CD16⁺ Macrophages: An Emerging Biomarker for Combined CTLA-4 and PD-1 Blockade

In this issue of Clinical Cancer Research, Lee and colleagues report an association between CD16⁺ macrophage cell density in pretreatment melanoma tissue specimens and clinical outcomes following administration of combined CTLA-4 and PD-1 blockade to patients with treatment-naïve metastatic disease (1). CD16⁺ macrophage densities were higher in patients that experienced partial or complete responses to therapy compared with those with stable or progressive disease (P = 0.0041), and patients with higher CD16⁺ macrophage densities were also found to have significantly improved 3-year progression-free survival (87% vs. 42%; P = 0.0056). Importantly, no association between CD16⁺ macrophage density and clinical outcome was observed in a separate cohort of 50 patients treated with PD-1 monotherapy, indicating specificity of this predictive biomarker to combination CTLA-4 plus PD-1 checkpoint blockade.

Predictive biomarkers for combined CTLA-4 and PD-1 blockade, administered in melanoma as ipilimumab and nivolumab (ipi-nivo), remain a major unmet clinical need in the treatment of unresectable melanoma. Across tumor types including melanoma, tumor-intrinsic and immune biomarkers of response to PD-1 monotherapy have been well characterized including PD-L1 expression, IFNγ gene expression signatures, tumor mutational burden, and microsatellite instability (2, 3). In clinical practice, however, these features are not typically used to select patients with melanoma for anti-PD-1 monotherapy as response rates to pembrolizumab and nivolumab are considerably higher in unselected melanoma populations (4, 5) compared with other options such as chemotherapy. While most patients with unresectable melanoma will be treated with an initial regimen including a PD-1 inhibitor, oncologists must decide which patients warrant initial treatment with ipi-nivo, which is associated with higher response rates but also higher rates of severe immune-related toxicity compared with anti-PD-1 monotherapy (6). The recent regulatory approval of the LAG-3 inhibitor relatlimab in combination with nivolumab for previously untreated, unresectable melanoma has further complicated decision-making. There are currently no widely accepted biomarkers to guide the choice of upfront therapy, and clinicians rely on a combination of patient factors including overall fitness and the presence of high-risk features such as brain metastases (7) to identify patients that may need ipi-nivo as their first treatment. In this context, the findings by Lee and colleagues may represent an important step toward developing a clinically applicable, biological rationale for choosing among different first-line ICI regimens in the treatment of advanced melanoma.

Identification of CD16⁺ macrophages as a predictive biomarker for combined PD-1 and CTLA-4 blockade potentially supports a proposed role of the Fc region of ipilimumab in its mechanism of action (Fig. 1). While ipilimumab is known to facilitate costimulatory signaling during T-cell priming by antigen-presenting cells (8), some studies have highlighted the importance of immune microenvironment remodeling through the binding of ipilimumab's Fc region to Fc gamma receptors (FcγR), expressed on innate immune cells. CD16 is a low-affinity FcγR that is expressed on monocytes and macrophages and plays a role in antibody-dependent cellular toxicity (9). Downstream sequelae of ipilimumab-FcγR binding in murine tumor models includes depletion of suppressive regulatory T cells (T_reg; ref. 10), and in T_reg-ablated conditions, reduction in suppressive macrophages and concomitant activation of type I IFN signaling (11). In humans, while T_reg depletion has not been demonstrated following treatment with ipilimumab (12), polymorphisms in the CD16a gene that increase binding to IgG1 antibodies are associated with improved overall survival after treatment with ipilimumab (13). Further translational research elucidating changes in the tumor immune microenvironment with CD16 engagement by ipilimumab may inform the identification of additional biomarkers and potential targets for novel immunotherapies.

As most research on tumor-associated macrophages (TAM) has focused on immunosuppressive populations that abrogate immune checkpoint inhibitor (ICI) efficacy through molecules such as IL10 and TGFβ, it may be noteworthy that CD16 expression in macrophages was found to be associated with benefit from combined CTLA-4 and PD-1 blockade (1). TAMs exist on a spectrum from a proinflammatory M1-like phenotype expressing MHC type II (MHC II) to an immunosuppressive M2 CD163⁺ phenotype (14), and have been shown to both help and hinder ICI efficacy. In a YUMM1.7 murine model of melanoma, selective depletion of CD163⁺ macrophages led to an increase in effector T-cell infiltrates and concomitant tumor regression (15). Studies in other model systems have suggested that PD-1 blockade itself may repolarize macrophages from an M2- to an M1-like phenotype, and that macrophages are necessary for antitumor efficacy of PD-1 blockade (16). Given their expression of complement pathway genes and CD86, the CD16⁺ macrophages described by Lee and colleagues are likely more M1-like, but functional characterization of this population in model systems and patients are needed to understand their role in mediating combined ICI efficacy.

