). Consistently, outside the context of cancer, noma (PDA) by augmenting T cell function (39). CD40 is known high CXCL12 levels are known to maintain hematopoietic stem to regulate antitumor immune responses, in part by enhancing and progenitor cells within bone marrow reservoirs (47–49), immune activation via antigen-presenting cells (40). Beatty and perhaps providing some explanation for why CXCL12-expressing colleagues demonstrated that treatment of PDA tumors with an monocytes were found to be abundant in perivascular niches. agonistic CD40 antibody (FGK45) increased macrophage expres- Given the nonspecific role for this axis during macrophage sion of MHC class II (MHCII) and CD86 (a T cell costimulatory recruitment, however, it is unclear whether disrupting the molecule expressed on antigen-presenting cells that is important CXCL12/CXCR4 axis in cancer would be beneficial or cause broad for T cell activation) and caused them to infiltrate tumors, adopt systemic inflammation. an antitumorigenic phenotype, and disrupt the dense tumor Another example of an important chemokine axis in tumor- stroma (39). Consistently, in the KPC model of PDA, Zhu and associated recruitment of macrophages is CCL2/CCR2. In one colleagues targeted immunosuppressive macrophages via CSF-1R study on colorectal and lung cancers, tumor cell–derived CCL2 þ inhibition in combination with checkpoint blockade (41). In this supported the recruitment of CD11b Gr1mid myeloid cells to the model, treatment of tumor-bearing animals with a CSF-1R inhib- metastatic liver niche, and inhibition of CCL2 blocked recruitment itor did not cause a depletion of macrophages; rather, TAMs were of these cells and subsequently reduced metastatic burden (50–52). reprogrammed to adopt enhanced antigen-presenting capacity In another study on MMTV-PyMT breast cancer, Qian and collea- and triggered antitumorigenic T cell activation, including mod- gues showed that lung metastases specifically support infiltration of ulation of checkpoint molecules. Subsequent combination treat- inflammatory monocytes expressing CCR2 to facilitate metastatic ment of PDA with CSF-1R blockade and checkpoint immuno- seeding (53). In this model, the chemokine CCL2 was found to be therapy [including programmed cell death protein 1 (PD1) or produced by both tumor cells and the lung stroma, and blockade of cytotoxic T lymphocyte–associated protein 4 (CTLA4) antag- the CCL2/CCR2 axis was sufficient to reduce monocyte recruitment onists] enhanced efficacy, compared with checkpoint inhibitor and metastatic seeding, ultimately resulting in prolonged overall treatment alone (41). survival (53). However, animals still eventually succumbed to their Finally, in a BRAF-driven melanoma study by Mok and collea- disease despite continued blockade of the CCL2/CCR2 axis (53). gues, pharmacologic inhibition of CSF-1R via PLX3397 inhibited Moreover, others have reported accelerated progression of immunosuppressive myeloid cells in tumors and favored an disease after CCL2 blockade is stopped (54), indicating that MHCIIhi antigen-presenting population, resulting in improved caution should be taken, as this treatment regimen may be only efficacy of adoptive cell therapy of melanoma-targeted T cells temporarily effective at keeping metastases at bay. (31). The authors suggest that targeting macrophages in combination with enhancing T cell responses is likely to provide optimal Targeting angiogenesis through TIE-2–expressing monocytes. Final- therapeutic outcomes. From these collective studies, it follows ly, another means to target macrophages in cancer is to block that the abundance and nature of immunosuppressive myeloid subpopulations that regulate specific protumorigenic behaviors. cell populations need to be considered when evaluating which One phenotype that is frequently associated with macrophages is

