The Evaluation Committee (section 8) considered evidence submitted by the company of elosulfase alfa, a review of this submission by the Evidence Review Group (ERG; section 9), evidence submitted by clinical experts, patient experts and NHS England, and the responses to consultation on the first and second evaluation consultation documents from all consultees and commentators and members of the public. The Evaluation Committee also considered the managed access agreement.
4.1 It was estimated at the time of the evaluation that 88 people were living with mucopolysaccharidosis type IVa (MPS IVa) in England, with about 3 new diagnoses expected per year. About 74–77 people were anticipated to be eligible for enzyme replacement therapy and may want treatment with elosulfase alfa. Patient experts and patient groups highlighted the substantial impacts of MPS IVa on the quality of life of people with the condition and their families.
Most people with MPS IVa have seriously reduced mobility and stamina, making many daily activities challenging and leading to dependence on a wheelchair.
Spinal instability puts people at a high risk of injury and medical complications, and often needs high‑risk surgery. In particular, compression of the spinal cord needs surgical management by highly experienced surgical teams. Without careful monitoring, or because of an accident, patients can become paralysed or die.
Bone and joint problems often cause chronic pain.
Progressive hearing loss and frequent periods of diarrhoea can be debilitating and isolating.
MPS IVa causes reduced life expectancy; the average life expectancy in people with this condition is about 25 years.
The combination of symptoms in MPS IVa, including physical features and short stature, can cause considerable anxiety, depression and low self‑esteem.
Expensive wheelchairs and home adaptations carry a financial burden, which can be exacerbated by limitations on income caused by difficulty working.
MPS IVa also significantly affects the families of people with the condition. Many parents struggle to continue working as care needs increase.
Patient groups highlighted testimonies from families involved in clinical trials of elosulfase alfa. They describe substantial improvements in quality of life, and anticipate that stabilising disease progression and maintaining mobility and stamina could have a particularly important impact.
4.2 The company also presented evidence from an observational study of the natural history of MPS IVa (MOR‑001), which followed 325 people with MPS IVa for up to 10 years. This study suggested that the progression of MPS IVa over time leads to decline in endurance, restricted growth and limitations in the activities of daily living. Evidence was also presented from cross‑sectional surveys of 63 people with the condition and of 56 families. The results showed that the effect of MPS IVa on quality of life and its effect on carers are both related to how much the person relies on a wheelchair. Quality of life is also correlated with endurance, pulmonary function and height. The survey of families found that people with the condition need up to 15 hours of care per day. Carers often report stress, lack of sleep, back pain, anxiety and depression, as well as effects on family and social life and finances.
4.3 The testimonies from families of children and adults who took part in clinical trials of elosulfase alfa gave individual illustrations of the benefits of treatment. They noted that some children who had elosulfase alfa continued to grow, experienced less physical deterioration, avoided surgery and did not become dependent on a wheelchair. Adults experienced a significant increase in energy levels and reduction in pain, enabling them to work and have a more normal life. Patient groups emphasised the positive impact of stabilisation of the disease symptoms on quality of life and life expectancy. They highlighted the importance of outcomes such as improved energy levels, reduced disability and wheelchair dependence. They acknowledged that the testimonies are subjective and that elosulfase alfa will not cure MPS IVa nor undo existing skeletal damage and disability, and stated that it is unknown how well the short‑term benefits seen in clinical trials predict long‑term outcomes. Clinical experts noted that elosulfase alfa is expected to slow the progression of disease, reduce the need for surgery and improve quality of life. Elosulfase alfa is unlikely to reduce certain complications, such as spinal cord damage, and may not remove the need for supportive interventions.
4.4 The company identified 7 clinical studies of elosulfase alfa for MPS IVa, involving 255 patients. These included 1 randomised, placebo‑controlled phase III trial (MOR‑004) and its long‑term extension (MOR‑005), 1 open‑label ascending dose study (MOR‑002) and its long‑term extension (MOR‑100), 1 randomised phase II trial (MOR‑008), and 2 single‑arm studies in specific population groups (people under 5 years and people with limited mobility; MOR‑006 and ‑007).
