Inhibitors targeting programmed death-1 (PD-1) and its ligand PD-L1 have profoundly reshaped cancer therapy. Since the first PD-1 inhibitor received regulatory approval in 2014, these agents have significantly improved overall survival (OS) and progression-free survival (PFS) rates (1), becoming standard treatments for a wide range of malignancies, including non-small cell lung cancer (NSCLC), melanoma, head and neck squamous cell carcinoma, and renal cell carcinoma (2).

Beyond their immunomodulatory effects on T cells, the PD-1/PD-L1 axis plays a multifaceted role in bone metabolism. Murine models with PD-1 or PD-L1 gene knockout exhibit pronounced osteoporotic phenotypes, including decreased trabecular bone volume, disrupted microarchitecture, elevated osteoclastogenesis, and increased RANKL/OPG ratios (3). These outcomes are mediated via multiple signaling cascades that finely regulate bone-resorbing and bone-forming cells (4–6). However, the complexity of these mechanisms has led to conflicting results. While some clinical studies suggest that PD-1/PD-L1 inhibitors exert bone-protective effects (4), others report the opposite, documenting bone-related adverse events in patients receiving immune checkpoint inhibitors (ICIs) (7, 8), even as some individuals maintain stable bone mineral density (9).

At the molecular level, similarly inconsistent findings are observed. Some investigations describe a pro-osteogenic role for PD-1/PD-L1 blockade (10), while others note inhibitory effects on osteoblast differentiation (6). These contradictions are likely influenced by a convergence of factors, including patient-specific clinical features, the immune microenvironment, cell-type-specific responses, soluble PD-1/PD-L2 activity (11), and the bidirectional nature of the Wnt/β-Catenin pathway (5, 12) and PD-L1–PKM2 metabolic signaling (13).

Furthermore, clinical evidence links PD-1/PD-L1 inhibitors to a heightened risk of pathological fractures, vertebral compression fractures, and femoral neck fractures (14), contributing to a cumulative incidence of bone-related adverse events as high as 69% (15). Despite their prevalence and clinical significance, conditions such as osteoporosis and fragility fractures are not formally recognized as immune-related adverse events (irAEs) (16), underscoring a gap in clinical classification and management.

Figure 1 outlines the opportunities and challenges associated with PD-1/PD-L1 blockade in bone metabolism, emphasizing its potential protective effects alongside its regulatory complexity and the under-recognition of skeletal irAEs such as osteoporosis.

To elucidate the dual role of PD-1/PD-L1 inhibitors in bone metabolism and explore their implications in osteoporosis prevention, this review addresses the following key questions:

PD-1/PD-L1 immune checkpoint inhibitors have become a cornerstone of modern cancer immunotherapy. However, their effects on bone metabolism remain underrecognized and exhibit contradictory findings across studies. Gassner et al. reported that in cancer patients without bone metastases, treatment with PD-1/PD-L1 inhibitors led to a significant early reduction in the bone resorption marker CTX (from a baseline mean of 0.51 ng/ml to 0.42 ng/ml at week 3), while bone formation markers such as PINP and osteocalcin (OCN) increased after 4 months of treatment, suggesting a bone-protective effect (4). Conversely, Pantano et al. observed a marked increase in CTX-I levels and a downward trend in PINP levels after 3 months of immune checkpoint inhibitor (ICI) therapy, which correlated with poor treatment response and decreased survival (7).

Moreover, some studies have documented adverse skeletal events during ICI treatment, including vertebral compression fractures and osteolytic lesions (8). In contrast, others have reported relatively stable bone mineral density (BMD) in ICI-treated patients compared to non-ICI counterparts, suggesting a long-term bone-preserving effect (9). These conflicting outcomes underscore the complex and multifactorial nature of PD-1/PD-L1 blockade on bone metabolism, likely influenced by clinical heterogeneity such as age, sex, tumor type, treatment regimen (monotherapy vs. combination therapy), and baseline skeletal health (17).

Importantly, there remains a lack of standardized bone health monitoring protocols for patients receiving ICIs, potentially delaying the detection and intervention of subclinical bone metabolic disorders (18). This gap may contribute to skeletal-related adverse events (SREs), ultimately impairing therapeutic efficacy and quality of life. Therefore, the impact of PD-1/PD-L1 inhibition on bone health warrants greater clinical attention and systematic evaluation.

