CADTH Reimbursement Review

Mirikizumab (Omvoh)

Sponsor: Eli Lilly Canada

Therapeutic area: Ulcerative colitis

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Stakeholder Input

Clinical Review

Executive Summary

An overview of the submission details for the drug under review is provided in Table 1.

Table 1: Background Information on Application Submitted for Review

JAK = Janus kinase; NOC = Notice of Compliance; SC = subcutaneous; UC = ulcerative colitis.

Introduction

Inflammatory bowel disease (IBD) is a term used to describe disorders that involve chronic inflammation of the digestive tract. Ulcerative colitis (UC) is 1 such disease. UC causes inflammation and ulcers in the digestive tract, affecting the innermost lining of the large intestine (colon) and rectum.^1,2 UC is characterized by blood in the stool with mucus, frequent diarrhea, loss of appetite, and tenesmus (a strong urge to use the bathroom without necessarily having a bowel movement), in addition to abdominal pain, rectal bleeding (RB), and weight loss.^3-5 The most common initial manifestation of UC is bloody diarrhea with or without mucus. While the etiology of UC is not completely understood, there is growing evidence to suggest genetic and environmental factors may contribute to the irregular immune response that aberrantly recruits activated immune cells to the colon,³ resulting in chronic inflammation that damages the colon and causes UC symptoms. UC generally develops in young adulthood^6-8 and persists throughout life, marked by periods of spontaneous remission and relapse.⁹ Though most patients experience this relapsing-remitting disease course, up to 24% of patients report experiencing continuous UC symptoms.⁹ The majority of individuals living with UC have a mild to moderate disease course, generally with active disease at diagnosis followed by alternating exacerbations and longer periods of remission.¹⁰ However, aggressive disease course is experienced in 10% to 15% of patients, with a cumulative risk of relapse of between 70% to 80% at 10 years postdiagnosis.¹⁰ Regardless of severity, UC is associated with a substantial reduction in quality of life (QoL) for patients, with considerable impact on many aspects of their lives, including emotional and psychological functioning, social and physical functioning, and work and academic life.^11,12 UC is diagnosed clinically, with endoscopy, biopsy, and stool sampling being common tests used in ruling out other causes of symptoms.¹³ Treatment strategies for UC are dependent on the presence of active disease, severity and extent of the UC, and patient preference with the goal of achieving complete remission. Conventional therapies for UC include 5-aminosalicylic acid (5-ASA) products, corticosteroids, and immunomodulators (such as azathioprine, 6-mercaptopurine, and methotrexate); advanced therapies consist of adalimumab, golimumab, infliximab, ustekinumab, tofacitinib, ozanimod, and vedolizumab. However, current treatments are unable to meet all current needs of patients in terms of short-term or long-term treatment. Remission with treatment is not universal and a patient’s UC can lose response to treatment after an initial period of improvement and relapse even after long periods of remission on an existing therapy. Accordingly, there is a need for novel therapies targeting alternative pathways. An estimated 322,600 patients in Canada are living with IBD.¹⁴ In 2030, the number of people living in Canada with IBD is anticipated to be 470,000, accounting for 1.1% of the population with a prevalence for UC specifically of 0.44%.^14,15

Mirikizumab is a humanized immunoglobulin G4 (IgG4) monoclonal antibody that binds with high affinity and specificity to the p19 subunit of human interleukin-23 (IL-23) cytokine and inhibits its interactions with the IL-23 receptor.¹⁶ Mirikizumab is indicated for the treatment of adult patients with moderately to severely active UC who have had an inadequate response, had a loss of response, or were intolerant to conventional therapy, a biologic treatment, or a Janus kinase (JAK) inhibitor.

The objective of this report is to review and critically appraise the evidence submitted by the sponsor on the beneficial and harmful effects of mirikizumab, 300 mg/15 mL IV (induction) and 100 mg/1 mL subcutaneous (SC) injection (maintenance) in the treatment of adult patients with moderately to severely active UC who have had an inadequate response, had a loss of response, or were intolerant to conventional therapy, a biologic treatment, or a JAK inhibitor.

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient and clinician groups that responded to CADTH’s call for input and from a clinical expert consulted by CADTH for the purpose of this review.

Patient Input

Patient input was collected from the Gastrointestinal Society (GI Society) and Crohn’s and Colitis Canada. Patient input was collected through a variety of questionnaires (n = 4 to 432), focus groups, and individual interviews. In addition, 1-to-1 interviews were conducted with 4 patients with UC who received mirikizumab in a clinical trial. After collating responses, it was noted that UC has a profound effect on daily life — physically, emotionally, and socially — at home and at school or in the workplace. Many patients surveyed by Crohn’s and Colitis Canada revealed that they hid aspects of their diagnosis from their friends, coworkers, and classmates and almost two-thirds of patients (63%) agreed that their family and friends do not know or understand what they are going through. Patients noted that symptoms can be relentless, embarrassing, and scary. Based on the surveys conducted by Crohn’s and Colitis Canada, the most frequently reported UC-related complications reported were mental health and stress (65%), joint inflammation and arthritis (51%), anal fissures and hemorrhoids (40%), anemia (33%), skin conditions (30%), malnutrition (30%), and weight loss (30%). Patients stated that more than anything, sustained remission and/or treatment response was more important than relieving any 1 symptom of UC. The constant concern that there would be future flares, possibly worse than the last, at unpredictable times, was noted as being disastrously disruptive.

Regarding current treatments for UC, it was noted that although there are several available options, most patients have difficulty obtaining remission or adequate symptom relief. Based on survey data from the GI Society, only 24% of patients with IBD found available medications to be adequate, 56% of patients found them to be only somewhat adequate, and 20% of patients found them not at all adequate. More than half of patients (56%) surveyed by Crohn’s and Colitis Canada believed that different treatment options could make them feel better. While steroid use is an important part of symptom management for UC, patients surveyed by Crohn’s and Colitis Canada reported that they were not particularly supportive of the treatment option. Almost all patients (93%) surveyed by Crohn’s and Colitis Canada stated that they only take systemic steroids if absolutely necessary. Patient input from the GI Society stressed that treatment response varies across patients, and in some cases response to medication may stop after prolonged use. For these reasons, patients noted it is important to have a variety of treatment options for UC. Patients noted that there is a need for new and effective options to achieve mucosal healing and reduce the debilitating symptoms of UC, as well as provide good QoL. Patients interviewed by Crohn’s and Colitis Canada added that any new treatment must be able to protect a patient’s ability to work productively, attend school and social events, and conduct basic necessities such as leaving the home to run errands. However, patients interviewed by Crohn’s and Colitis Canada also added that potential risks and side effects, especially those related to heart and liver function, are a major source of concern when considering new treatment options. The 4 patients interviewed by the GI Society who received mirikizumab in clinical trials reported at the time they were interviewed that they continue to take the medication. All 4 patients experienced improved gut healing and expressed improvement in QoL following treatment with mirikizumab. Regarding the administration of mirikizumab, 2 patients reported that the initial induction of treatment by infusion was exhausting and time consuming, and although 2 patients did not particularly like the SC administration of mirikizumab, they were willing to tolerate it and described it as manageable.

Clinician Input

Input From Clinical Experts Consulted by CADTH

The clinical expert noted that there is an unmet need for treatments that are better tolerated, improve convenience and compliance, and take into consideration special populations such as those with previous or current malignancies. The clinical expert stated that having available treatments with different administration methods is also important to patients (e.g., IV, SC, oral) as well as having multiple treatment options, given that a patient’s UC will often lose response to treatment and require another therapy. The clinical expert noted that not all patients’ UC will respond to available treatments and patients’ disease often becomes refractory to current treatment options. According to the clinical expert, mirikizumab is an anti–IL-23 drug for UC and would offer a novel treatment mechanism for the disease that is more targeted compared to other therapies such as ustekinumab. The clinical expert anticipated that the place in therapy of mirikizumab would be similar to other biologics and would be recommended as a second-line therapy after 5-ASA, instead of immunomodulators. It was the opinion of the clinical expert that patients do not need to initiate other therapies and have their UC fail to respond to these other therapies before being prescribed mirikizumab, given the limitations and risks of other therapies. The clinical expert noted that patients whose UC is most likely to respond to treatment with mirikizumab would be those with moderate to severe UC (who are biologic-naive or biologic-experienced) that has not responded to conventional therapy. Patients described as biologic-experienced had tried at least 1 biologic therapy and/or tofacitinib. The clinical expert noted that patients least suited for treatment with mirikizumab are patients with active infections, malignancy, and/or severe hepatic impairment, and patients who are pregnant. The clinical expert felt that the patients in greatest need of mirikizumab would be those whose UC had failed to respond to first-line therapy with 5-ASA.

According to the clinical expert, the outcomes used in clinical practice align with those used in clinical trials, such as clinical remission and clinical response (measured by Partial Mayo Score), endoscopic remission and response, and biomarkers (e.g., fecal calprotectin). The clinical expert noted that clinicians routinely schedule a colonoscopy to check for endoscopic healing 6 months to 9 months after a patient has been started on a biologic therapy or a small molecule drug. If concerned about patient response, the expert indicated that some physicians may try to book a flexible sigmoidoscopy soon after the induction period has been completed.

According to the clinical expert, a clinically meaningful response to treatment would be no further RB, no rectal incontinence, rectal urgency that has been reduced or no longer exists, bowel movement frequency that has been reduced or is normal, stools becoming more solid, and abdominal pain that has been reduced or no longer exists. The clinical expert would expect clinical improvement within 4 weeks and clinical remission within 12 weeks; however, depending on the severity of disease and previous medication exposure, the clinical expert noted that patients may have a slower response or a delay to remission. In this case, the expert indicated that they would be comfortable with an extended induction period of 12 weeks for those with disease that does not respond to treatment, which is aligned with the product monograph. The clinical expert noted that most gastroenterologists use standard clinical scores (e.g., Partial Mayo Score, Modified Mayo Score [MMS]) for UC in clinics with an endoscopic component if performing colonoscopy.

Regarding the discontinuation of treatment, the clinical expert suggested that mirikizumab should be discontinued in the event of serious adverse events (SAEs), disease progression, or the inability to taper off steroids. The clinical expert would consider stopping treatment after 24 weeks of therapy if the patient’s UC does not respond. This would include an extended induction phase if the patient’s UC was not responding to initial induction (e.g., 12 weeks). According to the clinical expert, it would be expected that approximately 30% of patients who did not have an initial induction response might have a delayed response to induction treatment.

Clinician Group Input

Clinician group input was received by a group of gastroenterologists in Canada. Input from the clinician group was compiled by 9 gastroenterologists recognized as experts in the management of IBD. Based on input from the clinician group, the goals of UC therapy are multifaceted, ranging from controlling symptoms to preventing disease progress, surgery, and disability with early intervention and a treat-to-target approach. The clinician group identified the following unmet needs in a therapy that treats moderate to severe UC: a therapy that induces and maintains symptomatic remission, is safe with long-term use, and can rapidly improve endoscopic appearance of the bowel and maintain this in the long term. The clinician group emphasized that none of the currently available therapies for UC meet all of the current needs of patients in the short term or long term. Remission with treatment is not universal and a patient’s UC can lose response to therapy after an initial period of improvement and relapse even when in deep remission on an existing therapy. Accordingly, the clinician group advocated for the need for novel therapies targeting alternative pathways. Overall, the clinician group found that mirikizumab has the potential for a broad range of uses in clinical practice from first-line advanced therapy to the treatment of patients with inadequate response or intolerance to multiple advanced therapies.

With regard to treatment with mirikizumab, the clinician group suggested that the aim of treatment should be remission. The clinician group suggested that a meaningful improvement in symptoms as measured by the resolution of stool frequency (SF) and RB should be demonstrated in the first 3 months of therapy. The clinician group expected patients to be in symptomatic remission and off corticosteroids by 6 months after the initiation of mirikizumab. The clinician group added that symptomatic improvement should be accompanied by a decrease in biomarkers of inflammatory activity (C-reactive protein and fecal calprotectin) in the first 3 months after initiating mirikizumab. The clinician group suggested discontinuing treatment with mirikizumab in the event of worsening symptoms or inadequate response. In circumstances where there was an inadequate response to mirikizumab as a first-line biologic, the clinician group indicated that a switch to another class of drugs, such as an anti–tumour necrosis factor (anti-TNF) drug, is warranted. Based on clinical experience, the clinician group suggested that mirikizumab be administered in a clinic by a trained health care professional during the induction phase.

Drug Program Input

Input was obtained from the drug programs that participate in the CADTH reimbursement review process. The following were identified as key factors that could potentially impact the implementation of a CADTH recommendation for mirikizumab:

• relevant comparators

• consideration for the initiation of therapy

• consideration for the continuation or renewal of therapy

• consideration for the prescribing of therapy

• care provision issues

• system and economic issues.

The clinical expert consulted by CADTH provided advice on the potential implementation issues raised by the drug programs. Refer to Table 4.

Clinical Evidence

Pivotal Studies and Randomized Controlled Trial Evidence

Description of Studies

Two double-blind, multicentre, parallel-arm, randomized placebo-controlled trials, the LUCENT-1 and LUCENT-2 trials, were submitted by the sponsor.

The LUCENT-1 trial (N = 1,281) was a 12-week induction trial in which patients were randomized 3:1 to either mirikizumab 300 mg IV every 4 weeks or placebo. The aim of the study was to assess whether mirikizumab 300 mg IV would induce clinical remission at week 12 in adult patients with moderately to severely active UC. Major secondary objectives included alternate clinical remission, clinical response, clinical response in patients who are biologic-experienced, endoscopic remission, symptomatic remission, bowel urgency improvement, and histologic endoscopic mucosal improvement (HEMI), all at week 12. Health-related quality of life (HRQoL) was also evaluated at week 12 using the Inflammatory Bowel Disease Questionnaire (IBDQ), the 5-Level EQ-5D (EQ-5D-5L), and the Short Form (36) Health Survey (SF-36). The Work Productivity and Activity Impairment Questionnaire: Ulcerative Colitis (WPAI:UC), for patients with UC, was also evaluated at week 12 among patients employed at baseline. At baseline, patients had a mean age of 42.5 years (standard deviation [SD] = 13.92), with the majority being male (59.8%) and white (72.3%). There was an equal number of patients with moderate UC and severe UC based on MMSs. The proportion of patients reporting prior biologic or tofacitinib failure was also similar between treatment groups (41.6% and 40.1% of patients randomized to mirikizumab and placebo, respectively).

The LUCENT-2 trial (N = 544 in the primary analysis) was a 40-week maintenance trial in which patients were randomized 2:1 to either mirikizumab 200 mg SC every 4 weeks or placebo. The aim of the trial was to assess whether mirikizumab 200 mg SC would achieve clinical remission from baseline to week 40 in adult patients with moderately to severely active UC who had previously attained a clinical response at week 12 of the LUCENT-1 trial. Major secondary objectives included alternate clinical remission, corticosteroid-free remission, durable clinical remission, endoscopic remission, bowel urgency remission and improvement, and histologic endoscopic mucosal remission (HEMR), all at week 40 (i.e., 52 weeks of treatment in total). The WPAI:UC was also evaluated at week 40 among patients employed at baseline. The baseline characteristics in the LUCENT-2 trial were similar to that of the LUCENT-1 trial. The majority of the patients in the main cohort had a mean age of 42.3 (SD = 13.5) years, and were male (58.4%) and white (71.3%). Based on the MMS, approximately half of the patients in each treatment arm were categorized as moderate UC severity and 35.1% of patients in the mirikizumab group and 35.8% of patients in the placebo group had a history of biologic or tofacitinib failure. Overall, the baseline characteristics were well-balanced between treatment arms.

The LUCENT-2 study also enrolled 405 patients from the LUCENT-1 study whose UC had not responded to 12 weeks of induction dosing with either mirikizumab or placebo. These patients received open-label mirikizumab (300 mg administered intravenously) for 12 weeks. This was referred to as an extended induction period for patients who had previously received 12 weeks of induction dosing (i.e., 24 weeks of continuous therapy).

Efficacy Results

A summary of key efficacy results from the LUCENT-1 and LUCENT-2 trials are available in Table 2.

Induction Period: LUCENT-1 Trial

Clinical Response

In the LUCENT-1 trial, clinical response was evaluated using the MMS. After 12 weeks of treatment, a greater proportion of patients on mirikizumab 300 mg IV than placebo attained clinical response, with a common risk difference of 21.4% (99.875% confidence interval [CI], 10.8% to 32.0%; P < 0.00001). Results of the sensitivity analyses in the intention-to-treat (ITT) population were consistent with the modified intention-to-treat (mITT) population results. In both the biologic-naive and biologic-experienced subgroups, more patients attained clinical response on mirikizumab than on placebo, with a common risk difference of 19.8% (95% CI, 11.3% to 28.3%; P < 0.001) and 23.9% (95% CI, 14.3% to 33.5%; P < 0.001), respectively. The biologic-naive subgroup consisted of patients who did not have experience with biologic therapy and tofacitinib, while the biologic-experienced subgroup consisted of patients whose UC had failed to respond to at least 1 biologic therapy or tofacitinib. The magnitude of effect for both subgroups was consistent with the primary analysis. In the subgroups of patients using corticosteroids or immunomodulators at baseline, the observed effect sizes for clinical response in the mirikizumab group compared to the placebo group were numerically positive but small when compared to the overall population. No subgroup differences were observed.

Clinical Remission

Clinical remission was assessed using 2 different outcomes in the LUCENT-1 study: clinical remission and alternate clinical remission. Clinical remission was based on the MMS and defined as the following: SF subscore equals 0 or SF subscore equals 1 with at least a 1-point decrease from baseline; RB subscore equals 0; and endoscopic subscore (ES) equals 0 or 1 (excluding friability). Alternate clinical remission used the same definition except that it excluded the need for at least a 1-point decrease from baseline in the SF subscore. These were considered appropriate measures by the clinical expert.

Clinical Remission Rate

A greater proportion of patients on mirikizumab 300 mg IV (24.2%) versus placebo (13.3%) experienced clinical remission at week 12, with a common risk difference of 11.1% (99.875% CI, 3.2% to 19%; P = 0.00006). Results of the analyses in the per-protocol (PP) and ITT populations were consistent with the mITT population results. Results of the sensitivity analyses assessing the impact of attrition and missing data were consistent with the results from the primary analysis. In terms of the tipping point analysis, there was no significant difference between groups when imputing missing data as “responder” for the placebo group and as “nonresponder” for the mirikizumab group.

In both the biologic-naive and biologic-experienced subgroups, more patients experienced clinical remission on mirikizumab than on placebo with a common risk difference of 15.1% (95% CI, 8.3% to 21.9%; P < 0.001) and 6.8% (95% CI, 0.5% to 13.0%; P = 0.065), respectively. The magnitude of effect for the biologic-naive subgroup was consistent with the primary analysis. In the subgroups of patients using corticosteroids or immunomodulators at baseline, the observed effect sizes for clinical remission in the mirikizumab group compared to the placebo group were numerically positive but small when compared to the overall population.

Alternate Clinical Remission Rate

The results for alternate clinical remission, a slightly less stringent definition of remission, were aligned with the results for clinical remission.

Endoscopic Remission

At week 12 of the LUCENT-1 study, 36.3% of patients on mirikizumab experienced endoscopic remission versus 21.1% of patients on placebo, with a common risk difference of 15.4% (99.875% CI, 6.3% to 24.5%; P value < 0.00001). Results of the analyses in the ITT population were consistent with the mITT population results.

In both the biologic-naive and biologic-experienced subgroups, more patients experienced endoscopic remission at week 12 on mirikizumab than on placebo, with a common risk difference of 17.9% (95% CI, 9.8% to 25.9%; P < 0.001) and 12.3% (95% CI, 5.2% to 19.4%; P = 0.003), respectively.

In the subgroups of patients using corticosteroids or immunomodulators at baseline, the observed effect sizes for endoscopic remission in the mirikizumab group compared to the placebo group were numerically positive but small when compared to the overall population. No subgroup differences were observed.

Ulcerative Colitis Endoscopic Index of Severity Score of 1 or More

The Ulcerative Colitis Endoscopic Index of Severity (UCEIS) is a physician-reported measure of the endoscopic disease activity of UC on flexible sigmoidoscopy or colonoscopy, which calculates a score ranging from 0 to 8 based on vascular pattern (scored 0 to 2); bleeding (scored 0 to 3); and erosions and ulcers (scored 0 to 3), with higher scores indicating worse outcomes. || |||| || || ||||||||| ||||| || |||||||| || ||||||||||| ||| || ||||| ||||| ||| || |||| |||||| ||| || |||||||| || |||||||| |||||||||| |||| |||||||||| || ||||| |||||||| ||| ||| || ||||| |||||||||

Symptomatic Remission

At week 12 of the LUCENT-1 study, 45.5% of patients on mirikizumab and 27.9% of patients on placebo experienced symptomatic remission, with a common risk difference of 17.5% (99.875% CI, 7.5% to 27.6%; P < 0.00001). Results of the analyses in the ITT population were consistent with the mITT population results.

In both the biologic-naive and biologic-experienced subgroups, more patients experienced symptomatic remission at week 12 on mirikizumab than on placebo, with a common risk difference of 17.1% (95% CI, 8.7% to 25.4%; P < 0.001) and 18.8% (95% CI, 10.1% to 27.4%; P < 0.001), respectively.

In the subgroups of patients using corticosteroids or immunomodulators at baseline, the observed effect sizes for symptomatic remission in the mirikizumab group compared to the placebo group were numerically positive but small when compared to the overall population. No subgroup differences were observed.

Bowel Urgency Improvement: Urgency Numeric Rating Scale

The Urgency Numeric Rating Scale (UNRS) is an instrument used to assess patient-reported severity of bowel urgency in adults with UC with a 24-hour recall period.¹⁷ Using the UNRS, the least squares mean (LSM) change from baseline at week 12 in the mirikizumab group was –2.59 points and –1.63 points in the placebo group, a difference of –0.95 points (99.875% CI, –1.5 points to –0.4 points; P value < 0.00001). Results of the analyses in the ITT population were consistent with the mITT population results.

In the biologic-naive subgroup, greater improvement in UNRS was seen in the mirikizumab group (–2.7 points) versus the placebo group (–2.1 points), with an LSM mean difference of –0.6 points at week 12 (95% CI, –1.0 points to –0.2 points; P = 0.002). Similar results were seen in the biologic-experienced subgroup, but with a larger LSM mean difference of –1.5 points at week 12 (95% CI, –2.0 points to –1.0 points; P < 0.001).

In the subgroups of patients using corticosteroids or immunomodulators at baseline, the observed effect sizes for bowel urgency improvement in the mirikizumab group compared to the placebo group were numerically positive but small when compared to the overall population. No subgroup differences were observed.

Health-Related Quality of Life

HRQoL was assessed in the LUCENT-1 trial based on the IBDQ score, EQ-5D-5L score, and SF-36 score.

