Reslizumab for treating eosinophilic asthma inadequately controlled by

Reslizumab for treating eosinophilic asthma inadequately controlled by inhaled corticosteroids 3 rd Appraisal Committee meeting Committee A Lead team: Jeremy Braybrooke, David Thomson Stephen Sharp ERG: Southampton Health Technology Assessments Centre NICE technical team: Sana Khan, Joanna Richardson Company: Teva 8 th June 2017 Slides for public only 1

Reslizumab Marketing authorisation Reslizumab is indicated as add-on therapy in adult patients with severe eosinophilic asthma inadequately controlled despite high-dose inhaled corticosteroids plus another medicinal product for maintenance treatment European marketing authorisation was granted in August 2016 Administration & dose Mechanism of action Intravenous infusion 3 mg/kg body weight once every 4 weeks Inhibits interleukin-5 which reduces eosinophil numbers and activity 2

History of appraisal • 1 st committee meeting 18 th October 2016 – ‘minded no’ recommendation requesting further analyses including: – the effect of reslizumab on exacerbations for subgroups of people with 3 or more or with 4 or more exacerbations in the previous year. These should not include an adjustment for a placebo effect. – appropriate administration costs, including the need to go to hospital for cannula insertion and supervised infusion – drug wastage using only the licensed 100 -mg vial – evaluation of response to treatment at periods that reflect clinical practice (such as 6 months from the start of treatment) – the committee recommends that the company also considers how reslizumab may affect oral corticosteroid usage and its consequent adverse effects and their costs. • 2 nd committee meeting 11 th January 2017 – Reslizumab is not recommended, within its marketing authorisation, for treating severe eosinophilic asthma inadequately controlled despite highdose inhaled corticosteroids plus another medicinal product for maintenance 3 treatment in adults

Reslizumab clinical studies Name Inclusion criteria Study 3082 Patients aged 12– 75 years with asthma and elevated blood eosinophils (≥ 400/µL) inadequately controlled with medium to high dose ICS Study 3083 ≥ 1 asthma exacerbation requiring corticosteroid use for at least 3 days in the 12 months prior to screening Intervention Comparator No. pts Duration 489 Reslizumab Placebo 3. 0 mg/kg 52 weeks 464 3082 and 3083 included patients aged 12 -75 years (although mean age from main trials was 44 -47 years) No UK centres for 3082 or 3083 Oral corticosteroid use: 19%(study 3082), 12% (study 3083) 4

CONFIDENTIAL Clinical effectiveness (1) XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX Number of exacerbations in the previous year ≥ 3 ≥ 4 n Exacerbation rate reduction [adjusted rates Placebo versus Reslizumab] 158 XXXXXX [XXXXXX] 94 XXXXXXX [XXXX

Clinical effectiveness (2) Transition probabilities • Transition probabilities for reslizumab and BSC were estimated based on patients who had experienced 3 exacerbations or more (N=158) in both active and placebo arm combined. Rate of exacerbation in the BSC arm • Exacerbation rate in placebo arm less than preceding year in trial. Committee discussed whether this could be due to placebo effect, better overall care in specialist setting of trial, or people not optimised before entering trial. • Company noted that the exacerbation rate went down, therefore ‘adjusted up’ the placebo rate of exacerbations to ‘expected’ rate and then calculated the number of exacerbations with reslizumab relative to that i. e. did not use the trial data for either arm directly • Committee preferred use of trial data directly, not adjusted to ‘real world’ levels • ERG indicated that there could be a placebo effect which was real in year one (duration of trial), but which gradually ‘wore off’ over 10 years 6

CONFIDENTIAL Key ICERs Scenario Company’s original base case (patients with ≥ 3 exacerbations in previous year) Combined effects of all amendments (company’s revised base case) Combined effects of all amendments except for no adjustment to real world rate of exacerbations (committee’s preferred base case) Subgroup of patients with ≥ 4 exacerbations in previous year, no adjustment to real world evidence (committee’s preferred) Total costs Reslizumab BSC Total QALYs Reslizumab BSC ICER XXXXXXX £ 24, 907 XXXXXXX £ 25, 408 XXXXXXX £ 43, 064 XXXXXX XXX £ 40, 715 XXXXXX

Committee conclusions (1) • Inadequately controlled severe eosinophilic asthma is associated with substantial morbidity – need for alternative treatment options. • In clinical practice, patients considered for this treatment may have lower eosinophil counts than in the trials and a higher percentage will be on oral corticosteroids. • A criterion based on the number of exacerbations is not unreasonable: the more frequent the exacerbations, the greater the clinical need. • Reslizumab may be considered for people who are not taking maintenance oral corticosteroids but it would be most beneficial for people who have multiple exacerbations despite maintenance oral corticosteroid use. • Reslizumab is effective in reducing the rate of clinically significant exacerbations compared with placebo. • Committee was unable to consider making a recommendation for reslizumab similar to NICE’S guidance for mepolizumab as no comparative evidence had been presented to committee and the evidence base for mepolizumab differed from reslizumab in terms of eosinophil count, number of exacerbations and oral steroid use. 8