www.aacrjournals.org Clin Cancer Res; 23(4) February 15, 2017 OF5

Quail and Joyce

their ability to regulate vascularization. This is often attributed to been shown using murine models that depletion of TAMs prior to angiopoietin receptor (TIE-2)–expressing monocytes/macrophages chemotherapy results in enhanced response to treatment (17, 23– (TEMs; ref.55), which can be recruited to tumors through classical 25). Similar approaches are being implemented in patient studies macrophage cytokine axes, such as CXCL12/CXCR4 (56), and act in using CSF-1R inhibitors in combination with paclitaxel, eribulin, a paracrine manner with angiopoietins (Ang) to modulate angio- vemurafenib, or temozolomide, among others, with the goal genesis in the perivascular space (Fig. 1E; refs. 57–60). TEM deple- being that macrophage depletion will similarly improve tion in different cancer models (e.g., gliomas, breast tumors, therapeutic efficacy of chemotherapeutic agents. In addition, pancreatic tumors, melanoma, lung tumors, among others) is antibodies against CSF-1R are being used in combination with associated with reduced angiogenesis and delayed tumor growth immunotherapies such as nivolumab in patients with solid with minimal adverse systemic effects on hematopoiesis (56, 58, tumors (18). These trials extend from findings in both mice and 59). Given the inherent complications in broadly blocking angio- patients, whereby targeting macrophages has been reported to þ þ genic pathways in patients, targeting TEMs in cancer patients may shift CD4 helper T cell and CD8 Tc cell populations in a variety be beneficial, as these cells are more specific to tumor-associated of cancers, including endometrial, breast, colon, and prostate angiogenesis compared with physiologic angiogenic processes. cancer, among others (17, 41, 65). In light of these initiatives, However, in light of the highly compensatory and conserved it will be important going forward to determine how to balance mechanisms of angiogenesis during tissue homeostasis, one both macrophage and T cell populations effectively during treat- might predict that resistance could emerge whereby tumor angio- ment to optimize therapeutic efficacy and disease outcome. genesis no longer involves TEMs (61). For instance, hypoxia may In clinical trials, the rationale for using CSF-1R inhibitors is result from TEM depletion/reprogramming, causing a heightened largely based on the goal of depleting macrophage populations. dependency on VEGF, FGF, or ephrin signaling, or pericytes may However, as discussed here, rather than depleting macrophages, increase vascular remodeling and maturation of existing imma- reprogramming or altering mobilization of progenitor cells might ture vessels (62). This hypothesis is supported by the findings result in improved disease outcomes (16, 63, 67). Indeed, many reported by De Palma and colleagues that although breast and of the preclinical strategies discussed here have now been extend- lung tumor growth is delayed in response to TEM depletion, it is ed to human studies (Table 1). For example, in patients, the not fully blocked and tumors eventually grow (59). It will be agonistic CD40 mAb CP 870893 is being used to boost antitumor important to consider whether TEM-targeted therapies will expe- immune responses in pancreatic cancer, melanoma, and other rience similar shortcomings to several of the original antiangio- solid tumors (39, 68, 69). Compounds targeting CXCR4, such as genic drugs, which originally held considerable promise for AMD3100, are also being developed in clinical trials, mainly for multiple cancer types. Of note, it has been shown that dual blood-based cancers to mobilize immune cells away from the targeting of VEGF and Ang2 with a bispecific antibody in ortho- supportive bone marrow niche and thus sensitize them to che- topic glioma models prolongs survival compared with targeting motherapy (70), or to increase hematopoietic stem cell yields for VEGF alone by reprogramming TAMs toward an M1-like anti- therapeutic transplantation (71–78), which has now been tumorigenic phenotype (63). These data suggest that combina- approved for marketing. Targeting the CCL2/CCR2 axis in tion targeting of both VEGF and TEMs might help improve long- patients is also underway: CCR2 inhibitors are being used to term response to antiangiogenic therapy and that macrophage alter mobilization of bone marrow–derived cells, including reprogramming may enhance response to VEGF inhibitors, a MLN1202 (NCT01015560, NCT02723006) and PF-04136309 notion that is supported by preclinical data discussed herein [NCT01413022 (79), NCT02732938], which are in earlier phases regarding combined inhibition of CSF-1R and VEGFR (27). of clinical trial development for bone metastases, melanoma, and pancreatic cancer. Similarly, the CCL2 antibody CNTO888 has Clinical–Translational Advances been shown to be well tolerated in prostate cancer studies; however, early results have not demonstrated efficacy in the The majority of clinical trials currently in progress that aim to patient setting as a single agent (80). Finally, compounds target- target macrophages in solid tumors are focused on CSF-1R inhi- ing angiopoietins or TIE-2 are being developed clinically to bition (Table 1; ref.18). These are predominantly in the context of control angiogenesis in solid tumors, including CEP-11981 phase I and II trials; therefore, tolerability and efficacy are still (NCT00875264, which has now been discontinued; ref.81) and under active investigation in patients. Perhaps the greatest success AMG386, which has shown some efficacy in a variety of tumors, so far has come from diffuse-type giant cell tumors (gtGCT) particularly ovarian cancer (82–94). As such, it will be important arising from the synovium within joints, which secrete high levels to start considering reprogramming strategies for myeloid cell of CSF-1 resulting in a robust recruitment of macrophages that populations in the context of cancer therapy for patients, in support disease progression. Ongoing trials with PLX3397 have addition to the depletion strategies that are currently being reported a partial response in more than 50% of patients and implemented. In our opinion, the insights from preclinical stud- disease stabilization in approximately 30% of patients (64). In ies may critically inform additional reprogramming strategies in similar studies, use of the CSF-1R neutralizing antibody RG7155 gliomas, pancreatic cancer, and other tumor types, as discussed resulted in macrophage depletion and objective clinical responses here, with the goal of enlisting these cells to actively fight cancer. in more than 70% of gtGCT patients (65, 66). Given the high abundance of macrophages in these tumors and the success of Disclosure of Potential Conflicts of Interest these trials thus far, gtGCT has served as a model for broadening No potential conflicts of interest were disclosed. the use of CSF-1R inhibitors to an expanded panel of tumor types as monotherapy or in combination with additional therapies. Disclaimer Of note, several studies are underway combining CSF-1R The content is solely the responsibility of the authors and does not neces- inhibitors with chemotherapy (18). As discussed above, it has sarily represent the official views of the NIH.

