Burosumab for treating X-linked hypophosphataemia in children and young people

The symptoms of X‑linked hypophosphataemia (XLH) usually start in early childhood. The clinical experts highlighted that XLH is heterogeneous, meaning that people with the condition can have a variety of symptoms of different severities, and that symptoms may start later in life. The clinical experts also noted that XLH leads to bone-related and non-bone-related manifestations. The symptoms of XLH in children can include bone defects, pain, functional impairment, fatigue, dental problems and, in some people, hearing loss. The committee heard that children with XLH are often unable to participate in educational and social activities, which has a substantial emotional effect on the child and their family. The clinical experts explained that many manifestations of XLH mean children with the condition are vulnerable to bullying, which further exacerbates its emotional and psychological effects. The committee recognised that bone defects will become permanent if the condition is untreated while bones are growing. This can result in skeletal abnormalities continuing into adulthood, and lifelong disability and pain. The patient experts noted from their experiences of XLH that painful symptoms result from established bone deformities and associated surgeries. The clinical experts explained that there can also be progressive bone disease in adults, including osteomalacia and accompanying stress fractures, pain and disability. The committee acknowledged that some manifestations of XLH will continue after the skeleton matures. The clinical experts noted that the metabolic manifestations of XLH that occur in children, such as fatigue and muscle weakness, also occur in adults. The committee acknowledged comments received during consultation that dental abscesses cause significant detrimental effects on the quality of life of adults and children. The committee agreed that, in the absence of evidence, it was difficult to determine the relative importance of pre-existing skeletal deformities, and ongoing bone and metabolic manifestations in adults. It recognised that XLH affects people from a young age and continues to affect them into adulthood.

The patient experts noted that conventional therapy (vitamin D and phosphate) must be taken 4 to 6 times a day, which interferes with usual activities (school or work). This treatment can also disturb sleep because of night-time doses (although these are not always used). The clinical experts agreed that impractical dosing regimens, and an unpleasant taste and side effects, contribute to poor adherence to this therapy (see section 4.16), and the committee recognised that this was an issue. The patient experts explained that, even when children have treatment with vitamin D and phosphate from a young age, severe skeletal deformities can still occur. Surgical intervention is often needed to correct skeletal deformities. The committee heard from the patient experts that surgery is distressing and can be disruptive to education or work. One of the patient experts also explained that they have had continued bone and joint pain after having corrective surgery (see section 4.1). The committee concluded that there is an important unmet need for an effective and practical treatment for XLH.

The patient and clinical experts explained that the earlier treatment is started, the better the outcomes. The committee recalled that skeletal deformities would continue into later life if untreated in childhood (see section 4.1). The clinical experts explained that they aim to start treatment as early as possible, and delays in starting treatment can increase the need for corrective surgery. The committee accepted that starting treatment for XLH early is likely to be beneficial.

The committee understood that currently all children and young people with XLH are offered conventional therapy. However, the committee recalled that XLH is heterogeneous both in symptoms and severity, and in its response to treatment (either vitamin D and phosphate, or new treatments such as burosumab). The clinical and patient experts highlighted that, if recommended, they would expect to consider burosumab for all people with XLH. The clinical experts noted that it is not possible to predict how well the condition will respond to treatment. The committee recognised that, because XLH is a heterogeneous condition, it is challenging to understand who will benefit most from treatment. It concluded that, given the evidence presented, it was not able to identify any subgroups who would benefit more from treatment or who should be considered separately.

The committee recalled that skeletal deformities can continue into adulthood in people who had been taking conventional therapy from a young age (see section 4.2). The clinical experts noted that people are given the opportunity to stop conventional therapy when bone growth stops. A patient expert explained that, in this situation, people make the choice between continuing the treatment and having unpleasant side effects, or living with symptoms. They noted that some people with XLH can manage without the treatment after childhood but, for some people, the benefit of treatment outweighs the unpleasant side effects. The clinical experts suggested that there may be some benefit to continuing treatment into adulthood, and noted that about half of people who stop conventional therapy will start again when symptoms return. They noted that it is not uncommon for people who stop treatment to have stress fractures over the following 9 to 12 months, which causes significant pain. The committee acknowledged comments received during consultation stating that prolonged use of conventional therapy can cause serious side effects (kidney stones and hypercalcaemia). It concluded that XLH is a heterogeneous condition, and agreed that the benefit of continuing conventional therapy into adulthood would vary between individuals.

