Blocking SIRPα can enhance antibody-dependent cellular phagocytosis (ADCP) by macrophages, features that have attracted significant attention for research (29–31) ( Figure 2 ). Microglia play a similar functional role to macrophages in central nervous system tumors. They function as the effector cells in the disruption of the CD47-SIRPα anti-phagocytic axis (32, 33). Generally, promoting ADCP is achieved by blocking the binding of SIRPα to CD47 to abolish the “don’t eat me” signal. Furthermore, in chimeric antigen receptor macrophages, SIRPα inhibition in macrophages can activate inflammatory pathways and the cGAS–STING signaling cascade, leading to an elevated production of proinflammatory cytokines, such as interleukin-1 (IL-1), tumor necrosis factor-alpha (TNF-α), reactive oxygen species (ROS), and nitric oxide, which increase the anticancer activity (34, 35). Moreover, preventing the expression of SIRPα in macrophages induces the recruitment and migration of T-cells via increased secretion of chemokines (e.g., C-C motif chemokine ligands CCL3 and CCL4) (26). In SIRPα-knockout (SIRPα-KO) mice, SIRPα-KO macrophages were found to display robust anticancer activity and antigen-presenting capacity, which was associated with enhanced T-cell activation and proliferation. Notably, SIRPα-KO macrophages were found to promote T-cell recruitment in cancers via a Syk–Btk-dependent mechanism involving CCL8 secretion, transforming tumor-associated macrophages and granulocytic myeloid-derived suppressor cells into subsets expressing high levels of CCL8 and H2-Q10, respectively, with enhanced antigen presentation, phagocytosis, inflammatory response, and chemotaxis capacities (36). Therefore, targeting SIRPα has the potential to reprogram the tumor immune microenvironment, promoting systemic anticancer responses and preventing solid cancer progression. In brief, by blocking the expression of SIRPα in macrophages, the traditional “don’t eat me” signaling pathway can be suppressed, which will improve phagocytosis and stimulate macrophages to secrete chemokines and cytokines via additional signaling pathways. Simultaneously, it has the potential to block the SIRPα-mediated non-CD47-dependent pathway, reprogramming the suppressive tumor immune microenvironment.

In cancer therapeutics, anti-SIRPα antibodies exert their antitumor effects by disrupting the CD47-SIRPα interaction and relieving inhibitory signaling on neutrophils.) (37). When combined with tumor-targeting antibodies, the Fc region of these therapeutic antibodies engages activating Fcγ receptors (e.g., FcγRIIIa) on neutrophils, triggering antibody dependent cell-mediated cytotoxicity (ADCC) and subsequently enhancing neutrophil-mediated tumor cell killing (3, 38). Sodium stibogluconate (SSG), a selective SHP-1 inhibitor, enhances neutrophil cytotoxicity by blocking phosphatase-mediated suppression of Vav1 and PLCγ2 signaling. Co-administration of SSG with CD47-SIRPα blockade amplifies ADCC efficacy through dual inhibition of immunosuppressive pathways (39). SIRPα signaling suppresses neutrophil phagocytic activity and cytotoxicity through the SHP-1/p38 MAPK/STAT3 pathway while promoting IL-6 and IL-17 secretion. After SIRPα-KO, neutrophils polarize toward the anti-tumor N1 phenotype, with enhanced phagocytic function and reduced inflammatory cytokine secretion, thereby inhibiting the growth of lung cancer (40). However, compared with IgA, IgG-mediated ADCC exhibits relatively low efficiency (41–45). Blocking SIRPα on neutrophils with anti-SIRPα antibodies significantly enhances ADCC mediated by IgA2 variants of cetuximab and trastuzumab against HER2-positive breast cancer cells and EGFR-positive epidermoid carcinoma cells (46). Paradoxically, SIRPα overexpression in autoimmune lesions (e.g., rheumatoid arthritis and inflammatory bowel disease) exacerbates inflammation through dysregulated innate immunity (47). During chronic inflammation, neutrophil-derived serine proteases cleave the SIRPα ITIM domain in an IL-17-dependent manner. The resultant truncated SIRPα retains CD47-binding capacity but loses inhibitory signaling, unleashing neutrophil chemotaxis, ROS production, and phagocytic activity (48).

