==> picture [215 x 22] intentionally omitted <==

Single Technology Appraisal

Mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] Committee Papers

==> picture [215 x 22] intentionally omitted <==

NATIONAL INSTITUTE FOR HEALTH AND CARE EXCELLENCE

SINGLE TECHNOLOGY APPRAISAL

Mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

Contents:

The following documents are made available to stakeholders:

Access the final scope and final stakeholder list on the NICE website.

1. Company submission from Eli Lilly:

• a. Full submission

• b. Summary of Information for Patients (SIP)

2. Addendum to the company submission

3. Clarification questions and company responses

4. Patient group, professional group, and NHS organisation submission from:

• a. a. Crohn’s & Colitis UK

5. External Assessment Report prepared by Southampton Health Technology Assessments Centre

6. External Assessment Group response to factual accuracy check of EAR

Any information supplied to NICE which has been marked as confidential, has been redacted. All personal information has also been redacted.

NATIONAL INSTITUTE FOR HEALTH AND CARE EXCELLENCE

Single technology appraisal: cost-comparison

Mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

Document B

Company evidence submission

December 2022

Company budget impact analysis submission for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly and Company (2022). All rights reserved

Instructions for companies

This is the template for submission of evidence to the National Institute for Health and Care Excellence (NICE) when a cost-comparison case is made as part of the single technology appraisal process. Please note that the information requirements for submissions are summarised in this template; full details of the requirements for pharmaceuticals and devices are in the user guide.

This submission must not be longer than 100 pages, excluding appendices and the pages covered by this template. If it is too long it will not be accepted.

Companies making evidence submissions to NICE should also refer to the NICE health technology evaluation guidance development manual.

In this template any information that should be provided in an appendix is listed in a box.

Highlighting in the template (excluding the contents list)

Square brackets and xxxx highlighting are used in this template to indicate text that should be replaced with your own text or deleted. These are set up as form fields, so to replace the prompt text in xxxxxxxxxxxxxxxxxxx with your own text, click anywhere within the highlighted text and type. Your text will overwrite the highlighted section.

To delete grey highlighted text, click anywhere within the text and press DELETE.

Grey highlighted text in the footer does not work as an automatic form field, but serves the same purpose – as prompt text to show where you need to fill in relevant details. Replace the text highlighted in xxxxxx in the footer with appropriate text. (To change the footer, double click over the footer text. Double click back in the main body text when you have finished.)

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Contents

Contents ........................................................................................................................................ 3 List of Tables ................................................................................................................................. 4 List of Figures ................................................................................................................................ 5 B.1 Decision problem, description of the technology and clinical care pathway ............................ 7 Decision problem ............................................................................................................. 7 Description of the technology being evaluated .............................................................. 13 Health condition and position of the technology in the treatment pathway .................... 15 Disease overview .................................................................................................... 15 Burden of disease ................................................................................................... 17 Clinical pathway of care .......................................................................................... 18 Equality considerations .................................................................................................. 21 B.2 Key drivers of the cost effectiveness of the comparator(s) ................................................... 22 Clinical outcomes and measures ................................................................................... 22 Resource use assumptions ............................................................................................ 39 B.3 Clinical effectiveness ............................................................................................................ 41 Identification and selection of relevant studies ............................................................... 42 List of relevant clinical effectiveness evidence ............................................................... 42 Summary of methodology of the relevant clinical effectiveness evidence...................... 43 Trial design and methodology ................................................................................. 43 Baseline characteristics .......................................................................................... 50 Outcome definitions ................................................................................................ 55 Statistical analysis and definition of study groups in the relevant clinical effectiveness evidence .................................................................................................................................. 56 Trial populations ..................................................................................................... 56 Patient disposition ................................................................................................... 58 Statistical methods .................................................................................................. 58 Critical appraisal of the relevant clinical effectiveness evidence .................................... 62 Clinical effectiveness results of the relevant studies ...................................................... 62 LUCENT-1 .............................................................................................................. 63 LUCENT-2 .............................................................................................................. 68 Subgroup analysis ......................................................................................................... 76 Meta-analysis ................................................................................................................. 76 Indirect and mixed treatment comparisons .................................................................... 77 Identification and selection of relevant studies ....................................................... 77 Feasibility assessment ............................................................................................ 77 Methodology ........................................................................................................... 81 Results .................................................................................................................... 83 Uncertainties in the indirect and mixed treatment comparisons ............................ 102 Conclusions .......................................................................................................... 103 Adverse reactions ...................................................................................................... 103 Overview of adverse events ............................................................................... 104 Treatment-emergent adverse events .................................................................. 104 Serious adverse events ...................................................................................... 105 Adverse events of special interest ...................................................................... 107 Discontinuations due to adverse events ............................................................. 107 Conclusions about comparable health benefits and safety ........................................ 108 Ongoing studies ......................................................................................................... 110 B.4 Cost-comparison analysis ................................................................................................... 111 Changes in service provision and management .......................................................... 111 Cost-comparison analysis inputs and assumptions ..................................................... 111 Features of the cost-comparison analysis ............................................................ 111 Intervention and comparators’ acquisition costs ................................................... 117 Intervention and comparators’ healthcare resource use and associated costs ..... 121

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Adverse reaction unit costs and resource use ...................................................... 122 Miscellaneous unit costs and resource use .......................................................... 122 Model validation .................................................................................................... 122 Uncertainties in the inputs and assumptions ........................................................ 122 Base-case results ........................................................................................................ 123 Sensitivity and scenario analyses ................................................................................ 125 One-way sensitivity analysis ................................................................................. 125 Scenario analyses ................................................................................................ 128 Subgroup analysis ....................................................................................................... 130 Interpretation and conclusions of economic evidence ................................................. 130 B.5 References ......................................................................................................................... 131

List of Tables

Table 1: The decision problem ...................................................................................................... 9 Table 2: Technology being appraised ......................................................................................... 13 Table 3: The Mayo score for ulcerative colitis ............................................................................. 16 Table 4: Clinical outcomes and measures appraised in published NICE guidance for the comparators included in the NICE final scope ............................................................................. 24 Table 5: Clinical effectiveness evidence...................................................................................... 42 Table 6: Summary of LUCENT-1 trial methodology .................................................................... 45 Table 7: Summary of LUCENT-2 trial methodology .................................................................... 48 Table 8: Baseline demographic characteristics for patients in the mITT population of the LUCENT-1 trial ............................................................................................................................ 50 Table 9: Baseline disease characteristics and prior therapies of patients in the mITT population of the LUCENT-1 trial ...................................................................................................................... 51 Table 10: Baseline demographic characteristics for patients in the mITT population of the LUCENT-2 trial ............................................................................................................................ 52 Table 11: Baseline disease characteristics and prior therapies of patients in the mITT population of the LUCENT-2 trial .................................................................................................................. 53 Table 12: Definitions of clinical effectiveness outcomes used in the LUCENT-1 and LUCENT-2 trials ............................................................................................................................................ 55 Table 13: Trial populations used for the analysis of outcomes in the LUCENT-1 and LUCENT-2 trials ............................................................................................................................................ 56 Table 14: Analysis cohorts used in LUCENT-2, dependent on clinical response at Week 12 of LUCENT-1 ................................................................................................................................... 57 Table 15: Summary of the statistical methods employed in the LUCENT-1 and LUCENT-2 trials .................................................................................................................................................... 58 Table 16: Quality assessment results for the LUCENT-1 and LUCENT-2 trials .......................... 62 Table 17: Alternative clinical remission rates at Week 12 (NRI) .................................................. 64 Table 18: Endoscopic remission rates at Week 12 (NRI) ............................................................ 65 Table 19: Symptomatic remission rates at Week 12 (NRI) ......................................................... 66 Table 20: Bowel movement urgency NRS change from baseline at Week 12 (MMRM) ............. 66 Table 21: Alternate clinical remission rates at Week 40 in a randomised withdrawal maintenance period .......................................................................................................................................... 69 Table 22: Endoscopic remission rates at Week 40 (NRI) in a randomised withdrawal maintenance period ..................................................................................................................... 70 Table 23: Corticosteroid-free remission without surgery rates at Week 40 (NRI) in a randomised withdrawal maintenance period ................................................................................................... 71 Table 24: Bowel movement urgency NRS improvement (change from induction baseline) at Week 40 ...................................................................................................................................... 72 Table 25: Urgency remission (urgency NRS=0 or 1) at Week 40 (NRI) in a randomised withdrawal maintenance period in patients with urgency NRS ≥3 at induction baseline ............. 73 Table 26: Symptomatic response rates through the re-induction (NRI) ....................................... 74

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Table 27: Clinical remission at 12 Weeks (NRI) in an open-label extended induction period (mITT population) .................................................................................................................................. 75 Table 28: Clinical response rates at 12 Weeks (NRI) in an open-label extended induction period (mITT population) ........................................................................................................................ 76 Table 29: Endoscopic remission at 12 Weeks (NRI) in an open-label extended induction period (mITT population) ........................................................................................................................ 76 Table 30: NMA models used in the base case analysis .............................................................. 81 Table 31: Overview of adverse events (LUCENT-2) ................................................................. 104 Table 32: TEAEs occurring in ≥2% of patients in the mirikizumab induction responders cohort of LUCENT-2 ................................................................................................................................. 104 Table 33: SAEs in patients in the mirikizumab induction responders cohort of LUCENT-2 ....... 105 Table 34: Treatment-emergent serious infection events experienced by patients in LUCENT-2 .................................................................................................................................................. 106 Table 35: AESIs in patients in the mirikizumab induction responders cohort of LUCENT-2 ...... 107 Table 36: Discontinuations due to adverse events in patients in the mirikizumab induction responders cohort of LUCENT-2 ............................................................................................... 108 Table 37: Clinical response and response at the end of induction ............................................ 115 Table 38: Clinical response at the end of the delayed response assessment period ................ 115 Table 39: Patient baseline characteristics ................................................................................. 116 Table 40: Acquisition costs of the intervention and comparator technologies ........................... 118 Table 41: Drug acquisition costs for mirikizumab and the comparators during the induction phase .................................................................................................................................................. 120 Table 42: Drug acquisition costs for mirikizumab and the comparators during the maintenance phase ........................................................................................................................................ 120 Table 43: Unit cost of treatment administration for IV therapies ................................................ 121 Table 44: Drug administrations for mirikizumab and the comparators during the induction and maintenance phase ................................................................................................................... 121 Table 45: Settings and values used in the base case analysis ................................................. 122 Table 46: Key model assumptions ............................................................................................ 123 Table 47: Base case results for a 10-year time horizon at mirikizumab list price (biologic-naïve population) ................................................................................................................................ 124 Table 48: Base case results for a 10-year time horizon at mirikizumab list price (biologic- failed population) ................................................................................................................................ 124 Table 49: Summary of one-way sensitivity analyses ................................................................. 125 Table 50: Scenario analyses ..................................................................................................... 128 Table 51: Scenario analysis for a 10-year time horizon in the biologic-naïve population .......... 129 Table 52: Scenario analysis for a 10-year time horizon in the biologic-failed population .......... 129

List of Figures

Figure 13: Histologic-endoscopic mucosal remission rates at Week 40 (NRI) in a randomised withdrawal maintenance period ................................................................................................... 72 Figure 14: Bowel urgency improvement by treatment week in induction responders (MMRM, full mirikizumab induction responder population) .............................................................................. 73 Figure 15: Symptomatic remission rates through the re-induction period (NRI) .......................... 75 Figure 16: All pairwise odds ratios with random treatment effects and a half-Normal (location: 0, scale: 5) prior for the between-trial standard deviation: clinical response, induction, biologic-naïve population .................................................................................................................................... 84 Figure 17: All pairwise odds ratios with random treatment effects and a half-Normal (location: 0, scale: 5) prior for the between-trial standard deviation: clinical remission, induction, biologicnaïve population .......................................................................................................................... 85 Figure 18: Odds ratios with random treatment effects with baseline risk meta-regression and a half-Normal (location: 0, scale: 5) prior for the between-trial standard deviation: mucosal healing, induction, biologic-naïve population ............................................................................................ 86 Figure 19: Odds ratios with fixed treatment effects with baseline risk meta-regression: clinical response, maintenance, biologic-naïve population ..................................................................... 88 Figure 20: Odds ratios with fixed treatment effects with baseline risk meta-regression: clinical remission, maintenance, biologic-naïve population ..................................................................... 89 Figure 21: All pairwise odds ratios with fixed treatment effects with baseline risk meta-regression: mucosal healing, maintenance, biologic naïve population .......................................................... 91 Figure 22: All pairwise odds ratios with fixed treatment effects (response and response with remission): clinical response and remission, induction, biologic-failed population ...................... 93 Figure 23: All pairwise odds ratios with fixed treatment effects: mucosal healing, induction, biologic-failed population ............................................................................................................. 94 Figure 24: Odds ratios with fixed treatment effects with baseline risk meta-regression: clinical response, maintenance, biologic-failed population ...................................................................... 96 Figure 25: Odds ratios with fixed treatment effects with baseline risk meta-regression: clinical remission, maintenance, biologic-failed population. .................................................................... 97 Figure 26: All pairwise odds ratios with fixed treatment effects with baseline risk meta-regression: mucosal healing, maintenance, biologic-failed population .......................................................... 99 Figure 27: All pairwise odds ratios with random treatment effects and a half-Normal (location: 0, scale: 5) prior for the between-trial standard deviation: all cause discontinuation, induction, mixed population .................................................................................................................................. 101 Figure 28: All pairwise odds ratios with random treatment effects and a half-Normal (location: 0, scale: 5) prior for the between-trial standard deviation: serious adverse events, induction, mixed population .................................................................................................................................. 102 Figure 29: Decision tree schematic for the induction phase ...................................................... 113 Figure 30: Markov structure for the maintenance phase ........................................................... 114 Figure 31: Tornado plot with results from the one-way sensitivity analysis – mirikizumab (PAS price) versus ustekinumab in the biologic-naïve population ...................................................... 126 Figure 32: Tornado plot with results from the one-way sensitivity analysis – mirikizumab (PAS price) versus ustekinumab in the biologic-failed population ...................................................... 126 Figure 33: Tornado plot with results from the one-way sensitivity analysis – mirikizumab (PAS price) versus vedolizumab (IV) in the biologic-naïve population ................................................ 127 Figure 34: Tornado plot with results from the one-way sensitivity analysis – mirikizumab (PAS price) versus vedolizumab (IV) in the biologic-failed population ................................................ 127 Figure 35: Tornado plot with results from the one-way sensitivity analysis – mirikizumab (PAS price) versus vedolizumab (IV/SC) in the biologic-naïve population .......................................... 128 Figure 36: Tornado plot with results from the one-way sensitivity analysis – mirikizumab (PAS price) versus vedolizumab (IV/SC) in the biologic-failed population .......................................... 128

