The Appraisal Committee reviewed the data available for the clinical and cost effectiveness of denosumab, having considered evidence on the nature of osteoporotic fractures and the value placed on the benefits of denosumab by postmenopausal women with the condition, those who represent them, and clinical specialists. It also took into account the effective use of NHS resources.
The Committee discussed the clinical need of postmenopausal women for the prevention of osteoporotic fragility fractures. It heard from the clinical specialists that the main aim of primary prevention is the opportunistic identification of postmenopausal women who are at risk of osteoporotic fragility fractures, and the aim of secondary prevention is to provide the most effective treatment for women who have already had an osteoporotic fragility fracture and are at risk of further fractures.
The Committee heard from the clinical specialists that current UK clinical practice is to start treatment with oral bisphosphonates (first with alendronate and then either risedronate or etidronate if alendronate is unsuitable). These treatments are not suitable for all women because some women are unable to comply with the special instructions for the administration of oral bisphosphonates, or have a contraindication to or are intolerant of oral bisphosphonates. The Committee also heard from the clinical specialists that women for whom oral bisphosphonates are unsuitable receive either no treatment or strontium ranelate for primary prevention (as set out in NICE's technology appraisal guidance on raloxifene), or no treatment, strontium ranelate or raloxifene for secondary prevention (as set out in NICE's technology appraisal guidance on raloxifene and teriparatide). The clinical specialists stated that the management of osteoporosis usually takes place in primary care (both strontium ranelate and raloxifene are given in primary care). Women who have severe osteoporosis may receive more potent agents such as zoledronate or intravenous ibandronate but there is limited capacity for treatment in secondary care because of the need for day-case facilities for these intravenous treatments. The patient experts stated that some women who cannot take or cannot tolerate oral bisphosphonates have a preference for strontium ranelate or raloxifene as they do not like the intravenous infusion used for zoledronate treatment, whereas others prefer the convenience of a 12-monthly intravenous infusion (zoledronate) over taking oral treatments daily (strontium ranelate and raloxifene). The Committee accepted that the great majority of treatment for the primary and secondary prevention of osteoporotic fragility fractures is provided in primary care. It also accepted that women for whom oral therapies are unsuitable or who have severe osteoporosis may receive more potent agents such as zoledronate or intravenous ibandronate in secondary care and that teriparatide is also used for secondary prevention when women are unable to take other therapies. The Committee concluded that the relevant comparators for primary prevention are no treatment and strontium ranelate, and for secondary prevention are no treatment, strontium ranelate and raloxifene, because both the administration and supervision of strontium ranelate and raloxifene are organised in primary care. The Committee also concluded that potential comparators for denosumab are zoledronate (for severe osteoporosis) and teriparatide (for women who have sustained a clinically apparent osteoporotic fracture and who are defined by age, T score and number of osteoporotic fractures and who are unable to take all oral bisphosphonates, strontium and raloxifene, as defined in NICE's technology appraisal guidance on raloxifene and teriparatide).
The Committee was aware that the licensed indication for denosumab is for the treatment of osteoporosis in postmenopausal women at increased risk of fractures. The Committee noted that the manufacturer's decision problem focused on postmenopausal women diagnosed with osteoporosis for whom oral bisphosphonates are unsuitable. The manufacturer stated that denosumab was not expected to compete with oral bisphosphonates in clinical practice, given the wide availability of generic oral bisphosphonates in the UK and the need for efficient use of NHS resources. The Committee also noted that the manufacturer did provide an analysis of denosumab compared with oral bisphosphonates for completeness. It accepted that it was reasonable to base its considerations on women for whom oral bisphosphonates are unsuitable and the subsequent discussion focused on this population only.
The Committee heard from the clinical specialists that denosumab is a monoclonal antibody that reduces osteoclast activity and hence reduces bone breakdown, that it is the first drug of its class, and that its biological mechanism of action results in targeted therapy with fewer adverse events than other treatments. The Committee considered that a treatment administered by subcutaneous injection once every 6 months, such as denosumab, offers an advantage because it may improve adherence with therapy, particularly for women who have problems swallowing or standing to take oral bisphosphonates. The Committee also accepted evidence from the patient experts that many women stop taking oral bisphosphonates because of adverse events, and often do not go back to their GP. Therefore a 6-monthly subcutaneous injection of denosumab could provide women with a pre-arranged opportunity to discuss their treatment and any adverse events with a healthcare professional, and this support may improve compliance and persistence with treatment.
