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

Cost-comparison evaluation

Upadacitinib for previously treated moderately to severely active Crohn's disease [ID4027]

Committee Papers

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

NATIONAL INSTITUTE FOR HEALTH AND CARE EXCELLENCE

COST-COMPARISON EVALUATION

Upadacitinib for previously treated moderately to severely active Crohn's disease [ID4027]

Contents:

The following documents are made available to stakeholders:

The final scope and final stakeholder list are available on the NICE website.

1. Company submission from AbbVie:

• a. Full submission

• b. Summary of Information for Patients (SIP)

2. Clarification questions and company responses

3. Patient group, professional group and NHS organisation submissions from:

• a. Crohn’s & Colitis UK

4. Expert personal perspectives from:

• a. James Lindsay, professor of inflammatory bowel disease – clinical expert, nominated by AbbVie

• b. Ruth Rudling, advanced clinical pharmacist – clinical expert, nominated by UKCPA

• c. Greg Collins, policy lead – patient expert, nominated by Crohn’s & Colitis UK (*see item 3a)

• d. Derek Fraser – patient expert, nominated by Crohn’s & Colitis UK

• e. Peter Irving, consultant gastroenterologist – clinical expert, nominated by AbbVie

5. External Assessment Report prepared by Kleijnen Systematic Reviews Ltd

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 Upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

Document B

Company evidence submission

October 2022

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Contents

Tables and figures

Table 1: The decision problem .............................................................................................. 8 Table 2: Technology being evaluated .................................................................................. 10 Table 3: Definitions of clinical response and remission in VDZ and UST trials .................... 19 Table 4: Clinical outcomes and measures appraised in published NICE guidance for the comparator(s) ..................................................................................................................... 21 Table 5: Resource use considered in relevant NICE technology appraisals ........................ 24 Table 6: Clinical effectiveness evidence – U-EXCEL (induction study) ................................ 29 Table 7: Clinical effectiveness evidence – U-EXCEED (induction study) ............................. 29 Table 8: Clinical effectiveness evidence – U-ENDURE (maintenance study) ...................... 30 Table 9: Definitions of analysis sets – U-EXCEL and U-EXCEED ....................................... 34 Table 10: U-ENDURE study cohorts ................................................................................... 35 Table 11: Definitions of analysis sets – U-ENDURE ............................................................ 36 Table 12: Comparative summary of trial methodology......................................................... 37

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 13: Primary and secondary efficacy endpoints in U-EXCEL, U-EXCEED and U- ENDURE ............................................................................................................................. 43 Table 14: Definition of disease-specific endpoints used in U-EXCEL, U-EXCEED and U- ENDURE ............................................................................................................................. 44 Table 15: Characteristics of study participants across treatment groups (ITT1 population) . 45 Table 16: Summary of statistical analysis approach in U-EXCEL, U-EXCEED, and U- ENDURE ............................................................................................................................. 50 Table 17: Quality assessment results for RCTs ................................................................... 53 Table 18: CDAI clinical remission at Week 12 (NRI-C) – overall and Bio-IR (U-EXCEL ITT1 population) .......................................................................................................................... 55 Table 19: Endoscopic response at Week 12 (NRI-C) – overall and Bio-IR (U-EXCEL ITT1 population) .......................................................................................................................... 56 Table 20: CDAI clinical remission at Week 4 (NRI-C) – overall and Bio-IR (U-EXCEL ITT1 population) .......................................................................................................................... 57 Table 21: CDAI clinical response (CR-100) at Week 2 (NRI C) – overall and Bio-IR (UEXCEL ITT1 population) ..................................................................................................... 58 Table 22: CDAI clinical response (CR-100) at Week 12 (NRI-C) – overall and Bio-IR (UEXCEL ITT1 population) ..................................................................................................... 59 Table 23: Summary of achievement of endoscopic remission at Week 12 (NRI-C) – overall and Bio-IR (U-EXCEL ITT1 population) ............................................................................... 60 Table 24: Discontinuation of corticosteroid use for CD and achievement of CDAI clinical remission at Week 12 in subjects taking corticosteroids for CD at baseline (NRI-C) – overall and Bio-IR (U-EXCEL ITT1 population) ............................................................................... 61 Table 25: CD-related hospitalisations during the 12-week induction period (AO) – overall and Bio-IR (U EXCEL ITT1 population) ...................................................................................... 62 Table 26: Change from baseline in EQ-5D-5L index value and EQ-5D VAS at Weeks 4 and 12 (MMRM) (U-EXCEL ITT1 population) ............................................................................. 63 Table 27: CDAI clinical remission at Week 12 (NRI-C) – overall and by prior biologic failure status (U-EXCEED ITT1 population) ................................................................................... 64 Table 28: Endoscopic response at Week 12 (NRI-C) – overall and by prior biologic failure status (U-EXCEED ITT1 population) ................................................................................... 65 Table 29: CDAI clinical remission at Week 4 (NRI-C; U-EXCEED ITT1 population) ............ 66 Table 30: CDAI clinical response (CR-100) at Week 2 and Week 12 (NRI-C; U-EXCEED ITT1 population) .................................................................................................................. 66 Table 31: Endoscopic remission at Week 12 (NRI-C; U-EXCEED ITT1 population) ............ 67 Table 32: Discontinuation of corticosteroid use and achievement of CDAI clinical remission at Week 12 in subjects taking corticosteroids for CD at baseline (NRI-C; U-EXCEED ITT1 population) .......................................................................................................................... 67 Table 33: CD-related hospitalisation during the 12-week induction period (AO; U-EXCEED ITT1 population) .................................................................................................................. 68 Table 34: Change from baseline in EQ-5D-5L index value and EQ-5D VAS at Weeks 4 and 12 (MMRM; U-EXCEED ITT1 population) ........................................................................... 69 Table 35: CDAI clinical remission at Week 52 (NRI-C) – overall and Bio-IR (U-ENDURE ITT1 population) .......................................................................................................................... 70 Table 36: Endoscopic response at Week 52 (NRI-C) – overall and Bio-IR (U-ENDURE ITT1 population) .......................................................................................................................... 71 Table 37: CDAI clinical response (CR-100) at Week 52 (NRI-C; U-ENDURE ITT1 population) ........................................................................................................................................... 72 Table 38: Endoscopic remission at Week 52 (NRI-C; U-ENDURE ITT1 population) ........... 72 Table 39: CDAI clinical remission and endoscopic remission at Week 52 (NRI-C; U- ENDURE ITT1 population) .................................................................................................. 73

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 40: Without corticosteroid use for CD ≥90 days prior to Week 52 and achievement of CDAI clinical remission at Week 52 (NRI-C; U-ENDURE ITT-1 population) ........................ 74 Table 41: Discontinuation of corticosteroid use for CD ≥90 days prior to Week 52 and achievement of CDAI clinical remission at Week 52 in subjects taking corticosteroids for CD at induction baseline (NRI-C; U-ENDURE ITT1 population) ................................................ 74 Table 42: CD-related hospitalisation during the 52-week maintenance period (AO; U-ENDURE ITT1 population) .............................................................................................. 75 Table 43: Change from induction baseline in EQ-5D-5L at Week 52 (MMRM; U-ENDURE ITT1 population) .................................................................................................................. 76 Table 44: Summary of trials used in the NMA ..................................................................... 78 Table 45: CDAI outcomes assessed in NMA ....................................................................... 79 Table 46: Trials reporting CDAI outcomes used in the NMA................................................ 80 Table 47: Trials reporting safety outcomes assessed in the NMA ....................................... 80 Table 48: CDAI inclusion criteria and mean baseline CDAI scores of included BF induction populations ......................................................................................................................... 82 Table 49: Results for CDAI clinical response (CR-100) in BF induction NMA (FE model) ... 85 Table 50: Results for CDAI clinical remission in BF induction NMA (FE model) .................. 85 Table 51: Results for CDAI clinical remission in BF maintenance NMA (FE model) ............ 86 Table 52: Results for serious AEs in induction NMA (FE model) ......................................... 87 Table 53: Results for serious AEs in maintenance NMA (FE model) ................................... 88 Table 54: Results for discontinuation due to AEs in induction NMA (FE model) .................. 89 Table 55: Results for discontinuation due to AEs in maintenance NMA (FE model) ............ 90 Table 56: Overview of TEAEs and deaths during the 12-week double-blind and open-label induction periods (U EXCEL and U-EXCEED SA1 and SA2 populations) ........................... 94 Table 57: TEAEs reported in ≥5% of subjects in any treatment group of U-EXCEL or U EXCEED during the 12-week double-blind and open-label induction periods (U-EXCEL and U-EXCEED SA1 and SA2 populations) ............................................................................... 95 Table 58: Overview of treatment-emergent AESI during the 12-week double-blind and openlabel induction periods (U-EXCEL and U-EXCEED SA1 and SA2 populations)................... 96 Table 59: Overview of TEAEs and deaths during the 52-week maintenance period (UENDURE SA1 population) .................................................................................................. 97 Table 60: TEAEs reported in ≥5% of subjects during the 52-week maintenance period (UENDURE SA1 population) .................................................................................................. 98 Table 61: Overview of treatment-emergent AESI during the 52-week maintenance period (UENDURE SA1 population) .................................................................................................. 99 Table 62: Population characteristics for cost comparison model ....................................... 108 Table 63: Proportion of patients on standard and high dose maintenance therapy in cost comparison model ............................................................................................................. 108 Table 64: Acquisition costs of the intervention and comparator technologies .................... 109 Table 65: Administration costs used in cost comparison model ......................................... 110 Table 66: Resource costs of the intervention and comparator technologies (Year 1) ........ 111 Table 67: Summary of key model inputs ........................................................................... 112 Table 68: Key assumptions of the analysis ....................................................................... 113 Table 69: Base case results .............................................................................................. 114 Table 70: Sensitivity analysis 1 results (Year 2+ costs, per year) ...................................... 115 Table 71: Sensitivity analysis 2a results (100% on UPA 15 mg) ........................................ 115 Table 72: Sensitivity analysis 2b results (0% on UPA 15 mg) ............................................ 115 Table 73: Sensitivity analysis 3a results (0% on UST standard dose) ............................... 116 Table 74: Sensitivity analysis 3b results (20% on UST standard dose) ............................. 116 Table 75: Sensitivity analysis 3c costs (30% on UST standard dose) ................................ 116 Table 76: Sensitivity analysis 4 costs (extended induction) ............................................... 116

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Figure 1: Treatment pathway based on CD management guidance by NICE ...................... 15 Figure 2: Proposed positioning of upadacitinib in UK treatment pathway for CD ................. 18 Figure 3: U-EXCEL study design......................................................................................... 32 Figure 4: U-EXCEED study design ...................................................................................... 34 Figure 5: U-ENDURE study design ..................................................................................... 36 Figure 6: Network diagram of included induction studies in the BF population .................... 83 Figure 7: Network diagram of included maintenance studies in the BF population .............. 84

Abbreviations

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

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

B.1.1 Decision problem

The population considered in this appraisal is people with moderately to severely active Crohn’s disease (CD) in whom tumour necrosis factor (TNF)-alpha inhibitors are deemed unsuitable; or where biological treatment is not tolerated or not working well enough.

The anticipated license for upadacitinib in this indication is expected to be ************************************************************************************************ ************************************************************************************************

************************************************. As such, the submission represents a subpopulation to that specified in the NICE pre-invitation scope and licensed indication. The submission covers the anticipated licensed doses of upadacitinib for CD, i.e., oral induction dose of 45 mg once daily for 12 weeks followed by an oral maintenance dose of 15 mg or 30 mg once daily based on individual patient presentation. The decision problem addressed in this appraisal is outlined in Table 1.

Upadacitinib currently holds marketing authorisation in Great Britain (GB) for rheumatoid arthritis, psoriatic arthritis, axial spondyloarthritis, atopic dermatitis, and ulcerative colitis (UC) (1). NICE has recommended upadacitinib in[1] : severe active rheumatoid arthritis (TA665) (2); moderate active rheumatoid arthritis (TA744) (3); active psoriatic arthritis (TA768) (4); and ankylosing spondylitis (TA829) (5). Appraisals of upadacitinib in non-radiographic axial spondyloarthritis (ID3958) (6) and ulcerative colitis (ID3953) (7) are ongoing. The SMC has accepted upadacitinib for use in[1] : active psoriatic arthritis (SMC2361) (8); severe active rheumatoid arthritis (SMC2315) (9); moderate to severe atopic dermatitis in adults and adolescents aged 12 years and older (SMC2417) (10). An appraisal of upadacitinib in ulcerative colitis (SMC2510) and moderate to severe rheumatoid arthritis (SMC2495) are ongoing (11).

1 Restrictions apply to the NICE and SMC recommendations in each indication.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 1: The decision problem

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 8 of 125

Abbreviations: ADA, adalimumab; BF, biologic failure; Bio-IR, biologic inadequate response/intolerance; BSC, best supportive care; CD, Crohn’s disease; CCF, conventional care failure; IFX, infliximab; NA, not applicable; TNF, tumour necrosis factor; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab. †Subjects who had an inadequate response or intolerance to conventional therapy (defined as one or more of the following: aminosalicylates, oral locally acting steroids [e.g., budesonide, beclomethasone], systemic corticosteroids [prednisone or equivalent], or immunomodulators). This population may include patients who had received biologic therapy in the past but stopped therapy based on reasons other than inadequate response (IR) or intolerance (e.g., change in reimbursement coverage, well-controlled disease); ‡Subjects with documented intolerance or inadequate response (either failure to respond to induction treatment, or loss of response to maintenance therapy) to one or more biologics for CD (infliximab, adalimumab, certolizumab, natalizumab, vedolizumab, and/or ustekinumab).[§] The BF population is considered to include those contraindicated to TNF-alpha inhibitors.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 9 of 125

B.1.2 Description of the technology being evaluated

Details of the technology being appraised in the submission are provided in Table 2. The draft summary of product characteristics (SmPC) for upadacitinib is provided in Appendix C (13).

Table 2: Technology being evaluated

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

UPA treatment should be interrupted in patients with ALC <0.5 x 10[9] cells/L, ANC <1 x 10[9] cells/L, or Hb levels <8 g/dL; UPA treatment can be restarted once these levels are restored. UPA should be temporarily interrupted if drug-induced liver injury is suspected. Patients should be managed according to international clinical guidelines for Additional tests hyperlipidaemia. or investigations As such, routine blood workup would be performed on patients with active disease who are eligible to receive UPA. However, patients would receive these tests as part of routine clinical practice and so additional tests or investigations beyond this would not be needed for patients receiving UPA. UPA is commercially available as a pack of 28 x 15 mg tablets at a list price of £805.56 List price and per pack, and as pack of 28 x 30 mg tablets at a list price of £1,281.54 per pack. UPA is average cost of also anticipated to be commercially available as a pack of 28 x 45 mg tablets at a list a course of price of ********* per pack. treatment Average cost of course of treatment (1 year): ****** There is a simple PAS agreed with NHS England which is reflected in the submission. Patient access The PAS equates to an approximate *** discount for each 15 mg, 30 mg, and 45 mg scheme (if packet. This ******* the per packet cost to ******* for a 28 x 15 mg packet, to ********for a applicable) 28 x 30 mg packet, and to ******* for a 28 x 45 mg packet.

Abbreviations: ALC, absolute lymphocyte count; ANC, absolute neutrophil count; CD, Crohn’s disease; CHMP, Committee for Medicinal Products for Human Use; EMA, European Medicines Agency; Hb, haemoglobin; IFN, interferon; IL, interleukin; JAK, Janus kinase; MHRA, Medicines and Healthcare Products Regulatory Agency; PAS, Patient Access Scheme; STAT, signal transducer and activator of transcription; TYK, tyrosine kinase; UPA, upadacitinib.

B.1.3 Health condition and position of the technology in the treatment pathway

B.1.3.1 Disease burden

CD is a chronic relapsing systemic inflammatory bowel disease (IBD) that can cause inflammation and mucosal ulceration to the entire gastrointestinal tract (from the mouth to the anus), but most commonly affects the distal small intestine. Inflammation affects the whole thickness of the bowel wall (23, 24). The pathogenesis of CD involves the complex interaction of immunological, microbiological, environmental, and genetic factors (23, 25, 26).

Presenting symptoms can be heterogeneous and insidious; common CD symptoms include abdominal pain, diarrhoea, fatigue, weight loss, and blood or mucus in stools (26, 27). Individuals with CD typically suffer from recurrent relapses, with acute exacerbations interspersed with periods of remission (28-30). The symptoms of CD can significantly adversely affect individuals’ lives, negatively impacting educational achievements, work productivity, mental health, and quality of life (QoL) (31-36).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Psychological disorders are more prevalent in people with CD compared with matched controls; 60% of people with CD have been reported to experience mental health problems, such as depression and anxiety, when symptoms are active (35-37). Additionally, increased disease activity has been reported to negatively affect individuals’ feelings about relationships; people with CD may experience embarrassment and feel socially restricted as a result of symptoms (38).

Symptoms can also lead to extensive use of health services for disease management (33-35, 39). Timely intervention is required to promote mucosal healing and reduce the risk of long-term complications, e.g., development of fibrotic strictures leading to bowel obstruction and penetrating disease resulting in the development of an abscess and/or fistula (abnormal connection between the inflamed intestine and other body sites) (40, 41). Inadequate treatment of mucosal inflammation leads to disease progression and increased likelihood of surgery (42).

B.1.3.2 Epidemiology

The prevalence of CD in the UK in 2021 was 0.35 and 0.44 in men and women, respectively (43). Given the latest population estimates (44), there are an estimated 178,797 people aged ≥18 years with CD in England and 186,067 in England and Wales. Approximately 40% of people with CD in the UK are estimated to have moderately to severely active disease at any time post diagnosis (45, 46). Based on these data, the estimated total prevalent number of people with moderately to severely active CD in England was approximately 71,519 in 2022. In England and Wales, this figure is estimated to be 74,427 people.

It is estimated that *** of patients with moderately to severely active CD are eligible for biologic treatment (47) (more details on the pathway of care, including biologic treatment, are presented in Section B.1.3.4). The proportion of these patients who have failed at least one prior biologic therapy is 66% (48).