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

B.1 Decision problem, description of the technology and clinical care pathway

==> picture [32 x 11] intentionally omitted <==

Decision problem

Population

Mirikizumab is anticipated to receive a marketing authorisation for the treatment of

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx.[1] Mirikizumab is positioned for use as an alternative to ustekinumab (Stelara[®] ) and vedolizumab (Entyvio[®] ), which have received the following positive recommendations for reimbursement by NICE:

• Ustekinumab: as an option for treating moderately to severely active ulcerative colitis in adults when conventional therapy or a biological agent cannot be tolerated, or the disease has responded inadequately or lost response to treatment, only if:

  • A tumour necrosis factor-alpha inhibitor has failed (that is the disease has responded inadequately or has lost response to treatment), or

  • A tumour necrosis factor-alpha inhibitor cannot be tolerated or is not suitable

• Vedolizumab: as an option for treating moderately to severely active ulcerative colitis in adults

In alignment with the anticipated use of mirikizumab in UK clinical practice and with the patient populations in which ustekinumab is recommended by NICE, this submission focusses on a subpopulation of the full anticipated marketing authorisation of mirikizumab. This population is defined as adult patients with moderately to severely active ulcerative colitis for whom:

• Conventional treatment cannot be tolerated or is not working well enough and other biologic treatment is not suitable (“biologic-naïve”), or

• Biological treatment cannot be tolerated or is not working well enough (“biologic-failed”)

Throughout the submission, the term “biologic-failed” encompasses prior failure with anti-TNF antibodies, anti-integrin antibodies or tofacitinib. This is in line with the definition of “biologicfailed” patients within the pivotal LUCENT trials, as outlined in Section B.3.4.1.

Comparator

The manufacturer is proposing that the appraisal of mirikizumab be considered under the NICE proportional approach to technology appraisals sub-process, making use of a cost comparison economic evaluation. The NICE user guide states that a cost-comparison case can be made if a health technology is likely to provide similar or greater health benefits at similar or lower cost than technologies already recommended in published technology appraisal guidance for the same indication.[2] An indirect treatment comparison was conducted to estimate the relative efficacy of mirikizumab against the full range of comparators specified in the final scope (see Section B.3.9.6) with the results of this analysis indicating that that mirikizumab has similar or better efficacy than ustekinumab and vedolizumab in the induction and maintenance phases.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Additionally, the criteria for the selection of an appropriate comparator state that the selected comparator must fulfil the following:

• It adequately represents the NICE recommended treatments as a whole both in terms of costs and effects.

• It has significant market share.

• It is recommended in published NICE technology appraisal guidance for the same indication.

• As such, ustekinumab and vedolizumab are deemed to be the most appropriate comparators for this appraisal.

The decision problem addressed by this submission, and as compared with the decision problem defined in the final NICE scope, is summarised in Table 1.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Table 1: The decision problem

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 9 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 10 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 11 of 136

Abbreviations: AE: adverse event; ERG: evidence review group; EQ-5D-5L: European Quality of Life 5 Dimensions 5 Level; HRQoL: health-related quality of life; IBD-Q: Inflammatory Bowel Disease-Questionnaire; JAK: Janus kinase; N/A: not applicable; NICE: National Institute for Health and Care Excellence; NRS: numeric rating scale; PAS: patient access scheme; PSS: Personal and Social Services; TNF: tumour necrosis factor; UC: ulcerative colitis.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 12 of 136

==> picture [33 x 12] intentionally omitted <==

Description of the technology being evaluated

A description of the technology being appraised, mirikizumab, is presented in Table 2.

Table 2: Technology being appraised

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Abbreviations: EMA: European Medicine Agency; IgG4: immunoglobulin G4; IL: interleukin; IV: intravenous; MHRA: Medicines and Healthcare products Regulatory Agency; PASLU: Patient Access Schemes Liaison Unit; SC: subcutaneous; SmPC: summary of product characteristics.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Health condition and position of the technology in the treatment pathway

==> picture [43 x 11] intentionally omitted <==

Disease overview

Disease background

Ulcerative colitis (UC) is a chronic inflammatory bowel disease (IBD) characterised by continuous mucosal inflammation of the colon, starting at the rectum and extending proximally.[13] The severity of inflammation varies and ulceration, oedema and haemorrhaging may be present along the length of the colon.[14] The clinical course of UC is unpredictable and consists of periods of exacerbation (flares) and remission; during flares, common symptoms include rectal bleeding, diarrhoea, abdominal pain, bowel urgency and tenesmus.[13, 15, 16]

In addition to the primary symptoms of UC, between one-third and half of patients experience a range of complications and manifestations outside of the gastrointestinal (GI) tract, known as extraintestinal manifestations (EIM).[13, 17] The risk of developing an EIM increases with duration of the disease and if a patient already has an EIM.[17] EIMs can affect multiple organ systems, including the skin, eyes, liver or spine and may occur as a result of malabsorption, chronic inflammation, medication or genetic risk factors.[18] The most common EIM is anaemia, affecting approximately 21% of patients, followed by arthropathy (peripheral and axial) in 20% of patients.[19]

Epidemiology and diagnosis

The exact cause of the inflammation underlying UC is unclear, but factors related to genetics, the environment and the gastrointestinal microbiota have been identified as potential risk factors, with a family history of the disease representing the most common risk factor.[20] It has been estimated that 146,000 people in the UK are affected by UC, equivalent of approximately 1 in 420 people.[21] Additionally, worldwide prevalence of UC is increasing: a large study conducted across seven major countries, including the UK and US, predicted that the prevalence may rise by a further 9% in 10 years from 1.8 million in 2017 to 1.9 million in 2027.[22] This is supported by data from the Health Improvement Network (THIN) on the prevalence of IBD, which showed an increase in point prevalence from 390 per 100,000 in 200 to 570 per 100,000 in 2017, an approximate increase of 2.5% annually.[23]

UC affects both sexes relatively equally, as evidenced by the UK IBD registry which found that 48% of patients with UC in the UK were female, and by a large UK observational study in which the adjusted incidence rate ratio for females in UC was 0.87 (95% CI: 0.84, 0.90).[24, 25] Although it can present at any age, UC primarily presents during early adulthood, with a secondary peak for presentations observed after the age of 50 in some populations.[19, 24] Diagnosis of UC is typically made once other potential causes of symptoms, such as pathogens, vascular and medicationrelated causes, have been excluded.[19] Guidelines from the European Crohn’s and Colitis Organisation (ECCO) state that no ‘gold standard’ for diagnosis exists and instead recommend that UC is diagnosed using clinical, laboratory and endoscopic parameters, including histopathology.[19]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Disease staging

Severity

Advancing disease severity is associated with a worsening of symptoms, with more severe disease typically leading to an increase in bloody diarrhoea and the presence of systemic symptoms such as weight loss, fever, nausea and vomiting.[13, 19] The severity of UC is typically categorised as mild, moderate or severe and is generally assessed using criteria such as stool frequency (with or without blood) and a combination of endoscopic and histological assessments and the presence of systemic symptoms.[13, 19] The ECCO and the British Society of Gastroenterology (BSG) outline several indexes by which disease severity in UC may be assessed and categorised.[19, 26, 27] A well-established scoring system is the Mayo score, which has been recommended by the BSG for use in clinical practice as a composite clinical and endoscopic tool, and which has been implemented in several clinical trials.[16, 26-28] The score consists of four variables, each scored from 0–3 and summed to a maximum of 12; a higher score corresponds to more severe disease as follows:[29]

• Remission: ≤2, with no individual subscore >1

• Mild: 3–5

• Moderate: 6–10

• Severe 11–12

Table 3: The Mayo score for ulcerative colitis

Source: Lamb et al ., (2019).[26]

Extent

In addition to severity, UC can be stratified depending on the extent to which the colon is involved. This is typically separated into three categories:[30]

• Proctitis: involvement limited to the rectum (extent of inflammation is distal to the rectosigmoid junction).

• Left-sided colitis (distal UC): involvement of the rectum, sigmoid colon and descending colon.

• Extensive (pancolitis): involvement of the left colon and some or all of the colon proximal to the splenic flexure.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

A review of population-based studies estimated that, upon presentation, 30–60% of patients have proctitis, 16–45% have left-sided colitis and 14–35% have pancolitis.[31] A recent systematic review of studies in patients with UC found that 28–30% of patients with proctitis at diagnosis progressed to left-sided colitis and 14–16% to pancolitis; the rate of progression from left-sided colitis to pancolitis was 21–34%.[32] The severity and extent of disease affects treatment decisions, prognosis and patient outcomes, with more extensive disease at diagnosis more likely to result in surgical interventions such as removal of all or part of the colon (colectomy) or hospitalisation.[19, 26]

==> picture [45 x 12] intentionally omitted <==

Burden of disease

Impact on health-related quality of life

The chronic, lifelong and progressive nature of UC results in a significant negative impact to patients’ health-related quality of life (HRQoL).[33] This impact is multi-faceted and extends across patients’ psychological wellbeing, education and employment, social life and daily activities.[34] In particular, the physical symptoms of the disease can result in substantial disability and significantly impacted daily life, including the ability to carry out ordinary tasks.[34, 35] Of note, patients have reported bowel urgency, stool frequency and rectal bleeding as the most bothersome symptoms of UC.[36] The additional symptoms of abdominal pain and fatigue also affect a significant number of patients, impacting their quality of life.[37, 38]

Approximately 40% of patients with UC will experience a relapse/flare each year, during which the physical symptoms are exacerbated, directly affecting HRQoL.[39, 40] In contrast to those experiencing symptoms and disease flares, patients in remission have improved HRQoL and reduced work impairment, highlighting the importance of adequate disease control. In addition to the physical burden of the disease, patients with UC experience increased incidence of anxiety and depression as compared with matched controls.[41]

Comorbidities

Aside from symptoms directly related to UC itself, the condition is associated with considerable comorbidities, with nearly half (45.6%) of patients with UC presenting with comorbidities at baseline in a recent global study.[42] The most commonly reported comorbidities were fatigue (27.7%), anxiety/depression (24.8%), sleep disorders (20.6%) and cardiac abnormalities/cardiovascular disease.[42] These findings are supported by the results of a realworld evidence (RWE) study of 208 UC patients in the UK between 2020–2021.[43] Through this disease specific programme (DSP) in UC, it was identified that xxx of patients suffered from comorbidities, xxx of which were autoimmune conditions.[43] The most common autoimmune comorbidity was psoriasis (xxx), but patients also reported axial spondylarthritis (xxx), psoriatic arthritis (xxx), rheumatoid arthritis (xxx) and coeliac disease (xxx).[43]

As compared with people without UC, patients with UC are also at higher risk of developing issues associated with malabsorption, as well as diabetes (5%), hypertension (11%), and irritable bowel syndrome (IBS) (10%).[43, 44] An increased risk of non-alcoholic liver disease and hepatic fibrosis, resulting from increased gut permeability, altered microbiome and chronic inflammation has also been shown to affect patients with UC.[45]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Economic burden

In addition to the considerable physical and psychological burdens for patients with UC, it is associated with an economic burden due to the increased healthcare resource utilisation required for disease management, including the potential for surgery and treatment of EIMs and flares.[46] European patients with all severities of active UC have been demonstrated to have a high cost burden, with higher costs found to be associated with more severe disease states.[47]

==> picture [45 x 11] intentionally omitted <==

Clinical pathway of care

In the UK, moderately to severely active UC is treated in a stepwise manner based on factors such as disease severity (mild, moderate, severe), prior medication response, relapse frequency/remission status, and patient suitability for available treatments, as recommended by NICE guidance (NG130).[48] The BSG guidelines recommend that the primary treatment goal for UC should be symptomatic remission combined with mucosal healing, although treatment decisions are made based on a variety of factors, meaning that there is no single pathway of care adopted by all clinicians and patients.[26]