OF6 Clin Cancer Res; 23(4) February 15, 2017 Clinical Cancer Research

Macrophages and Cancer Therapy

Authors' Contributions Grant Support Conception and design: J.A. Joyce This work was supported by the NCI of the NIH under award number Writing, review, and/or revision of the manuscript: D.F. Quail, J.A. Joyce R01CA181355. Other [reviewed literature and interpreted data from different sources (review article)]: D.F. Quail Received August 23, 2016; revised October 26, 2016; accepted October 26, 2016; published OnlineFirst November 28, 2016.

References 1. Solinas G, Germano G, Mantovani A, Allavena P. Tumor-associated macro- 21. Coniglio SJ, Eugenin E, Dobrenis K, Stanley ER, West BL, Symons MH, et al. phages (TAM) as major players of the cancer-related inflammation. J Microglial stimulation of glioblastoma invasion involves epidermal Leukoc Biol 2009;86:1065–73. growth factor receptor (EGFR) and colony stimulating factor 1 receptor 2. Quail DF, Joyce JA. Microenvironmental regulation of tumor progression (CSF-1R) signaling. Mol Med 2012;18:519–27. and metastasis. Nat Med 2013;19:1423–37. 22. Brennan CW, Verhaak RG, McKenna A, Campos B, Noushmehr H, Salama 3. Sieweke MH, Allen JE. Beyond stem cells: self-renewal of differentiated SR, et al. The somatic genomic landscape of glioblastoma. Cell 2013;155: macrophages. Science 2013;342:1242974. 462–77. 4. Amano SU, Cohen JL, Vangala P, Tencerova M, Nicoloro SM, Yawe JC, et al. 23. DeNardo DG, Brennan DJ, Rexhepaj E, Ruffell B, Shiao SL, Madden SF, et al. Local proliferation of macrophages contributes to obesity-associated Leukocyte complexity predicts breast cancer survival and functionally adipose tissue inflammation. Cell Metab 2014;19:162–71. regulates response to chemotherapy. Cancer Discov 2011;1:54–67. 5. Jenkins SJ, Ruckerl D, Cook PC, Jones LH, Finkelman FD, van Rooijen N, 24. Shree T, Olson OC, Elie BT, Kester JC, Garfall AL, Simpson K, et al. et al. Local macrophage proliferation, rather than recruitment from the Macrophages and cathepsin proteases blunt chemotherapeutic response blood, is a signature of TH2 inflammation. Science 2011;332:1284–8. in breast cancer. Genes Dev 2011;25:2465–79. 6. Robbins CS, Hilgendorf I, Weber GF, Theurl I, Iwamoto Y, Figueiredo JL, 25. Ruffell B, Chang-Strachan D, Chan V, Rosenbusch A, Ho CM, Pryer N, et al. et al. Local proliferation dominates lesional macrophage accumulation in Macrophage IL-10 blocks CD8þ T cell-dependent responses to chemo- atherosclerosis. Nat Med 2013;19:1166–72. therapy by suppressing IL-12 expression in intratumoral dendritic cells. 7. Wei S, Nandi S, Chitu V, Yeung YG, Yu W, Huang M, et al. Functional Cancer Cell 2014;26:623–37. overlap but differential expression of CSF-1 and IL-34 in their CSF-1 26. Patwardhan PP, Surriga O, Beckman MJ, de Stanchina E, Dematteo RP, Tap receptor-mediated regulation of myeloid cells. J Leukoc Biol 2010;88: WD, et al. Sustained inhibition of receptor tyrosine kinases and macro- 495–505. phage depletion by PLX3397 and rapamycin as a potential new approach 8. Yoshida H, Hayashi S, Kunisada T, Ogawa M, Nishikawa S, Okamura H, for the treatment of MPNSTs. Clin Cancer Res 2014;20:3146–58. et al. The murine mutation osteopetrosis is in the coding region of the 27. Priceman SJ, Sung JL, Shaposhnik Z, Burton JB, Torres-Collado AX, macrophage colony stimulating factor gene. Nature 1990;345:442–4. Moughon DL, et al. Targeting distinct tumor-infiltrating myeloid cells by 9. Wiktor-Jedrzejczak W, Bartocci A, Ferrante AWJr, Ahmed-Ansari A, Sell KW, inhibiting CSF-1 receptor: combating tumor evasion of antiangiogenic Pollard JW, et al. Total absence of colony-stimulating factor 1 in the therapy. Blood 2010;115:1461–71. macrophage-deficient osteopetrotic (op/op) mouse. Proc Natl Acad Sci 28. Mok S, Tsoi J, Koya RC, Hu-Lieskovan S, West BL, Bollag G, et al. Inhibition U S A 1990;87:4828–32. of colony stimulating factor-1 receptor improves antitumor efficacy of 