As specialized APCs, dendritic cells (DCs) are crucial in facilitating T-cell activation and maintaining immune tolerance (49, 50). When DCs come into contact with cancer cells, they send a “don’t eat me” signal through the classic ITIM–SHP1 complex that mediates anti-phagocytic effects but also through SIRPα, which detects cancer mitochondrial DNA for cross-priming or activate the STAT3 signaling pathway to suppress the production of cytokines (such as IL-12, TNF-α, and interferon-γ) and consequently inhibit DC maturation. Additionally, the PI3K–AKT signaling pathway also plays a pivotal role in regulating the activation and maturation of DCs through SIRPα (51, 52). Combined therapy with radiotherapy/anti-SIRPα/anti-PD-1 for colorectal cancer was shown to effectively induce cGAS–STING signaling in DCs both in vitro and in vivo, facilitating efficient cross-presentation of tumor-associated antigens (53, 54). Moreover, when SIRPα was silenced in DCs, increased secretion of cytokines (e.g., TNF-α, IL-12, and IL-6), enhanced the secretion of interferon-γ by CD8+ T lymphocytes, and effectively killed cervical cancer cells in vitro ( 55). Of note, the interaction between SIRPα on DCs and CD47 on T-cells modulates the differentiation of naïve T-cells into T-helper (Th) cells. Mice lacking SIRPα exhibit enhanced resistance to autoimmune diseases caused by Th1 or Th17 cells, such as encephalomyelitis and colitis (56–59). Besides regulating T-cells by presenting tumor antigens, SIRPα can further influence the differentiation and function of T-cells by regulating their own maturation. Thus, blocking SIRPα can promote DC maturation and enhance their antigen-presenting function, thereby facilitating the function of cytotoxic T-cells.

Multiple anti-SIRPα antibodies developed for solid tumor treatment in preclinical studies have demonstrated significant efficacy in suppressing tumor progression (38, 69) ( Table 1 ). Yanagita et al. validated the tumor-inhibitory effect of the mouse-derived anti-SIRPα monoclonal antibody MY-1, which showed enhanced cytotoxicity against HER2-positive breast cancer cells in vitro. It significantly inhibited the growth of SIRPα-expressing renal cell carcinoma and melanoma cells, but not of non-SIRPα-expressing cells. Combination with rituximab or anti-PD-1 antibody further enhanced the ability of MY-1 to suppress the growth of Burkitt lymphoma and colorectal cancer cells. Moreover, when used as a monotherapy, MY-1’s anticancer activity against renal cell carcinoma and melanoma was mediated by macrophages, but also NK and CD8+ T -cells (60). In SIRPα-deficient mice, MY-1 monotherapy showed inhibition of cancer growth by binding to SIRPβ and promoting ADCP (70). The effects of MY-1 differ between tumors with and without SIRPα expression, indicating that endogenous SIRPα in cancer cells is involved in certain regulatory mechanisms.

Humanized SIRPα antibodies can effectively block various human SIRPα variants. Several antibody monotherapies each have their own characteristics. KWAR23 alone fails to induce macrophage phagocytosis. Moreover, no immune cell infiltration or obvious neurological abnormalities were observed in the brains of mice treated with KWAR23; however, it binds to SIRPγ and affects T-cell function (38). The phagocytic activity of SIRP-1 and -2 is important as monotherapy depends on the “eat me” receptor CD32 (FcγRII) in macrophages. SIRP-1 functions by directly blocking SIRPα and inducing internalization of the SIRPα/antibody complex, thereby reducing the levels of SIRPα in macrophages, while SIRP-2 alters the affinity of SIRPα for CD47 by affecting its dimerization/aggregation in macrophages (69). BR105 is ineffective when used alone; although it can mildly bind to SIRPγ, it does not inhibit T-cell activation. Toxicity studies in non-human primates showed that BR105 is well-tolerated, with no treatment-related adverse reactions observed (71). 1H9 exhibits a similar effect in inhibiting cancer progression without affecting T-cell function. When comparing anti-SIRPα and anti-CD47 antibodies using CD47/SIRPα double-humanized mice, it was found that 1H9 exhibits significantly reduced antigen sink effect and enhanced biosafety owing to the limited tissue distribution of SIRPα expression (72).