The committee acknowledged that, if the symptoms of XLH were to return after stopping burosumab, there may be consideration of whether continued treatment would be beneficial. It acknowledged that continued treatment with burosumab into adulthood might seem desirable, but noted that conventional therapy would have to be considered because the marketing authorisation for burosumab does not include adults. The committee agreed that it was unable to consider burosumab in adults because this is outside its marketing authorisation. It concluded that burosumab could only be considered for use in children and young people.

One parent recalled that their child had struggled emotionally and physically when having conventional therapy. They explained that there had been significant improvements in their child's symptoms, emotional state, independence and involvement in sporting activities and social events since starting treatment with burosumab. The clinical experts explained that it is challenging to heal bone defects with conventional therapy. They highlighted that only a small proportion of people – perhaps those with more mild disease and good adherence to treatment – are expected to have improvements with conventional therapy, but that burosumab is expected to provide significant bone healing. The clinical experts also explained that one of the main benefits with burosumab is the improved convenience of dosing compared with conventional therapy. The committee recalled that the complex dosing regimen of conventional therapy made it highly challenging (see section 4.2). It concluded that, based on the patient and clinical experts' experiences, burosumab could provide substantial benefits to people with XLH, including improved convenience.

The main clinical evidence submitted by the company came from 4 studies:

• CL205: an ongoing multicentre single-arm study including 13 children aged 1 to 4 years with XLH

• CL201: an ongoing multicentre open-label dose-finding study including 52 children aged 5 to 12 years with XLH, in which they were randomised to have burosumab either monthly or bi‑weekly (26 people had burosumab bi‑weekly, consistent with the marketing authorisation)

• CL301: an ongoing multicentre open-label randomised controlled trial including 61 children aged 1 to 12 years, comparing the efficacy and safety of burosumab with conventional therapy

• CL002: a retrospective natural history study in people aged 5 to 12 years who have had treatment with conventional therapy.
  
  The ERG explained that there were important differences between the studies, which made a comparison of results difficult to interpret. It noted that CL301 only included people with severe bone defects (that is, a Rickets Severity Score [RSS] of 2 or more), and that there were different dosing regimens for burosumab across studies. The committee understood that these differences could be associated with differences in treatment effect, and agreed that it should be taken into account when interpreting the clinical-effectiveness results.

The committee discussed the generalisability of CL301 to the XLH population in England, and noted that a population with an RSS of 2 or more may not be representative of the XLH population treated in clinical practice. The committee acknowledged that it would be more difficult to show a statistically significant treatment effect in people with less severe disease at baseline. It recognised that only including people with more severe disease in the studies could have affected the size of the estimated treatment effect. The committee noted that, before initial results from CL301 were available, there was no comparative evidence for children aged 1 to 4 years, and that the comparison of children aged 5 to 12 years was done by naive and matched comparisons of CL201 and CL002, which were highly uncertain. The committee acknowledged that CL301 provided comparative data from a randomised controlled trial for children aged 1 to 12 years, so went some way to addressing the uncertainties in the clinical evidence. However, it agreed that, because of restrictive inclusion criteria and small patient numbers, there was still uncertainty about the magnitude of the treatment effect. It concluded that CL301 provided a more robust comparison of burosumab with conventional therapy than was previously available, but that the magnitude of the relative treatment benefit was still uncertain.

The committee noted that there were no data in young people aged 13 to 17 years, but that this age group is covered by burosumab's marketing authorisation. It recognised that evidence collection in this age group would be challenging. The committee noted that people in this age group might stop burosumab treatment during a clinical study because their bones would mature, but that the timing of this would vary between people. It accepted that this could interfere with the collection of clinical-effectiveness data in this age group. The committee considered whether there might be a biological reason why XLH in people aged over 12 years would have a different response to treatment. The clinical experts explained that puberty is an important time for bone growth, and one of the patient experts explained that sometimes the biggest deterioration in rickets and leg bowing occurs in people aged between 11 and 16 years. The committee considered that people may benefit more from burosumab treatment in this period, but emphasised that there was no evidence to support this. The company noted that it is developing a European registry and running disease-monitoring programmes in the United States and Japan to generate more data in young people. The committee agreed that it would be beneficial to consider any data available in this age group. It concluded that the lack of evidence for the effects of burosumab in people aged between 13 and 17 years was a significant limitation.