4.5 The pivotal clinical effectiveness trials for elosulfase alfa were the MOR‑004 and ‑005 studies; the MOR‑002, ‑100, ‑007 and ‑008 studies provided supportive evidence, and no results were available from MOR‑006 at the time of the evaluation. On enrolment into study MOR‑004, 176 patients were randomised (1:1:1, stratified by age and endurance) to elosulfase alfa 2 mg/kg/week every week (licensed dose; n=58), elosulfase alfa 2 mg/kg/fortnight (unlicensed dose; n=59) or placebo (n=59), all for 24 weeks. Patients and investigators were blinded to treatment allocation. After 24 weeks, patients could enter the long‑term extension, MOR‑005 (n=173). At this stage, patients in the placebo group were randomised to elosulfase alfa 2 mg/kg/week or 2 mg/kg/fortnight; patients in the elosulfase alfa groups continued with their original treatment assignment, which remained double‑blind. When the primary efficacy analysis of MOR‑004 was complete (based on week 24 data), all patients in MOR‑005 switched to open‑label elosulfase alfa 2 mg/kg/week for the rest of the study (up to 240 weeks). Outcomes included endurance (measured using the 6‑minute walk test [6MWT, primary outcome measure] and 3‑minute stair climb test [3MSCT]), urine keratan sulphate (uKS), pulmonary and cardiac function, anthropometric measures, vision and hearing. The baseline and disease characteristics were broadly comparable between treatment groups in MOR‑004 and ‑005, although there were some small differences between groups, particularly in endurance scores in MOR‑005.
4.6 The primary efficacy analysis of MOR‑004 showed that elosulfase alfa 2 mg/kg/week was associated with a statistically significant improvement in 6MWT at week 24 compared with placebo. Patients treated with elosulfase alfa 2 mg/kg/week had a mean improvement from baseline in 6MWT of 23.7 metres at week 12 and 36.5 metres at week 24, equating to a mean difference between elosulfase alfa and placebo of 22.5 metres at week 24 (95% confidence interval [CI] 4.0–40.9 metres, p=0.017). A pre‑specified subgroup analysis showed consistent improvements in 6MWT with elosulfase alfa 2 mg/kg/week compared with placebo across subgroups based on age, sex, endurance at baseline, geographical region and race.
4.7 Longer‑term analyses of endurance were presented from MOR‑005, ‑002 and ‑100. In the interim analysis of MOR‑005 at week 72 (that is, 72 weeks after enrolment to study MOR‑004 and 48 weeks after enrolment to MOR‑005), continuous treatment with elosulfase alfa 2 mg/kg/week was associated with a sustained increase from baseline in 6MWT distance (week 72: 30.1 metres and 46.0 metres in the intention‑to‑treat and per‑protocol analyses respectively). In MOR‑002 and ‑100, elosulfase alfa appeared to be associated with improvements in 6MWT, although analyses were based on small patient numbers and had wide confidence intervals.
4.8 The company presented analyses of secondary and tertiary outcomes from studies MOR‑004, ‑005, ‑100, ‑007 and ‑008. The results of MOR‑004 suggested that elosulfase alfa 2 mg/kg/week provided improvements in endurance, pulmonary function, anthropometrics and quality of life compared with placebo at week 24. Statistical significance was not reached, although the company noted that the study was not powered to detect differences in secondary or tertiary outcomes. In this study, no substantial changes in corneal clouding were seen with elosulfase alfa compared with placebo, whereas the hearing results showed some numerical changes based on a small population (n=6–9 per group). The number of people using a wheelchair at week 24 increased by 5 in the placebo group but did not increase in the elosulfase alfa 2 mg/kg/week group (although fewer people in the placebo group used a wheelchair at baseline than in the elosulfase alfa 2 mg/kg/week group). In MOR‑005, there was a reduction in wheelchair dependency at week 72 with elosulfase alfa 2 mg/kg/week compared with baseline. Patients treated continuously with elosulfase alfa 2 mg/kg/week had fewer surgical operations and those that did had them later in the study, compared with the other treatment groups. In children under 5 years (MOR‑007), the mean height z‑score showed no significant change from baseline to week 52 (−2.0 at baseline and −2.2 at week 52). The company noted that, in untreated patients of a similar age in the observational study MOR‑001, the mean height z‑score decreased from −2.3 to −3.1 over 52 weeks. Evidence for the effect of elosulfase alfa on sleep apnoea was collected in study MOR‑008, and the results showed no clear trends in the change from baseline to week 24. However, this study included patients with slowly progressive disease (attenuated) who have less obstructive sleep apnoea than classical patients. Analyses of urine keratan sulphate levels showed a consistent decrease from baseline associated with elosulfase alfa treatment across the clinical studies. The decrease was about 40% over 24 weeks, and was statistically significantly larger than with placebo (MOR‑004). The company stated that this confirms the biological activity of the enzyme replacement. In analyses of the effect of treatment on a combination of multiple endpoints in MOR‑004, the results were statistically significantly or borderline significantly in favour of elosulfase alfa 2 mg/kg/week compared with placebo (p values from 0.002 to 0.053).
