CADTH Reimbursement Review

Tralokinumab (Adtralza)

Sponsor: LEO Pharma Inc.

Therapeutic area: Atopic dermatitis

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Clinical Review

Executive Summary

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

Table 1: Background Information of Application Submitted for Review

AD = atopic dermatitis; NOC = Notice of Compliance.

^aA 300 mg per 2 mL prefilled pen is currently not marketed in Canada.

Introduction

Atopic dermatitis (AD), also referred to as eczema, is a chronic, heterogeneous inflammatory relapsing-remitting skin condition that occurs most frequently in early childhood.¹ It is estimated that the prevalence of eczema in Canada is 8.9% in adolescents (aged 13 to 14 years), and 3.5% in adults.^2,3 Acute worsening of AD, commonly referred to as flares, presents as dry, red, itchy skin that can lead to lesions that blister, ooze, and crust. An intense and debilitating itch and chronically relapsing eczematous lesions are the key clinical hallmarks of moderate-to-severe disease⁴ and could lead to sleep disturbances, psychosocial distress, and reduced quality of life (QoL) in patients and caregivers.

Conventional treatment options for moderate-to-severe AD include topical therapies, phototherapy, and off-label systemic immunosuppressants. Newer systemic treatments, including dupilumab (a biologic), abrocitinib and upadacitinib (oral small molecules, each of which is a Janus kinase inhibitor [JAKi]), are effective options that are currently available for patients who did not progress on conventional treatments, although some patients do not achieve an adequate response to dupilumab and JAKi treatments. Dupilumab is associated with conjunctivitis, which may necessitate treatment discontinuation for some patients.⁵ Upadacitinib and abrocitinib treatments require baseline and routine laboratory monitoring and have black-box warnings in the product monograph related to infections, malignancies, thrombosis, and major adverse cardiovascular events.^6,7

The objective of this report is to review and critically appraise the evidence submitted by the sponsor on the beneficial and harmful effects of tralokinumab, 150 mg per 1 mL prefilled syringe and 300 mg per 2 mL prefilled pen, administered by subcutaneous injection, in the treatment of moderate-to-severe AD in patients aged 12 years and older whose disease is not adequately controlled with topical prescription therapies or when those therapies are not advisable. Tralokinumab can be used with or without topical corticosteroids (TCS). Tralokinumab was previously reviewed by CADTH for the treatment of adults with moderate-to-severe AD and received a recommendation not to reimburse.⁸

Stakeholder Perspectives

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

Patient Input

Two separate patient group inputs were received — 1 from the Eczema Society of Canada and another from Eczema Québec and the Canadian Skin Patient Alliance. The Eczema Society of Canada’s input was based on a survey of 3,000 patients, caregivers and family members; questionnaires; and 1-on-1 interviews (number not reported) with patients and caregivers. Eczema Québec’s input was based on patient testimonials (n = 6), interviews (n = 10), and 2 group discussions (n = 13 in total), as well as insights gleaned from the McGill University Health Centre’s Centre of Excellence for Atopic Dermatitis and a report (The Skin I’m in: 2022 Update) from 2021 to 2023. The groups noted that symptoms of moderate-to-severe AD include inflamed, red, and dry skin that cracks, oozes, bleeds, and in some cases involves thickening and/or infections of the skin. Often, patients experience “flare-ups’ of worsening symptoms. Some patients experience remission, but others never experience relief. The input noted that itch is frequently reported as the most burdensome symptom and has been described as “incapacitating,” “debilitating,” or “bugs crawling all over,” leading to disrupted sleep, fatigue, decreased functionality, and significant impacts on daily life, work, and school. Skin rashes were reported to be not only painful but a source of embarrassment and stigmatization, affecting self-esteem and social relationships. Family members and/or caregivers shared that they experience negative impacts on intimacy, family dynamics, and relationships, as well as feelings of anxiety, depression, and sleep loss. Patients with moderate-to-severe AD also reported that their choices of work, clothing, foods, environments, hobbies, regular activities, travel, and hygiene routines are limited due to AD. Some patients reported contemplating suicide due to uncontrollable AD. The joint input by Eczema Québec and the Canadian Skin Patient Alliance quoted data from the Canadian Institute for Health Information showing that patients sometimes end up in the emergency department or become hospitalized when AD is not well controlled. Patients expressed a need for treatments that can result in improvement in symptoms (dryness, flaking, inflammation, blistering, and cracking), reduction in itch frequency and/or intensity, long-term improvement in QoL (sleep, prevention of flares, discomfort, psychological burden), and ability to carry out daily activities (work, school, leisure, personal hygiene), and that are safe (reducing infections with minimal short-term and long-term adverse effects), affordable, flexible, and easy to administer.