In 2021, there were approximately 6,458 diagnosed incident cases of CD (based on a rate of 9 cases per 100,000) (49). Most individuals are diagnosed between 17 and 40 years of age, with incidence peaking at 14.3 per 100,000 person years in this age category in England (49).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.1.3.3 Economic burden of CD

In 2006, IBD treatment cost the NHS in excess of £700 million (50). The medical requirements of people with CD place a significant burden on healthcare resources, with the average annual cost of care for treating CD in the UK estimated to be £6,156 per person (51). CD is associated with higher rates of primary care visits and emergency attendances compared with matched controls (35). In 2020–2021, there were 132,648 finished consultant episodes and 123,138 hospital admissions related to CD in England (ICD code K50 as the primary diagnosis; note that these data may be slightly different to a ‘typical’ year due to the impact of COVID-19) (52). Additionally, many people with CD require surgery, which contributes to their healthcare resource use (HCRU); the risk of surgery 5 and 10 years after diagnosis of CD has been reported to be 33.3% and 46.6%, respectively (27, 53, 54). Compared with remission, relapse or flare-up is associated with higher costs for treatment, adverse events and complications (total cost per year: £10,513 versus £1,800 for relapse versus remission) (51), while worsening disease severity is associated with increasing healthcare costs (55).

In addition, CD negatively affects the educational achievements and work productivity of individuals (31). In a study of the long-term impact of CD and ulcerative colitis on the career aspirations, opportunities and choices of individuals aged 16–25 in the UK (N=91), 67% reported that their IBD had delayed or was delaying their education and/or training, while 69% felt IBD prevented them from reaching their full educational potential (31). Similarly, in individuals with CD who were in paid employment (N=744), 40% stated that CD prevented them from pursuing their job of choice, 59% had to reduce working hours due to CD, and 54% reported that CD had an impact on their career progression (31). The negative effect of CD on careers translates into a perceived loss of earning for individuals (31). In a retrospective analysis of people with IBD (N=233), 50% of employed people with CD had some loss of employment days, with a median loss of earnings of £299 over a 6-month period (55). Similarly, significantly higher loss of productivity costs has been reported for the caregivers of people with CD aged ≤18 years compared with controls, highlighting that the indirect costs of CD can extend to caregivers (56).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.1.3.4 Current pathway of care

CD is not medically or surgically curable. Treatment choices are made according to clinical judgement and individual preference (57). The aim of medical treatment in CD has been focused on maintaining a symptom-free remission state whilst controlling inflammation, reducing risk of complications, and minimising surgery to preserve the patient’s nutritional independence by maintaining sufficient intestinal luminal length. Endoscopic outcomes, such as mucosal healing, are now recognised as an important treatment target (evident from the recent update in British Society of Gastroenterology [BSG] guidance) (57-59). This is because mucosal healing is associated with better long-term outcomes, such as reduced risk of relapse, decreased hospitalisation rates, steroid-free remission in follow-up examination, resection free intervals and improved QoL (60, 61).

The treatment guidelines that are considered relevant for moderately to severely active CD in UK clinical practice are listed below:

• Crohn’s disease: management (NICE guidance, NG129, 2019) (62)

• BSG consensus guidelines on the management of IBD in adults (2019) (57)

• European Crohn's and Colitis Organisation Guidelines on Therapeutics in Crohn's Disease: Medical Treatment (2020) (63)

In England, current NICE guidelines (NG129) for adults recommend initial pharmacotherapy with conventional care to induce remission (initial presentation or during a flare-up), which typically includes corticosteroids (e.g., prednisolone) or aminosalicylates, typically followed by immunomodulators (IMM), such as azathioprine, to maintain remission. IMM can also be given in addition to corticosteroids in the presence of continued inflammatory exacerbations (62). Of note, aminosalicylates are rarely used in UK clinical practice (64) and other treatment guidelines do not recommend their use for CD (57, 63).

Advanced therapies (i.e., biologics in the current pathway of care) are introduced if there is a poor response to initial therapy with conventional care, or if the conventional care therapy is not tolerated, or is contraindicated. Figure 1 summarises the treatment options for patients with moderately to severely active CD who have failed

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

conventional care (green box) or biologics (orange box) or for whom specific therapies, e.g., TNF-alpha inhibitors, are contraindicated. Guidance from the BSG generally aligns with NICE guidance (57).

Figure 1: Treatment pathway based on CD management guidance by NICE

==> picture [386 x 278] intentionally omitted <==

Abbreviations: CC, conventional care; CD, Crohn’s disease; IR, inadequate response/treatment failure; TNF, tumour necrosis factor. Figure adapted from NICE guidance. Source: NICE (2019), Crohn’s disease: management (NG129) (62). †Biosimilars are also available. ‡TNF-alpha contraindicated people with CD are considered as part of the biologic failure population. For severe disease, stronger immunosuppressive add-on therapies, such as azathioprine and methotrexate, are used (65).

NICE recommends starting biologic therapy treatment with the least expensive option.

After 12 months of treatment with a biologic therapy, clinicians are recommended to assess individuals to determine if they are responding and should continue on the same therapy (62). BSG guidance recommends that the choice between TNF-alpha inhibitor treatment, ustekinumab and vedolizumab should be made on an individual basis, considering individual preference, cost, likely adherence, safety data and speed of response to the drug (57).

Ultimately, clinical management of CD depends on disease activity, site, behaviour of disease (inflammatory, fistulising or stricturing), response to previous treatments, side effect profiles of treatments, patient preference, and extra-intestinal manifestations, such as uveitis and arthritis (57, 62, 66).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Surgery is another treatment option for people with CD (62); the most common reasons for surgery include poor response to drug or nutritional treatment, strictures, and fistulas. The benefits of surgery can include relief from pain, reduction of symptoms (e.g., diarrhoea, vomiting and fatigue), reduction or cessation of treatments which may cause side effects (e.g. steroids), and prevention of delayed growth (67). However, avoidance of surgery is preferable because surgery is not curative and use of biologic therapies may still be required (68). Furthermore, multiple surgeries for CD can result in short bowel syndrome, in which the intestine is shortened and nutrient absorption is impaired (69).

B.1.3.4.1 Dosing of biologic therapies

Dosing of currently available biologic treatments requires induction therapy, where the drug is administered at a higher dose initially to reduce inflammation and improve CD symptoms (i.e., achieve remission). Following induction, a standard dose is administered at regular intervals to maintain control of the disease. There is clinical flexibility for dose escalation of biologics (i.e., IL-12/23 [ustekinumab] and integrin α4β7 [vedolizumab] inhibitors) (70, 71). Specifically, ustekinumab may be initiated at a standard dose (Q12W) or a higher dose (Q8W), and may also be increased from the standard dose (Q12W) to a higher dose during treatment (Q8W) (71, 72). Vedolizumab may also be increased from the standard dose (Q8W) to a higher dose (Q4W) during treatment (70). Vedolizumab SC as maintenance treatment is administered as 108 mg Q2W with no recommended dose escalation (73). Feedback from clinical experts indicates that dose escalation may be necessary in 30% and 92.5% of individuals receiving vedolizumab or ustekinumab, respectively (12).

B.1.3.5 Limitations in current treatment pathway

CD is not always adequately controlled by currently available conventional and biologic therapies; it is estimated that there are approximately 4,000 people in the UK who have had an inadequate response to currently available conventional and biologic therapies[2] (45). Approximately 50% of people with moderately to severely active CD do not respond to or cannot tolerate conventional treatment (74). Although biologic

2 Figure estimated before the introduction of ustekinumab.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

therapies offer additional treatment options, individuals may still experience disease flares resulting in the appearance or worsening of disease symptoms, such as abdominal pain and fatigue, which may require dose escalation, therapy change or treatment with additional therapies, such as corticosteroids (72, 75-78).

Primary non-response to treatment is an issue with biologic therapies. Up to 30% of individuals do not respond to TNF-alpha inhibitor therapy (e.g., infliximab, adalimumab, and their biosimilars) (76). A loss of response rate of approximately 30% at 52 weeks has been reported for vedolizumab (integrin α4β7 inhibitor) and ustekinumab (IL-12/23 inhibitor) (72, 78). Furthermore, in an international online survey (Canada, France, Germany, Italy, Spain, UK and USA) of people with CD, loss of response to prior treatment was reported in 69% of individuals (48).

Another key limitation of existing treatments is their association with adverse side effects leading to potentially increased HCRU and costs. For example, long-term exposure to corticosteroids (often used alongside biologic therapies) may result in an increased risk of numerous adverse events, including infection, psychological disturbances, diabetes, hypertension and osteoporosis (79). TNF-alpha inhibitors are associated with an increased risk of malignancy, demyelination and infection, including tuberculous infection (70, 72, 76).

A further limitation of existing treatments is the development of anti-drug antibodies, which may lead to a loss of clinical efficacy (51, 70). The rates of anti-drug antibody development with TNF-alpha inhibitors are high; anti-drug antibody rates of 28.5% and 62.8% have been reported for adalimumab and infliximab, respectively (80).

B.1.3.6 Upadacitinib for the treatment of CD

The proposed positioning of upadacitinib is for people with moderately to severely active CD in whom TNF-alpha inhibitors are deemed unsuitable; or where biological treatment is not tolerated or not working well enough. This is in line with the NICE recommendations for the clinical pathway in moderately to severely active CD (see Section B.1.3.4) and is similar to that of ustekinumab and vedolizumab (Figure 1).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Upadacitinib fulfils an important unmet need by providing a novel therapeutic option for the management of CD patients with an incomplete response to currently available biologic therapies (Figure 2); it will be the first JAK inhibitor and the only oral advanced therapy available for CD. As an oral advanced therapy, upadacitinib can rapidly improve symptoms that significantly impact on the lives of people with CD.

Figure 2: Proposed positioning of upadacitinib in UK treatment pathway for CD

==> picture [381 x 274] intentionally omitted <==

Abbreviations: CC, conventional care; CD, Crohn’s disease; IR, inadequate response/treatment failure; TNF, tumour necrosis factor. Figure adapted from NICE guidance. Source: NICE (2019), Crohn’s disease: management (NG129) (62). † Biosimilars are also available. ‡ TNF-alpha contraindicated people with CD are considered as part of the biologic failure population, in line with CEM and BIM. For severe disease, stronger immunosuppressive add-on therapies, such as azathioprine, are used (65).

B.1.4 Equality considerations

No equality issues associated with the use of upadacitinib in this indication have

been identified or are foreseen.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.2 Key drivers of the cost effectiveness of the comparator(s)

B.2.1 Clinical outcomes and measures

There are two biologic treatments for CD that would be displaced by the introduction of upadacitinib. The relevant NICE technology appraisals for the treatment of adult patients with moderately to severely active CD after prior therapy are:

• TA456: Ustekinumab for moderately to severely active Crohn’s disease after prior therapy (published 2017) (45)

• TA352: Vedolizumab for treating moderately to severely active Crohn's disease after prior therapy (published 2015) (46)

B.2.1.1 Clinical outcomes used in the cost-effectiveness analyses

B.2.1.1.1 Clinical effectiveness

Key outcomes evaluated in RCTs to assess efficacy of treatments for CD include clinical response and clinical remission, with CD disease activity historically using the CDAI score. The definitions of clinical response and remission were consistent across the ustekinumab and vedolizumab CD RCTs and are presented in Table 3. More details of these severity measures are presented in Appendix K.

Table 3: Definitions of clinical response and remission in VDZ and UST trials

Abbreviations: CDAI, Crohn’s Disease Activity Index; RCT, randomised controlled trial; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab.

The ustekinumab and vedolizumab submissions used two definitions of clinical response. CR-100 (decrease of ≥100 points in CDAI from baseline) was used as a definition of clinical response in both the vedolizumab and ustekinumab clinical trials. In the cost-effectiveness analysis, vedolizumab used CR-70 as the definition of clinical response (to aid comparison with other treatments) while ustekinumab used CR-100 in the base-case analysis, with CR-70 used in a scenario.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Although CDAI is not commonly used as a measure to assess disease severity in UK clinical practice, it is the measure most frequently used in clinical trials for this indication, including in the vedolizumab and ustekinumab trials (57, 81, 82). In UK clinical practice, the Harvey Bradshaw Index (HBI), a simplified and less comprehensive tool than the CDAI is more frequently used to measure disease severity due to its ease of use (83). Studies have shown that results from the HBI correlate with CDAI results (83, 84). Furthermore, committees accepted the use of CDAI in the ustekinumab and vedolizumab submissions given its historic use in assessing response to other biologic treatments, as well as it being a clinically valid and comprehensive tool for CD assessment.

B.2.1.1.2 Safety

In addition to clinical response and remission, the incidence of AEs and discontinuation rates were included in the cost-effectiveness models. Both ustekinumab and vedolizumab appraisals incorporated AEs (selected based on expert opinion and sourced using the same criteria) in their cost-effectiveness analyses. Ustekinumab excluded AEs in a scenario analysis to assess the impact on the incremental cost-effectiveness ratio (ICER), which was negligible. Discontinuation rates due to lack of efficacy were included and based on clinical trial data. Vedolizumab also included discontinuations due to AEs. Discontinuation rates or AE rates did not have a major impact on ICERs across the different populations. Table 4 summarises the key clinical outcome measures in the relevant NICE TAs.

B.2.1.2 Key clinical drivers of the cost-effectiveness analyses

In both the appraisals listed (TA456 and TA352), sensitivity and scenario analyses were conducted to identify key drivers of cost effectiveness. For vedolizumab (TA352) in comparison with TNF-alpha inhibitors (infliximab and adalimumab), the variables with the largest impact on the ICER were the transition probabilities (particularly for remission), the vedolizumab/infliximab/adalimumab efficacy data, and health state costs. For ustekinumab (TA456) the variables with the largest impact on the ICER were treatment duration (largest impact; versus adalimumab) as well as efficacy and extent of resource use for the moderate-to-severe health state (versus adalimumab and vedolizumab).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 4: Clinical outcomes and measures appraised in published NICE guidance for the comparator(s)

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 21 of 125

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 22 of 125

Abbreviations: AE, adverse event; CD, Crohn’s disease; CDAI, Crohn’s Disease Activity Index; ERG, Evidence Review Group; HES, Hospital Episode Statistics; ICER; incremental cost-effectiveness ratio; IFX, infliximab; NA, not applicable; ONS, Office for National Statistics; QALY, quality-adjusted life year; TA; technology appraisal, UST, ustekinumab; VDZ, vedolizumab.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 23 of 125

B.2.2 Resource use assumptions

Resource use considered in the relevant NICE technology appraisals listed in Section B.2.1 is shown in Table 5.

Table 5: Resource use considered in relevant NICE technology appraisals

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 24 of 125

Abbreviations: ADA, adalimumab; AE, adverse event; CQs, clarification questions; ERG, Evidence Review Group; ICER, incremental cost-effectiveness ratio; IFX, infliximab; IV, intravenous; SC, subcutaneous; TB, tuberculosis; UST, ustekinumab; VDZ, vedolizumab.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 25 of 125

B.3 Clinical effectiveness

Evidence for upadacitinib in moderately to severely active CD patients

The Phase 3 pivotal induction studies (U-EXCEL and U-EXCEED) and maintenance study (U-ENDURE) provide the clinical evidence for upadacitinib for the treatment of moderately to severely active CD.

• U-EXCEL and U-EXCEED were Phase 3, multicentre, randomised, double-blind 12-week induction studies that evaluated the efficacy and safety of upadacitinib 45 mg QD versus placebo in adults with moderately to severely active CD

• U-EXCEL enrolled subjects with either inadequate response/intolerance to prior biologic therapy (Bio-IR) or with inadequate response/intolerance to conventional therapy (non-Bio-IR) for CD, while U-EXCEED enrolled subjects with only a documented inadequate response or intolerance to ≥1 biologic therapy for CD (Bio-IR)

  • Subjects in the double-blind upadacitinib 45 mg arms who did not achieve clinical response at Week 12 subsequently received upadacitinib 30 mg QD for 12 weeks (i.e., until Week 24) in an extended induction period. Subjects who received doubleblind placebo and did not achieve clinical response subsequently received upadacitinib 45 mg QD for 12 weeks (i.e., until Week 24)

• U-ENDURE is a Phase 3, multicentre, randomised, double-blind maintenance study (with an ongoing open-label extension phase) to evaluate the efficacy and safety of upadacitinib 30 mg or 15 mg QD in subjects with moderately to severely active CD. The study enrolled subjects who achieved clinical response in U-EXCEL or U-EXCEED

• Subjects who achieved clinical response to upadacitinib 45 mg after 12 weeks of induction treatment (at either Week 12 or Week 24) in U-EXCEL or U-EXCEED were rerandomised 1:1:1 to receive upadacitinib 30 mg QD, upadacitinib 15 mg QD, or placebo

Definition of subpopulation of interest

The naming convention used to describe the population of interest in the pivotal upadacitinib clinical trials in CD (Bio-IR) is different from that used in previous TAs for this indication (biologic failure [BF]). The definition of this specific population is as follows: subjects with documented intolerance or IR (either failure to respond to induction treatment, or loss of response to maintenance therapy) to one or more biologics for CD (infliximab, adalimumab, certolizumab, natalizumab, vedolizumab, and/or ustekinumab). This population is analogous to the BF population which has been described in previous submissions. Consequently, the Bio-IR naming convention is used in the context of the upadacitinib clinical trials only, whilst through the remainder of the submission (i.e., the NMA and cost comparison model), BF is used for consistency with previous TAs.