The current treatment pathway for patients with moderately to severely active UC in the UK is shown in Figure 2. Initially, patients receive conventional therapies, such as aminosalicylates, corticosteroids, and thiopurines. However, the effectiveness of these conventional therapies varies in between patients, with a systematic review of conventional therapies in IBD finding that some show no statistical benefit over placebo.[49] In addition, during the COVID-19 pandemic, weaning of corticosteroids for patients with IBD was recommended due to the observance that their use may increase risk of adverse outcomes of COVID-19, whereas biologics such as IL12/23 inhibitors (ustekinumab) and TNF alpha inhibitors (TNFis), such as adalimumab and golimumab, were found to be safe to continue, and this may have accelerated patients through the conventional therapy stage.[50]

Patients who have had an inadequate response to, lose response to, or are contraindicated or otherwise unsuitable to receive conventional therapies will switch treatments, with the initiation of biologic treatments representing the next line of treatment.[51] TNFis such as infliximab, adalimumab and golimumab are considered as the first line biologic treatment for many patients, particularly since biosimilar treatment options are available for infliximab and adalimumab.[52] However, treatment options such as ustekinumab, vedolizumab, and Janus kinase (JAK) inhibitors (tofacitinib or filgotinib) may be used as a first line biologic in some circumstances, often where patients are unsuitable to receive TNFis, although the NICE recommendations for ustekinumab and ozanimod are the only treatment options for which such a restriction is specified.[3, 53-55]

Patients are treated until remission occurs, after which maintenance of the treatment with or without concomitant conventional therapies is used to maintain remission. Medication reviews are recommended to occur every 12 months to assess the suitability of the current treatment.[51] patients lose response, treatment switching or cycling is common; therefore, a patient may be prescribed several treatments in order to induce or maintain remission.[56] Typically, patients will cycle to an alternative mechanism of action, with limited within-class switching due to typically poorer efficacy outcomes with this approach.[56]

If patients continue to experience inadequate disease control, or are unsuitable to receive these treatment options, the final stage of the treatment pathway involves surgery to remove the colon

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

(colectomy), although patients may elect to undergo surgery at an earlier stage if they deem this to be a more suitable option to improve their quality of life, or reduce symptoms.[48] Despite this, surgery can lead to a number of serious complications requiring further treatment, with recent clinical expert opinion received by NICE during TA828 emphasising the importance of the availability of a range of effective treatment options in order to delay or avoid surgery becoming necessary.[19, 55]

Figure 2: Current treatment pathway for moderately to severely active UC in UK clinical practice and the anticipated positioning of mirikizumab within it

==> picture [451 x 297] intentionally omitted <==

Patients with a response or in remission remain on the same therapy with a 12-month review. In the biologicnaïve setting, ustekinumab is restricted for use only where a TNFi has failed (that is, the disease has responded inadequately or has lost response to treatment) or cannot be tolerated, and ozanimod is for use where conventional treatment cannot be tolerated or is not working well enough and infliximab is not suitable. Abbreviations : IL: interleukin; JAK: Janus kinase; S1P: sphingosine-1-phosphate; TNFi: tumour necrosis factor alpha inhibitors; UC: ulcerative colitis.

Limitations of current treatments

The treatment landscape for UC currently provides several treatment options for clinicians and patients to consider. Despite this, a number of challenges are yet to be adequately addressed. Conventional therapies, typically used at first line, are associated with numerous disadvantages, including limited response, low rates of maintained remission and both short- and long-term side effects.[19, 57, 58] In addition, corticosteroid-free remission is typically sought, given that long-term corticosteroid use can be associated with development of corticosteroid-dependent disease, as well as side effects such as increased risk of infection, osteoporosis, diabetes, weight gain and cardiovascular disease amongst others.[26]

Biologic options considered following treatment failure with conventional therapies may be associated with poor initial response, loss of response over time and safety concerns. Of particular note, a substantial proportion of patients, approximately a third, show no initial

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

response to TNFi induction therapy (primary non-response), while it is estimated that up to 46% of patients who do respond lose response over time (secondary non-response).[59, 60] Other advanced treatments with mechanisms of action distinct from inhibition of TNF, such as vedolizumab, ustekinumab or JAK inhibitors (tofacitinib and filgotinib) may be used, and changing to other mechanisms of action has been identified as a potential solution to overcome non-response to TNFi therapy in IBD.[60] However, these treatments are also associated with several disadvantages, including the common prospect of secondary non-response and the continued experience of debilitating UC symptoms, despite ongoing biologic treatment.[61-64] Additionally, remission rates for these therapies are also suboptimal, and some currently available therapies are associated with safety concerns, such as the association of lymphoma with TNFi treatment.[65, 66]

Due to these limitations, the cycling of several treatment options is required in a significant number of patients; the primary reasons for switching include primary non-response, secondary non-response, and continued pain, and the result is suboptimal treatment for patients with UC.[67] Furthermore, it has been reported that patients who do not adequately respond to primary TNFi induction are less likely to achieve remission following subsequent treatment with a currently available biologic therapy such as another TNFi or ustekinumab.[62, 68]

Unmet need

The combination of factors described above means that there is currently an unmet need in UK clinical practice for a novel treatment for patients with UC that is effective, with a tolerable safety profile and an alternative mechanism of action. This is supported by a Delphi panel consisting of nine gastroenterologists from European countries and with a specialism in IBD, which identified an unmet need for a new treatment which improves disease control and obtains the common treatment goals of maintained long-term clinical and endoscopic remission.[69] These unmet needs clearly underscore the value that more treatment options to alleviate this situation would bring.

Additionally, a common symptom mentioned in the American and European treatment guidelines, but which has not been widely addressed with current treatment options, is bowel urgency.[19, 70] Bowel urgency is defined as the sudden and immediate need to have a bowel movement and is a distinct symptom from stool frequency. Affecting the majority of patients with UC, urgency is an impactful and disruptive symptom which can have a significant negative effect on patient quality of life, which is reflected by urgency being ranked as the most bothersome symptom in a survey of over 750 patients with UC.[36, 71] The mechanisms underlying bowel urgency are linked to rectal inflammation and a lack of mucosal healing and the inflammation and alteration in rectal wall functioning is thought to lead to decreased compliance of the rectum, resulting in a propensity for constant rectal spasms.[72, 73] This link between inflammation and bowel urgency may provide an explanation for observations that clinical and endoscopic outcomes are improved in patients exhibiting lower levels of bowel urgency, thereby signifying reduced disease activity.[74] However, despite being a common symptom, it is often not discussed by patients due to the associated embarrassment, and it may not be addressed by clinicians due to an expectation for patients to proactively raise this sensitive topic themselves.

Mirikizumab

Mirikizumab is positioned for use in adult patients with moderately to severely active ulcerative colitis for whom:

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

• Conventional treatment cannot be tolerated or is not working well enough and other biological treatment is not suitable (“biologic-naïve”), or

• Biologic treatment cannot be tolerated or is not working well enough (“biologic-failed”)

As outlined above, the term “biologic-failed” is used throughout this submission to encompass prior failure with anti-TNF antibodies, anti-integrin antibodies or tofacitinib. This is in line with the definition of “biologic-failed” patients within the pivotal LUCENT trials, as outlined in Section B.3.4.1.

This positioning is represented in Figure 2. Here, patients in the UK may be eligible to receive a range of biologic treatment options, such as ustekinumab, vedolizumab and ozanimod. Clinician choice is likely to depend on factors including failure to respond or loss of response, or contraindication or unsuitability. Furthermore, clinician choice often depends on prior treatment received, and for some treatment options, this restriction is explicit; for example, as per its NICE recommendation, ustekinumab may be considered a first-line biologic option only where a patient has responded inadequately or has lost response to TNFi therapy, or TNFi therapy cannot be tolerated.

The efficacy and safety of mirikizumab versus placebo have been evaluated in the LUCENT-1 and LUCENT-2 trials where it demonstrated rapid and sustained induction of clinical remission and a tolerable safety profile (the methodology and results of the LUCENT trials are presented in Section B.3.3). Given its novel mechanism of action, clinical efficacy and safety, a NICE recommendation for mirikizumab as a treatment in this population in England and Wales would fulfil a considerable unmet clinical need in this group of patients and provide clinicians with another option in their armamentarium for the treatment of adult patients with moderately to severely active UC.

==> picture [34 x 12] intentionally omitted <==

Equality considerations

No equality considerations have been identified.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

B.2 Key drivers of the cost effectiveness of the

comparator(s)

==> picture [32 x 11] intentionally omitted <==

Clinical outcomes and measures

The six NICE technology appraisals relating to biologic treatments for moderately to severely active UC included in the NICE final scope for this appraisal for which final NICE guidance has been published are listed below, alongside whether they followed the NICE Single Technology Appraisal (STA) route or the Multiple Technology Appraisal (MTA) route. The main comparators considered in this submission are bolded.

• Infliximab, adalimumab and golimumab for treating moderately to severely active ulcerative colitis after the failure of conventional therapy (MTA; TA329)[51]

• Vedolizumab for treating moderately to severely active ulcerative colitis (STA; TA342)[53]

• Tofacitinib for moderately to severely active ulcerative colitis (STA; TA547)[75]

• Ustekinumab for treating moderately to severely active ulcerative colitis (STA; TA633)[3]

• Filgotinib for treating moderately to severely active ulcerative colitis (STA; TA792)[54]

• Ozanimod for treating moderately to severely active ulcerative colitis (STA; TA828)[55]

Note that upadacitinib is not considered here given that NICE are yet to published final guidance for its use in UC and therefore it is not considered to represent current clinical practice.[76]

For the six appraisals listed above, the economic models typically consisted of two phases, reflecting the design of the clinical trials informing them: a short-term induction phase and a longer-term maintenance phase. The induction phase has been modelled using two-week tunnel states in order to permit the varying lengths of induction for different treatment options to be modelled as per their respective SmPC.[3, 55]

For all six appraisals, the definitions of clinical remission and response were based on the Mayo scoring system to assess patients’ disease activity after the induction period and during the maintenance phase. Surgery and post-surgery were included in all economic analyses although variations are present in their categorisation in the different appraisals. For example, in the appraisal of tofacitinib (TA547), surgery was modelled as a transient event rather than as a health state, whereas the Committee-accepted approach in TA329, TA342, TA633 and TA828 modelled surgery as a distinct health state.[3, 51, 53, 55] In addition, the TA329 appraisal modelled a single post-surgery state which captured all patients who did had not moved to the ‘death’ state, regardless of whether they experienced post-surgery complications or not, whereas each of the other five appraisals modelled separate post-surgery health states dependent upon whether patients experienced post-surgery complications or not.[3, 51, 53-55, 75]

The health state utility values (HSUVs) considered in these appraisals have typically been sourced from the literature, with values derived from Woehl et al , variously supplemented by values from Arsenau et al or Swinburn et al , being accepted for use in several appraisals.[3, 51, 53, ] 55, 75 Committees have accepted the use of HSUVs from the literature based on the source study being UK-specific or including a reasonably large number of patients, and EQ-5D utility values

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

being reported for most health states.[51] However, utility values derived from the key pivotal trial have also been considered.[3, 54]

Consistent with the ustekinumab appraisal (TA633), the filgotinib (TA792) and ozanimod (TA828) appraisals assumed 30% of patients received an escalated dose in the maintenance period at any one time, and the corresponding higher drug acquisition costs were applied to these patients.[3, 54, 55] A constant risk of discontinuation in the maintenance phase was assumed, but scenario analyses were presented in which a 25% reduction in the loss of response rate after the first year or two years was implemented.[3, 54, 55] The impact of these scenarios was minimal, and in each case, the Committee accepted the constant risk as per the Company’s approach, with one citing a lack of data to inform the model otherwise.[54]

Due to the high costs incurred during the management of serious infections, the associated costs were included in the model in several appraisals (TA828, TA792, TA633 and TA547).[3, 54, 55, 75] They were calculated as a weighted average of six different infections included in the National Schedule of NHS costs: sepsis, tuberculosis, pneumonia, soft tissue infections, bone and joint infections and urinary tract infections. This assumption was deemed to be appropriate for modelling adverse events in UC submissions.

Clinical outcomes and measures in each of these appraisals are discussed in Table 4.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Table 4: Clinical outcomes and measures appraised in published NICE guidance for the comparators included in the NICE final scope

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 24 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 25 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 26 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 27 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 28 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 29 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 30 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 31 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 32 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 33 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 34 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 35 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 36 of 136

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 37 of 136

Abbreviations: ERG: Evidence Review Group; EQ-5D: European Quality of Life 5 Dimensions; HSUV: health state utility value; ICER: incremental cost-effectiveness analysis; ITT: intent-to-treat; NICE: National Institute of Health and Care Excellence; NMA: network meta-analysis; SAE: serious adverse event; SLR: systematic literature review; SmPC: summary of product characteristics; TA: technology appraisal; TNFi: tumour necrosis factor alpha inhibitor; QoL: quality of life; UC: ulcerative colitis.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 38 of 136

==> picture [33 x 12] intentionally omitted <==

Resource use assumptions

The resource use and cost elements included in previous NICE technology appraisals in adults with moderately to severely active UC (TA342 and TA633) that are most relevant to the current appraisal were:

• Drug acquisition costs

• Drug administration costs

• Health-state costs

• Costs of surgery

• Costs associated with the management of AEs (discussed in Section B.2.1)

These costs categories are broadly in line with those included in the cost-effectiveness models of the previous technology appraisals mentioned in Section B.2.1. Disease-related monitoring costs are captured within the health state costs, and except for ozanimod, existing products have minimal to no treatment-specific monitoring requirements outlined in their SmPCs, so costs related to treatment-specific monitoring have not been included in their respective models.[55]

A summary of the healthcare resource use and related cost assumptions and Committee comments on these assumptions are presented below.

