Included Studies

One published, manufacturer-sponsored, double-blind (DB), randomized controlled trial (RCT) was included in the systematic review. Other Solid Tumours Study 20050244 (Other Solid Tumours Study 244) (n = 1,776)^{4, 5} evaluated the non-inferiority and superiority of denosumab compared with zoledronic acid based on the first occurrence of an SRE in patients with advanced cancer (excluding breast and prostate cancer) and bone metastases. Patients were randomized to receive either denosumab 120 mg administered by subcutaneous (SC) injection every four weeks, or zoledronic acid 4 mg administered by intravenous (IV) injection every four weeks. All patients received concomitant treatment with calcium (≥ 500 mg) and vitamin D (≥ 400 international units [IU]). The primary efficacy outcome was the time to the first occurrence of an SRE, defined as any of the following: pathological fracture, radiation therapy to the bone, surgery to the bone, or spinal cord compression.

One limitation of Other Solid Tumours Study 244 was the fact that a high proportion of patients in both treatment groups (80%) discontinued from the study, although event rates for death (35% versus 36%) and disease progression (14% versus 12%) were similar in both treatment groups. In addition, pain and health-related quality of life (HRQoL) outcomes were likely confounded by unbalanced use of radiation to the bone, an effective treatment for bone pain due to metastasis, which was significantly more common in the zoledronic acid group (▬ P = 0.03), potentially biasing results in favour of zoledronic acid. Similar results were obtained with denosumab and zoledronic acid for these outcomes; therefore, this confounding factor undermines the potential for denosumab to show a between-group difference compared with zoledronic acid regarding pain and HRQoL outcomes. Another limitation of the study was generalizability: approximately 10% of the trial population had multiple myeloma, a condition that is excluded from the Health Canada–approved indication for denosumab. In addition, the trial involved patients with a relatively good performance status at baseline compared with real-life patients, according to the clinical expert consulted by CDR; therefore, the effectiveness and safety observed may not be generalizable to patients with a poorer performance status.

Efficacy

Results from Other Solid Tumours Study 244 demonstrate the non-inferiority (NI) of denosumab compared with zoledronic acid to reduce the risk of a first SRE in patients with advanced cancer and bone metastases. The NI criterion was designed to demonstrate that denosumab would preserve at least 50% of the effect of zoledronic acid compared with placebo. With a hazard ratio (HR) of 0.84 (95% CI, 0.71 to 0.98), results for Other Solid Tumours Study 244 achieved the criteria for NI (P = 0.0007), but failed to demonstrate the superiority of denosumab over zoledronic acid (following adjustment for multiplicity (P = 0.0619). Nevertheless, in a secondary analysis in which patients with multiple myeloma were excluded from the analysis population, denosumab was associated with a significant reduction in the risk of a first SRE compared with zoledronic acid (HR 0.81; 95% CI, 0.68 to 0.96), which in this case achieved the criteria for superiority (P = 0.02).

Pain and HRQoL were identified as important outcomes for patients according to the patient input received by CADTH. These outcomes were measured using reliable and validated tools. However, it is uncertain whether the reduced risk of SREs observed with denosumab translates into improvements in pain or HRQoL, or into reductions in analgesic usage, due to mixed findings and limitations regarding the trial population and confounding factors. Other Solid Tumours Study 244 showed a statistically significant reduction in the risk of a ≥ 2-point increase from baseline on the Brief Pain Inventory (Short Form) (BPI-SF) “Worst Pain” item with denosumab; a finding considered clinically meaningful as a minimal clinically important difference (MCID) of 1.5 to 2.0 was estimated for this instrument. However, results for other patient-related outcomes did not differ between denosumab and zoledronic acid. Patients entering Other Solid Tumours Study 244 had a relatively good performance status, with consequently relatively low pain levels and limited analgesic use at baseline, as well as relatively little impairment in HRQoL compared with real-life patients, based on experience from specialists’ clinical practice. In such circumstances, it may be difficult to detect significant improvements from baseline in these patient-reported outcomes throughout the course of the trial. An absence of worsening in pain or deterioration of HRQoL compared with baseline may also be perceived as a benefit for patients, as suggested by patient input, because the natural disease history in patients with metastatic cancer typically evolves toward disease progression. Finally, pain and HRQoL outcomes were also likely confounded by unbalanced use of concurrent treatments to control pain such as radiation to the bone, which was significantly more common in the zoledronic acid group, potentially biasing results in favour of zoledronic acid. Similar results were obtained with denosumab and zoledronic acid for these outcomes; therefore, this confounding factor undermines the potential for denosumab to show a between-group difference compared with zoledronic acid regarding pain and HRQoL outcomes.