The Committee accepted that the clinical-effectiveness evidence presented in the manufacturer's submission was derived from a large, high-quality trial of adequate duration (FREEDOM) that studied treatment with denosumab compared with placebo. The Committee noted that, because the FREEDOM study did not provide a head-to-head comparison of denosumab against all relevant comparators, the manufacturer carried out a random-effects meta-analysis to obtain direct estimates for each treatment compared with placebo (denosumab, strontium ranelate, raloxifene, teriparatide and zoledronate).
The Committee noted that the FREEDOM trial showed a statistically significant 68% reduction in the relative risk p<0.001) of the 36-month incidence of new radiographically diagnosed vertebral fractures. The Committee also noted that denosumab significantly reduced the risk of non-vertebral fracture (HR 0.80; 95% CI 0.67 to 0.95; relative reduction of 20%) and of hip fracture (HR 0.60, 95% CI 0.37 to 0.97; relative reduction of 40%). The Committee concluded that the evidence from the FREEDOM trial demonstrated that denosumab was effective in reducing the risk of fracture in postmenopausal women compared with placebo.
The Committee discussed the meta-analysis that was undertaken by the manufacturer to obtain direct estimates for each treatment compared with placebo. However, the Committee noted the lack of a direct comparison of denosumab with active comparators. It was, therefore, unable to make a conclusion about the relative clinical effectiveness of denosumab, but was satisfied with the evidence on the direct estimates for each treatment compared with placebo and concluded that the methods used in the meta-analysis were sufficiently robust for use in the economic analysis.
The Committee discussed the adverse events experienced by postmenopausal women receiving denosumab for the prevention of osteoporotic fragility fractures. The Committee noted that in the FREEDOM trial treatment with denosumab was associated with fewer adverse events than placebo. The patient experts stated that the main concern women have about treatment for osteoporosis is the duration for which therapies are taken and whether they will experience adverse events over a long period of time. The Committee heard that there may be a risk of infection associated with denosumab treatment; however, it accepted evidence from the clinical specialists that this risk was low and that if substantial evidence became available, it may be necessary to assess women with severe infections before considering the use of denosumab. The Committee noted that studies of denosumab for other indications have shown that treatment may be associated with osteonecrosis of the jaw, but it was satisfied with the clinical specialists' statement that there was no evidence of this from the clinical studies of denosumab in women with osteoporosis. The clinical specialists confirmed that 14,000 women have received denosumab and that it was well tolerated. The Committee concluded that the available clinical evidence on the adverse effects associated with denosumab indicated that it was a well-tolerated treatment for the prevention of osteoporotic fragility fractures in postmenopausal women.
The Committee considered the manufacturer's economic model, and the critique and exploratory sensitivity analyses performed by the ERG. It noted that the manufacturer used a Markov economic model to evaluate the cost effectiveness of denosumab compared with a range of comparators, and that the clinical data were derived from the manufacturer's direct comparison of each comparator with placebo in their random-effects meta-analysis. The Committee considered that the methods used in the analysis were robust.
The Committee considered the ERG's concerns about a number of aspects of the economic model, such as the long-term effects of fractures on mortality, the setting where denosumab is likely to be given, and the associated administration and monitoring costs modelled. The Committee also discussed issues raised from consultation on the appraisal consultation document: the use of quality-of-life data from the FREEDOM study in the economic model, the use of the FRAX tool to estimate fracture risk, the cost and setting of care for women receiving denosumab, the cost of zoledronate, the relative risk for wrist fracture for zoledronate, and the relative risk for hip fracture with strontium ranelate.
The Committee considered the long-term effects of fractures on mortality and the need for nursing home care. It noted that the manufacturer assumed that 30% of observed mortality for all fracture types is causally related to osteoporotic fracture, and this estimate was varied to 100% in sensitivity analyses. The Committee was aware that the manufacturer's model also assumed that women were at increased risk of entry into nursing home care following hip fracture and that the lowest cost of private residential care was applied. The Committee noted that some women may already be in a nursing home, and some may also be self-funding their nursing home care. After consultation, the manufacturer carried out additional sensitivity analyses, using the assumption that nursing home admission was zero. The Committee concluded that the long-term effects of fractures on mortality and nursing home care had only a minor impact on the cost-effectiveness estimates for denosumab.
The Committee noted that the manufacturer's model assumed treatment duration of 5 years, and that the FREEDOM study was of 3 years' duration. The Committee accepted that the 5-year treatment duration assumption was appropriate and reflected clinical practice. The Committee noted that the manufacturer's model assumed that women would return in a linear fashion to baseline fracture risk levels over 1 year after treatment stops. It considered whether the return to baseline fracture risk should be different according to treatment type, and heard from the ERG that the duration of benefit in terms of fracture risk (as opposed to bone mineral density) is unknown after cessation of osteoporosis treatments. The Committee concluded that there was little evidence on the duration of effect on fracture risk for osteoporosis treatments and that this was an area of uncertainty.