The committee noted that the key outcomes from the clinical trials, RSS and Radiographic Global Impression of Change (RGI‑C), focused on the bone manifestations of XLH. It recognised that RSS and RGI‑C would not capture changes in metabolic manifestations of XLH. The ERG explained that these measures do not capture all aspects of XLH symptoms and progression. The patient experts explained that bone defects (in particular bowing of arms and legs) are important aspects of XLH, but that other aspects of the condition also substantially affect their lives. The company noted that outcomes relating to walking ability, pain, functional mobility, growth, fatigue and quality of life were all captured in CL301. The committee recalled that dental abscesses substantially affect quality of life in children and adults with XLH (see section 4.1). It agreed that it would have liked to have considered the effect of treatment on dental problems. The company noted that dental problems had not been captured as a study endpoint in the studies presented, but noted that it had plans to collect data on dental problems in a proposed long-term real-world-data programme. A clinical expert advised that there is a correlation between changes in bone symptoms (assessed using X‑rays) and other aspects of the condition. The committee acknowledged this, and noted that some of the wider effects of XLH would also be captured in measures of quality of life. It concluded that RSS and RGI‑C are informative outcome measures to quantify the clinical effectiveness of burosumab in children, and that other outcomes are also relevant to capture manifestations of XLH beyond rickets.

The committee discussed the clinical-effectiveness results presented for bone defects (RSS and RGI‑C); in particular:

• Children aged 1 to 4 years (CL205): treatment with burosumab resulted in an improvement in RSS and RGI‑C. Total RSS improved by 59% from baseline (2.92) to week 40 (1.19) with burosumab.

• Children aged 5 to 12 years (CL201 and CL002): there was a 58% improvement in total RSS score from baseline (1.92) to week 64 (0.81) in patients who had burosumab. These findings were supported by improvements in RGI‑C.

• Children aged 1 to 12 years (CL301): at week 40 there was a greater improvement in RGI‑C global scores in patients who had burosumab (1.92) compared with those who had conventional therapy (0.77). These findings were supported by improvements in RSS.

The committee noted that results from the naive and matched comparisons of CL201 and CL002 consistently showed that burosumab improved bone defects (measured by RSS and RGI‑C) more than conventional therapy. It considered that the results from CL301 were broadly consistent with those seen in the comparisons of CL201 and CL002. It agreed that the findings from CL301 were reassuring, and recognised that the clinical evidence suggested that treatment with burosumab was associated with improvements in bone defects. Further findings related to bone defects cannot be presented here because they have been marked academic in confidence by the company. The committee also considered data on growth (change in standing height and velocity), mobility (6‑minute walk test distance), fatigue, physical function, mobility and pain. It noted that there was a trend towards improvement in all of these outcome measures, and that improvements were larger for people who had burosumab compared with conventional therapy; the results are academic in confidence and cannot be reported here. The committee recalled that there is a correlation between changes in bone symptoms and other aspects of the condition (see section 4.11). It accepted that, while people are on treatment, burosumab is likely to improve symptoms of XLH beyond bone defects. The committee concluded that the clinical evidence suggested that burosumab was more clinically effective than conventional therapy.

The committee considered the clinical significance of the findings from the studies. It recalled its concern about the limitations of the clinical evidence base (see section 4.8). It also recalled the comments from patient and clinical experts, which highlighted experiences suggesting that the improvements associated with burosumab had a substantial effect (see section 4.7). It acknowledged that data from CL301 had resolved some of the uncertainty in the clinical evidence base, although some uncertainties remained. While acknowledging these, the committee concluded that the clinical evidence suggested that burosumab provided meaningful clinical benefits for people with XLH.