Clinician Input

Input From Clinical Experts Consulted by CADTH

The clinical experts noted that there is an unmet need for more treatment options for moderate-to-severe AD that are effective and safe, given that some patients do not respond or are refractory to the newer systemic treatments (dupilumab, upadacitinib, abrocitinib) and that JAKi options are associated with safety concerns. One clinical expert also noted that there is a need for treatment options that could improve adherence and convenience of drug administration for patients who are averse to needles (dupilumab is administered as a subcutaneous injection) or have difficulty adhering to daily administration of oral upadacitinib and abrocitinib.

The clinical experts expected tralokinumab to have the same place in therapy as dupilumab, serving as an additional biologic option for the treatment of moderate-to-severe AD after failure of off-label immunosuppressants. In the clinical experts’ opinion, any patient with moderate-to-severe AD could be a candidate for tralokinumab treatment. The clinical experts noted that tralokinumab would most likely be used in patients with AD in the absence of comorbid conditions such as asthma, chronic rhinosinusitis with nasal polyposis, and eosinophilic esophagitis, as these patients could benefit from dupilumab treatment instead, given that dupilumab is also indicated for the treatment of these conditions.

The clinical experts noted that disease improvement is assessed in clinical practice using instruments, such as the Physician’s Global Assessment (PGA; also referred to as the Investigator’s Global Assessment [IGA] in clinical trials), Eczema Area and Severity Index (EASI), Dermatology Life Quality Index (DLQI), Children’s Dermatology Life Quality Index (CDLQI), and worst daily pruritus numeric rating scale (NRS). In the clinical experts’ clinical experience, it takes approximately 6 months to observe optimal benefits from tralokinumab treatment. They noted that significant improvements in QoL and ability to perform daily activities are indicators of a meaningful response to treatment even if the skin is not completely clear of all erythema or lichenification. The clinical experts noted that it would be appropriate to consider switching therapy in patients who experience no improvement in clinical or patient-reported outcomes, or in those who have intolerable side effects. Tralokinumab could be prescribed by a dermatologist, allergist, immunologist, and pediatrician with expertise in the diagnosis, treatment, and monitoring of patients with AD, in the clinical experts’ opinion.

Clinician Group Input

Three clinician groups, the Atlantic Specialist Group Managing AD (7 clinicians), Dermatology Association of Ontario (16 clinicians), and Canadian Dermatology Association (unknown number of clinicians) provided separate inputs. The 3 clinician groups and the clinical experts consulted by CADTH agreed that the goals of therapy are to improve symptoms (long-term and durable relief of chronic itch; minimization of dry and inflamed skin; clear or almost clear skin’ less oozing, scaling, cracking, or fissures; and improved QoL (better sleep) and function (focus on work and school). The clinical experts added that a reduction in anxiety or depressive symptoms and caregiver burnout are goals of therapy. As for unmet needs, the clinician groups and the clinical experts consulted by CADTH all agreed that not all patients respond to or tolerate the existing systemic treatments. Each JAKi has safety and contraindication issues (black-box warnings for patients with risk factors for cardiovascular events, cancers, and infections), and dupilumab is associated with conjunctivitis. New treatments are therefore needed to provide more options for patients whose AD is not well controlled with existing systemic therapies. The clinician groups stated that tralokinumab would have the same place in therapy as dupilumab, after phototherapy and/or off-label systemic therapies (if required by insurance or public plans) and may be trialled if patients fail to respond to dupilumab and oral JAKi treatment. The clinician groups reported that the suitable patient population aligns with the reimbursement request. They also noted that those who did not respond to biologics and/or JAKi therapy; have a history of conjunctivitis and/or risk factors associated with cardiovascular events, thrombosis, malignancy, serious infections and/or significant drug-drug interactions; or find it challenging to adhere to stricter dosing schedules, and those over the age of 65 years would be best suited for tralokinumab treatment. The clinical experts added that tralokinumab would most likely be used in patients with “pure” AD without comorbid asthma or eosinophilic esophagitis and those with special site involvement. The 3 clinician groups and the clinical experts consulted by CADTH indicated that they would assess response to treatment based on body surface area (BSA) affected, pruritus NRS, PGA (in clinical practice) and/or EASI, if required by an insurance company or payers, at 6 months after initiation of tralokinumab. According to the clinician groups, a lack of response or efficacy, worsening disease, deterioration of QoL, increased affected BSA, presence of adverse events (AEs), unacceptable intolerance, and allergies would prompt clinicians to consider discontinuation of tralokinumab treatment. Last, the clinician groups and the clinical experts agreed that a dermatologist, allergist, pediatrician, or immunologist well versed in managing moderate-to-severe AD should be allowed to prescribe tralokinumab. The 3 clinician groups raised concerns regarding differential access to tralokinumab, which is currently only funded by private insurance, and the need to try off-label immunosuppressants with lower efficacy and increased risk before accessing newer systemic agents.