The population referred to in previous TAs as conventional care failure (CCF) is referred to as non-Bio-IR in the upadacitinib clinical trials; data on this population are presented in Appendix J for completeness†.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Efficacy

• In the induction studies (U-EXCEL and U-EXCEED), a significantly greater proportion of patients achieved the co-primary endpoints (CDAI clinical remission[‡] and endoscopic response[§] ) with upadacitinib 45 mg compared with placebo

• Symptomatic improvements were observed as early as Week 2 (CDAI clinical response) and Week 4 (CDAI clinical remission) with upadacitinib 45 mg

• Significantly more patients receiving upadacitinib achieved endoscopic remission at Week 12 compared with the placebo group in both U-EXCEL and U-EXCEED

• In the maintenance study (U-ENDURE), symptomatic and endoscopic improvements were observed after 52 weeks of treatment with upadacitinib 30 mg or 15 mg

• Rates of steroid-free remission (discontinuation of corticosteroids and achievement of CDAI clinical remission) were significantly higher with upadacitinib versus placebo in both induction studies (U-EXCEL and U-EXCEED) and the maintenance study (U-ENDURE)

• Improvements in HRQoL (assessed using EQ-5D-5L) with upadacitinib were observed as early as Week 4 in the induction studies and persisted to Week 52 of the maintenance study

Safety

• Across the induction (U-EXCEL and U-EXCEED) and maintenance (U-ENDURE) studies, upadacitinib was generally safe and well-tolerated with no new safety risks observed compared with the known safety profile of upadacitinib

Indirect treatment comparison

• In induction NMAs in the BF population, upadacitinib demonstrated superior efficacy to ustekinumab and vedolizumab (as well as to placebo) for CDAI clinical remission and CDAI clinical response

• In the maintenance NMA in the BF population, upadacitinib generally showed superior efficacy to comparators for CDAI clinical remission (the only efficacy outcome assessed)

• In the safety NMAs, upadacitinib showed comparable rates of serious AEs and discontinuation due to AEs compared with all comparators, including placebo

‡CDAI clinical remission defined as CDAI score <150. §Endoscopic response defined as decrease in SES-CD of >50% from baseline of the induction study or for subjects with an SES-CD of 4 at baseline, ≥2-point reduction from baseline, as scored by central reviewer. †Non-Bio-IR: Subjects who had an inadequate response or intolerance to conventional therapy, defined as one or more of the following: aminosalicylates, oral locally acting steroids [e.g., budesonide, beclomethasone], systemic corticosteroids [prednisone or equivalent], or immunomodulators. This population may also include subjects who had received biologic therapy in the past but stopped therapy based on reasons other than inadequate response or intolerance (e.g., change in reimbursement coverage, well-controlled disease); however, the majority of subjects had not received a prior biologic therapy (in U-EXCEL, 8.5% and 9.2% of the upadacitinib and placebo non-Bio-IR groups had received a prior biologic; in U-ENDURE, 14.6% of the upadacitinib 30 mg group and 11.1% of the upadacitinib groups had received a prior biologic; no patients in the placebo group of U-ENDURE had received a prior biologic). This population is analogous to the CCF population which has been described in previous submissions.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.1 Identification and selection of relevant studies

A systematic literature review (SLR) was conducted to identify relevant clinical evidence on the efficacy and safety of upadacitinib and relevant comparators for the treatment of people aged ≥16 years with moderately to severely active CD; the aim was to identify data to facilitate indirect comparison (via NMA) between upadacitinib and comparators. An overview of the methodology, including search strategy, PRISMA flow diagram, list of included studies and list of excluded studies at full text review is provided in Appendix D.

The SLR identified clinicaltrial.gov entries for the upadacitinib clinical trials; however, data for upadacitinib were derived from clinical study reports (CSRs).

B.3.2 List of relevant clinical effectiveness evidence

A summary of the clinical effectiveness evidence for upadacitinib is provided in Table 6, Table 7 and Table 8.

In all of the clinical trials, upadacitinib was compared against placebo. No head-tohead data were available for upadacitinib versus ustekinumab and vedolizumab (the comparators of interest for this submission). Therefore, an NMA was performed to indirectly compare upadacitinib with ustekinumab and vedolizumab (see Section B.3.9).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 6: Clinical effectiveness evidence – U-EXCEL (induction study)

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CD, Crohn’s disease; CSR, clinical study report; non-Bio-IR, conventional therapy inadequate response/intolerance; PBO, placebo; p.o., per os (orally); QD, once daily; UPA, upadacitinib.

Source: U-EXCEL CSR (85).[†] Part 1 refers to the initial induction period of 12 weeks into which all eligible subjects were enrolled.[‡] Part 2 refers to the extended treatment period of 12 weeks that only included clinical non-responders from Part 1.

Table 7: Clinical effectiveness evidence – U-EXCEED (induction study)

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CD, Crohn’s disease; PBO, placebo; p.o., per os (orally); QD, once daily; UPA, upadacitinib. Source: U-EXCEED CSR (86).[†] Part 1 refers to the initial 12-week double-blind induction period into which all eligible subjects were enrolled until sufficient subject numbers were achieved. ‡Part 2 refers to the open-label 12-week induction period that was included to ensure sufficient subject numbers for the U-ENDURE maintenance study.[§] Part 3 refers to the extended treatment period of 12 weeks that only included clinical non-responders from Part 1 and Part 2.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 8: Clinical effectiveness evidence – U-ENDURE (maintenance study)

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CD, Crohn’s disease; CSR, clinical study report; non-Bio-IR, conventional therapy inadequate response/intolerance; PBO, placebo; p.o., per os (orally); QD, once daily; UPA, upadacitinib. Source: U-ENDURE CSR (87)

B.3.3 Summary of methodology of the relevant clinical effectiveness evidence

B.3.3.1 Upadacitinib randomised controlled trials

B.3.3.1.1 Induction studies (U-EXCEL and U-EXCEED)

U-EXCEL and U-EXCEED were Phase 3, randomised, double-blind induction studies that evaluated the efficacy and safety of upadacitinib 45 mg QD versus matched placebo in adults with moderately to severely active CD (85, 86). The population of U- EXCEL comprised subjects with an inadequate response or intolerance to prior biologic therapy (Bio-IR) or with an inadequate response or intolerance to conventional therapy (non-Bio-IR). The population of U-EXCEED comprised subjects with an inadequate response or intolerance to prior biologic therapy (Bio-IR). These populations were defined as follows:

• Bio-IR population: included subjects with a documented inadequate response or intolerance to one or more prior biologic therapies for CD (infliximab, adalimumab,

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

certolizumab, natalizumab, vedolizumab, and/or ustekinumab) (see Appendix J for full details of inclusion criteria).

• Non-Bio-IR population : included subjects with a documented inadequate response or intolerance to one or more prior conventional therapies for CD, defined as oral locally acting steroids (budesonide, beclomethasone); IV or oral corticosteroids (prednisone or equivalent); and/or immunosuppressants (azathioprine, mercaptopurine, methotrexate, tacrolimus). The non-Bio-IR population included subjects who may have received a prior biologic therapy for up to 1 year but discontinued for reasons other than intolerance or inadequate response, e.g., a change in insurance or achieving well-controlled disease.

B.3.3.1.1.1 U-EXCEL

U-EXCEL consisted of two parts:

• Part 1: randomised, double-blind, placebo-controlled induction period

• Part 2: extended induction period for non-responders from Part 1

In Part 1, subjects were randomised in a 2:1 ratio to receive either upadacitinib 45 mg QD or matched placebo for 12 weeks. Randomisation was stratified by baseline corticosteroid use (yes or no), endoscopic disease severity (SES-CD <15 or ≥15), and number of prior biologic treatments (0, 1, >1). At Week 12, subjects who achieved a clinical response were eligible to enter the 52-week U-ENDURE maintenance study (see Section B.3.3.1.2). Clinical response was defined as ≥30% decrease in average daily very soft or liquid stool frequency (SF) and/or ≥30% decrease in average daily abdominal pain (AP) score (both not worse than baseline) (note that this is different from the definition of clinical response as a trial endpoint, see Section B.3.3.3).

Subjects who did not achieve a clinical response at Week 12 of Part 1 were enrolled in Part 2, a 12-week extended induction period. The objectives of Part 2 were to offer blinded upadacitinib induction treatment to placebo non-responders from Part 1 and evaluate delayed response in subjects who did not initially respond to upadacitinib during Part 1.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Treatment allocation in Part 2 was as follows:

• Cohort 1 (subjects who received placebo in Part 1) : double-blind induction treatment with upadacitinib 45 mg QD for 12 weeks (i.e., until Week 24)

• Cohort 2 (subjects who received upadacitinib in Part 1) : double-blind upadacitinib 30 mg QD for 12 weeks (i.e., until Week 24)

At Week 24, subjects who achieved clinical response in Part 2 were also eligible to enter U-ENDURE (see Section B.3.3.1.2). Subjects who did not achieve clinical response were discontinued from the programme and received standard or care at the investigator’s discretion. Data collected from subjects in Part 2 were not part of the primary efficacy analyses for U-EXCEL and are presented in Appendix J.

The study design for U-EXCEL is summarised in Figure 3.

Figure 3: U-EXCEL study design

==> picture [462 x 149] intentionally omitted <==

Abbreviations: DB, double blind; OL, open label; QD, once daily; Rand, randomisation; UPA, upadacitinib.

B.3.3.1.1.2 U-EXCEED

U-EXCEED was divided into three parts:

• Part 1: randomised, double-blind, placebo-controlled induction period

• Part 2: open-label, single arm active induction period

• Part 3: extended induction period for non-responders from Part 1 or Part 2

In Part 1, subjects were randomised in a 2:1 ratio to receive either upadacitinib 45 mg

QD or matched placebo for 12 weeks. Randomisation was stratified by baseline corticosteroid use (yes or no), endoscopic disease severity (SES-CD <15 or ≥15), and number of prior biologic treatments (>1 or ≤1).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Once Part 1 enrolment was completed, subjects were enrolled in Part 2 to receive open-label upadacitinib 45 mg QD for 12 weeks. The objective of Part 2 was to have sufficient subjects with a clinical response who could be re-randomised in the doubleblind maintenance phase of the 52-week U-ENDURE study, while also minimising unnecessary exposure to placebo. Data collected from subjects in Part 2 were not part of the primary efficacy analyses for U-EXCEED and are not presented in this submission.

At Week 12 in Part 1 and Part 2, subjects who achieved clinical response were eligible to enter the 52-week U-ENDURE maintenance study (see Section B.3.3.1.2). Clinical response was defined as ≥30% decrease in average daily very soft or liquid SF and/or ≥30% decrease in average daily AP score (both not worse than baseline). Subjects who did not achieve clinical response were eligible to enter Part 3 of U-EXCEED, an extended 12-week induction period with three cohorts. The objectives of Part 3 were to offer blinded upadacitinib induction treatment to placebo non-responders from Part 1 and to evaluate delayed clinical response to upadacitinib in subjects who did not initially respond to upadacitinib during Part 1 or Part 2.

Treatment allocation for each cohort in Part 3 was as follows:

• Cohort 1 (subjects who received placebo in Part 1): double-blind induction treatment with upadacitinib 45 mg QD for 12 weeks (i.e., until Week 24)

• Cohort 2 (subjects who received upadacitinib 45 mg in Part 1): double-blind upadacitinib 30 mg QD for 12 weeks (i.e., until Week 24)

• Cohort 3 (subjects who received upadacitinib 45 mg in Part 2): open-label upadacitinib 30 mg QD for 12 weeks (i.e., until Week 24)

Subjects in Cohort 1 and Cohort 2 remained blinded to their treatment allocation to avoid unmasking the treatment received in Part 1. At Week 24, subjects who achieved clinical response were eligible to enter U-ENDURE. Subjects without clinical response were discontinued from the program and received standard of care at the investigator’s discretion. Data collected from subjects in Part 3 were not part of the primary efficacy analyses for U-EXCEED.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

The study design for U-EXCEED is summarised in Figure 4.

Figure 4: U-EXCEED study design

==> picture [447 x 194] intentionally omitted <==

Abbreviations: DB, double blind; OL, open label; QD, once daily; Rand, randomisation; UPA, upadacitinib.

Efficacy and safety analyses for U-EXCEL and U-EXCEED were performed using the ITT1 and SA1 populations, respectively. These populations are defined in Table 9 and all analysis sets from the induction studies are defined in Appendix J.

Table 9: Definitions of analysis sets – U-EXCEL and U-EXCEED

Abbreviations: ITT, intention to treat; SA, safety analysis. Note: For safety populations, subjects were assigned to a treatment group based on the ‘as treated’ treatment group, which was determined by the most frequent dose regimen received in the analysis period.

B.3.3.1.2 Maintenance study (U-ENDURE)

U-ENDURE comprised two substudies, Substudy 1 (maintenance phase) and Substudy 2 (long-term extension). Long-term extension data are not presented in this submission and therefore ‘U-ENDURE’ is used to refer to the 52-week maintenance phase of the trial (i.e., Substudy 1).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

The U-ENDURE population included subjects who achieved clinical response and completed U-EXCEL or U-EXCEED. As such, the population comprised both Bio-IR and non-Bio-IR subjects.

For U-ENDURE, baseline was defined as the baseline visit of the induction studies (UEXCEL and U-EXCEED) and Week 0 was defined as the first study visit of U- ENDURE. At Week 0, all subjects were enrolled in U-ENDURE in a blinded fashion to one of 3 cohorts according to their induction treatment history in U-EXCEL or U-EXCEED, as shown in Table 10.

Table 10: U-ENDURE study cohorts

Abbreviations: ITT, intention to treat; QD, once daily; SA, safety analysis. Source: U-ENDURE CSR (87)

Randomisation in Cohort 1 was stratified by Bio-IR and non-Bio-IR status from the induction studies, as well as by patient-reported outcome (PRO) clinical remission and endoscopic response status on entry to U-ENDURE. In line with the study protocol, all primary and secondary efficacy endpoints were analysed in Cohort 1 (ITT1 population, see Table 11), which was planned to include the first 501 patients who were randomised and received at least one dose of study drug. Safety analyses were also conducted in Cohort 1, using the SA1 population (Table 11).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 11: Definitions of analysis sets – U-ENDURE

Abbreviations: ITT, intention to treat; SA, safety analysis.

Subjects with persistent symptoms or worsening CD were permitted to discontinue at any time. Subjects were discontinued from the study if they withdrew consent or if they were deemed unsuitable to continue for any reason by the investigator.

At or after Week 4 of U-ENDURE, subjects who met the criteria for inadequate response and required medical treatment were eligible to receive rescue therapy with open-label upadacitinib 30 mg QD from that point forward until the end of follow-up.

The study design for U-ENDURE is summarised in Figure 5.

Figure 5: U-ENDURE study design

==> picture [465 x 189] intentionally omitted <==

Abbreviations: BL, baseline; DB, double-blind; LTE, long-term extension; PBO, placebo; QD, once daily; RR, rerandomisation; UPA, upadacitinib.

B.3.3.2 Comparative summary of trial methodology

An overview of the methodology of the three pivotal upadacitinib studies informing this submission is presented in Table 12.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 12: Comparative summary of trial methodology

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 37 of 125

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 38 of 125

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 39 of 125

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 40 of 125

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 41 of 125

Abbreviations: ADA, adalimumab; AZA, azathioprine; AP, abdominal pain; BAR, baricitinib; bio-IR, biologic inadequate response/intolerance; CD, Crohn’s Disease; CDAI, Crohn’s Disease Activity Index; CER, certolizumab pegol; EQ-5D-5L, EuroQol-5 Dimensions 5-level; FIL, filgotinib; GOL, golimumab; Hct, haematocrit; IFX, infliximab; IR, inadequate response; IRT, interactive response technology; JAK, Janus kinase; MTX, methotrexate; MP, mercaptopurine; NAT, natalizumab; non-Bio-IR, conventional therapy inadequate response/intolerance; NSAID, non-steroidal anti-inflammatory; PBO, placebo; QD, once daily; SES-CD, Simplified Endoscopic Score for Crohn’s Disease; SF, stool frequency; TNF; tumour necrosis factor; TOF, tofacitinib; UK, United Kingdom; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab; USA, United States of America.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 42 of 125

B.3.3.3 Trial endpoints

The co-primary and secondary endpoints for U-EXCEL, U-EXCEED, and U-ENDURE are presented in Table 13. Definitions and interpretations of these endpoints are provided in Table 14. Both induction (U-EXCEL, U-EXCEED) and maintenance (UENDURE) trials utilised two different co-primary endpoints to address regional differences in regulatory requirements (CDAI is specifically required for FDA) (Table 13). Clinical remission was assessed using PRO and CDAI for the EMA/FDA, respectively, as defined in Table 14.

The main body of this submission presents results for CDAI outcomes; CD clinical trials have historically used this measure and its use is consistent with outcomes reported in the ustekinumab and vedolizumab clinical trials for CD (57, 81, 82).

PRO outcomes are presented in Appendix J for completeness.

Table 13: Primary and secondary efficacy endpoints in U-EXCEL, U-EXCEED and U- ENDURE

Abbreviations: CD, Crohn’s Disease; CDAI, Crohn’s Disease Activity Index; EMA, European Medicines Agency; EQ-5D-5L, EuroQol-5 Dimensions 5-level; EU, European Union; FDA, Food and Drug Administration; PRO, patient-reported outcome; US, United States.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 14: Definition of disease-specific endpoints used in U-EXCEL, U-EXCEED and U-ENDURE

Abbreviations: AP, abdominal pain; CDAI, Crohn’s Disease Activity Index; EQ-5D-5L, EuroQol-5 Dimensions 5- level; PRO, patient-reported outcome; SES-CD, Simplified Endoscopic Score for Crohn’s Disease; SF, stool frequency; VAS, visual analogue scale.

B.3.3.4 Baseline characteristics and demographics

The baseline characteristics from the pivotal induction (U-EXCEL, U-EXCEED) and maintenance (U-ENDURE) trials are summarised in Table 15. The baseline demographics and clinical characteristics of subjects were well balanced between the treatment groups of each trial and were generally similar across the studies.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 15: Characteristics of study participants across treatment groups (ITT1 population)

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 45 of 125

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 46 of 125

Abbreviations: AP, abdominal pain; Bio-IR, biologic inadequate response/intolerance; BMI, body mass index; CD, Crohn’s Disease; CDAI, Crohn’s Disease Activity Index; CSR, clinical study report; FCP, faecal calprotectin; hs-CRP, high sensitivity C-reactive protein; NA, not applicable; non-Bio-IR, conventional therapy inadequate response or intolerance; PBO, placebo; SD, standard deviation; SES-CD, Simple Endoscopic Score for Crohn’s Disease; SF, stool frequency; TNF, tumour necrosis factor; UPA, upadacitinib. Source: U-EXCEL CSR (85); U-EXCEED CSR (86); U-ENDURE CSR (87). † AP score: 0 = no pain; 1 = mild; 2 = moderate; 3 = severe. ‡ U-EXCEED included patients who had an inadequate response or intolerance to biologic therapy, with the eligibility criteria stating that demonstration of intolerance required no minimum dose or duration of use. Therefore, a small proportion of the trial population (1 patient in the placebo arm) failed <1 prior biologic.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 47 of 125

B.3.3.5 Expert elicitation/opinion

Expert opinion was gathered through review of this submission document by 3 clinical and 2 economic experts. The criteria for selecting suitable experts were expertise and experience of treating CD in the UK (clinicians) and specialised technical expertise in economic evaluation and health technology assessment (health economic experts).