4.9 The company presented a responder analysis for endurance and pulmonary outcomes at week 72 in people treated with elosulfase alfa 2 mg/kg/week throughout MOR‑005. Patients were classed as 'multi‑domain responders' if they had any improvement from baseline in endurance (measured using 6MWT or 3MSCT) and any improvement from baseline in pulmonary function, as 'single‑domain responders' if they had improvements in either endurance or pulmonary function, and as 'non‑responders' if they had no improvement in either outcome. Most patients treated with elosulfase alfa 2 mg/kg/week were 'multi‑domain responders', and the remainder were 'single‑domain responders'. The number of 'single‑ and multi‑domain responders' were reported as academic in confidence and so cannot be reported here.
4.10 The company presented adverse event data from 235 people who had elosulfase alfa during the clinical trial programme, of whom 222 had it at the licensed dose of 2 mg/kg/week. Of people who had elosulfase alfa 2 mg/kg/week, 77% reported at least 1 adverse event; the most common included vomiting (35%), fever (34%) and headache (34%). Serious adverse events occurred in 29% of people who had elosulfase alfa 2 mg/kg/week; in the MOR‑004 trial most adverse events were mild (48%) or moderate (45%). The company reported that most serious adverse events were related to the underlying disease or the intravenous administration of elosulfase alfa. There were no adverse events that led to the treatment being permanently stopped, and there were no deaths during the clinical trials. About 20% of people taking elosulfase alfa 2 mg/kg/week had a hypersensitivity reaction, and 71% had an infusion‑associated reaction, the most common of which included fever, vomiting and headache. These infusion‑associated reactions decreased in frequency with longer‑term treatment. Three infusions were interrupted or stopped because of hypersensitivity reactions, and 14% of people had their infusions interrupted or stopped because of infusion‑associated reactions. Clinical experts stated that infusion‑associated reactions are an accepted complication of enzyme replacement therapy and respond to treatment.
4.11 The company presented a cost–consequence analysis comparing elosulfase alfa 2 mg/kg/week with established clinical management. The analysis was based on a Markov model with 7 states, representing the progression of MPS IVa based on increasing use of a wheelchair. The model included people with MPS IVa, with the demographic and disease characteristics (for example, body weight) based on the population in the MOR‑001 observational study. The company noted that this was a younger group with less‑advanced disease than people with MPS IVa in England. The model had a lifetime time horizon (that is, it simulated costs and benefits over a maximum of 100 years) and a 1‑year cycle length. The analysis was conducted from the perspective of the NHS and Personal Social Services (PSS), and costs and benefits were discounted at a rate of 1.5% per year.
4.12 At the start of the model, patients were distributed between the 7 health states according to the range of symptoms and wheelchair use seen at the start of MOR‑001. Patients then progressed through the model. Patients who started in the 'asymptomatic' state transitioned to the 'no wheelchair' state at age 3 years if treated with established clinical management (the 'established management group'), and 5 years later if treated with elosulfase alfa (the 'elosulfase alfa group'). This assumption was based on expert opinion. Patients then moved between health states at a rate based on changes in wheelchair use, 6MWT and forced vital capacity (FVC) seen in MOR‑001 and ‑005. In the elosulfase alfa group, people whose disease responded in both endurance and pulmonary domains ('multi‑domain responders'; see section 4.9) were assumed not to have any progression in wheelchair use, 6MWT or FVC score. This assumption was based on expert opinion and experience of the effect of enzyme replacement therapy in related disorders such as MPS VI. People whose disease responded in 1 domain ('single‑domain responders'; see section 4.9) had a slowing of progression to 50% of the rate in the established management group. In each state, some patients had surgery, and either recovered (after a recovery period with lower quality of life), developed paraplegia because of complications (and transitioned to the 'paraplegic' state) or died (and transitioned to the 'death' state). Mortality was based on the assumption that people treated with elosulfase alfa have the same mortality risk as the general population, and that people in the established management group have a 3.03‑fold greater risk of death than those in the elosulfase alfa group.