The Committee discussed health-related quality-of-life benefits and utility values in the economic model. The Committee noted that no significant differences were seen in health-related quality of life between the denosumab and placebo arms of the FREEDOM trial. The manufacturer justified the omission from the economic model of EQ-5D data from FREEDOM on the grounds that the number of fracture events with associated EQ-5D scores was low and there was relative infrequency of health-related quality of life measurement. The ERG accepted that the lack of statistical difference in EQ-5D scores between the denosumab and placebo arms of the trial was explained by the above factors (low number of fracture events with associated EQ-5D scores and relative infrequency of health-related quality of life measurement), rather than an adverse effect of denosumab masking the health benefit of fracture prevention. The ERG stated that omitting the FREEDOM trial EQ-5D data from the economic analysis was justified given the quality and depth of the manufacturer's systematic review of health-related quality-of-life data. The Committee concluded that the manufacturer's approach to modelling health-related quality of life was acceptable.
The Committee considered that the manufacturer's model allowed sensitivity analyses to be carried out using the FRAX tool. The Committee accepted that FRAX is a potentially useful tool in clinical practice, but it was mindful that the tool is presently unvalidated. The Committee was not persuaded that recommendations about treatment should be based on absolute risk as calculated using FRAX and that the stepped approach of assessing fracture risk is needed to ensure the effective allocation of NHS resources. This is because absolute fracture risk is the total risk for all fracture sites, but different fracture sites have different impacts on quality of life, costs and mortality. Therefore, cost effectiveness is dependent on the contribution from each fracture site to the total fracture risk. The Committee concluded that using a combination of T-score, age and a number of independent clinical risk factors for fracture remained more appropriate for defining treatment recommendations in this appraisal.
The Committee considered the key drivers of cost effectiveness and noted that alterations to most key parameters in the manufacturer's sensitivity analyses had limited impact on comparisons of denosumab with raloxifene, strontium ranelate and no treatment. Comparisons with ibandronate, zoledronate and teriparatide were most sensitive to changes in the assumptions about the cost of administering denosumab. When the manufacturer increased the cost of administering denosumab (by assuming that 1 administration per year would be delivered in secondary care), this increased the ICER for denosumab compared with no treatment from £29,200 to £36,200 per QALY gained for primary prevention, and from £12,400 to £15,700 per QALY gained for secondary prevention. The Committee noted that given the similar cost and efficacy of denosumab and zoledronate, changes to this assumption also resulted in zoledronate dominating denosumab (that is, zoledronate was less costly and more effective than denosumab) for both primary and secondary prevention. The Committee concluded that with the exception of administration costs for denosumab, alterations to most key parameters had limited impact on comparisons between denosumab and the primary and secondary comparators.
The Committee noted the ERG's view that administration of denosumab may not be provided in primary care. However, the clinical specialists stated that there is no reason why denosumab should only be used in secondary care. The clinical specialists highlighted that because denosumab is a new biological agent they expected that, initially, treatment would be started in secondary care, but with follow-up almost exclusively in primary care (except for women with severe osteoporosis, who may be followed up in secondary care in line with current UK clinical practice). The Committee discussed whether administering denosumab would be part of general medical services or whether it would be regarded as an enhanced service for which an additional payment would be negotiated, and it noted the comments received during consultation on the ACD. The clinical specialists stated that women would not need to go through a screening process before starting treatment with denosumab, and that women receiving denosumab are not likely to be at high risk of side effects and so follow-up in secondary care would not be necessary. The clinical specialists also stated that although denosumab is a biological agent and also has effects on the immune system, it is specifically targeted for regulating bone cells. The clinical specialists, therefore, thought that the potential safety concerns associated with other biological agents (such as those targeting tumour necrosis factor) may not be applicable to denosumab. The clinical specialists stated that because treatment with denosumab would not involve substantial additional activities to standard practice in managing osteoporosis, it would probably be provided as part of general medical services. The Committee accepted the views of the clinical specialists that there were no specific safety concerns around the use of denosumab and that follow-up in secondary care would not be necessary. Therefore, it was not persuaded to alter its opinion that denosumab is likely to be provided as part of general medical services in primary care. The Committee concluded that while treatment with denosumab may be started in secondary care, it would be subsequently delivered almost exclusively in primary care. The relatively small proportion of women with severe osteoporosis would continue to be followed-up in secondary care, in line with current UK clinical practice.