Drug acquisition costs

Drug acquisition costs for active and concomitant treatments were included in the costeffectiveness analyses in TA547, TA792 and TA828.[54, 55, 75] In TA633, the costs for concomitant therapies were not included in the company base case, which did not align with ERG preferences.[3] Prior appraisals have derived unit costs from standard sources including the British National Formulary (BNF), the Drugs and Pharmaceutical Electronic Market Information Tool (eMIT), Monthly Index of Medical Specialities (MIMS), previous NICE submissions and published literature. Costs were also modelled separately during the induction phase and maintenance phase of the treatment cycles in TA342, TA633 and TA828.[3, 53, 55] This approach was deemed reasonable by the ERG for each appraisal.

Dose escalation

The ERG considered the modelling of a 30% dose escalation in the maintenance phase as being a reasonable assumption in TA828, particularly given that this was in line with the assumptions for dose escalation accepted in TA633.[3, 55] In addition, this assumption is supported by a multinational chart review conducted in Europe and Canada of patients with IBD who received treatment with TNFi which found that 25.8% of patients with UC needed dose escalation.[77]

Conventional therapy costs

In prior appraisals, patients on active treatment were expected to receive concomitant treatment, which may include treatments typically considered under the term ‘conventional therapy’ such as corticosteroids.[53, 54] Upon discontinuation of active treatment, patients were modelled to proceed to conventional therapy.[3, 53, 54] As such, prior appraisals applied costs of conventional therapy to patients both in the “active treatment” and “post-active treatment” states of the model.[53, 54]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

In prior appraisals, the proportion of patients modelled to be receiving conventional therapy have been derived from TA342 (such as TA633), or from TA547 (such as TA792).[3, 53, 54, 75] In the most recent appraisal, TA828, the ERG and Committee considered that the proportions derived from TA342 were more appropriate for use, in line with those accepted in TA633.[55]

Administration costs

Successive Committees have acknowledged that the costs incurred during drug administration are reliant upon the method of administration.[3, 54, 55, 75] Orally administered drugs have been assumed to not incur any costs to the NHS (TA792).[54] Additionally, in TA633 and TA828, the Committees concluded that it is appropriate for drugs administered subcutaneously to be assumed as having no cost to the NHS, besides the initial cost associated with a nurse training the patient in how to self-administer treatment, due to the possibility of self-administration.[3, 55] In the case of TA828, the ERG noted that approximately 2% of patients may require assistance when using subcutaneous therapy, but it was considered that the inclusion of administration costs for such a small proportion of patients would not have a meaningful impact on results. As such, the Company’s base case assumption was deemed to be acceptable.[55] Drugs requiring IV administration have been assumed to incur the cost of an outpatient visit, based on a weighted average of the NHS reference costs for consultant-led non-admitted, face-to-face follow-up appointments and non-consultant led non-admitted, face-to-face follow-up appointments (TA633, TA792 and TA828).[3, 54, 55]

Health state costs

Various health states were modelled in previous technology appraisals in UC (TA828, TA792, TA633, TA547 and TA342), including: Active UC, Response without remission, Remission, first surgery, Post-first surgery, Post-first surgery complications, Second surgery, and Post-second surgery. Differences between appraisals in how surgery was modelled are outlined in Section B.2.1, but across all appraisals outlined (TA329, TA342, TA547, TA633, TA792 and TA828), a UK cost-effectiveness study by Tsai et al (2008) has represented the accepted source of health care resource use and costs for all non-surgery health states, for which no costs were reported.[3, ] 51, 53-55, 75, 78 The assumption in TA633 that the resource use for first and second surgery health states would be equivalent to the active UC health state was deemed appropriate by the ERG.[3] Modelled costs were also validated by clinical expert opinion in TA828 which elicited costs that were broadly in line with those reported in TA633.[55]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

B.3 Clinical effectiveness

==> picture [452 x 642] intentionally omitted <==

----- Start of picture text -----
Summary of clinical effectiveness evidence
The LUCENT trials
• The efficacy and safety data for mirikizumab versus placebo are derived from two
randomised, double-blind, placebo-controlled Phase III trials (LUCENT-1 and LUCENT-2).
• Baseline characteristics of patients in both trials were well-balanced across treatment
groups, and in the mITT study population, biologic-naïve patients constituted xxxxx and
xxxxx of the mirikizumab and placebo arms in LUCENT-1, respectively, and xxxxx in the
primary analysis cohort (mirikizumab induction responders) in LUCENT-2.
Efficacy data from LUCENT-1 (induction)
• Treatment with mirikizumab in the LUCENT-1 trial showed improvements in all efficacy
outcomes at Week 12 compared to treatment with placebo.
• In the mITT cohort, a significantly higher proportion of patients receiving mirikizumab
achieved clinical remission at Week 12 as compared with those receiving placebo
(xxxxxxxx), and clinical response at the same timepoint was similarly statistically
significantly improved (xxxxxxxx).
• Bowel urgency, an endpoint noted as being of importance to patients, was significantly
improved in mITT patients receiving mirikizumab as compared with those receiving placebo
at Week 12 (xxxxxxxx), with statistically significant improvements in urgency observed
consistently from as early as Week 2 (xxxxx).
• Consistent efficacy results were observed for patients in the biologic-naïve and biologic-
failed subgroups.
Efficacy data from LUCENT-2 (maintenance)
• As in LUCENT-1, treatment with mirikizumab in the LUCENT-2 trial demonstrated
improvements in all efficacy outcomes at Week 40 compared to treatment with placebo
amongst mirikizumab induction responders.
• In the cohort of mirikizumab induction responders, significantly higher rates of clinical
remission were observed at Week 40 in patients re-randomised to receive mirikizumab as
compared with those re-randomised to receive placebo (xxxxxxx), and mirikizumab was
additionally associated with a statistically significant benefit over placebo in maintaining
response in those who have previously responded to mirikizumab induction therapy
(xxxxxxx).
• At Week 40, histologic endoscopic mucosal remission rates and improvements in bowel
urgency were both statistically significantly higher for patients receiving mirikizumab than
those receiving placebo (both xxxxxxx).
• The rates of corticosteroid-free remission without surgery, at Week 40, were also
significantly improved among mirikizumab induction responders when compared to placebo
(xxxxxxx), and this improvement held true across both of the subgroups analysed (both
xxxxxxx).
• In alignment with LUCENT-1, efficacy results from subgroup analyses by prior biologic
exposure status were broadly consistent with the mITT population.
• Mirikizumab re-induction, for patients who achieved a response in LUCENT-1 but
subsequently lost it during LUCENT-2, and open-label extended mirikizumab induction
therapy, for patients who did not achieve a response (regardless of treatment allocation) in
LUCENT-1, also evidenced mirikizumab efficacy.
Safety data from LUCENT-1 and LUCENT-2
• The frequencies of adverse events (AEs) in the mirikizumab-treated patients of LUCENT-2
----- End of picture text -----

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

compared to those receiving placebo were similar, with the majority of TEAEs observed being mild to moderate in nature in both treatment arms. • In both trials, frequencies of serious adverse events (SAEs) and treatment discontinuation due to an AE were broadly comparable between arms, although in LUCENT-2, rates were marginally higher in the group of mirikizumab induction responders receiving placebo than in those receiving mirikizumab (xxxx versus xxxx and xxxx versus xxxx, respectively). • xxxxxxxxxxxxxx occurred throughout either study, which was in the placebo group of LUCENT-2. Efficacy data from the network meta-analyses (NMAs) • In the absence of direct head-to-head data for the efficacy of mirikizumab versus relevant comparators in UC, indirect efficacy analyses were performed for induction and maintenance timepoints for the biologic-naïve and biologic-failed populations • The results of the analyses found that regardless of biologic exposure, mirikizumab offered similar efficacy to most treatments at induction, and clinical benefits in terms of clinical response, clinical remission and mucosal healing versus all other comparators in the maintenance phase. Conclusion • In summary, the introduction of mirikizumab to UK clinical practice would provide clinicians with an additional, effective option with a tolerable safety profile in their armamentarium of biologic therapies to treat patients with moderately to severely active UC in the UK.

==> picture [32 x 11] intentionally omitted <==

Identification and selection of relevant studies

A de novo systematic literature review (SLR) was conducted in November 2018 to identify relevant clinical evidence from randomised controlled trials (RCTs) describing the clinical efficacy and safety of biologic treatments (including JAK inhibitors) for patients with moderately to severely active UC. The SLR was updated in July 2020, April 2021, October 2021, May 2022 and June 2022 using identical methodology to ensure recently published evidence was included.

In total, the overall SLR, including all updates, included 94 publications reporting on 68 unique studies. Full details of the SLR search strategy, study selection process, and results can be found in Appendix D.

==> picture [33 x 12] intentionally omitted <==

List of relevant clinical effectiveness evidence

The SLR identified two randomised, double-blind, placebo-controlled Phase III trials (LUCENT-1 and LUCENT-2) for mirikizumab in UC. The results of these trials are presented from the final clinical study reports (CSRs).[79, 80] A summary of the clinical effectiveness evidence from LUCENT-1 and LUCENT-2 are presented in Table 5.

Table 5: Clinical effectiveness evidence

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Abbreviations: AE: adverse event; CRP: C-reactive protein; Q4W: every 4 weeks; UC: ulcerative colitis.

==> picture [33 x 11] intentionally omitted <==

Summary of methodology of the relevant clinical

effectiveness evidence

==> picture [43 x 11] intentionally omitted <==

Trial design and methodology

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Two Phase III studies, LUCENT-1 and LUCENT-2, were conducted to evaluate the clinical effectiveness and safety of mirikizumab in UC. LUCENT-1 was a 12-week induction study, after which patients could enrol into the LUCENT-2 maintenance study. Further details of both studies are presented below.

B.3.3.1.1 LUCENT-1

Trial design

The trial design of LUCENT-1 is shown in Figure 3. LUCENT-1 was a multicentre, randomised, double-blind, parallel-arm, placebo-controlled Phase III trial designed to evaluate the safety and efficacy of mirikizumab, compared with placebo, over a 12-week induction period. The trial was conducted at 163 centres that screened 2079 patients; screening lasted up to 28 days prior to trial baseline. As outlined in Section B.3.2, the study population included patients with moderately to severely active UC who had an inadequate response to, loss of response to, or intolerance to conventional therapy for UC (“conventional-failed”), and those who had an inadequate response to, loss of response to, or intolerance to biologic therapy for UC (“biologic-failed”).

After screening, patients were randomised 3:1 to intravenous (IV) mirikizumab 300 mg Q4W or IV placebo Q4W, respectively, and were stratified by biologic-failed status, baseline corticosteroid use, baseline disease activity (as defined by a modified Mayo score of 4–6 or 7–9) and region. Patients received their allocated study treatment at Weeks 0, 4 and 8. Patients who completed Week 12 of LUCENT-1 were eligible to enrol into the LUCENT-2 maintenance trial, regardless of their clinical response status at Week 12 of LUCENT-1, while patients who discontinued LUCENT-1 before Week 12, or completed LUCENT-1 but did not enrol in LUCENT2, completed a post-treatment follow-up period for 16 weeks subsequent to their last visit.

Figure 3: The trial design of LUCENT-1

==> picture [452 x 157] intentionally omitted <==

aPatients who completed LUCENT-1 through Week 12 either completed post-treatment follow-up within the study or were eligible to participate in the maintenance study LUCENT-2. Abbreviations: IV: intravenous; Q4W: every 4 weeks.

Trial methodology

The primary endpoint of LUCENT-1 was the proportion of patients in clinical remission at Week 12, based on the modified Mayo score. Secondary endpoints included further assessments of remission, the proportion of patients with clinical response at Week 12, the proportion of biologicfailed patients with clinical response at Week 12 and assessment of improvement in bowel urgency. A summary of the methodology of the LUCENT-1 trial is presented in Table 6.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Table 6: Summary of LUCENT-1 trial methodology

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Abbreviations : CSR: clinical study report; IBD: inflammatory bowel disease; JAK: Janus kinase; TNFi: tumour necrosis factor alpha inhibitor; UC: ulcerative colitis.

B.3.3.1.2 LUCENT-2

Trial design

LUCENT-2 was a multicentre, randomised, double-blind, parallel-arm, placebo-controlled Phase III maintenance study which followed LUCENT-1 and evaluated the safety and efficacy of mirikizumab in maintaining treatment response at Week 40 (after 52 weeks of continuous study drug treatment). The trial was conducted at 368 centres with 1178 patients and comprised five treatment arms. The treatment received in LUCENT-2 was dependent on the treatment arm patients were initially randomised to at Week 0 of LUCENT-1, and the achievement of clinical response at Week 12 of LUCENT-1. The LUCENT-2 study design is shown in Figure 4. Patients who completed LUCENT-2 could be enrolled into an open-label extension (LUCENT-3).

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Figure 4: The trial design of LUCENT-2

==> picture [452 x 210] intentionally omitted <==

aPatients for whom re-induction (“rescue therapy”) with open-label mirikizumab was not deemed to demonstrate clinical benefit discontinued treatment and were not eligible to enter the open-label extension. Abbreviations: IV: intravenous; NR: non-responder OL: open-label; Q4W: every 4 weeks; R: responder; SC: subcutaneous; W: week.