IBDQ Score

The IBDQ consists of a 32-item list subdivided into 4 dimensions: systemic symptoms, bowel symptoms, emotional function, and social function. Total scores range from 32 to 224, with a higher score indicating a better HRQoL. The IBDQ has been consistently shown to have good internal consistency and test-retest reliability, as well as showing responsiveness to change in IBD.^18-20 Available studies have suggested that an improvement of 30 points from baseline or an improvement of at least 15 points above placebo may constitute a minimal important difference (MID).^21-24 In the LUCENT-1 trial, the mean change from baseline to week 12 in the IBDQ score was 38.4 points for patients in the mirikizumab group and 25.2 points for those in the placebo group, representing a difference of 13.2 points (P < 0.001) in favour of mirikizumab during the induction phase. The MID for the IBDQ score was defined as a change greater than 30 points from baseline as well as an MID of greater than 15 points over placebo. While the improvement within the mirikizumab group demonstrated clinical benefit (exceeding the MID threshold of 30 points), it fell short of meeting the MID threshold of at least 15 points compared to the placebo group as defined by previous studies.^21-24 Nevertheless, the clinical expert believed that the observed change was clinically meaningful in terms of improving QoL.

EQ-5D-5L Score

In the LUCENT-1 study, the mean change from baseline to week 12 in the EQ visual analogue scale (EQ VAS) score was 14.6 points in the mirikizumab group and 9.4 points in the placebo group, with an LSM difference of 5.2 points (95% CI, ||| || |||; P < 0.001). The change in the mirikizumab group was clinically important, but the clinical importance of the difference between groups was uncertain.²⁵

SF-36 Score

In the LUCENT-1 study, the mean change from baseline to week 12 in the SF-36 physical component summary (PCS) score was 5.97 points in the mirikizumab group and 3.90 points in the placebo group, with an LSM difference of 2.07 points (95% CI, 1.21 points to 2.93 points; P < 0.001). The mean change from baseline to week 12 in the SF-36 mental component summary (MCS) score was 5.02 points in the mirikizumab group and 3.42 points in the placebo group, with an LSM mean difference of 1.60 points (95% CI, 0.56 points to 2.63 points; P = 0.002). The change in the mirikizumab group was clinically important, but the clinical importance of the difference between groups was uncertain (i.e., MID = 3 points to 5 points).

Mucosal Healing: HEMI

In the LUCENT-1 study, mucosal healing was assessed based on the HEMI outcome, which considers both histologic and endoscopic outcomes. At week 12, 27.1% of patients on mirikizumab attained HEMI versus 13.9% of patients on placebo, with a common risk difference of 13.4% (99.875% CI, 5.5% to 21.4%; P < 0.00001). The results of the analyses in the ITT population were consistent with the mITT population results.

In both the biologic-naive and biologic-experienced subgroups, more patients attained HEMI at week 12 on mirikizumab than on placebo, with a common risk difference of 17.1% (95% CI, 9.8% to 24.3%; P < 0.001) and 8.4% (95% CI, 2.5% to 14.3%; P = 0.022), respectively. The magnitude of the difference in the biologic-naive subgroup was consistent with the mITT population.

In the subgroups of patients using corticosteroids or immunomodulators at baseline, the observed effect sizes for HEMI in the mirikizumab group compared to the placebo group were numerically positive but small when compared to the overall population. No subgroup differences were observed.

Work Productivity

In the LUCENT-1 study, work productivity was assessed using the WPAI:UC score at week 12. The WPAI:UC is a self-administered, disease-specific scale aimed at measuring the level of work impairment due to UC.²⁶ WPAI:UC considers the 4 domains of absenteeism, presenteeism, overall work performance, and nonwork activities.²⁷ The 4 domains comprise a total of 6 items. The final scores for each domain are a percentage of total impairment, ranging from 0 to 100%, with a higher number indicating greater impairment in that domain. The WPAI:UC is a valid and responsive instrument for use in UC.²⁷

Among those employed at baseline (n = 566), patients on mirikizumab experienced a mean change in WPAI:UC of –20.65, compared to –14.91 for patients in the placebo group (LSM difference = –5.74 points; 95% CI, –10.06 points to –1.42 points; P = 0.009).

Maintenance Period: LUCENT-2 Study

Clinical Remission

Clinical remission was assessed using 3 different outcomes in the LUCENT-2 trial: clinical remission, alternate clinical remission, and durable clinical remission. The same definitions of clinical remission and alternate clinical remission used in the LUCENT-1 trial were used in the LUCENT-2 study. Durable clinical remission was attained if patients who experienced clinical remission at week 12 in the LUCENT-1 trial had ongoing remission at week 40 in the LUCENT-2 trial (i.e., 52 weeks of continuous clinical remission). These were considered appropriate measures by the clinical expert.

Clinical Remission

A greater proportion of patients on mirikizumab 200 mg SC (49.9%) versus placebo (25.1%) experienced clinical remission after 40 weeks of maintenance therapy (common risk difference = 23.2%; 95% CI, 15.2% to 31.2%; P < 0.001). Analyses in the PP and ITT populations were consistent with the mITT population results. In addition, results of the sensitivity analyses were consistent with the results from the primary analysis. In terms of the tipping point analysis, the difference between groups did not reach statistical significance when imputing missing data as “responder” for the placebo group and “nonresponder” for the mirikizumab group.

In terms of subgroups, more patients on mirikizumab than on placebo were clinical remitters at the end of the LUCENT-2 study for both the biologic-naive subgroup (51.5% versus 30.7%) and the biologic-failed subgroup (46.1% versus 15.6%; P < 0.001) with a common risk difference of 20.8% (95% CI, 10.2% to 31.5%; P < 0.001) and 30.5% (95% CI,18.1% to 42.9%; P < 0.001), respectively. The magnitude of the effect in both subgroups was consistent with the primary analysis. Note that the biologic-failed subgroup consisted of patients whose disease had an inadequate response to or loss of response to biologic therapy for UC, or they were intolerant to biologic therapy for UC.

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (no), and patients with severe UC at baseline were consistent with the results of the primary analysis. Subgroup results for immunomodulator use (yes) and patients with moderate UC at baseline were numerically positive but small when compared to the overall population. No subgroup differences were observed.

Alternate Clinical Remission

Results for alternate clinical remission, a slightly less stringent definition of remission, were very similar to those of clinical remission.

Durable Clinical Remission

Of patients who experienced clinical remission at week 12 of the LUCENT-1 study, 63.6% who were randomized to mirikizumab 200 mg SC were still in clinical remission at week 40 of the LUCENT-2 trial, compared to 36.9% of those randomized to placebo SC, with a common risk difference of 24.8% (95% CI, 10.4% to 39.2%; P < 0.001). Analyses in the ITT population were consistent with the mITT population results.

In the biologic-naive subgroup, a greater number of patients in the mirikizumab group (62.5%) versus the placebo group (46.8%) attained durable clinical remission at the end of the LUCENT-2 study, with a common risk difference of 15.7% (95% CI, –1.3% to 32.7%; P = 0.078), although the effect size was small compared to the overall population. In the biologic-failed subgroup, a greater proportion of patients in the mirikizumab group versus the placebo group (66.7% versus 11.1%) experienced clinical remission at week 40, with a common risk difference of 55.6% (95% CI, 34.4% to 76.7%; P < 0.001). However, sample sizes for this subgroup were quite small.

Subgroup results for baseline corticosteroid use (no), baseline immunomodulator use (no), and patients with moderate and severe UC at baseline were consistent with the results of the primary analysis. Subgroup results of patients with baseline corticosteroid use (yes) and immunomodulator use (yes) were numerically positive but small when compared to the overall population. No subgroup differences were observed.

Corticosteroid-Free Remission

Corticosteroid-free remission was defined as clinical remission at week 40, symptomatic remission at week 28, and no corticosteroid use for at least 12 weeks before week 40.

A greater number of patients randomized to mirikizumab experienced corticosteroid-free remission at week 40 (44.9%) than those randomized to placebo (21.8%) (common risk difference = 21.3%; 95% CI, 13.5% to 29.1%; P < 0.001).

In both the biologic-naive and biologic-failed subgroups, more patients experienced corticosteroid-free remission on mirikizumab than on placebo, with a common risk difference of 20.4% (95% CI, 10.1% to 30.8%; P < 0.001) and 26.6% (95% CI, 14.5% to 38.6%; P < 0.001), respectively. The magnitude of effect in both subgroups was consistent with the primary analysis.

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (yes and no), and patients with moderate and severe UC at baseline were consistent with the results of the primary analysis.

Endoscopic Remission

At week 40 of the LUCENT-2 study, 58.6% of patents on mirikizumab experienced endoscopic remission versus 29.1% of patients on placebo, with a common risk difference of 28.5% in favour of mirikizumab (95% CI, 20.2% to 36.8%; P < 0.001). Analyses in the ITT population were consistent with the mITT population results.

In both the biologic-naive and biologic-failed subgroups, more patients experienced endoscopic remission at week 40 on mirikizumab than on placebo, with a common risk difference of 28.2% (95% CI, 17.5% to 39.0%, P < 0.001) and 30.5% (95% CI, 17.3% to 43.6%, P < 0.001), respectively.

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (yes and no), and patients with moderate and severe UC at baseline were consistent with the results of the primary analysis.

Bowel Urgency

In the LUCENT-2 study, bowel urgency outcomes consisted of bowel urgency remission and bowel urgency improvement as measured by the UNRS. The UNRS is an instrument used to assess patient-reported severity of bowel urgency in adults with UC with a 24-hour recall period.¹⁷

Bowel Urgency Remission

Of the patients with a UNRS score of at least 3 at the LUCENT-1 study baseline, 42.9% of patients on mirikizumab and 25% of patients on placebo at week 40 experienced bowel urgency remission, with a common risk difference of 18.1% in favour of mirikizumab (95% CI, 9.8% to 26.4%; P < 0.001). Analyses in the ITT population were consistent with results in the mITT population.

In both the biologic-naive and biologic-failed subgroups, more patients experienced bowel urgency remission at week 40 on mirikizumab than on placebo, with a common risk difference of 17.9% (95% CI, 7.0% to 28.8%; P = 0.002) and 30.5% (95% CI, 17.3% to 43.6%; P < 0.001), respectively. The magnitude of the effect was similar with the primary analysis.

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (yes and no), and patients with moderate and severe UC at baseline were consistent with the results of the primary analysis.

Bowel Urgency Improvement (Change in UNRS)

At week 40 of the LUCENT-2 trial, patients on mirikizumab experienced a –3.80-point change in UNRS versus the LUCENT-1 study baseline, while patients randomized to placebo had a –2.74-point change in score from the LUCENT-1 study baseline (LSM difference = –1.06; 95% CI, –1.51 to –0.61; P < 0.001). Patients on mirikizumab experienced a clinically significant improvement in bowel urgency from baseline while those on placebo did not meet the MID threshold (MID = 3 points from baseline)²⁸ for clinically significant improvement.

Health-Related Quality of Life

HRQoL was assessed in the LUCENT-2 study based on the IBDQ score, EQ-5D-5L score, and SF-36 score.

IBDQ Score

The IBDQ consists of a 32-item list subdivided into 4 dimensions: systemic symptoms, bowel symptoms, emotional function, and social function. Total scores range from 32 to 224, with a higher score indicating a better HRQoL. The IBDQ has been consistently shown to have good internal consistency and test-retest reliability, as well as showing responsiveness to change in IBD.^18-20 Available studies have suggested that an improvement of 30 points from baseline or an improvement of at least 15 points above placebo may constitute an MID.^21-24 In the LUCENT-2 study, the LSM mean change from the LUCENT-1 study’s baseline to week 40 in the IBDQ score was 49.8 points and 25.4 points for those in the mirikizumab group and placebo group, respectively, representing a statistically significant difference of 25.2 points in favour of mirikizumab (95% CI, 19.2 points to 31.3 points; P < 0.001).The difference between groups was considered clinically meaningful as the difference was above the MID of at least 15 points above placebo.^21-24

EQ-5D-5L Score

In the LUCENT-2 study, the LSM difference at week 40 between groups in the EQ VAS score was 20.1 points in the mirikizumab group and 8.8 points in the placebo group, representing a statistically significant difference of 11.3 points between groups (MID = 14.6 points) (95% CI, ||| || ||||; P < 0.001).The change from baseline in the mirikizumab group appeared clinically important (with an MID of 14.6 points on the EQ VAS),²⁵ but the clinical importance of the difference between groups was uncertain.

SF-36 Score

Only the LUCENT-2 study evaluated change in the health outcome SF-36. At week 40, patients randomized to mirikizumab experienced an LSM change in SF-36 PCS of 9.0 points, compared to 6.7 points in patients randomized to placebo, a 2.3-point difference between the groups (P < 0.001). In the MCS of SF-36, mirikizumab patients had an LSM change of 7.0 points, compared to 5.5 points in the placebo group (LSM change difference between groups = 1.5; P = 0.031). The change from baseline in the mirikizumab group appeared clinically important (MID threshold of at least 3 points),²⁹ but it was unclear whether the difference between groups was clinically important.

Mucosal Healing: HEMR

A greater proportion of patients randomized to mirikizumab attained HEMR (a stricter outcome than HEMI) more often than those randomized to placebo at week 40: 43.3% versus 21.8%, respectively (common risk difference = 19.9%; 95% CI, 12.1% to 27.6%; P < 0.001). Analyses in the ITT population were consistent with the mITT population results.

In both the biologic-naive and biologic-failed subgroups, HEMR occurred more often in patients on mirikizumab than in patients on placebo, with a common risk difference of 20.8% for patients who were biologic-naive (95% CI, 10.5% to 31.2%) and 21.2% for the biologic-failed population (95% CI, 10.9% to 31.4%). The magnitude of the effect was consistent with the primary analysis.

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (yes and no), and patients with moderate and severe UC at baseline were consistent with the results of the primary analysis.

Work Productivity

The WPAI:UC is a self-administered, disease-specific scale aimed at measuring the level of work impairment due to UC.²⁶ WPAI:UC considers the 4 domains of absenteeism, presenteeism, overall work performance, and nonwork activities.²⁷ The 4 domains comprise a total of 6 items. The final scores for each domain are a percentage of total impairment, ranging from 0 to 100%, with a higher number indicating greater impairment in that domain. The WPAI:UC is a valid and responsive instrument for use in UC.²⁷ At week 40 of the LUCENT-2 trial, patients randomized to the mirikizumab group had an LSM change of –31.72 points from the LUCENT-1 trial baseline, and placebo patients had an LSM change of –22.59 points from the LUCENT-1 trial baseline, equating to an LSM difference of –9.13 points between the groups (95% CI, –14.26 points to –4.01 points; P < 0.001). An MID was not identified for this outcome.

LUCENT-2 Study Extended Induction

Patients whose UC did not respond to mirikizumab or placebo during the 12-week induction period in the LUCENT-1 study went on to the LUCENT-2 study to receive extended induction (an additional 12 weeks) with open-label mirikizumab 300 mg IV for 3 doses. Of the mirikizumab induction nonresponders from the LUCENT-1 study, 272 patients entered the open-label extended induction arm of the LUCENT-2 study in the mITT population. Of these, 146 patients (53.7%) attained a delayed clinical response (95% CI, 47.8% to 59.6%) at week 12 of the LUCENT-2 trial (i.e., after 24 weeks of continuous mirikizumab 300 mg IV every 4 weeks, for a total of 6 doses). Additionally for this cohort of 272 patients, at week 12 of the LUCENT-2 trial, the rates of clinical remission, endoscopic remission, and symptomatic remission were 11.4% (95% CI, 7.6% to 15.2%), 16.5% (95% CI, 12.1% to 21.0%), and 37.1% (95% CI, 31.4% to 42.9%), respectively. When considering clinical response at the end of the initial 12-week induction period and the extended induction period, it can be noted that 80% of patients (697 of 868 patients) on mirikizumab 300 mg IV attained a clinical response by the end of the 24 weeks.

A total of 146 patients were considered delayed responders at week 12 in the LUCENT-2 study. Of the 146 patients, 144 (99%) entered the open-label maintenance period, and 104 (72.2%) maintained clinical response at week 40 versus || ||| || || ||||||| of patients from the placebo group who entered the maintenance period. Clinical response at week 40 was not evaluated in the LUCENT-2 study for induction responders from the LUCENT-1 study. Hence, no comment can be made on the difference in treatment effects between these 2 cohorts at week 40.

Harms Results

The key harms results from the pivotal trials are summarized in Table 2.

For both the LUCENT-1 and LUCENT-2 trials, the overall rate of adverse events (AEs) was similar between groups, though numerically slightly higher in the placebo groups compared to the respective mirikizumab treatment groups. In the LUCENT-1 trial, 44.5% and 46.1% of patients reported an AE in the mirikizumab and placebo groups, respectively. In the LUCENT-2 trial, 64.5% and 68.8% of patients reported an AE in the mirikizumab and placebo groups, respectively. In the LUCENT-1 study, the most common AEs for patients on mirikizumab 300 mg IV included nasopharyngitis (mirikizumab = 4.1%; placebo = 3.1%), anemia (mirikizumab = 3.3%; placebo = 5.9%), and headache (mirikizumab = 3.3%; placebo = 2.8%). In the LUCENT-2 study, the most common AEs for patients on mirikizumab 200 mg SC included nasopharyngitis (mirikizumab = 7.2%; placebo = 5.7%), arthralgia (mirikizumab = 6.7%; placebo = 4.2%), and UC (mirikizumab = 6.7%; placebo = 20.8%).

The rate of SAEs in the LUCENT-1 study was found to be lower in patients treated with mirikizumab than in those treated with placebo (2.8% versus 5.3%); however, this was due to UC being included as a harm. In the LUCENT-2 study, 3.3% and 7.8% of patients reported an SAE in the mirikizumab and placebo groups, respectively. In the LUCENT-1 trial, the most common SAEs in those on mirikizumab IV included UC (mirikizumab = 0.8%; placebo = 3.1%) and pneumonia (mirikizumab = 0.2%; placebo = 0%). In the LUCENT-2 trial, no SAE (at the “preferred term” level) occurred in more than 1 patient on mirikizumab SC.

Withdrawals due to AEs occurred at a lower rate in mirikizumab-treated patients compared to placebo-treated patients in the LUCENT-1 and LUCENT-2 trials. In the LUCENT-1 study, 1.6% and 7.2% of patients withdrew from the trial due to an AE in the mirikizumab and placebo groups, respectively. In the LUCENT-2 study, 1.5% and 8.3% of patients withdrew from the trial due to an AE in the mirikizumab and placebo groups, respectively. |||||||||| ||||||| ||||||||||||||||| ||||||||||||| |||||||||||||||| |||||||||||||||| ||||||||||||||||| ||||||||||| ||| ||||||||||| ||||||||||||||||| |||||||||| |||| ||| |||| |||||| || ||||| || ||||||||||| || || ||||||||| |||||||||| ||||||| ||| ||| |||| || ||||||||| |||| |||| ||||||| ||||||||||| |||||| |||||| ||||| || ||||||||| ||| |||| || ||||||||||| || |||||||| ||| |||| || ||||||| |||||||| ||||||||| |||| || || |||||| ||| |||||||| ||||||||||

In the LUCENT-1 study, no deaths were recorded. In the LUCENT-2 study, 1 (0.5%) death was recorded in the placebo group due to COVID-19.

Most adverse events of special interest (AESIs) occurred at a similar rate between mirikizumab and placebo patients in the LUCENT-1 and LUCENT-2 studies. One exception was the rate of injection site reactions in the LUCENT-2 trial, where 8.7% of patients on mirikizumab SC experienced this AESI compared to 4.2% of patients on placebo SC. The rates of opportunistic infection, cerebrocardiovascular events, malignancy, depression, suicide/self-injury, and hepatic-related AEs were low overall and similar between groups for both the LUCENT-1 study and the LUCENT-2 study.

Table 2: Summary of Key Results From Pivotal Studies and RCT Evidence (mITT Population)

CI = confidence interval; EQ-5D-5L = 5-Level EQ-5D; HEMI = histologic endoscopic mucosal improvement; HEMR = histologic endoscopic mucosal remission; IBDQ = Inflammatory Bowel Disease Questionnaire; LSM = least squares mean; MCS = mental component summary; mITT = modified intention-to-treat; NA = not applicable; PCS = physical component summary; q.4.w. = every 4 weeks; RCT = randomized controlled trial; SC = subcutaneous; SE = standard error; SF-36 = Short Form (36) Health Survey; TEAE = treatment-emergent adverse event; UNRS = Urgency Numeric Rating Scale; WDAE = withdrawal due to adverse event; WPAI:UC = Work Productivity and Activity Impairment Questionnaire: Ulcerative Colitis.

Notes: Details in Table 2 have been taken from the sponsor’s Summary of Clinical Evidence.³⁰

Results for the LUCENT-1 trial outcomes are at 12 weeks while the LUCENT-2 trial outcomes are at 40 weeks post–LUCENT-2 trial baseline, unless otherwise specified. Biologic-experienced was defined as patients whose UC had failed to respond to at least 1 or more biologic therapy or tofacitinib.

^aIn the LUCENT-2 study, UNRS, IBDQ, EQ-5D-5L, and WPAI:UC scores were reported as change from the LUCENT-1 study baseline to week 40 (e.g., 52 continuous weeks).

^bThe ANCOVA model included treatment, baseline value, prior biologic or tofacitinib failure (yes or no), baseline corticosteroid use (yes or no), baseline disease activity (modified Mayo score of 4 to 6 vs. 7 to 9), and region (North America, Europe, or other).

^cThe outcome, WPAI:UC, was the overall work impairment score that combined absenteeism and presenteeism.

Sources: Clinical Study Report for the LUCENT-1 study and the LUCENT-2 study.^31,32

Critical Appraisal

Internal Validity

Overall, the LUCENT-1 and LUCENT-2 trials were well conducted. They were adequately powered to detect a difference between mirikizumab and placebo in the primary end point and employed an appropriate prespecified graphical multiple testing approach to control key secondary outcomes for multiplicity. Many of the primary and secondary outcomes, including clinical remission, alternate clinical remission, clinical response, symptomatic remission, bowel urgency remission and improvement, HRQoL, and work productivity, may have been at risk of reporting bias and recall bias, due to the subjective nature of the patient electronic reporting diary. However, the direction and magnitude of the bias is unknown. As well, there was a risk of attrition bias against mirikizumab due to higher attrition in the placebo arm compared with the intervention; however, sensitivity analyses of the primary end point and the key secondary end points of clinical remission and clinical response assessed the impact of missing data and showed that the results were consistent with the primary analysis, increasing certainty of the findings. MIDs were provided by the sponsor for the IBDQ, EQ-5D-5L, and SF-36 (PCS and MCS), which were in line with thresholds reported in the literature. Numerically, statistically significant improvements were observed in the mirikizumab group compared to the placebo group |||||| ||| ||| |||||||||||||| ||||| |||||||||||. Notably, the IBDQ did reach the MID threshold for the change from baseline score (i.e., > 30 points) in the mirikizumab treatment groups in both the LUCENT-1 study and LUCENT-2 study, which the clinical expert acknowledged as a meaningful improvement. As for the between-group treatment difference, in the LUCENT-2 study, there was a greater change in the mirikizumab group versus the placebo group in the IBDQ score, exceeding the MID threshold mentioned in the literature of greater than 15 points over placebo. However, in the LUCENT-1 study, the IBDQ score fell short of reaching this 15-point MID threshold over the placebo group.