Drug Program Input

The drug programs identified recent safety warnings for abrocitinib and upadacitinib that may preclude these drugs from being true comparators for adolescent patients. The clinical experts noted that these treatments are appropriate comparators for tralokinumab given that JAKi therapies are used to treat AD in adolescent patients in clinical practice, although most clinicians generally prefer to prescribe biologics first due to a better safety profile in this patient population.

The drug programs expressed interest in understanding whether tralokinumab should be reimbursed when used in patients who lost response to, or never achieved clinical benefit from, a trial of dupilumab. The clinical experts described these patients as reasonable candidates for tralokinumab treatment. The CADTH review team noted that the benefits of tralokinumab in patients who had prior systemic dupilumab and/or JAKi treatments were inconclusive in the 2 sponsor-submitted observational studies due to important limitations of the studies, including the open-label, retrospective, and noncomparative study designs, and small sample sizes.

The drug programs noted that consideration might be given to aligning the initiation, renewal, and prescribing criteria for tralokinumab with the existing criteria for dupilumab.

Clinical Evidence

Pivotal Studies and Randomized Controlled Trial Evidence

Description of Studies

Five phase III, double-blind, randomized controlled trials (RCTs) assessed whether tralokinumab increased the proportion of patients with an IGA score of 0 (clear) or 1 (almost clear) and the proportion of patients with a reduction of at least 75% in an Eczema Area and Severity Index score from baseline (EASI-75) at week 16 compared to placebo in patients with moderate-to-severe AD, were included in the submission; 1 of which included adolescent patients (ECZTRA 6, N = 301)⁹ and 4 of which included adult patients (ECZTRA 1, N = 802; ECZTRA 2, N = 794, ECZTRA 3, N = 380; ECZTRA 7, N = 277).^10-12 The 4 studies in adults were previously reviewed by CADTH, and no new data from these studies were submitted for the current review. All enrolled patients had previously not progressed with topical therapy for AD. Patients in the ECZTRA 7 study had previously experienced uncontrolled disease on, or were not candidates for, systemic cyclosporine A treatment. Tralokinumab was compared with placebo, as monotherapy in the ECZTRA 1, 2, and 6 studies, and tralokinumab with TCS was compared with placebo plus TCS in the ECZTRA 3 and 7 studies. The proportion of patients with a reduction of at least 4 points in worst daily pruritus NRS, change from baseline in Scoring Atopic Dermatitis (SCORAD) score, and change from baseline in DLQI or CDLQI score were assessed at week 16 as key secondary end points in the ECZTRA 1, 2, 3, and 6 studies. In the ECZTRA 7 study, these were assessed as secondary end points at weeks 16 and 26. The mean age of the study population was 14.6 years (standard deviation [SD] = 1.7) in the ECZTRA 6 study and ranged between 36.5 years (SD = 14.1) and 39.1 years (SD = 15.2) in the ECZTRA 1, 2, 3, and 7 studies. The majority of patients were white and male in all studies. In the ECZTRA 6 study, prior systemic immunosuppressant, monoclonal antibody, and phototherapy treatments for AD were reported in 21.1%, 2.4%, and 25.6% of patients, respectively. In the ECZTRA 1, 2, 3, and 7 studies, prior phototherapy was noted in 43.7% to 58.8% of patients. Prior systemic immunosuppressant treatment was more common in the ECZTRA 7 study than in other studies in adults, with cyclosporine A being the most frequently used across studies (74.7% in the ECZTRA 7 study and 31.1% to 36.4% in the ECZTRA 1, 2, and 3 studies). Small proportions of patients in the ECZTRA 3 and 7 studies received prior monoclonal antibody treatment for AD (6.3% and 7.6%, respectively).