4.13 Quality of life was captured in the model by assigning utility scores to each health state and a utility decrement in the recovery period after surgery. The utility scores in the established management group were based on the general population (asymptomatic state) and the surveys of the people with the condition and their families (see section 4.2). In addition, each health state included an effect of the disease on carers' quality of life (a disutility). An increase in utility score (utility increment) was applied to the elosulfase alfa group in each health state. This utility increment was based on the improvement in 6MWT and FVC seen in clinical trials of elosulfase alfa, combined with the correlation between 6MWT and FVC and quality of life seen in the survey of people with MPS IVa (see section 4.2). The model did not include any effect of adverse events on quality of life.
4.14 The model included costs associated with treatment (acquisition cost of elosulfase alfa 2 mg/kg/week and infusion‑related costs) and the disease (health‑state costs). The acquisition cost of elosulfase alfa was based on the list price and the discount agreed in the patient access scheme; the discount is commercial in confidence and so cannot be reported here. Because elosulfase alfa is dosed on a per kilogram basis, the cost of elosulfase alfa was also influenced by patient weight. The model took this into account by assigning an average weight to people in each health state (based on MOR‑001) and assuming that weight stayed the same throughout each health state (that is, patients' weights changed only when they transitioned between health states).The health‑state costs included primary and secondary care appointments, emergency treatment and surgery. The frequency of these appointments was based on expert opinion. In addition, each health‑state cost included costs for daily care, based on the number of hours of care reported in the survey of the parents of people with MPS IVa (see section 4.2); the company assumed that 50% of this care was given by professional carers. No costs associated with adverse events were included. NHS costs were taken from the Personal and Social Services Research Unit, NHS reference costs and literature sources.
4.15 In the company's base case, established clinical management was associated with £618,812 in costs and 9.75 quality‑adjusted life years (QALYs) over the lifetime of the model. Elosulfase alfa was associated with 27.93 QALYs, amounting to 18.18 additional QALYs compared with established clinical management. The acquisition cost of elosulfase alfa over the lifetime of the model was £14,014,636 (based on the list price). The total costs in the elosulfase alfa group were provided as commercial in confidence and so cannot be reported here.
4.16 The company presented a one‑way (deterministic) sensitivity analysis and probabilistic sensitivity analysis to assess the uncertainty in the model. The deterministic sensitivity analysis showed that the results were most sensitive to the discount rate for benefits and costs, body weights, and utility scores in the 'no wheelchair', 'sometimes wheelchair' and 'wheelchair‑dependent' states and the progression of disease in 'single‑domain responders'.
4.17 The company also presented scenario analyses to explore the effects of assumptions about the population, perspective, effect on carers, mortality, utilities and the treatment benefit associated with elosulfase alfa. The incremental costs and QALYs were most influenced by the use of a 'birth cohort' (that is, modelling patients from birth rather than their current health state), the modelling of mortality and the assumptions about the effect of elosulfase alfa on disease progression.
4.18 The company presented a budget impact analysis to predict the costs of elosulfase alfa in the NHS and PSS. The company stated that, of the 88 people living with MPS IVa in England, 74–77 are expected to want elosulfase alfa treatment. The company assumed 50 people would have treatment in the first year and the remainder would start in the second year. The results of the budget impact analysis suggested that the net budget impact of elosulfase alfa (based on the list price) would be £17,310,564 in the first year. The analysis suggested that the net budget impact would rise to £28,823,649 in year 5. These figures do not take into account the arrangements included in the managed access agreement.
4.19 The company stated that people with MPS IVa can gain high achievements in education and employment, but may be limited by progression of the disease. It also emphasised the effects of the MPS IVa on the parents, siblings and carers of people with the condition, and suggested that elosulfase alfa may provide benefits in quality of life, education, and employment for families and carers.