Committee conclusions (2) • The adjustment to the exacerbation rate for the placebo arm requested by the committee (at its first meeting) was not transparent, and the relative treatment effect demonstrated in the trial had not been applied as requested. • Updated analysis demonstrate minimal difference in cost-effectiveness at 16 week and 6 month time points for reassessment. In addition, 16 week reassessments is used for other asthma drugs and it would be helpful to use this same reassessment time point for reslizumab. • The 25 -mg vial could be considered but any positive recommendation would only be based on the availability of this size of vial. • Some benefits related to avoiding the significant adverse effects of oral corticosteroid use had not been fully captured in the QALY calculations. 9

ACD consultation responses • Consultee comments from: – Company – Asthma UK – British Thoracic Society – Royal College of Physicians (endorsing BTS comments) • Commentator comments from: – Novartis 10

Key updates in this appraisal • The company has made 4 changes: – Removed the higher rates of exacerbation in the placebo arm – Introduced differential utility values in the severe exacerbation state for the placebo and reslizumab arm based on data from trials on duration of severe exacerbations – Enhanced PAS – Vial based dosing schedule including use of 25 mg vials to reduce wastage- SPC variation request • Company has also provided justification for 16 week evaluation period • No inclusion of oral corticosteroid sparing • Scenario analysis of ERG’s placebo effect ‘wearing off’ 11

Consultation comment – British Thoracic Society • Disagree that previous exacerbations are not a predictor of subsequent exacerbations as this is a strong predictor – Company should provide specific data on exacerbation reduction by looking at patients with frequent exacerbations NOT just in the previous year • Disagree that reslizumab is more appropriate for patients with nasal polyps or rhinitis than omalizumab • Evidence from studies such as MENSA and Heaney et al study support blood eosinophil count as a good biomarker for eosinophilic asthma. – If eosinophils have been suppressed by steroids, the only argument for using an anti-IL-5 treatment is to allow withdrawal of steroids (which are effective at reducing IL-5). • Reslizumab introduces extra administration costs compared to other monoclonal antibodies such as omalizumab – Dosing of reslizumab also needs to be simplified. • Acknowledge limited data on oral steroid withdrawal and urge company to complete its steroid reduction trials 12

Consultation comments - Asthma UK • Note that costs considered in appraisal do not consider the impact on quality of life and restrictions people experience due to ill-health resulting from oral corticosteroid use (OCS) • Highlighted recent study (Sweeney et al. 2016) reporting comorbidities resulting from severe asthma requiring OCS using data from the Optimum Patient Care Research Database and the British Thoracic Difficult Asthma Registry which could be used to demonstrate a steroid sparing effect. Other existing literature (such as Bel et al. 2014) that reports the steroid-sparing effects of monoclonal antibody also cited. • Final guidance issued by committee should make a recommendation to the NHS to support an independent programme of research that seeks to fully analyse the potential cost savings to the NHS resulting from reductions. 13

Consultation comments - Novartis • Consider 16 week assessment for reslizumab inappropriate. Omalizumab has a 16 week reassessment period based on robust evidence for response criteria from RCT – Omalizumab and reslizumab have different mechanism of actions and populations so inappropriate to consider assessment time-point – Clinical expert highlighted 16 week too early to assess response to reslizumab • Mepolizumab and reslizumab have similar mechanism of action and would be clinically relevant to consider similar assessment response time of 12 months • Recommendations should be based on 100 mg vial till regulatory approval for 25 mg received. 14

CONFIDENTIAL Company response Population – 3 or more exacerbation in the previous year Efficacy from pivotal trials (3082 and 3083) for patients who experienced ≥ 3 exacerbations in the year preceding enrolment in the trial over 16 and 52 weeks n 16 weeks Exacerbation rate reduction XXX Rate ratio (95%CI) XXXXXXX XX 52 weeks XXX XXXXXXX FEV 1 Gain ACQ-7 Gain AQLQ Gain [L] (95%CI) XXX XXXX XXXX XXX XXXXX