10. Ginhoux F, Schultze JL, Murray PJ, Ochando J, Biswas SK. New insights into BRAF inhibition. BMC Cancer 2015;15:356. the multidimensional concept of macrophage ontogeny, activation and 29. Schmid MC, Avraamides CJ, Dippold HC, Franco I, Foubert P, Ellies LG, function. Nat Immunol 2015;17:34–40. et al. Receptor tyrosine kinases and TLR/IL1Rs unexpectedly activate 11. Murray PJ, Allen JE, Biswas SK, Fisher EA, Gilroy DW, Goerdt S, et al. myeloid cell PI3kgamma, a single convergent point promoting tumor Macrophage activation and polarization: nomenclature and experimental inflammation and progression. Cancer Cell 2011;19:715–27. guidelines. Immunity 2014;41:14–20. 30. Kaneda MM, Messer KS, Ralainirina N, Li H, Leem C, Gorjestani S, et al. 12. Ostuni R, Piccolo V, Barozzi I, Polletti S, Termanini A, Bonifacio S, et al. PI3Kg is a molecular switch that controls immune suppression. Nature Latent enhancers activated by stimulation in differentiated cells. Cell 2016;539:437–42. 2013;152:157–71. 31. Mok S, Koya RC, Tsui C, Xu J, Robert L, Wu L, et al. Inhibition of CSF-1 13. Mosser DM, Edwards JP. Exploring the full spectrum of macrophage receptor improves the antitumor efficacy of adoptive cell transfer immu- activation. Nat Rev Immunol 2008;8:958–69. notherapy. Cancer Res 2014;74:153–61. 14. Ruffell B, Affara NI, Coussens LM. Differential macrophage programming 32. Yaddanapudi K, Putty K, Rendon BE, Lamont GJ, Faughn JD, Satoskar A, in the tumor microenvironment. Trends Immunol 2012;33:119–26. et al. Control of tumor-associated macrophage alternative activation 15. Ruffell B, Coussens LM. Macrophages and therapeutic resistance in cancer. by macrophage migration inhibitory factor. J Immunol 2013;190: Cancer Cell 2015;27:462–72. 2984–93. 16. Pyonteck SM, Akkari L, Schuhmacher AJ, Bowman RL, Sevenich L, Quail 33. Loke P, Allison JP. PD-L1 and PD-L2 are differentially regulated by Th1 and DF, et al. CSF-1R inhibition alters macrophage polarization and blocks Th2 cells. Proc Natl Acad Sci U S A 2003;100:5336–41. glioma progression. Nat Med 2013;19:1264–72. 34. Joyce JA, Fearon DT. T cell exclusion, immune privilege, and the tumor 17. Strachan DC, Ruffell B, Oei Y, Bissell MJ, Coussens LM, Pryer N, et al. microenvironment. Science 2015;348:74–80. CSF1R inhibition delays cervical and mammary tumor growth in 35. Geissmann F, Manz MG, Jung S, Sieweke MH, Merad M, Ley K. Develop- murine models by attenuating the turnover of tumor-associated macro- ment of monocytes, macrophages, and dendritic cells. Science 2010;327: phages and enhancing infiltration by CD8 T cells. Oncoimmunology 656–61. 2013;2:e26968. 36. MacDonald KP, Palmer JS, Cronau S, Seppanen E, Olver S, Raffelt NC, et al. 18. Ries CH, Hoves S, Cannarile MA, Ruttinger D. CSF-1/CSF-1R targeting An antibody against the colony-stimulating factor 1 receptor depletes the agents in clinical development for cancer therapy. Curr Opin Pharmacol resident subset of monocytes and tissue- and tumor-associated macro- 2015;23:45–51. phages but does not inhibit inflammation. Blood 2010;116:3955–63. 19. Quail DF, Bowman RL, Akkari L, Quick ML, Schuhmacher AJ, Huse JT, et al. 37. Varvel NH, Grathwohl SA, Baumann F, Liebig C, Bosch A, Brawek B, The tumor microenvironment underlies acquired resistance to CSF-1R et al. Microglial repopulation model reveals a robust homeostatic inhibition in gliomas. Science 2016;352:aad3018. process for replacing CNS myeloid cells. Proc Natl Acad Sci U S A 2012; 20. Butowski N, Colman H, De Groot JF, Omuro AM, Nayak L, Wen PY, et al. 109:18150–5. Orally administered colony stimulating factor 1 receptor inhibitor 38. Bowman RL, Klemm F, Akkari L, Pyonteck SM, Sevenich L, Quail DF, et al. PLX3397 in recurrent glioblastoma: an ivy foundation early phase clinical Macrophage ontogeny underlies differences in tumor-specific education in trials consortium phase II study. Neuro Oncol 2015;18:557–64. brain malignancies. Cell Rep 2016;17:2445–59.