While the PD-1/PD-L1 axis plays a pivotal role in immune regulation, it also significantly influences bone metabolism, particularly osteoclast differentiation and function. Genetic deletion of PD-1 or PD-L1 leads to osteoporotic phenotypes, implicating this pathway in the maintenance of bone homeostasis (3). Within the tumor microenvironment (TME), upregulation of PD-L1 and CCL2 facilitates osteoclastogenesis by activating the JNK pathway and enhancing CCL2-mediated RANKL signaling, thereby promoting bone resorption (5). Additionally, soluble PD-1 (sPD-1), which is elevated in inflammatory settings, stimulates IL-17A production—a key mediator of osteoclast activation—resulting in accelerated bone destruction. In contrast, PD-L2 expression under inflammatory conditions appears to suppress osteoclastogenesis and confer bone-protective effects (11).

Osteoclasts, particularly in their activated state, can upregulate PD-L1 expression to inhibit T cell proliferation and cytotoxicity, contributing to an immunosuppressive microenvironment. This PD-L1 upregulation is itself modulated by pro-inflammatory cytokines such as IFN-γ and IL-6, forming a feedback regulatory loop (19, 20).

PD-1/PD-L1 inhibitors interrupt this axis and exert both direct and indirect effects on osteoclast activity. By inhibiting the JNK pathway, these agents reduce osteoclast proliferation and resorptive capacity, leading to decreased CTX levels (4, 5). They can also interfere with the STAT3/NFATc1 signaling cascade, impeding pre-osteoclast maturation and reversing osteoclast-mediated immunosuppression (20). This dual mechanism results in a bidirectional modulation of bone remodeling (4), with short-term treatment reducing TRAP+ osteoclasts by approximately 60%, lowering bone destruction scores by 40%, and decreasing CTX by 23%, while increasing PINP levels by 8% at 4 months (p = 0.02) (4, 5). For a detailed summary of these findings across preclinical and clinical contexts, see Table 1 .

However, prolonged ICI therapy has been associated with increased fracture risk, with adjusted hazard ratios nearing 1.6 (21), suggesting a shift toward net bone loss over time. Chronic inflammation during extended PD-1 blockade can enhance RANKL expression and osteoclast activation, further contributing to skeletal damage (3).

Wnt/β-Catenin signaling also plays a complex, bidirectional role in mediating the skeletal effects of PD-1/PD-L1 inhibition ( Table 2 ). The canonical Wnt/β-Catenin pathway synergizes with short-term PD-1 blockade to suppress osteoclastogenesis and increase bone mass (5). In contrast, the non-canonical Wnt5a–Ror2 axis promotes osteoclast formation and reduces BMD, with Wnt5a overexpression in the TME antagonizing the bone-preserving effects of PD-1 inhibitors (12). Moreover, chronic PD-1 deficiency can suppress canonical Wnt signaling through sustained inflammation, leading to a 72% increase in osteoclast numbers and a 12% decrease in bone density (22).

To mitigate this deleterious effect, concurrent administration of anti-DKK1 antibodies—targeting inhibitors of canonical Wnt signaling—has shown promise in reducing skeletal damage and restoring bone homeostasis in the context of long-term ICI therapy (23).

Although the PD-1/PD-L1 pathway has been shown to promote osteogenic gene expression and calcium deposition (10), it can also inhibit osteogenic differentiation (6). This inhibition may occur through suppression of the SHP2 signaling pathway, which relieves its inhibitory effect on NF-κB activation, thereby enhancing osteoblast differentiation and bone formation (24). These seemingly contradictory effects may be related to the PD-1/PD-L1–PKM2 axis (13) (see Table 3 ). The balance within this axis plays a critical role: PD-L1 promotes osteogenic differentiation by upregulating RUNX2 expression, increasing ALP activity by 42%, enhancing mineralization by 30%, reducing inflammation, and promoting oxidative phosphorylation metabolism (10). In contrast, activation of PKM2 leads to increased tetramer formation, inhibition of glycolysis, a 60% reduction in RUNX2 expression, and a 35% decrease in ALP activity, collectively impairing osteoblast differentiation (25). Notably, PKM2 inhibition can reduce osteoclast numbers by 41%, improve oxidative metabolism and mitochondrial function, and enhance bone volume/tissue volume (BV/TV) ratio by 25%, thus promoting osteogenesis (26).

