Real-world use of denosumab and bisphosphonates in patients with solid tumours and bone metastases in Germany

Bone metastases are common in patients with advanced solid tumours [1‐3]; they affect 68% of patients with prostate cancer (> 90% of patients with metastatic castration-resistant disease), 73% with breast cancer and 36% with lung cancer [2, 3]. Bone metastases can lead to debilitating bone complications, known as skeletal-related events (SREs), which include pathologic fractures, spinal cord compression and the need for radiation or surgery to the bone [4]. Bone complications can cause pain, reduce patient quality of life (QoL) and be associated with an increased risk of death [5‐8]. They are also linked to high healthcare resource use, thereby placing a considerable burden on healthcare systems [9]. Early treatment to prevent bone complications is, therefore, paramount; results from real-world studies and randomised controlled trials (RCTs) in breast cancer suggest that, on average, a first bone complication occurs as early as 8–9 months after a diagnosis of bone metastases [10, 11].

As the cancer treatment landscape evolves and survival outcomes improve, the long-term implications of supportive care become increasingly relevant to clinical decision-makers. Growing awareness of the prevalence and impact of bone complications has led to greater emphasis on their prevention within clinical cancer guidelines [5, 12, 13]. Bisphosphonates (e.g. zoledronic acid, pamidronate disodium, ibandronate and clodronate) and denosumab are approved for prevention of bone complications in patients with advanced malignancies involving bone [14‐18]. Bisphosphonates are synthetic analogues of pyrophosphonate (a natural regulator of bone metabolism) that are incorporated into the bone, and reduce bone resorption by inhibiting the differentiation and activation of osteoclasts [14, 16‐20]. They are administered as an intravenous (IV) infusion every 3–4 weeks (zoledronic acid, pamidronate disodium, ibandronate) or orally every day (ibandronate and clodronate) [14, 16‐19]. Denosumab is a fully human, monoclonal antibody against the receptor activator of nuclear factor kappa B (RANK) ligand (RANKL). By disrupting RANK signalling, denosumab prevents the fusion and activation of osteoclasts, thus reducing bone resorption [15, 21]. Denosumab is administered as a subcutaneous (SC) injection every 4 weeks [15]. Bisphosphonates and denosumab have different pharmacokinetic profiles owing to their differing mechanisms of action; zoledronic acid has a half-life of 2–189 days and may remain in bone for up to 10 years, whereas denosumab is not incorporated into the bone and has a mean half-life of 14–55 days [14, 15, 22]. Bisphosphonates and denosumab reduce the incidence of bone complications in patients with bone metastases from solid tumours [23‐25], with denosumab showing superiority compared with zoledronic acid in patients with breast [26] and prostate cancer [27] and non-inferiority in patients with multiple myeloma and other solid tumours [28]. These agents have also been shown to prevent pain progression and the worsening of patient QoL [5], with denosumab shown to be more effective than zoledronic acid in these dimensions [29].

The European Society for Medical Oncology recommends that a bisphosphonate or denosumab is initiated at the time of bone metastasis diagnosis to reduce pain and delay bone complications [5]. Clinical guidelines draw heavily on RCTs to provide robust evidence on the safety and efficacy of therapies. RCTs offer high internal validity, but their findings are based on the use of agents in closely managed, highly selected patients often treated over the short term [30]. In contrast, routine clinical practice typically involves diverse patient populations and care settings, and long-term treatment periods. Thus, the use and effectiveness of therapies must also be evaluated in real-world practice. Importantly, the full benefit of medications will only be achieved if patients follow prescribed regimens [31]; understanding treatment persistence and compliance is critical to interpretation of real-world outcomes. Persistence can be defined as the duration of time from initiation to discontinuation of therapy, whereas compliance can be defined as the extent to which a patient acts in accordance with the prescribed interval and dosing of a regimen [32]. Two studies using data from the Oncology Services Comprehensive Electronic Records (OSCER) in the United States of America (USA) have shown that in patients with bone metastases from solid tumours, compliance and persistence with denosumab are higher than with IV bisphosphonates [33, 34]. To build on these findings from a European perspective, this retrospective study was conducted to provide real-world data on bisphosphonate and denosumab usage (persistence, compliance, switching and drug re-initiation rates) in a large sample of patients with bone metastases, from breast, prostate or lung cancer, receiving current anti-tumour treatment in routine clinical practice in Germany. Exploratory analysis of potential implications of time to initiation of denosumab or a bisphosphonate on prevention of bone complications was also conducted.

This study updates the limited evidence available on bisphosphonate and denosumab usage patterns in the current anti-tumour treatment landscape, by providing real-world data from patients with bone metastases from breast, prostate and lung cancer diagnosed between 2011 and 2015 in Germany. Across all tumour types and time points, persistence with denosumab was higher than with any bisphosphonate. Persistence and compliance with denosumab were lower for patients with prostate cancer than for those with breast cancer. This may reflect the male gender and older age of these patients, factors that have previously been shown to influence compliance [33]. In general, time to discontinuation was longer, compliance greater and switch rates lower in patients receiving denosumab than in those receiving bisphosphonates. The higher levels of persistence and compliance with denosumab than with bisphosphonates may be associated with improved outcomes in routine clinical practice [34]. Median time to discontinuation with denosumab was not reached and compliance data not reported in the lung cancer group owing to a small patient sample. These data may suggest a reluctance to prescribe supportive care treatments to patients who have a poor prognosis. Although there were no class-specific patterns for drug re-initiation, our data show that treatments were continued even after large gaps. The exploratory analyses suggest that prompt initiation of bone-protecting therapy after bone metastasis diagnosis may delay time to first and subsequent bone complications.