4.20 The company anticipated that treatment with elosulfase alfa could result in cost savings outside the NHS. These could include savings in the education, local government and welfare budgets, for example, through reductions or delays in school and home adaptations, and reduced welfare payments. In addition, if people with MPS IVa remain in employment for longer, they may contribute more income tax.
4.21 The company noted that people with MPS IVa incur substantial personal costs. These include adaptations of their homes and cars, bespoke clothes and specialist mobility equipment (including wheelchairs). In addition, families may incur costs for travelling to hospital appointments, time off work to care for their family, privately funded care or treatment, and special help with schooling.
4.22 The company described the Morquio A Registry Study (MARS), an ongoing registry of people with MPS IVa collecting natural history data, and clinical and patient‑reported outcomes worldwide, including England. The registry will also collect long‑term efficacy and safety outcomes for elosulfase alfa. The company stated that this registry represents a significant development for strengthening the evidence base in this condition.
4.23 The company stated that no additional infrastructure is needed to ensure the safe and effective use of elosulfase alfa in centres with experience in diagnosing and treating lysosomal storage disorders. It noted that there are 8 designated national centres for the diagnosis and management of lysosomal storage disorders, 7 of which have been involved in clinical studies of MPS IVa. The company highlighted that the specialist centres have extensive experience of enzyme replacement therapies. NHS England also anticipated that the service impact of the technology would be small, although clinical experts noted that giving the drug outside specialist centres may need additional staff training.
4.24 The ERG considers that, although there were some methodological shortcomings, the systematic review captured all relevant evidence including several reasonable quality clinical studies. It noted some challenges in interpreting the clinical trial data (which are not uncommon in clinical research for rare diseases), including:
the limited number and relatively short duration of randomised trials
heterogeneity in the natural history of MPS IVa and the observed treatment effect
limitations in the outcome measures
an apparent beneficial effect associated with placebo in MOR‑004
a potential confounding effect of surgical interventions during MOR‑005.
Nevertheless, the ERG noted that there was a statistically significant benefit in 6MWT associated with elosulfase alfa in MOR 004. It also noted that the observed benefit was clinically meaningful (although there is no empirical evidence to support this).
4.25 The ERG commented that the longer‑term data from MOR‑005 suggest that improvements with treatment are sustained (although the gains in 6MWT decline with time, as may be expected in a progressive condition). However, the results should be interpreted with caution because of the possible placebo effect. The ERG also noted that the company presented an analysis of the per‑protocol group to minimise the confounding effect of surgery. This excluded people who needed surgery and the ERG considered that the intention‑to‑treat group was more representative of the patient population to be treated in the NHS.
4.26 The ERG noted that the 6MWT is a surrogate outcome, but accepted that it is appropriate for this condition. However, it highlighted limitations in this outcome about the meaningfulness of observed changes, sensitivity to improvements, validity of the test and potential for confounding. Similar limitations were noted for the 3MSCT. In addition, the ERG noted that both the 6MWT and 3MSCT may be affected by the test conditions, but that the 3MSCT in particular was not standardised across centres and studies. Clinical experts noted in their submissions that the surrogate and composite measures used in trials were not satisfactory but were the only available options. A patient expert noted in their submission that the improvement in quality of life associated with elosulfase alfa might be greater than the increase in 6MWT, and noted that even a small improvement in endurance could make a substantial difference to the quality of life of a person with MPS IVa.
4.27 The ERG highlighted that the responder analysis for studies MOR‑004 and ‑005 (see section 4.9) included no statistical comparisons, and that it was unclear whether the changes in 'responders' were clinically significant. The ERG noted that the change in FVC in 'single‑domain responders' was counterintuitive. The clinical significance of the changes in height and growth rate were unknown, and the analyses of wheelchair use and audiometry were limited by small population sizes. The ERG noted that urine keratan sulphate may be a useful surrogate outcome as a marker of lysosomal function, but that it is not correlated with clinical outcomes.
4.28 The ERG considered that the company's economic model was broadly appropriate and followed methodological guidance and the NICE reference case. The progression of disease was coherently modelled and reflected the natural history, and the time horizon and cycle length were appropriate. The ERG noted that the discounting rate of 1.5% per year might be considered reasonable, in the context of the NICE guide to the methods of technology appraisal 2013 (which it considered to be relevant to the highly specialised technologies programme), but explored this in a scenario analysis.