www.aacrjournals.org Clin Cancer Res; 23(4) February 15, 2017 OF7

Quail and Joyce

39. Beatty GL, Chiorean EG, Fishman MP, Saboury B, Teitelbaum UR, Sun W, 64. Tap WD, Wainberg ZA, Anthony SP, Ibrahim PN, Zhang C, Healey JH, et al. et al. CD40 agonists alter tumor stroma and show efficacy against pan- Structure-guided blockade of CSF1R kinase in tenosynovial giant-cell creatic carcinoma in mice and humans. Science 2011;331:1612–6. tumor. N Engl J Med 2015;373:428–37. 40. Vonderheide RH. Prospect of targeting the CD40 pathway for cancer 65. Ries CH, Cannarile MA, Hoves S, Benz J, Wartha K, Runza V, et al. Targeting therapy. Clin Cancer Res 2007;13:1083–8. tumor-associated macrophages with anti-CSF-1R antibody reveals a strat- 41. Zhu Y, Knolhoff BL, Meyer MA, Nywening TM, West BL, Luo J, et al. CSF1/ egy for cancer therapy. Cancer Cell 2014;25:846–59. CSF1R blockade reprograms tumor-infiltrating macrophages and improves 66. Cassier PA, Gomez-Roca C, Italiano A, Cannarile MA, Ries CH, Brillouet A, response to T-cell checkpoint immunotherapy in pancreatic cancer models. et al. Phase 1 study of RG7155, a novel anti-CSF1R antibody, in patients Cancer Res 2014;74:5057–69. with locally advanced pigmented villonodular synovitis (PVNS). J Clin 42. Balkwill F. Cancer and the chemokine network. Nat Rev Cancer 2004; Oncol 32:5s, 2014 (suppl; abstr 10504). 4:540–50. 67. Hagemann T, Lawrence T, McNeish I, Charles KA, Kulbe H, Thompson RG, 43. Homey B, Muller A, Zlotnik A. Chemokines: agents for the immunotherapy et al. "Re-educating" tumor-associated macrophages by targeting NF-kap- of cancer? Nat Rev Immunol 2002;2:175–84. paB. J Exp Med 2008;205:1261–8. 44. Balkwill F. The significance of cancer cell expression of the chemokine 68. Vonderheide RH, Burg JM, Mick R, Trosko JA, Li D, Shaik MN, et al. Phase I receptor CXCR4. Semin Cancer Biol 2004;14:171–9. study of the CD40 agonist antibody CP-870,893 combined with carbo- 45. Boimel PJ, Smirnova T, Zhou ZN, Wyckoff J, Park H, Coniglio SJ, et al. platin and paclitaxel in patients with advanced solid tumors. Oncoimmu- Contribution of CXCL12 secretion to invasion of breast cancer cells. Breast nology 2013;2:e23033. Cancer Res 2012;14:R23. 69. Bajor DL, Xu X, Torigian DA, Mick R, Garcia LR, Richman LP, et al. Immune 46. Sanchez-Martin L, Estecha A, Samaniego R, Sanchez-Ramon S, Vega MA, activation and a 9-year ongoing complete remission following CD40 Sanchez-Mateos P. The chemokine CXCL12 regulates monocyte-macro- antibody therapy and metastasectomy in a patient with metastatic mela- phage differentiation and RUNX3 expression. Blood 2011;117:88–97. noma. Cancer Immunol Res 2014;2:1051–8. 47. Petit I, Szyper-Kravitz M, Nagler A, Lahav M, Peled A, Habler L, et al. G-CSF 70. Uy GL, Rettig MP, Motabi IH, McFarland K, Trinkaus KM, Hladnik LM, et al. induces stem cell mobilization by decreasing bone marrow SDF-1 and up- A phase 1/2 study of chemosensitization with the CXCR4 antagonist regulating CXCR4. Nat Immunol 2002;3:687–94. plerixafor in relapsed or refractory acute myeloid leukemia. Blood 48. Mercier FE, Ragu C, Scadden DT. The bone marrow at the crossroads of 2012;119:3917–24. blood and immunity. Nat Rev Immunol 2012;12:49–60. 