Optimizing antiresorptive treatment in patients with bone metastases: time to initiation, switching strategies, and treatment duration

Bone is a common site of metastasis that occur in most tumor types, especially in breast, prostate, and lung cancer [1, 2]. The presence of bone metastases can lead to complications known as skeletal-related events (SRE) such as pathological fractures, spinal cord compression, need of radiation therapy due to bone pain, hypercalcemia of malignancy, and need of orthopedic or neurosurgical procedures [1]. These events can negatively affect patients’ quality of life [3] and have been associated with increased mortality [4‐6] and increased resource utilization [7].

Treatment with antiresorptive agents, including bisphosphonates or denosumab, is highly recommended in cancer patients with bone metastases, since it reduces the risk of SRE [8, 9]. However, several issues on the optimal use of antiresorptive agents in respect of time to initiation, therapeutic strategy after a SRE while on antiresorptive treatment, and duration of treatment remain unclear due to the paucity of evidence. The current international guidelines recommend that therapy with an antiresorptive agent should be initiated in the presence of a documented metastatic bone lesion, a switching strategy from bisphosphonates to denosumab can be considered if a SRE occurs while on bisphosphonate therapy, whereas the treatment duration should be life-long until impairment in the patient’s general performance status or occurrence of severe adverse events [10‐13]. The expert panels recognize, however, that these recommendations are based on scarce evidence [10‐13].

The purpose of the present study was to investigate the efficacy of antiresorptive treatment in patients with direct initiation of the therapy compared with patients with delayed initiation, the efficacy of switching antiresorptive agents due to symptomatic skeletal event (SSE) or skeletal disease progression, and whether cancer patients with bone metastases benefit from treatment with antiresorptive agents for a period of more than 2 years.

In our study cohort of 255 consecutive patients with metastatic cancer treated with antiresorptive treatment, we investigated the optimal use of antiresorptive agents through three clinical questions. We found that time to antiresorptive treatment initiation did not influence treatment outcome. In addition, a switching strategy from bisphosphonates to denosumab at skeletal disease progression or SRE reduced the risk for new SRE compared with continuing with bisphosphonate treatment. Furthermore, treatment continuation for more than 2 years had a positive impact on the risk for SSE. However, our results should be interpreted with caution considering the study limitations.

The first clinical question we tried to answer was whether time to antiresorptive treatment initiation upon diagnosis of bone metastases influences the treatment effect. The current treatment recommendation is to initiate antiresorptive treatment directly after the diagnosis of bone metastases [10‐13]. This recommendation is, however, not supported by clinical evidence but rather reflects how the pivotal trials on antiresorptive treatment in metastatic setting were designed [9]. In clinical practice, delays in starting antiresorptive treatment are relatively common for instance in patients with poor dental hygiene that are in need for dental care before treatment initiation. Our study is the first that specifically address this clinical question. We found no detrimental effect on the time to SSE or the incidence of SSE in the delayed group (median time 7 months) compared with the direct group. Our results are hypothesis-generating and support further prospective studies investigating a more individualized approach on the time to antiresorptive treatment initiation. The second clinical question was whether switching strategy from bisphosphonates to denosumab in case of skeletal disease progression or SSE is more effective than continuation with bisphosphonates. We found a benefit for switching strategy in terms of time to SSE. However, due to the small sample size, our study was unable to investigate whether a switching strategy from denosumab to bisphosphonates could be beneficial as well in comparison with continuation of denosumab. Switching strategy has been investigated in some randomized phase II studies using surrogate endpoints as changes in bone turnover markers but without clinically relevant outcomes and found indirect evidence that the switching approach might be more effective [16‐18]. Our results support the hypothesis that has been tested by these studies in a real-world setting by using clinically relevant outcomes such as time to SSE. Our results are also in accordance with indirect evidence from pivotal randomized studies about denosumab in which time to subsequent SRE in patients treated with denosumab (that could correspond to the switching strategy to denosumab after SRE in our study) was significantly longer than in patients treatment with bisphosphonates (corresponds to continuation of bisphosphonates after SRE in our study) [19]. Based on the current evidence, the switching strategy has a theoretical and early clinical potential to be the preferred treatment strategy in case of disease progression but further studies with clinically relevant outcomes are necessary.

The third clinical question was the treatment duration of antiresorptive treatment and whether continuing antiresorptive treatment for more than 2 years would add any clinically significant benefit. To minimize the risk for immortal time bias, we used the landmark approach and chose 2 years as landmark time, i.e., we included in the analysis patients that survived for at least 2 years and divided the patients into two groups based on the exposure to antiresorptive treatment at the landmark time or not. Continuing antiresorptive treatment for more than 2 years seemed to be beneficial in terms of time to first SSE after 2 years and time to subsequent SSE. As expected, the risk for drug-specific adverse events was higher in patients with treatment continuation of more than 2 years but we found no significant difference in serious adverse events (severe renal impairment, osteonecrosis of the jaw, hospitalization due to hypocalcaemia) or irreversible adverse events (need for renal replacement therapy due to irreversible renal impairment). Our results are in line with prior studies in which long-term use of antiresorptive treatment was found to be effective with increased, but manageable, toxicity [20‐23]. Combining our results on the clinical benefit of continuing antiresorptive treatment for more than 2 years with recent evidence that enables a reduction in the dosing frequency of zoledronic acid to every 3 months [24‐26], a reasonable treatment approach would be to continue with infusion every 3 months for more than 2 years.

An interesting observation from our study cohort was that a considerably large number of patients received antiresorptive treatment until their death. Current recommendations suggest that antiresorptive treatment should be continued until a substantial decline in the patient’s general performance status is detected and the reason behind this recommendation is the potential analgesic effect of such agents. Considering the lack of high-quality evidence on the analgesic role of antiresorptive treatment [27], the well-documented risk for aggressive non-beneficial cancer treatment at the end-of-life [28, 29], and the risk for adverse events, future studies should investigate the potential role of antiresorptive treatment at the end-of-life care.

There are several limitations in our study that need to be taken into account when interpreting the results. First, the retrospective nature of the study makes the results prone to selection and misclassification bias. Furthermore, the relative small sample size, especially in subgroups, did not enable further analyses that could shed light on other clinically relevant aspects on antiresorptive treatment, e.g., the potential benefit of such treatment for more than 3 years, the identification of subgroups in which antiresorptive treatment should be initiated directly, and the potential effect of switching from denosumab to bisphosphonates after disease progression or SSE. Finally, whether adverse events were related only to antiresorptive treatment or whether there were other factors that led to adverse events could not be retrospectively defined.

The current treatment approach with antiresorptive therapy in patients with bone metastases is based on the “one-size-fits-all” concept. Considering the obvious limitations of our study, our results are hypothesis-generating but support the need for further studies to better optimize the antiresorptive treatment in patients with bone metastases. Based on our results and current literature, antiresorptive treatment is an important supportive cancer therapy that should be offered in patients with bone metastases. Some hypotheses towards a more individualized approach on antiresorptive treatment are supported by our results including the lack of detrimental effect when the treatment is delayed, the potential benefit of switching strategy after skeletal disease progression or SSE, and the benefit of continuing antiresorptive treatment beyond 2 years. Future efforts should be made to a more personalized antiresorptive treatment approach regarding the choice of antiresorptive agent, the time to treatment initiation, the switching strategy, and the duration of treatment.