Company response Removal of upward adjustment in placebo arm No adjustment of exacerbation rates • As per committee preference, revised base case does not include adjustment to reflect the rate of exacerbations observed in clinical practice in the UK. Baseline exacerbation rate in placebo arm was matched to trial data. • Same transition probabilities that were submitted in the response to the first meeting consultation submitted but without adjustment used. – Based on these transition probabilities without adjustment, model uses mean annual exacerbation rate of 2. 68. Slightly lower compared to the mean rate of exacerbation of 2. 73 reported in BSC and placebo arms. Base case submitted in response to ACD 1 Revised base case predicted by the model – no adjustment Observed in the clinical trials at 52 weeks* Mean annual exacerbation rate 4. 85 2. 68 2. 73 *Within a subgroup of patients in two pivotal trials (studies 3082 & 3083) who experienced 3 or more exacerbations in the year preceding enrolment in the trial 16

CONFIDENTIAL Company response No adjustment approach conservative • Restricted population of adults who experience 3 or more exacerbations in the previous year likely to experience higher rate of exacerbation over subsequent years in clinical practice compared to those observed within a controlled setting of 52 -week clinical trials. • Using mean annual exacerbation rate of <3 (2. 68) is conservative estimate • Number of UK studies demonstrate severe asthma patients attending specialised centres experience high level of exacerbations: Study Evidence reported Gibeon et al. Report higher mean the 2. 68 exacerbations in base case XXXXXXXXXXbased on an analysis of the Portsmouth cohort Baseline risk of exacerbations in the updated base case analysis submitted in response to ACD 1 uses mean exacerbation rate of XXX per year from this analysis XXXXXXXXXXXXXXXXXXXc cohort study of XXX severe asthma patients Patients with eosinophil levels of 400 or more had a mean number of OCS exacerbation, emergency room visits and hospital admissions numerically higher than in patients with less than 400 eosinophils (XXXXXXXXXXXXXXXXXX, respectively).

CONFIDENTIAL Company response Further evidence to support higher rate of exacerbation • New evidence available to address committee concern that only XXXX of the XXXX cohort presented with severe eosinophilic asthma and if mean rate of exacerbations of XXXX was applicable to these patients. • Considering all patients with 1 or more exacerbation in the previous year, data show that patients with severe eosinophilic asthma experienced on average XXX exacerbations per year compared to XXX for noneosinophilic patients. – additional results submitted for presentation at the upcoming ERS presentation (XXXXX). • This expected rate in clinical practice explored in scenario analysis in the model based on approach by ERG: rate of exacerbations observed in the BSC + Placebo IV arm of clinical trials assumed to apply to BSC for duration of trial (i. e. one year) and assumed to increase linearly over following nine years until reaching the exacerbation rate of the ‘real world’ data from WSAC (i. e. XXX exacerbations per year) at year 10. • Note: this approach lowers ICER

CONFIDENTIAL Company response Utility values for severe exacerbations Updated utility for severe exacerbation based on shorter duration data • Post-hoc analyses from pivotal trials show that for patients with severe eosinophilic asthma and 3 or more exacerbations in the previous year, mean length of a severe exacerbation is XXX days for the reslizumab arm versus XXX days for patients on placebo XXXXXXX (reflecting total of XX versus XXX severe exacerbations respectively) – may indicate patients on reslizumab recover quicker from severe exacerbation than patients on BSC + placebo IV with positive impact on their quality of life in these additional XX days • In revised base case analysis, mean utility associated with the severe exacerbation health state was estimated by treatment to reflect the impact of different length of exacerbation (compared to assuming to last for same duration regardless of treatment arm) Reslizumab BSC + Placebo IV Uncontrolled asthma 0. 728 Controlled asthma 0. 920 Moderate exacerbation 0. 570 Severe exacerbation 0. 54 0. 50

ERG response (1) Utility values for severe exacerbations • Company provided new data from studies 3082 and 3083 on the duration of severe exacerbations in patients who had severe eosinophilic asthma and 3 or more exacerbations in the previous year to determine updated utility values. • ERG highlighted the following limitations to this exacerbation data: – Data may not be reflective of the responses of patients who have lower eosinophil counts and a need for oral corticosteroids (section 4. 4 in the 2 nd ACD) – ERG unable to verify the data reported by company on severe exacerbation durations as it is new (not available in the company’s submission or clinical study reports) – ERG was unable to find any comparable data on severe exacerbation durations experienced in clinical practice against which to compare the company’s data. Effect of reslizumab in reducing duration of severe exacerbations is clinically plausible, but there is uncertainty as to how closely the company’s data would match ‘real world’ severe exacerbation durations and the variability associated with them. 20

ERG response (2) Utility values for severe exacerbations • New analysis provides specific utility values for each comparator which are weighted according to duration of the severe exacerbations to account for the fact that severe exacerbations do not last for the full model cycle. • In the company’s original analysis a single utility value was used for both resilizumab and best supportive care and this was applied to the duration of the full model cycle (4 weeks). • Overall mean utility for severe exacerbation in each model cycle is calculated from the weighted utility for the time with severe exacerbation plus the weighted utility for the exacerbation-free (‘uncontrolled’ utility) remainder time of the model cycle. – ERG considers the calculation to be appropriate. – Uncertainty in utility value estimates for severe exacerbations highlighted due to the lack of robust health-related quality of life data – changing the utility values in this way reduces the ICER by about £ 1000. 21

