Osteoporotic fractures: an unrecognized adverse event of immune checkpoint inhibitors?

Abstract

The widespread use of immune checkpoint inhibitors (ICIs) in clinical practice has broadened our understanding of their immune-related adverse events (irAEs). IrAEs, including musculoskeletal adverse events, remain a significant concern. While ICI-associated arthritis is a well-documented musculoskeletal side effect of ICI therapy, the direct effects of ICIs on bone in patients with cancer are poorly understood. There is emerging evidence to support the hypothesis that ICIs adversely impact bone turnover and can lead to osteoporosis and fragility fractures, which are not currently recognized as irAEs.

The development of immunotherapy caused a paradigm shift in treating different types of metastatic cancers. Immune checkpoint inhibitors (ICIs) are one of the most widespread immunotherapeutic treatments for cancer. ICIs activate T lymphocytes by interfering with the immune checkpoint proteins: cytotoxic T-lymphocyte-associated antigen 4, programmed cell death protein 1 (PD-1), and programmed death-ligand 1 (PD-L1), leading to immune activation that recognizes and attacks tumor cells. Unfortunately, ICIs can have off-target effects called immune-related adverse events (irAEs), which have been reported in numerous organ systems.¹ While ICI-associated arthritis is a well-documented musculoskeletal side effect of ICI therapy, the effect of ICI on bone is not well understood, and fractures are not reported as ICI-associated adverse events in clinical trials. Based on well-established biological pathways between T cell activation and bone resorption and emerging animal and human data, there is increasing evidence that osteoporosis and fractures are unrecognized irAEs of ICIs.

Considering the mechanistic aspect, it is well established that bone metabolism is affected by various inflammatory cytokines.² Activated T cells directly impact bone remodeling by producing receptor activator of nuclear factor kappa-Β ligand (RANKL). In turn, RANKL and its receptor RANK, promote osteoclastogenic activity. Moreover, RANK activation triggers the expression of tumor necrosis factor (TNF) receptor-associated factors, causing osteoclast differentiation. In addition, T cells produce several cytokines, such as TNF-α, interleukin (IL)-18, IL-17, and IL-12, which contribute to RANK signaling pathways, enhancing osteoclast differentiation. Hence, the imbalance of osteoclast formation and maturation surpassing osteoblast formation leads to osteoporosis development, as shown in figure 1.

Figure 1. Graphical scheme of the hypothesized relationship between immune checkpoint inhibitors and the development of fractures. RANKL, receptor activator of nuclear factor kappa-Β ligand.

In an in vitro study of mice, the impact of ICIs on bone loss and fractures was investigated in the presence of bone metastases.³ Eight-week-old female C57Bl/6 mice injected with E0771 mouse mammary carcinoma cells were treated with anti-PD-1 for 2 weeks. Anti-PD-1 therapy resulted in a decrease in bone mineral density (BMD) and bone volume fraction in the treated mice compared with the control group. Moreover, the CD8+ effector memory T cells, a subset of immune cells capable of stimulating osteoclastogenesis through the release of interferon (IFN)-γ, were elevated in the bone marrow in the treated group by 1.8-fold (p<0.03). In addition, Ifng, the gene encoding IFN-γ, was highly expressed from 5.8-fold to 8.2-fold in the treated group (p<0.002) compared with the control group. These data indicate that anti-PD-1 may negatively impact bone turnover and microarchitecture.

Human data also supports the potential adverse impact of ICIs on bone turnover. In a prospective cohort study, 44 patients with renal cell carcinoma and non-small cell lung cancer treated with anti-PD-1 or anti-PD-L1 had their plasma levels of N-terminal propeptide of type I procollagen (P1NP), a marker of bone formation, and type I collagen C-terminal telopeptide (CTx), a marker of bone resorption, measured at baseline and after 3 months of ICI therapy.⁴ After 3 months of treatment, CTx increased significantly (p=0.045), and P1NP decreased with a trend toward statistical significance (p=0.073). This suggests that ICIs can result in an upregulation of bone resorption, coupled with a downregulation of bone formation, which could result in the development of osteoporosis. However, the clinical significance of these changes in bone turnover markers is unknown.