After consultation on the appraisal consultation document, the Committee discussed comments that outlined a change in the cost of zoledronate and that an alternative relative risk could be used for the effect of zoledronate on wrist fracture. The ERG was requested to carry out exploratory analyses that showed that denosumab was less effective and less costly than zoledronate. The Committee had already concluded that although treatment with denosumab may be started in secondary care, it will be subsequently delivered almost exclusively in primary care, unlike the administration of zoledronate, use of which will remain in secondary care. As the Committee regarded the main comparators for denosumab to be those treatments delivered in primary care when oral bisphosphonates were unsuitable (no treatment, strontium ranelate, raloxifene), it did not regard these issues to be central to the decision problem.
The Committee, therefore, accepted that the most likely cost-effectiveness estimates would lie between the manufacturer's base case (using primary care assumptions) and the ERG's additional analyses (using secondary care assumptions), and that the most plausible ICERs were likely to be closer to the manufacturer's base case estimates given that care would mostly be in the primary setting.
The Committee considered the relative risks for hip fracture that were used in the manufacturer's meta-analysis. The Committee was aware that NICE recommendations on strontium ranelate are due to be reconsidered following a court of appeal judgement related to assumptions about relative risk of hip fracture. It noted that the figure that the manufacturer used as the 5-year relative risk of hip fracture for strontium ranelate was 0.89 compared with placebo and this was derived from the published TROPOS study, but that alternative relative risk figures of 0.64 (obtained over 3 years) or 0.57 (obtained over 5 years) for the effect of strontium ranelate on hip fracture were suggested during consultation. These alternative, and lower relative risk values, were based on a post-hoc subgroup analysis in the TROPOS study of women over the age of 74 with a T score of -2.4. The Committee also noted a suggestion by the manufacturer of denosumab that if a subgroup relative risk figure for hip fracture was to be used for 1 treatment in comparative analysis, then similar subgroup figures should also be used for denosumab for this and other outcomes.
The Committee noted that when using a relative risk of hip fracture of 0.89 for strontium ranelate in the manufacturer's base case, denosumab was dominant compared with strontium ranelate for both primary and secondary prevention. The Committee heard from the ERG that exploratory analyses applying the relative risk estimate of 0.64 over the modelled 5-year treatment period in the manufacturer's model resulted in a base-case ICER of £10,200 per QALY gained for denosumab compared with strontium ranelate for primary prevention and an ICER of £5,100 per QALY gained for secondary prevention. When the relative risk estimate of 0.57 for strontium ranelate was applied over the modelled 5-year treatment period in the manufacturer's model, this resulted in an ICER of £16,300 per QALY gained for denosumab compared with strontium ranelate for primary prevention and an ICER of £8,600 per QALY gained for denosumab compared with strontium ranelate for secondary prevention. The Committee considered these exploratory analyses and concluded that it did not need to make a decision on which relative risk for strontium ranelate was the most appropriate one to apply, because for any of the suggested relative risk values for hip fracture for strontium ranelate, the ICERs for denosumab compared with strontium ranelate fell within a range that was still considered to be a cost-effective use of NHS resources.
The Committee noted that for the primary prevention of osteoporotic fragility fractures, denosumab dominated strontium ranelate in the manufacturer's base case, and the ICER was £15,900 per QALY gained for denosumab compared with strontium ranelate in the ERG's additional analyses (assuming denosumab is always given in secondary care). The Committee noted that in NICE's technology appraisal guidance on raloxifene, strontium ranelate is recommended for postmenopausal women:
who are unable to comply with the special instructions for the administration of alendronate and either risedronate or etidronate, or have a contraindication to or are intolerant of alendronate and either risedronate or etidronate and
who also have a combination of T-score, age and number of independent clinical risk factors for fracture as indicated in the following table.
| Age | 0 independent clinical risk factors for fracture | 1 independent clinical risk factor for fracture | 2 independent clinical risk factors for fracture |
|---|---|---|---|
| 65 to 69 years |
Treatment with strontium ranelate is not recommended |
-4.5 |
-4.0 |
| 70 to 74 years |
-4.5 |
-4.0 |
-3.5 |
| 75 or older |
-4.0 |
-4.0 |
-3.0 |
For the purposes of NICE's technology appraisal guidance on raloxifene, independent clinical risk factors for fracture are parental history of hip fracture, alcohol intake of 4 or more units per day, and rheumatoid arthritis.
