The immune environment of bone sarcomas

Immune environment of osteosarcoma

Osteosarcoma (OsA) is the most frequent bone sarcoma affecting mainly children and teenagers (median age at diagnosis: 15 years) [ ].

The biology of OsA is complex. It is classified as a sarcoma with complex genomic and is characterized by chromosomal instability with high levels of somatic structural variations and copy number alterations. Recurrent DNA alterations affecting a variety of genes are encountered, the most frequent one being alteration of RB and TP53.

Besides these intrinsic genetic factors, microenvironmental factors seem to play an important role in osteosarcomagenesis and progression. Indeed, the OsA microenvironment is heterogeneous and consists of multiple cell types: bone cells (osteoclasts and osteoblasts), hematopoietic precursors, and immune cells. All these cells interact with OsA cells and with each other thus affecting tumor progression.

Recently, more insight into OsA microenvironment shed light on the role played by the immune effectors encountered in OsA environment [ , ] and which will be discussed in the next sections.

Tumor-infiltrating lymphocytes and OsA progression

Tumor-infiltrating lymphocytes (TILs) encompass all CD3+ T cells. These cells can exert a specific cytotoxic antitumor activity (CD8+ CTL), promote the antitumor response (CD4+ Th1), or limit it (CD4+ Th2, CD4+/FOXP3+Treg).

It was shown that a high CD8+ T cells/FOXP3+ T cells ratio is associated with better survival for OsA. Thus, a balance in favor of proinflammatory CD8+ CTL over antiinflammatory FOXP3+ LTreg is a good prognostic factor in OsA [ ]. Those results were confirmed first in vitro , as TILs derived from patient biopsies show direct cytotoxic activity against allogenic OsA cell lines, and in vivo in a rat orthotopic OsA model [ ]. More recently, a large study analyzed the expression of immune cells including CD8+, CD4+ T cells, and CD4+ T reg in soft tissue and bone sarcoma and confirmed the presence of these immune effectors in all OsAs [ ].

Tumor-associated macrophages and OsA progression

Tumor-associated macrophages (TAMs) play role in invasion, angiogenesis, and metastasis processes. According to microenvironment stimuli, macrophages adopt either a proinflammatory (M1) or an antiinflammatory (M2) phenotype, displaying different functions. Classically activated M1 macrophages produce proinflammatory cytokines and can exert cytotoxic antitumor activity. On the opposite, alternatively activated M2 macrophages produce proinflammatory cytokines and inhibit antitumor immunity.

Although the abundance of TAMs has been validated in OsA by several studies, their role in OsA progression is less clear. A study showed that a high density of total CD14+ macrophages was associated with better overall and metastatic-free survival [ ]. However, no impact of the density of M1 macrophages (CD14+/HLADRa) and M2 macrophages (CD14+/CD163+) on patient outcome was demonstrated. In opposite, another study showed that a significant proportion of CD68+ macrophages was correlated with poorer event-free survival at 5 years (45.5% for “CD68 high” vs. 84.4% for “CD68 low”) [ ]. Some studies have focused on the implication of primary OsA's immune environment on metastatic dissemination. One showed a significant increase of M1 macrophages in tumors of nonmetastatic patients compared to metastatic ones [ ] whereas a more recent study established that a high proportion of CD163+ M2 macrophages in primary OsA was associated with better patient survival and slower metastatic dissemination [ ].

The discrepancies observed between these studies can be partly explained by different markers used to characterize M2 TAMs (CD68+, CD14+, iNOS+, CD163+) and by the high plasticity of these macrophages moving from one phenotype to another. It is difficult to characterize these immune effectors at a given time point knowing that the immune environment is constantly changing. Nevertheless, all these studies indicate that TAMs present in primary tumor could contribute to prevent metastatic dissemination of OsA.

Immune checkpoints and OsA progression

Immune responses are tightly regulated by costimulatory or coinhibitory molecules known as ICPs. The interaction of ICPs with their receptor expressed on T cells will activate or inhibit T cell activity, either in the secondary lymphoid organs or in the tumor microenvironment. The prognostic value of ICP expressed in OsA has been studied. Initial studies focused on PD-1/PD-L1 and TIM3; most recent work focuses on B7H3.

• PD-1/PD-L1

The interaction of the ligand PD-L1 with its receptor PD-1 inhibits the proliferation of CTL, their production of cytokines, and thus their cytotoxic activity [ ].

Prognosis value of PD-L1 expression in OsA has been exhaustively studied. While the different studies show variable expressions of PD-1/PD-L1 in OsA—possibly related to the analytical methods and positivity threshold (set either at >1% or> 5%s of PD-L1+ cells) used—PD-L1 appears to be expressed in OsA ( Table 16.1 ). A recent meta-analysis involving 868 sarcoma patients showed that PD-L1 expression in the primary tumor is a poor prognostic factor in OsA [ ]. The expression of PD-L1 in OsA pulmonary metastases is less studied, but a higher frequency of PD-L1 expression is found in metastases. In fact, 48% of OsA patients express PD-L1 in lungs metastases while only 13% show a PD-L1 expression in the primary tumor [ ].