Mirikizumab responders from LUCENT-1 (primary study population)

The primary study population comprised patients, randomised to the mirikizumab arm of LUCENT-1, who showed a clinical response at Week 12 (for definition see Section B.3.3.3). These patients were re-randomised 2:1 to subcutaneous mirikizumab 200 mg Q4W (maintenance therapy) or subcutaneous placebo Q4W, respectively. Randomisation was stratified based on biologic-failed status, induction remission status, baseline corticosteroid use, and region. Patients continued their treatment assignment throughout the LUCENT-2 trial unless they developed secondary loss of response (for definition, see Section B.3.3.3).

If a loss of response was confirmed (including the use of endoscopy results), patients received three doses of open-label IV mirikizumab 300 mg Q4W re-induction therapy. If patients were deemed to have achieved clinical benefit from the re-induction therapy after the three doses, the patients were considered for enrolment into LUCENT-3 but could not continue in LUCENT-2.

Placebo responders from LUCENT-1

Patients randomised to placebo in LUCENT-1 who achieved clinical response at Week 12 continued to receive placebo in LUCENT-2. If a loss of response was confirmed, patients followed the same procedures as for the mirikizumab responders, described above.

Mirikizumab and placebo non-responders from LUCENT-1

Patients who did not achieve clinical response to IV mirikizumab 300 mg Q4W or placebo during LUCENT-1 received open-label extended induction therapy (IV mirikizumab 300 mg Q4W) at Weeks 0, 4 and 8 of LUCENT-2. At Week 12, following extended induction, these patients underwent delayed clinical response assessment. Patients who achieved delayed clinical response, as compared with LUCENT-1 baseline, after extended induction therapy could subsequently receive open-label subcutaneous mirikizumab 200 mg Q4W from Week 12. Patients continued this treatment and underwent clinical response evaluation via endoscopy at Week 40 unless they discontinued from the study. If patients were deemed to have achieved Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

clinical benefit at Week 40, they were considered for enrolment into LUCENT-3 to continue subcutaneous maintenance therapy. Patients who did not achieve clinical response after extended induction therapy at Week 12 of LUCENT-2 discontinued the study.

Post-treatment follow-up period

Patients underwent a maximum 16-week post-treatment follow-up period. Patients who discontinued the study having last received IV mirikizumab returned for post-treatment follow-up visits at 4 and 16 weeks after the end-of-treatment visit. Patients who discontinued the study having received subcutaneous mirikizumab returned for post-treatment follow-up at 4 and 12 weeks after the end-of-treatment visit. Patients who subsequently entered the open-label extension study did not need complete the post-treatment follow-up period.

Trial methodology

The primary endpoint of LUCENT-2 was the proportion of patients in clinical remission at Week 40, based on the modified Mayo score. Secondary endpoints included further assessments of remission, maintenance of remission from Week 12 of LUCENT-1 to Week 40 of LUCENT 2 and assessment of improvement and remission in bowel urgency. A summary of the methodology of the LUCENT-2 trial is presented in Table 7.

Table 7: Summary of LUCENT-2 trial methodology

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Abbreviations : IBD: inflammatory bowel disease; JAK: Janus kinase; Q4W: every 4 weeks; TNFi: tumour necrosis factor alpha inhibitor; UC: ulcerative colitis.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

==> picture [45 x 12] intentionally omitted <==

Baseline characteristics

B.3.3.2.1 LUCENT-1

Summaries of the demographic characteristics and baseline disease characteristics for patients included in the LUCENT-1 trial are provided below in Table 8 and Table 9, respectively.

Overall, baseline characteristics were well-balanced across treatment groups. Respectively, patients in the mITT population receiving mirikizumab and placebo had a mean age of xxxx years and xxxx years, and xxxxx and xxxxx were male. A similar proportion of patients in each arm had severe disease at baseline as per their modified Mayo score category (7–9) (xxxxx for the mirikizumab arm, xxxxx for the placebo arm) and the mean total Mayo score was xxx in both arms. Additionally, in line with the decision problem, biologic-naïve and biologic-failed patients were included, with biologic-naïve patients constituting xxxxx and xxxxx of the mirikizumab and placebo arms, respectively.

Table 8: Baseline demographic characteristics for patients in the mITT population of the LUCENT-1 trial

Abbreviations: BMI: body mass index; IV: intravenous; Miri: mirikizumab; mITT: modified intent-to-treat; PBO: placebo; SD: standard deviation. Source: Eli Lilly and Company (data on file): LUCENT-1 Clinical Study Report Table AMAN 8.3 (page 108)[79]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Table 9: Baseline disease characteristics and prior therapies of patients in the mITT population of the LUCENT-1 trial

Abbreviations : IV: intravenous; Miri: mirikizumab; mITT: modified intent-to-treat; PBO: placebo; SD: standard deviation; TNFi: tumour necrosis factor inhibitor.

Source: Eli Lilly and Company (data on file): LUCENT-1 Clinical Study Report Table AMAN 8.3 (page 108)[79]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

B.3.3.2.1 LUCENT-2

Summaries of the demographic characteristics and baseline disease characteristics for patients included in the LUCENT-2 trial and grouped by their response status at the end of the LUCENT-1 trial are provided below in Table 10 and Table 11, respectively.

As in LUCENT-1, baseline demographic and disease characteristics were well-balanced across all treatment arms. The mean age of all study patients in the mITT population was xxxx years, ranging from xxxx years to xxxx years across treatment groups, and just over half of patients were male in all groups (mean: xxxxx). Disease severity at baseline was similarly consistent between groups, with xxxxx to xxxxx (mean: xxxxx) with modified Mayo score of 7–9, and a mean total Mayo score of between xxx and xxx (mean: xxx). Of all patients in the mITT study population, xxxxx were biologicnaïve and xxxxx were biologic-failed.

Table 10: Baseline demographic characteristics for patients in the mITT population of the LUCENT-2 trial

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 52 of 136

Abbreviations: BMI: body mass index; IV: intravenous; Miri: mirikizumab; mITT: modified intent-to-treat; PBO: placebo; SD: standard deviation. Source: Eli Lilly and Company (data on file): LUCENT-2 Clinical Study Report (Additional Analyses 1), Table AMBG.8.6 (page 23).[81]

Table 11: Baseline disease characteristics and prior therapies of patients in the mITT population of the LUCENT-2 trial

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 53 of 136

aAll Study Patients pooling together LUCENT-1 responders and non-responders.

Abbreviations : IV: intravenous; Miri: mirikizumab; OL: open-label; PBO: placebo; SC: subcutaneous; SD: standard deviation; TNFi: tumour necrosis factor inhibitor. Source: Eli Lilly and Company (data on file): LUCENT-2 Clinical Study Report (Additional Analyses 1), Table AMBG.8.6 (page 23).[81]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973] © Eli Lilly (2022). All rights reserved Page 54 of 136

==> picture [45 x 12] intentionally omitted <==

Outcome definitions

Definitions for the modified Mayo score and clinical effectiveness outcomes used in the LUCENT-1 and LUCENT-2 trials are presented in Table 12.

Table 12: Definitions of clinical effectiveness outcomes used in the LUCENT-1 and LUCENT-2 trials

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Abbreviations: EQ-5D 5L: European Quality of Life 5 Dimensions 5 Level index; IBDQ: Inflammatory Bowel Disease Questionnaire; NRS: numerical rating scale; SF-36: Medical Outcomes Study 36-Item Short Form Health Survey UC: ulcerative colitis; WPAI:UC: Work Productivity and Activity Impairment Questionnaire ulcerative Colitis.

==> picture [34 x 12] intentionally omitted <==

Statistical analysis and definition of study groups in the

relevant clinical effectiveness evidence

==> picture [43 x 11] intentionally omitted <==

Trial populations

LUCENT-1 and LUCENT-2

The description and number of patients in each analysis population for the LUCENT-1 and LUCENT-2 trials are presented in Table 13. The modified intention-to-treat (mITT) population was used in the analysis of efficacy and health outcomes. The safety population was used for safety-related analysis.

Table 13: Trial populations used for the analysis of outcomes in the LUCENT-1 and LUCENT-2 trials

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

aFor details of the eCOA transcription error, see Section 3.1.2.2 of the LUCENT-1 and LUCENT-2 CSRs.79, 80 Abbreviations: CSR: clinical study report; eCOA: electronic outcome assessment; ITT: intention-to-treat; mITT: modified intention-to-treat; PP; per-protocol.

LUCENT-2

In addition to the populations described above, LUCENT-2 comprised additional cohorts which were based on the clinical response at Week 12 of LUCENT-1 and the subsequent treatment received in LUCENT-2. These analysis cohorts are presented in Table 14. The mirikizumab induction responder cohort was used for inferential comparisons in the analysis of LUCENT-2 trial data. Data reported for all other cohorts were descriptive.

Table 14: Analysis cohorts used in LUCENT-2, dependent on clinical response at Week 12 of LUCENT-1

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Abbreviations: IV: intravenous; Q4W: every 4 weeks.

Subgroup definitions

In line with the NICE final scope, subgroup analyses based on prior biological agent use were performed. The trial definition of conventional-failed patients, henceforth referred to as “biologicnaïve”, included patients who had an inadequate response to, loss of response to, or were intolerant to conventional therapy. The trial definition of biologic-failed patients, referred to as “biologic-failed”, included patients who had failed and thus discontinued prior biologic therapy, including tofacitinib, due to loss of response, inadequate response, or intolerance. An additional 5 patients on placebo and 15 patients on mirikizumab were previously exposed to but did not fail a biologic or JAK inhibitor.

==> picture [45 x 11] intentionally omitted <==

Patient disposition

Patient flow diagrams for LUCENT-1 and LUCENT-2 are presented in Appendix D.2.

==> picture [45 x 11] intentionally omitted <==

Statistical methods

The statistical methods employed for the LUCENT-1 and LUCENT-2 trials are presented in Table 15.

Table 15: Summary of the statistical methods employed in the LUCENT-1 and LUCENT-2 trials

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

aFor patients impacted by the eCOA transcription error, modified NRI was used; for more details see Section 5.3.4 of the LUCENT-1 SAP and LUCENT-2 SAP.[84, 85] Abbreviations: ANCOVA: analysis of covariance; CMH: Cochran–Mantel–Haenszel; eCOA: electronic outcome assessment; mBOCF: modified baseline observation carried forward; MMRM: mixed-effects model for repeated measures; NRI: non-responder imputation; SAP: statistical analysis plan. Source: Eli Lilly (Data on File): LUCENT-1 Statistical Analysis Plan;[84] Eli Lilly (Data on File): LUCENT-2 Statistical Analysis Plan.[85]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Figure 5: Graphical approach to controlling Type I error rate in LUCENT-1

==> picture [285 x 323] intentionally omitted <==

Source: Eli Lilly (Data on File): LUCENT-1 Statistical Analysis Plan.[84]

Figure 6: Graphical approach to controlling Type I error rate in LUCENT-2

==> picture [438 x 276] intentionally omitted <==

Source: Eli Lilly (Data on File): LUCENT-2 Statistical Analysis Plan.[89]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

==> picture [34 x 12] intentionally omitted <==

Critical appraisal of the relevant clinical effectiveness

evidence

RCTs captured in the clinical SLR were assessed for quality using the NICE clinical effectiveness quality assessment checklist. The results of these quality assessments are presented in Appendix D.3, and a summary of the quality assessment for LUCENT-1 and LUCENT-2 is presented in Table 16.

Table 16: Quality assessment results for the LUCENT-1 and LUCENT-2 trials

==> picture [34 x 11] intentionally omitted <==

Clinical effectiveness results of the relevant studies

Summary of key efficacy outcomes from the LUCENT-1 and LUCENT-2 trials

LUCENT-1:[79]

• Treatment with mirikizumab in the LUCENT-1 trial was associated with improvements in all efficacy outcomes at Week 12 as compared with receipt of placebo.

• • Clinical remission at Week 12 was a primary endpoint in LUCENT-1 and in the mITT cohort, a significantly higher proportion of patients receiving mirikizumab achieved clinical remission as compared with those receiving placebo: xxxx% versus xxxx %, respectively (99.875% CI: xxxxxxxx x xxxxxxxx x xxxxxx). A higher rate of clinical remission was similarly observed in the biologic-naïve and biologic-failed subgroups, although only the former reached statistical significance (xxxxxxxx and xxxxxxx, respectively).

• • Similarly, for a major secondary endpoint of clinical response, a significantly greater proportion of patients receiving mirikizumab achieved clinical response at Week 12 as compared with those receiving placebo (xxxx% versus xxxxxx, respectively; RD: xxxx; 98.785% CI: xxxxxxxx x xxxxxxxxxxxx) in the mITT cohort. This result remained consistent in the biologic-naïve and biologic-failed subgroups (both xxxxxxxx).

• • Bowel urgency was another key secondary endpoint; after 12 weeks of treatment, patients receiving mirikizumab showed significant improvement in bowel urgency as compared with those receiving placebo in the mITT (xxxxxxxx x xxxx), biologic-naïve (xxxxxxxx) and biologic-failed (xxxxxxxx) groups.

• • Furthermore, the statistically significant improvement in bowel urgency associated with mirikizumab treatment as compared with placebo treatment was observed as early as Week 2 in the mITT cohort (xxxxxxxx) and the biologic-failed subgroup (xxxxxxx), and at

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

The Phase III LUCENT-1 and LUCENT-2 trials provide the key source of efficacy and safety data for mirikizumab versus placebo. Efficacy results in this submission are presented for the modified ITT population, the biologic-naïve and biologic-failed populations, as defined in Section B.3.4, from these trials. Presented efficacy results for LUCENT-2 are for the primary study population as defined in Section B.3.4.1: mirikizumab responders from LUCENT-1 who were re-randomised 2:1 to receive mirikizumab or placebo, respectively, throughout LUCENT-2.