4.29 The ERG considered that the assumptions used to model clinical effectiveness were uncertain. It stated that the assumptions were not fully consistent with the evidence and overestimated the benefits associated with elosulfase alfa. The ERG noted that the model assumed that elosulfase alfa would delay the onset of symptoms by 5 years and stabilise established disease (that is, prevent any progression) in 'multi‑domain responders'. However, it considered these assumptions to have been optimistic and based on limited evidence, and to have predicted an implausible increase in clinical benefits during the model. The ERG explored these assumptions in scenario analyses. The ERG also noted that progression between health states was based on evidence from MOR‑005 and so was limited by the non‑randomised nature of the study and small population sizes.
4.30 The ERG noted that the model included a mortality benefit associated with elosulfase alfa both through delaying disease progression and through the relative risk of death. It considered that this double‑counted the mortality benefit associated with treatment.
4.31 The ERG considered that the modelling of quality of life was mostly reasonable, but noted some concerns. It highlighted that elosulfase alfa was assumed to improve quality of life in each health state, and also to delay progression of disease. The ERG considered that this double‑counted the utility benefit associated with elosulfase alfa treatment. The ERG also noted that the utility value for the 'paraplegic' health state was the same as that for the 'wheelchair‑dependent' health state, despite feedback from experts that quality of life may be worse in patients with paraplegia. It also commented on the marked decline in utility between occasional and constant wheelchair use. The study from which the health state utility scores were obtained reported utility values for both children and adults (although only the values for adults were used in the model). The ERG noted that the 2 sets of utility values were different, and the reasons for this were largely unknown. The ERG highlighted that quality‑of‑life evidence from MOR‑004 and ‑005 does not appear to have been used in the model.
4.32 The ERG considered the resource use and costs included in the model to be generally appropriate. However, it noted that the company included a reduction in drug costs of 20% associated with home administration, to reflect the VAT waiver for home care. It considered that VAT should have been excluded from all analyses, so this reduction was inappropriate.
4.33 The ERG presented exploratory analyses to explore the effects of key assumptions on the company's cost–consequence analysis and budget impact estimates. It noted that these scenarios substantially affected the model results. In the scenario in which elosulfase alfa was assumed to slow disease progression in 'multi‑domain responders' by 50% compared with established clinical management, elosulfase alfa was associated with a total of 19.79 QALYs, equating to 10.03 additional QALYs compared with conventional management. The ERG presented a combined scenario analysis in which elosulfase alfa slowed disease progression by 50% in 'multi‑domain responders' but had no effect on progression in 'single‑domain responders', the VAT waiver for home care was omitted, and the utility increment and mortality benefit associated with elosulfase alfa were removed. In this analysis, elosulfase alfa was associated with a total of 15.04 QALYs, equating to 5.04 additional QALYs compared with conventional management. For the budget impact analysis, the ERG presented a scenario based on the number of people that NHS England anticipates would be treated. In this analysis, it assumed that 38.5 people currently living with MPS IVa would be treated in the first year (that is, 50% of people who have the condition and wish to have treatment would get it in the first year); this was assumed to rise to 77 people from the prevalent population in the second and subsequent years, as well as 2.6 additional people diagnosed with MPS IVa each year. The results of this analysis were provided as commercial in confidence and so cannot be reported here.
4.34 Full details of all the evidence are in the submissions received for this evaluation, and in the ERG report, which are all available in the Committee papers.
4.35 Following the conclusion of the first Committee meeting, the Committee requested the company to provide additional economic analyses exploring:
plausible rates of disease progression in 'single‑ and multi‑domain responders', with appropriate sensitivity analyses
alternative estimates for the utility benefit (increment) associated with elosulfase alfa in each health state
exclusion of the VAT waiver for home care
alternative modelling of mortality to accurately reflect the mortality risks associated with mucopolysaccharidosis type IVa (excluding the double‑counting of the mortality benefit associated with elosulfase alfa)
budget impact of administering elosulfase alfa at people's homes.