The complex interplay between tumor-infiltrating T-cell and myeloid populations complicates a simplistic conception of “hot” versus “cold” immune microenvironments as predictors of ICI response. Canonically, T cell–inflamed RNA signatures (17, 18) and higher levels of CD8⁺ tumor-infiltrating lymphocytes (19) have been associated with better outcomes of PD-1 response. However, how T cells influence the infiltration and behavior of TAM subsets is an area of ongoing investigation. T cell–induced melanoma expression of CSF1 has been associated with higher levels of CD163⁺ TAMs and impaired response to PD-1 blockade (20). Conversely, in the current study, CD16⁺ macrophages are associated with expression with T cell–recruiting cytokines (CXCL9, CXCL10, CXCL11), and are higher in on-treatment biopsies of responding than nonresponding patients. Dissecting the specificity and directionality of cross-talk between tumor-infiltrating T cells and macrophage subpopulations may provide additional opportunities to enhance antitumor immune responses.

In parallel with lab-based experiments to elucidate the functional role of CD16⁺ macrophages in facilitating response to combined CTLA-4 and PD-1 blockade, there are concrete steps that can be taken to translate this observation into a clinical assay to inform patient care. In the immediate future, this candidate biomarker will need further validation in other ipi-nivo–treated melanoma cohorts, and can be prospectively incorporated into ongoing trials of novel agents used in combination with CTLA-4 and PD-1 blockade. A standardized scoring system of CD16⁺ macrophage staining will need to be developed collaboratively with melanoma pathologists to deploy an IHC assay at scale without dependence on imaging software. While not necessary for validation a predictive biomarker, additional on-treatment biopsies at later timepoints may also be informative to determine whether upregulation in CD16⁺ macrophages is durable over time, and to track concomitant changes in T_reg populations.

Beyond the further development of CD16 macrophage expression as a predictive biomarker, there are related opportunities to leverage emerging high-plex spatial imaging technologies to identify additional biomarkers to address this important clinical question. For instance, identifying immune cell populations that positively correlate with response to combined ipi-nivo and are agnostic or negatively correlate with response to PD-1 monotherapy may be the most useful in identifying patients that truly need up-front addition of CTLA-4 blockade. Alternatively, biopsies of PD-1 refractory melanoma that subsequently respond to ipi-nivo may represent another rich setting for biomarker discovery. In these and other contexts, robust bedside-to-bench characterization of the tumor immune microenvironment will ultimately allow for the application of precision medicine concepts to the field of immuno-oncology.

J.J. Luke reports service on data and safety monitoring boards for AbbVie, Agenus, Immutep, and Evaxion; participation on scientific advisory boards for (no stock) 7Hills, Affivant, BioCytics, Bright Peak, Exo, Fstar, Inzen, RefleXion, and Xilio and (stock) Actym, Alphamab Oncology, Arch Oncology, Duke Street Bio, Kanaph, Mavu, NeoTx, Onc.AI, OncoNano, physIQ, Pyxis, Saros, STipe, and Tempest; consultancy with compensation for AbbVie, Agenus, Alnylam, AstraZeneca, Atomwise, Bayer, Bristol Myers Squibb, Castle, Checkmate, Codiak, Crown, Cugene, Curadev, Day One, Eisai, EMD Serono, Endeavor, Flame, G1 Therapeutics, Genentech, Gilead, Glenmark, HotSpot, Kadmon, KSQ, Janssen, Ikena, Inzen, Immatics, Immunocore, Incyte, Instil, IOBiotech, Macrogenics, Merck, Mersana, Nektar, Novartis, Partner, Pfizer, Pioneering Medicines, PsiOxus, Regeneron, Replimmune, Ribon, Roivant, Servier, STINGthera, Synlogic, and Synthekine; and research support (all to institution for clinical trials unless noted) from AbbVie, Astellas, AstraZeneca, Bristol Myers Squibb, Corvus, Day One, EMD Serono, Fstar, Genmab, Hot Spot, Ikena, Immatics, Incyte, Kadmon, KAHR, Macrogenics, Merck, Moderna, Nektar, Next Cure, Numab, Palleon, Pfizer, Replimmune, Rubius, Servier, Scholar Rock, Synlogic, Takeda, Trishula, Tizona, and Xencor. In addition, J.J. Luke holds patents (both provisional) #15/612,657 (Cancer Immunotherapy) and PCT/US18/36052 (Microbiome Biomarkers for Anti-PD-1/PD-L1 Responsiveness: Diagnostic, Prognostic and Therapeutic Uses Thereof). No disclosures were reported by the other author.

J.W. Smithy acknowledges NIH grant P30CA008748. J.J. Luke acknowledges NIH grants UM1CA186690-06, P50CA254865-01A1, P30CA047904-32, and R01DE031729-01A1.