The primary and key secondary outcomes for both trials are presented below.[79, 80] Additional secondary outcomes from LUCENT-1 and LUCENT-2 can be found in Appendix M and Appendix N, respectively.

Phase II efficacy and safety results for mirikizumab are not considered in this submission due to the availability of Phase III data but are presented in Sandborn et al (2020).[9]

==> picture [43 x 12] intentionally omitted <==

LUCENT-1

B.3.6.1.1 Clinical remission at Week 12

Clinical remission rates at Week 12 for patients receiving mirikizumab or placebo in the LUCENT-1 trial are presented in Figure 7. These results show that a higher proportion of patients receiving mirikizumab achieved clinical remission at Week 12 as compared with those receiving Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

placebo: xxxxxxx versus xxxxxxx, respectively. This translated to a common risk difference (RD) of xxxxxx (99.875% CI: xxxxxxxx) which was statistically significant (xxxxxxxxx).

In the biologic-naïve subgroup, the proportion of patients who achieved clinical remission was similarly higher in those treated with mirikizumab compared to the placebo group (RD: xxxx; 95% CI: xxxxxxxx; xxxxxxxx). Similarly, in the biologic-failed subgroup, a higher rate of clinical remission was achieved by patients receiving mirikizumab than those receiving placebo (RD: xx; 95% CI: xxxxxxxx; xxxxxxxx).

Figure 7: Clinical remission rates at Week 12 (NRI)

==> picture [460 x 197] intentionally omitted <==

Abbreviations: CI: confidence interval; Miri: mirikizumab; NRI: non-responder imputation; PBO: placebo. Source: Eli Lilly (data on file): LUCENT-1 Clinical Study Report, Tables AMAN.5.1 and AMAN.5.2 (pages 52– 53).[79]

B.3.6.1.2 Alternate clinical remission at Week 12

As outlined in Section B.3.3.3, an alternative definition of clinical remission was considered based on feedback from the FDA on the mirikizumab paediatric programme proposal. In LUCENT-1, xxxxx of the mITT population receiving mirikizumab achieved alternate clinical remission at Week 12, as compared with xxxxx patients receiving placebo. The associated RD was xxxx, identical to the RD observed with the primary outcome, which remained statistically significant (xxxxxxxxx) (Table 17). Statistically significantly higher rates of alternate clinical remission at Week 12 were also observed with mirikizumab as compared with placebo in the biologic-naïve subgroup (RD: xxxx; 95% CI: xxxxxxxxx; xxxxxxx) whereas statistical significance was not achieved in the biologic-failed subgroup (RD: xxx; 95% CI: xxxxxxxxx; xxxxxxx).

Table 17: Alternative clinical remission rates at Week 12 (NRI)

a 99.875% CI reported. b 95% CI reported. Abbreviations: Cl: confidence interval; IV: intravenous; Miri: mirikizumab; mITT: modified intent-to-treat; NRI: non-responder imputation; Q4W: every 4 weeks.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Source: Eli Lilly (data on file): LUCENT-1 Clinical Study Report, Tables AMAN.5.3 and AMAN.5.4 (pages 54– 55).[79]

B.3.6.1.3 Clinical response at Week 12

At Week 12 of the LUCENT-1 trial, a greater proportion of patients receiving mirikizumab than placebo achieved clinical response: xxxxx versus xxxxx respectively (RD: xxxx; 98.785% CI: xxxxxxxxxx), which translated to a statistically significant effect (xxxxxxxxx) (Figure 8). Similarly, in both the biologic-naïve and biologic-failed subgroups, a significantly higher proportion of patients achieved clinical response at Week 12 in the group receiving mirikizumab as compared with placebo (both xxxxxxx).

Figure 8: Clinical response rates at Week 12 (NRI)

==> picture [452 x 196] intentionally omitted <==

Abbreviations: CI: confidence interval; IV: intravenous; Miri: mirikizumab; NRI: non-responder imputation; PBO: placebo.

Source: Eli Lilly (data on file): LUCENT-1 Clinical Study Report, Tables AMAN.5.5 and AMAN.5.6 (pages 55– 56).[79]

B.3.6.1.4 Endoscopic remission at Week 12

Endoscopic remission at Week 12, as defined in Section B.3.3.3, was measured in patients receiving mirikizumab or placebo at the end of Week 12 of the LUCENT-1 trial. The results are summarised in Table 18.

Patients in the mITT population receiving mirikizumab showed significantly higher rates of endoscopic remission at Week 12 when compared to those receiving placebo (xxxxx versus xxxxx). The associated RD (xxxx) was statistically significant (99.875% CI: [xxxxxxxxx; xxxxxxxxx]), and remained statistically significant in the subgroups by prior treatment: in the biologic-naïve subgroup, xxxxx of patients receiving mirikizumab achieved endoscopic remission at Week 12 as compared with xxxxx receiving placebo (RD: xxxx; xxxxxxx), and in the biologicfailed subgroup, xxxxx of patients in the mirikizumab arm achieved this outcome as compared with xxxxx in the placebo arm (RD: xxxx; xxxxxxx).

Table 18: Endoscopic remission rates at Week 12 (NRI)

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

a99.875% CI reported. b 95% CI reported.

Abbreviations: CI: confidence interval; IV: intravenous; Miri: mirikizumab; mITT: modified intent-to-treat; NRI: non-responder imputation; Q4W: every 4 weeks

Source: Eli Lilly (data on file): LUCENT-1 Clinical Study Report, Tables AMAN.5.7 and AMAN.5.8 (pages 57– 58).[79]

B.3.6.1.5 Symptomatic remission at Week 12

Considering symptomatic remission, a significantly greater proportion of patients in the mITT population achieved symptomatic remission at Week 12 following mirikizumab treatment as compared with placebo (xxxxx versus xxxxx). This RD was statistically significant (RD: xxxx, xxxxxxxxx), as presented in Table 19. This result remained consistent in the prior treatment subgroups, both xxxxxxx.

Table 19: Symptomatic remission rates at Week 12 (NRI)

a 99.875% CI reported. b 95% CI reported Abbreviations: CI: confidence interval; IV: intravenous; Miri: mirikizumab; mITT: modified intent-to-treat; NRI: non-responder imputation; Q4W: every 4 weeks

Source: Eli Lilly (data on file): LUCENT-1 Clinical Study Report, Tables AMAN.5.9 and AMAN.5.10 (page 58– 59).[79]

B.3.6.1.6 Bowel urgency NRS at Week 12

As presented in Table 20, bowel urgency at Week 12 was found to be statistically significantly improved in patients in the mITT population receiving mirikizumab as compared with those receiving placebo, as assessed by improvements in the Numeric Rating Scale (NRS) from baseline. Patients receiving mirikizumab demonstrated a greater least squares mean (LSM) change from baseline when compared to the corresponding group receiving placebo (xxxxxx versus xxxxx, respectively). This LSM difference of xxxxx was statistically significant at xxxxxxxxx. Similar statistically significant improvements in bowel urgency associated with mirikizumab treatment as compared with placebo were observed in the biologic-naïve subgroup (LSM change from baseline xxxx; xxxxxxx) and in the biologic-failed subgroup (xxxx; xxxxxxx) at the same timepoint.

As demonstrated in Figure 9, the statistically significant improvement in bowel urgency associated with mirikizumab treatment as compared with placebo treatment was observed as early as Week 2 in the mITT cohort (xxxxxxx) and the biologic-failed subgroup (xxxxxx), and at Week 8 in the biologic-naïve subgroup (xxxxxx).

Table 20: Bowel movement urgency NRS change from baseline at Week 12 (MMRM)

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

==> picture [469 x 88] intentionally omitted <==

----- Start of picture text -----
xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxx [x] xxx
mITT
xxxxxxxxx xxxxxxx xxxxxxxxxx
xxxxxxxxxxxxxxx xxxxxxxxxxxxxxx xxxxxxxxxxxxxxxx [x] xxxx
Biologic-naïve
xxxxxxx xxxxxxx xxxxxxx
xxxxxxxxxxxxxxx xxxxxxxxxxxxxxx xxxxxxxxxxxxxxxx [x] xxxx
Biologic-failed
xxxxxxx xxxxxxx xxxxxxx
----- End of picture text -----

a 99.875% CI reported. b 95% CI reported. Abbreviations: CI: confidence interval; IV: intravenous; LSM: least squares mean; Miri: mirikizumab; mITT: modified intent-to-treat; NRS: numeric rating scale; MMRM: mixed model for repeated measures; Q4W: every 4 weeks; SE: standard error.

Source: Eli Lilly (data on file): LUCENT-1 Clinical Study Report, Tables AMAN.5.11 and AMAN.5.12 (pages 60– 61).[79]

Figure 9: Bowel urgency improvement by treatment week (MMRM)

==> picture [452 x 418] intentionally omitted <==

*p≤0.05; **p≤0.01; ***p≤0.001. MMRM analysis performed only for data at scheduled visits ie, Week 2, Week 4, Week 8, and Week 12.

Abbreviations: CI: confidence interval; IV: intravenous; LSM: least squares mean; Miri: mirikizumab; MMRM: mixed-effects model for repeated measures; NRI: non-responder imputation; PBO: placebo. Source: Eli Lilly (data on file): LUCENT-1 Clinical Study Report, Tables AMAN 8.26 and 8.27 (pages 323 and 326) .[79]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

B.3.6.1.7 Histologic-endoscopic mucosal improvement at Week 12

Histological-endoscopic mucosal improvement at Week 12 was compared in the group of mITT patients receiving mirikizumab and the group receiving placebo. Histologic-endoscopic mucosal improvement was defined as having achieved both histological improvement and endoscopic remission, as outlined in Section B.3.3.3.

Results for histologic-endoscopic mucosal improvements at Week 12 of the LUCENT-1 trial are presented in Figure 10. In the mITT population, a statistically significantly greater proportion of patients receiving mirikizumab achieved histological-endoscopic mucosal improvement at Week 12 (xxxxx) as compared to those receiving placebo (xxxxx) (RD: xxxx; xxxxxxxxx). This statistical significance was similarly observed in the biologic-naïve subgroup (RD: xxxx; xxxxxxx) and the biologic-failed subgroup (RD: xxx; xxxxxxx).

Figure 10: Histologic-endoscopic mucosal improvement rates at Week 12 (NRI)

==> picture [452 x 193] intentionally omitted <==

Abbreviations: CI: confidence interval; IV: intravenous; Miri: mirikizumab; NRI: non-responder imputation; PBO: placebo.

Source: Eli Lilly (data on file): LUCENT-1 Clinical Study Report, Tables AMAN.5.13 and AMAN.5.14 (pages 61– 62).[79]

==> picture [45 x 11] intentionally omitted <==

LUCENT-2

B.3.6.2.1 Clinical remission at Week 40

The proportion of patients who responded to mirikizumab in the LUCENT-1 trial, and were subsequently re-randomised to receive mirikizumab or placebo (mirikizumab induction responders), who achieved clinical remission at Week 40 of the LUCENT-2 trial are presented in Figure 11. In the group receiving mirikizumab, xxxxx of patients achieved clinical remission at Week 40; by comparison, xxxxx of patients receiving placebo achieved clinical remission at the same time point. This translated to a statistically significant benefit for mirikizumab over placebo (RD: xxxx; xxxxxxx). This result remained consistent for the biologic-naïve (RD: xxxx) and the biologic-failed subgroups (RD: xxxx) (both xxxxxxx).

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Figure 11: Clinical remission rates at Week 40 (NRI)

==> picture [452 x 203] intentionally omitted <==

aIncluding tofacitinib-failed patients.

Abbreviations : CI: confidence interval; NRI: non-responder imputation; SC: subcutaneous. Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report, Tables AMBG.5.1 and AMBG.5.2 (pages 70 and 71).[80]

B.3.6.2.2 Alternate clinical remission at Week 40

Rates of clinical remission defined using an alternate definition (see Section B.3.3.3) for mirikizumab induction responders at Week 40 of the LUCENT-2 trial are presented in Table 21. In the full mirikizumab induction responder population, xxxxx of patients achieved alternate clinical remission at Week 40, which was significantly higher than the proportion of patients achieving this response following receipt of placebo (xxxxx). This RD (xxxx) was statistically significant at xxxxxxx). Additionally, patients receiving mirikizumab were found to have significantly higher rates of alternate clinical response at Week 40 in both the biologic-naïve and biologic-failed subgroups when compared with those receiving placebo (both xxxxxxx).

Table 21: Alternate clinical remission rates at Week 40 in a randomised withdrawal maintenance period

Abbreviations: CI: confidence interval; Miri: mirikizumab; Q4W: every 4 weeks; SC: subcutaneous Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report, Tables AMBG.5.3 and AMBG.5.4 (pages 72 and 73).[80]

B.3.6.2.3 Maintenance of clinical remission rates at Week 40

In addition to achievement of clinical remission at Week 40 (see Section B.3.6.2.1), the efficacy of mirikizumab versus placebo in maintaining the clinical remission to Week 40 of patients in the LUCENT-2 trial who had achieved clinical remission in LUCENT-1, was measured in the LUCENT-2 study (Figure 12). In the full mirikizumab induction responder population, more patients achieved maintenance of clinical remission at Week 40 in the mirikizumab group as

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

compared with the group who received placebo (xxxxx versus xxxxx, respectively; RD: xxxx; 95% CI: xxxxxxxxxx). This translated to a statistically significant benefit for mirikizumab over placebo in maintaining remission in those who have previously responded to treatment (xxxxxxx). Similarly, the number of patients achieving maintenance of clinical remission at Week 40 were higher in both the biologic-naïve and biologic-failed subgroups. However, this improvement was statistically significant in the biologic-failed subgroup only (xxxxxxx and xxxxxxx, respectively).