The Committee accepted that the above were appropriate when making recommendations on treatment with denosumab for the prevention of primary and secondary osteoporotic fractures. The Committee was mindful that the clinical effectiveness evidence was based on a meta-analysis of each treatment compared with placebo rather than a direct comparison of denosumab with active comparators, and of the areas of uncertainty it had discussed in the economic modelling. However it was satisfied with the robustness of the clinical evidence and, when taken together with the low ICERs presented, the Committee concluded that, as an option, denosumab was a cost-effective use of NHS resources for the primary prevention of osteoporotic fragility fractures only for postmenopausal women at increased risk of fractures for whom oral bisphosphonates are unsuitable, and who have the same level of fracture risk as described in the recommendations for strontium ranelate in NICE's technology appraisal guidance on raloxifene.
The Committee was aware that for women for whom oral bisphosphonates are unsuitable and who do not currently fulfil the criteria for treatment with strontium ranelate, there is no treatment currently recommended by NICE for the primary prevention of osteoporotic fragility fractures. The Committee noted that the ICER for denosumab compared with no treatment was £29,200 per QALY gained in the manufacturer's base-case analysis, and this increased to £40,600 per QALY gained in the ERG's additional analyses. It concluded that the ICER for the base-case population (women 70 years and over with a T-score of -2.5 SD or below) for denosumab compared with no treatment was likely to lie within this range. The Committee explored the results for subgroups of women at higher risk by age and T-score, for whom oral bisphosphonates are unsuitable and in circumstances in which none of the treatments appraised by NICE are recommended. The ICERs for denosumab compared with no treatment from the manufacturers model varied between £19,300 and £71,300 per QALY gained. The Committee agreed that the most plausible ICERs were likely to be higher, based on the ERG's amended assumptions. Taking into account the uncertainties around the long-term effects of fractures on mortality and nursing home care, and the setting where denosumab is likely to be given, the Committee concluded that denosumab was not a cost-effective use of NHS resources for women for whom oral bisphosphonates are unsuitable and the treatments appraised by NICE are not recommended.
For the secondary prevention of osteoporotic fragility fractures, the Committee noted that the ICER for denosumab compared with no treatment in women for whom oral bisphosphonates are unsuitable was £12,400 per QALY gained in the manufacturer's base-case analysis, which increased to £17,900 per QALY gained in the ERG's additional analyses. Denosumab dominated raloxifene or had an ICER of £2,000 per QALY gained in the manufacturer's base-case analysis, which increased to £12,200 per QALY gained in the ERG's additional analyses. The cost-effectiveness results for denosumab compared with strontium ranelate ranged from strontium ranelate being dominated by denosumab in the manufacturer's base-case analysis to an ICER of £6,600 per QALY gained in the ERG's exploratory analyses. The Committee also noted the results of the subgroup analysis by age and T-score for women for whom oral bisphosphonates are unsuitable and in circumstances in which the treatments appraised by NICE are not recommended, in which the ICER for denosumab compared with no treatment varied between £12,289 and £22,957 per QALY gained. The Committee noted that teriparatide is recommended in NICE's technology appraisal guidance on raloxifene and teriparatide as a treatment option for the secondary prevention of osteoporotic fragility fractures in women with severe osteoporosis for whom oral bisphosphonates and strontium ranelate are unsuitable. The Committee noted that in the manufacturer's base-case analysis, denosumab was slightly less effective and much less costly than teriparatide.
The Committee concluded that treatment with denosumab was a cost-effective use of NHS resources and may be an option for the secondary prevention of osteoporotic fragility fractures in women for whom oral bisphosphonates are unsuitable (that is, in women who are unable to comply with the special instructions for administering alendronate and either risedronate or etidronate, or have an intolerance of, or a contraindication to, those treatments). The Committee also concluded that denosumab may provide an alternative treatment option that would be a cost-effective use of NHS resources for women who are eligible for treatment with teriparatide as defined in NICE's technology appraisal guidance on raloxifene and teriparatide.
The Committee considered whether its recommendations were associated with any potential issues related to equality. Following consultation on the appraisal consultation document, the Committee was aware that 1 area of potential discrimination was the primary prevention of osteoporotic fractures in postmenopausal women with swallowing problems as a result of disabling stroke disease, who would otherwise be eligible for treatment with oral bisphosphonates, but do not fulfil the criteria for denosumab or other treatments. The Committee concluded that its considerations were based on the population of postmenopausal women for whom oral bisphosphonates are unsuitable which included women with and without a disability.