In solid tumors such as those in breast
In solid tumors, such as those in breast and pancreatic cancer, infiltrating CD68+ or CD163+ tumor-associated macrophages (TAMs) correlate with poor outcome (DeNardo et al., 2011, Kurahara et al., 2011, Shabo et al., 2008). The tumor-promoting function of TAMs is based on their capacity to secrete proangiogenic and growth factors, as well as to potently suppress T cell effector function by releasing immunosuppressive cytokines and affecting their metabolism (Wynn et al., 2013, Biswas and Mantovani, 2010, Gordon and Martinez, 2010, Hoves et al., 2006). Macrophages exerting these protumorigenic functions are also termed M2-type macrophages, in contrast to the antitumorigenic M1 subtype (Mantovani et al., 2002).
Discussion Here we describe a specific and potent CSF-1R targeting antibody, RG7155, that shows significant clinical activity in Dt-GCT patients and reduces TAMs in tumors of patients suffering from various cancers. In contrast to RG7155’s blocking both ligand-dependent and -independent receptor activation, all other blocking CSF-1R mao inhibitors currently known to be in phase 1 clinical development for advanced solid tumors (antibody AMG820, identifier NCT01444404; antibody IMC-CS4, identifier NCT01346358) target the ligand binding domains (Amgen, human C-FMS antigen binding proteins US2008/073611, 2009; Imclone, antibodies against CSF-1R, US2011/030148, 2011). Further clinical evaluation is needed to explore whether targeting of different CSF-1R epitopes will impact efficacy and safety as shown for other therapeutic antibodies targeting an RTK (King and Wong 2012) and the impact of direct versus indirect competition in settings in which the ligand concentration increases as much as 1,000-fold after receptor blockade. Additional toxicities triggered by less-specific CSF-1R small-molecule tyrosine kinase inhibitors have been discussed as another limiting factor of CSF-1R blockade in tumor tissue (El-Gamal et al., 2013), such as imatinib, which is a weak CSF-1R inhibitor compared with its primary targets, ABL and c-KIT (Taylor et al., 2006, Dewar et al., 2005). Imatinib mesylate and its more potent and better-tolerated successor, nilotinib, have been tested in small series of Dt-GCT patients with only limited clinical activity (Gelderblom et al., 2013, Cassier et al., 2012) compared to the objective clinical responses in 74% of Dt-GCT patients treated with RG7155 as described here. The overall safety profile for RG7155 was acceptable, with the most frequent adverse event being periorbital edema, which mirrored the findings from the monkey study. Thus far, none of these patients was reported to have progressive disease, with the longest follow-up being 12 months. The molecular pathology of Dt-GCT represents a model disease for CSF-1R targeting agents, with a few aberrant cells producing CSF-1, resulting in massive recruitment of CSF-1R positive macrophages. Additionally, in solid malignancies such as sarcoma or breast cancer CSF-1, overexpression by tumor cells and an extensive CD68+ or CSF-1R+ macrophage infiltrate are associated with poor prognosis (Espinosa et al., 2009, Kluger et al., 2004). In solid malignancies, additional chemokines such as MCP-1 or SDF-1 might support recruitment of monocytes (Pollard, 2009, Murdoch et al., 2008). However, GM-CSF or CSF-1 control TCM macrophage survival, as shown in our in vitro studies. Since CSF-1 (unlike GM-CSF) is systemically available (Hamilton and Achuthan 2013), we postulate that TAMs are amenable to anti-CSF-1R therapy independently of their recruitment mechanism. Hence, patients enrolled in the phase 1 trial with RG7155 were not preselected for CSF-1 levels in tumor, serum, or TAM infiltrate. Yet, in all paired tumor biopsies, a significant reduction of CSF-1R+ TAMs upon RG7155 therapy was detectable. Analysis into whether patients with a stronger macrophage infiltrate benefit more from RG7155 therapy is ongoing, a question that will require treatment of a larger patient cohort to answer. As our in vitro data show that the presence of GM-CSF during macrophage differentiation inhibits RG7155 induced cell death, which has been demonstrated previously also for a small-molecule CSF-1R inhibitor (Pyonteck et al., 2013), it will be important to address in future clinical studies the role of tumor-derived GM-CSF on the efficacy of RG7155-mediated TAM depletion. In particular, pancreatic cancer has been described to express GM-CSF, and the role of this growth factor in inducing immunosuppressive myeloid-derived suppressor cells in mouse models has been reported (Bayne et al., 2012, Dolcetti et al., 2010, Bronte et al., 2000). Even if TAM depletion is diminished in GM-CSF-expressing tumors, inhibition of CSF-1R signaling can alter the polarization of macrophages and hence be therapeutically beneficial (Pyonteck et al., 2013). However, tumor types to be treated with RG7155 will have to be selected carefully, taking into account that TAMs have been also attributed good prognostic relevance, e.g., in colorectal cancer (Zhang et al., 2012).