Medulloblastoma is the most common primary pediatric malignancy of the central nervous system, which is divided into wingless (WNT), sonic hedgehog (SHH) and non-WNT/non-SHH molecular subgroups (1). Multimodal treatment consisting of surgery, radiation therapy and chemotherapy has significantly improved its long-term survival (2). However, approximately 30% of medulloblastoma patients present with metastases at diagnosis and their survival has remained unfavorable (3, 4).

Tumorigenesis is a complex and dynamic process. In recent years, the role played by the tumor microenvironment in promoting or inhibiting brain tumor growth has caused more concerns (5, 6). Tumor-associated macrophages (TAMs) are a key component of the tumor microenvironment that can either support tumor growth by promoting angiogenesis and immune suppression or suppressing tumor growth via pro-inflammatory effect (7). Studies have suggested that TAMs are the major immune cells in brain tumor microenvironment; moreover, macrophage polarization plays an essential role in the growth and progression of brain tumors (8–10).

At present, the role of TAMs in prognosis of medulloblastoma is still controversial. Margol et al. (11) first reported that TAMs was significantly higher in SHH medulloblastomas compared to other subgroups. Other study also showed that TAMs played an active role in SHH medulloblastoma by inhibiting tumor growth (12). A reduction or repolarization of TAMs can result in accelerated tumor progression. However, a recent study revealed that effective reduction of TAMs had a negligible role in medulloblastoma recurrence and metastatic spread (13). On the other hand, Lee et al. (14) suggested that high percentage of M1 rather than M2 showed worse outcome in SHH medulloblastoma patients, contrary to the common view on role of M1 phenotype in prognosis.

In our study, we described the characteristics of the activated phenotype (M1/M2) of TAMs in medulloblastoma and investigated the correlation between TAM recruitment and prognosis.

At present, the potential role of immune microenvironment in promoting or inhibiting human medulloblastoma progression has not been fully clarified (23). In this study, we found that TAMs increased specifically in the SHH subgroup. We also investigated the correlation between TAM recruitment and treatment outcome. It was showed that high M_total and M1 macrophages predicted better PFS, however, TAMs made no significant effect on OS.

The SHH subgroup is common and the most investigated type in medulloblastoma, approximately accounting for 30% of all patients (15, 24). The SHH signaling pathway plays an important role in actively orchestrating many aspects of cerebellar development and maturation (25). SHH subgroup is characterized by constitutive activation of the SHH signaling pathway leading to abnormal proliferation of the cerebellar cells (26). Our study showed that the quantity of TAMs was abundant in SHH medulloblastoma, which was consistent with the reports in other literatures (10, 11, 14). Moreover, the proportion of M2 macrophages increase in SHH medulloblastoma. These may be due to the significant increase of monocyte chemotactic protein-1 (MCP-1/CCL2), which could recruit TAMs and promoting M2 macrophage polarization, in human SHH medulloblastoma (12, 27). Furthermore, it was reported that tumor-derived SHH directly acts on TAMs to promote M2 polarization, mediated by the transcription factor Krüppel-like factor 4 (Klf4) (28). In addition, other signaling pathways are also involved, such as signal transducer and activator of transcription 3 (STAT3), which is necessary to maintain SHH signaling and can enhance macrophage proliferation and survival through multiple pathways (29, 30). Augmented M1 recruitment in SHH subgroup was not observed. The underlying mechanism is unclear and needs to be explored. Meanwhile, the finding of higher percentage of macrophages in patients younger than 3-year-old in this study may be explained by the fact that SHH cases were mainly under 3 years old.

In this study, high M_total and M1 macrophages were associated with better short-term prognosis, which was consistent with the classical anti-tumor effect of M1 macrophages in various tumor types (31–33). However, there are controversial results in other studies. No significant association between survival and macrophage infiltration in any subgroups was found in one study (10), whereas a negative correlation between M1 and prognosis of SHH subgroup patients was observed in another (14). Furthermore, macrophage abrogation decreased the survival of murine models (12). It has been showed that genetic aberrations and stages of disease make different effect on immune microenvironments including macrophages (34, 35). For instance, TAMs are mainly composed of different proportions of both microglia and monocyte-derived macrophages, and the extent and composition of TAMs are different in molecular subtypes (36). It was noteworthy that there were many differences in the proportion of molecular subtypes, age and number of patients with metastatic disease between Lee et al’s and our study (14). Moreover, Lee et al. mainly investigated the correlation between TAM recruitment and prognosis in SHH subgroup, while we validated it in all the groups of patients. Thus, these contradictory findings may be related to the heterogeneity of patients with different proportion of molecular subtypes in different studies including ours. It also implies the complexity and the lack of understanding of macrophages. With regard to the noncorrelation of TAM counts with OS, TAMs at diagnosis may not affect the overall outcome as the composition of TAMs would change with treatment and disease progression (13, 36). Our data represented just the TAMs characterization before surgery and chemoradiotherapy. Dynamic observations on TAMs during treatment may contribute to prognosis evaluation more helpfully. Further research is needed while cerebrospinal fluid might be a potential alternative (37).