B.3.4 Statistical analysis and definition of study groups in the relevant clinical effectiveness evidence

B.3.4.1 Participant flow in the relevant randomised controlled trials See Appendix D for details of participant flow.

B.3.4.2 Definitions of subject population analysis sets

Definitions of subject population analysis sets for the induction (U-EXCEL and U- EXCEED) and maintenance (U-ENDURE) studies are presented in Table 9 and Table 11, respectively, and in Appendix D.

Analyses of the co-primary endpoints and secondary endpoints for Part 1 (12-week induction period) were performed using the ITT1 analysis sets from U-EXCEL and U-EXCEED. Similarly, analyses of the co-primary and secondary endpoints for Cohort 1 (52-week maintenance period) were performed using the ITT1 analysis set from U- ENDURE. Subject numbers for each data set are presented in Appendix D.

B.3.4.3 Statistical analysis

A summary of the statistical analysis methods used in U-EXCEL, U-EXCEED, and U- ENDURE is provided in Table 16.

For all trials, a multiple testing procedure was used to provide strong control of the type I error rate at alpha = 0.05 (2-sided) across analyses comparing upadacitinib 45 mg (U-EXCEL, U-EXCEED) or upadacitinib 30 mg and 15 mg (U-ENDURE) with placebo for the co-primary endpoints and ranked secondary endpoints. Specifically, testing used a sequence of hypothesis testing for the co-primary endpoints followed by the ranked secondary endpoints (see rankings in Appendix J). If both co-primary endpoints achieved statistical significance, continued testing followed a pre-specified

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

weight of alpha allocation between individual hypotheses, as well as between families of hypotheses.

In handling missing data for analysis of the co-primary endpoints, the primary approach was non-responder imputation while incorporating MI to handle missing data due to COVID-19 (NRI-C). The NRI-C categorised any subjects who did not have an evaluation during a pre-specified visit window (either due to missing assessment or early withdrawal from the study) as a non-responder for the visit. The only exception was that subjects with missing data due to COVID-19 infection or logistical restriction were handled by MI and the subjects were categorised as responders or nonresponders based on MI imputed values. In U-EXCEL and U-EXCEED, at/after the CD-related corticosteroids intercurrent event and on/after the date of initiation of CDrelated confounding medications after premature discontinuation of the study drug, subjects were considered non-responders. In U-ENDURE, subjects were considered non-responders at/after the occurrence of the CD-related rescue medications intercurrent event or on/after the date of initiation of CD-related confounding medications after premature discontinuation of the study drug. As observed (AO) analysis was used for some endpoints, including CD-related hospitalisation; in AO analysis, values were not imputed for missing evaluations and therefore a subject without an evaluation on a scheduled visit was excluded from the analysis for that visit. AO included all values collected in the study.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 16: Summary of statistical analysis approach in U-EXCEL, U-EXCEED, and U-ENDURE

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 50 of 125

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 51 of 125

Abbreviations: ANCOVA, analysis of covariance; AO, as observed; Bio-IR, biologic inadequate response/intolerance; CD, Crohn’s disease; CDAI, Crohn’s Disease Activity Index; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; EU, EMA, European Medicines Agency; EU, European Union; FDA, Food and Drug Administration; MAR, missing at random; MI, multiple imputation; MMRM, mixed-effect model repeat measurement; non-Bio-IR, conventional therapy inadequate response/intolerance; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; PRO, patient-reported outcome; SES-CD, Simple Endoscopic Score for Crohn’s disease; UPA, upadacitinib; US, United States.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 52 of 125

B.3.5 Critical appraisal of the relevant clinical effectiveness evidence

A summary of quality assessment results for the upadacitinib trials is provided in Table

17. A complete quality assessment for each trial is provided in Appendix D.

Table 17: Quality assessment results for RCTs

Abbreviations: RCT, randomised controlled trial.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.6 Clinical effectiveness results of the relevant studies

Summary

Key results from upadacitinib induction studies (U-EXCEL and U-EXCEED)

• In both induction studies, a significantly greater proportion of subjects in the upadacitinib 45 mg arm achieved the co-primary endpoints of CDAI clinical remission and endoscopic response at Week 12 compared with the placebo arm

• Greater efficacy with upadacitinib 45 mg compared with placebo was observed as early as Week 2 for CDAI clinical response and Week 4 for CDAI clinical remission

• Endoscopic remission rates were significantly higher with upadacitinib versus placebo at Week 12 of U-EXCEL and U-EXCEED

• Subjects treated with upadacitinib 45 mg had significant improvements in HRQoL (assessed using EQ-5D-5L) at Week 4 and Week 12 compared with placebo

• Clear treatment effects were also observed in the Bio-IR population

Key results from upadacitinib maintenance study (U-ENDURE)

• In the overall population of U-ENDURE, a significantly greater proportion of subjects in the upadacitinib 30 mg and 15 mg groups achieved the co-primary endpoints of CDAI clinical remission and endoscopic response at Week 52 compared with the placebo group

• Significantly more patients achieved endoscopic remission at Week 52 with both doses of upadacitinib versus placebo

• Significant improvements in EQ-5D-5L VAS from induction baseline to Week 52 were observed with both doses of upadacitinib

• Upadacitinib was effective in the overall population and in the Bio-IR population

This section presents the results from the pivotal upadacitinib induction (U-EXCEL and U-EXCEED) and maintenance (U-ENDURE) studies. As described in Section B.3.3.3, CDAI outcomes are presented across all studies as this endpoint has been used in previous trials of treatments for CD and therefore facilitates indirect treatment comparisons (81, 82). PRO outcomes (defined using SF and AP scores) are not used in the model but are presented in Appendix J for completeness. Upadacitinib met all co-primary endpoints (CDAI clinical remission or PRO clinical remission in addition to endoscopic response) across the induction and maintenance studies.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

All study outcome definitions are presented in Section B.3.3.3 and provided in the footnotes of the results tables. Data are presented for the anticipated licensed doses of upadacitinib (45 mg for induction therapy and 30 mg or 15 mg for maintenance therapy).

B.3.6.1 U-EXCEL

B.3.6.1.1 Co-primary efficacy outcome: proportion of subjects with CDAI clinical remission and endoscopic response at Week 12

In U-EXCEL, the co-primary endpoints of CDAI clinical remission and endoscopic response were met for upadacitinib 45 mg QD compared with placebo (85).

At Week 12, a significantly greater proportion of subjects in the upadacitinib 45 mg arm achieved CDAI clinical remission (see definition in Table 14) compared with the placebo arm () (Table 18). A clear treatment effect was observed in the Bio-IR population, with a greater proportion of patients achieving CDAI clinical remission with upadacitinib than with placebo (**********) (Table 18).

Table 18: CDAI clinical remission at Week 12 (NRI-C) – overall and Bio-IR (U-EXCEL ITT1 population)

==> picture [454 x 274] intentionally omitted <==

----- Start of picture text -----
Population Responder (NRI-C) Response rate difference vs PBO
Treatment N n (%) 95% CI [†] Missing Diff. Adj. 95% CI [§] P value
due to diff.
COVID- (%) [‡]
19, n
All subjects
UPA 45 mg *** ********** **********
*** **** **********
PBO *** ********* **********
*
UPA 45 mg *** ********* ********** *
* ** ********** *
PBO ** ********* ********* *
Abbreviations: Bio-IR, biologic inadequate response/intolerance; CDAI, Crohn’s Disease Activity Index; CI,
confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study
report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple
imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-EXCEL CSR
(85). Note: CDAI clinical remission defined as CDAI score <150. [†] 95% CI for response rate is the synthetic result
based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19
or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-
19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for
stratification factors. [§] 95% CI for adjusted difference and p-value are calculated according to the CMH test
adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple
imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data
due to COVID-19.
**----- End of picture text -----*

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

At Week 12, a significantly greater proportion of subjects achieved endoscopic response (see definition in Table 14) in the upadacitinib 45 mg arm compared with the placebo arm () (Table 19). A clear treatment effect was also observed in the Bio-IR population, with a greater proportion of patients achieving endoscopic response with upadacitinib than with placebo (**********) (Table 19).

Table 19: Endoscopic response at Week 12 (NRI-C) – overall and Bio-IR (U-EXCEL ITT1 population)

==> picture [452 x 164] intentionally omitted <==

----- Start of picture text -----
Population Responder (NRI-C) Response rate difference vs PBO
Treatment Missing
Adj.
due to
N n (%) 95% CI [†] Diff. diff. 95% CI [§] P value
COVID-
(%) [‡]
19, n
All subjects
UPA 45 mg *** ********** **********
*** **** **********
PBO *** ********* *********
*
UPA 45 mg *** ********* ********** *
* ** ********** *
PBO ** ******* ********* *
**----- End of picture text -----*

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CI, confidence interval; CMH, Cochran-MantelHaenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; SES-CD, Simple Endoscopic Score – Crohn’s Disease; UPA, upadacitinib. Source: U-EXCEL CSR (85). Note: endoscopic response defined as decrease in SES-CD of >50% from baseline of the induction study or for subjects with an SES-CD of 4 at baseline, ≥2-point reduction from baseline, as scored by central reviewer.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

B.3.6.1.2 Secondary efficacy outcomes – U-EXCEL

B.3.6.1.2.1 CDAI clinical remission at Week 4

At Week 4 of U-EXCEL, a significantly greater proportion of patients achieved CDAI clinical remission in the upadacitinib arm compared with the placebo arm (************************) (Table 20). A between-treatment difference of ***** was observed in the Bio-IR population for upadacitinib versus placebo (Table 20).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 20: CDAI clinical remission at Week 4 (NRI-C) – overall and Bio-IR (U-EXCEL ITT1 population)

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

----- Start of picture text -----
Populat Responder (NRI-C) Response rate difference vs PBO
ion
Missing
Treat Adj.
due to
ment N n (%) 95% CI [†] Diff. diff. 95% CI [§] P value
COVID-19,
(%) [‡]
n
All subjects
UPA
*** ********** **********
45 mg **** **** *********
PBO *** ********* **********

UPA
* ********* ********** *
45 mg **** ** ********* *
PBO ** ********* ********* *
**----- End of picture text -----*

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-EXCEL CSR (85). Note: CDAI clinical remission defined as CDAI score <150.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

B.3.6.1.2.2 CDAI clinical response (CR-100) at Week 2 and Week 12

At Week 2, a significantly greater proportion of patients achieved CDAI clinical response (see definition in Table 14) with upadacitinib compared with placebo (************************) (Table 21). A between-treatment difference of ***** was observed in the Bio-IR population (Table 21).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 21: CDAI clinical response (CR-100) at Week 2 (NRI C) – overall and Bio-IR (UEXCEL ITT1 population)

==> picture [452 x 198] intentionally omitted <==

----- Start of picture text -----
Population Responder (NRI-C) Response rate difference vs PBO
Treatment
Missing
Adj.
due to P
N n (%) 95% CI [†] Diff. diff. 95% CI [§]
COVID- value
(%) [‡]
19, n
All subjects
UPA 45 mg *** ********** **********
*** **** *********
PBO *** ********* **********
Prior biologic failure status
*
UPA 45 mg *** ********* ********** *
** ** ********* *
PBO ** ********* ********* *
**----- End of picture text -----*

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-EXCEL CSR (85). Note: CR-100 defined as decrease of ≥100 points in CDAI from baseline.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on nonresponder imputation incorporating multiple imputation to handle missing data due to COVID-19 or nonresponder imputation only if there are no missing data due to COVID-19.

At Week 12, a significantly greater proportion of patients achieved CDAI clinical response (CR-100) with upadacitinib compared with placebo (************************) (Table 22). A between-treatment difference of ***** for upadacitinib versus placebo was observed in the Bio-IR (Table 22).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 22: CDAI clinical response (CR-100) at Week 12 (NRI-C) – overall and Bio-IR (UEXCEL ITT1 population)

==> picture [452 x 180] intentionally omitted <==

----- Start of picture text -----
Population Responder (NRI-C) Response rate difference vs PBO
Treatment
Missing
Adj.
due to
N n (%) 95% CI [†] Diff. diff. 95% CI [§] P value
COVID-
(%) [‡]
19, n
All subjects
UPA 45 mg *** ********** **********
*** **** **********
PBO *** ********* **********
*
UPA 45 mg *** ********* ********** *
* ** ********** *
PBO ** ********* ********** *
**----- End of picture text -----*

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-EXCEL CSR (85). Note: CR-100 is defined as decrease of ≥100 points in CDAI from baseline.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on nonresponder imputation incorporating multiple imputation to handle missing data due to COVID-19 or nonresponder imputation only if there are no missing data due to COVID-19.

B.3.6.1.2.3 Endoscopic remission at Week 12

At Week 12, a significantly greater proportion of patients achieved endoscopic remission (see definition in Table 14) in the upadacitinib arm compared with the placebo arm (***********************) (Table 23). A between-treatment difference of ***** was observed for upadacitinib versus placebo in the Bio-IR population (Table 23).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 23: Summary of achievement of endoscopic remission at Week 12 (NRI-C) – overall and Bio-IR (U-EXCEL ITT1 population)

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CI, confidence interval; CMH, Cochran-MantelHaenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib.

Source: U-EXCEL CSR (85). Note: endoscopic remission defined as SES-CD ≤4 and ≥2-point reduction from baseline and no subscore >1 in any individual variable, as scored by central reviewer.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

B.3.6.1.2.4 Discontinuation of corticosteroid use and CDAI clinical remission at Week 12

Among subjects who were taking corticosteroids for CD at baseline, a significantly greater proportion discontinued corticosteroid use and achieved CDAI clinical remission at Week 12 with upadacitinib compared with placebo (**************; ********) (Table 24). A clear treatment effect was also observed in the Bio-IR population, with a between-treatment difference for upadacitinib versus placebo of ***** (Table 24).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 24: Discontinuation of corticosteroid use for CD and achievement of CDAI clinical remission at Week 12 in subjects taking corticosteroids for CD at baseline (NRI-C) – overall and Bio-IR (U-EXCEL ITT1 population)

==> picture [452 x 181] intentionally omitted <==

----- Start of picture text -----
Population Responder (NRI-C) Response rate difference vs PBO
Treatment
Missing
Adj.
due to
N n (%) 95% CI [†] Diff. diff. 95% CI [§] P value
COVID-
(%) [‡]
19, n
All subjects
UPA 45 mg *** ********* **********
*** **** ********** *******
PBO ** ********* *********
Bio-IR
UPA 45 mg ** ********* **********
** ** ********** *
PBO ** ******* ********* *
**----- End of picture text -----*

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-EXCEL CSR (85). Note: CDAI clinical remission defined as CDAI score <150.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

B.3.6.1.2.5 CD-related hospitalisation

Table 25 shows the occurrence of CD-related hospitalisations during the 12-week induction period of U-EXCEL. No significant difference was observed between the upadacitinib and placebo arms (**********************). A between-treatment difference of **** in favour of upadacitinib was observed in the Bio-IR population (Table 25).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 25: CD-related hospitalisations during the 12-week induction period (AO) – overall and Bio-IR (U EXCEL ITT1 population)

Abbreviations: AO, as observed; Bio-IR, biologic inadequate response/intolerance; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; non-Bio-IR, conventional therapy inadequate response/intolerance; NR, not reported; PBO, placebo; UPA, upadacitinib. Source: U-EXCEL CSR (85). Note: For all subjects in the ITT1 population, occurrence of hospitalisation due to CD was a binary variable. The value was ‘yes’ for subjects who had at least one hospitalisation due to CD during the 12-week induction period and ‘no’ for subjects who did not have any hospitalisation during the 12-week induction period. †95% CI for response rate is based on the normal approximation to the binomial distribution. ‡Risk difference = UPA – PBO.[ §] 95% CI for treatment difference was based on normal approximation of the binomial proportions. P-value was calculated according to the Chi-squared test or Fisher’s exact test if more than 20% of the cells have expected counts of less than 5.

B.3.6.1.2.6 EQ-5D-5L at Week 4 and Week 12

Statistically significant improvements in the EQ-5D-5L index value and EQ-5D VAS were observed with upadacitinib compared with placebo. At Week 4, the improvement from baseline (LS mean) was ***** in the upadacitinib arm and ***** in the placebo arm with a statistically significant between-treatment difference of **************** (Table 26). At Week 12, the improvement from baseline was ***** and ***** in the upadacitinib and placebo arms, respectively, resulting in a significant between-treatment difference of **************** (Table 26).

Similar results were observed for the EQ-5D VAS score; at Week 4, subjects in the upadacitinib arm had a greater improvement from baseline compared with the placebo arm (LS mean ************); a between-treatment difference of ************** (Table 26). At Week 12, the improvement from baseline was **** and **** with upadacitinib and placebo, respectively, resulting in a significant between-treatment difference of ************** (Table 26).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 26: Change from baseline in EQ-5D-5L index value and EQ-5D VAS at Weeks 4 and 12 (MMRM) (U-EXCEL ITT1 population)

==> picture [452 x 312] intentionally omitted <==

----- Start of picture text -----
Outcome Between group difference vs
Within group change from baseline
PBO
Timepoint
Treatment Baseline Visit LS
N 95% CI LS mean 95% CI P value
mean mean mean
EQ-5D-5L index value
Week 4
UPA 45 mg *** ***** ***** ***** ************
***** ************ ******
PBO *** ***** ***** ***** ************
Week 12
UPA 45 mg *** ***** ***** ***** ************
***** ************ *******
PBO *** ***** ***** ***** ************
EQ-5D VAS
Week 4
UPA 45 mg *** **** **** **** **********
*** ******** ******
PBO *** **** **** **** *********
Week 12
UPA 45 mg *** **** **** **** **********
*** ********* *******
PBO *** **** **** **** *********
----- End of picture text -----

Abbreviations: CI, confidence interval; CSR, clinical study report; EQ-5D-5L, EuroQol-5 Dimensions 5-level; LS, least squares; MMRM, mixed effect model repeat measurement; PBO, placebo, SES-CD, Simple Endoscopic Score for Crohn’s Disease; UPA, upadacitinib; VAS, visual analogue scale. Source: U-EXCEL CSR (85). MMRM was the mixed effect model repeat measurement with baseline value, stratification factors (baseline corticosteroid use [yes or no], endoscopic disease severity [SES-CD <15 or ≥15], and number of prior biologics failed [0, 1, or >1]), treatment visit, treatment-by-visit interaction included in the model. An unstructured covariance matrix was used.