• TIM3

TIM3 is expressed on the surface of CD8+ CTL and CD4+ LTh1. The binding of TIM3 to its ligand Gal-9 induces LTh1 depletion and apoptosis. In OsA, the high concentration of soluble TIM3 in serum was found to be correlated with larger tumors and more frequent metastases [ ]. As another study showed that TIM3 is expressed in 69% of OsA patients, the level of its soluble circulating form could be a prognosis marker for OsA. Further research studies are needed to better understand TIM3 role, particularly when expressed by intratumoral TILs as a study showed that TIM3 is expressed by intratumor TIL in 69% of OsA patients [ ].

• B7H3

B7H3 is expressed on the surface of antigen-presenting cells (APCs) or tumor cells. The interaction of B7H3 to its yet unknown T cell receptor induces a decrease in T cell proliferation and IL2 and IFNγ secretion [ ]. A study showed an expression of B7H3 by 86% OsA patients; and its high expression (both in intensity and density) being a poor prognosis factor and correlated with a risk of relapse [ ].

• CD47/SIRPA

The interaction of SIRPA, expressed on macrophages, with CD47, expressed on tumor cell is called the “don't eat me signal” preventing the phagocytosis of tumor cells by macrophages. A study of tumor-associated ICP expression was conducted in a large cohort of bone and soft tissue sarcoma containing 226 OsA ²¹⁶ . Interestingly, it showed that SIRPA was expressed by TAM in 30% of OsA patients and CD47 was expressed by tumor cells in 47% patients. In addition, the expression of CD47 is associated with poorer progression-free survival [ ].

It appears that numerous immune effectors either cellular or molecular are involved in OsA progression but the particular context in which this tumor arises must also be taken into account and the interactions of bone homeostasis effectors with immune ones gave rise to the concept of osteoimmunology.

Osteoimmunology

The bone environment is known to be interconnected with immune cells.

Indeed some cytokines produced by immune cells regulate bone remodeling. T cells particularly influence the activity of osteoclasts. The secretion of IFNγ by CTL and LTh1 inhibits osteoclastogenesis. Indeed, IFNγ inhibits the expression of CSF1R by osteoclastic precursors, thus preventing the effect of CSF1 [ ]. LTreg and LTH2 cells produce TGF-β and IL-4 which also inhibit the formation of osteoclasts [ , ]. APCs, such as dendritic cells (DCs) and macrophages, can transdifferentiate into osteoclasts under the action of CSF1 and RANKL. They therefore act as osteoclastic precursors [ , ].

Resident macrophages from the bone environment are also called osteomacs (contraction of osteoclasts and macrophages) [ ]. Under physiological conditions, these osteomacs are mainly M2 macrophages and are involved in the maintenance of bone homeostasis, remodeling, and resolution of inflammation. In the early inflammatory stages, the osteomacs are activated as M1 proinflammatory macrophages [ ].

Osteomacs, preferably of the M2 phenotype, are known to interact with osteoblasts and activate bone formation via the secretion of BMP-2 and TGF-β. Indeed, it has been shown that in the absence of osteomacs, bone mineralization resulting from osteoblast activity is drastically decreased [ ]. M1 macrophages secrete proinflammatory cytokines such as TNF-α, IL-6, and IL-1β responsible for the expression of RANKL, which activates bone resorption [ ]. While M2 macrophages produce antiinflammatory cytokines IL-10 and TGF-β which inhibit bone resorption [ ]. The effects of different immune cells on bone remodeling are shown in Table 16.2 .

This illustrates the interconnections between bone remodeling and immunity and the variety of macrophages actions depending on their phenotype. It is therefore likely that cells from the OsA immune environment interact with cells from the bone environment to regulate tumor progression.

To summarize, OsA presents intratumor immune infiltrates composed of different actors that regulate the antitumor immune response: CTL, LTReg, TAMs, and ICPs.

A higher proportion of CD8+ TILs than LTReg is a good prognostic factor. TAMs remain the major immune population of OsA since they can represent up to 80% of the tumor proportion. A higher proportion of TAMs appears to correlate with better survival and prevent metastatic spread. Finally, to escape immune monitoring, OsA cells express ICP, such as PD-L1 and B7H3, whose expression is a poor prognostic factor. However, the description of ICPs remains largely incomplete in OsA and needs to be further explored.

Fig. 16.1 below recapitulates the current knowledge of OsA immune environment.

The knowledge we have of OsA' s immune environment shows that immunotherapies could be used in this bone sarcoma. Some of these will be reviewed in the section dedicated to bone sarcoma immunotherapies.