It is reported that the immunoreactivity pattern is age-related in medulloblastoma (38). In order to investigate whether and how patient age affects the TAMs profile, the patients were divided into three age groups. We found that 0-3-year-old and 11-18-year-old children had higher M_total than 4-10-year-old children, and 11-18-year-old children had higher M1 than 4-10-year-old children. It implies that differentially targeting TAMs is crucial in different age groups of patients, and we need to focus on 4-10-year-old patients who may have a poorer immunotherapeutic efficacy on TAMs due to the low macrophage count (39).

It is well-known that the patients with metastasis had a poor prognosis. In this study, however, we found that M1 macrophages increased significantly in patients with metastatic disease. Probably, M1 macrophages were highly recruited by some tumor suppressors to enhance the phagocytesis ability (40), although this mechanism could not eventually resist the invasion of the tumor. The exact underlying mechanisms remain to be clarified. There was relatively high proportion of patients with metastatic disease, therefore PFS correlation with M1 might be confounded by the number of metastatic samples in this case. However, Cox proportional hazard regression model showed that M1 and metastasis were both independently prognostic factors of PFS, which avoided the confounding factors.

In this study, there were high percentages of CD68⁺ CD163⁺ HLA-DR⁺ cells with a mixed signature of M1/M2. The markers of M1 and M2 can be co-expressed on the same cell. A high amount of mixed M1/M2-like polarized macrophages exists in a variety of tumors (41, 42). The value of these macrophages for prognostic evaluation is not clear. We did not find any correlation between M_mix and prognosis in this study. A recent research showed that tumor reduction was associated with the percentage of macrophages with mixed M1/M2 phenotypes, and the outcome of these therapeutic strategies targeting TAMs was influenced by the half-life of various macrophage phenotypes (43). The abundance of these macrophages indicates the plasticity of TAMs and the potential of targeting TAMs for immunotherapy in medulloblastoma. It also reflects that the paradigm of categorizing TAMs into M1 and M2 does not fully represent the phenotypic diversity and comprehensive functions of TAMs, especially in brain immunobiology (44, 45). TAMs with current widely accepted biomarkers could play a role in both pro-and anti-tumor functions (46). For example, TAMs expressing M2-associated markers may display M1-like functions under a wide variety of stimuli (47). However, one limitation of the present study is lacking of functional interrogation of TAMs. Further studies are necessary for a more in-depth investigation of their functions.

TAMs are a promising target for tumor immunotherapy. Existing therapeutic strategies targeting TAMs mainly include adjusting the TAM population, reverting TAMs into M1 macrophages, regulating macrophagic phagocytosis signals and engineering macrophages to trigger phagocytosis (48). M1 and M2 macrophages have high degree of plasticity and can be converted into each other upon tumor microenvironment changes or therapeutic interventions (49–51). Therefore, the attempts to convert TAMs into M1 macrophages or reduce M2 macrophages have potential value in treatment of medulloblastoma. At present, animal trials have showed an anti-tumor effect of STAT3 inhibitors and CSF1R inhibitors in medulloblastoma (30, 52, 53). Thus, this therapy strategy would probably improve the patient prognosis in the future.

Higher total macrophages and M1 macrophages recruitment predicted a good outcome in medulloblastoma patients. TAMs might be a target of therapy for medulloblastoma.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by Ethics Committee of Beijing Shijitan Hospital, Capital Medical University. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin.

TW, SD, WW, and LS. designed experiments; JZ. carried out experiments; YW. assisted with multiplex immunofluorescence staining; XY and JZ. conducted image analysis; JZ, JL, YS, XG, SL, and YL. contributed to the acquisition of data and statistical analysis; JZ. wrote the manuscript; ZL and SD. revised the text of the manuscript. All authors contributed to the article and approved the submitted version.

This work was supported by National Science and Technology Key Projects (2017ZX09304029001). Capital Health Development Scientific Research Special Project (2020-2-2047) and Fund of Beijing Shijitan Hospital of Capital Medical University(2021-C04).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.