The committee discussed the long-term effects of improving skeletal deformities from a young age. The company highlighted that burosumab may reset people's bone architecture, and that correcting or preventing skeletal deformities before bone growth stops is expected to result in lifelong benefits. The committee recognised that the only opportunity to correct skeletal deformities was in childhood, and that doing so could provide important long-term benefits. The clinical experts reiterated that skeletal deformity is not the only manifestation of XLH, explaining that people will continue to have symptoms and be at risk of new bone defects into adulthood even if bone defects are corrected in childhood. The company acknowledged that burosumab cannot directly affect XLH in the long term after treatment stops, but noted that there will be lifelong benefits from correcting skeletal deformities in children. The committee acknowledged that this implied that burosumab would not affect the progression of bone manifestations of XLH during adulthood (for example, increased risk of osteomalacia and accompanying stress fractures and pain) or other ongoing aspects of the condition such as metabolic effects of low phosphate (for example, muscle weakness and fatigue; see section 4.1). The company stated that painful symptoms in adults with XLH are mostly caused by bone deformity, not osteomalacia. The committee recalled comments from one of the patient experts that pain was a result of established deformity and surgery (see section 4.1). Conversely, it recalled that many patients who stop having conventional management have painful symptoms, including stress fractures, soon after (see section 4.5), and noted that a similar effect might be seen after stopping burosumab (when bone growth stops). The committee acknowledged that the manifestations and severity of XLH would vary in adults, and agreed that the cause of painful symptoms could not be isolated. It concluded that there would be important long-term benefits from fixing or preventing childhood skeletal deformities, although burosumab would not affect other aspects of XLH in the long term.

The committee considered the adverse effects of burosumab and conventional therapy. It acknowledged that the frequently occurring adverse events in the clinical trials were consistent with those expected in a trial setting. It noted that the only notable treatment-related adverse reactions for burosumab were due to the subcutaneous administration of the treatment. The committee recalled comments from the patient and clinical experts stating that conventional therapy results in a variety of unpleasant side effects (see section 4.2). The committee concluded that the adverse effects associated with burosumab were more manageable than those with conventional therapy.

The company presented an economic analysis comparing burosumab with conventional therapy. This was based on a Markov model with 4 health states using RSS: healed (RSS of 0), mild (RSS of 0.5 to 1.0), moderate (RSS of 1.5 to 2.0), and severe (RSS of 2.5 or more). The committee discussed in detail whether defining health states using a measure of bone defects (RSS) was a suitable proxy for disease severity and progression in people with XLH. It recalled that RSS is a measure relevant to children with growing bones (see section 4.11), and recognised that it would not capture new bone disease in adults or other manifestations of XLH. However, the clinical experts explained that RSS and other aspects of XLH are correlated in children. The committee considered that, in the absence of an alternative measure of disease severity in XLH, RSS was an acceptable and measurable proxy. In response to the evaluation consultation document, the company presented a revised analysis using an amended version of the economic model. This incorporated a commercial arrangement and changes to transition probabilities, costs and utility values. The committee concluded that the model structure adequately captured the disease progression of XLH in children, and agreed that the revised analysis results from the economic model could be considered in its decision-making.

To model transitions between the health states of the economic model, the company used data from CL205, CL201, CL301 and from a UK chart review study. Transition probabilities for people having burosumab were based on RSS data from CL205 (children aged 1 to 4 years), CL201 (children aged 5 to 12 years; licensed dose group only) and CL301 (children aged 1 to 12 years). The company explained that the inclusion criterion in CL301 of an RSS of 2 or more meant that data from CL205 and CL201 were needed to model transitions in less severe health states. The ERG highlighted that differences in the study populations reduced the suitability of pooling results from the different studies. It also noted that, in its revised analysis, the company calculated a single transition probability matrix for all ages, so assumed that treatment effect does not depend on age. The committee acknowledged the ERG concerns, but agreed that it did not expect this change to have a substantial effect on the analysis. It recognised that the inclusion of data from CL301 allowed for more information to be used in calculating the transition probabilities and agreed this was appropriate. The committee recalled that there was no clinical evidence for people aged between 13 and 17 years (see section 4.10). It agreed that extrapolating transition probabilities based on trial data for people aged 5 to 12 years to young people aged over 12 years introduced additional uncertainty. Data from CL301 and the review of UK patient charts were used to generate transition probabilities in the conventional therapy arm. The committee highlighted that the assumed transition probabilities implied that people having burosumab could not deteriorate (as measured by RSS). It considered that generating the transition probabilities from the trial data was appropriate, although it acknowledged that the very small population sizes in the trials introduced uncertainty into the model. The committee was also aware that the transition probabilities were based on trial data over about 1 year, but were applied to all model cycles during childhood. It noted that there was no evidence for the effects of burosumab over 2 or more years that could inform later transition probabilities. The committee concluded that the transition probabilities were uncertain but acceptable for decision-making.