Vial based dosing Company response: • Favourable and early regulatory approval by EMA demonstrated by company for 25 mg vial and change in SPC to incorporate vial-based dosing by bodyweight. VBD accepted in CHMP Rapporteur Assessment Report. • VBD will simplify process of determining dose and reduce time needed to prepare reslizumab. Wastage will be minimised and total cost of treatment reduced as patients in each dosing group will receive a dose marginally lower than the 3 mg/kg weight-based dosing. – Modelling and simulation approach estimated predicted drug exposures and simulated clinical responses on VBD will be comparable to weight-based dosing while maintaining the same efficacy. ERG response • If 25 -mg vials are made available (and are acceptable to clinicians), use of vial-based dosing would be reasonable. 22

CONFIDENTIAL Revised patient access scheme Company response: • Revised PAS submitted by company to include both 100 -mg and 25 mg vials. • Inclusion of vial-based dosing in the revised base case together with the revised PAS, reduces the total average cost per patient by XXX from XXXXXX to £XXXXXX at an annual cost of £ XXXXXX less per patient compared with the initial submission. ERG response: • Amendments to analyses checked and results replicated. ERG consider the company’s amendments and their results presented are reasonable, given the limitations of the available data.

CONFIDENTIAL Company’s revised base case Total costs Total QALYs ICER Scenario Reslizumab BSC Committee’s preferred base case (no adjustment on exacerbation rate) XXXXXXX £ 43, 064 Utility adjustment for severe exacerbation XXXXXXX £ 42, 025 Vial-based dosing XXXXXXX XXXXX Revised PAS XXXXXXX XXXXX Company’s revised deterministic base case with all amendments XXXXXXX £ 29, 870 Company’s revised probabilistic base case with all amendments - - £ 27, 509

Company’s Deterministic Sensitivity Analysis 25

CONFIDENTIAL Company’s Scenario Analysis Rate of exacerbations to increase linearly from year 2 to real world estimate of 4. 85 exacerbations in year 10 ICER Costs QALYs £/QALY BSC + Reslizumab placebo IV Company’s revised base case in XXXXXXX response to ACD 2 XXXX £ 29, 870 Increase in rate of XXXXXXX exacerbation XXXX £ 17, 748 Oral corticosteroid reduction associated with the initiation of omalizumab resulted in a decrease in the reported base case ICER by between £ 4 -£ 6 K per QALY gained.

Key issues for consideration • Does the committee accept the changes made by the company? That is: – use of lower baseline rate of exacerbations in the placebo arm than in original submission – differential utilities for severe exacerbations – vial based dosing • Does the committee consider that the evidence for reslizumab allows for a positive recommendation corresponding with NICE’s recommendation for mepolizumab? • Are there any additional issues for this appraisal e. g. OCS sparing that need to be considered? 27

Additional slides 28

Mepolizumab FAD 1. 1 Mepolizumab, as an add-on to optimised standard therapy, is recommended as an option for treating severe refractory eosinophilic asthma in adults, only if: • The blood eosinophil count is 300 cells/microlitre or more in the previous 12 months • The person has agreed to and followed the optimised standard treatment plan and: – The person has had 4 or more asthma exacerbations needing systemic corticosteroids in the previous 12 months or – The person has had continuous oral corticosteroids of at least the equivalent of prednisolone 5 mg per day over the previous 6 months […] Stopping rule: 1. 2 At 12 months of treatment: • stop mepolizumab if the asthma has not responded adequately or • continue treatment if the asthma has responded adequately and assess response each year (adequate response defined in FAD) 29

Mepolizumab FAD evidence base • Data mainly from 3 RCTs in patients with: – Severe refractory eosinophilic asthma on high-dose oral corticosteroids and a history of 2 or more exacerbations in the previous 12 months. – Blood eosinophil level of either 300 cells/microlitre or more in the 12 months before screening or 150 cells/microlitre or more at screening. – One trial (SIRIUS, n=135) included people with asthma who needed regular treatment with maintenance systemic (oral or injectable) corticosteroids and high-dose inhaled corticosteroids. • Clinical experts agreed that: – Population based on a blood eosinophil count of 300 cells/microlitre or more in the previous 12 months would be relevant to clinical practice. – Criterion based on 4 or more exacerbations per year would identify the most severe patient group which would gain the most benefit. – Agreed that population should be defined as in the SIRIUS trial, that is, having continuous OCS of at least the equivalent of prednisolone 5 mg/day in the 6 months before the start of treatment. 30