External Validity

In general, the clinical expert consulted by CADTH considered the baseline demographic and disease characteristics in the pivotal trials to be reflective of patients with moderate to severe UC seen in Canadian clinical practice. Concomitant medication use was also reflective of Canadian clinical practice except for prednisolone, which is not typically used in Canada. In the LUCENT-2 trial, a corticosteroid taper was trialled on all patients in the main cohort. Patients who did not taper their steroid use were allowed to continue their treatment; however, this is in contrast to the input received from the clinical expert whereby patients would be considered treatment failures and discontinue therapy if they could not taper or stop concomitant corticosteroid use by the time of the maintenance phase (i.e., after the induction or extended induction period).³³ Therefore, the efficacy of mirikizumab in the trials may appear to be biased, given that patients who could not taper were included in the primary analysis, even though they would have been considered treatment failures in clinical practice. However, the direction and magnitude of this bias is unknown, given that both groups underwent the same tapering protocol. Furthermore, the generalizability of the results may be limited to Canadian clinical practice, given the discrepancy in tapering protocol. The number of screening failures was quite high in the LUCENT-1 study (35%); however, this is similar to other UC trials.^34,35 According to the clinical expert, potential reasons for the higher screening failure rate could be due to how patients were referred to the trial.

To be eligible for enrolment in the primary cohort of the LUCENT-2 trial, patients were required to attain clinical response following 12 weeks of induction treatment in the LUCENT-1 trial. This requirement may have resulted in an enriched patient population that was included in the primary analysis of the maintenance trial as it does not take into consideration delayed responders. As per the product monograph, mirikizumab is indicated for patients who experience delayed response. Hence, by excluding these patients in the primary analysis, there is uncertainty about the efficacy of maintenance treatment in the broader population of patients with moderately to severely active UC. Other UC trials have similar concerns regarding enrichment, given that they have used a similar study design. Patients who entered the LUCENT-2 study as nonresponders received open-label mirikizumab, and therefore the results should be interpreted with caution, considering the potential risk of detection or performance bias due to the open-label nature. The clinical expert noted that the duration of follow-up in the LUCENT-1 trial (12 weeks) was not a sufficient amount of time to see a difference in endoscopic remission. However, the issue of insufficient duration is addressed by the LUCENT-2 trial, which measures end points to week 40 (i.e., 52 weeks of continuous therapy). Long-term data beyond 52 weeks is not available; hence, long-term outcomes (e.g., loss of response, harms) may not be sufficiently captured between the 2 trials.

Long-Term Extension Studies

There are currently no published or unpublished long-term extension phase III or phase IV randomized controlled trials (RCTs) or real-world evidence studies evaluating mirikizumab. The sponsor noted that there is an ongoing phase III, open-label, long-term extension trial enrolling patients from the LUCENT-2 study and the phase II study (NCT02589665) into the LUCENT-3 study (I6T-MC-AMAP), with an expected primary completion date of June 6, 2025.⁹⁷

Indirect Comparisons

Description of Studies

One sponsor-conducted indirect treatment comparison (ITC) indirectly comparing the treatment effect of mirikizumab to other advanced therapies in adult patients with moderate to severe UC via a network meta-analysis (NMA) was included in the sponsor’s submission. In total, || studies evaluating || different treatment regimens were included in the NMA. Of note, among the comparators eligible for inclusion in the NMA, filgotinib and upadacitinib are currently not approved for use in Canada. Moreover, the ustekinumab maintenance regimen of 90 mg every 12 weeks is not used in Canada. Accordingly, these treatment regimens were not reported in the CADTH clinical review. The outcomes assessed in the NMA efficacy analysis included clinical response, clinical remission, and mucosal healing at induction and maintenance, as well as overall SAEs and all-cause discontinuation of treatment.

Efficacy Results

Efficacy results of the NMA are presented for biologic-naive, JAK inhibitor–naive, biologic-experienced, and JAK inhibitor–experienced populations by time points (i.e., induction and maintenance).

||||||||| |||||||| |||||||| ||| ||||||||| ||| ||| ||||||||||||| ||||| ||||||||||| ||| |||||||| ||| |||||||||||| || |||||||||||||| || ||||||||| |||||||| |||||||| |||| ||||| ||| |||| |||| || ||||| ||| ||||||||| |||||||| ||| |||| |||| || ||||| ||||||| |||||||||| ||| |||||||||||| ||| |||| ||| ||| ||||||||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||||||||| |||||||| |||||||| ||| |||||||| |||||||| ||||||| ||||||||||| ||| ||| |||||||| |||||||| |||||| ||||||||||||||| ||| ||||||||||||| ||||||||||| ||||||||||| ||||||||||| || |||||||| |||||||| |||| ||||| ||| |||| |||| || ||||| ||| |||||||| ||||||||| |||| ||||| ||| |||| |||| || ||||| |||||||| ||||||||||| |||||||| || ||||||||||| ||| |||| ||| ||| ||||||||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||||||||| |||||||| |||||||| ||| |||||||| ||||||||| ||||||| ||||||||||| ||| ||| ||||||||| |||||||| |||||| |||||||||||||||||||||||| |||||||| |||||||| ||| ||||||||| ||| ||| ||||||||||||| ||||| ||||||||||| ||||||||| |||| ||||||||||| ||| |||||||||| |||| |||||||||| |||||||||||||||| |||||||| || |||||||||| |||| ||||| ||| |||| |||| || |||||| |||||||||| |||| ||||| ||| |||| |||| || |||||| ||||||||| || || |||| ||||| ||| |||| |||| || ||||| ||||||||| ||| || |||| ||||| ||| |||| |||| || |||||| |||||||| |||| ||||| ||| |||| |||| || |||||| |||||||||||| |||| ||||| ||| |||| |||| || |||||| ||||||||||| || || |||| ||||| ||| ||| |||| || |||||| ||||||||||| ||| || ||| |||| ||||| ||| ||| |||| || |||||| ||||||||||| ||| || ||| |||| ||||| ||| |||| |||| || |||||| ||| ||||||||||| ||| || ||| |||| ||||| ||| |||| |||| || |||||| ||||||||| |||| ||||||||||| ||| |||||||||| |||| |||||||||| |||||||| ||||||||| |||||||| || |||||||||| |||| ||||| ||| |||| |||| || |||||| |||||||||| |||| ||||| ||| |||| |||| || |||||| ||||||||| || || |||| ||||| ||| |||| |||| || ||||| ||||||||| ||| || |||| ||||| ||| |||| |||| || |||||| |||||||| |||| ||||| ||| |||| || |||||| |||||||||||| |||| ||||| ||| ||| |||| || |||||| ||||||||||| || || |||| ||||| ||| |||| |||| || |||||| ||||||||||| ||| || ||| |||| ||||| ||| |||| |||| || |||||| ||||||||||| ||| || ||| |||| ||||| ||| |||| |||| || ||||||||||||||||||| ||| || ||| |||| ||||| ||| |||| |||| || |||||| ||| |||| ||| ||| ||||||||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||||||||||| |||||||| |||||||| ||| ||||||||| ||||||| ||||||||||| ||| ||||||||||| |||||||| |||||| ||||||||||||||| ||| ||||||||||||| ||||||||||| |||||||||| ||| |||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||||||||||| |||||||| |||||||| ||| |||||||| ||||||||| |||||||||||||||||| ||| |||||||| |||||| ||||||||||||

||||||||| ||||||| ||||||||||| ||||||||| ||||||| ||||||| || ||| ||||||||||||| ||||| ||||||||||| ||| |||||||| ||| |||||||||||| || |||||||||||| ||||||| |||||||||| ||| ||||||||||||||| ||||| ||| |||| |||| || |||||| ||| |||| ||| ||| ||||||||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||||||||| ||||||| ||||||| ||||||| ||||||||||| ||| ||| ||||||||| |||||||| |||||| |||||||||||| ||| ||| ||||||||||||| ||||||||||| ||||||||||| ||| |||| ||| ||| |||||||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||||||||| ||||||| ||||||| ||||||| ||||||||||| ||| ||||| |||||||| |||||| |||||||||||||||||||||||| ||||||| |||||||||| ||| ||||||||||||| ||||| ||||||||||| ||| |||| ||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||||||||||| ||||||| ||||||| ||||||| ||||||||||| ||| ||| ||||| |||||| ||||||||||||||| ||| ||||||||||||| ||||||||||| ||||||||||| ||| |||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||||||||||| ||||||| ||||||| ||||||| ||||||||||| ||| ||| ||||| |||||||| |||||| ||||||||||||

Harms Results

Harms outcomes were presented for the overall mixed population regardless of prior exposure to biologic and/or JAK inhibitor therapy.

||| ||||| |||||||||||||||||||||||||||| ||||||||| ||| ||||||||||| |||||||||||| ||||| |||| || ||||||||| ||||||||||||||| |||||||| || |||||||| ||||| |||| || ||||||||| ||||||||||||||| ||| |||||||| ||||||||| ||||||||| |||| ||||||||||| |||||| ||||||||||| |||| ||||| ||| |||| |||| || ||||| ||| |||||||||| |||| ||||| ||| |||| |||| || |||||| ||| |||| ||| ||| ||||||||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||| ||||| ||||||||||||||| ||||||| ||||||||||| ||| ||| ||||| |||||||| |||||| |||||||||||| |||||||| ||||||| |||||||||| |||| ||| ||| |||||||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || |||||||| |||| ||||||| ||||||||||| ||| ||| ||||| |||||||| |||||||||||| ||||||||| ||||||||

Critical Appraisal

The NMA was based on studies identified from a sponsor-conducted systematic literature review of relevant randomized evidence of European Medicines Agency (EMA)-approved and FDA-approved treatments for adult patients with moderately to severely active UC.³⁶ The systematic literature review was based on a PICOS defined a priori and the literature search involved multiple electronic databases, clinical registries, and supplementary manual searches, thereby minimizing error and bias in the study selection and data extraction process. The sponsor identified 3 other sources of heterogeneity across the included UC studies: race, trial design, and prior exposure to biologics and/or JAK inhibitors. To account for racial disparity in UC, the NMA evaluated an “only Asian” subgroup, regardless of prior experience with biologics or JAK inhibitors. The network for this subgroup, however, was small, consisting of 9 and 8 studies at the induction and maintenance periods, respectively, evaluating 6 interventions. To mitigate heterogeneity due to trial design (treat-through design versus rerandomized design), statistical adjustments were employed to make treat-through trials comparable to efficacy data from rerandomized trials, and sensitivity analyses were conducted excluding treat-through study design. However, the CADTH review team was unable to confirm whether the method employed adequately adjusted for differences in trial design without introducing bias. Moreover, follow-up sensitivity analyses excluding studies with treat-through study design were unlikely to account for the potential issues since the network was different and associated with validity issues of its own. To account for the potential for heterogeneity due to treatment history (biologic-naive versus biologic-experienced), the sponsor conducted separate analyses for biologic-naive and biologic-experienced subgroups. However, the CADTH review team determined that the definitions of biologic-naive and biologic-experienced varied across studies (definitions included tumour necrosis factor [TNF]-naive versus TNF-experienced, no biologic or JAK inhibitor failure versus biologic or JAK inhibitor failure, biologic-naive versus biologic-experienced, and no biologic failure versus biologic failure), creating heterogeneity within each classification group. The use of separate analyses and reporting for efficacy results by prior exposure to biologics would not account for these differences. The CADTH review team identified several other sources of heterogeneity that could not be adjusted for in the NMA, including differences in definitions of clinical response and remission, prior biologics exposure (due to time periods in which the studies occurred), permitted concomitant medications, outcome assessment methods and definitions, and the duration of the maintenance period. The inclusion of comparator treatments not relevant to the Canadian setting (i.e., filgotinib, upadacitinib, and maintenance ustekinumab 90 mg every 12 weeks) provided information to the network and was not expected to significantly impact the heterogeneity of the NMA above the other sources of heterogeneity mentioned previously. The violation of the exchangeability assumption for efficacy outcomes is likely due to heterogeneity, and several estimates were affected by wide credible intervals (CrIs) that increased uncertainty. Moreover, network consistency or coherence could not be assessed due to the lack of relevant closed loops when comparing mirikizumab to other active treatments. As a result, the NMA evidence was considered to be indirect, thus reducing certainty in the study findings.

Studies Addressing Gaps in the Pivotal and RCT Evidence

No relevant studies addressing gaps in the pivotal and RCT evidence were submitted.

Conclusions

Two pivotal, multinational, double-blind, randomized placebo-controlled trials — the LUCENT-1 (N = 1,281) and LUCENT-2 (N = 554) trials — and 1 ITC informed the assessment of mirikizumab in this review. Both pivotal trials demonstrated the superiority of mirikizumab over placebo across all end points. The evidence from the LUCENT-1 trial (the induction trial) demonstrated the efficacy of mirikizumab 300 mg IV over placebo in achieving induction clinical remission, alternate clinical remission, clinical response, HRQoL, endoscopic remission, symptomatic remission, bowel urgency improvement, mucosal healing, and work productivity in patients with moderately or severely active UC over 12 weeks. The evidence from the LUCENT-2 study, the maintenance trial, further supported these results after 40 weeks of maintenance dosing. Additionally, the LUCENT-2 trial demonstrated the efficacy of mirikizumab 200 mg SC in achieving corticosteroid-free remission, the maintenance of clinical remission, bowel urgency remission, and mucosal remission at week 40 of the LUCENT-2 study among patients who attained a clinical response by week 12 of the LUCENT-1 study. The results from the primary analysis were considered generalizable to the Canadian landscape; however, it should be noted that the patient population for the LUCENT-2 study may have been enriched, as only responders were rerandomized to the trial, excluding delayed responders who represent a subset of the general population for this indication. Regarding the extended induction, although mirikizumab was able to capture delayed response, the data were considered observational due to the absence of a comparison to the main cohort. Clinically meaningful improvements in HRQoL based on the IBDQ were observed in patients receiving mirikizumab for 40 weeks in the maintenance phase. The NMA comparison between mirikizumab and relevant comparators (i.e., adalimumab, golimumab, infliximab, ozanimod, tofacitinib, ustekinumab, and vedolizumab) did not demonstrate a difference in favour of 1 treatment over another in induction clinical remission and response, mucosal healing, all-cause discontinuation, and SAEs. The apparent benefit of mirikizumab in the maintenance period among patients in the biologic-naive population may be explained by heterogeneity between the trials. Definitive conclusions related to treatment effect and harms of mirikizumab compared to other relevant treatments for UC could not be drawn from the NMA analysis due to substantial heterogeneity in patient characteristics, inclusion criteria (e.g., the definition of prior biologic exposure), prior treatment exposure, and outcome definitions, which likely challenged the underlying exchangeability assumption, and wide CrIs for most estimates. Overall, the LUCENT-1 and LUCENT-2 studies demonstrated clinical efficacy and minimal safety concerns for up to 52 weeks of treatment with mirikizumab in patients with moderate to severe UC and relative to placebo. Evidence of efficacy and safety of mirikizumab beyond 52 weeks, as well as direct comparisons with other treatments for UC, are necessary to further understand the long-term benefits and comparative effectiveness of mirikizumab.

Introduction

The objective of this report is to review and critically appraise the evidence submitted by the sponsor on the beneficial and harmful effects of mirikizumab, 20 mg/mL administered by IV infusion for induction and 100 mg/mL administered by SC injection for maintenance, in the treatment of UC in adult patients with moderately to severely active disease who have had an inadequate response, had a loss of response, or were intolerant to conventional therapy, a biologic treatment, or a JAK inhibitor.

Disease Background

Content in this section has been informed by materials submitted by the sponsor and clinical expert input. The following has been summarized and validated by the CADTH review team.

IBD is a term used to describe disorders that involve chronic inflammation of the digestive tract. There are 2 main types of IBD: Crohn disease and UC. UC causes inflammation and ulcers in the digestive tract, affecting the innermost lining of the large intestine (colon) and rectum.^1,2 UC is characterized by blood in the stool with mucus, frequent diarrhea, loss of appetite, and tenesmus (a strong urge to use the bathroom without necessarily having a bowel movement), in addition to abdominal pain, RB, and weight loss.^3-5 The most common initial manifestation of UC is bloody diarrhea with or without mucus. In addition, patients with UC report high rates of fatigue and sleep difficulties.¹⁰ While the etiology of UC is not completely understood, there is growing evidence to suggest genetic and environmental factors may contribute to the irregular immune response that aberrantly recruits activated immune cells to the colon,³ resulting in chronic inflammation that damages the colon and causes UC symptoms. UC generally develops in young adulthood^6-8 and persists throughout life, marked by periods of spontaneous remission and relapse.⁹ Though most patients experience this relapsing-remitting disease course, up to 24% of patients report experiencing continuous UC symptoms.⁹ The majority of individuals living with UC have a mild to moderate disease course, generally with active disease at diagnosis followed by alternating exacerbations and longer periods of remission.¹⁰ However, aggressive disease course is experienced in 10% to 15% of patients, with a cumulative risk of relapse of between 70% to 80% at 10 years postdiagnosis.¹⁰ Regardless of severity, UC is associated with a substantial reduction in QoL for patients, with considerable impact on many aspects of their lives, including emotional and psychological functioning, social and physical functioning, and work and academic life.^11,12 Chronic active UC may lead to structural damage of the colon; this in turn can lead to dysmotility, chronic symptoms, reduced QoL, and the risk of colon cancer, which requires a colectomy.

UC is diagnosed clinically, with endoscopy, biopsy, and stool sampling being common tests used to rule out other causes of symptoms.¹³ These investigations are typically performed in an outpatient setting, with a gastroenterologist ultimately confirming the diagnosis.^9,10 UC has a worldwide annual incidence rate of 1.2 cases to 20.3 cases per 100,000 people and a prevalence of 7.6 cases to 246.0 cases per 100,000 people.³ The highest age-standardized prevalence rate of IBD in 2017 occurred in high-income countries in North America,³⁷ with Canada having 1 of the highest rates in the world.⁶ The estimated annual incidence rates for UC in Canada range from a low of 8.4 per 100,000 people in Alberta to a high of 21.4 per 100,000 people in Nova Scotia.^6-8 There are an additional 15,000 individuals living with IBD in Canada who do not have a confirmed diagnosis of Crohn disease or UC (termed indeterminate colitis).³⁸ In 2030, the Canadian IBD prevalence is anticipated to rise to 0.98%, equating to a 0.44% prevalence for UC specifically.¹⁵ The incidence of premature mortality associated with UC is no greater than in the general population, but the condition is associated with significant morbidity and an increased risk of colorectal cancer.^13,33 Other potentially severe complications associated with UC that may require hospitalization include severe blood loss, fulminant colitis, perforated bowel, or toxic megacolon.³⁹ About 50% of patients with UC will require hospitalization at some point in their lives, with about a 50% risk of rehospitalization within 5 years for these patients.¹⁰

Standards of Therapy

Content in this section has been informed by materials submitted by the sponsor and clinical expert input. The following has been summarized and validated by the CADTH review team.

UC treatment strategies are dependent on the presence of active disease, severity and extent of the UC, and patient preference. Treatments for UC are commonly separated into 2 groups: conventional therapies and advanced therapies. Conventional therapies include 5-ASA products, corticosteroids, and immunomodulators (such as azathioprine, 6-mercaptopurine, and methotrexate). Corticosteroids are recommended as an initial treatment to bring about complete remission for patients with moderate to severe UC; however, due to serious side effects and a lack of long-term efficacy, corticosteroids are mainly used for the acute treatment of flares and are not recommended for long-term use. Immunomodulators can be considered next, but the use of biologic therapy may be indicated immediately after steroid failure (or prolonged steroid dependence) in many cases. Biologics and JAK inhibitors are often grouped together as “advanced therapies.” Advanced therapies used in UC include adalimumab, golimumab, infliximab, ustekinumab, tofacitinib, ozanimod, and vedolizumab. However, ustekinumab and ozanimod are not publicly reimbursed for UC in Canada. Tofacitinib, a JAK inhibitor, is recommended for use only in patients considered biologic-failed.⁴⁰ For patients whose UC fails on a biologic therapy, a biologic with a different mechanism of action or tofacitinib are reasonable next options.^41,42 Clinicians will aim to optimize treatment by confirming adherence, checking drug trough levels, and adjusting the dose if subtherapeutic. If despite optimization the patient continues to flare, the clinician will switch therapies. When advanced therapies have been exhausted, surgery is the next therapeutic option or referral to a clinical trial.³³ Though surgical treatment of UC is considered curative,⁴³ it comes with the risk of high complication rates.³³ As such, surgery is usually reserved for patients who cannot be managed medically, patients with acute severe UC (toxic megacolon, perforation, and uncontrolled severe hematochezia), or patients who develop colorectal cancers.³³ The clinical expert on this review noted that treatment decisions are often dictated by access to currently available therapies via health insurance rather than clinical guidelines or expert recommendations.

According to the 2015 Canadian guidelines, the treatment goal of UC is “complete remission,” defined as symptomatic remission (normal SF and no blood in the stool) and endoscopic healing (Mayo ES of 0 or 1).⁴² Note that the parameters assessed in determining complete remission (SF, RB, and findings on endoscopy) are the same 3 considered in evaluating the MMS.⁴⁴ Another important treatment goal according to the international 2021 initiative Selecting Therapeutic Targets in Inflammatory Bowel Disease–II (known as STRIDE-II)⁴⁵ is clinical response, defined as at least a 50% improvement in RB and SF. Clinical response is the most immediate target. An intermediate target is clinical remission, defined as Mayo RB and SF subscores of 0, or a Partial Mayo Score of less than 3 with no Mayo subscore greater than 1. Suggested long-term targets include endoscopic healing and improvement in QoL. According to the clinical expert consulted on this review, treatment goals vary by timeline. Short-term goals are aimed at improving clinical response, improving symptoms, and preventing progression. Medium-term goals are to reduce inflammatory burden, discontinue corticosteroid use, and improve endoscopic healing. Long-term goals are sustained steroid-free clinical remission, endoscopic healing, and mucosal remission. However, according to the clinical expert, treatment decisions are often dictated by access to currently available therapies via health insurance rather than clinical guidelines or expert recommendations.

According to the clinical expert, treatments are needed that work both quickly and in the long term, improve convenience and adherence, provide a variety of options, and are well tolerated and safe.

Drug Under Review

Key characteristics of mirikizumab are summarized in Table 3, along with other treatments available for UC.

IL-23 is an important driver of mucosal inflammation in UC and affects the differentiation, expansion, and survival of T-cell subsets and innate immune cell subsets, which represent sources of proinflammatory cytokines.¹⁶ Mirikizumab is a humanized IgG4 monoclonal antibody that binds with high affinity and specificity to the p19 subunit of human IL-23 cytokine and inhibits its interactions with the IL-23 receptor.¹⁶

Mirikizumab is indicated for the treatment of adult patients with moderately to severely active UC who have had an inadequate response, had a loss of response, or were intolerant to conventional therapy, a biologic treatment, or a JAK inhibitor. Mirikizumab received a positive opinion from the EMA for the indication under review.⁴⁶

In the induction phase, mirikizumab is infused intravenously at 300 mg for at least 30 minutes at week 0, week 4, and week 8, and then administered by SC injection at 200 mg every 4 weeks after completion of induction dosing at week 12. For patients who do not have adequate therapeutic response at week 12 after induction dosing, extended induction dosing may be considered by administering 300 mg mirikizumab by IV infusion at week 12, week 16, and week 20. Patients who do not show evidence of therapeutic benefit from extended induction therapy by week 24 should discontinue mirikizumab. If therapeutic benefit is achieved with the additional IV therapy, patients may initiate mirikizumab SC maintenance dosing every 4 weeks.