When combined with therapeutic antibodies, such as rituximab, all antibodies demonstrate significant inhibitory effects on the growth of hematological malignancies and solid cancers both in vitro and in vivo. Additionally, several antibodies when used in combination with ICIs exhibit good safety and therapeutic effects. Competition between hAB21 and cetuximab for macrophage FcγR limits the ability of anti-SIRPα antibodies to enhance macrophage phagocytosis. Alternatively, hAB21 with an active Fc structure can co-bind to SIRPα and FcγR in macrophages, leading to heterotrimeric interactions that restrict the binding of cetuximab to macrophage FcγR, thereby reducing phagocytic signaling. This phenomenon is known as the “scorpion effect.” When combined with anti-PD-1 or anti-PD-L1 antibody blockade therapy, hAB21 significantly inhibits the growth of tumor cells and does not cause anemia or other adverse outcomes when used in cynomolgus monkeys (11). CTX-5861 is a bispecific antibody targeting both SIRPα and PD-L1, designed to enhance macrophage phagocytosis and improve the efficiency of antigen presentation by DCs (73). AL008, a specific antibody targeting pan-alleles of SIRPα, demonstrates monotherapy efficacy by triggering SIRPα degradation and stimulating the activation of FcγR on bone marrow cells via its Fc domain. Additionally, the antitumor activity of anti-PD-L1 drugs has also been enhanced (74).

Although the aforementioned anti-SIRPα antibodies have not entered the clinical trial stage, some anti-SIRPα antibodies have demonstrated good biosafety and cancer treatment efficacy in preclinical studies and have thus entered clinical trials ( Table 2 ). ADU-1805, a humanized IgG2 anti-SIRPα antibody, does not affect T-cell activation or bind to red blood cells/platelets. In non-human primates, ADU-1805 exhibited no toxicity. Furthermore, ADU-1805 does not bind to macrophage FcγRIIA to trigger the “scorpion effect,” nor does it induce NK cell-mediated ADCC, lacks activity in mediating complement-dependent cytotoxicity, and does not stimulate cytokine secretion in human whole blood, further substantiating its clinical viability. ADU-1805 is undergoing clinical trials (NCT05856981), and the results are yet to be announced (27, 75). BI 765063 is a humanized IgG4 monoclonal antibody antagonist of SIRPα that binds with high affinity to SIRPαV1 but not to SIRPγ, thereby preserving T-cell function. Ongoing research (NCT05249426) to test whether different combinations of BI 765063, Ezabenlimab, chemotherapy, cetuximab and BI 836880 are helpful for patients with head and neck or liver cancer. Another clinical trial (NCT03990233) is currently evaluating the safety and efficacy of BI 765063 as monotherapy or in combination with ezabenlimab in patients with advanced solid tumors. BI 765063 monotherapy was found to be well-tolerated and showed activity, with treatment biopsies from responders demonstrating increased CD8+ T-cell infiltration and activation (76). Additionally, a clinical trial in Japan (NCT04653142) assessed the safe dose of BI 765063 in Japanese patients and found that its safety and pharmacokinetic parameters were consistent with those observed in Caucasian patients (77). A study (NCT05446129) aimed at evaluating the safety, feasibility, efficacy, and biological activity of the neoadjuvant treatment with Ezabenlimab combined with BI 765063 and pembrolizumab combined with BI 765063 in newly diagnosed patients with locally regional colorectal cancer has been dropped by the pharmaceutical company. BI 770371 is a pan-specific monoclonal antibody against SIRPα currently being evaluated the tolerability of different doses of BI 770371 when used alone or in combination with ezabbenlimab (NCT05327946). It is considered that the toxicity profile of BI 770371, both as a monotherapy and in combination therapy, is manageable. Another study (NCT05068102) aimed at finding out how the two drugs, BI 765063 and BI 770371, are absorbed in tumors and how they are distributed in the body is underway (78). CC-95251(BMS-986351) is a fully human monoclonal antibody targeting SIRPα, with preclinical studies showing its ability to enhance macrophage phagocytic activity when combined with the therapeutic antibody rituximab (79). A clinical trial (NCT03783403) is evaluating CC-95251 as a monotherapy and in combination with cetuximab and rituximab for safety, tolerability, and preliminary clinical activity in participants with advanced solid and hematological malignancies. Unfortunately, the clinical trial has been terminated owing to changes in business objectives (80). DS-1103a, a recombinant humanized IgG4 antibody targeting SIRPα, is currently being assessed in combination with T-DXd for its efficacy, recommended dosage, and pharmacokinetic properties in patients with advanced solid tumors (NCT05765851). IBI397, a pan-allelic antibody against SIRPα, underwent clinical trials for advanced malignant tumors but the trial (NCT05245916) was withdrawn owing to changes in the company’s development strategy.