B.3.6.2 U-EXCEED

B.3.6.2.1 Co-primary efficacy outcome: proportion of subjects with CDAI clinical remission and endoscopic response at Week 12

In U-EXCEED, the co-primary endpoint of CDAI clinical remission and endoscopic response at Week 12 was met. At Week 12, a significantly greater proportion of subjects achieved CDAI clinical remission in the upadacitinib arm compared with the

placebo arm (***** vs ***********) (Table 27). A clear treatment effect with upadacitinib

versus placebo was also observed in the ≤1 biologic failure and >1 biologic failure populations, with between-treatment differences of **************** respectively (Table 27).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 27: CDAI clinical remission at Week 12 (NRI-C) – overall and by prior biologic failure status (U-EXCEED ITT1 population)

==> picture [452 x 252] intentionally omitted <==

----- Start of picture text -----
Population Responder (NRI-C) Response rate difference vs PBO
Treatment
Missing
Adj.
due to
N n (%) 95% CI [†] Diff. diff. 95% CI [§] P value
COVID-
(%) [‡]
19, n
All subjects
UPA 45 mg *** ********** **********
*** **** **********
PBO *** ********* **********
Prior biologic failure status

UPA 45 mg *** ********* **********
* ** ********* **
PBO ** ********* **********
*********
UPA 45 mg *** ********* **********
*** ** ********* *
PBO *** ********* ********* *
**----- End of picture text -----*

Abbreviations: CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-EXCEED CSR (86). Note: CDAI clinical remission defined as CDAI score <150.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

At Week 12, a significantly greater proportion of subjects achieved endoscopic response in the upadacitinib arm compared with the placebo arm (***********************) (Table 28). A clear treatment effect with upadacitinib versus placebo was also observed in the ≤1 biologic failure and >1 biologic failure populations, with between-treatment differences of **************** respectively (Table 28).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 28: Endoscopic response at Week 12 (NRI-C) – overall and by prior biologic failure status (U-EXCEED ITT1 population)

==> picture [452 x 225] intentionally omitted <==

----- Start of picture text -----
Population Responder (NRI-C) Response rate difference vs PBO
Treatment Missing
Adj.
due to
N n (%) 95% CI [†] Diff. diff. 95% CI [§] P value
COVID-
(%) [‡]
19, n
All subjects
*** ********** **********
UPA 45 mg **** **** **********
PBO *** ******* ********
Prior biologic failure status

* ********* ********** *
UPA 45 mg **** ** **********
PBO ** ******* ********
********
*** ********* ********** *
UPA 45 mg **** ** **** ****
PBO *** ******* ******** *
----- End of picture text -----

Abbreviations: CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-EXCEED CSR (86). Note: endoscopic response defined as decrease in SES-CD of >50% from baseline of the induction study or for subjects with an SES-CD of 4 at baseline, ≥2-point reduction from baseline, as scored by central reviewer.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

B.3.6.2.2 Secondary efficacy outcomes – U-EXCEED

B.3.6.2.2.1 CDAI clinical remission at Week 4

At Week 4, approximately *** of subjects in the upadacitinib arm achieved CDAI clinical remission. A significantly greater proportion of subjects achieved CDAI clinical remission with upadacitinib compared with placebo at this timepoint (************************) (Table 29). A significant difference in favour of upadacitinib was also observed at Week 2 (see Appendix J).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 29: CDAI clinical remission at Week 4 (NRI-C; U-EXCEED ITT1 population)

Abbreviations: CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-EXCEED CSR (86). Note: CDAI clinical remission defined as CDAI score <150.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

B.3.6.2.2.2 CDAI clinical response (CR-100) at Week 2 and Week 12

Significantly greater proportions of patients achieved CDAI clinical response with upadacitinib compared with placebo at Week 2 and Week 12 (*********************************, respectively: both ********) (Table 30).

Table 30: CDAI clinical response (CR-100) at Week 2 and Week 12 (NRI-C; U-EXCEED ITT1 population)

==> picture [454 x 291] intentionally omitted <==

----- Start of picture text -----
Treatment Responder (NRI-C) Response rate difference vs PBO
Missing
Adj.
due to
N n (%) 95% CI [†] Diff. diff. 95% CI [§] P value
COVID-
(%) [‡]
19, n
Week 2
UPA 45 mg *** ********** **********
*** **** ********** *******
PBO *** ********* *********
Week 12
UPA 45 mg *** ********** **********
** **** ********** *******
PBO *** ********* ********** *
Abbreviations: CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel;
CR-100, clinical response 100; COVID-19, Coronavirus Disease 2019; CR-100, clinical response 100; CSR,
clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while
incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib.
Source: U-EXCEED CSR (86). Note: CR-100 defined as decrease of ≥100 points in CDAI from baseline. [†] 95% CI
for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if
there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if
there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is
calculated based on the CMH test adjusting for stratification factors. [§] 95% CI for adjusted difference and p-value
are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-
responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-
responder imputation only if there are no missing data due to COVID-19.
----- End of picture text -----

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.6.2.2.3 Endoscopic remission at Week 12

At Week 12, a significantly greater proportion of patients achieved endoscopic remission in the upadacitinib arm compared with the placebo arm (***********************) (Table 31).

Table 31: Endoscopic remission at Week 12 (NRI-C; U-EXCEED ITT1 population)

Abbreviations: CI, confidence interval; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-EXCEED CSR (86). Note: endoscopic remission defined as SES-CD ≤4 and ≥2-point reduction from baseline and no subscore >1 in any individual variable, as scored by central reviewer.[ †] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

B.3.6.2.2.4 Discontinuation of corticosteroid use and CDAI clinical remission at Week 12

Among subjects taking corticosteroids for CD at baseline, a significantly greater proportion discontinued corticosteroids and achieved CDAI clinical remission with upadacitinib compared with placebo at Week 12 (*************** ********) (Table 32).

Table 32: Discontinuation of corticosteroid use and achievement of CDAI clinical remission at Week 12 in subjects taking corticosteroids for CD at baseline (NRI-C; U-EXCEED ITT1 population)

Abbreviations: CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-EXCEED CSR (86). Note: CDAI clinical remission defined as CDAI score <150.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.6.2.2.5 CD-related hospitalisation

Table 33 shows CD-related hospitalisations occurring during the 12-week induction period of U-EXCEED. No significant differences were observed between the treatment groups (Table 33).

Table 33: CD-related hospitalisation during the 12-week induction period (AO; U- EXCEED ITT1 population)

==> picture [452 x 108] intentionally omitted <==

----- Start of picture text -----
Responder Response rate difference vs PBO
Missing
Treatment Adj.
due to
N n (%) 95% CI Diff. diff. 95% CI P value
COVID-
(%)
19, n
UPA 45 mg *** ******** ********
** ** ********* ******
PBO *** ******** *********
----- End of picture text -----**

Abbreviations: AO, as observed; CI, confidence interval; COVID-19, coronavirus disease 2019; CSR, clinical study report; ITT, intention to treat; NR, not reported; PBO, placebo; UPA, upadacitinib. Source: U-EXCEED CSR (86). Note: For all subjects in the ITT1 population, occurrence of hospitalisation due to CD was a binary variable. The value was ‘yes’ for subjects who had at least one hospitalisation due to CD during the 12-week induction period and ‘no’ for subjects who did not have any hospitalisation during the 12-week induction period. †95% CI for response rate is based on the normal approximation to the binomial distribution. ‡Risk difference = UPA – PBO.[ §] 95% CI for treatment difference was based on normal approximation of the binomial proportions. P-value was calculated according to the Chi-squared test or Fisher’s exact test if more than 20% of the cells have expected counts of less than 5.

B.3.6.2.2.6 EQ-5D-5L at Week 4 and Week 12

Significant improvements in the EQ-5D-5L index value and EQ-5D VAS score were observed with upadacitinib compared with placebo at both Week 4 and Week 12. At Week 4, the improvement from baseline (LS mean) was ***** in the upadacitinib arm and ***** in the placebo arm; a significant between-treatment difference of **************** (Table 34). At Week 12, the improvement from baseline was ***** and ***** with upadacitinib and placebo, respectively; a between-treatment difference ******************* (Table 34).

Similar results were observed for the EQ-5D VAS score. At Week 4, subjects in the upadacitinib arm had a greater improvement from baseline compared with the placebo arm (LS mean ***********); a between-treatment difference of ************** (Table 34). At Week 12, the improvement from baseline was **** and *** with upadacitinib and placebo, respectively; a between-treatment difference of *************** (Table 34).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 34: Change from baseline in EQ-5D-5L index value and EQ-5D VAS at Weeks 4 and 12 (MMRM; U-EXCEED ITT1 population)

==> picture [452 x 312] intentionally omitted <==

----- Start of picture text -----
Score Between group difference vs
Within group change from baseline
PBO
Timepoint
Treatment Baseline Visit LS
N 95% CI LS mean 95% CI P value
mean mean mean
EQ-5D-5L index value
Week 4
UPA 45 mg *** ***** ***** ***** ************
***** ************ ******
PBO *** ***** ***** ***** ************
Week 12
UPA 45 mg *** ***** ***** ***** ************
***** ************ *******
PBO *** ***** ***** ***** ************
EQ-5D VAS
Week 4
UPA 45 mg *** **** **** **** **********
*** ********* *******
PBO *** **** **** *** *********
Week 12
UPA 45 mg *** **** **** **** **********
**** ********* *******
PBO *** **** **** *** *********
----- End of picture text -----

Abbreviations: CI, confidence interval; CSR< clinical study report; EQ-5D-5L, EuroQol-5 Dimensions 5-Levels; LS, least squares; MMRM, mixed effect model repeat measurement; PBO, placebo; SES-CD, Simple Endoscopic Score for Crohn’s disease; UPA, upadacitinib; VAS, visual analogue scale. Source: U-EXCEED CSR (86). MMRM was the mixed effect model repeat measurement with baseline value, stratification factors (baseline corticosteroid use [yes or no], endoscopic disease severity [SES-CD <15 or ≥15], and number of prior biologics failed [>1 or ≤1]), treatment visit, treatment-by-visit interaction included in the model. An unstructured covariance matrix was used.

B.3.6.3 U-ENDURE

B.3.6.3.1 Co-primary efficacy outcome: proportion of subjects with CDAI clinical remission and endoscopic response at Week 52

In U-ENDURE, the co-primary endpoints of CDAI clinical remission and endoscopic response at Week 52 were met. A significantly greater proportion of subjects achieved CDAI clinical remission with upadacitinib 30 mg and upadacitinib 15 mg compared with placebo (**********************************) (Table 35). In the Bio-IR population, the between-treatment difference with upadacitinib 30 mg and upadacitinib 15 mg versus placebo was ***** and ****** respectively (Table 35).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 35: CDAI clinical remission at Week 52 (NRI-C) – overall and Bio-IR (U-ENDURE ITT1 population)

==> picture [452 x 216] intentionally omitted <==

----- Start of picture text -----
Population Responder (NRI-C) Response rate difference vs PBO
Treatment
Missing
Adj.
due to
N n (%) 95% CI [†] Diff. diff. 95% CI [§] P value
COVID-
(%) [‡]
19, n
All subjects
UPA 30 mg *** ********* ********** * **** **** ********** *******
UPA 15 mg *** ********* ********** * **** **** ********** *******
PBO *** ********* ********* * ** ** **
****
UPA 30 mg *** ********* ********** * **** ** ********** **
UPA 15 mg *** ********* ********** * **** ** **********
PBO *** ********* ********* * ** ** **
----- End of picture text -----

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NA, not applicable; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-ENDURE CSR (87). Note: CDAI clinical remission defined as CDAI score <150.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

At Week 52, a significantly greater proportion of patients achieved endoscopic response with upadacitinib 30 mg and 15 mg compared with placebo (**************************************) (Table 36). In the Bio-IR population, the betweentreatment difference with upadacitinib 30 mg and upadacitinib 15 mg versus placebo was ***** and *****, respectively (Table 36).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 36: Endoscopic response at Week 52 (NRI-C) – overall and Bio-IR (U-ENDURE ITT1 population)

==> picture [452 x 216] intentionally omitted <==

----- Start of picture text -----
Population Responder (NRI-C) Response rate difference vs PBO
Treatment
Missing
Adj.
due to
N n (%) 95% CI [†] Diff. diff. 95% CI [§] P value
COVID-
(%) [‡]
19, n
All subjects
UPA 30 mg *** ********* ********** * **** **** ********** *******
UPA 15 mg *** ********* ********** * **** **** ********** *******
PBO *** ******** ********* * ** ** **
****
UPA 30 mg *** ********* ********** * **** ** ********** **
UPA 15 mg *** ********* ********** * **** ** **********
PBO *** ******* ******** * ** ** **
----- End of picture text -----

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CI, confidence interval; CMH, Cochran-MantelHaenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NA, not applicable; NR, not reported; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-ENDURE CSR (87). Note: endoscopic response defined as decrease in SES-CD of >50% from baseline of the induction study or for subjects with an SES-CD of 4 at baseline, ≥2-point reduction from baseline, as scored by central reviewer.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

B.3.6.3.2 Secondary efficacy outcomes – U-ENDURE

B.3.6.3.2.1 CDAI clinical response (CR-100) at Week 52

At Week 52, ************************** receiving upadacitinib 30 mg achieved CDAI clinical response (CR-100); a significantly greater proportion than those receiving placebo (************************) (Table 37). In the upadacitinib 15 mg arm, ***** of patients achieved CDAI clinical response, which was also significantly greater than the placebo arm (********) (Table 37).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 37: CDAI clinical response (CR-100) at Week 52 (NRI-C; U-ENDURE ITT1 population)

Abbreviations: CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CR, clinical response; CSR, clinical study report; ITT, intention to treat; NA, not applicable; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-ENDURE CSR (87). Note: CR-100 defined as decrease of ≥100 points in CDAI from baseline.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on nonresponder imputation incorporating multiple imputation to handle missing data due to COVID-19 or nonresponder imputation only if there are no missing data due to COVID-19.

B.3.6.3.2.2 Endoscopic remission at Week 52

Statistically significantly more patients achieved endoscopic remission at Week 52 with upadacitinib 30 mg or 15 mg compared with placebo (**************************************) (Table 38).

Table 38: Endoscopic remission at Week 52 (NRI-C; U-ENDURE ITT1 population)

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.6.3.2.3 CDAI clinical remission and endoscopic remission at Week 52

At Week 52, almost ******************************* receiving upadacitinib 30 mg achieved both CDAI clinical remission and endoscopic remission; a significantly greater proportion than in the placebo arm **************** (Table 39). In the upadacitinib 15 mg arm, ***** of subjects achieved this outcome, which was also significantly greater than the placebo arm (********) (Table 39).

Table 39: CDAI clinical remission and endoscopic remission at Week 52 (NRI-C; U- ENDURE ITT1 population)

Abbreviations: CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CR, clinical response; CSR, clinical study report; ITT, intention to treat; NA, not applicable; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; SES-CD, Simple Endoscopic Score for Crohn’s Disease; UPA, upadacitinib. Source: U-ENDURE CSR (87). Note: CDAI clinical remission defined as CDAI score <150. endoscopic remission defined as SES-CD ≤4 and ≥2-point reduction from baseline and no subscore >1 in any individual variable, as scored by central reviewer.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

B.3.6.3.2.4 Discontinuation of corticosteroid use and CDAI clinical remission at Week 52

Table 40 shows the proportion of subjects who were steroid-free for at least 90 days prior to Week 42 and achieved CDAI clinical remission at Week 52. The proportion of patients achieving this outcome was ***** in the upadacitinib 30 mg arm and ***** in the upadacitinib 15 mg arm, which was significantly better than in the placebo arm (*****; **********) (Table 40).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 40: Without corticosteroid use for CD ≥90 days prior to Week 52 and achievement of CDAI clinical remission at Week 52 (NRI-C; U-ENDURE ITT-1 population)

Abbreviations: CD, Crohn’s Disease; CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NA, not applicable; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib. Source: U-ENDURE CSR (87). Note: CDAI clinical remission defined as CDAI score <150.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

Table 41 shows the proportion of subjects who were receiving corticosteroids for CD

and induction baseline and who were steroid-free for ≥90 days prior to Week 52 and achieved CDAI clinical remission at Week 52. The proportion of subjects achieving this outcome was the same with both maintenance doses of upadacitinib () and statistically significantly better than the placebo arm ( both p<0.0001) (Table 41).

Table 41: Discontinuation of corticosteroid use for CD ≥90 days prior to Week 52 and achievement of CDAI clinical remission at Week 52 in subjects taking corticosteroids for CD at induction baseline (NRI-C; U-ENDURE ITT1 population)

Abbreviations: CD, Crohn’s Disease; CDAI, Crohn’s Disease Activity Index; CI, confidence interval; CMH, Cochran-Mantel-Haenszel; COVID-19, Coronavirus Disease 2019; CSR, clinical study report; ITT, intention to treat; NA, not applicable; NRI-C, non-responder imputation while incorporating multiple imputation to handle missing data due to COVID-19; PBO, placebo; UPA, upadacitinib.