Several case series have reported clinical evidence of bone resorption, osteoporosis, and fragility fractures occurring in patients on ICI treatment.4,6 In a case series of six patients with skeletal adverse effects related to ICI therapy for metastatic melanoma, non-small cell lung cancer and renal cell carcinoma, three patients developed focal bone resorptive lesions, and the other three patients developed new osteoporotic fractures. Vertebral compression was observed in all patients with osteoporotic fractures, and two out of three had fractures in multiple sites.⁵ At the biochemical level, bone resorption markers were elevated in five out of the six patients. Another prospective cohort study found that 4 out of 44 patients (9%) undergoing ICI therapy experienced new lumbar fractures.⁴ In 2021, another case series reported four patients with cancer receiving ICI who developed osteoporotic fractures, of which three developed vertebral fractures.⁶ Taken alone, these case series do not establish causation between ICIs and osteoporotic fractures.

However, large database studies have also shown an association between ICI therapy and increased fragility fracture risk. Using the United States Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS) pharmacovigilance database, Filippini et al conducted a disproportionality analysis to examine the development of bone and joint injuries, including fractures, in 95,787 patients receiving at least one ICI drug. They compared reporting odd ratio (ROR) of fractures in ICI users and those treated with other therapies.⁶ 650 out of 95,787 patients experienced bone and joint injuries, including femoral neck fracture (26 cases, ROR=2.38), pathological fracture (46 cases, ROR=3.17), and spinal compression fracture (42 cases, ROR=2.51) as the most reported ICI-associated adverse events. A subsequent study, also using FAERS, examined musculoskeletal adverse events associated with ICI use and found 1010 reports of fractures, with the most prevalent being femoral, hip, and spinal fractures. Among these fractures, 42% were osteoporotic. Compared with other drugs, atezolizumab, nivolumab, and pembrolizumab were all associated with RORs greater than 1 for osteoporotic fractures, and falls were the most frequently reported comorbidity among all fractures.⁷ The authors concluded that their findings contribute to the hypothesis that ICIs may increase the risk of fractures by inducing osteoporosis and falls.

Fractures of ICI users have been examined at a population level. A retrospective administrative database study from Canada examined 1600 individuals with cancer treated with ICI and assessed their fracture risk pretreatment and post-treatment.⁸ The rate of major osteoporotic fractures (hip, vertebral, humerus, and wrist) was 11.3 per 1000 person-years in the year preceding ICI initiation, increasing to 27.3 per 1000 person-years in the year following ICI initiation, resulting in an incidence rate ratio of 2.43 (95% CI 1.34 to 4.27). These results have been reproduced using a large US insurance database, in which ICI users’ adjusted HR of fracture post-ICI initiation was 1.8 compared with the year before ICI treatment.⁹ Older age, female sex, prior fracture, and combination ICI therapy were risk factors for fracture. Fracture risk was independent of bone metastases.

Despite osteoporosis not being recognized as an irAE, osteoporosis treatments have been shown to potentially improve outcomes in ICI users. In an observational cohort study of 307 individuals on both denosumab and ICIs from the USA, the authors found that there was a statistically significant association between denosumab treatment duration, tumor response, and overall survival (p<0.0001).¹⁰ As such, antiresorptive agents, including denosumab and zoledronic acid, are actively being studied in ICI combination regimens.

Although a causal relationship has not been established between ICI and osteoporosis, and there is the possibility of residual confounding in some studies by important factors such as glucocorticoid exposure, bone metastases, fragility and falls, consideration of the available evidence should at the very least alert clinicians and researchers to the potential of long-term adverse effects of ICIs on bone health. Prospective long-term comparative studies of changes in BMD, bone strength, bone turnover markers, and fracture rates, including osteoporotic and pathologic fractures, will help us better understand the effects of ICIs on bone. While randomized control trials may have inadequate follow-up or sample size to ascertain differences in risk of osteoporotic fractures between ICI and placebo groups, we suggest that future clinical trials of ICI drugs should report fractures as potential irAEs so that their risk can be studied in a controlled setting. A better understanding of the impact of ICIs on bone can lead to earlier screening and interventions to prevent serious osteoporotic fractures. As indications for ICIs expand rapidly and patients are living longer with late-stage cancers, bone health will be an increasingly important factor in the quality of life of cancer survivors.