Table 3: Key Characteristics of Mirikizumab and Main Comparators

6-MP = 6-mercaptopurine; AE = adverse event; CNS = central nervous system; IgG1 = immunoglobulin G1; IgG1k = immunoglobulin G1k; IgG4 = immunoglobulin G4; IL = interleukin; JAK = Janus kinase; MS = multiple sclerosis; S1P = sphingosine-1-phosphate; SC = subcutaneous; TNF = tumour necrosis factor; UC = ulcerative colitis.

^aHealth Canada–approved indication.

Sources: Product monographs for mirikizumab (Omvoh),¹⁶ ozanimod (Zeposia),⁴⁷ ustekinumab (Stelara),⁴⁸ infliximab (Remicade),⁴⁹ vedolizumab (Entyvio),⁵⁰ golimumab (Simponi),⁵¹ tofacitinib (Xeljanz),⁵² and adalimumab (Humira).⁵³

Stakeholder Perspectives

Patient Group Input

This section was prepared by the CADTH review team based on the input provided by patient groups. The full original patient input received by CADTH has been included in the stakeholder section at the end of this report.

Input from patients with moderate to severe UC was received from the GI Society and Crohn’s and Colitis Canada. Patient input from the GI Society was collected using a series of surveys conducted between 2015 and 2022 (n = 54 to 432), focus groups, and 1-to-1 contact with patients with IBD at lectures, patient roundtables, support groups, and via phone, email, and social media interactions, including individual interviews with 3 patients with UC who received mirikizumab in a clinical trial. Patient input from Crohn’s and Colitis Canada was compiled from 2 online surveys conducted in 2022 (n = 354), from individual interviews with 2 patients with UC who participated in the Rinvoq clinical trial, and from 1 phone interview with a patient who participated in the mirikizumab clinical trial.

It was noted that UC has a profound effect on daily life — physically, emotionally, and socially — at home and at school or in the workplace. Many patients surveyed by Crohn’s and Colitis Canada revealed that they hid aspects of their diagnosis from their friends, coworkers, and classmates. Almost two-thirds (63%) of respondents agreed that their family and friends do not understand what they are going through. Patients noted that symptoms can be relentless, embarrassing, and scary. Based on the surveys conducted by Crohn’s and Colitis Canada, the most frequently reported UC-related complications were mental health and stress (65%), joint inflammation and arthritis (51%), anal fissures and hemorrhoids (40%), anemia (33%), skin conditions (30%), malnutrition (30%), and weight loss (30%). Patients stated that sustained remission and/or treatment response is more important than relieving any 1 symptom. The constant concern that there will be future flares, possibly worse than the last, at unpredictable times, was noted as being disastrously disruptive. As expressed by 1 patient surveyed by the GI Society, “The worse part is fear of irreversible permanent damage that will affect your day-to-day life forever.”

Regarding current treatments for UC, it was noted that although there are several available options, most patients have difficulty obtaining remission or adequate symptom relief. Based on survey data from the GI Society, only 24% of patients with IBD found available medications to be adequate, 56% of patients found them to be only somewhat adequate, and 20% of patients found them not at all adequate. More than half of patients (56%) surveyed by Crohn’s and Colitis Canada believed that different treatment options could make them feel better. While steroid use is an important part of symptom management for UC, patients reported not being particularly supportive of the treatment option. Patient input from the GI Society stressed that a variety of treatment options is important because response varies across patients, and there may be a loss of response after prolonged use. Patients noted that there is a need for new and effective options to achieve mucosal healing and reduce the symptoms of UC, and to provide good QoL. Patients interviewed by Crohn’s and Colitis Canada added that any new treatment must be able to protect a patient’s ability to work, attend school and social events, and leave the home to run errands. Patients interviewed by Crohn’s and Colitis Canada also added that potential risks and side effects, especially those related to heart and liver function, are a major source of concern when considering new treatment options.

Among the 4 patients who received mirikizumab in clinical trials, all experienced improved gut healing and continued to take the medication. The initial induction of treatment via infusion was found to be exhausting and time consuming by 2 patients interviewed by the GI Society. Although 2 patients did not particularly like the SC administration of mirikizumab, they added that the injections were manageable. All 4 patients expressed having improved QoL. As expressed by 1 patient interviewed by the GI Society, “My experience has been really good with this drug, and I recommend it. I have no anxieties on going out to public events. I’m definitely a lot more comfortable leaving the house and going about my normal day to day activities.” The 1 patient interviewed by Crohn’s and Colitis Canada stated that treatment with mirikizumab was “[…] a life changer. It's hard sometimes to know what success necessarily looks like, but I would say if I had. If someone asked me, I absolutely have had great success on this medication. I’m thrilled that I have no side effects to deal with. And I would do it all over again in a second. And I'm hoping that it will be available for me to continue and that it continues to work for me.”

Clinician Input

Input From Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise regarding the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, providing guidance on the potential place in therapy). The following input was provided by 1 clinical specialist with expertise in the diagnosis and management of UC.

Unmet Needs

The clinical expert noted that there is an unmet need for treatments that are better tolerated, improve convenience and compliance, and take into consideration special populations such as those with previous or current malignancies. The clinical expert stated that having available treatments with different administration methods is also important to patients (e.g., IV, SC, oral) as well as having multiple treatment options, given that patients will often lose response to treatment and require another therapy. The clinical expert noted that not all patients’ UC responds to available treatments and patients often become refractory to current treatment options. Treatment success (e.g., clinical response, clinical remission) varies from 15% to 60%. The clinical expert stated that it is difficult to predict which patients with UC will respond to treatment, as there is currently no available clinical test. According to the clinical expert, no treatments are available to reverse the progression of IBD or UC (e.g., damage to the colon). However, if UC is treated and controlled early in the disease course, it is possible to stop or delay progression. The expert provided the example of achieving endoscopic remission, which may be able to prevent structural damage of the colon and reduce risk of dysplasia and/or colon cancer.

Place in Therapy

According to the clinical expert, mirikizumab is an anti–IL-23 drug for UC and would offer a novel treatment mechanism for the disease that is considered more selective. The clinical expert stated that mirikizumab’s place in therapy would be similar to other biologics and would be recommended as a second-line therapy after 5-ASA, instead of immunomodulators. For patients whose UC has failed on all available therapies but require a biologic therapy for another condition, mirikizumab may be used as a dual therapy with a biologic. For patients who are biologic-naive, the expert felt that many clinicians may recommend mirikizumab for induction and maintenance as it may be even more targeted than other therapies. Given the available treatments, mirikizumab may be recommended before other treatments that have more side effects or risks, such as tofacitinib. It was the opinion of the clinical expert that patients do not need to initiate and experience treatment failure with other therapies before being prescribed mirikizumab, given the limitations of other therapies. For example, azathioprine and methotrexate have side effect profiles that include risks to the liver and the risk of lymphoma and take longer to act. In addition, the clinical expert noted that previous trials have shown that patients who are biologic-naive have a better chance of achieving a response to treatment compared to patients who are biologic-exposed. Hence, the clinical expert recommends that mirikizumab be offered both to patients with UC who are biologic-naive and who are biologic-experienced.

Patient Population

The clinical expert noted that patients whose UC is most likely to respond to treatment with mirikizumab would be those with moderate to severe UC (biologic-naive or biologic-experienced) that has failed to respond to conventional therapy. In the opinion of the clinical expert, patients who are in most need of mirikizumab would be those whose UC has failed to respond to first-line therapy with 5-ASA. The clinician indicated that patients who have UC and comorbid psoriasis may have further benefit from mirikizumab as research has demonstrated its efficacy in treating psoriasis.⁵⁴ According to the clinical expert, these patients can be identified through clinical examination, lab tests, and colonoscopy. No clinical tests are available to determine whether a patient will be more likely to exhibit a response to mirikizumab. Patients least suited for treatment with mirikizumab, according to the clinical expert, are patients with active infections, malignancy, and severe hepatic impairment, and patients who are pregnant.

Assessing the Response Treatment

According to the clinical expert, the outcomes used in clinical practice align with those used in clinical trials such as clinical remission and response (measured by Partial Mayo Score), endoscopic remission and response, and biomarkers (e.g., fecal calprotectin). The clinical expert noted that clinicians routinely schedule colonoscopy 6 months to 9 months after starting biologics or small-molecule drugs to check for endoscopic healing. The clinical expert indicated that it is difficult to schedule or expect a patient to return for a colonoscopy to assess response in the induction phase due to the short time span (e.g., 12 weeks); therefore, surrogate markers such as fecal calprotectin may be used to approximate objective response. If concerned about patient response, the expert indicated that some physicians may try to book a flexible sigmoidoscopy soon after the induction period has been completed.

According to the clinical expert, a clinically meaningful response to treatment would be no further RB, no rectal incontinence, rectal urgency that has been reduced or no longer exists, bowel movement frequency that has been reduced or is normal, stools becoming more solid, and abdominal pain that has been reduced or no longer exists. The clinical expert would expect clinical improvement within 4 weeks, and clinical remission within 12 weeks; however, depending on the severity of disease and previous medication exposure, the clinical expert noted that patients may have slower response or a delay to remission. In this case, the expert indicated that they would be comfortable with an extended induction period of 12 weeks for nonresponders, which is aligned with the product monograph. The clinical expert stated that there will not be much variation in assessing response among physicians as most gastroenterologists treating IBD use the standard clinical scores (e.g., Partial Mayo Score, MMS) for UC in clinics with an endoscopic component if performing a colonoscopy. Additionally, physicians will monitor with C-reactive protein and fecal calprotectin.

Discontinuing Treatment

According to the clinical expert, mirikizumab should be discontinued in the event of SAEs, disease progression, or the inability to taper off steroids. The clinical expert would consider stopping treatment after 24 weeks of treatment if the patient’s UC does not respond. This would include an extended induction phase if the patient’s UC was not responding to initial induction (e.g., 12 weeks). According to the clinical expert, it might be expected that approximately 30% of patients who did not have an initial induction response would have a delayed response to induction treatment. If a patient had some response from the first induction period (e.g., 12 weeks), but perhaps could not completely taper off steroids, they could be given an extended induction (e.g., additional 12 weeks). At the end of this additional 12-week period, if their UC continues to not respond or they are unable to stop steroids, they would be considered nonresponders. At that point, the clinical expert stated that the patient would be changed to another therapy.

Prescribing Considerations

The clinical expert noted that the community, hospitals (outpatient clinics), and specialty clinics are all appropriate settings for treatment administration. The clinical expert stated that specialists, such as gastroenterologists, should prescribe mirikizumab. In rural settings, a general practitioner or endoscopist trained for IBD management may also prescribe mirikizumab.

Clinician Group Input

This section was prepared by the CADTH review team based on the input provided by clinician groups. The full original clinician group input received by CADTH has been included in the stakeholder section at the end of this report.

One clinician group input submission was received from a group of gastroenterologists in Canada. The clinician group consists of gastroenterologists in Canada who are recognized nationally and internationally as experts in the management of IBD. Input from the clinician group was compiled by 9 clinicians and was based on the literature, clinical practice guidelines for the management of UC, consultation with Canadian and international expert groups, and a review of individual product monographs and primary trial publications, and information from manufacturers of current and future treatments.

Based on input from the clinician group, the goals of UC therapy have evolved from simply controlling symptoms to preventing disease progression, surgery, and disability with early intervention and a treat-to-target approach. The clinician group noted that treatment for UC is influenced by disease characteristics and may involve multiple medications, including 5-ASA, corticosteroids, immunosuppressants, biologics (which include anti-TNF therapy, anti-integrin therapy, and anti–IL-12 and anti–IL-23 therapy) and advanced small molecule drugs (which include JAK inhibitors and S1P receptor modulators). The clinician group stressed that treatment with corticosteroids is reserved for periods of disease flare-up and not prescribed for long-term use due to side effects and poor effectiveness for maintaining remission. The clinician group noted that surgery is necessary in UC when medications prove ineffective, for fulminant disease which is medically refractory, or if complications arise, such as dysplasia or strictures. The clinician group emphasized that the ongoing need for surgery in patients with UC reflects the ongoing need for medical therapies that can prevent complications and meet treatment goals. Accordingly, the clinician group identified unmet needs for the treatment of moderate to severe UC. This includes the need for therapy that induces symptomatic remission, can be used to maintain symptomatic remission, is safe with long-term use, and can rapidly improve endoscopic appearance of the bowel and maintain this in the long term. The clinician group emphasized that none of the currently available therapies for UC meet all current needs of patients in terms of short-term or long-term treatment. Remission with treatment is not universal and patients can lose response after an initial period of improvement and relapse, even after long periods of remission on an existing therapy. Accordingly, the clinician group stressed the need for novel therapies targeting alternative pathways.

Overall, the clinician group suggested that mirikizumab has the potential for a broad range of uses in clinical practice, from first-line advanced therapy to treatment of patients with inadequate response or intolerance to multiple advanced therapies. The clinician group anticipates that mirikizumab may become a treatment of choice for most patients with UC as it appears to have a favourable benefit-risk ratio. The general consensus among the clinician group members was that mirikizumab may be a good candidate to help address unmet needs as a potential first-line option or for individuals whose UC has failed to respond to other advanced therapies. The clinician group recommended that management strategies using mirikizumab aim for remission, which is defined as both symptomatic and endoscopic remission and aligns with Canadian consensus guidelines. The clinician group suggested that a meaningful improvement in symptoms as measured by the resolution of SF and RB should be demonstrated in the first 3 months of therapy. The clinician group expected patients to be in symptomatic remission and off corticosteroids by 6 months. In addition, the clinician group indicated that symptomatic improvement should be accompanied by a decrease in biomarkers of inflammatory activity (C-reactive protein and fecal calprotectin) in the first 3 months after initiating mirikizumab. The clinician group suggested that it would be appropriate to discontinue treatment with mirikizumab in the event of worsening symptoms or inadequate response. In circumstances where there was an inadequate response to mirikizumab as a first-line biologic, the clinician group indicated that a switch to another class of drugs, such an anti-TNF drugs, may be warranted. Based on clinical experience, the clinician group suggested that it would be appropriate for a trained health care professional to administer mirikizumab in a clinic during the induction phase.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may impact their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 4.

Table 4: Summary of Drug Plan Input and Clinical Expert Response

5-ASA = 5-aminosalicylic acid; CDEC = CADTH Canadian Drug Expert Committee; IBD = inflammatory bowel disease; UC = ulcerative colitis.

Clinical Evidence

The objective of CADTH’s Clinical Review Report is to review and critically appraise the clinical evidence submitted by the sponsor on the beneficial and harmful effects of mirikizumab, 300 mg/15 mL IV (induction) and 100 mg/1 mL SC injection (maintenance), in the treatment of adult patients with moderately to severely active UC who have had an inadequate response, had a loss of response, or were intolerant to conventional therapy, a biologic treatment, or a JAK inhibitor.The focus will be on comparing mirikizumab to relevant comparators and identifying gaps in the current evidence.

A summary of the clinical evidence included by the sponsor in the review of mirikizumab is presented in 2 sections, and CADTH’s critical appraisal of the evidence is included after each section. The first section, the systematic review, includes pivotal studies and RCTs that were selected according to the sponsor’s systematic review protocol. The second section includes indirect evidence from the sponsor.

Included Studies

Clinical evidence from the following is included in the CADTH review and appraised in this document:

• two pivotal studies (placebo-controlled RCTs)

• one systematic review with an NMA.

Pivotal Studies and RCT Evidence

Content in this section has been informed by materials submitted by the sponsor. The following has been summarized and validated by the CADTH review team.

Description of Studies

The LUCENT-1 and LUCENT-2 trials were pivotal phase III trials evaluating the safety and efficacy of mirikizumab for UC in the induction phase (the LUCENT-1 study) and maintenance phase (the LUCENT-2 study). Characteristics of the included studies are summarized in Table 5.

Table 5: Details of Pivotal Studies and RCT Evidence Identified by the Sponsor

EQ-5D-5L = 5-Level EQ-5D; ES = endoscopic Mayo subscore; HEMI = histologic endoscopic mucosal improvement; HEMR = histologic endoscopic mucosal remission; IBDQ = Inflammatory Bowel Disease Questionnaire; LOR = loss of response; MCS = mental component summary; MMS = Modified Mayo Score; NA = not available; NR = not reported; q.4.w. = every 4 weeks; PCS = physical component summary; RCT = randomized controlled trial; SC = subcutaneous; SF-36 = Short Form (36) Health Survey; UC = ulcerative colitis; UCEIS = Ulcerative Colitis Endoscopic Index of Severity; UNRS = Urgency Numeric Rating Scale; WPAI:UC = Work Productivity and Activity Impairment Questionnaire: Ulcerative Colitis.

Note: Details in Table 5 have been taken from the sponsor’s Summary of Clinical Evidence.^20,30

Sources: Clinical Study Reports for the LUCENT-1 and LUCENT-2 studies.^31,32

LUCENT-1 Study

The LUCENT-1 trial (N = 1,281) was a 12-week, multicentre, randomized, double-blind, parallel-arm induction trial evaluating the safety and efficacy of mirikizumab 300 mg IV versus placebo in adults with moderately to severely active UC. The primary objective of the LUCENT-1 study was to evaluate the ability of mirikizumab to induce clinical remission at week 12. Major secondary objectives included the ability of mirikizumab to induce alternate clinical remission, clinical response, endoscopic remission, symptomatic remission, clinical response in patients who are biologic-experienced, bowel urgency improvement, and HEMI, all at week 12, as well as symptomatic remission at week 4. The LUCENT-1 study consisted of a 4-week screening period and a 12-week induction dosing period. At the end of the induction period, patients had the choice of transitioning to the LUCENT-2 study (a maintenance trial) or discontinuing study treatment and completing the 16-week posttreatment follow-up period. A study design flow diagram of the LUCENT-1 and LUCENT-2 studies is outlined in Figure 1.

The LUCENT-1 study consisted of patients from North America, Europe, Asia, South America, and Australia, and included patients from 9 Canadian sites. Patients were randomized to receive either blinded mirikizumab 300 mg IV or placebo IV every 4 weeks in a 3:1 ratio. Randomization was stratified by biologic-experienced status (yes or no), baseline corticosteroid use (yes or no), baseline disease activity (MMS of 4 to 6 or 7 to 9), and region (North America, Europe, or other). The LUCENT-1 study data discussed is based on a database lock date of June 11, 2021. Patients who completed the LUCENT-1 trial were given the option to enrol in the LUCENT-2 trial.

LUCENT-2 Study

The LUCENT-2 study (N = 554, the primary analysis) was a 40-week, multicentre, randomized, double-blind, parallel-arm maintenance trial evaluating the safety and efficacy of mirikizumab 200 mg SC in adults with moderately to severely active UC, in the main cohort analysis. The primary objective of the LUCENT-2 trial’s main cohort was to evaluate the ability of mirikizumab to induce clinical remission at the LUCENT-2 study’s week 40 (week 52 of continuous therapy). Major secondary objectives included the ability of mirikizumab to induce alternate clinical remission, durable clinical remission, endoscopic remission, corticosteroid-free remission, HEMR, bowel urgency improvement, and bowel urgency remission, all at week 40. In addition to the main cohort analysis, subcohort and other cohort analyses in the LUCENT-2 study were aimed at determining the safety and efficacy of extended induction with mirikizumab IV in mirikizumab induction nonresponders from the LUCENT-1 and LUCENT-2 trials, and the safety and efficacy of mirikizumab IV reinduction in the loss of response cohort originally randomized in the LUCENT-2 trial to mirikizumab SC. Of note, treatment of the reinduction cohort is not aligned with the recommended dosing in the Health Canada–approved product monograph and therefore will not be discussed further in this report.

Figure 1: Study Design of LUCENT-1 and LUCENT-2 Studies, Including All Cohorts

AMAN = LUCENT-1 study; AMAP = LUCENT-3 study; AMBG = LUCENT-2 study; clin. = clinical; D/C = discontinue; LOR = loss of response; miri = mirikizumab; OL = open-label; PBO = placebo; Q4W = every 4 weeks; R = randomization; RB = rectal bleeding; SC = subcutaneous; SF = stool frequency; W0 = week 0; W12 = week 12; W40 = week 40; W52 = week 52.

Note: If patients had clinical benefit with OL rescue therapy with mirikizumab after a loss of response during the treatment period, those patients were eligible to enter the LUCENT-3 study. However, if patients did not experience a clinical benefit after a loss of response during the treatment period, those patients discontinued mirikizumab and entered the posttreatment follow-up period. Similarly, if delayed clinical response was lost during the OL maintenance period after OL extended induction, those patients discontinued mirikizumab and entered the posttreatment follow-up period.

^a Loss of response at or after week 12 and up to and including week 28. LOR is defined as a 2-point or greater increase from the maintenance baseline in the combined SF plus RB scores, and a combined SF plus RB score of at least 4 on 2 consecutive visits and confirmed by an endoscopic subscore of 2 or 3.

Source: Clinical Study Report for the LUCENT-2 study.³²

The main cohort of the LUCENT-2 study included 544 patients from the LUCENT-1 study who received and whose UC responded to blinded mirikizumab IV (i.e., induction dosing) at 12 weeks. These patients were randomized to receive either blinded mirikizumab 200 mg SC or placebo SC in a 2:1 ratio. Randomization was stratified by biologic-experienced status (yes or no), baseline corticosteroid use (yes or no), region (North America, Europe, or other), and induction remission status (yes or no). Although the LUCENT-2 study main cohort included 544 patients, the total number of patients who entered the LUCENT-2 study was 1,178, including patients from the same 9 Canadian clinics as in the LUCENT-1 trial. Additional cohorts of patients included the following.

• Placebo induction responders: Patients who received blinded placebo IV in the LUCENT-1 trial and whose UC responded to placebo were continued on blinded placebo SC every 4 weeks throughout the LUCENT-2 study.

• Extended induction: Patients whose UC did not respond to mirikizumab or placebo during the 12-week induction period in the LUCENT-1 study went on to the LUCENT-2 trial to receive extended induction (an additional 12 weeks) with open-label mirikizumab 300 mg IV for 3 doses. Those whose UC responded 12 weeks after the first extended induction dose were referred to as “delayed clinical responders” and received open-label mirikizumab 200 mg SC every 4 weeks (e.g., maintenance dosing) for the remainder of the LUCENT-2 study. Those whose UC did not respond after 24 weeks of induction therapy discontinued mirikizumab.

After completing the LUCENT-2 trial, patients from any cohort who had not yet discontinued mirikizumab were given the option of moving to the long-term extension study (the LUCENT-3 study), or of discontinuing treatment and entering the 16-week posttreatment follow-up period. The LUCENT-2 study data discussed is based on a database lock date of December 6, 2021.