Current developments of CD47–SIRPα signaling pathway inhibitors can be roughly categorized into three types: (i) blockers of CD47 molecules in target cells, which includes anti-CD47 antibodies and SIRPα-Fc fusion antibodies, (ii) blockers of SIRPα molecules in immune effector cells, and (iii) inhibitors of glutaminase-like proteases (81). Anti-CD47 antibodies have been shown to achieve objective (total or partial) remission in 50% of patients by showing considerable anticancer activity in hematological malignancies. However, treatment of solid cancers has led to adverse effects, including anemia (57% of patients) and lymphocytopenia (34% of patients) (82–84). Although strong effects in preclinical studies were observed, especially those that retain large Fc receptor (FcR) inactivation potential in human IgG1 molecules, their clinical value may be limited by non-tumor toxicity (18). The primary reason is that CD47 lacks cancer specificity and is widely distributed in healthy tissues, leading to a substantial “antigen sink”; thus, high doses of anti-CD47 drugs are required to attain anticancer efficacy. Moreover, many anti-CD47 antibodies retain effector functions via their immunoglobulin Fc domains, which may trigger macrophages to engage in ADCP against healthy cells (85–87). Indeed, anemia and thrombocytopenia are common side effects of such anti-CD47 antibodies, often requiring red blood cell transfusions and low-dose initiation strategies to mitigate the adverse situation (82, 88). To manage these risks, current research on anti-CD47 antibodies is focused on molecules that reduce FcγR binding ability, such as IgG4 antibodies. Most of these molecules can still induce severe anemia in non-human primates and cancer patients. Moreover, anti-CD47 antibodies may affect how much CD47 interacts with other receptors, such as integrins, vascular endothelial growth factor receptor-2 (89), thrombospondin-1 (90), and SIRPγ (91). Notably, blocking the interaction between CD47 and SIRPγ can inhibit T-cell extravasation and activation, thereby diminishing the anticancer response. Hence, CD47 signals appear to have a more complex biological functions and its blockade may elicit unexpected cellular responses (3, 27). Additionally, the anticancer activity of anti-CD47 antibodies depends on CS1 glycoprotein antigen (SLAMF7) phagocytic signaling, which is generally absent in solid cancers but is expressed in hematological malignancies (86, 92). Since 2022, multiple Phase III clinical trials of magrolimab were terminated or suspended owing to a lack of survival benefits or adverse reactions, with the regulatory agency also pausing some clinical studies of magrolimab in solid cancers (93). The side effects caused by non-targeted cancer cells and the negative impact on the interaction between CD47 and other receptors have become major obstacles limiting the widespread application of first-type antibodies in the treatment of solid cancers (94).

SIRPα is predominantly expressed in myeloid cells, including monocytes, granulocytes, DCs, macrophages, and microglia, which demonstrates a more limited histological distribution than CD47. SIRPα blocking agents are less likely to be influenced by constraints on antigen expression. Therefore, therapies targeting SIRPα have the potential to avoid side effects associated with targeting CD47 (95, 96) ( Table 3 ). Research on anti-SIRPα antibodies indicates that, similar to CD47 blocking antibodies devoid of Fc, SIRPα blocking agents lacking Fc can effectively induce anticancer immune responses when used along with T-cell-targeted therapies (11). Moreover, monotherapy with anti-SIRPα can alter the composition of the immune cell population in the tumor microenvironment, as evidenced by a significant increase in the proportion of M1 macrophages and a decrease in M2 macrophages (60, 97). SIRPα can negatively regulate DC activation and maturation, thus inhibiting SIRPα can enhance DC responses (52). Anti-SIRPα antibody therapy can stimulate an influx of tumor-infiltrating NK cells and CD8+ T-cells, as well as induce DC activation and promote T-cell effector function when used in combination with anti-PD-1 antibodies (36). The blockade of the SIRPα–CD47 signaling pathway combined with T-cell ICIs can enhance adaptive immune responses. Although the strategy of inhibiting SIRPα has advantages, such as increased antitumor responses and lack of red blood cell toxicity, the high polymorphism rate of the distal IgV domain in the extracellular region of SIRPα raises a risk of cross-reactivity with other members of the SIRP family. This makes the development of clinically beneficial SIRPα inhibitors particularly challenging (27).When glutaminase-like proteases are inhibited, newly synthesized CD47 molecules are unable to effectively bind to their natural binding partners owing to the lack of pyroglutamate modification. Unlike antagonistic molecules targeting CD47 or SIRPα directly, small-molecule inhibitors for this pathway do not compete with natural binding partners in the tumor microenvironment. Moreover, small-molecule inhibitors have high tissue penetration and potential oral bioavailability, which makes them an attractive option. However, the risk of blocking other functions of CD47 persists with small-molecule inhibitors (81). Collectively, anti-SIRPα antibodies capable of blocking all SIRPα alleles hold promise as competitive candidates for achieving the clinical goal of halting the progression of solid cancers.