The denosumab and bisphosphonate usage patterns in our study are similar to those found in the two OSCER studies conducted in the USA in patients with bone metastases from breast, prostate and lung cancer [33, 34]. In a 1-year study of 3569 patients, those prescribed denosumab received a higher median number of annual doses and were less likely to switch to an alternative therapy than those initially prescribed a bisphosphonate [33]. These results were confirmed in a 3-year study in which patients receiving denosumab were less likely to switch agents and more likely to demonstrate compliance with treatment at 1, 2 and 3 years across all solid tumour subtypes. Median time to non-persistence was also significantly longer for denosumab than for zoledronic acid [34].

Although the determinants of persistence and compliance can only be inferred, there are several potential reasons why persistence and compliance are greater with denosumab in this study. First, denosumab has demonstrated superiority to zoledronic acid in preventing bone complications in patients with solid tumours; therefore, physicians may be more motivated to ensure compliance [23]. Second, in an integrated analysis of three phase 3 clinical trials in patients with bone metastases from solid tumours, denosumab was more effective than zoledronic acid at extending the time to significant increases in pain and use of strong opioids [29]. Third, discrete-choice studies have suggested that patients and physicians prefer the 4-weekly SC injection of denosumab to the 3–4-weekly IV infusion of bisphosphonates [35, 36]. Finally, side effects, particularly renal toxicity and acute-phase reactions, are important factors in determining physician and patient treatment preferences [35, 36]. The use of IV bisphosphonates requires routine renal monitoring and dose adjustments, complicating the care of these patients [14, 37]. In contrast, denosumab is not excreted by the kidneys and does not require renal monitoring, which may make optimum compliance and persistence easier to achieve [15]. Given that this study comprised an elderly patient population (mean age ≥ 65 years), poorer persistence and compliance recorded for patients receiving IV bisphosphonates may reflect physicians’ decisions to skip or delay doses because of renal impairment [35, 36]. Of note, 3-monthly zoledronic acid infusions have been shown to be non-inferior to monthly infusions in clinical trials of patients with multiple myeloma and breast or prostate cancer with bone metastasis [38, 39]. It is possible that less frequent dosing may affect persistence and compliance with zoledronic acid; however, this has not been formally assessed in clinical studies, and this dosing regimen is currently off-label.

Exploratory analysis on the time to denosumab or bisphosphonate initiation showed large differences among tumour types, with a much longer median time to initiation recorded for patients with lung cancer than for those with breast or prostate cancer. The shortest time to treatment initiation was seen in those with breast cancer (median time 4 months). Similarly, a real-world study assessing the treatment of bone metastases in patients with breast cancer across six European countries, including Germany, showed that 81% of patients treated with denosumab or bisphosphonates received therapy in the 3 months after bone metastasis diagnosis [40]. The long median time to treatment initiation in patients with lung cancer (23 months) may reflect poor prognosis in this patient group; studies of other cancer types have shown that physicians cite poor prognosis as a reason for not initiating treatment to prevent bone complications [40]. Our data suggest that improvements in bone healthcare are needed to ensure that all patients receive treatment at the time of bone metastasis diagnosis to prevent bone complications, in line with the current guidelines [5].

The main strength of this study was that it incorporated a large patient sample representative of the wider German population and, unlike the OSCER studies, was not limited to patients treated in the oncology clinic setting only, so improving the generalisability of the results. The study has some limitations. The study was descriptive and not designed to demonstrate significant differences in outcomes between different tumour types, drug classes or time to initiation subgroups; therefore, these results should be interpreted with some caution. As with all claims studies, data are based on prescription claims records and may not correspond to actual drug consumption. Additionally, determinants of differential usage patterns between agents can only be inferred. Furthermore, usage definitions must be meaningful; the 90-day gap definition for persistence and the 12 prescription definition for compliance were considered appropriate for this study and are in line with previously published studies [33, 34]. Sensitivity analyses using alternative gaps (45, 60 and 120 days, and 10, 11 and 13 prescriptions per year) broadly confirmed the result of the primary analysis. The time between bone metastasis diagnosis and treatment initiation may be underestimated; the occurrence of a bone complication, or initiation of denosumab or bisphosphonate therapy, may trigger the (retrospective) recording of bone metastases, resulting in a delay between the documentation of the diagnosis in the patient’s medical chart and its appearance in their administrative claims record. Finally, there are inherent difficulties in defining bone complications using claims data, and the exploratory analyses on the time to first bone complication should be interpreted with caution.

In this real-world study of German clinical practice, patients initiated on denosumab following a bone metastasis diagnosis from breast, prostate or lung cancer had higher medication persistence, longer time to discontinuation, improved compliance and lower switch rates compared with those initiated on a bisphosphonate. Median time to first and subsequent bone complications was shorter for late vs early initiators of either agent, thereby supporting clinical recommendations to introduce bone-protecting therapy at the time of bone metastasis diagnosis to provide optimal patient care and maximise patient QoL.