Populations

Inclusion and Exclusion Criteria

Key inclusion criteria for the LUCENT-1 study included being aged 18 years to 80 years, and having an established diagnosis of UC at least 3 months before baseline, endoscopic evidence of UC, a histopathology report that supported UC, an MMS of 4 to 9 with an ES of at least 2 within at least 14 days from baseline, UC extending beyond the rectum, and an inadequate response, a loss of response, or an intolerance to a conventional therapy (corticosteroids and/or immunomodulators), a biologic (approved anti-TNF and/or anti-integrin antibodies), or tofacitinib. Patients were excluded from the LUCENT-1 trial if they had UC limited to the rectum, any other forms of IBD, an immunodeficiency syndrome that would cause UC-like colonic inflammation, extensive colonic resection, stricture or stenosis in the small bowel or colon, toxic megacolon, colonic adenoma not yet removed, dysplasia of colonic mucosa, gastrointestinal cancer, and/or an adequate trial of 3 or more biologic therapies for UC (excluding tofacitinib).

Key inclusion criteria for the LUCENT-2 study included patients who completed the LUCENT-1 trial, received at least 1 study drug dose in the LUCENT-1 study, and participated in all necessary evaluations to assess MMS at the end of the LUCENT-1 study. If patients met any of the LUCENT-1 trial exclusion criteria during or at the end of the LUCENT-1 trial, they would be excluded from the LUCENT-2 trial.

Interventions

In the LUCENT-1 study, patients received either mirikizumab 300 mg IV or placebo IV every 4 weeks for 3 doses, infused over 30 minutes, at the study site (either a hospital or infusion centre) by study personnel. Mirikizumab IV was supplied to study sites in a 15 mL vial at a concentration of 20 mg/mL. Placebo IV was also supplied in 15 mL syringes.

In the LUCENT-2 study main cohort, patients received either mirikizumab 200 mg SC or placebo SC every 4 weeks. SC administration for placebo or mirikizumab was done by study personnel at the study site via 2 sequential injections to different body sites, with each injection containing 100 mg of drug in 1 mL. Self-administration at home was allowed in the case of COVID-19–related restrictions.

In both studies, pharmacists who prepared the study drug and nurses and pharmacists who administered the study drug were all blinded. IV mirikizumab and the placebo were visually indistinguishable from each other. Dose modifications were not permitted in either pivotal trial. During the LUCENT-2 study, if a main cohort patient or placebo induction responder lost response between week 12 and week 28 of the LUCENT-2 trial, they were given open-label mirikizumab 300 mg IV every 4 weeks for 3 doses as a rescue therapy, interchangeably referred to as “reinduction.” Patients were stopped from taking the drug if toxicity developed. Patients who completed the LUCENT-2 trial through week 40 (with an acceptable safety profile) were eligible to participate in the long-term extension study (the LUCENT-3 study).

Study treatment could be permanently discontinued due to AEs or patient decision. Patients who discontinued study treatment early underwent early termination procedures, which included an early termination visit and posttreatment follow-up visits.

Concomitant Medications

Concomitant UC therapies that were permitted in both the LUCENT-1 study and the LUCENT-2 study included oral 5-ASA, oral corticosteroids, azathioprine, 6-mercaptopurine, and methotrexate. Patients who required and used a prohibited medication had to discontinue the study drug. The use of nonprohibited concomitant medication was documented at each patient visit.

In the LUCENT-2 trial, a corticosteroid taper was trialled on all patients in the main cohort (i.e., mirikizumab induction responders) depending on the steroid and formulation they were using, as follows (beginning at baseline):

• oral budesonide multimatrix system 9 mg per day — reduce to 9 mg every other day for 2 weeks, then 9 mg every third day for 2 weeks, then discontinue

• oral beclomethasone dipropionate 5 mg per day — 5 mg every other day for 4 weeks, then discontinue

• all other oral steroids:

  • if dose is greater than 10 mg per day prednisone or equivalent — reduce daily dose by 5 mg per week until 10 mg per day, then continue tapering daily dose by 2.5 mg per week until 0 mg per day

  • if dose is less than 10 mg per day prednisone or equivalent — taper daily dose by 2.5 mg per week until 0 mg per day.

If patients could not tolerate the taper without the return of clinical symptoms, the taper was paused or the dose was increased to the original LUCENT-1 study baseline dose. If this occurred, an attempt was made to reinitiate the taper within 2 weeks of taper interruption. The suggested goal was to complete tapering by week 12 of the LUCENT-2 study.

Amendments

One amendment was made to the protocol of the LUCENT-1 and LUCENT-2 trials, referred to as Amendment (a), on September 12, 2019. Major changes included the addition, modification, and deletion of major secondary and other secondary end points, as well as clarifications to the objectives and end points. In addition, updates were made to the schedule of activities (expanding the time window for assessments), the permitted and prohibited medications list, and entry criteria (e.g., criteria excluded patients requiring systemic corticosteroids for non-UC indications except for premedication for study drug infusion or locally). These were minor changes.

A second amendment was made during the LUCENT-1 study on August 21, 2020. It included extending the window for the week-12 endoscopy assessment and study drug administration, permitting the use of local laboratories if a central laboratory was not available, permitting the use of virtual telephone visits, and lastly, allowing a patient missing the week-12 endoscopy to continue to the maintenance study.

Outcomes

A list of efficacy end points assessed in this Clinical Review Report are summarized in Table 6. The summarized end points are based on those included in the sponsor’s Summary of Clinical Evidence³⁰ as well as any end points identified as important to this review according to stakeholders — for example, the clinical expert, clinician groups, or patient groups.

Table 6: Outcomes Summarized From Pivotal Studies and RCT Evidence Identified by the Sponsor

EQ-5D-5L = 5-Level EQ-5D; HEMI = histologic endoscopic mucosal improvement; HEMR = histologic endoscopic mucosal remission; IBDQ = Inflammatory Bowel Disease Questionnaire; MCS = mental component summary; NA = not applicable; PCS = physical component summary; RCT = randomized controlled trial; SF-36 = Short Form (36) Health Survey; UCEIS = Ulcerative Colitis Endoscopic Index of Severity; UNRS = Urgency Numeric Rating Scale; vs. = versus; WPAI:UC = Work Productivity and Activity Impairment Questionnaire: Ulcerative Colitis.

Note: Details in Table 6 have been taken from the sponsor’s Summary of Clinical Evidence.^20,30

^aThe LUCENT-2 study time points mentioned assume that week 0 is the start of the LUCENT-2 trial, unless otherwise indicated. Week 40 of the LUCENT-2 trial represents 52 weeks of continuous therapy from the LUCENT-1 study’s baseline.

^bStatistical testing for primary and key secondary end points was adjusted for multiple comparisons via a prespecified graphical testing approach.

Sources: Clinical Study Reports for the LUCENT-1 and LUCENT-2 studies.^31,32

Efficacy Outcomes

A description of the efficacy outcome measures and their measurement properties that were used in both the LUCENT-1 study and the LUCENT-2 study are presented in Table 7.

Many outcomes in the LUCENT-1 and LUCENT-2 trials were defined by subscores of the Mayo scoring system. The pivotal trials used the MMS as opposed to the overall Mayo Score. Scoring of the MMS subscores are further detailed in Table 7.

Clinical Response

Clinical response was measured in the LUCENT-1 trial at week 12 and in the extended induction cohort of the LUCENT-2 trial at week 12. Clinical response was based on the MMS and defined as:

• a decrease in the MMS total score (the sum of the SF subscore, RB subscore, and ES) of at least 2 points and at least a 30% decrease from baseline

• a decrease of 1 point or less in the RB subscore from baseline or an RB score of 0 or 1.

The SF and RB subscores were recorded by patients in an electronic diary daily during the treatment period. Patient diary adherence reviews were conducted at each treatment visit. For SF, a stool was defined as a trip to the toilet when the patient had either a bowel movement, passed blood alone, passed blood and mucus, or passed mucus only. The total number of stools passed in a 24-hour period was recorded by the patient in a daily electronic diary.

For RB, patients recorded their subscore daily and an average of the subscores was used to calculate the MMS score and RB subscore at the end-of-treatment visit.

The ES is a physician-reported measure that reports the worst appearance of the mucosa on flexible sigmoidoscopy or colonoscopy on a 4-point scale. Endoscopy was used to determine the Mayo ES at screening and week 12 (or end-of-treatment visit) in the LUCENT-1 study and week 40 in the LUCENT-2 study. A flexible sigmoidoscopy or colonoscopy was performed on all patients at each time point by a licensed physician. Site and blinded central reading of endoscopies were used to determine ES. Disagreement between the site and central read were adjudicated by an additional blinded central reader. Scoring of the MMS subscores are further detailed in Table 7.

Clinical Remission

Clinical Remission

Clinical remission was the primary end point in the LUCENT-1 study at week 12 and the LUCENT-2 study at week 40. Clinical remission was based on the MMS and defined as:

• SF subscore equals 0, or SF equals 1 with at least a 1-point decrease from baseline

• RB subscore equals 0

• ES equals 0 or 1 (excluding friability).

The MMS subscores were calculated using the same procedures described for the clinical response.

Alternate Clinical Remission

Alternate clinical remission was measured in the LUCENT-1 study at week 12 and the LUCENT-2 study at week 40. Alternate clinical remission was defined as:

• SF subscore equals 0, or SF equals 1

• RB subscore equals 0

• ES equals 0 or 1 (excluding friability).

The MMS subscores were calculated using the same procedures described for the clinical response.

Durable Clinical Remission

Durable clinical remission at week 40 was assessed in the LUCENT-2 study only. Durable clinical remission was attained if patients who experienced clinical remission at week 12 in the LUCENT-1 study had ongoing remission at week 40 in the LUCENT-2 study (i.e., 52 weeks of continuous clinical remission).

Corticosteroid-Free Remission

Corticosteroid-free remission at week 40 was assessed in the LUCENT-2 study only. Corticosteroid-free remission was defined as:

• clinical remission at week 40

• symptomatic remission at week 28

• no corticosteroid use for at least 12 weeks before week 40.

Endoscopic Outcomes

Endoscopic Remission

Endoscopic remission was assessed in the LUCENT-1 trial at week 12 and the LUCENT-2 trial at week 40. Endoscopic remission was defined as an ES of 0 or 1 (excluding friability) on the MMS at week 12 and week 40 for the LUCENT-1 trial and the LUCENT-2 trial, respectively. The endoscopies were performed by a licensed physician qualified to perform colonoscopies. The site endoscopist determined the Mayo ES at each endoscopy and recorded the result in the electronic case report form. The ES was calculated using the same procedures described for the clinical response.

Endoscopic Disease Activity

Endoscopic disease activity, as measured by the UCEIS score, was assessed in the LUCENT-1 study only. The UCEIS is a physician-reported measure of the endoscopic disease activity of UC on a flexible sigmoidoscopy or colonoscopy. The UCEIS consists of the following 3 descriptors calculated as a simple sum: vascular pattern (scored 0 to 2), bleeding (scored 0 to 3), and erosions and ulcers (scored 0 to 3).^56,57 The 3 descriptors are summed, resulting in a range of UCEIS scores of 0 to 8, with higher scores indicating worse outcomes (i.e., 0 = remission; 8 = severe).⁵⁷ Only blinded central reading of endoscopies was used to determine the UCEIS score for each endoscopy. The proportion of patients with a UCEIS score less than or equal to 1 at week 12 was reported.

Symptoms

Symptomatic Remission

The symptomatic remission rate at week 12 was assessed in the LUCENT-1 study only. The symptomatic remission rate was defined as:

• SF equals 0, or SF equals 1 with at least a 1-point decrease from baseline

• RB equals 0.

The MMS subscores were calculated using the same procedures described for the clinical response.

Bowel Urgency Remission (in Patients With UNRS Score of 3 or More at Baseline)

Bowel urgency remission at week 40, as measured by the UNRS, was assessed in the LUCENT-2 study only. Bowel urgency remission was defined as a UNRS score of 0 or 1 among patients who experienced clinical response and who had a baseline UNRS score of at least 3 points. The UNRS is a patient-reported, validated, 11-point numeric rating scale aimed at measuring the degree of bowel movement urgency.¹⁷ Scores range from 0 (no urgency) to 10 (worst possible urgency). Analyses of patients in the LUCENT-1 study suggested that a within-patient 3-point improvement or decrease in the UNRS score from baseline would be considered clinically important.⁵⁸ Responses were collected using a patient electronic diary daily. Refer to Table 7 for further details on the UNRS.

Bowel Urgency Improvement

Bowel urgency improvement as measured by UNRS was assessed in the LUCENT-1 study at week 12 and the LUCENT-2 study at week 40. This end point was defined as the change in UNRS score from induction baseline to the end of study treatment.

Health-Related Quality of Life

Inflammatory Bowel Disease Questionnaire

The IBDQ⁵⁹ was used to assess disease-specific HRQoL in the LUCENT-1 trial at week 12 and the LUCENT-2 trial at week 40. IBDQ questionnaires were self-administered, with responses recorded electronically using a tablet device at appropriate visits. The IBDQ consists of a 32-item list subdivided into 4 dimensions: systemic symptoms, bowel symptoms, emotional function, and social function. Total scores range from 32 to 224, with a higher score indicating a better HRQoL. The IBDQ has been consistently shown to have good internal consistency and test-retest reliability, as well as showing responsiveness to change in IBD.^18-20 Available studies have suggested that an improvement of 30 points from baseline or an improvement of at least 15 points or greater above placebo may constitute an MID.^21-24

5-Level EQ-5D

The EQ-5D-5L was used as a generic HRQoL measure in the LUCENT-1 study at week 12 and the LUCENT-2 study at week 40. The EQ-5D-5L is a widely used patient-reported measure. Its 5 dimensions include mobility, self-care, usual activities, pain/discomfort, and anxiety/depression, with each dimension including 5 levels ranging from “no problems” to “extreme problems.”⁶⁰ A 1-digit score is assigned for each dimension, which is then combined into a 5-digit score code to represent the total score on the EQ-5D-5L. This 5-digit score code is then converted to a utility from 0 to 1, where 1 represents perfect health and 0 represents death. The second, complementary piece of the EQ-5D-5L is a patient-reported visual analogue scale (the EQ VAS); with the EQ VAS, the patient ranks their overall health on a continuum, converted to a score from 0 to 100, where a higher number suggests better self-reported health.⁶⁰ Though the EQ-5D-5L is a generic scale, it has been validated specifically for the IBD population, which includes the UC population.^25,61,62 The EQ-5D-5L index score (i.e., the utility) was shown to have an MID ranging from 0.050 to 0.076 in patients with IBD.²⁵ The EQ VAS MID was found to be a 10.9-point change from baseline in patients with IBD.²⁵ According to the evidence, the type of IBD (Crohn disease or UC) did not have an impact on the MID threshold.²⁵

Responses were completed at baseline and week 12 (or the end-of-treatment visit) in the LUCENT-1 study and week 40 in the LUCENT-2 study.

Table 7: Summary of Outcome Measures and Their Measurement Properties

CI = confidence interval; EQ VAS = EQ visual analogue scale; EQ-5D-5L = 5-Level EQ-5D; FACIT-F = Functional Assessment of Chronic Illness Therapy–Fatigue; GRS = Guyatt’s responsiveness statistic; HRQoL = health-related quality of life; IBD = inflammatory bowel disease; IBDQ = Inflammatory Bowel Disease Questionnaire; ICC = intraclass correlation; MCS = mental component summary; MID = minimal important difference; OR = odds ratio; OWI = overall work impairment; PCS = physical component summary; PGA = Physician’s Global Assessment; PGR-S = Patient Global Rating of Severity; RB = rectal bleeding; SCCAI = Simple Clinical Colitis Activity Index; SE = standard error; SES = standardized effect size; SF = stool frequency; SF-12 = Short Form (12) Health Survey; SF-36 = Short Form (36) Health Survey; SIBDQ = Short Inflammatory Bowel Disease Questionnaire; UC = ulcerative colitis; UCDAI = Ulcerative Colitis Disease Activity Index; UNRS = Urgency Numeric Rating Scale; VAS = visual analogue scale; WPAI = Work Productivity and Activity Impairment Questionnaire; WPAI:UC = Work Productivity and Activity Impairment Questionnaire: Ulcerative Colitis.

Short Form (36) Health Survey

The SF-36 was used as another generic HRQoL measure in the LUCENT-1 study at week 12 and the LUCENT-2 study at week 40. The SF-36 is a non–disease-specific, self-reported questionnaire.⁶³ The 8 domains evaluated by the SF-36 are physical function, role limitations due to physical health, role limitations due to emotional health, bodily pain, general health, vitality, social functioning, and mental health. The output is 2 separate summary scores — the physical and mental component summaries, measured on a scale from 0 to 100, with a higher score indicative of a better health state.⁶³ Using data from 43 studies, a systematic review indicated that SF-36 has acceptable validity and reliability (for most scales) for UC.⁶⁴ For the PCS and MCS summaries, an MID of 3 points to 5 points was identified.⁶⁴ An absolute score increase of 2 points to 4 points from baseline was determined as the MID to capture improvement in disease activity and symptoms.⁶⁴

Mucosal Healing

Histologic Endoscopic Mucosal Improvement

HEMI at week 12 was assessed in the LUCENT-1 trial only. HEMI was defined using the Geboes scoring system (refer to Table 7 for details) with neutrophil infiltration in fewer than 5% of crypts; no crypt destruction; no erosions, ulcerations, or granulation tissue; and endoscopic remission, defined as ES equals 0 or 1 (excluding friability) on the MMS. Biopsies were obtained at each endoscopy to support assessment of the histopathology end points and scoring was performed by a blinded central reader. The ES was calculated using the same procedures described for the clinical response.

Histologic Endoscopic Mucosal Remission

HEMR with resolution of mucosal neutrophils at week 40 was assessed in the LUCENT-2 study only. HEMR was defined as achieving both endoscopic remission and histologic remission. Endoscopic remission was defined as an ES of 0 or 1 (excluding friability). Histologic remission was characterized as the resolution of mucosal neutrophils, defined using the Geboes scoring system with subscores of 0 for grades:

• 2b (lamina propria neutrophils)

• 3 (neutrophils in the epithelium)

• 4 (crypt destruction)

• 5 (erosion or ulceration).

Work Impairment

Work Productivity and Activity Impairment Questionnaire: Ulcerative Colitis

Work impairment, as evaluated by the WPAI:UC, was assessed in the LUCENT-1 trial at week 12 and the LUCENT-2 trial at week 40. The WPAI:UC is a self-administered, disease-specific scale aimed at measuring the level of work impairment due to UC.²⁶ WPAI:UC considers the 4 domains of absenteeism, presenteeism, overall work performance, and nonwork activities.²⁷ The 4 domains comprise a total of 6 items. The final scores for each domain are a percentage of total impairment, ranging from 0 to 100%, with a higher number indicating greater impairment in that domain. The WPAI:UC is a valid and responsive instrument for use in the assessment of UC.²⁷ An established MID for WPAI:UC was not identified in this patient population.

Harms Outcomes

AEs were predefined and reported in both studies, including SAEs, suspected unexpected serious adverse reactions, AESIs, withdrawal due to AEs, and deaths due to AEs. AESIs for the LUCENT-1 and LUCENT-2 studies were defined as, but not limited to, opportunistic infections, hypersensitivity events, injection site reactions, cerebrocardiovascular events, malignancies, depression and/or suicidal ideation/behaviour, and hepatic AEs.

Safety data were captured during the treatment periods — during a 16-week follow-up after the LUCENT-1 study (for those who did not go on to the LUCENT-2 trial) and during a 16-week follow-up after the LUCENT-2 trial (for those who did not go on to the LUCENT-3 trial). Given that most patients from the LUCENT-1 study went on to the LUCENT-2 study and most patients from the LUCENT-2 study transitioned to the LUCENT-3 trial, few patients completed the 16-week posttreatment follow-up periods. As such, safety data for mirikizumab from the LUCENT-1 and LUCENT-2 trials is mostly limited to 52 weeks of therapy.

Statistical Analysis

The statistical analysis of efficacy end points in the pivotal trials assessed in this Clinical Review Report are summarized in Table 8.

Sample Size and Power Calculation

For the LUCENT-1 study, a sample size of 1,160 patients was deemed appropriate by the sponsor based on the following power calculation assumptions:

• about 50% of randomized patients will be considered conventional-failed and 50% of patients will be considered biologic-failed

• at week 12, the predicted clinical remission rate is expected to be 23% for mirikizumab and 7.8% for placebo

  • for patients who are considered conventional-failed, this is expected to be 30% for mirikizumab versus 12% for placebo

  • for patients who are considered biologic-failed, this is expected to be 16% for mirikizumab versus 3.5% for placebo

• desired power was greater than 90% with a chi-square test with a 2-sided significance level of 0.00125 for the end point of clinical remission at 12 weeks.

For the LUCENT-2 study, a sample size of 470 patients in the main cohort was expected to satisfy the following power calculation assumptions:

• at week 40, clinical remission rates for patients randomized to mirikizumab and placebo would be 47% and 27%, respectively

• desired power was greater than 95% with a chi-square test with a 2-sided significance level of 0.05 for the end point of clinical remission at 40 weeks

• desired power for 2 key secondary end points, endoscopic remission and corticosteroid-free remission, at week 40 was greater than 80% with a 2-sided significance level of 0.05.

In addition, in the LUCENT-2 study’s statistical analysis plan, the following assumptions about the LUCENT-1 study were made:

• the LUCENT-1 study would randomize at approximately 1,160 patients, 90% of whom (1,044 patients) would complete the LUCENT-1 study and enrol in the LUCENT-2 study (i.e., there would be a 10% dropout rate from the LUCENT-1 trial to the LUCENT-2 trial)

• 50% of patients in the LUCENT-1 study would be biologic-experienced

• the LUCENT-1 study would have a clinical response rate of 60% and clinical remission rate of 23% for patients randomized to mirikizumab

• 75% of LUCENT-1 study patients would receive mirikizumab (as per the 3:1 randomization ratio).

Statistical Test or Model

In both the LUCENT-1 trial and the LUCENT-2 trial, all primary analyses of efficacy end points were based on the mITT population, described in Table 9, while some sensitivity analyses may have used other analysis populations, as outlined in Table 8.

For the primary end point in the LUCENT-1 and LUCENT-2 trials, as well as other binary efficacy and health outcome end points, the Cochran-Mantel-Haenszel test was used. For continuous efficacy and health outcome end points, comparisons were made using mixed-effects model of repeated measures analysis. An analysis of covariance was used for continuous efficacy and health outcome end points with a single postbaseline time point. For each end point, adjustment factors including specific covariates as well as sensitivity analyses are listed in Table 8.

Sensitivity Analyses

In the primary and sensitivity analyses, various techniques were used to handle dropouts or missing data. Nonresponder imputation was used for missing data for analyses of binary efficacy and health outcome variables, whereby patients with missing clinical efficacy data at a time point of interest were deemed to be nonresponders. Some sensitivity analyses in Table 8 discuss an “alternative estimand.” This is a conservative estimate that considers each of the following to be a treatment failure: failing to meet protocol-defined primary end point criteria, missing data to calculate primary end point criteria, initiating the use of systemic corticosteroids, increasing systemic corticosteroid dose beyond baseline, switching between corticosteroid therapies, or having a UC surgery.