71. DiPersio JF, Micallef IN, Stiff PJ, Bolwell BJ, Maziarz RT, Jacobsen E, et al. 49. Lavin Y, Mortha A, Rahman A, Merad M. Regulation of macrophage Phase III prospective randomized double-blind placebo-controlled trial of development and function in peripheral tissues. Nat Rev Immunol plerixafor plus granulocyte colony-stimulating factor compared with pla- 2015;15:731–44. cebo plus granulocyte colony-stimulating factor for autologous stem-cell 50. Zhao L, Lim SY, Gordon-Weeks AN, Tapmeier TT, Im JH, Cao Y, et al. mobilization and transplantation for patients with non-Hodgkin's lym- Recruitment of a myeloid cell subset (CD11b/Gr1 mid) via CCL2/CCR2 phoma. J Clin Oncol 2009;27:4767–73. promotes the development of colorectal cancer liver metastasis. Hepatol- 72. Devine SM, Vij R, Rettig M, Todt L, McGlauchlen K, Fisher N, et al. Rapid ogy 2013;57:829–39. mobilization of functional donor hematopoietic cells without G-CSF using 51. Lim SY, Gordon-Weeks AN, Zhao L, Tapmeier TT, Im JH, Cao Y, et al. AMD3100, an antagonist of the CXCR4/SDF-1 interaction. Blood 2008; Recruitment of myeloid cells to the tumor microenvironment supports 112:990–8. liver metastasis. Oncoimmunology 2013;2:e23187. 73. Fruehauf S, Ehninger G, Hubel K, Topaly J, Goldschmidt H, Ho AD, et al. 52. Lim SY, Yuzhalin AE, Gordon-Weeks AN, Muschel RJ. Targeting the CCL2- Mobilization of peripheral blood stem cells for autologous transplant in CCR2 signaling axis in cancer metastasis. Oncotarget 2016;7:28697–710. non-Hodgkin's lymphoma and multiple myeloma patients by plerixafor 53. Qian BZ, Li J, Zhang H, Kitamura T, Zhang J, Campion LR, et al. CCL2 and G-CSF and detection of tumor cell mobilization by PCR in multiple recruits inflammatory monocytes to facilitate breast-tumour metastasis. myeloma patients. Bone Marrow Transplant 2010;45:269–75. Nature 2011;475:222–5. 74. Stiff P, Micallef I, McCarthy P, Magalhaes-Silverman M, Weisdorf D, Territo 54. Bonapace L, Coissieux MM, Wyckoff J, Mertz KD, Varga Z, Junt T, et al. M, et al. Treatment with plerixafor in non-Hodgkin's lymphoma and Cessation of CCL2 inhibition accelerates breast cancer metastasis by multiple myeloma patients to increase the number of peripheral blood promoting angiogenesis. Nature 2014;515:130–3. stem cells when given a mobilizing regimen of G-CSF: implications for the 55. Lewis CE, Harney AS, Pollard JW. The multifaceted role of perivascular heavily pretreated patient. Biol Blood Marrow Transplant 2009;15: macrophages in tumors. Cancer Cell 2016;30:18–25. 249–56. 