Source: U-ENDURE CSR (87). Note: CDAI clinical remission defined as CDAI score <150.[†] 95% CI for response rate is the synthetic result based on Student’s t-distribution from PROC MIANALYZE procedure if there are missing data due to COVID-19 or is based on the normal approximation to the binomial distribution if there are no missing data due to COVID-19. ‡Risk difference = (UPA – PBO). Adjusted risk difference is calculated based on the CMH test adjusting for stratification factors.[§] 95% CI for adjusted difference and p-value are calculated according to the CMH test adjusted for stratification factors. The calculations are based on non-responder imputation incorporating multiple imputation to handle missing data due to COVID-19 or non-responder imputation only if there are no missing data due to COVID-19.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.6.3.2.5 CD-related hospitalisation during the 52-week maintenance period

Table 42 shows the occurrence of CD-related hospitalisations during the 52-week maintenance period. No statistically significant differences were observed between the treatment groups (Table 42).

Table 42: CD-related hospitalisation during the 52-week maintenance period (AO; U-ENDURE ITT1 population)

B.3.6.3.2.6 EQ-5D-5L at Week 52

Statistically significant improvements in the EQ-5D-5L index value and EQ-5D VAS were observed with upadacitinib 30 mg compared with placebo at Week 52. The improvement from baseline (LS mean) was ***** in the upadacitinib 30 mg arm and ***** in the placebo arm, with a statistically significant between-treatment difference of **************** (Table 43). The improvement from baseline with upadacitinib 15 mg was *****, which was not statistically significantly different from placebo (LS mean difference ***************) (Table 43).

Similar results were observed for the EQ-5D VAS score. Subjects in the upadacitinib 30 mg arm had a significantly greater improvement from baseline compared with the placebo arm (LS mean ************), giving a between-treatment difference of ************** (Table 43). Similarly, the improvement from baseline with upadacitinib 15 mg was ****, which was significantly different from placebo (LS mean difference *************) (Table 43).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 43: Change from induction baseline in EQ-5D-5L at Week 52 (MMRM; U-ENDURE ITT1 population)

==> picture [452 x 181] intentionally omitted <==

----- Start of picture text -----
Outcome Between group difference vs
Within group change from baseline
Treatment PBO
Baseline Visit LS LS
N 95% CI 95% CI P value
mean mean mean mean
EQ-5D-5L index value
UPA 30 mg ** ***** ***** ***** ************ ***** ************ ******
UPA 15 mg ** ***** ***** ***** ************ ***** ************* ******
PBO ** ***** ***** ***** ************ ** ** **
EQ-5D VAS
UPA 30 mg ** **** ***** ***** *********** *** ********* ******
UPA 15 mg ** **** ***** ***** *********** *** ********** *******
PBO ** **** ***** **** ********** ** **
----- End of picture text -----**

Abbreviations: Bio-IR, biologic inadequate response/intolerance; CI, confidence interval; CSR, clinical study report; DB, double-blind; EQ-5D-5lL, EuroQol-5 dimensions-5 Levels; LS, least squares; MMRM, mixed effect model repeat measurement; non-Bio-IR, conventional therapy inadequate response/intolerance; OL, open-label; PBO, placebo, PRO, patient-reported outcome; UPA, upadacitinib; VAS, visual analogue scale. Source: U-ENDURE CSR (87). Note: MMRM is the mixed effect model repeat measurement with induction baseline value, Week 0 value, stratification factors (prior induction population [non-Bio-IR, DB bio-IR, OL bio-IR), PRO clinical remission [yes/no], and endoscopic response status [yes/no]), treatment visit, visit, treatment-by-visit interaction included in the model. Induction baseline was defined as the last non-missing observation prior to the first dose of study drug in U-EXCEL/U-EXCEED. An unstructured covariance matrix was used.

Additional outcomes beyond those presented in Section B.3.6.3 are presented in Appendix J and listed below:

• CDAI clinical remission at Week 2

• PRO clinical remission at Week 52

• CDAI clinical remission at Week 52 in subjects who achieved CDAI clinical remission at Week 0

• PRO clinical remission in subjects at Week 52 in subjects who achieved PRO clinical remission at Week 0

• PRO clinical remission and endoscopic remission at Week 52

• Change from induction baseline in FACIT-F total score at Week 52

• Change from induction baseline in IBDQ total score at Week 52

• Resolution of EIMs at Week 52 in subjects with any EIMs at induction baseline

• Without corticosteroids/discontinuation of corticosteroids and CDAI clinical remission at Week 52

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.7 Subgroup analysis

As described in Section B.1, this submission presents evidence for upadacitinib in the BF (Bio-IR) population, a subpopulation of the anticipated licensed population for upadacitinib. Data on the CCF (non-Bio-IR) subpopulation are presented in Appendix J for completeness.

In addition, subgroup data are presented for subjects with prior TNF-alpha inhibitor failure (failed 1 or >1 TNF-alpha inhibitor) and for subjects who had draining fistulas at baseline (see Appendix E).

B.3.8 Meta-analysis

The absence of head-to-head data prevented a standard meta-analysis of RCTs from being performed. Instead, a comprehensive network meta-analysis (NMA) was conducted; this enabled comparisons with other biologic therapies included in the NICE scope and allowed for more precise estimates of treatment effects to be calculated compared with a naïve comparison of trials. The NMA is presented in Section B.3.9.

B.3.9 Indirect and mixed treatment comparisons

B.3.9.1 Methodology

Full details of the methodology for the indirect/mixed treatment comparison are provided in Appendix D. A brief overview of the methodology is presented in Section B.3.9.1.

B.3.9.1.1 Analysis scope

As discussed in Section B.3.1, an SLR was conducted to identify all relevant clinical evidence on the efficacy and safety of upadacitinib and comparators for the treatment of moderately to severely active CD in adults. In the absence of head-to-head RCTs between all comparators specified in the NICE scope, an NMA was performed to assess the relative efficacy of upadacitinib compared with relative comparators (ustekinumab, vedolizumab) in adults with moderately to severely active CD who experienced BF. The Bio-IR population in the upadacitinib clinical trials is considered

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

analogous to the BF population (see Section B.3.3.1 for more details). NMAs were also performed for the non-Bio-IR population (considered analogous to the CCF population) and results are presented in Appendix L. The methodology of the SLR that identified studies used in the NMAs is described in Appendix D.

B.3.9.1.2 Study selection for NMA

As described in Appendix D, a total of 290 records met the inclusion criteria of the clinical SLR, reporting on 71 original studies. After applying the inclusion/exclusion criteria, 10 unique trials reported by 11 records were included in the NMA. A list of all studies excluded from the NMA (including reason for exclusion) is available in Appendix D.

The interventions and doses of interest included in the NMAs for the induction and maintenance phases are presented in Appendix D. For each of the interventions, only licensed UK doses were included in the analysis. A summary of the trials used to conduct the NMA is presented in Table 44.

Table 44: Summary of trials used in the NMA

Abbreviations: ADA, adalimumab; BF, biologic failure; CCF, conventional care failure; CDAI, Crohn’s Disease Activity Index; CR-100, clinical response ≥100-point decrease from baseline in CDAI score; IFX, infliximab; Ind, induction; IV, intravenous; Maint, maintenance; NA, not applicable; NMA, network meta-analysis; QxW, every x weeks; SC, subcutaneous; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab. † CSR data used in NMA.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.9.1.3 Outcomes of interest

Outcomes assessed using NMA included CDAI clinical remission and CDAI clinical response (CR-100), as defined in Table 45. The definitions for these outcomes used in the NMA align with those used in the upadacitinib clinical trials (see Section B.3.3.3). CDAI outcomes were of interest as they facilitate comparison with comparator therapies. Furthermore, these outcomes have historically been used in CD clinical studies and their use in the NMA is consistent with clinical trials of ustekinumab and vedolizumab in CD (unlike endoscopic outcomes, which were not universally assessed in the ustekinumab and vedolizumab trials) (57, 81, 82). CDAI outcomes were generally assessed after an induction phase of 4 to 12 weeks and a maintenance phase of 44 to 60 weeks (see Appendix D for more details).

Table 45: CDAI outcomes assessed in NMA

Abbreviations: CDAI, Crohn’s Disease Activity Index; CR, clinical response; NMA< network-meta-analysis. Note: CR-100 defined as decrease of ≥100 points in CDAI from baseline.

In the NMAs, CDAI clinical remission and CDAI clinical response (CR-100) were assessed for induction trials, while CDAI clinical remission was assessed for maintenance trials (clinical response was not generally reported in the maintenance trials as only patients with a clinical response at the end of induction continued to the maintenance trials). Outcomes were assessed by CCF and BF subgroups (Table 46).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 46: Trials reporting CDAI outcomes used in the NMA

Abbreviations: BF, biologic failure; CCF, conventional care failure; CDAI, Crohn’s Disease Activity Index.

In addition to CDAI outcomes, NMAs were conducted for serious AEs and discontinuation due to AEs. Most trials did not report safety outcomes by subpopulation (CCF or BF) and therefore safety NMAs were conducted in the overall population only (Table 47, for full methods and results, see Appendix D).

Table 47: Trials reporting safety outcomes assessed in the NMA

Abbreviations: ADA, adalimumab; AE, adverse event; IFX, infliximab; Ind., induction; IV, intravenous; Maint., maintenance; QxW, every x weeks; SC, subcutaneous; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.9.1.4 Summary of trials included in NMA

A summary of the trials included in the base case and sensitivity analysis NMAs, as well as the reporting of outcomes from each study considered for inclusion, is detailed in Appendix D.

B.3.9.1.5 Overview of NMA methodology

A Bayesian NMA approach was selected to indirectly compare upadacitinib with relevant comparators using an evidence base of published RCTs. Binary outcomes were modelled with a binomial likelihood and a risk difference (RD) link (NICE decision support unit [DSU] technical support document [TSD] 2 (97, 98)). An NMA feasibility assessment was performed based on the included RCTs; full details of the methodology and feasibility assessment are presented in Appendix D.

B.3.9.1.5.1 Model selection

Model selection was made after comparing model fit statistics, leverage plots, and density plots of posterior SDs for each set of two risk difference models (FE and RE; see Appendix D for more details). Due to heterogeneity in reported placebo rates in the included trials, a risk difference approach was utilised rather than an unadjusted approach as this approach was less sensitive to differences in placebo response rates; see Section B.3.9.1.5.3 for further information. FE models were selected for all outcomes in the base case; justification for this approach is provided in Appendix D with the main reason being the small number of trials eligible for inclusion in each network leading to implausible estimates of the between-study standard deviation when RE models were used.

B.3.9.1.5.2 Upadacitinib post-hoc CDAI 220–450 inclusion criterion

For robust NMAs, a key assumption is that populations are sufficiently similar and comparable across the included trials. All trials included in the NMAs applied a similar CDAI score inclusion criterion, except for the upadacitinib studies. Instead of the CDAI score, the upadacitinib induction trials (U-EXCEL and U-EXCEED) used two key CD symptoms of abdominal pain and stool frequency as inclusion criteria (both are components of CDAI; see Appendix K for full details). To align with the CDAI score inclusion criterion (220–450) applied in the other CD trials in the NMA, post-hoc

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

analyses were conducted to obtain upadacitinib trial results in a population with a baseline CDAI score of 220–450.

After applying the post-hoc inclusion analysis, approximately 80% of patients were retained in each arm of the upadacitinib trials. In general, baseline CDAI values for the whole population and for the post-hoc CDAI 220–450 restricted population of U- EXCEL and U-EXCEED were comparable to the values observed in other CD trials (Table 48). As the trials had a re-randomisation design, only patients who were included in the induction trials (and had a clinical response) moved to the maintenance trials. Therefore, only the baseline CDAI scores for the induction studies are presented in Table 48. Baseline characteristics and treatment efficacy also remained similar between the whole trial population and the post-hoc CDAI 220–450 restricted population (see Appendix C). Therefore, data from the CDAI 220–450 restricted population from the upadacitinib trials was used in the base case NMA to maximise the potential comparability of the trial populations.

Table 48: CDAI inclusion criteria and mean baseline CDAI scores of included BF induction populations

Abbreviations: BF, biologic failure; CCF, conventional care failure; CDAI, Crohn’s Disease Activity Index; IV, intravenous; PBO, placebo; SD, standard deviation; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab. *indicates that baseline CDAI is only reported for the overall population and not by CCF/BF subgroup.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.9.1.5.3 Risk-difference approach

The risk-difference approach was selected to minimise the impact of different placebo rates which were observed across the included trials (efficacy data for the placebo groups across the different trials are presented in Appendix D) and because of the intuitive presentation of results when generated on the risk-difference scale. The implications of this approach are discussed in further detail in section B.3.9.3.1.

B.3.9.1.6 Sensitivity analysis methodology

Sensitivity analyses were conducted using RE NMA (FE was used in all base case analyses) and the results are presented in Appendix L. Overall, the RE NMA results were comparable to the FE model results, although there were larger CrIs across most comparisons. This is to be expected as the RE NMA incorporates between-study differences in its efficacy estimates.

B.3.9.1.7 NMA networks

Separate analyses were performed for the induction and maintenance trials for the BF population. In all networks, placebo was the common comparator. NMAs were also performed for the CCF population and are presented in Appendix L.

B.3.9.1.7.1 BF population

The BF population networks for the induction and maintenance phases are presented in Figure 6 and Figure 7, respectively.

Figure 6: Network diagram of included induction studies in the BF population

==> picture [336 x 162] intentionally omitted <==

----- Start of picture text -----
UST
GEMINI 2 (BF) U-EXCEL (CDAI restricted BF)
VDZ IV PBO UPA 45
GEMINI 3 (BF) U-EXCEED (CDAI restricted BF)
Watanabe 2020 (BF)
UNITI-1 (Pure BF)
----- End of picture text -----

Abbreviations: BF, biologic failure; IV, intravenous; PBO, placebo; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Figure 7: Network diagram of included maintenance studies in the BF population

==> picture [229 x 223] intentionally omitted <==

Abbreviations: BF, biologic failure; IV, intravenous; PBO, placebo; SC, subcutaneous; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab.

B.3.9.2 Results

The following sections report results from the NMA for CDAI outcomes, which have been used to inform the assumption that upadacitinib is likely to provide similar or greater health benefits than its comparators. The results are reported as RD with credible intervals (CrI, 95% CrIs presented throughout). Please note that ‘significance’ in these results is defined by CrIs not crossing zero; these analyses should not be interpreted in a frequentist manner. In addition, safety outcomes are presented to support comparison of upadacitinib with relevant active comparators.

For each combination of outcome and NMA, league tables of the relative effect estimate for all possible pair-wise comparisons are presented.

Additional results for base-case analysis are presented in Appendix L:

• Relative effect estimates for each relevant comparator versus placebo on the RD scale

• Predicted absolute outcomes for each treatment

• Surface Under the Cumulative RAnking (SUCRA) values for each treatment (99)[3]

3 SUCRA would be 100% when a treatment is certain to be the best and 0% when a treatment is certain to be the worst

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.9.2.1 Base-case analysis – induction CDAI-100 clinical response (CR-100)

Table 49 presents the base-case NMA results for CDAI clinical response (CR-100) with induction upadacitinib versus comparators in a BF population.

The NMA results show a RD of ********************** for upadacitinib versus placebo,

showing that there is a ***** greater probability of CDAI clinical response in patients receiving upadacitinib versus placebo.

Table 49: Results for CDAI clinical response (CR-100) in BF induction NMA (FE model)

==> picture [447 x 80] intentionally omitted <==

----- Start of picture text -----
VDZ IV UPA UST PBO
PBO ***************** ******************* ******************
UST ****************** ****************** *********************
UPA ********************* ********************* **********************
VDZ IV ****************** ***************** ********************
----- End of picture text -----

Abbreviations: BF, biologic failure; CDAI, Crohn’s Disease Activity Index; FE, fixed effects; IV, intravenous; NMA, network meta-analysis; PBO, placebo; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab. Asterisks indicate risk difference scale credible intervals do not cross zero, which may be considered ‘significant’.

B.3.9.2.2 Base case analysis – induction CDAI clinical remission

Table 50 presents the base case NMA results for CDAI clinical remission with induction upadacitinib versus comparators in a BF population.

The NMA results show a RD of ********************** for upadacitinib versus placebo, showing that there is a ***** greater probability of CDAI clinical remission in patients receiving upadacitinib versus placebo.

Table 50: Results for CDAI clinical remission in BF induction NMA (FE model)

==> picture [447 x 83] intentionally omitted <==

----- Start of picture text -----
VDZ IV UPA UST PBO
PBO **************** ******************* *****************
UST ****************** ****************** ********************
UPA ********************** ********************* **********************
VDZ IV ******************* ***************** *****************
----- End of picture text -----

Abbreviations: BF, biologic failure; CDAI, Crohn’s Disease Activity Index; FE, fixed effects; IV, intravenous; NMA, network meta-analysis; PBO, placebo; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab. Asterisks indicate risk difference scale credible intervals do not cross zero, which may be considered ‘significant’.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.9.2.3 Base case analysis – maintenance CDAI clinical remission

Table 51 presents the base case NMA results for CDAI clinical remission with maintenance upadacitinib versus comparators in a BF population.

The NMA results show a RD of ****************** for upadacitinib 30 mg and ****************** for upadacitinib 15 mg versus placebo, showing that there is a ***** and ***** greater probability of CDAI clinical remission in patients receiving upadacitinib 30 mg and 15 mg versus placebo, respectively.

Table 51: Results for CDAI clinical remission in BF maintenance NMA (FE model)

==> picture [499 x 244] intentionally omitted <==

----- Start of picture text -----
VDZ IV Q4W VDZ IV Q8W UST Q12W UST Q8W UPA 15 UPA 30 VDZ SC PBO
*************** *************** *************** *************** *************** ***************
PBO
* *** *** *** **** **** ***
** *************** *************** *************** *************** *************** ***************
VDZ SC
**** **** **** **** *** ***
********* *************** *************** *************** *************** *************** ***************
UPA 30
****** ****** ****** ****** ****** ******
******** *************** *************** *************** *************** *************** ***************
UPA 15
*** **** **** **** *** **** *******
UST *************** *************** *************** *************** *************** *************** ***************
Q8W *** *** **** *** *** *** ****
UST *************** *************** *************** *************** *************** *************** ***************
Q12W *** *** *** *** *** *** ****
VDZ IV *************** *************** *************** *************** *************** *************** ***************
Q8W **** **** **** *** *** *** ******
VDZ IV *************** *************** *************** *************** *************** *************** ***************
Q4W **** ***** ***** *** **** **** *******
Abbreviations: BF, biologic failure; CDAI, Crohn’s Disease Activity Index; FE, fixed effects; IV, intravenous; NMA,
network meta-analysis; PBO, placebo; QxW, every x weeks; SC, subcutaneous; UPA, upadacitinib; UST,
ustekinumab; VDZ, vedolizumab.
----- End of picture text -----

Asterisks indicate risk difference scale credible intervals do not cross zero, which may be considered ‘significant’.