End points with continuous variables that used the mixed model of repeated measures to make comparisons assumed that missing data were missing at random. For patients who discontinued the study drug, the modified baseline observation carried forward approach may have been used. With modified baseline observation carried forward, those who discontinue due to AEs had their baseline observation carried forward for all missing observations after discontinuation. For discontinuation not due to AEs, the last nonmissing, postbaseline observation before discontinuation was carried forward. Some sensitivity analyses applied a modified nonresponder imputation (mNRI) method for dealing with missing data, which included the imputation of data for patients who discontinued treatment due to COVID-19 reasons, were lost to follow-up, or had a protocol deviation. Patients with sporadically missing data or who had had an endoscopy outside of day 71 to day 113 for the LUCENT-1 study or day 267 to day 337 for the LUCENT-2 study had data imputed. With mNRI, patients who discontinued for other reasons, including AEs, were considered nonresponders.

A final method of dealing with missing data was tipping point analysis, conducted as a sensitivity analysis on primary end points in the LUCENT-1 and LUCENT-2 studies. The tipping point analysis assumed the most extreme conservative case, imputing the worst possible outcomes for missing data in patients randomized to mirikizumab, but the best possible outcomes for patients with missing data randomized to placebo. Table 8 highlights where each of these missing data techniques was used.

Multiplicity Adjustments

To control for multiplicity for the primary and key secondary end points, a prespecified graphical multiple testing approach was used, controlling the type I error rate at a 2-sided alpha of 0.00125 for the LUCENT-1 trial and 0.05 for the LUCENT-2 trial. Figure 2 and Figure 3 show the graphical testing approach to control type I error rate in the LUCENT-1 study and the LUCENT-2 study, respectively.

Subgroup Analyses

Subgroup analyses in both the LUCENT-1 study and the LUCENT-2 study were prespecified in the statistical analysis plans and were conducted for primary and key secondary efficacy end points. No analysis for the comparability of baseline covariates across subgroups was performed. Subgroup analyses did not control for multiplicity, except for those specified as key secondary end points (i.e., clinical response in the biologic-failed population at week 12 in the LUCENT-1 trial). In the LUCENT-2 trial, subgroup analyses were conducted for the same factors as the LUCENT-1 trial as well as by country, number of failed biologics or tofacitinib (≤ 2 or > 2), and prior failure of conventional therapies but not biologics or tofacitinib (yes or no).

The following subgroups, planned a priori in the statistical analyses plans, aligned with the subgroups deemed relevant for this CADTH review: patients treated previously with conventional therapy (i.e., steroids or immunomodulators), patients treated previously with advanced UC therapy (i.e., biologics, TNF, or JAK inhibitors), baseline MMS (moderate versus severe), and disease extension (extensive colitis versus nonextensive colitis). Only the subgroups identified as relevant are reported here.

Figure 2: Graphical Testing Approach to Control Type I Error Rate in LUCENT-1 Study

Source: Clinical Study Report for the LUCENT-1 study.³¹

Figure 3: Graphical Testing Approach to Control Type I Error Rate in LUCENT-2 Study

AMAN = LUCENT-1 study.

Source: Clinical Study Report for the LUCENT-2 study.³²

Table 8: Statistical Analysis of Efficacy End Points

ANCOVA = analysis of covariance; CMH = Cochran-Mantel-Haenszel; EQ-5D-5L = 5-Level EQ-5D; HEMI = histologic endoscopic mucosal improvement; HEMR = histologic endoscopic mucosal remission; HRQoL = health-related quality of life; IBDQ = Inflammatory Bowel Disease Questionnaire; ITT = intention-to-treat; MAR = missing at random; mBOCF = modified baseline observation carried forward; MCS = mental component summary; mITT = modified intention-to-treat; MMRM = mixed model of repeated measures; MMS = Modified Mayo Score; mNRI = modified nonresponder imputation; NR = not reported; NRI = nonresponder imputation; PCS = physical component summary; PP = per-protocol; SF-36 = Short Form (36) Health Survey; UNRS = Urgency Numeric Rating Scale; vs. = versus; WPAI:UC = Work Productivity and Activity Impairment Questionnaire: Ulcerative Colitis.

Note: Details in Table 8 have been taken from the sponsor’s Summary of Clinical Evidence.^20,30

Source Clinical Study Reports for the LUCENT-1 and LUCENT-2 studies.^31,32

Analysis Populations

Analysis populations of the LUCENT-1 and LUCENT-2 studies are summarized in Table 9.

Note that the safety populations in the LUCENT-1 and LUCENT-2 studies refer to those who were randomized and received at least 1 dose of any study treatment. For the LUCENT-1 trial, this includes all patients randomized to mirikizumab IV or placebo, while in the LUCENT-2 trial, it refers to main cohort patients randomized to mirikizumab SC or placebo.

Table 9: Analysis Populations

GCP = good clinical practice; ITT = intention-to-treat; mITT = modified intention-to-treat; PP = per-protocol.

Note: Details in Table 9 have been taken from the sponsor’s Summary of Clinical Evidence.^20,30

Sources: Clinical Study Reports for the LUCENT-1 and LUCENT-2 studies.^31,32

Results

Patient Disposition

A summary of patient disposition in the LUCENT-1 study and the LUCENT-2 study is in Table 10. In addition, the patient disposition of the mirikizumab group in the induction nonresponders (extended induction cohort) from the LUCENT-2 study is presented in Table 11.

LUCENT-1: Induction Trial

Of the 2,079 patients screened in the LUCENT-1 trial, 730 (35%) patients were screen failures and 33 (1.6%) patients were patient withdrawals. In the LUCENT-1 trial, 959 patients were randomized to the mirikizumab arm and 322 patients were randomized to the placebo arm. Study treatment discontinuation occurred in 39 (4.1%) patients in the mirikizumab arm and 37 (11.5%) patients in the placebo arm. The most common reason for treatment discontinuation was AEs in both the mirikizumab arm (1.6%) and the placebo arm (7.1%).

LUCENT-2: Maintenance Trial

From the LUCENT-1 trial, 1,178 patients entered the LUCENT-2 trial. Of the responders in the mirikizumab and placebo groups from the LUCENT-1 study, 581 patients were randomized to the primary efficacy analysis cohort in the LUCENT-2 study. Of these patients, 389 patients were randomized to the mirikizumab arm and 192 patients were randomized to the placebo arm in the LUCENT-2 trial. Study treatment discontinuation occurred in 42 (10.8%) patients in the mirikizumab arm and 73 (38.0%) patients in the placebo arm. The most common reason for treatment discontinuation was loss of response (i.e., the patient was switched to open-label rescue IV mirikizumab) in both the mirikizumab arm (4.9%) and the placebo arm (21.9%). Of the patients receiving rescue therapy, 4 (1.0%) patients in the mirikizumab arm and 4 (2.1%) patients in the placebo arm discontinued rescue mirikizumab.

Of the mirikizumab nonresponders from the LUCENT-1 study in the ITT population, 313 patients (272 patients according to the mITT population) entered the open-label extended induction period; of these, 171 (54.4%) patients (146 [53.7%] patients in the mITT population) responded and entered the maintenance period with open-label 200 mg mirikizumab SC. Of the placebo nonresponders, ||| |||||||| entered the open-label extended induction period; of these patients ||| ||||||| responded and entered the maintenance period with open-label 200 mg mirikizumab SC.

In the LUCENT-2 trial, 135 patients from the placebo induction responders in the LUCENT-1 study continued placebo SC in the extended induction period. Of these, 90 (66.7%) patients completed blinded placebo SC maintenance.

Table 10: Summary of Patient Disposition From Pivotal Studies Submitted by the Sponsor

ITT = intention-to-treat; mITT = modified intention-to-treat; NA = not applicable; NR = not reported; PP = per-protocol.

Sources: Clinical Study Reports for the LUCENT-1 and LUCENT-2 studies.^31,32

Table 11: Summary of Patient Disposition of Induction Nonresponders and Placebo Induction Responders From LUCENT-2 Study

NA = not applicable; SC = subcutaneous.

Notes: Details in Table 11 have been taken from the sponsor’s Summary of Clinical Evidence.^20,30

No comparisons were drawn between the mirikizumab induction nonresponders and placebo induction nonresponders.

Source: Clinical Study Report for the LUCENT-2 study.³²

Baseline Characteristics

A summary of baseline patient demographics, disease characteristics, and medication history in the pivotal trials is shown in Table 12. The baseline characteristics outlined in Table 12 are limited to those that are most relevant to this review or were felt to impact the outcomes or interpretation of the study results.

LUCENT-1: Induction Trial

The study population in the LUCENT-1 trial had a mean age of 42.5 (SD = 13.92) years. The majority of patients were male (59.8%) and white (72.3%). Based on the MMS, patients’ UC severity was either categorized as moderate (46.5% of mirikizumab patients and 47.1% of placebo patients) or severe (53.3% of mirikizumab patients and 52.9% of placebo patients). In terms of prior UC therapies, the proportion of patients reporting prior biologic or tofacitinib failure was similar between treatment groups (41.6% and 40.1% of patients randomized to mirikizumab and placebo, respectively). Loss of response to a biologic or tofacitinib was present in 22.6% of mirikizumab patients and 22.1% of placebo patients. Baseline corticosteroid use was present in 40.4% of patients in the mirikizumab arm and 38.4% of patients in the placebo arm. Overall, the baseline characteristics were well-balanced between treatment arms.

LUCENT-2: Maintenance Trial

The study population in the LUCENT-2 study main cohort had a mean age of 42.3 (SD = 13.5) years. The majority of patients were male (58.4%) and white (71.3%). Based on the MMS, approximately half of the patients in each treatment arm was categorized as moderate UC severity (MMS score of 4 points to 6 points). In terms of prior UC therapies, 35.1% of patients in the mirikizumab group and 35.8% of patients in the placebo group had a history of biologic or tofacitinib failure. Loss of response to a biologic or tofacitinib was reported in 20% of mirikizumab patients and 13.7% of placebo patients. Baseline corticosteroid use was present in 37% of patients in the mirikizumab arm and 38% of patients in the placebo arm. Overall, the baseline characteristics were well-balanced between treatment arms.

Table 12: Summary of Baseline Characteristics of Pivotal Studies and RCT Evidence Submitted by the Sponsor (mITT Populations)

5-ASA = 5-aminosalicylic acid; BMI = body mass index; mITT = modified intention-to-treat; q.4.w. = every 4 weeks; Q1 = first quartile; Q3 = third quartile; RCT = randomized controlled trial; SC = subcutaneous; SD = standard deviation; TNF = tumour necrosis factor; UC = ulcerative colitis; UNRS = Urgency Numeric Rate Score.

Note: Details in Table 12 have been taken from the sponsor’s Summary of Clinical Evidence.^20,30

Sources: Clinical Study Reports for the LUCENT-1 and LUCENT-2 studies.^31,32

Exposure to Study Treatments

Treatment exposure from the pivotal trials is summarized in Table 13. Study treatment durations occurred at 4-week intervals. In the LUCENT-1 study’s mITT population, 98.6% of patients randomized to placebo and 98.0% of patients randomized to mirikizumab were deemed adherent, defined as missing no study doses. || |||||||||| |||| |||||| |||| ||||||||||| ||||||| ||||||||| ||| ||||| || || ||||| ||| |||||||| |||||||||| || ||||||||||| ||| ||||| ||| ||||| |||||||||| || |||||||| ||||||| || ||||||||| ||| ||||||||| |||| |||| || ||||| |||||||||| ||| || ||||| ||| || ||||||| In the LUCENT-2 study, a greater proportion of patients in the mirikizumab group (90.5%) compared to the placebo group (64.6%) had at least 32 weeks of treatment exposure. Fewer patients had a treatment exposure of at least 40 weeks, with a greater proportion in the mirikizumab group (67.6%) compared to the placebo group (47.4%).

Table 13: Summary of Patient Exposure From Pivotal Studies and RCT Evidence Submitted by the Sponsor (Safety Populations)

Q.4.w. = every 4 weeks; RCT = randomized controlled trial; SC = subcutaneous.

Note: Details in Table 13 have been taken from the sponsor’s Summary of Clinical Evidence.^20,30

Sources: Clinical Study Reports for the LUCENT-1 and LUCENT-2 studies.^31,32

Concomitant Medications and Cointerventions

Concomitant medication use in the pivotal trials is summarized in Table 14 ||||||||||| |||||||||| ||| ||| ||||||||| ||||||| |||||| ||||||||| |||| ||| ||||||| ||||||.

Table 14: Redacted

||| |||||||||||| |||| ||||||||||| ||| ||||||||||||| |||||||| |||||||| ||||| ||||||| ||| |||||||| ||| ||||||||^31,32 |||||||||| |||| ||| ||||| |||| |||| ||||| |||| ||| ||||||||| ||||||| || |||||||| |||||||||²⁰

Protocol Deviations

Protocol deviations in the pivotal trials for the mITT population are summarized in Table 15.

|| ||||||||| ||| |||||||||| || |||||||| |||| || |||||||||||||| |||||||| ||||||||| ||| ||| ||||||| || ||| ||||||||||| ||| ||| || ||||||| || ||| ||||||| |||| ||| |||| |||||||| |||||||| |||||||||| || ||| ||||||| ||||| |||| |||||||||||||||||||| |||| ||| |||||||| ||||||||||| |||||||| ||||||||| ||||| ||| ||||||||| |||||| |||||| |||||||| ||| || |||| |||| || |||||| |||||||| |||||| |||||| ||| ||||||||| |||||||||||||| || |||||||| |||||| |||||| || ||| ||||||||||||||||||||| |||||| ||| |||| |||||||| |||||||| |||||||||| |||| |||||||| ||| || |||| |||| || |||||| |||||||| |||||| |||||| |||||||||||||||||||| |||| ||| |||||||| |||||||||| ||||||||| |||||| |||||||||| ||||||||||| ||| ||||||||| |||||||||||||| || |||||||| |||||| ||||||||| ||||||||| ||||| ||||||||||| ||||||||| ||||||||||| ||| ||||||||||| || |||||||| |||| || |||||||||||||| |||||||| ||||||||| ||||| ||||| |||||| ||| ||| ||||||||||| ||| ||| ||||| |||||| ||| ||| ||||||| |||| ||| |||||||| || ||| ||||||| |||||| ||| |||| |||||||| |||||||| ||||||||| |||||||||| ||||| |||||||||||||||||||| ||| |||||||| ||| ||||||| ||| ||||||||| ||| ||||| || || || |||||||||||||||| |||||| || ||||| ||| |||||| |||||| |||||||| ||||||||| ||||| ||| ||||||||| || || || |||||||||||||||| ||||||||| || |||||||| |||||| |||||| ||| |||||| || |||||||| || ||| || || ||||| |||| |||||||| ||| ||| ||||||||||| || ||| |||||| |||||||||| ||||||| ||||| |||| |||| || ||| |||||||| |||||||||| |||||||| |||||||| |||||||| ||| ||||| ||||||| || |||||||||||||| || ||| ||||| ||||||| || |||||||||||| ||

Table 15: Redacted

|||||||||||| ||||||||||||||||||||||| |||||||| ||||| ||||||| ||| |||||||| ||| |||||||||^31,32

Transcription Error in the Electronic Clinical Outcome Assessment Devices in Poland and Turkey Impacting Outcomes Assessment

In the LUCENT-2 trial, an error was identified in the way that the electronic clinical outcomes assessment devices worded daily self-reported SF and RB assessments for patients in Turkey and Poland, respectively. The error was corrected when noticed. Overall, 104 impacted patients (4 patients from Turkey and 100 patients from Poland), whose LUCENT-1 study baseline data were collected using incorrectly transcribed patient-reported outcome questions on the electronic clinical outcomes assessment devices, were excluded from the primary efficacy analysis. Impacted patients were, however, still included in the primary safety analysis. Sensitivity analyses for efficacy were performed on primary and selected key secondary end points at week 40 by including impacted patients using the ITT population with the mNRI method, with RB score imputed for impacted patients in Poland and SF scores imputed for impacted patients in Turkey.

Efficacy

Summarized end points are based on those included in the sponsor’s Summary of Clinical Evidence³⁰ as well as any end points identified as important to this review according to stakeholders — for example, the clinical expert, clinician groups, or patient groups. These end points are summarized in Table 16 for the LUCENT-1 trial and Table 17 for the LUCENT-2 trial. Results are presented by induction period (the LUCENT-1 study) and maintenance period (the LUCENT-2 study).

LUCENT-1 Study: Induction Period

Clinical Response

A greater proportion of patients on mirikizumab 300 mg IV (63.5%) than on placebo (42.2%) experienced clinical response at week 12 (common risk difference = 21.4%; 99.875% CI, 10.8% to 32.0%; P < 0.00001). Sensitivity analyses in the ITT population were consistent with the mITT population results.

In the biologic-naive subgroup, a greater proportion of patients on mirikizumab 300 mg IV (70.1%) versus placebo (50.3%) experienced clinical response (common risk difference = 19.8%; 95% CI, 11.3% to 28.3%; P < 0.001). Similarly, patients in the biologic-experienced subgroup also demonstrated greater clinical response on mirikizumab (54.6%) versus placebo (30.9%), with a common risk difference of 23.9% (95% CI, 14.3% to 33.5%; P < 0.001).

In terms of baseline conventional therapy use, of the 464 patients receiving corticosteroids and 280 patients receiving immunomodulators at baseline in the mITT population, the observed effect sizes for the mirikizumab group compared to the placebo group were numerically positive but small when compared to the overall population. The subgroup of patients without baseline corticosteroid and immunomodulator use had similar results to patients in the overall mITT population. The interaction P values did not suggest a difference between the groups. The LUCENT-1 study baseline therapy subgroup data are in Appendix 1.

The subgroup analysis for baseline MMS score was prespecified in the LUCENT-1 study; however, the analysis was not available from the sponsor.

Clinical Remission

Clinical remission was assessed using 2 different outcomes in the LUCENT-1 trial: clinical remission and alternate clinical remission. A summary of key remission outcomes is presented in Table 16.

Clinical Remission Rate

A greater proportion of patients on mirikizumab 300 mg IV (24.2%) versus placebo (13.3%) experienced clinical remission at week 12, with a common risk difference of 11.1% (99.875% CI, 3.2% to 19.1%; P = 0.00006). Analyses in the PP and ITT populations were similar to the mITT population results. In addition, results of the sensitivity analyses using different imputation methods were consistent with the results from the primary analysis. In terms of the tipping point analysis, there was no significant difference between groups when imputing missing data as “responder” for the placebo group and as “nonresponder” for the mirikizumab group.

In terms of the biologic-naive subgroup, a greater number of patients in the mirikizumab group (30.9%) versus the placebo group (15.8%) were clinical remitters at the end of the LUCENT-1 trial (common risk difference = 15.1%; 95% CI, 8.3% to 21.9%; P < 0.001). The magnitude of the effect was consistent with the primary analysis. Similar results were seen in the biologic-experienced subgroup between patients receiving mirikizumab (15.4%) versus placebo (9.8%), although the effect was small (common risk difference = 5.7%; 95% CI, –0.7% to 12.1%; P = 0.135).

In the subgroups of patients using corticosteroids or immunomodulators at baseline, the observed effect sizes for clinical remission in the mirikizumab group compared to the placebo group were numerically positive but small when compared to the overall population. The subgroup of patients without baseline corticosteroid and immunomodulator use had similar effect sizes to patients in the overall mITT population. The interaction P values did not suggest a difference between the groups.

Appendix 1 includes the LUCENT-1 study baseline therapy subgroup data. The subgroup analysis for baseline MMS score was prespecified in the LUCENT-1 trial; however, data were not available in the submission provided by the sponsor.

Alternate Clinical Remission Rate

Alternate clinical remission used a definition similar to that of clinical remission except it did not require a 1-point decrease in the SF subscore. Results for alternate clinical remission were very similar to those of clinical remission. At week 12 of the LUCENT-1 study, 25.6% of patients on mirikizumab versus 14.6% of patients on placebo experienced alternate clinical remission, with a common risk difference of 11.1% in favour of mirikizumab (99.875% CI, 3.0% to 19.3%; P < 0.001). Analyses in the ITT population were consistent with the mITT population results.

In the biologic-naive subgroup, a greater number of patients in the mirikizumab group (32.5%) versus the placebo group (18.1%) experienced alternate clinical remission at the end of the LUCENT-1 study (common risk difference = 14.4%; 95% CI, 7.3% to 21.5%; P < 0.001). Results were similar in the biologic-experienced subgroup between patients receiving mirikizumab (15.4%) versus placebo (9.8%), although the effect was small (common risk difference = 6.7%; 95% CI, 0.3% to 13.2%; P = 0.079).

|| ||| ||||||||| || |||||||| ||||| ||||||||||||||| || |||||||||||||||| || ||||||||| ||| |||||||| |||||| ||||| ||| ||||||||| |||||||| ||||||||| || ||| ||||||||||| ||||| |||||||| || ||| ||||||| ||||| |||| ||||||||||| |||||||| ||| ||||| |||| |||||||| || ||| ||||||| ||||||||||| ||| |||||||| || |||||||| ||||||| |||||||| |||||||||||||| ||| ||||||||||||||| ||| ||| ||||||| |||||| ||||| || |||||||| || ||| ||||||| |||| ||||||||||| ||| ||||||||||| |||||||| ||| ||| ||||||||||||||||| ||||||| ||| ||||||| Appendix 1 |||||||| ||| |||||||| |||||||| ||||||| |||||||| ||||.

The subgroup analysis for baseline MMS score was prespecified in the LUCENT-1 study; however, data were not available from the sponsor.

Endoscopic Remission

At week 12 of the LUCENT-1 trial, 36.3% of patients on mirikizumab experienced endoscopic remission versus 21.1% of patients on placebo (common risk difference = 15.4%; 99.875% CI, 6.3% to 24.5%; P value < 0.00001). Analyses in the ITT population were consistent with the mITT population results.

In the biologic-naive subgroup, a greater proportion of patients on mirikizumab (45.9%) versus placebo (28.1%) experienced endoscopic remission at week 12 (common risk difference = 17.9%; 95% CI, 9.8% to 25.9%; P < 0.001). Similarly, within the biologic-experienced subgroup, a greater proportion of patients on mirikizumab (23.7%) versus placebo (11.4%) experienced endoscopic remission at week 12 (common risk difference = 12.3%; 95% CI, 5.2% to 19.4%; P = 0.003).

In the subgroups of patients using corticosteroids or immunomodulators at baseline, the observed effect sizes for endoscopic remission in the mirikizumab group compared to the placebo group were numerically positive but small when compared to the overall population. The subgroup of patients without baseline corticosteroid and immunomodulator use had similar effect sizes to patients in the overall mITT population. The interaction P values did not suggest a difference between the groups.

Appendix 1 includes the LUCENT-1 study baseline therapy subgroup data. The subgroup analysis for baseline MMS score was prespecified in the LUCENT-1 trial; however, data were not available in the submission provided by the sponsor.

|||||||| |||| || || ||||||||| ||||| || |||||||| || ||||||||||| ||| || ||||| ||||| || || ||||||||||| ||| || |||||||| || |||||||| ||||||||| |||| |||||||||| || ||||| |||||||| ||| ||| || ||||| |||||||||

Symptomatic Remission

At week 12 of the LUCENT-1 study, 45.5% of patients on mirikizumab and 27.9% of patients on placebo experienced symptomatic remission, with a common risk difference of 17.5% (99.875% CI, 7.5% to 27.6%; P < 0.00001). Analyses in the ITT population were consistent with the mITT population results.