56. Welford AF, Biziato D, Coffelt SB, Nucera S, Fisher M, Pucci F, et al. TIE2- 75. Deol A, Abrams J, Masood A, Al-Kadhimi Z, Abidi MH, Ayash L, et al. Long- expressing macrophages limit the therapeutic efficacy of the vascular- term follow up of patients proceeding to transplant using plerixafor disrupting agent combretastatin A4 phosphate in mice. J Clin Invest mobilized stem cells and incidence of secondary myelodysplastic syn- 2011;121:1969–73. drome/AML. Bone Marrow Transplant 2013;48:1112–6. 57. De Palma M, Naldini L. Angiopoietin-2 TIEs up macrophages in tumor 76. DiPersio JF, Stadtmauer EA, Nademanee A, Micallef IN, Stiff PJ, Kaufman angiogenesis. Clin Cancer Res 2011;17:5226–32. JL, et al. Plerixafor and G-CSF versus placebo and G-CSF to mobilize 58. De Palma M, Venneri MA, Galli R, Sergi Sergi L, Politi LS, Sampaolesi M, hematopoietic stem cells for autologous stem cell transplantation in et al. Tie2 identifies a hematopoietic lineage of proangiogenic monocytes patients with multiple myeloma. Blood 2009;113:5720–6. required for tumor vessel formation and a mesenchymal population of 77. Kumar SK, Mikhael J, Laplant B, Lacy MQ, Buadi FK, Dingli D, et al. Phase 2 pericyte progenitors. Cancer Cell 2005;8:211–26. trial of intravenously administered plerixafor for stem cell mobilization in 59. De Palma M, Venneri MA, Roca C, Naldini L. Targeting exogenous genes to patients with multiple myeloma following lenalidomide-based initial tumor angiogenesis by transplantation of genetically modified hemato- therapy. Bone Marrow Transplant 2014;49:201–5. poietic stem cells. Nat Med 2003;9:789–95. 78. Russell N, Douglas K, Ho AD, Mohty M, Carlson K, Ossenkoppele GJ, et al. 60. De Palma M, Murdoch C, Venneri MA, Naldini L, Lewis CE. Tie2-expressing Plerixafor and granulocyte colony-stimulating factor for first-line steady- monocytes: regulation of tumor angiogenesis and therapeutic implica- state autologous peripheral blood stem cell mobilization in lymphoma tions. Trends Immunol 2007;28:519–24. and multiple myeloma: results of the prospective PREDICT trial. Haema- 61. Ferrara N. VEGF and the quest for tumour angiogenesis factors. Nat Rev tologica 2013;98:172–8. Cancer 2002;2:795–803. 79. Nywening TM, Wang-Gillam A, Sanford DE, Belt BA, Panni RZ, Cusworth 62. Bergers G, Hanahan D. Modes of resistance to anti-angiogenic therapy. Nat BM, et al. Targeting tumour-associated macrophages with CCR2 inhibition Rev Cancer 2008;8:592–603. in combination with FOLFIRINOX in patients with borderline resectable 63. Kloepper J, Riedemann L, Amoozgar Z, Seano G, Susek K, Yu V, et al. Ang-2/ and locally advanced pancreatic cancer: a single-centre, open-label, dose- VEGF bispecific antibody reprograms macrophages and resident microglia finding, non-randomised, phase 1b trial. Lancet Oncol 2016;17:651–62. to anti-tumor phenotype and prolongs glioblastoma survival. Proc Natl 80. Pienta KJ, Machiels JP, Schrijvers D, Alekseev B, Shkolnik M, Crabb SJ, et al. Acad Sci U S A 2016;113:4476–81. Phase 2 study of carlumab (CNTO 888), a human monoclonal antibody