Additionally, the bidirectional modulation of the Wnt/β-Catenin pathway influences osteoblast function and bone metabolism in patients receiving immune checkpoint inhibitors (ICIs) (see Table 4 ). Suppression of Wnt signaling significantly reduces bone mass (BV/TV by approximately 60%) and bone strength (by 33%), suggesting the need for bone-protective interventions during ICI therapy (27). Conversely, DKK1 can relieve Wnt inhibition, increasing osteoblast density (2.4-fold), osteoblast surface coverage (by 65%), and bone mineral density (by 9%). When combined with anti-PD-1 therapy, it further enhances antitumor immunity (28, 29), as β-catenin not only promotes osteogenesis but also induces PD-L1 expression (3–4-fold increase), contributing to a “self-limiting” immunosuppressive feedback loop. PD-1/PD-L1 blockade can reverse this suppression (30). From an ICI-centered perspective, PD-1 receptor activation promotes osteogenic markers via the ERK/β-catenin signaling pathway (ALP ↑ 61%, OCN ↑ 54%, mineralization ↑ 83%), whereas PD-1 blockade may weaken this osteogenic effect (10). Therefore, in cancer patients receiving ICIs, bone metabolism assessment must take into account the net outcome of these bidirectional regulatory mechanisms.

Nevertheless, clinical evidence suggests that the primary bone-related effects of ICIs stem from inhibition of osteoclast activity and indirect promotion of bone formation, rather than direct osteoblast activation (4). These findings offer important implications for future osteoporosis management strategies.

Figure 2 provides an integrated overview of the bidirectional regulatory mechanisms of the PD-1/PD-L1 axis in bone metabolism and the effects of its inhibition. It illustrates how the PD-1/PD-L1 pathway interacts with multiple signaling axes to regulate both osteoclasts and osteoblasts, and highlights the temporal differences in short- vs. long-term blockade, the influence of the bone microenvironment, and crosstalk with the Wnt/β-Catenin pathway.

Both clinical and mechanistic studies have reported contradictory findings regarding the role of the PD-1/PD-L1 pathway in bone metabolism regulation (4, 6, 10, 18). These discrepancies highlight the complexity of this pathway and the bidirectional effects of its inhibitors on bone homeostasis, suggesting a potential bone-protective role for PD-1/PD-L1 blockade. Compared with traditional bone-protective agents ( Table 7 ), PD-1/PD-L1 inhibitors act further upstream in the RANKL axis (5), exerting a dual regulatory effect—initial inhibition of osteoclastogenesis followed by promotion of osteogenesis (4). This complements the unidirectional effects of conventional agents such as bisphosphonates and denosumab (34, 35).

Notably, combining PD-1 inhibitors with denosumab has been shown to increase the objective response rate in tumors (from 37% to 54%) and reduce the incidence of skeletal-related events by approximately 30% (36). Similarly, their combination with zoledronic acid can synergistically enhance anti-tumor immunity and anti-osteoporotic effects through the activation of γδ T cells (37). These findings suggest that PD-1/PD-L1 inhibitors may offer a novel therapeutic avenue for osteoporosis prevention and the management of skeletal-related adverse events (SREs) (14, 15).

As evidence continues to accumulate, it is anticipated that within the next five years, osteoporosis and fragility fractures will be increasingly recognized as part of the spectrum of immune-related adverse events (irAEs) associated with immune checkpoint inhibitors (ICIs) (16). Looking further ahead, over the next decade, the integration of bone-targeted strategies with ICI-specific bone-protective protocols may emerge (33), aiming to balance anti-tumor efficacy with bone health preservation and reduce the risk of SREs.

In the coming 5–10 years, an evidence-based framework for comprehensive skeletal health monitoring and stratified prevention is expected to be developed and refined. As summarized in Table 8 , several emerging findings support this trend: baseline CT-derived BMD values have been shown to correlate with survival outcomes following ICI therapy (32), suggesting that CT-BMD should be incorporated into pre-treatment screening. The newly developed “Mel-ICI Fracture Score” can help identify high-risk individuals, with scores ≥4 warranting the initiation of bone-protective agents (31). Given that fracture risk increases 2.6-fold within the first year of ICI treatment, DXA assessments are recommended at therapy initiation (21). Moreover, AI-assisted QCT techniques, coupled with dual-track monitoring of BMD changes and serum bone turnover markers, can enable precise and timely surveillance of bone health (4, 9).

The establishment of a “Bone-Oncology Joint Clinic” model is also proposed, facilitating a closed-loop management approach—from baseline evaluation to regular follow-up and comprehensive intervention (38), as well as stratified care for patients with bone metastases (39). These innovations aim to address bone loss and provide continuity of care for patients who develop skeletal complications during ICI treatment (15), ultimately mitigating poor prognostic outcomes (7). Collectively, these strategies may enable dual optimization of tumor control and skeletal health during ICI therapy and significantly improve the quality of life and long-term health outcomes of cancer survivors.

whereas the non-canonical branch promotes osteoclast formation. Clinical observations suggest that immune checkpoint inhibitors primarily promote bone formation indirectly by suppressing osteoclast activity rather than by directly enhancing osteoblast differentiation.