In the biologic-naive subgroup, a greater proportion of patients on mirikizumab (50.4%) versus placebo (33.3%) experienced symptomatic remission at week 12 (common risk difference = 17.1%; 95% CI, 8.7% to 25.4%; P < 0.001). Similarly, within the biologic-experienced subgroup, a greater proportion of patients on mirikizumab (38.6%) versus placebo (18.8%) experienced endoscopic remission at week 12 (common risk difference = 18.8%; 95% CI, 10.1% to 27.4%; P < 0.001).

|| ||| ||||||||| || |||||||| ||||| ||||||||||||||| || |||||||||||||||| || ||||||||| ||| |||||||| |||||| ||||| ||| ||||||||||| ||||||||| || ||| ||||||||||| ||||| |||||||| || ||| ||||||| ||||| |||| ||||||||||| |||||||| ||| ||||| |||| |||||||| || ||| ||||||| ||||||||||| ||| |||||||| || |||||||| ||||||| |||||||| |||||||||||||| ||| ||||||||||||||| ||| ||| ||||||| |||||| ||||| || |||||||| || ||| ||||||| |||| ||||||||||| ||| ||||||||||| |||||||| ||| ||| ||||||||||||||||| ||||||| ||| |||||||

Appendix 1 includes the LUCENT-1 study baseline therapy subgroup data. The subgroup analysis for baseline MMS score was prespecified in the LUCENT-1 study; however, data were not available from the sponsor.

Bowel Urgency Improvement: UNRS

The LSM change from baseline at week 12 in the mirikizumab group was –2.59 points and –1.63 points in the placebo group, a difference of –0.95 points (99.875% CI, –1.5 points to –0.4 points; P < 0.00001). Analyses in the ITT population were consistent with the mITT population results.

In the biologic-naive subgroup, greater improvement in UNRS was seen in the mirikizumab group (–2.7 points) versus the placebo group (–2.1 points), with an LSM mean difference of –0.6 points at week 12 (95% CI, –1.0 points to –0.2 points; P = 0.002). Similar results were seen in the biologic-experienced subgroup, but with a larger LSM mean difference of –1.5 points at week 12 (95% CI, –2.0 points to –1.0 points; P < 0.001).

|| ||| ||||||||| || |||||||| ||||| ||||||||||||||| || |||||||||||||||| || ||||||||| ||| |||||||| |||||| ||||| ||| ||||| ||||||| ||||||||||| || ||| ||||||||||| ||||| |||||||| || ||| ||||||| ||||| |||| ||||||||||| |||||||| ||| ||||| |||| |||||||| || ||| ||||||| ||||||||||| |||||||| ||||||| |||||||| |||||||||||||| ||| ||||||||||||||| ||| ||| ||||||| ||||||| || |||||||| || ||| ||||||| |||| ||||||||||| ||| ||||||||||| |||||||| ||| ||| ||||||||||||||| ||||||| ||| |||||||

Appendix 1 includes the LUCENT-1 study baseline therapy subgroup data. The subgroup analysis for baseline MMS score was prespecified in the LUCENT-1 study; however, data were not available from the sponsor.

Health-Related Quality of Life

HRQoL was assessed in the LUCENT-1 trial based on the IBDQ score, EQ-5D-5L score, and SF-36 score (PCS and MCS). A summary of HRQoL outcomes is presented in Table 16.

IBDQ Score

In the LUCENT-1 study, the mean change from baseline to week 12 in the IBDQ score was 38.4 points for patients in the mirikizumab group and 25.2 points for those in the placebo group, representing a difference of 13.2 points (95% CI, 9.3 points to 17.2 points; P < 0.001) in favour of mirikizumab. The change from baseline in the mirikizumab group met the MID of at least 30 points, representing a meaningful improvement from baseline to week 12. However, the difference between groups was not considered to be clinically important as it did not meet the suggested MID of at least 15 points over placebo.^21-24

EQ-5D-5L Score

In the LUCENT-1 study, the mean change from baseline to week 12 in the EQ VAS score was 14.6 points in the mirikizumab group and 9.4 points in the placebo group, with an LSM difference of 5.2 points (95% CI, ||| || |||; P < 0.001). The change from baseline in the mirikizumab group met the MID of 10.9 points on the EQ VAS;²⁵ however, it was not clear whether the difference between groups was clinically important given that no MID was identified for between-group differences.

SF-36 Score

In the LUCENT-1 study, the mean change from baseline to week 12 in the SF-36 PCS score was 5.97 points in the mirikizumab group and 3.90 points in the placebo group, with an LSM difference of 2.07 points (95% CI, 1.21 points to 2.93 points; P < 0.001). The mean change from baseline to week 12 in the SF-36 MCS score was 5.02 points in the mirikizumab group and 3.42 points in the placebo group, with an LSM mean difference of 1.60 points (95% CI, 0.56 points to 2.63 points; P = 0.002). The change from baseline in the mirikizumab group appeared to meet the identified MID threshold (i.e., MID = 3 points to 5 points) for the PCS and MCS. However, it was not clear whether the difference between groups was clinically important given that no MID was identified for between-group differences.

Mucosal Healing: HEMI

In the LUCENT-1 study, mucosal healing was assessed based on the HEMI outcome, which considers both histologic and endoscopic outcomes. At week 12, 27.1% of patients on mirikizumab experienced HEMI versus 13.9% of patients on placebo, with a common risk difference of 13.4% (99.875% CI, 5.5% to 21.4%; P < 0.00001). Analyses in the ITT population were consistent with the mITT population results.

Within the biologic-naive subgroup, the percentage of patients who experienced HEMI was higher in the mirikizumab treatment group (35.8%) compared to the placebo group (18.7%), with a common risk difference of 17.1% (95% CI, 9.8% to 24.3%; P < 0.001). The magnitude of the difference was consistent with the mITT population. Within the biologic-experienced subgroup, the percentage of patients who experienced HEMI was greater in the mirikizumab treatment group (15.7%) compared to the placebo group (7.3%), with a common risk difference of 8.4% (95% CI, 2.5% to 14.3%; P = 0.022).

|| ||| ||||||||| || |||||||| ||||| ||||||||||||||| || |||||||||||||||| || ||||||||| ||| |||||||| |||||| ||||| ||| |||| || ||| ||||||||||| ||||| |||||||| || ||| ||||||| ||||| |||| ||||||||||| |||||||| ||| ||||| |||| |||||||| || ||| ||||||| ||||||||||| |||||||| ||||||| |||||||| |||||||||||||| ||| ||||||||||||||| ||| ||| ||||||| ||||||| || |||||||| || ||| ||||||| |||| ||||||||||| ||| ||||||||||| |||||||| ||| ||| ||||||||||||||||| ||||||| ||| |||||||

Appendix 1 includes the LUCENT-1 trial subgroup data. The subgroup analysis for baseline MMS score was prespecified in the LUCENT-1 study; however, data were not available from the sponsor.

Work Productivity

In the LUCENT-1 study, work productivity was assessed using the WPAI:UC score at week 12. Among those employed at baseline (n = 566), patients on mirikizumab experienced a mean change in WPAI:UC of –20.65, compared to –14.91 in the placebo group (LSM difference = –5.74 points; 95% CI, –10.06 points to –1.42 points; P = 0.009). An MID was not identified for the WPAI:UC in patients for UC; hence, the clinical importance of the results was uncertain.

Table 16: Summary of Key Efficacy Results From LUCENT-1 Study (mITT Population at Week 12)

CI = confidence interval; EQ-5D-5L = 5-Level EQ-5D; HEMI = histologic endoscopic mucosal improvement; IBDQ = Inflammatory Bowel Disease Questionnaire; LSM = least squares mean; MCS = mental component summary; mITT = modified intention-to-treat; PCS = physical component summary; q.4.w. = every 4 weeks; SE = standard error; SF-36 = Short Form (36) Health Survey; UCEIS = Ulcerative Colitis Endoscopic Index of Severity; UNRS = Urgency Numeric Rating Scale; WPAI:UC = Work Productivity and Activity Impairment Questionnaire: Ulcerative Colitis.

Notes: Details in Table 16 have been taken from the sponsor’s Summary of Clinical Evidence.^20,30

The 2-sided alpha was 0.00125 for the LUCENT-1 trial.

^aP value has not been adjusted for multiple testing.

Source: Clinical Study Report for the LUCENT-1 study.³¹

LUCENT-2 Study: Maintenance Period

Clinical Remission

Clinical remission was assessed using 4 different outcomes in the LUCENT-2 study: clinical remission, alternate clinical remission, corticosteroid-free remission, and durable clinical remission. A summary of key remission outcomes is presented in Table 17.

Clinical Remission

A greater proportion of patients on mirikizumab 200 mg SC (49.9%) versus placebo (25.1%) experienced clinical remission after 40 weeks of maintenance therapy (common risk difference = 23.2%; 95% CI, 15.2% to 31.2%; P < 0.001). Analyses in the PP and ITT populations were similar to the mITT population results. In addition, results of the sensitivity analyses were consistent with the results from the primary analysis.

In the tipping point analysis, all missing data for the placebo group was imputed as a responder and all missing data for the mirikizumab group were imputed as a nonresponder. When considering this extreme case, there was no difference between the mirikizumab group and placebo the group.

Clinical remission was analyzed by subgroups. In the biologic-naive subgroup, 51.5% of patients on mirikizumab and 30.7% of patients on placebo experienced clinical remission at week 40 of the LUCENT-2 trial, corresponding to a common risk difference of 20.8% (95% CI, 10.2% to 31.5%; P < 0.001). In the biologic-failed subgroup, 46.1% of patients on mirikizumab and 15.6% of patients on placebo experienced clinical remission with a common risk difference of 30.5% (95% CI,18.1% to 42.9%; P < 0.001).

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (no), and patients with severe UC at baseline were consistent with the results of the primary analysis. Subgroup results for immunomodulator use (yes) and patients with moderate UC at baseline were numerically positive but small when compared to the overall population. ||| ||||||||||| |||||||| ||| ||| ||||||||||||||||| ||||||| ||| ||||||. Appendix 1 contains the LUCENT-2 study subgroup data.

Alternate Clinical Remission

Alternate clinical remission used a definition similar to that of clinical remission except it did not require a 1-point decrease in the SF subscore. Results for alternate clinical remission were very similar to those of clinical remission. At week 40 of the LUCENT-2 study, 51.8% of mirikizumab patients and 26.3% of placebo patients experienced alternate clinical remission with a common risk difference of 24.1% (95% CI, 16.0% to 32.2%; P < 0.001). |||||||| || ||| ||| |||||||||| |||| |||||||||| |||| ||| |||| |||||||||| |||||||.

Alternate clinical remission was analyzed by subgroups. In the biologic-naive subgroup, 54.1% of patients on mirikizumab and 32.5% of patients on placebo experienced alternate clinical remission, corresponding to a common risk difference of 21.7% (95% CI, 10.9% to 32.4%; P < 0.001). ||||| ||| ||||||||||||||| ||||||||| ||||| || |||||||| || ||||||||||| ||| ||||| || |||||||| || ||||||| ||||||||||| ||||||||| |||||||| ||||||||| |||||||||| |||| |||||||||| || ||||| || |||||| || ||||||||||| |||| ||| |||| || ||||| ||||||||| ||| ||||||||| || |||||| ||| |||||||||| |||| ||| ||||||| |||||||||

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (yes and no), and patients with moderate and severe UC at baseline were |||||||||| |||| ||| ||||||| || ||| ||||||| ||||||||. Appendix 1 contains the LUCENT-2 study subgroup data.

Durable Clinical Remission

Of patients who experienced clinical remission at week 12 of the LUCENT-1 study (n = 210), 63.6% of patients who were randomized to mirikizumab 200 mg SC were still in clinical remission at week 40 of the LUCENT-2 trial compared to 36.9% of those randomized to placebo SC, with a common risk difference of 24.8% (95% CI, 10.4% to 39.2%; P < 0.001). Analyses in the ITT population were consistent with the mITT population results.

In the biologic-naive subgroup, 62.5% of patients on mirikizumab and 46.8% of patients on placebo experienced durable clinical remission at week 40 of the LUCENT-2 study, with a common risk difference of 15.7% (95% CI, –1.3% to 32.7%; P = 0.078), although the effect size was small compared to the overall population. In the biologic-failed subgroup, 66.7% of patients in the mirikizumab group and 11.1% of patients in the placebo group experienced durable clinical remission at week 40, with a common risk difference of 55.6% (95% CI, 34.4% to 76.7%; P < 0.001). However, sample sizes for this subgroup were quite small.

Subgroup results for baseline corticosteroid use (no), baseline immunomodulator use (no), and patients with moderate and severe UC at baseline were consistent with the results of the primary analysis. Subgroup results of patients with baseline corticosteroid use (yes) and immunomodulator use (yes), were numerically positive but small when compared to the overall population. ||| ||||||||||| |||||||| ||| ||| ||||||||||||||||| ||||||| ||| ||||||. Appendix 1 contains the LUCENT-2 study subgroup data.

Corticosteroid-Free Remission

A greater number of patients randomized to mirikizumab experienced corticosteroid-free remission than those randomized to placebo at week 40 — 44.9% versus 21.8%, respectively (common risk difference = 21.3%; 95% CI, 13.5% to 29.1%; P < 0.001). This between-group difference was consistent in both the biologic-naive and biologic-failed subgroups (P < 0.001 for both).

In the biologic-naive subgroup, 46.7% of patients on mirikizumab and 26.3% of patients on placebo experienced corticosteroid-free remission at week 40 with a common risk difference of 20.4% (95% CI, 10.1% to 30.8%; P < 0.001). In the biologic-failed subgroup, 40.6% of patients on mirikizumab and 14.1% of patients on placebo experienced corticosteroid-free remission at week 40 with a common risk difference of 26.6% (95% CI, 14.5% to 38.6%; P < 0.001), respectively. The magnitude of effect was consistent with the primary analysis.

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (yes and no), and patients with moderate and severe UC at baseline were consistent with the results of the primary analysis. ||| ||||||||||| |||||||| ||| ||| |||||||||||||||| ||||||| ||| ||||||. Appendix 1 contains the LUCENT-2 study subgroup data.

Endoscopic Remission

At week 40 of the LUCENT-2 trial, 58.6% of patents on mirikizumab experienced endoscopic remission versus 29.1% of patients on placebo, with a common risk difference of 28.5% in favour of mirikizumab (95% CI, 20.2% to 36.8%; P < 0.001). Analyses in the ITT population were consistent with the mITT population results.

In the biologic-naive subgroup, 62.4% of patients on mirikizumab and 34.2% of patients on placebo experienced endoscopic remission at week 40, with a common risk difference of 28.2% (95% CI, 17.5% to 39.0%; P < 0.001). In the biologic-failed subgroup, 50.8% of patients on mirikizumab and 20.3% of patients on placebo experienced endoscopic remission at week 40, with a common risk difference of 30.5% (95% CI, 17.3% to 43.6%; P < 0.001).

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (yes and no), and patients with moderate and severe UC at baseline were consistent with the results of the primary analysis. Appendix 1 contains the LUCENT-2 study subgroup data.

Bowel Urgency

In the LUCENT-2 study, bowel urgency outcomes consisted of bowel urgency remission and bowel urgency improvement. The results are presented in Table 17.

Bowel Urgency Remission

Of the patients with a UNRS score of at least 3 at the LUCENT-1 study baseline (n = 420), 42.9% of patients on mirikizumab and 25% of patients on placebo at week 40 experienced bowel urgency remission, corresponding to a common risk difference of 18.1% (95% CI, 9.8% to 26.4%; P < 0.001). Analyses in the ITT population were consistent with results in the mITT population.

In the biologic-naive subgroup, 46.6% of patients on mirikizumab and 28.7% of patients on placebo experienced bowel urgency remission at week 40, with a common risk difference of 17.9% (95% CI, 7.0% to 28.8%; P = 0.002). In the biologic-failed subgroup, 35.2% of patients on mirikizumab and 19% of patients on placebo experienced bowel urgency remission at week 40, with a common risk difference of 30.5% (95% CI, 17.3% to 43.6%; P < 0.001).

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (yes and no), and patients with moderate and severe UC at baseline were consistent with the results of the primary analysis. Appendix 1 contains the LUCENT-2 study subgroup data.

Bowel Urgency Improvement (Change in UNRS)

At week 40 of the LUCENT-2 study, patients on mirikizumab experienced a –3.80-point change in UNRS versus the LUCENT-1 study baseline while patients randomized to placebo had a –2.74-point change in score from the LUCENT-1 study baseline (LSM difference = –1.06; 95% CI, –1.51 to –0.61; P < 0.001). Patients on mirikizumab experienced a clinically significant improvement in bowel urgency from baseline while those on placebo did not meet the MID threshold (MID = 3 points from baseline)²⁸ for clinically significant improvement.

Health-Related Quality of Life

HRQoL was assessed in the LUCENT-2 study based on the IBDQ score, EQ-5D-5L score, and SF-36 score.

IBDQ Score

In the LUCENT-2 study, the LSM mean change from the LUCENT-1 study baseline to week 40 in the IBDQ score was 49.8 points and 25.4 points for those in the mirikizumab and placebo groups, respectively; this represented a difference of 25.2 points in favour of mirikizumab (95% CI, 19.2 points to 31.3 points; P < 0.001).The difference between groups was considered clinically meaningful as the difference was above the MID of at least 15 points or greater above placebo.^21-24

EQ-5D-5L Score

In the LUCENT-2 study, the LSM difference at week 40 between groups in the EQ VAS score was 20.1 points in the mirikizumab group and 8.8 points in the placebo group, representing a difference of 11.3 points between the groups (95% CI, ||| || ||||; P < 0.001). The change from baseline in the mirikizumab group was clinically important based on an MID of 10.9 points on the EQ VAS.²⁵ It is unclear whether the difference between groups was clinically important given that no MID was identified for between-group differences.

SF-36 Score

At week 40 of the LUCENT-2 study, patients randomized to mirikizumab experienced an LSM change from induction baseline in SF-36 PCS of 9.0 points compared to 6.7 points in patients randomized to placebo, a 2.3-point difference between groups (95% CI, 1.1 points to 3.5 points; P < 0.001). In SF-36 MCS, mirikizumab patients had an LSM change from induction baseline of 7.0 points compared to 5.5 points in the placebo group (LSM change difference between groups = 1.5 points; 95% CI, 0.1 points to 2.8 points; P = 0.031). The change from baseline was clinically important in the mirikizumab group based on an MID threshold at least 3 points.²⁹ The clinical importance of the difference between groups is uncertain given that no MID was identified for between-group differences.

Mucosal Healing: HEMR

A greater proportion of patients randomized to mirikizumab experienced HEMR (a stricter outcome than HEMI) than those randomized to placebo at week 40 — 43.3% versus 21.8%, respectively (common risk difference = 19.9%; 95% CI, 12.1% to 27.6%; P < 0.001). Analyses in the ITT population were consistent with the mITT population results.

HEMR was analyzed by subgroups. In the biologic-naive subgroup, 47.2% of patients on mirikizumab and 26.3% of patients on placebo experienced HEMR at week 40, with a common risk difference of 20.8% (95% CI, 10.5% to 31.2%). In the biologic-failed subgroup, 36.8% of patients on mirikizumab and 13.8% of patients on placebo experienced HEMR at week 40, with a common risk difference of 21.2% (95% CI, 10.9% to 31.4%). The magnitude of the effect was consistent with the primary analysis.

Subgroup results for baseline corticosteroid use (yes and no), immunomodulator use (yes and no), and patients with moderate and severe UC at baseline were consistent with the results of the primary analysis. Appendix 1 contains the LUCENT-2 trial subgroup data.

Work Productivity

At week 40 of the LUCENT-2 study, patients randomized to the mirikizumab group had an LSM change from induction baseline of –31.72 points from the LUCENT-1 study baseline and placebo patients had an LSM change from induction baseline of –22.59 points from the LUCENT-1 study baseline, equating to an LSM difference of –9.13 points between groups (95% CI, –14.26 points to –4.01 points; P < 0.001). An MID was not identified for the WPAI:UC in patients for UC; hence, the clinical importance of the results is uncertain.

LUCENT-2 Study Extended Induction

Of the mirikizumab induction nonresponders from the LUCENT-1 trial, 272 patients entered the open-label extended induction arm of the LUCENT-2 trial in the mITT population. Of the 272 patients who underwent extended induction in the LUCENT-2 trial, 146 (53.7%) patients (95% CI, 47.8% to 59.6%) experienced a delayed clinical response after 24 weeks of continuous therapy with mirikizumab 300 mg IV, for a total of 6 doses. Additionally for this cohort of 272 patients, the rates of clinical remission, endoscopic remission, and symptomatic remission were 11.4% (95% CI, 7.6% to 15.2%), 16.5% (95% CI, 12.1% to 21.0%), and 37.1% (95% CI, 31.4% to 42.9%), respectively, at week 12 of the LUCENT-2 study. Furthermore, of the 146 mirikizumab induction nonresponders who experienced delayed clinical response at week 12 in the LUCENT-2 study, 144 patients entered the open-label maintenance period and 72.2% (n = 104) of patients maintained clinical response at week 40 versus ||||| ||||||| of placebo induction nonresponders who entered the maintenance period. Clinical response at week 40 was not evaluated in the LUCENT-2 trial for induction responders from the LUCENT-1 trial. Hence, no comment can be made on the difference in treatment effects between these 2 cohorts at week 40.

Table 17: Summary of Key Efficacy Results From LUCENT-2 Study (mITT Population at Week 40)

CI = confidence interval; EQ-5D-5L = 5-Level EQ-5D; HEMR = histologic endoscopic mucosal remission; IBDQ = Inflammatory Bowel Disease Questionnaire; LSM = least squares mean; MCS = mental component summary; mITT = modified intention-to-treat; PCS = physical component summary; q.4.w. = every 4 weeks; SC = subcutaneous; SE = standard error; SF-36 = Short Form (36) Health Survey; UNRS = Urgency Numeric Rating Scale; WPAI:UC = Work Productivity and Activity Impairment Questionnaire: Ulcerative Colitis.

Notes: Details in Table 17 have been taken from the sponsor’s Summary of Clinical Evidence.^20,30

The 2-sided alpha was 0.05 for the LUCENT-2 study.

^aIn the LUCENT-2 trial, UNRS, IBDQ, EQ-5D-5L, and WPAI:UC scores are reported as change from the LUCENT-1 study’s baseline to week 40 (e.g., 52 continuous weeks).

^bP value has not been adjusted for multiple testing.

Source: Clinical Study Report for the LUCENT-2 study.³²

Harms

A summary of harms reported in the LUCENT-1 and LUCENT-2 trials is provided in Table 18.

Adverse Events

For both the LUCENT-1 trial and the LUCENT-2 trial, the overall rate of AEs was similar between groups. In the LUCENT-1 study, 44.5% and 46.1% of patients reported an AE in the mirikizumab and placebo groups, respectively. In the LUCENT-2 study, 64.5% and 68.8% of patients reported an AE in the mirikizumab and placebo groups, respectively. Over 12 weeks of treatment in the LUCENT-1 study, the most common AEs for patients on mirikizumab 300 mg IV included nasopharyngitis (mirikizumab group = 4.1%; placebo group = 3.1%), anemia (mirikizumab group = 3.3%; placebo group = 5.9%), and headache (mirikizumab group = 3.3%; placebo group = 2.8%).