In its revised analysis, the company estimated the age at which growth stops to be 14 years in girls and 16 years in boys. This was based on the mean age of final height defined as a growth velocity of less than 2 cm per year. The committee noted that, in the company's original base case, the treatment stopping age was 16 years in girls and 17 years in boys. It acknowledged that stopping treatment earlier had a substantial effect on the estimated treatment costs and hence the incremental cost-effectiveness ratios (ICER). It was aware that the earlier treatment stopped in the model, the lower the ICER, so the original base case was more conservative than the revised base case. The ERG stated that the company had misinterpreted the growth charts and that, based on the proposed definition of final height, the treatment stopping ages should be 15 years in girls and 17 years in boys. The committee considered how final height would be determined in clinical practice. One clinical expert explained that they used radiographs of the hands to identify closure of the growth plate. The committee understood that bone growth stops after the growth plate has closed. It recognised that there was value in using an objective measure, and recommended that an objective measure of growth completion should be used in clinical practice. Alternative clinical advice was that growth completion as identified by a hand radiograph would be expected to happen at about the same time as when it is identified by final height defined as a growth velocity of less than 2 cm per year; the expert considered that radiographs are not necessary. The committee was reassured that, regardless of which approach were to be adopted in practice, the model would be suitable for decision-making because treatment would be unlikely to continue for longer than estimated in the company's economic modelling. The committee concluded that treatment stopping age was an important variable in the assessment of cost effectiveness, and agreed that it would consider the effect of the different stopping ages on the results.

The company assumed that, when treatment was stopped (see section 4.19), people would remain in their current health state for the rest of the modelled time horizon. The committee noted that, in the model, nearly all people having burosumab were in the healed health state by the time treatment was stopped, whereas there was a distribution of people across the different health states for conventional therapy. It queried the plausibility of this prediction, recalling its conclusion that there would be long-term benefits from fixing or preventing skeletal deformities in childhood. However, it agreed that burosumab would not affect other aspects of XLH in the long term (see section 4.15). The committee also recalled that skeletal aspects of XLH are expected to progress during adulthood (see section 4.1), and considered that the fact that some people continue to have conventional therapy in adulthood (despite its side effects) implied that the condition had not stabilised. The committee was aware that there was no evidence on the long-term progression of XLH after burosumab treatment in childhood. It concluded that assuming lifetime disease stabilisation after stopping treatment was unrealistic, but that the nature of disease progression was unknown.

The committee considered how to incorporate long-term disease progression in the economic modelling. It noted the ERG's scenario analysis in which it assumed that, after 20 years, people in the burosumab arm would move down to the next (more severe) health state utility value (see section 4.24). The ERG explained that it adopted this approach because of limitations in the model and the lack of evidence to inform transition probabilities in adulthood, but acknowledged that this was simplistic. The company highlighted that there was no evidence to support a greater long-term utility decline for people who had burosumab in childhood than for those who had conventional management. It stated that, if applied, a utility decrement should be applied to both groups. The committee accepted that the ERG's approach was intended to capture a waning of burosumab treatment effect, but that the company had not intended to imply a direct effect of treatment after it stopped (that is, the long-term benefit of burosumab was solely through improving the patient's health state at the start of adulthood). It acknowledged that, if there was a long-term direct treatment effect, the ERG's approach might be reasonable, but that there was no evidence to support this. Conversely, the committee considered that, if the decline in utility captured the long-term progression of disease, an assumption of disease progression should be applied to both burosumab and conventional therapy. It considered the ERG's approach, and agreed that the duration of disease stabilisation and persisting benefits of burosumab would vary between individuals and were uncertain. It also noted that the ERG's approach applied a utility decrement that was the same size as the decrement seen with normal ageing over 40 years. The committee recognised that there was no evidence available to estimate the size of the most appropriate utility decrement, and it agreed that the magnitude of the decrement was uncertain. Overall, the committee considered that the ERG's approach was uncertain and was purely illustrative of the potential effect of reducing the long-term treatment benefit of burosumab, and therefore did not inform the most plausible scenario for decision-making.