OF8 Clin Cancer Res; 23(4) February 15, 2017 Clinical Cancer Research

Macrophages and Cancer Therapy

against CC-chemokine ligand 2 (CCL2), in metastatic castration-resistant 88. Dieras V, Wildiers H, Jassem J, Dirix LY, Guastalla JP, Bono P, et al. prostate cancer. Invest New Drugs 2013;31:760–8. Trebananib (AMG 386) plus weekly paclitaxel with or without bevacizu- 81. Pili R, Carducci M, Brown P, Hurwitz H. An open-label study to determine mab as ﬁrst-line therapy for HER2-negative locally recurrent or metastatic the maximum tolerated dose of the multitargeted tyrosine kinase inhibitor breast cancer: a phase 2 randomized study. Breast 2015;24:182–90. CEP-11981 in patients with advanced cancer. Invest New Drugs 2014;32: 89. Herbst RS, Hong D, Chap L, Kurzrock R, Jackson E, Silverman JM, et al. 1258–68. Safety, pharmacokinetics, and antitumor activity of AMG 386, a selective 82. Atkins MB, Gravis G, Drosik K, Demkow T, Tomczak P, Wong SS, et al. angiopoietin inhibitor, in adult patients with advanced solid tumors. J Clin Trebananib (AMG 386) in combination with sunitinib in patients with Oncol 2009;27:3557–65. metastatic renal cell cancer: an open-label, multicenter, phase II study. 90. Peeters M, Strickland AH, Lichinitser M, Suresh AV, Manikhas G, Shapiro J, J Clin Oncol 2015;33:3431–8. et al. A randomised, double-blind, placebo-controlled phase 2 study of 83. Doi T, Ohtsu A, Fuse N, Yoshino T, Tahara M, Shibayama K, et al. Phase 1 trebananib (AMG 386) in combination with FOLFIRI in patients with study of trebananib (AMG 386), an angiogenesis targeting angiopoietin-1/ previously treated metastatic colorectal carcinoma. Br J Cancer 2013;108: 2 antagonist, in Japanese patients with advanced solid tumors. Cancer 503–11. Chemother Pharmacol 2013;71:227–35. 91. Hong DS, Kurzrock R, Mulay M, Rasmussen E, Wu BM, Bass MB, et al. A 84. Liontos M, Lykka M, Dimopoulos MA, Bamias A. Proﬁle of trebananib phase 1b, open-label study of trebananib plus bevacizumab or motesanib (AMG386) and its potential in the treatment of ovarian cancer. Onco in patients with solid tumours. Oncotarget 2014;5:11154–67. Targets Ther 2014;7:1837–45. 92. Eatock MM, Tebbutt NC, Bampton CL, Strickland AH, Valladares-Ayerbes 85. SP DA, Mahoney MR, Van Tine BA, Adkins DR, Perdekamp MT, Condy M, Swieboda-Sadlej A, et al. Phase II randomized, double-blind, placebo- MM, et al. Alliance A091103 a phase II study of the angiopoietin 1 and 2 controlled study of AMG 386 (trebananib) in combination with cisplatin peptibody trebananib for the treatment of angiosarcoma. Cancer Che- and capecitabine in patients with metastatic gastro-oesophageal cancer. mother Pharmacol 2015;75:629–38. Ann Oncol 2013;24:710–8. 86. Vergote I, Oaknin A, Baurain JF, Ananda S, Wong S, Su X, et al. A phase 1b, 93. Rini B, Szczylik C, Tannir NM, Koralewski P, Tomczak P, Deptala A, et al. open-label study of trebananib in combination with paclitaxel and carbo- AMG 386 in combination with sorafenib in patients with metastatic clear platin in patients with ovarian cancer receiving interval or primary debulk- cell carcinoma of the kidney: a randomized, double-blind, placebo-con- ing surgery. Eur J Cancer 2014;50:2408–16. trolled, phase 2 study. Cancer 2012;118:6152–61. 87. Hong DS, Gordon MS, Samlowski WE, Kurzrock R, Tannir N, Friedland D, 94. Lu JF, Rasmussen E, Karlan BY, Vergote IB, Navale L, Kuchimanchi M, et al. et al. A phase I, open-label study of trebananib combined with sorafenib or Exposure-response relationship of AMG 386 in combination with weekly sunitinib in patients with advanced renal cell carcinoma. Clin Genitourin paclitaxel in recurrent ovarian cancer and its implication for dose selection. Cancer 2014;12:167–77e2. Cancer Chemother Pharmacol 2012;69:1135–44.

www.aacrjournals.org Clin Cancer Res; 23(4) February 15, 2017 OF9

Molecular Pathways: Deciphering Mechanisms of Resistance to Macrophage-Targeted Therapies

Daniela F. Quail and Johanna A. Joyce

Clin Cancer Res Published OnlineFirst November 28, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-16-0133

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2017/01/24/1078-0432.CCR-16-0133. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.