4.36 The company maintained that, based on evidence from MOR‑005 at week 72 (see section 4.9) and the experience from using enzyme replacement therapy for other MPS disorders, elosulfase alfa would be expected to stop disease progression in 'multi‑domain responders' and reduce it by 50% in 'single‑domain responders'. It also presented results at week 120 from MOR‑005, which were not available at the time of the company's original submission. These showed that the benefits seen at week 72 were sustained up to week 120. To address the Committee's request, the company presented 2 additional analyses in which it varied the rates of disease progression across health states for multi‑domain and single‑domain 'responders' (rates were constant across health states in the base case). It stated that both analyses resulted in small changes to the base‑case incremental costs and QALYs.
4.37 The company explored an alternative approach to modelling the utility benefit associated with elosulfase alfa. It obtained from the literature a utility estimate for each additional benefit reported by patients but not captured by the 6MWT or FVC; namely, pain, visual acuity, sleep, wrist function and hearing. It used those estimates in 4 analyses:
'Optimistic scenario': the utility values for each benefit were added up to estimate the total utility benefit patients would receive (constant utility value across health states of 0.147).
'Conservative scenario': the utility benefit was estimated as a proportion (20–40%) of the total possible utility benefit (constant utility value across health states of 0.074).
The utility benefit increased as the patient moved to a worse health state (variable utility values for each health state ranging from 0.074 to 0.147).
The utility benefit decreased as the patient moved to a worse health state (variable utility values for each health state ranging from 0.147 to 0.074).
The incremental QALYs from the above analyses ranged from 17.11 to 19.96 compared with 18.18 in the base case (that is, the incremental QALYs changed between −1.07 and +1.78 compared with the base case).
4.38 The company presented an analysis showing the effect of excluding the VAT waiver for home care. The incremental costs increased by 20.7% compared with the base case.
4.39 The company modelled 2 alternative scenarios excluding the mortality benefit of elosulfase alfa. Firstly, it applied a constant mortality rate across all health states, equal to the mortality rate seen in the general population. Secondly, it applied a mortality rate that increased as patients moved through worsening health states; this was estimated based on clinical opinion. The incremental QALYs were 17.93 in the first scenario (decrease of 0.25 [−1.4%] compared with the base case) and 17.42 in the second scenario (decrease of 0.76 [−4.2%] compared with the base case). Incremental costs decreased by 0.2% and 3.6% respectively compared with the base case.
4.40 To assess the combined effect of the additional analyses (that is, the alternative modelling of disease progression, the utility benefit associated with elosulfase alfa, the VAT waiver for home care and mortality), the company presented a 'best case scenario' and a 'worst case scenario'. In these, it used the scenario that produced the best or worst QALYs respectively from each additional analysis. Compared with the base case, the incremental QALYs increased by 2.63 in the best case scenario and decreased by 3.23 in the worst case scenario. The incremental costs increased by 10.8% and 3.5% compared with the base case in the best and worst case scenarios respectively.
4.41 The company provided an analysis of the budget impact for the NHS and PSS of administering elosulfase alfa in people's homes. The base case assumed that 90% of patients would have infusions at home after an initial 3 months of having infusions in hospital. The company considered this to be conservative because all MPS VI patients who receive galsulfase receive it at home. It expected that all patients with MPS IVa will also have home care. The costs used in the analysis were actual costs incurred by patients currently receiving elosulfase alfa at home. The costs of providing elosulfase alfa at home were significantly greater than the tariff costs of providing it in hospital, but these differences were overwhelmed by the VAT that would be payable if the drug was supplied via the hospital, such that the difference in cost for all patients treated would be £3,069,111 less for home care in the first year increasing to £5,941,015 less for home care in the fifth year.
4.42 In response to the second evaluation consultation document, and facilitated by NICE, a managed access agreement was developed and submitted by a group of stakeholders comprising the company, NHS England, the MPS Society and a group of clinical experts.
4.43 The managed access agreement takes effect with publication of this guidance and will remain in force until a review of the guidance has been published or after 5 years (whichever is earlier). The company has agreed to provide the relevant data to inform the review of the guidance during the fourth year of the agreement. The managed access agreement states that, if NICE does not recommend elosulfase alfa for NHS funding when the guidance is reviewed, NHS England funding for elosulfase alfa will no longer be available for any patient. Those involved in the managed access agreement will ensure that any patient receiving elosulfase alfa whose treatment is funded by NHS England under this managed access agreement is made aware of these funding limitations and accepts them upon signing the patient agreement on the terms set out in the managed access agreement.