Figure 12: Maintenance of clinical remission at Week 40 (NRI) in a randomised withdrawal maintenance period

==> picture [452 x 204] intentionally omitted <==

Abbreviations: CI: confidence interval; IV: intravenous; Miri: mirikizumab; NRI: non-responder imputation; SC: subcutaneous.

Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report, Tables AMBG.5.7 and AMBG.5.8 (pages 75 and 77).[80]

B.3.6.2.4 Endoscopic remission at Week 40

As compared with those receiving placebo, a statistically significantly greater proportion of mirikizumab induction responders receiving mirikizumab in LUCENT-2 achieved endoscopic remission at Week 40 regardless of previous treatments received (all xxxxxxx; Table 22). In the full mirikizumab induction responder population, xxxxx and xxxxx of patients receiving mirikizumab and placebo achieved endoscopic remission, respectively (RD: xxxx). Similarly, in the biologic-naïve and biologic-failed subgroups respectively, xxxxx and xxxxx of patients receiving mirikizumab achieved endoscopic remission as compared with xxxxx and xxxxx patients receiving placebo (RDs: xxxx and xxxx, respectively; both xxxxxxx).

Table 22: Endoscopic remission rates at Week 40 (NRI) in a randomised withdrawal maintenance period

Abbreviations: CI: confidence interval; Miri: mirikizumab; NRI: non-responder imputation; Q4W: every 4 weeks; SC: subcutaneous

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report, Tables AMBG.5.5 and AMBG.5.6 (pages 74 and 75).[80]

B.3.6.2.5 Corticosteroid-free remission without surgery at Week 40

Corticosteroid-free remission without surgery at Week 40 was measured among mirikizumab induction responders. As shown in Table 23, a significantly greater proportion of patients in the full mirikizumab induction responder population achieved corticosteroid-free remission without surgery at Week 40 in the mirikizumab-treatment group as compared with the placebo group (xxxxx versus xxxxx, respectively. The resulting RD of xxxx (95% CI: xxxxxxxxxx) was statistically significant at xxxxxxx. This was similarly observed in both the biologic-naïve and biologic-failed subgroups, where statistically significantly more patients attained corticosteroidfree remission without surgery at Week 40 following mirikizumab treatment than placebo treatment (both xxxxxxx).

Table 23: Corticosteroid-free remission without surgery rates at Week 40 (NRI) in a randomised withdrawal maintenance period

Abbreviations: CI: confidence interval; Miri: mirikizumab; NRI: non-responder imputation; Q4W: every 4 weeks; SC: subcutaneous

Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report, Tables AMBG.5.9 and AMBG.5.10 (pages 78 and 79).[80]

B.3.6.2.6 Histologic-endoscopic mucosal remission rates at Week 40

At Week 40, a significantly greater proportion of mirikizumab induction responders who received mirikizumab in LUCENT-2 achieved histologic-endoscopic mucosal remission as compared with those receiving placebo: xxxxx of patients in the mirikizumab arm as compared with just xxxxxxin the placebo group (RD: xxxx; xxxxxxx) (Figure 13). The same result was observed when considering the biologic-naïve and biologic-failed subgroups: the RDs of xxxx and xxxx, respectively, were both statistically significant at xxxxxxx.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Figure 13: Histologic-endoscopic mucosal remission rates at Week 40 (NRI) in a randomised withdrawal maintenance period

==> picture [452 x 193] intentionally omitted <==

Abbreviations: CI: confidence interval; Miri: mirikizumab; NRI: non-responder imputation; PBO: placebo; SC: subcutaneous.

Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report, Tables AMBG.5.11 and AMBG.5.12 (pages 80 and 81).[80]

B.3.6.2.7 Bowel urgency NRS improvement at Week 40

Bowel movement urgency numeric rating scale (NRS) improvement was measured among mirikizumab induction responders in LUCENT-2. At Week 40, patients in the full mirikizumab induction responder population receiving mirikizumab had a significantly greater reduction in bowel movement urgency when compared to those receiving placebo (LSM change: xxxxx; xxxxxxx) (Table 24). At the same timepoint, similar improvements in bowel movement urgency were observed with mirikizumab treatment in the biologic-naïve (LSM change: xxxxx) and biologic-failed (LSM change: xxxxx), with both reaching statistical significance (xxxxxxx and xxxxxxx, respectively).

As demonstrated in Figure 14, the statistically significant improvement in bowel urgency associated with mirikizumab treatment as compared with placebo treatment in the full mirikizumab induction responder population was observed to Week 40 (xxxxxx).

Table 24: Bowel movement urgency NRS improvement (change from induction baseline) at Week 40

Abbreviations: CI: confidence interval; LSM: least squares mean; Miri: mirikizumab; NRS: numeric rating scale; Q4W: every 4 weeks; SC: subcutaneous; SE: standard error

Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report, Tables AMBG.5.13 and AMBG.5.14 (pages 82 and 83).[80]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Figure 14: Bowel urgency improvement by treatment week in induction responders (MMRM, full mirikizumab induction responder population)

==> picture [406 x 173] intentionally omitted <==

*p<0.05; **p<0.01; ***p≤0.001.

Abbreviations: CI: confidence interval; LSM: least squares mean; Miri: mirikizumab; MMRM: mixed-effects model for repeated measures; NRI: non-responder imputation; PBO: placebo; SC: subcutaneous. Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report (Additional Analyses 1), Table AMBG.8.33, page 377.[81]

B.3.6.2.8 Bowel urgency remission at Week 40 among clinical responders with urgency NRS ≥3 at induction baseline

Bowel urgency remission at Week 40, defined as a score of 0 or 1 in the urgency NRS, was measured amongst clinical responders who had an urgency NRS of ≥3 at induction baseline. As shown in Table 25, a statistically greater proportion of patients in the full mirikizumab induction responder mirikizumab group achieved urgency remission at Week 40 as compared with those in the placebo group (xxxxx versus xxxxx, respectively; RD: xxxx; xxxxxxx). Similar results were observed when considering the biologic-naïve and biologic-failed subgroups, with mirikizumab showing higher rates of urgency remission at Week 40 compared with placebo in both populations (RD: xxxx and xxxx, respectively). In line with the full mirikizumab induction responder population, this result was statistically significant for both subgroups (xxxxxxx and xxxxxxx, respectively).

Table 25: Urgency remission (urgency NRS=0 or 1) at Week 40 (NRI) in a randomised withdrawal maintenance period in patients with urgency NRS ≥3 at induction baseline

Abbreviations: CI: confidence interval; Miri: mirikizumab: NRI: non-responder imputation; NRS: numeric rating scale; Q4W: every 4 weeks; SC: subcutaneous

Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report, Tables AMBG.5.15 and AMBG.5.16 (pages 84 and 85).[80]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

B.3.6.2.9 Efficacy in patients with loss of response Symptomatic response and remission rates through the re-induction period

As an additional analysis, patients who achieved a clinical response in the LUCENT-1 trial, regardless of treatment allocation, but subsequently lost clinical response during the LUCENT-2 trial entered the Loss of Response (LOR) rescue period where they received open-label reinduction therapy: three doses of 300 mg intravenous mirikizumab (Q4W). This period is referred to throughout this document as the “re-induction” period, but, as noted in Section B.3.4.1, is referred to as “rescue therapy” throughout the LUCENT-2 CSR.

Among the mirikizumab induction responders, xx patients in the placebo group lost response during the Maintenance Period and entered the re-induction period. After receiving at least one dose of open-label IV mirikizumab re-induction, by Week 12 of the re-induction period, xx (xxxx%) regained symptomatic response (Table 26) and xx (xxxx%) achieved symptomatic remission (Figure 15).

Additionally, xx patients in the mirikizumab-treatment group lost response during the Maintenance Period and entered the re-induction period. After receiving at least one dose of open-label IV mirikizumab re-induction, by Week 12 of the re-induction period, xx (xxxx%) regained symptomatic response (Table 26) and x (xxxx%) achieved symptomatic remission (Figure 15).

Table 26: Symptomatic response rates through the re-induction (NRI)

==> picture [452 x 140] intentionally omitted <==

----- Start of picture text -----
Response, n (%) [95% CI]
Placebo Mirikizumab Mirikizumab
Mirikizumab
responder responder responder
responder, total
Placebo SC Placebo SC 200 mg Miri SC (N=xx)
(N=xx) (N=xx) (N=xx)
xxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxx
LOR Week 4
xxxxxxx xxxxxxx xxxxxxx xxxxxxx
xxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxx
LOR Week 12
xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx
----- End of picture text -----

Abbreviations: CI: confidence interval; Miri: mirikizumab; SC: subcutaneous. Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report (Additional Analyses 1), Table AMBG.8.102. (page 1106).[81]

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Figure 15: Symptomatic remission rates through the re-induction period (NRI)

==> picture [431 x 179] intentionally omitted <==

Abbreviations : NRI: non-responder imputation; SC: subcutaneous.

Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report (Additional Analyses 1), Table AMBG 8.101. (page 1101).[81 ]

B.3.6.2.10 Efficacy in mirikizumab induction non-responders

As an additional analysis, patients who did not achieve clinical response to blinded mirikizumab or blinded placebo in LUCENT-1 were entered into LUCENT-2 and received open-label extended induction therapy with 300 mg mirikizumab IV at Weeks 0, 4, and 8, and underwent delayed clinical response assessment via endoscopy at Week 12 of LUCENT-2. The results from these additional analyses are presented below.

Clinical remission

Clinical remission for LUCENT-1 non-responders in the mITT population was measured at Week 12 of the LUCENT-2 trial (Table 27). At this timepoint, following three initial doses of 300 mg, open-label IV mirikizumab therapy in LUCENT-2, xxxx% of patients previously treated with placebo in LUCENT-1 achieved clinical remission, while xxxx% of patients previously treated with mirikizumab demonstrated clinical remission.

Table 27: Clinical remission at 12 Weeks (NRI) in an open-label extended induction period (mITT population)

Abbreviations: CI: confidence interval; IV: intravenous; mITT: modified intent-to-treat; n: number of patients responding within each specific category; N: total number of patients in the population; NRI: non-responder imputation; Q4W: every 4 weeks

Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report (Additional Analyses 1), Table AMBG.8.94 (page 1084).[81]

Clinical response

Clinical response rates for LUCENT-1 non-responders in the mITT population was measured at Week 12 of the LUCENT-2 trial (Table 28). At this timepoint, following 3 initial doses of 300 mg, open-label IV mirikizumab therapy in LUCENT-2, xxxx% of patients treated with placebo in

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

LUCENT-1 achieved a clinical response and xxxx% of patients previously treated with mirikizumab.

Table 28: Clinical response rates at 12 Weeks (NRI) in an open-label extended induction period (mITT population)

Abbreviations: CI: confidence interval; IV: intravenous; mITT: modified intent-to-treat; n: number of patients responding within each specific category; N: total number of patients in the population; NRI: non-responder imputation; Q4W: every 4 weeks

Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report (Additional Analyses 1), Table AMBG.8.93 (page 1082).[81]

Endoscopic remission

Endoscopic remission for LUCENT-1 non-responders in the mITT population was measured at Week 12 in the LUCENT-2 trial (Table 29). At this timepoint, following 3 initial doses of 300 mg, open-label IV mirikizumab therapy in LUCENT-2, xxxx% of patients treated with placebo in LUCENT-1 achieved endoscopic remission as compare with xxxx% of patients previously treated with mirikizumab.

Table 29: Endoscopic remission at 12 Weeks (NRI) in an open-label extended induction period (mITT population)

Abbreviations: CI: confidence interval; IV: intravenous; mITT: modified intent-to-treat; n: number of patients responding within each specific category; N: total number of patients in the population; NRI: non-responder imputation; Q4W: every 4 weeks

Source: Eli Lilly (data on file): LUCENT-2 Clinical Study Report (Additional Analyses 1), Table AMBG.8.95 (page 1086).[81]

==> picture [33 x 11] intentionally omitted <==

Subgroup analysis

The pre-specified subgroup analyses relevant to the decision problem (previous systemic therapy and previous biologic therapy) are presented in Section B.3.6. No additional subgroup analyses are considered.

==> picture [33 x 11] intentionally omitted <==

Meta-analysis

The pre-planned analyses of clinical efficacy from the LUCENT-1 and LUCENT-2 trials are presented in Section B.3.6. A meta-analysis was not conducted as there was no head-to-head comparison between mirikizumab and the comparators within the scope of this submission. A network meta-analysis was conducted and is presented in Section B.3.9.6.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

==> picture [33 x 12] intentionally omitted <==

Indirect and mixed treatment comparisons

In the absence of RCTs providing direct head-to-head data, network meta-analyses (NMAs) to compare the efficacy and safety of mirikizumab with comparators were performed and are presented below. As outlined in Section B.1.1, ustekinumab and vedolizumab have been selected as the reference comparators for the cost-comparison analysis. However, indirect efficacy and safety data for mirikizumab versus a range of comparators are provided for completeness, with the relative efficacy data for ustekinumab and vedolizumab discussed in Section B.3.9.6 and considered further in the subsequent economic analysis presented in Section B.4.