B.3.9.2.4 Base case analysis – safety outcomes

Due to reporting limitations, safety events were taken as defined/reported in the relevant publications and not stratified by CCF or BF subgroup. Therefore, outcomes presented in this section are for the overall population (still with the CDAI 220–450 restriction, as described in Section B.3.9.1.5.2) and are presented for the induction and maintenance periods.

B.3.9.2.4.1 Serious AEs (induction)

Table 52 presents the base case NMA results for serious AEs with induction upadacitinib versus comparators in an overall population.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

The NMA results show a RD of ***************** with upadacitinib versus placebo, showing that the rate of serious AEs was comparable between upadacitinib and placebo.

Table 52: Results for serious AEs in induction NMA (FE model)

==> picture [447 x 86] intentionally omitted <==

----- Start of picture text -----
VDZ IV UPA UST PBO
PBO **************** ***************** *****************
UST **************** ***************** ****************
UPA **************** **************** ****************
VDZ IV ***************** ***************** *****************
----- End of picture text -----

Abbreviations: AE, adverse event; FE, fixed effects; IV, intravenous; NMA, network meta-analysis; PBO, placebo; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab. Asterisks indicate risk difference scale credible intervals do not cross zero, which may be considered ‘significant’.

B.3.9.2.4.2 Serious AEs (maintenance)

Table 53 presents the results for serious AEs with maintenance upadacitinib versus placebo in an overall population (still with the CDAI 220–450 restriction, as described in Section B.3.9.1.5.2).

The NMA results show a RD of –***************** with upadacitinib 30 mg versus placebo and ****************** with upadacitinib 15 mg versus placebo, showing that the rate of serious AEs was comparable between both upadacitinib doses and placebo. The results also showed a trend towards fewer serious AEs with any active treatment compared with placebo.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 53: Results for serious AEs in maintenance NMA (FE model)

Abbreviations: AE, adverse event; FE, fixed effects; IV, intravenous; NMA, network meta-analysis; PBO, placebo; QxW, every x weeks; SC, subcutaneous; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab.

Asterisks indicate risk difference scale credible intervals do not cross zero, which may be considered ‘significant’ (no ‘significant’ results were observed in this NMA).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 88 of 125

B.3.9.2.4.3 Discontinuation due to AEs (induction)

Table 54 presents the base case NMA results for discontinuation AEs with induction upadacitinib versus comparators in an overall population (still with the CDAI 220–450 restriction, as described in Section B.3.9.1.5.2).

The NMA results show a RD of ***************** for upadacitinib versus placebo, showing that the rate of discontinuation due to AEs was comparable between upadacitinib and placebo.

Table 54: Results for discontinuation due to AEs in induction NMA (FE model)

==> picture [447 x 86] intentionally omitted <==

----- Start of picture text -----
VDZ IV UPA UST PBO
PBO ***************** ***************** ******************
UST ***************** **************** ****************
UPA ***************** ***************** ****************
VDZ IV **************** **************** ****************
----- End of picture text -----

Abbreviations: CCF, conventional care failure; CDAI, Crohn’s Disease Activity Index; FE, fixed effects; IV, intravenous; NMA, network meta-analysis; PBO, placebo; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab.

Asterisks indicate risk difference scale credible intervals do not cross zero, which may be considered ‘significant’.

B.3.9.2.4.4 Discontinuation due to AEs (maintenance)

Table 53 presents the base case NMA results for discontinuation AEs with maintenance upadacitinib versus comparators in an overall population (still with the CDAI 220–450 restriction, as described in Section B.3.9.1.5.2).

The NMA results show a RD of **************** for upadacitinib 30 mg and **************** for upadacitinib 15 mg versus placebo, showing that the rate of discontinuation due to AEs was comparable between both doses of upadacitinib and placebo.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 55: Results for discontinuation due to AEs in maintenance NMA (FE model)

Abbreviations: AE, adverse event; FE, fixed effects; IV, intravenous; NMA, network meta-analysis; PBO, placebo; QxW, every x weeks; SC, subcutaneous; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab.

Asterisks indicate risk difference scale credible intervals do not cross zero, which may be considered ‘significant’.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 90 of 125

B.3.9.3 Uncertainties in the indirect and mixed treatment comparisons

B.3.9.3.1 RD NMA method

The NMAs used in this submission utilised the RD method, which was used in this instance as it is recommended where baseline risk-adjusted models are deemed inappropriate due to lack of convergence or face validity (97).

Like baseline risk adjustment, RD NMA is also recognised as valid framework by NICE (DSU TSD2, Section 3.7 (97)). It has been used in publications and prior submissions to NICE (100, 101). Cameron et al. (2018) (102) found that use of an NMA on the RD scale represents a viable alternative approach to account for the presence of crossstudy differences in placebo response rates. Per NICE DSU TSD2 (98, 103), RD NMA could be used as an alternative method to log-odds NMA when there are imbalances in the number of studies with low placebo response rates across pairwise contrasts in the network. The RD model code utilised was adapted from Dias et al. (2018) (98), which was based on modelling frameworks by Warn et al. (2002) (104).

In TA521, RD was used to adjust for cross-trial differences (100). Rather than calculating relative effects as ratios (such as odds ratios produced by traditional logitlink NMA frameworks), absolute probabilities of treatment response were subtracted across interventions in RD models to minimise the potential impacts of overly low or high placebo efficacy. This may help to minimise bias when there are imbalances in the number of studies with low placebo response rates across pairwise contrasts in the network. TA521 concluded that baseline-risk adjusted models and risk difference NMAs should yield less biased estimates of effect than the unadjusted NMA analyses on the relative scale.

Due to the general paucity of data in the relevant CD evidence networks leading to poor performance of baseline-risk adjusted logit-link NMAs, RD NMAs provided an attractive option to minimise impacts of placebo heterogeneity on NMA-produced treatment effect estimates.

Criticism of RD models stems from potential model instability, leading to lack of convergence and sensitivity to starting values (102). However, the RD models in the

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

CD NMA analysis in this submission converged and had appropriate fit (more details on model fit are presented in Appendix D). Appropriate vague prior distributions were utilised which corresponded to the RD scale. Starting values were utilised which are dispersed across the probability space.

In summary, the RD models addressed placebo rate variation of the sort observed in the biologic CD trials, yielded reasonable estimates, passed diagnostic tests based on their convergence and fit, are accepted by NICE, have been used in prior submissions, and have appeared in the published academic literature.

B.3.9.4 Conclusion

Induction upadacitinib was found to have superior efficacy to ustekinumab, vedolizumab, and placebo for CDAI clinical remission and clinical response. As maintenance therapy, upadacitinib was also superior to its comparators. Safety outcomes (serious AEs and discontinuation due to AEs) were comparable between upadacitinib and the comparators, including placebo.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.10 Adverse reactions

The primary safety data for upadacitinib in this submission is taken from CSRs. AEs were coded using MedDRA version 24.0. In this section, ‘Crohn’s disease’ refers to worsening of underlying CD compared with baseline.

B.3.10.1 U-EXCEL and U-EXCEED

The following section presents safety data from U-EXCEL and U-EXCEED separately; pooled safety data are presented in Appendix J.

B.3.10.1.1 TEAEs

TEAEs in U-EXCEL and U-EXCEED were defined as events that began or worsened either on or after the first dose of the study drug and:

• Within 30 days after the last dose of the study drug for subjects who did not participate in U-ENDURE

• Within 30 days after the last dose of the study drug in U-EXCEL or U-EXCEED and before the first dose of the study drug in U-ENDURE if the subject enrolled in U-ENDURE

An overview of TEAEs and deaths reported in U-EXCEL and U-EXCEED is presented in Table 56. Data are presented for the SA1 population, which included all subjects who received ≥1 dose of study drug in the double-blind induction period (Part 1) of the trials, and for the SA2 population, which included all subjects who received at least one dose of open-label upadacitinib 45 mg in Part 2. Results are presented for the anticipated licensed induction dose of upadacitinib only (i.e., 45 mg QD). A summary of TEAEs occurring in ≥5% of subjects in the upadacitinib or placebo arms of U-EXCEL or U-EXCEED is presented in Table 57.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 56: Overview of TEAEs and deaths during the 12-week double-blind and openlabel induction periods (U EXCEL and U-EXCEED SA1 and SA2 populations)

Abbreviations: COVID-19, Coronavirus Disease 2019; CSR, clinical study report; DB, double-blind; OL, openlabel; PBO, placebo; SA, safety analysis; TEAE, treatment-emergent adverse event; UPA, upadacitinib. Source: U-EXCEL (85) and U-EXCEED (86) CSRs. Note: In U-EXCEL, TEAEs in the induction period (Part 1) were defined as events that begin either on or after the first dose of the study drug in Part 1 and until (i) the first dose of study drug in U-ENDURE (if applicable), or (ii) until the first dose of study drug in Part 2 (if applicable), or (iii) within 30 days after the last dose administration of the study drug in Part 1, whichever is earlier. TEAEs for Part 2 were defined as events that began either on or after the first dose of study drug in Part 2 and until (i) first dose of study drug in U-ENDURE (if applicable) or (ii) within 30 days after the last dose of study drug in Part 2, whichever is earlier. In U-EXCEED, TEAEs for Part 1/Part 2 were defined as events that began either on or after the first dose of the study drug in Part 1/Part 2 and until (i) the first dose of study drug in U-ENDURE (if applicable), or (ii) until the first dose of study drug in Part 3 (if applicable), or until (iii) within 3 days after the last dose administration of the study drug in Part 1, whichever is earlier. Subjects are counted once in each row, regardless of the number of events they may have had.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 57: TEAEs reported in ≥5% of subjects in any treatment group of U-EXCEL or U EXCEED during the 12-week double-blind and open-label induction periods (U-EXCEL and U-EXCEED SA1 and SA2 populations)

Abbreviations: CD, Crohn’s disease; MedDRA, Medical Dictionary for Regulatory Activities; PBO, placebo; SA, safety analysis; TEAE, treatment-emergent adverse event; UPA, upadacitinib. Source: U-EXCEL (85) and U-EXCEED (86) CSRs. Note: In U-EXCEL, TEAEs in the induction period (Part 1) were defined as events that begin either on or after the first dose of the study drug in Part 1 and until (i) the first dose of study drug in U-ENDURE (if applicable), or (ii) until the first dose of study drug in Part 2 (if applicable), or (iii) within 30 days after the last dose administration of the study drug in Part 1, whichever is earlier. TEAEs for Part 2 were defined as events that began either on or after the first dose of study drug in Part 2 and until (i) first dose of study drug in U-ENDURE (if applicable) or (ii) within 30 days after the last dose of study drug in Part 2, whichever is earlier. In U-EXCEED, TEAEs for Part 1/Part 2 were defined as events that began either on or after the first dose of the study drug in Part 1/Part 2 and until (i) the first dose of study drug in U-ENDURE (if applicable), or (ii) until the first dose of study drug in Part 3 (if applicable), or until (iii) within 3 days after the last dose administration of the study drug in Part 1, whichever is earlier. Subjects are counted once in each row, regardless of the number of events they may have had.

B.3.10.1.2 Adverse events of special interest

Adverse events of special interest (AESI) were prespecified in the statistical analysis plan (SAP) for U-EXCEL and U-EXCEED. These events were selected based on safety concerns reported for other JAK inhibitors, as well as upadacitinib data from preclinical studies. An overview of AESI in U-EXCEL and U-EXCEED is presented in Table 58. Overall, rates were comparable between the upadacitinib and placebo treatment arms. Anaemia was more frequently reported in the upadacitinib arm of U- EXCEL compared with placebo, but comparable rates were observed between the treatment arms in U-EXCEED (Table 58). Rates of serious infections were *** in all

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

treatment groups, except for the open-label upadacitinib arm of U-EXCEED (****). Herpes zoster infection rates ranged from ************ in the upadacitinib arms of U- EXCEL and U-EXCEED, while no such infections were reported in the placebo group (Table 58). No cases of active TB or malignancies were reported in any of the trial arms.

Table 58: Overview of treatment-emergent AESI during the 12-week double-blind and open-label induction periods (U-EXCEL and U-EXCEED SA1 and SA2 populations)

Abbreviations: AESI, adverse events of special interest; CPK, creatine phosphokinase; CSR, clinical study report; DB, double-blind; GI, gastrointestinal; NMSC, non-melanoma skin cancer; OL, open-label; PBO, placebo; SA, safety analysis; TB, tuberculosis; UPA, upadacitinib. Source: U-EXCEL (85) and U-EXCEED (86) CSRs. Note: In U-EXCEL, TEAEs in the induction period (Part 1) were defined as events that begin either on or after the first dose of the study drug in Part 1 and until (i) the first dose of study drug in U-ENDURE (if applicable), or (ii) until the first dose of study drug in Part 2 (if applicable), or (iii) within 30 days after the last dose administration of the study drug in Part 1, whichever is earlier. TEAEs for Part 2 were defined as events that began either on or after the first dose of study drug in Part 2 and until (i) first dose of study drug in U-ENDURE (if applicable) or (ii) within 30 days after the last dose of study drug in Part 2, whichever is earlier. In U-EXCEED, TEAEs for Part 1/Part 2 were defined as events that began either on or after the first dose of the study drug in Part 1/Part 2 and until (i) the first dose of study drug in U-ENDURE (if applicable), or (ii) until the first dose of study drug in Part 3 (if applicable), or until (iii) within 3 days after the last dose administration of the study drug in Part 1, whichever is earlier.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.10.2 U-ENDURE

B.3.10.2.1 TEAEs

In U-ENDURE, TEAEs were defined as events that began either on or after the first dose of study drug in the maintenance phase and until (i) the first dose of study drug in the long-term extension phase (if applicable), or (ii) the first dose of open-label upadacitinib 30 mg QD rescue therapy (if applicable), or (iii) within 30 days after the last dose administration of the double-blinded drug in the maintenance phase, whichever is earlier. An overview of TEAEs and deaths occurring during U-ENDURE is presented in Table 59. A summary of TEAEs occurring in ≥5% of patients in U- ENDURE is presented in Table 60.

Table 59: Overview of TEAEs and deaths during the 52-week maintenance period (UENDURE SA1 population)

Abbreviations: COVID-19, coronavirus disease 2019; CSR, clinical study report; PBO, placebo; SA, safety analysis; TEAE, treatment-emergent adverse event; UPA, upadacitinib. Source: U-ENDURE CSR (87). Note: TEAEs were defined as events that began either on or after the first dose of study drug in the maintenance phase and until (i) the first dose of study drug in the long-term extension phase (if applicable), or (ii) the first dose of open-label upadacitinib 30 mg QD rescue therapy (if applicable), or (iii) within 30 days after the last dose administration of the double-blinded drug in the maintenance phase, whichever is earlier.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 60: TEAEs reported in ≥5% of subjects during the 52-week maintenance period (U-ENDURE SA1 population)

Abbreviations: COVID-19, coronavirus disease 2019; PBO, placebo; SA, safety analysis; TEAE, treatmentemergent adverse event; UPA, upadacitinib.

Source: U-ENDURE CSR (87). Note: TEAEs were defined as events that began either on or after the first dose of study drug in the maintenance phase and until (i) the first dose of study drug in the long-term extension phase (if applicable), or (ii) the first dose of open-label upadacitinib 30 mg QD rescue therapy (if applicable), or (iii) within 30 days after the last dose administration of the double-blinded drug in the maintenance phase, whichever is earlier. Subjects are counted once in each row, regardless of the number of events they may have had.

B.3.10.2.2 Adverse events of special interest

An overview of treatment-emergent AESI during the 52-week maintenance phase of U-ENDURE is presented in Table 61. Rates of serious infections were similar across the treatment groups, with the lowest rate () observed in the upadacitinib 15 mg group and the highest rate () observed in the upadacitinib 30 mg group (Table 61). Rates of anaemia and lymphopenia were higher in the placebo group than in both of the upadacitinib groups. Adjudicated GI perforations, malignancy, MACE, and VTE were infrequent and most of these events were considered by the investigator to either be unrelated to the study drug or not unexpected in a CD population (87). No events of active TB or lymphoma were reported in any treatment groups (Table 61).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 61: Overview of treatment-emergent AESI during the 52-week maintenance period (U-ENDURE SA1 population)

Abbreviations: AESI, adverse event of special interest; CPK, creatine phosphokinase; GI, gastrointestinal; MACE, major adverse cardiovascular event; NMSC, non-melanoma skin cancer; PBO, placebo; TB, tuberculosis; UPA, upadacitinib; VTE, venous thromboembolic event.

Source: U-ENDURE CSR (87). Note: MACE was defined as cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke. VTE was defined as deep vein thrombosis or pulmonary embolism (fatal and nonfatal). TEAEs were defined as events that began either on or after the first dose of study drug in the maintenance phase and until (i) the first dose of study drug in the long-term extension phase (if applicable), or (ii) the first dose of open-label upadacitinib 30 mg QD rescue therapy (if applicable), or (iii) within 30 days after the last dose administration of the double-blinded drug in the maintenance phase, whichever is earlier.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.11 Conclusions about comparable health benefits and safety

The clinical benefits of upadacitinib compared with placebo have been demonstrated in two pivotal induction studies (U-EXCEL and U-EXCEED) and one pivotal maintenance study (U-ENDURE). In the NMA, upadacitinib was superior to vedolizumab and ustekinumab for CDAI clinical remission and clinical response (CR100) in the induction period. For the maintenance period, upadacitinib was superior to its comparators for CDAI clinical remission (the only efficacy outcome assessed in the maintenance period). In NMAs for serious AEs and discontinuation due to AEs, upadacitinib showed comparable results to all comparators, including placebo.

