Targeting tumor-associated macrophages in cancer therapy and understanding their complexity

Abstract

Macrophage infiltration has been identified as an independent, poor prognostic factor for patients afflicted with various cancer entities. However, the characterization of tumor-associated macrophages (TAMs) prior to and following cancer patient treatment has been limited. Our study analyzed tumor biopsies before and after anti-CSF-1R antibody treatment unraveling the nature of TAMs and providing novel insights into their phenotypic and functional characteristics in cancer.

Over the course of the past decade, non-malignant cells and in particular tumor-associated macrophages (TAMs) have emerged as important players in the pathogenesis of cancer leading to the pursuit of new therapeutic intervention strategies in clinical trials. These macrophage-targeting immunotherapies primarily aim to block colony-stimulating factor-1 receptor (CSF-1R) signaling due to its essential, non-redundant function in differentiation and survival of macrophages.¹ Both small molecule tyrosine kinase antagonists (e.g., PLX3397) as well as antibodies (e.g., AMG820, IMC-CS4 and RG7155) are currently being tested clinically to evaluate the safety and therapeutic benefit of TAM depletion for patients with advanced solid malignancies, either as monotherapy or in combination with novel and standard treatment modalities.²

Macrophages are characterized by their highly plastic nature integrating and adapting to their locally encountered cytokine/chemokine milieu. The 2 extremes of a continuum of various macrophage phenotypes have been termed as “M1-like” tumor cell killing and antigen-presenting macrophages in contrast to the tumor-promoting “M2-like” macrophages providing growth and pro-angiogenic factors.³ Accordingly, the presence of macrophage infiltration ranges from either association with poor or favorable prognosis depending on the tumor type.⁴ Furthermore, tumors undergoing conventional treatments can alter the macrophage infiltrate and exploit the tissue repair functions of TAMs potentially driving the development of resistance to therapy-induced apoptosis.⁵ Thus, the prognostic relevance of TAM infiltration should be taken into consideration prior to cancer patient treatment. In support, it has been reported that TAMs in gastrointestinal stromal tumor (GIST) patients can be re-programmed by imatinib from an antitumor M1-like state toward a tumor-promoting phenotype.⁶ In light of the plastic nature of TAMs and considering that cancer patients from Phase 1 trials typically have undergone multiple lines of conventional and/or targeted therapies, we did not rely on archived primary tumor tissue to assess baseline TAM infiltration. Rather, we implemented a stringent biopsy sampling schedule in our first clinical trial investigating RG7155 induced effects in paired pre- and on-treatment biopsies from every patient.²

For the analysis of RG7155 treated tumors we evaluated several markers to determine the specific macrophage receptor expression pattern in patient tumors. The marker set was obtained by comparing in vitro differentiated macrophages to undifferentiated monocytes of healthy volunteers. CD163, a scavenger receptor, was confirmed as the most reliable macrophage-specific marker. To detect CSF-1R expression in situ, we generated a high-affinity antibody and optimized the immunohistochemistry (IHC) protocol including an amplification step that allowed detection of low levels of CSF-1R expressed on macrophages. Indeed, combination of CSF-1R and CD68/CD163 duplex IHC assays on consecutive slides enabled monitoring of TAM depletion in RG7155 treated patients.² In the course of our studies we found that different TAM subpopulations can be identified in patient specimens. Macrophages characterized either CD68⁺/CD163⁻/CSF-1R⁻ vs. CD68⁺/CD163⁺/CSF-1R⁺ revealed not only distinct morphological shapes but also differing localization within the tumor: the CSF-1R⁺ TAMs were found in the adjacent tumor stroma whereas CD68⁺CSF-1R⁻ macrophages were localized intratumorally (Fig. 1). Therefore, characterization of RG7155 susceptible TAM subpopulations demands a distinct surface marker expression profile in addition to analysis of cell morphology and spatial distribution patterns in situ. In our opinion the selected panel of markers and techniques of analysis were sufficient to demonstrate the mechanistic activity of RG7155 in various solid malignancies.

Figure 1.

Macrophage subsets show distinct morphology, surface marker profile and localization within the tumor and stromal microenvironment. (A) Consecutive slides of primary human breast cancer stained by immunohistochemistry for either colony-stimulating factor 1 receptor CSF-1R (left image) or CD68/CD163 duplex (right image), illustrating distinctive distribution patterns of macrophages. (B) Magnification of CD68/CD163 duplex staining showing that CD68⁺CD163⁺ double positive macrophages tend to cluster in the stromal compartments adjacent to cohesive tumor masses, whereas the majority of intra-tumorally localized macrophages can be assigned to a CD68⁺CD163−immunophenotype. Images provided by Friedrich Feuerhake, Roche Pharma Research and Early Development Penzberg (current affiliation: Hannover Medical School).

Given the high plasticity of macrophages,³ the use of only 3 markers is, however, not adequate to comprehensively discriminate phenotypic and, most importantly, functionally distinct macrophage subpopulations in vivo. Several additional markers have been described as characteristic for pro- vs. anti-tumoral macrophages and previously analyzed in patient-derived tumor tissues.⁷ However, to our knowledge, only one study reported on the successful isolation of TAMs from patients’ carcinomas and their subsequent functional profiling using in vitro assays.⁶ The approach used in this study represents a major advancement and complements the histological analyses potentially impacting the prognostic value of the TAM infiltrate assessed in many different tumor types. However, TAM isolation for functional characterization is technically challenging and requires rather large tumor specimens. This caveat is a significant limitation as in most clinical trial settings only small tumor biopsies can be obtained. Hence, the combination and the increasing sensitivity of technologies such as transcriptome sequencing, multiplex IHC, flow cytometry and single cell mass-spectroscopy will be key to enable us to better characterize macrophage infiltrates particularly with regards to their functional phenotypes. We hypothesize that using these novel technologies on RG7155 treated and untreated tissue samples will significantly improve interpretation of data in terms of separating background (e.g., stromal and tumor-derived) signals from true macrophage-derived signatures. A detailed characterization of TAMs, including their functional properties, marker expression, exact localization and polarity switching under therapeutic pressure, is essential to identify the patients who will benefit the most from RG7155 therapy. In particular, previous treatment-induced changes in the TAM infiltrate are relevant, because the macrophage-depleting agent RG7155 will mainly be developed as combination partner for conventional or targeted cancer therapies. Immunotherapies administered in combination with RG7155, in particular, might be clinically beneficial by virtue of the rewiring of the innate immune cell-signaling network mediated by pro-inflammatory cytokines. Additionally, we are also interested in the changes in the residual macrophage population after RG7155 treatment.

Furthermore, it should be considered whether it is useful to refer to the presence of a specific macrophage subtype as sufficient to predict patient outcome, as several other myeloid cell types such as tumor-associated neutrophils and myeloid-derived suppressor cells have been described as capable of providing similar immunostimulatory or immunosuppressive signals.^8,9 A systematic and harmonized characterization of the “myeloid collective” within the tumor microenvironment will improve our understanding of the biology of myeloid-driven immunosuppression in cancer patients.

Disclosure of Potential Conflicts of Interest

M.A.C., C.H.R., S.H. and D.R. are employees of Roche and are inventors of granted and pending patent applications relating to RG7155. M.A.C., C.H.R. and D.R. are stockholders in F. Hoffman La Roche.