As outlined in Section B.3.9.3.1, where the evidence base allowed, efficacy analyses were performed for induction and maintenance timepoints separately for the two populations of interest (biologic-naïve and biologic-failed) and analyses of safety outcomes were performed for the overall trial population regardless of prior exposure to biologic therapy for reasons described in Section B.3.9.3.1). In alignment with the approach outlined in Section B.1.1 and with the definition of “biologic-failed” patients within the pivotal LUCENT trials, the “biologic-failed” subgroup considered in these NMA analyses encompassed prior failed with JAKi as well as with a biologic therapy such as a TNFi.

==> picture [43 x 11] intentionally omitted <==

Identification and selection of relevant studies

The basis for the NMA was an SLR originally conducted on 12[th] November 2018 which has since been continuously updated to identify newly published studies of interest; the most recent update was conducted in June 2022. Full details on the methodology and results of the SLR, including a full list of search dates, are presented in Appendix D. The onrolling SLR periodically updates the evidence base to account for new studies and changes in approved treatment regimens and dosages.

The objectives of the SLR were to identify all eligible RCT evidence on relevant treatments for patients with moderately to severely active UC to inform the NMA of mirikizumab versus the respective treatment comparators. The scope of the SLR included RCT evidence for adult patients with moderately or severely active disease as defined by the Mayo score or by UC-DAI (further details of the study eligibility criteria are presented in Appendix D.1.3).

To date, a total of 68 unique studies have been included. A summary of the overall SLR search results alongside a PRISMA flow diagram is presented in Appendix D.1.4.1. In addition, while Phase II data for mirikizumab were identified in the SLR, data for the Phase III LUCENT-1 and LUCENT-2 trials were not published at the time of the latest SLR searches and were therefore provided by Lilly as data on file. Further, the PURSUIT-SC study of golimumab identified in the SLR as one study reported on two separate Phase II and Phase III studies and was therefore considered as two studies for the purpose of the NMA feasibility assessment. Thus, a total of 71 studies were considered for inclusion in the NMA and were investigated in the feasibility assessment presented below.

==> picture [45 x 11] intentionally omitted <==

Feasibility assessment

The comparability of the evidence identified from the 71 studies included in the SLR was investigated extensively through a NMA feasibility assessment prior to the conduct of the analyses. Heterogeneity with respect to patient characteristics, interventions, outcomes, and Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

study design were assessed and the potential implications of identified differences is summarised in sections below, and further details for each section are provided in Appendix D.1.6.1.

B.3.9.2.1 Population

As the patient population eligibility criteria for the SLR stipulated that patients must have moderately to severely active UC, as defined by the Mayo score or by UC-DAI, all studies included in the SLR have a patient population relevant for inclusion in the NMA.

A summary of the key baseline characteristics (such as age, the proportion of males, disease duration, baseline Mayo score and the proportion receiving concomitant glucocorticoids and/or immunosuppressants) of studies included in the NMA, by population (biologic-naïve or failed) and timepoint (induction or maintenance), are presented in Section 1.3 of the NMA report appendices, as provided in the reference pack. In addition, Section 1.4 of the NMA report appendices presents a summary of the reported imputation methods used by study.

A number of population and trial characteristics have previously been shown to impact placebo response rates (or baseline risk) within trials of patients with ulcerative colitis, including study location (European versus non-European) and duration of the induction phase, disease status, disease duration, and prior exposure to biologic therapy at enrolment.[90-93] To reduce heterogeneity observed across population characteristics discussed above, and aligning with the LUCENT trial populations, subgroup populations were considered for the NMA and were evaluated where the evidence base allowed as described in Section B.3.9.3.1. In addition, placebo response rates across trials identified for inclusion in the NMA were explored as is further described in Appendix D.1.6.2.

B.3.9.2.2 Study design

Induction

Of the 68 studies identified in the SLR, 28 compared an EMA or FDA approved dosing regimen (see Section B.3.9.2.3 and Appendix D.1.6.1) with another approved dosing regimen or placebo over the induction period. In addition, data for LUCENT are not yet published and thus were not identified by the SLR but were provided by Lilly as data on file and were also included in the feasibility assessment. Further, the study PURSUIT SC was considered as two separate studies (Phase II and Phase III) for the purpose of the feasibility assessment. Thus, a total of 30 induction studies were considered for inclusion in the NMA.

Most studies were multi-regional, with only 5 single-centre studies identified, 4 of which were conducted in Japan (3) and China (1). All included studies were double-blinded, however, substantial differences in sample size were observed ranging from 20 (Probert et al., 2003) to xxxx patients (LUCENT). In addition, differences in length of induction periods varied across studies from 6 to 14 weeks, as presented in Appendix D.1.5.

Maintenance

Of the 68 studies identified in the SLR, 21 compared an EMA or FDA approved dosing regimen (see Section B.3.9.2.3 and Appendix D.1.6.1) with another approved dosing regimen or placebo over the maintenance period. Again, LUCENT was also included in the feasibility assessment, thus, a total of 22 maintenance studies were considered for inclusion in the NMA.

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

Most studies were multinational, with 6 single-centre studies identified, all of which were conducted in Japan (5) and China (1). All included studies were double-blinded, however, again substantial differences in sample size were observed ranging from 31 (PURSUIT-J) to 386 patients (VARSITY).

Trial design heterogeneity

The identified clinical trials for the maintenance phase can be categorised into two groups: treatthrough and re-randomised responder trials. Patients in treat-through trials such as those of infliximab (ACT 1) and adalimumab (ULTRA 2) are randomised at baseline and outcomes are measured after induction and maintenance treatment phases. Patients in re-randomised responder trials on the other hand, continue to maintenance only if they had responded to induction treatment. The induction phase responders are re-randomised to the intervention or placebo/active comparator at maintenance doses. This trial design is more commonly used to evaluate newer treatments such as vedolizumab (GEMINI 1), tofacitinib (OCTAVE sustain), golimumab (PURSUIT), ustekinumab (UNIFI), upadacitinib (U-ACHIEVE), ozanimod (TRUE NORTH), filgotinib (SELECTION) and mirikizumab (LUCENT).

A summary of the included studies and the timepoint of assessment is provided in Appendix D.1.5. The approach to account for heterogeneity arising from these alternative trial designs, and from the timepoint of assessment in the maintenance phase, is discussed in Appendix D.1.6.1.

B.3.9.2.3 Approved doses and regimens for treatment and comparators

All EMA and FDA approved doses and regimens of targeted therapies for the treatment of moderately to severely active UC were included in the NMA. Different dosing arms of the same drug were treated as individual comparators within the NMA, and studies from the SLR that did not meet these criteria were not considered in the NMA feasibility assessment. The list of interventions included in the NMA is presented in Table 27 in Appendix D.1.6.1, and a list of all excluded studies, alongside reasons for exclusion, is provided in Table 26 in Appendix D.1.5.

B.3.9.2.4 Outcomes of interest

The primary goal of treatment for UC is to induce and maintain remission. Rates of clinical response and clinical remission are the most consistently reported outcomes across studies and are the most relevant efficacy parameter in UC to allow comparative analysis. Therefore, the NMA evaluated clinical response and remission for both induction and maintenance phase. Based on the previous HTA submissions and more recently published trials mucosal healing/endoscopic remission emerged as relevant outcome to assess efficacy.[3, 54, 61, 94, 95] To capture and compare the safety profiles of approved regimens to mirikizumab, all cause discontinuation and incidence of severe adverse events (SAEs) during the induction phase were analysed. The definition of the efficacy and safety endpoints are as per the clinical trials. Across the included trials the definitions were deemed as heterogenous. An overview of the trial definitions per outcome is presented in sections below.

Efficacy Outcomes

Clinical response (induction and maintenance)

In total, 30 induction studies and 15 maintenance studies reported clinical response. The most commonly reported definition of clinical response in both the induction and maintenance networks was a total Mayo score decrease of ≥3 points and ≥30% from baseline accompanied by Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

a decrease in the rectal bleeding subscore of ≥1 point or an absolute score of 0 or 1. The definitions of clinical response implemented in trials in the induction and maintenance networks are presented in Table 28 (induction) and Table 29 (maintenance) in Appendix D.1.6.1.

Clinical remission (induction and maintenance)

In total, 31 studies reported clinical remission for the induction phase and 24 studies reported clinical response for the maintenance phase. The most commonly reported definition of clinical remission in both networks, reported by 17 studies in the induction NMA and by 15 studies in the maintenance NMA, was a total Mayo score of ≤2 points with no individual subscore >1. The definitions of clinical remission implemented in trials in the induction and maintenance networks are presented in Table 30 (induction) and Table 31 (maintenance) in Appendix D.1.6.1.

Mucosal healing (induction and maintenance)

Terminology around mucosal healing and endoscopic improvement are often used interchangeably across studies. In addition, definitions of mucosal healing may differ. For the purpose of the NMA, the definition of mucosal healing was aligned with the definition used in the LUCENT trials: “Endoscopic subscore of 0 or 1”. In total, 13 induction studies used this definition when reporting on mucosal healing while 7 induction studies used this definition when reporting on endoscopic improvement. For the maintenance phase, mucosal healing was uniformly reported by 12 studies.

Safety outcomes

Due to different trial designs for the maintenance phase (outlined in Section B.3.9.2.2), patients with UC can be assigned to placebo or active treatment for the full length of the trial (treatthrough trials) and patients responding to active treatment after induction are re-randomised to active treatment, or placebo (withdrawal). In order to limit the exposure to inactive placebo in rerandomised response-based trials, there are variations in the maintenance treatment received following induction with placebo:

• Placebo induction responders are continued on placebo (UNIFI and PURSUIT studies)

• Placebo induction responders are re-randomised and placebo induction non-responders are treated separately (OCTAVE Sustain)

• Placebo induction responders and non-responders continue on placebo (GEMINI)

As a result, the ‘placebo’ safety population of these trials consist of various ‘placebo’ patients which differ due to the trial designs mentioned above. Therefore, for the safety outcomes, only all cause discontinuation (lack of efficacy, DAEs) and SAEs (adverse events grade 3–4) during the induction phase were considered to be of relevance.

B.3.9.2.5 Summary of trials included in the NMA

Included studies

In total, the NMA included 37 studies from the total of 71 identified in the SLR. For the induction study period, 21 and 13 studies reported data for at least one efficacy outcome for the biologicnaïve and biologic-failed populations, respectively. Further, 21 studies reported data for at least one safety outcome for the overall population (regardless of prior biologic exposure). For the

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved

maintenance study period, 15 and 11 studies reported data for at least one efficacy outcome for the biologic-naïve and biologic-failed populations, respectively.

The studies included in the induction and maintenance NMAs, and their full publication references, are provided in Tables 21 and 22 in Appendix D.1.5, respectively, and an overview of these included studies by population and outcome of interest is presented in Table 23 (induction) and Table 24 (maintenance) in Appendix D.1.5. A summary of the prior biologic therapy population definitions included in the NMA, where reported in the study publications, is presented in Table 25 in Appendix D.1.5.

Excluded studies

Following the NMA feasibility assessment, 34 studies were excluded from the NMA. Most (32 studies) were excluded because they evaluated an intervention and/or comparator which was not an EMA or FDA approved dosing regimen. Other reasons for exclusion from the NMA were relating to population: the RIVETING study (Vermeire et al., 2020) recruited patients who had received tofacitinib for two years and were in stable remission for six months prior to study entry, and the TOUCHSTONE study (Sandborn et al., 2016) did not present outcome data by subgroups for prior exposure to biologic treatment. A list of excluded studies, including further details of reasons for exclusion by study, is presented in Table 26 in Appendix D.1.5.

==> picture [45 x 12] intentionally omitted <==

Methodology

All NMAs were conducted under a Bayesian framework. NMAs of clinical response and remission were performed using a multinomial model with probit link, and a binomial model with logit link was used for NMAs of mucosal healing, all cause discontinuations, and serious adverse events. For each analysis (summarised in Table 30 and presented in full in Appendix D.1.7.1) both fixed effect and random-effects models were considered and are presented in Section B.3.9.4 below, Appendix D.1.10 and Section 2 of the NMA report appendices provided as part of the reference pack (additional results) and Section 3 of the NMA report approaches (supplementary analyses). Results from the best fitting model are reported for each outcome, population and timepoint combination.

Full descriptions of the statistical methods employed in the NMA, including the fixed and random effects models considered, the multinomial and binomial statistical models employed, and discussion of model convergence and selection, are presented in Appendix D.1.7.6.

In addition to the base case analyses, for NMAs of maintenance of clinical response and remission, sensitivity analyses were performed restricting the evidence base of re-randomised studies only (i.e., excluding treat-through studies). Exploratory analyses for baseline risk adjustment using meta-regression models were also conducted in recognition of differences in placebo response observed across trials of UC. Results from the best fitting model for each outcome, population and timepoint combination are reported throughout this section and Appendix D.1.10. Additional sensitivity analyses are presented in Sections 3.1 and 3.2 of the NMA report appendices provided in the reference pack.

Table 30: NMA models used in the base case analysis

Company evidence submission template for mirikizumab for treating moderately to severely active ulcerative colitis [ID3973]

© Eli Lilly (2022). All rights reserved