B.3.11.1 Principal (interim) findings from the clinical evidence highlighting the clinical benefits and harms of the technology

B.3.11.1.1 Efficacy

Upadacitinib is an oral treatment and, if approved, will be the first oral advanced therapy for CD. In clinical trials, upadacitinib has demonstrated that it is an effective and well-tolerated treatment for adults with moderately to severely active CD.

Across the three pivotal trials (U-EXCEL and U-EXCEED for induction, U-ENDURE for maintenance), upadacitinib met all co-primary endpoints of clinical remission (defined either by CDAI or PROs) and endoscopic response. In the overall populations of the upadacitinib arms, a significantly greater proportion of subjects achieved clinical remission and endoscopic response compared with placebo. A clear treatment effect was also observed in the BF subpopulation. Analysis of secondary efficacy outcomes demonstrated that upadacitinib is associated with higher rates of endoscopic remission than placebo at Week 12 of U-EXCEL and U-EXCEED. This effect was maintained to Week 52 of U-ENDURE.

Upadacitinib can rapidly improve symptoms that have a substantial impact on the lives of people with CD. In the induction studies, CDAI clinical remission was achieved as early as Week 4 in approximately *** of subjects receiving upadacitinib (***** in U- EXCEL and ***** in U-EXCEED; both significantly higher than placebo). In addition to

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

the clinical efficacy, statistically significant improvements in QoL (EQ-5D-5L) were observed in the upadacitinib arm at Week 4 in both induction studies.

The early (Week 2 and Week 4) onset of efficacy of upadacitinib is important for people with CD who experience a wide range of symptoms, including abdominal pain/distension, fatigue, and bowel obstruction or diarrhoea (23, 26, 27), as well a substantial impact on their QoL and daily activities (105, 106). Furthermore, CD is a progressive condition and disease duration correlates with accumulated bowel damage, meaning that an early treatment effect is important to achieve disease control and prevent structural damage (107). The time to response for currently available biologics is reported to range from 2 to 19 weeks in most patients (108), meaning that symptoms and QoL impairments are prolonged with the potential for more bowel damage to occur. Given that a proportion of patients will not respond to a specific (or any) biologic therapy, this length of time to onset can delay treatment switching (potentially to a more effective therapy) and achievement of improvements for patients.

In addition to its early efficacy onset, upadacitinib has demonstrated sustained efficacy in terms of clinical symptoms, endoscopic disease activity, and QoL measures. This durable effect is an area of unmet need with current CD biologic therapies as loss of response to biologic therapy has been reported in 66% of individuals (48). As described above, the primary efficacy endpoints of U-ENDURE were met and both upadacitinib maintenance doses (30 mg and 15 mg QD) demonstrated significantly better rates of clinical remission and endoscopic response compared with placebo at Week 52. Endoscopic remission rates at Week 52 were also significantly better with both doses of upadacitinib versus placebo. Meeting both clinical and endoscopic treatment targets over 52 weeks suggest that upadacitinib may facilitate long-term CD control. For example, improvements in endoscopic outcomes may reduce the future relapse rate, risk of penetrating complications, and surgery (61, 109) compared with people with severe ulcerations; mucosal healing has recently been recognised in UK clinical guidance as an important target for clinicians and people with CD (57).

Across all key secondary endpoints in U-ENDURE, both doses of upadacitinib showed statistically significant improvements compared with placebo, including for endoscopic remission and CDAI clinical response. QoL improvements were also maintained, with

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

statistically significant improvements in EQ-5D-5L observed with upadacitinib 30 mg and numerical improvements with upadacitinib 15 mg.

Finally, one of the key treatment aims in CD is the reduction or discontinuation of corticosteroids due to their association with AEs, including bone loss, mood disorders, insomnia, hypertension, elevated blood glucose, and hypoadrenalism (110). Corticosteroids are used as concomitant or bridging therapy in CD. However, in U-EXCEL, **** of patients who were receiving corticosteroids for CD at baseline achieved clinical remission and were able to discontinue corticosteroids at Week 12 (i.e., achieved steroid-free remission), with similar treatment effects observed in the overall and Bio-IR, as well as in the upadacitinib arm of U-EXCEED. The effect of upadacitinib on steroid-free remission was maintained to Week 52 of U-ENDURE (approximately *** steroid-free remission rate in both upadacitinib arms versus ***** in placebo).

B.3.11.1.2 Safety

Across the induction and maintenance studies, upadacitinib was generally safe and well-tolerated. The results of the upadacitinib CD trials did not identify any new safety risks compared with the known safety profile of upadacitinib, which has been established across six other indications with 19 Phase 3 trials and more than 5,000 patients.

In the 12-week induction and extended induction periods of U-EXCEL and U-EXCEED, upadacitinib 45 mg QD was well tolerated. Rates of serious AEs, severe AEs, and AEs leading to discontinuation of study drug were comparable between the upadacitinib 45 mg and placebo groups. One death was reported across the induction studies (in U-EXCEED) but was not considered to be related to the study drug. The most frequent AEs, occurring in ≥5% of subjects in the upadacitinib groups in either U-EXCEL or U-EXCEED, were acne, anaemia, nasopharyngitis, headache, CD, and upper respiratory tract infection.

Maintenance treatment with upadacitinib 30 mg or 15 mg QD for 52 weeks was also well tolerated. In the SA1 analysis set (representing Cohort 1), rates of overall AEs, serious AEs, severe AEs, and AEs leading to discontinuation of study drug were lower

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

in the upadacitinib groups than in the placebo group. The most frequent AEs, occurring in ≥5% of subjects in either of the upadacitinib groups, were worsening of CD, arthralgia, and pyrexia (CD worsening was reported more frequently in subjects receiving placebo compared with upadacitinib, which may reflect improvement in underlying CD among upadacitinib recipients). Rates of serious infection were comparable across the treatment groups. Herpes zoster, hepatic disorders, neutropenia, and creatine phosphokinase (CPK) elevations were reported more frequently in the upadacitinib 30 mg group than in the upadacitinib 15 mg and placebo groups. Anaemia and lymphopenia were more commonly reported in the placebo group than in the upadacitinib groups.

B.3.11.1.3 Indirect treatment comparison

In the absence of head-to-head RCTs between upadacitinib and the comparators specified in the NICE scope, a series of NMAs were performed to assess the relative efficacy of upadacitinib compared with relevant comparators (ustekinumab, vedolizumab) in people with moderately to severely active CD in the BF population. In the induction analyses, upadacitinib was superior to ustekinumab and vedolizumab for CDAI clinical remission and clinical response (CR-100). In the maintenance analyses, upadacitinib was generally superior to the comparators for CDAI clinical remission (the only outcome assessed). The safety NMAs, which were conducted for serious AEs and discontinuation due to AEs, showed that upadacitinib had a comparable safety profile to the comparators, including placebo.

B.3.11.2 Strengths and limitations of the clinical evidence base for the technology

B.3.11.2.1 Trial design

The U-EXCEL and U-EXCEED induction studies and the U-ENDURE maintenance study were large, multinational, placebo-controlled, well-conducted and methodologically robust studies with relevant and appropriate eligibility criteria. The upadacitinib clinical trial programme enrolled a total of ************************************************************************************************ ****************************************************************************

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

B.3.11.2.2 Intervention and comparators

Upadacitinib is an oral therapy and offers advantages through its mode of administration, including increased flexibility for patients in the timing and location of administration, and minimal invasiveness. Clinicians have noted that IV treatment is often delayed due to limited IV capacity in clinics (12); oral therapy is advantageous in this regard because it can be taken as soon as required. All currently available advanced therapies for CD are administered either IV or SC and therefore upadacitinib would be the first oral advanced therapy for CD in the UK.

The upadacitinib trials included treatment arms with different doses of upadacitinib; however, this submission only presents results for upadacitinib doses that are expected to be licensed in UK clinical practice, i.e., 45 mg QD for induction and 30 mg or 15 mg QD for maintenance. All three studies were placebo-controlled, which facilitates indirect treatment comparison with multiple other comparator treatments through the respective placebo arms.

B.3.11.2.3 Patient characteristics

Baseline subject characteristics in the upadacitinib clinical trials were representative of the moderately to severely active CD population in the UK. Demographics and clinical characteristics were generally well balanced across the treatment arms of each trial, as well as across the three trials. Furthermore, based on clinical expert opinion (64), the Bio-IR and non-Bio-IR populations of the upadacitinib clinical trials are broadly representative of populations seen in UK clinical practice, namely those with prior biologic therapy experience (BF population) and those with no prior biologic therapy experience (CCF population), respectively.

B.3.11.2.4 Outcomes

U-EXCEL, U-EXCEED, and U-ENDURE provide efficacy and safety data of direct relevance to the anticipated license for upadacitinib in CD. A key strength of the trials is that the co-primary endpoints determine improvements in both clinical symptoms (determined by CDAI/PRO clinical remission) and in mucosal healing (determined by endoscopic response). Although CDAI is not widely used to measure disease severity in UK clinical practice, it remains relevant because it is linked to the HBI, which is

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

commonly used as a disease severity measure in the UK (88). Use of the CDAI also facilitated comparison of upadacitinib with ustekinumab and vedolizumab as it was the assessment method used in these trials. Inclusion of endoscopic response as a primary endpoint reflects a paradigm shift in CD treatment with mucosal healing now considered a key treatment goal due to its association with improved long-term outcomes (57-59). For UK clinical practice, endoscopic outcomes may be more informative than CDAI for quantifying disease severity; however, as they are less frequently reported in trials of other therapies, there are limitations when making comparisons using these outcomes. Therefore, including both CDAI and endoscopic outcomes in the upadacitinib trials provides clinically relevant data that can be used to draw meaningful comparisons with other CD therapies.

B.3.11.2.5 Safety

Upadacitinib was generally well tolerated in clinical trials; however, as upadacitinib is administered orally, it can be easily stopped if AEs occur. AEs are also likely to be of limited duration as a short half-life of approximately 9–14 hours (13).

B.3.11.2.6 Limitations

As described above, the upadacitinib trial populations broadly represent the moderately to severely active CD population in the UK. However, as the three trials were multinational, some of the subjects may have received prior treatment with a therapy that is not approved in the UK.

A limitation of the upadacitinib trials is that they were all placebo-controlled, meaning that no head-to-head data with active comparators were available. This limitation was addressed by performing indirect treatment comparisons with NMAs conducted in the CCF and BF populations.

A limitation of U-ENDURE (maintenance study) is the re-randomised responderwithdrawal design, which meant that subjects with a clinical response to upadacitinib 45 mg in the induction period could be re-randomised to placebo for maintenance; as a result, a proportion of subjects receiving placebo in U-ENDURE had previous exposure to upadacitinib. However, this has also been a limitation of pivotal maintenance trials of other advanced therapies for CD (81, 82).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

The NMAs performed for induction upadacitinib versus comparators showed that upadacitinib tended to show improved efficacy versus comparators (***** improvement versus placebo for CDAI clinical response and ***** improvement versus placebo for CDAI clinical remission in the BF populations, respectively).

B.3.12 Ongoing studies

Publications of primary and post-hoc analyses from the three upadacitinib trials (UEXCEL, U-EXCEED, and U-ENDURE) are expected within the next 12 months. No results from other studies are expected.

B.4 Cost-comparison analysis

B.4.1 Changes in service provision and management

Upadacitinib is an oral therapy and is administered once-daily as a modified-release tablet for both the induction and maintenance doses. It is expected to be prescribed in secondary care with all administrations taking place at home. The comparator therapies, ustekinumab and vedolizumab, are administered either IV or SC. All IV infusions are administered in a hospital setting. The first dose of SC therapy is administered by a trained nurse and incurs an associated cost; no additional costs to the NHS are assumed for subsequent SC doses as these are typically selfadministered (see Section B.4.2.2). Therefore, despite requiring more frequent administration than its IV and SC comparators, upadacitinib requires fewer resources due to being an oral therapy.

B.4.2 Cost-comparison analysis inputs and assumptions

B.4.2.1 Features of the cost-comparison analysis

A cost comparison analysis was conducted to evaluate the cost to the NHS of using upadacitinib versus ustekinumab and vedolizumab for people with moderately to severely active CD in whom TNF-alpha inhibitors are deemed unsuitable; or where biological treatment is not tolerated or not working well enough. A cost-comparison model was developed in Microsoft Excel to facilitate comparison between the therapies.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

The model time horizon is 1 year, in line with clinical practice and NICE guidance, which states that patients should be reassessed at 12 months to determine whether continuing with biologic treatment is appropriate (111). This also aligns with the duration of the upadacitinib clinical trials, which assessed maintenance outcomes up to 52 weeks. Year 1 includes all induction treatment and associated costs, as well as maintenance therapy once the induction period is completed.

It has been noted during previous appraisals that a treatment continuation assessment means that it will be deemed appropriate for at least some patients to continue treatment beyond 1 year. Therefore, in Section B.4.4, a sensitivity analysis is presented that considers the cost of a patient continuing their treatment beyond Year 1. This time horizon is described as ‘Year 2+’ for the purposes of this submission and shows the costs of each year of treatment beyond Year 1.

In the model, upadacitinib is compared against ustekinumab and vedolizumab as these treatments are approved in the target BF and TNF-alpha inhibitorcontraindicated populations and would be displaced by the introduction of upadacitinib (See Section B.1.3.6). The BF population is considered to include those contraindicated to TNF-alpha inhibitors. Data on CCF patients from the upadacitinib clinical trials are presented for completeness in Appendix J.

Costs were not discounted in the analysis in line with the user guide for costcomparison appraisals (112).

B.4.2.2 Intervention and comparators’ acquisition costs

Ustekinumab is administered intravenously (IV) in the induction period and subcutaneously (SC) in the maintenance period. Vedolizumab administration is IV in the induction period and either IV or SC in the maintenance period. Both the IV and SC regimens of vedolizumab were considered in this analysis.

Ustekinumab IV dosing is weight-based. An average dose of 390 mg, equating to three 130 mg vials, was calculated based on trial population data presented in the ustekinumab SmPC (71). Dosing of ustekinumab SC, vedolizumab SC, and vedolizumab IV is not weight dependent. The weight characteristics of the model

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

population, which were used to calculate the ustekinumab IV induction dose, are presented in Table 62 and are based on the upadacitinib trial populations (113).

Table 62: Population characteristics for cost comparison model

Abbreviations: IV, intravenous; NA, not applicable; SmPC, Summary of Product Characteristics; UST, ustekinumab. Source: AbbVie data on file (113), ustekinumab IV SmPC (72)

Upadacitinib, ustekinumab, and IV vedolizumab are all available as either standard dose or high dose maintenance therapy. The proportion of patients on the standard and high doses is based on UK clinical expert input (114) and is shown in Table 63.

Table 63: Proportion of patients on standard and high dose maintenance therapy in cost comparison model

Abbreviations: IV, intravenous; NA, not applicable; SC, subcutaneous; UPA, upadacitinib; UST ustekinumab; VDZ, vedolizumab.

The acquisition costs of upadacitinib and comparators are presented in Table 64. The table also presents the dosing schedule used in the model for each treatment.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 64: Acquisition costs of the intervention and comparator technologies

Abbreviations: IV, intravenous; NA, not applicable; PAS, Patient Access Scheme; QD, once daily; QxW, once every x weeks; SC, subcutaneous; VAT, value-added tax. †For Year 2+ data, please refer to Section B.4.4

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 109 of 125

B.4.2.3 Intervention and comparators’ healthcare resource use and associated costs

A systematic literature review was undertaken to identify studies reporting healthcare resource use and costs associated with upadacitinib and relevant comparators. A total of 6 studies relevant to UK decision-making were identified and full details are reported in Appendix G.

Upadacitinib is administered as oral tablets for induction and maintenance doses; as such, there are no associated administration costs. Ustekinumab is administered via IV infusion for the induction and maintenance doses. Vedolizumab is administered by IV infusion for induction and then either IV or SC for maintenance therapy. Infusions (IV) were administered in a hospital setting; SC administrations assumed a cost for the first dose only (training by a nurse) and no additional cost to the NHS for subsequent doses (as these are typically self-administered).

Administration costs of IV, SC, and oral treatments are shown in Table 60.

Table 65: Administration costs used in cost comparison model

Abbreviations: IV, intravenous; SC, subcutaneous Source: IV costs were sourced from the National Tariff Payment System 2022/23 HRG code FD02H Inflammatory Bowel Disease without Interventions, with CC Score 0. SC costs were sourced from the Unit Costs of Health and Social Care 2021. Personal Social Services Research Unit. Cost per working hour or band 5 nurse. See Section 10.1, page 108 (115).

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027]

© AbbVie (2022). All rights reserved

Table 66: Resource costs of the intervention and comparator technologies (Year 1)

Abbreviations: HRG, Healthcare Resource Group; IV, intravenous; NA, not applicable; SC, subcutaneous; UPA, upadacitinib; UST, ustekinumab; VDZ, vedolizumab.

Company evidence submission template for upadacitinib for previously treated moderately to severely active Crohn’s disease [ID4027] © AbbVie (2022). All rights reserved Page 111 of 125

B.4.2.4 Adverse reaction unit costs and resource use

As described in Section B.3.9.2.4, results of the NMA analyses for serious AEs and discontinuation due to AEs showed similar AE rates between upadacitinib, ustekinumab, and vedolizumab. Therefore, it was assumed that the costs associated with treatment AEs would be similar for all therapies and AE costs were excluded from the analysis.

B.4.2.5 Miscellaneous unit costs and resource use

No other unit costs and resource use were considered.

B.4.2.6 Clinical expert validation

Several key assumptions in the model were validated with clinical experts (12, 64):

• Proportion of patients receiving standard and high doses of upadacitinib

• Proportion of patients receiving standard and high doses of ustekinumab

• Exclusion of extended induction from the base case analysis

B.4.2.7 Uncertainties in the inputs and assumptions

Key model inputs are summarised in Table 67; these assumptions (patient weight and proportion of patients on standard versus high dose) are considered key because they affect the required dose of upadacitinib for each patient and thus impact the associated costs. Model assumptions are presented in Table 68.

Table 67: Summary of key model inputs