In its response to the evaluation consultation document, the company implemented a new approach to model long-term disease progression. It asked a group of experts to value (using EQ‑5D‑5L) the quality of life of people with XLH aged 18 years, 40 years and 60 years, based on health state vignettes. Using these estimates, the company estimated a decrease in quality of life over time across the model health states. The ERG emphasised that this did not capture any waning in the long-term treatment effect of burosumab. The committee acknowledged that the company's approach incorporated more information than the ERG's illustrative method, and appeared to capture long-term disease progression, but recognised there were limitations with both methods. It concluded that, in the absence of long-term evidence, the company's approach was more appropriate for decision-making purposes, but acknowledged that the estimation of long-term outcomes remained highly uncertain.

The company's base case incorporated a non-reference-case discount rate of 1.5% for costs and benefits because the benefits of treatment are expected to be substantial and sustained over a lifetime. The committee was aware that NICE's guide to the methods of technology appraisal (2013) and its interim process and methods of the highly specialised technologies programme (2017) states that a non-reference-case rate of 1.5% may be used when: treatment restores people to full or near-full health when they would otherwise die or have severely impaired lives; when this is sustained over a very long period (normally at least 30 years); if it is highly likely that there will be long-term benefits; and if the treatment does not commit the NHS to significant irrecoverable costs. The committee recalled its conclusions about long-term treatment benefit (see sections 4.15, 4.19 and 4.21). It acknowledged that there would be a lifetime effect from correcting skeletal deformities in children, and therefore that burosumab had the potential to resolve one of the most important aspects of XLH. The committee agreed that burosumab might restore people's skeletons to normal or near-normal. However, it considered that it was unlikely that the drug would restore people's overall health to normal or near-normal because symptoms of XLH are expected to continue and even progress in adulthood. In particular, the committee highlighted that the fact that some people choose to continue with conventional therapy into adulthood (despite unpleasant, and potentially serious, side effects) illustrated that this was not consistent with 'normal or near-normal health'. Moreover, it noted that, if there is established bone deformity at the time burosumab treatment starts, this damage may not be resolved. The committee also recalled the substantial uncertainties about the long-term benefits of burosumab. It concluded that the criteria for deviating from the reference case discount rate of 3.5% for costs and benefits were not met.

To model health-related quality of life, the company conducted a utility study in which vignettes describing the modelled health states were developed. The vignettes were valued using EQ‑5D‑5L by 6 clinicians experienced in treating XLH. The committee noted that the utilities were scored by clinicians not patients, and were not taken directly from trials, which were limitations of the data. The vignettes assumed that all aspects of quality of life were worse in more severe health states (that is, there was perfect correlation between RSS and all aspects of quality of life). The ERG explained that asking experts to value the quality of life of hypothetical people is not ideal, and generates results that are substantially uncertain. The company noted that some responses were not provided, and explained that it had inferred 1 value for the healed health state. The committee accepted that this inference was logical, and agreed that the minor amendment to the utility values was appropriate. It concluded that the utility values were uncertain but, in the absence of an alternative, were acceptable for decision-making.

The committee considered the effect of XLH on caregivers. It recalled that, because of the genetic nature of XLH (see sections 2.1 and 2.2), many adults with the condition will often be carers for other affected family members. It recognised that this might make the effect on carers particularly acute. In its response to the evaluation consultation document, the company estimated the effect of XLH on caregivers of children with XLH using a 'disutility' value taken from a published study of people with limited mobility. The company highlighted that it was practically challenging to accurately model the full effect of XLH on caregivers. The ERG noted that the company's approach of using a published disutility value was broadly reasonable in this context. The committee questioned the appropriateness of the disutility value that was identified, although it welcomed the company's attempt to quantitatively estimate the effect on caregivers. The committee concluded that it was important to consider carer burden in its assessment of burosumab. It further concluded that it would take into account results including a quantitative estimate of carer burden, but because the estimate provided was not robust, it would also consider the burden qualitatively (using analyses in which the quantitative burden was omitted).