Over 40 weeks of treatment in the LUCENT-2 study, the most common AEs for patients on mirikizumab 200 mg SC included nasopharyngitis (mirikizumab group = 7.2%; placebo group = 5.7%), arthralgia (mirikizumab group = 6.7%; placebo group = 4.2%), and UC (mirikizumab group = 6.7%; placebo group = 20.8%).

Serious Adverse Events

Overall, the rate of SAEs in the LUCENT-1 and LUCENT-2 studies was found to be lower in patients treated with mirikizumab than with placebo; however, this was due to UC being included as a harm. In the LUCENT-1 study, 2.8% and 5.3% of patients reported an SAE in the mirikizumab and placebo groups, respectively. In the LUCENT-2 study, 3.3% and 7.8% of patients reported an SAE in the mirikizumab and placebo groups, respectively. In the LUCENT-1 trial, the most common SAEs in those on mirikizumab IV included UC (mirikizumab group = 0.8%; placebo group = 3.1%) and pneumonia (mirikizumab group = 0.2%; placebo group = 0%).

In the LUCENT-2 trial, no SAE (at the “preferred term” level) occurred in more than 1 patient on mirikizumab SC.

Withdrawals Due to Adverse Events

In the LUCENT-1 study, 1.6% and 7.2% of patients withdrew from the trial due to an AE in the mirikizumab and placebo groups, respectively. In the LUCENT-2 study, 1.5% and 8.3% of patients withdrew from the trial due to an AE in the mirikizumab and placebo groups, respectively. Withdrawals due to AEs occurred at a lower rate in mirikizumab-treated patients compared to placebo-treated patients in the LUCENT-1 and LUCENT-2 trials, ||||||| |||||||||| ||||||| |||||| ||| |||||||||| || || || ||||||| || ||||||||||||||||||| ||||||| |||||||||||||||| ||||||||||||| |||||||||||||||| |||||||||||||||| ||||||||||||||||| ||||||||||| ||| ||||||||||| |||||||||||||||| ||||||||| |||| ||| |||| |||||| || ||||| || ||||||||||| || || |||||||||

|||||||||| ||||||| ||| ||| |||| || ||||||||| |||| |||| ||||||| ||||||||||| |||||| |||||| ||||| || ||||||||| |||| |||| || ||||||||||| || |||||||| ||| |||| || ||||||| |||||||| ||||||||| |||| || || |||||| ||| |||||||| ||||||||||

Mortality

In the LUCENT-1 trial, no deaths were recorded. In the LUCENT-2 study, 1 (0.5%) death was recorded in the placebo group due to COVID-19.

Notable Harms

Most AESIs occurred at a similar rate between mirikizumab and placebo patients in the LUCENT-1 and LUCENT-2 studies. One exception was the rate of injection site reactions in the LUCENT-2 study, where 8.7% of patients on mirikizumab SC experienced this AESI compared to 4.2% of patients on placebo SC.

The rates of opportunistic infection, cerebrocardiovascular events, malignancy, depression, suicide/self-injury, and hepatic-related AEs were low overall and similar between groups for both the LUCENT-1 trial and the LUCENT-2 trial.

Table 18: Summary of Harms — Pivotal and RCT Evidence (Safety Population)

AE = adverse event; NA = not applicable; q.4.w. = every 4 weeks; RCT = randomized controlled trial; SAE = serious adverse event; SC = subcutaneous; TEAE = treatment-emergent adverse event; WDAE = withdrawal due to adverse event.

^aFrequency of 3% or more of patients in any treatment group during the LUCENT-1 or LUCENT-2 study.

^bFrequency of 2 or more patients in any treatment group during the LUCENT-1 or LUCENT-2 study.

Note: Details in Table 18 have been taken from the sponsor’s Summary of Clinical Evidence.^20,30

Sources: Clinical Study Reports for the LUCENT-1 and LUCENT-2 studies.^31,32

Critical Appraisal

Internal Validity

The LUCENT-1 and LUCENT-2 trials were parallel-arm, multicentre, double-blind, placebo-controlled randomized trials. Both trials employed appropriate methods for blinding, treatment allocation, and randomization. Randomization was stratified appropriately by important effect modifiers and used an interactive web response system for concealment of the randomization assignment. There were no notable differences between treatment arms for most baseline characteristics. The use of separate induction and maintenance studies was appropriate and consistent with other studies assessing other medications for the treatment of UC.⁷³ Statistical analyses were appropriate. In addition, a prespecified graphical multiple testing approach was used to control key secondary outcomes for multiplicity. The study was powered to detect a difference in the primary end point between treatment arms and the enrolled sample size was adequate. A hierarchical testing procedure was appropriately used to account for multiplicity in the primary and key secondary outcomes.

Many of the primary and major secondary end points as well as patient-reported outcomes (UNRS, IBDQ, EQ-5D-5L, and SF-36), were dependent on the complete and accurate daily recording of the Mayo SF and RB subscores and other data points by patients in their electronic diary. Daily reporting from patients is subjective and subject to error, and can lead to possible reporting bias and recall bias potentially in favour of mirikizumab if patients became unblinded. However, blinding was appropriate; hence, it is difficult to predict the direction of bias. The sponsor noted that there was a patient adherence review every 2 weeks at each study visit to ensure adherence to diary recordings. However, no details were provided on how adherence was reviewed. In addition, details were not provided on how patients were trained to use the electronic diary or tablet device that was used for the diary recordings to mitigate user error. In 2 of the study countries, Poland and Turkey, there was also a transcription error that led to 104 patients being excluded from the LUCENT-2 study’s primary efficacy analysis (but not the safety analysis). Overall, using multiple imputation to impute missing data due to transcription errors, COVID-19, loss to follow-up, or a protocol deviation did not have an impact on the primary or secondary end points.

Both pivotal trials used the mITT population for the main analysis rather than the ITT population. While the preferred approach would be to use the ITT population, results of the sensitivity analyses of the ITT population were aligned with the mITT results across all end points, which increases the certainty of the findings. The methods used for the sensitivity analyses to assess impact of attrition and missing data were valid and appropriate.

In both the LUCENT-1 study and the LUCENT-2 study, a notably higher proportion of patients in the placebo arm (11.5% and 38%, respectively) discontinued from the study compared with the mirikizumab arm (4.1% and 10.8%, respectively). The main reason for discontinuation was ||| || |||| |||||||||| ||| || ||| ||||||||||| || || ||||||. Prespecified sensitivity analyses were conducted to account for attrition and missing data and the results were consistent with the main analysis, increasing the certainty of the findings.

HRQoL and work productivity were identified as an important outcome by the patient and clinician groups providing input for this review. The IBDQ, EQ-5D-5L, and SF-36 were valid and reliable measuring instruments of HRQoL for patients with UC. MIDs were provided by the sponsor for the IBDQ, EQ-5D-5L, and SF-36 (PCS and MCS), which were in line with thresholds reported in the literature. MID estimates for these instruments have been established in patients with UC, although in the case of the EQ-5D-5L, sometimes the literature was based on IBD generally. An MID was provided for the WPAI:UC in patients with Crohn disease; however, this was not taken into consideration given the alternative patient population.

The LUCENT-2 trial enrolled 405 patients from the LUCENT-1 trial who had not responded to 12 weeks of induction dosing with either mirikizumab or placebo. These patients received open-label mirikizumab (300 mg administered intravenously) for 12 weeks. This was referred to as an extended induction period for patients who had previously received 12 weeks of induction dosing (i.e., 24 weeks of continuous therapy). As patients in the extended induction cohort received open-label mirikizumab, results should be interpreted with caution given the potential risk of detection or performance bias due to the open-label nature. In terms of subgroup analyses, the LUCENT-1 trial had included 1 subgroup analysis as part of its statistical testing hierarchy: clinical response in the biologic-experienced group. The results of the subgroup analysis were consistent with the results of the primary analysis, with mirikizumab being superior to placebo. No testing was done to compare between subgroup differences; hence, no conclusions could be drawn on whether mirikizumab is more beneficial in biologic-naive versus biologic-experienced populations. In terms of the other relevant prespecified subgroup analyses |||||| |||||||||||| ||||||||||| || |||||||| ||| ||| |||||||| || |||||||||| || ||||||||||| ||| || ||||| ||| || ||| |||| || |||||| |||| ||||||||||||| ||| |||| || ||||||| ||| |||||||||||| ||| |||||||| || ||| ||||||||| Hence, subgroup results should be considered exploratory.

External Validity

The clinical expert consulted by CADTH considered the baseline demographic and disease characteristics in the pivotal trials to be reflective of patients with moderate to severe UC seen in Canadian clinical practice. In addition, the clinical expert noted that the measurement instruments used were appropriate. Inclusion and exclusion criteria aligned with the selection criteria for candidates for mirikizumab treatment. However, the criteria that excluded patients who had received or whose UC had failed to respond to 3 or more biologic therapies may have excluded some patients unnecessarily, given that some patients may have received a biologic but stopped taking it due to reasons other than clinical failure (e.g., financial reasons). Hence, low socioeconomic patients may have been systemically excluded as a subgroup. Furthermore, there was a high number of screening failures (35%) in the LUCENT-1 study; hence, there may have been potential for selection bias in terms of how patients were being referred to the trial, according to the clinical expert.

Concomitant medication use was reflective of medications used in Canadian clinical practice with the exception of a few drugs, such as prednisolone, which is not typically used in Canada. In the LUCENT-2 trial, the maintenance trial, a corticosteroid taper was trialled on all patients in the main cohort. Patients who could not taper corticosteroids were not considered to be treatment failures; rather, tapering was reset or paused. This criterion contrasts with the input received from the clinical expert who indicated that patients would be considered treatment failures and discontinue treatment with the drug if they could not taper or stop concomitant corticosteroid use by the time of the maintenance phase (i.e., after the induction or extended induction period).³³ Therefore, the efficacy of mirikizumab in the trials may appear to be biased given that patients who could not taper were included in the primary analysis, even though they would have been considered treatment failures in clinical practice. However, the direction and magnitude of this bias is unknown, given that both groups underwent the same tapering protocol. Furthermore, the generalizability of the results may be limited to Canadian clinical practice due to the discrepancy in tapering protocol.

The trial criterion in the LUCENT-1 study of the achievement of clinical response, which directed entry into the primary cohort of the LUCENT-2 study, may have created an enriched patient population in the maintenance trial as it did not take into consideration delayed responders and only included patients who showed a timely response. As per the product monograph, extended induction dosing of mirikizumab may be considered in patients who experience delayed response (i.e., in adequate therapeutic response at week 12 after induction dosing); hence, by excluding these patients in the primary analysis, there is uncertainty about the efficacy of maintenance treatment in the broader population of patients with moderately to severely active UC.

The clinical expert noted that the duration of follow-up in the LUCENT-1 trial (i.e., 12 weeks) may not have been sufficient to adequately assess certain end points, such as endoscopic remission. The clinical expert noted that it was unlikely to see a difference in endoscopic outcomes by week 12 in the trial. However, the issue of insufficient duration was addressed by the LUCENT-2 trial, which measured end points at the week-40 time point (i.e., 52 weeks of continuous therapy). Long-term efficacy data beyond 52 weeks was not available; therefore, long-term outcomes such as durable clinical remission or endoscopic remission may not have been sufficiently captured between the 2 trials.

Long-Term Extension Studies

Content in this section has been informed by materials submitted by the sponsor. The following has been summarized and validated by the CADTH review team.

There are currently no published or unpublished long-term extension phase III or phase IV RCTS or real-world evidence studies. The sponsor did make note that there is an ongoing phase III, open-label, long-term extension trial — the LUCENT-3 study (I6T-MC-AMAP) — which is enrolling patients from the LUCENT-2 study and the phase II I6T-MC-AMAC study (NCT02589665).⁹⁷ The objective of the LUCENT-3 trial is to determine the efficacy and safety of mirikizumab 200 mg SC every 4 weeks in various clinical, patient-reported, and health outcomes at various time points up to 160 weeks (i.e., 160 weeks in the LUCENT-3 study, which represents up to 212 weeks of continuous mirikizumab therapy for patients coming from the LUCENT-2 trial). The expected primary completion date for the LUCENT-3 trial is June 6, 2025.

Indirect Evidence

Content in this section has been informed by materials submitted by the sponsor. The following has been summarized and validated by the CADTH review team.

Objectives for the Summary of Indirect Evidence

In the absence of direct head-to-head trials evaluating the comparative efficacy and safety of mirikizumab versus relevant comparators for moderately to severely active UC in adults, the sponsor submitted a systematic review with an NMA.⁷³ The sponsor-conducted NMA was used to inform the sponsor-submitted economic model for mirikizumab.

Description of Indirect Comparison

The sponsor submitted 1 NMA comparing the relative efficacy of induction and maintenance treatment with mirikizumab versus other advanced therapies on clinical response rate, clinical remission rate, and mucosal healing in adult patients with moderately to severely active UC. The NMA also assessed the relative safety of induction treatment with mirikizumab versus other advanced therapies on all-cause discontinuation and SAEs in the indication population.

ITC Design

Objectives

The objective of the sponsor-submitted NMA was to evaluate the relative efficacy and safety of mirikizumab versus relevant advanced therapies for moderately to severely active UC in adults.

Study Selection Methods

The studies eligible for inclusion in the sponsor-submitted NMA were selected according to a sponsor-conducted systematic review.³⁶ The sponsor’s systematic review was defined by the relevant population, intervention, comparators, outcomes, and study design (PICOS) described in Table 19. The scope of the systematic review included RCT evidence for adult patients undergoing treatment for moderately or severely active UC defined by a Mayo Score or by the Ulcerative Colitis Disease Activity Index.

Clinical evidence for the systematic review was identified using multiple electronic databases and trial registries as listed in Table 20, along with handsearching of conference proceedings. The reference lists of the 5 most relevant systematic reviews and meta-analysis were also scanned. The literature search is current to June 9, 2022. In addition to articles identified in the systematic review, the unpublished LUCENT-1 and LUCENT-2 trials were considered in the NMA. Articles were screened by 2 independent reviewers; a third reviewer was consulted to resolve any discrepancies. Data extraction was conducted by a single reviewer with a quality check by another. An assessment of risk of bias of included articles was conducted using the checklist for RCTs from CRD’s Guidance for Undertaking Reviews in Health Care (2009), courtesy of the Centre for Reviews and Dissemination.⁷⁴ Critical appraisals were not conducted for conference proceedings due to insufficient methodological data to assess the study quality. No studies were excluded based on the risk of bias assessment. The methods used for risk of bias assessment were not reported.

The study selection and methods for inclusion in the NMA are summarized in Table 20. For the purposes of the NMA, the population of interest was based on the population recruited in the LUCENT-1 and LUCENT-2 trials. All EMA-approved and FDA-approved doses and regimens of targeted therapies for the treatment of moderately to severely active UC were included as comparators in the NMA. Different dosing arms of the same medication were treated as individual comparators. Studies from the systematic review that did not meet the approved therapies, doses, and regimens criteria were not included in the NMA. Of note, among the comparators eligible for inclusion in the NMA, filgotinib and upadacitinib are currently not approved for use in Canada. Moreover, the ustekinumab maintenance regimen of 90 mg every 12 weeks is not used in Canada; the approved maintenance dose used in the Canadian practice setting is 90 mg SC every 8 weeks.

Outcomes evaluated in the NMA included clinical response and remission and mucosal healing and endoscopic remission for the induction and maintenance phases, as well as all-cause discontinuation (lack of efficacy and AEs) and the incidence of SAEs (grade 3 and grade 4 AEs) during the induction phase. The definitions of clinical response and remission and the definitions for safety end points were consistent with the definitions used in the included clinical trials. For the purposes of the NMA, the definition of mucosal healing was aligned with the LUCENT trials as being an ES of 0 or 1. Of note, due to different trial designs for the maintenance phase, the placebo safety populations in the maintenance phases across studies were not comparable. For example, mirikizumab patients in the LUCENT-1 study whose UC responded to blinded placebo were continued on blinded placebo in the LUCENT-2 study. However, in the tofacitinib phase III OCTAVE trial, placebo responders were rerandomized in the maintenance phase. Accordingly, only safety outcomes reported during the induction phase were analyzed.

Table 20: Study Selection Criteria and Methods for NMA Submitted by Sponsor

b.i.d. = twice a day; EMA = European Medicines Agency; NMA = network meta-analysis; p.o. = orally; q.2.w. = every 2 weeks; q.4.w. = every 4 weeks; q.8.w. = every 8 weeks; q.12.w. = every 12 weeks; RCT = randomized controlled trial; SC = subcutaneous; UC = ulcerative colitis; UCDAI = Ulcerative Colitis Disease Activity Index.

Note: Details in Table 20 have been taken from the sponsor’s Summary of Clinical Evidence.³⁰

^aHere, 1 mg of ozanimod refers to 1 mg of hydrochloride salt. 1 mg of ozanimod hydrochloride equals 0.92 mg of molecular ozanimod, which aligns with the dose form available in Canada.

^bCaptured at the end of the induction period of the respective study.

^cCaptured at the end of the maintenance period of the respective study.

Sources: Network meta-analysis technical report and systematic literature review technical report.^21,36

ITC Analysis Methods

Model Structure

Efficacy analyses were conducted for biologic-naive, biologic-experienced, and Asian-only subgroups of patients. Safety analyses were conducted on the overall population in the induction phase only. The NMA was powered by a Bayesian model using the Markov chain Monte Carlo method. A noninformative prior was used for unknown data. For the treatment effect prior, a normal density distribution was used (mean = 0; variance = 100). A normal prior (mean = 0; variance = 10,000) was used for the study-specific intercepts and for random-effects models, a half-normal prior (mean = 0; SD = 5) was used for the between-study SD.

A multinomial model with a probit link was used to account for the mutually exclusive nature of the relationship between response and remission outcomes (i.e., those who do not experience response cannot be in remission). For other dichotomous outcomes, a binomial model with a logit link was used.

A common baseline was estimated using all placebo arms, with the mean estimated using a normal prior (mean = 0; variance = 10,000). The between-study SD was estimated with a half-normal prior (mean = 0; SD = 5). Both fixed-effects and random-effects models were considered, and the deviance information criterion was used to select the most appropriate model.⁷⁵ Various graphical and statistical methods were used to assess model performance and facilitate selection.

Heterogeneity

There are 2 main trial designs related to the transition from induction to maintenance dosing: treat-through trial design and rerandomized responder design. To mitigate the heterogeneity caused by comparing patients from different trial designs, the treat-through trial results were adjusted for both response and remission end points. The adjustment incorporated these key assumptions:

• The total number of responders in treat-through trials during induction acted as a proxy for the number of patients entering maintenance.

• The adjusted clinical response in maintenance was based on the proportion of patients achieving durable or sustained response during maintenance in treat-through trials (prevents counting induction nonresponders).

• The adjusted clinical remission in maintenance was based on the number of patients achieving remission during the maintenance phase in treat-through trials (this assumed those in remission at maintenance were at least in response at induction).

To account for heterogeneity in patients and trial factors that are known to have a measurable impact on placebo response rates, baseline risk–adjusted models were considered that included study location, trial duration, disease status, disease duration, and prior exposure to biologic therapy.^76-80

Heterogeneity was assessed statistically by evaluating the posterior distribution for the between-study SD. If statistical heterogeneity was present, meta-regression, subgroup analysis, or random-effect modelling were considered.

Consistency was evaluated by comparing direct and indirect evidence for conflicting results. Tests using the unrelated mean effects model (i.e., the inconsistency model) were conducted, with results compared to that of the standard NMA using model fit statistics and the deviance contribution.

Sensitivity Analysis

Sensitivity analyses were conducted to explore the influence of potential outlier studies on the overall NMA result. Of note, data from treat-through design trials were excluded from the analysis of maintenance phase clinical response and remission in the sensitivity analysis. Given that the treat-through and rerandomized responder trial designs are so distinct from each other, the sensitivity analysis was used to provide insight on whether the approach in the base case NMA to standardize treat-through trials was sufficient in mitigating heterogeneity.

Table 21: NMA Analysis Methods

DIC = deviance information criterion; EMA = European Medicines Agency; MCMC = Markov chain Monte Carlo; NMA = network meta-analysis; UC = ulcerative colitis.

Note: Details in Table 21 have been taken from the sponsor’s Summary of Clinical Evidence.³⁰

Sources: Network meta-analysis technical report and systematic literature review technical report.^21,36

Results of NMA

Summary of Included Studies

In total, || publications reporting on || unique studies were included in the systematic literature review. The inclusion of the unpublished LUCENT trials resulted in || unique studies eligible for inclusion in the NMA. Of these, || studies were excluded from the NMA. In || of those, the reason for exclusion was the inclusion of a non–EMA-approved or non–FDA-approved dosing regimen. ||| ||||| was excluded due to not meeting the inclusion criteria for the study population and another for not presenting outcome data by subgroups for prior exposure to biologics and/or JAK inhibitors. In total, || studies were included in the NMA. An overview of the included studies is summarized in Table 21.

Of the included studies, || were induction studies (|| reported on induction clinical response; || on induction clinical remission and || on induction mucosal healing) and || maintenance studies (|| reported on maintenance clinical response, || reported maintenance clinical remission, and 12 on maintenance mucosal healing). At least 1 safety outcome for the overall population was reported in || studies. All included studies were double-blinded studies with sample size ranging from || || ||| patients. The duration of therapy ranged from | ||||| || || ||||| for induction and || ||||| || || ||||| for maintenance.

Baseline patient characteristics across the studies included in the NMA were not provided by the sponsor. The following details pertain to the studies included in the NMA feasibility assessment, part of the sponsor’s systematic literature review. Across the studies evaluating induction response, mean age ranged from |||| || |||| |||||, mean disease duration ranged from ||| || |||| |||||||| || ||||| of patients were male, and the mean total Mayo Score values ranged from ||| || |||. Across the studies evaluating maintenance response, mean age ranged from |||| || |||| years, mean disease duration ranged from ||| || |||| years, ||| || ||||| were male, and the mean total Mayo Score values ranged from ||| || |||.

Potential sources of heterogeneity across the included studies are summarized in Table 23. Two major sources of clinical and methodological heterogeneity identified by the sponsors were the differing rates of placebo group response across trials (I² = |||) and inconsistent maintenance phase treatment allocation (i.e., treat-through versus rerandomization responders) as detailed earlier in the Heterogeneity section.

Additional sources of heterogeneity were identified that were related to the definitions of clinical response, clinical remission, and prior experience with a biologic and/or JAK inhibitor applied across studies. The definition of induction and maintenance clinical response used in the LUCENT-1 trial differed from the commonly reported definition in the NMA (Table 24). Two definitions of induction and maintenance clinical remission were applied in the LUCENT-1 trial. Both these definitions differed from the most commonly reported definitions across studies included in the NMA (Table 25). Finally, the definition of prior experience with a biologic and/or JAK inhibitor varied across studies, as well as the prior therapies that patients may have been exposed to, due to the time period across which the studies occurred. Prior experience with a biologic and/or JAK inhibitor may be defined as TNF-experienced, TNF failure, biologic-experienced, biologic failure, or biologic and/or JAK inhibitor failure. As a result, the categorization of biologic-naive and JAK inhibitor–naive versus biologic-experienced and JAK inhibitor–experienced subgroups across studies was not consistent. Furthermore, permitted concomitant mediations differed across the included studies.

Table 22: Redacted