There are no data to inform on the sustainability of beneficial treatment effects observed with denosumab in patients with advanced cancer and bone metastases beyond the median time on study of seven months.

There is a lack of evidence with which to directly compare denosumab to drugs other than zoledronic acid used as treatment to reduce the risk of developing SREs in patients with bone metastases from solid tumours, including pamidronate and clodronate. To inform this evidence gap, CDR reviewed and critically appraised available indirect evidence. A literature search was undertaken by CDR to identify any relevant published indirect comparisons (IDCs). Two relevant publications were included, presenting data from one unique IDC. Ford et al.^{6, 7} assessed the comparative efficacy of denosumab versus zoledronic acid or best supportive care to reduce the risk of developing SREs in patients with bone metastases from solid tumours (except breast and prostate cancer). The network meta-analysis (NMA) results suggest that the use of denosumab was associated with a statistically significant reduction in the risk of SREs compared with zoledronic acid and placebo. This finding is consistent with the secondary analysis of the Other Solid Tumours Study 244 results after the exclusion of patients with multiple myeloma, but are not consistent with the primary analysis, in which denosumab was non-inferior to zoledronic acid. No data were available to compare denosumab versus pamidronate or clodronate.

Harms

Denosumab is approved for six indications and has been in use since 2010, and the overall harms in Other Solid Tumours Study 244 results did not raise any new safety concerns, as confirmed by the clinical expert consulted by CDR. However, there are no data to inform the long-term maintenance of safety of denosumab in patients with advanced cancer and bone metastases beyond the median time on study of seven months.

Mortality as well as the overall incidence of serious adverse events (SAEs) during Other Solid Tumours Study 244 did not differ significantly between denosumab and zoledronic acid, and were not higher than would be expected in this patient population; the clinical expert consulted by CDR highlighted the high burden of the malignancy. The most commonly reported SAEs for both treatments (< 12%) included neoplasm progression, dyspnea, pneumonia, respiratory failure, metastases to the central nervous system (CNS), dehydration, general deterioration, spinal cord compression, pyrexia, anemia, pleural effusion, febrile neutropenia, and vomiting. The proportion of patients experiencing adverse events (AEs) was high but similar between denosumab and zoledronic acid. The most common AEs included nausea, anemia, dyspnea, fatigue, constipation, vomiting, back pain, cough, and asthenia. Proportions of patients discontinuing due to AEs in the denosumab treatment group were, however, lower (10%), suggesting adequate tolerability.

Some AEs of particular interest were identified by CADTH based on the denosumab mechanism of action and Health Canada warnings, which have been issued with regard to the risks of hypocalcemia, infections, dermatologic AEs, osteonecrosis of the jaw, atypical femur fractures, and malignancies.³ Results for osteonecrosis of the jaw (ONJ), atypical femoral fractures, SAEs of infection, and dermatologic AEs are characterized by low and similar proportions of patients experiencing the event in both treatment groups. Cardiovascular events were relatively frequent but occurred in similar proportions of patients in both treatment groups. There were numerically more cases of hypocalcemia in the denosumab group compared with zoledronic acid; however, the difference did not seem to be clinically meaningful, according to the clinical expert consulted by CDR.

Experience from specialists’ clinical practice and patient input received by CADTH suggest the need for pharmacological drugs with added convenience and tolerability for use in patients with advanced cancer and bone metastases. The fact that denosumab is administered subcutaneously, compared with zoledronic acid that needs to be administered intravenously, may provide benefits for patients in terms of accessibility and convenience, and may also contribute to reducing the burden on the health care system by eliminating the need for a visit to a facility for administration. However, the double-dummy design prevented the objective assessment of whether the SC administration of denosumab is a significant benefit to patients compared with IV administration.

No data were available to directly compare the potential harms of denosumab versus other drugs used in patients with bone metastases from solid tumours. Potential harms were not analyzed in the IDC that was identified by CDR to compare the safety of denosumab with comparators other than zoledronic acid.