The committee discussed the probabilistic analyses presented by the company and the ERG, and was aware that the results differed substantially from each other and from the deterministic analysis. The company and the ERG explained that they explored the use of prior distributions in a Bayesian approach to generating transition probabilities in the probabilistic analysis. They did this to try and capture the uncertainty in the transition probabilities, and to resolve issues that resulted from having limited data, particularly about transitions that might be expected to occur occasionally but were not seen in the studies. The committee understood that both the company and the ERG used a uniform prior distribution. However, the company used a factor of 0.05, whereas the ERG provided results from a range of values with a preference for using a factor of 1.00. The committee noted that the company's approach increased the ICERs compared with the deterministic analysis, and that the ERG's preferred factor caused a substantially greater increase in the ICERs. The committee considered that, with few observed data points, the ERG's preferred factor placed too much emphasis on the prior distribution, and led to model results (ICERs) that lacked face validity. In particular, the committee highlighted that incorporating uncertainty in the transition probabilities through the probabilistic analysis was expected to increase the uncertainty around the cost effectiveness but not to dramatically increase the ICER (as was seen). The committee considered that the substantial differences between the probabilistic and deterministic ICERs, and between probabilistic scenarios, were driven by the choice of prior distribution, and not by other issues with the probabilistic analyses. It concluded that the probabilistic analyses outlined the magnitude of the uncertainty in the results, that the lack of observed data affected the robustness of the economic analyses, and that the mean ICERs from the probabilistic analyses were not suitable for decision-making.

The committee acknowledged that there were several other assumptions and uncertainties in the economic model. The ERG noted that the company based the starting distribution of health states and weight on the CL205 and CL201 studies, and UK growth charts. It suggested that this was unnecessary because data were available from the UK chart review. In response to the evaluation consultation document, the company corrected a methodological error in the burosumab transition probabilities and incorporated adverse events costs in the model. The committee recognised that there were additional uncertainties in the economic model, but concluded that the presented model was adequate for decision-making.

The committee considered the results of the economic analysis, taking into account the company base case, the ERG's exploratory analyses and the committee's preferred scenario. The committee considered that its preferred scenario was that based on the following amendments to the company's original base case:

• incorporating a cost for adverse events (see section 4.27)

• correcting burosumab transition matrices to account for a flaw in the methodology (see section 4.17)

• incorporating data from CL301 into the transition probabilities for burosumab and conventional therapy (see section 4.18)

• using utility values from the vignette study with an additional inferred value for the healed health state (see section 4.24)

• incorporating the company's estimate of progressive utility decline (see section 4.22)

• discount rate of 3.5% for costs and benefits (see section 4.23).
  
  The committee also took into account the age at which treatment is stopped (see section 4.19) and the effect of carer burden (see section 4.25). Based on the committee's preferred assumptions, and using the most conservative treatment stopping ages (16 years in girls and 17 years in boys) but omitting the caregiver disutility, the ICER was £149,565 per quality-adjusted life year (QALY) gained. Based on these assumptions and using the more optimistic treatment stopping ages (14 years in girls and 16 years in boys), and including the caregiver disutility, the ICER was £112,517 per QALY gained. In these scenarios, the incremental QALY gains associated with burosumab ranged from 5.52 to 15.99. The committee concluded that the ICER was in the range of £113,000 to £150,000 per QALY gained.

The committee understood that NICE's interim process and methods of the highly specialised technologies programme (2017) specifies that a most plausible ICER of below £100,000 per QALY gained for a highly specialised technology is normally considered an effective use of NHS resources. For a most plausible ICER above £100,000 per QALY gained, judgements about the acceptability of the highly specialised technology as an effective use of NHS resources must take account of the magnitude of the incremental therapeutic improvement, as revealed through the number of additional QALYs gained. Taking into account the incremental QALY gains (between 5.52 and 15.99), the committee concluded that burosumab could meet the criteria for applying a QALY weight.