Efficacy Results — Initial Treatment Period

The key efficacy results in the initial treatment period of the ECZTRA 6 study (adolescents) and the ECZTRA 1, 2, 3, and 7 studies (adults) are summarized in Table 2 and Table 3, respectively. Results presented in this section pertaining to the primary estimand (i.e., the COVID-19–modified composite in the ECZTRA 7 study, the composite estimand in other studies for binary end points, and the hypothetical estimand for continuous end points in all studies), unless otherwise specified.

Investigator’s Global Assessment of 0 or 1

Adolescents (Aged 12 to < 18 Years): In the ECZTRA 6 study, the difference between the tralokinumab 300 mg every 2 weeks group and the placebo group in the coprimary end point of an IGA of 0 or 1 (i.e., the proportion of patients achieving an IGA score of 0 [clear] or 1 [almost clear]) at week 16 was 13.8% (95% confidence interval [CI], 5.3% to 22.3%; P = 0.002), in favour of tralokinumab.

Adults: The between-group differences in the coprimary end point of an IGA of 0 or 1 at week 16 were 8.6% (95% CI, 4.1% to 13.1%; P = 0.002) in the ECZTRA 1 study and 11.1% (95% CI, 5.8% to 16.4%; P < 0.001) in the ECZTRA 2 study when comparing tralokinumab every 2 weeks with placebo, and 12.4% (95% CI, 2.9% to 21.9%; P = 0.015) in the ECZTRA 3 study when comparing tralokinumab every 2 weeks plus TCS against placebo plus TCS; all favouring tralokinumab (or tralokinumab plus TCS).

In the ECZTRA 7 study, the between-group difference in the secondary end point of an IGA score of 0 or 1 was ||||| |||| ||| |||| || |||||| at week 16, and ||||| |||| ||| |||| || |||||| at week 26 when comparing tralokinumab every 2 weeks plus TCS with placebo plus TCS. Neither end point was tested for superiority due to prior failure in the testing hierarchy (i.e., reduction of worst daily pruritus NRS of at least 4 points from baseline).

Eczema Area and Severity Index

Adolescents: In the ECZTRA 6 study, the between-group difference in the coprimary end point of EASI-75 at week 16 was 22.0% (95% CI, 12.0% to 32.0%; P < 0.001), in favour of tralokinumab 300 mg every 2 weeks over placebo. Analyses of a reduction of at least 90% in the Eczema Area and Severity Index score from baseline (EASI-90), a reduction of at least 50% in the Eczema Area and Severity Index score from baseline (EASI-50), and change from baseline in EASI also showed results in favour of tralokinumab; however, these end points were not adjusted for multiplicity and were therefore at an increased risk of type I error (false-positive results).

Adults: The between-group differences in the coprimary end point of EASI-75 at week 16 were 12.1% (95% CI, 6.5% to 17.7%; P < 0.001) in the ECZTRA 1 study and 21.6% (95% CI, 15.8% to 27.3%; P < 0.001) in the ECZTRA 2 study when comparing tralokinumab every 2 weeks with placebo, and 20.2% (95% CI, 9.8% to 30.6%; P < 0.001) in the ECZTRA 3 study when comparing tralokinumab every 2 weeks plus TCS with placebo plus TCS; all favouring tralokinumab (or tralokinumab plus TCS).

In the ECZTRA 7 study, the between-group difference in the primary end point of EASI-75 at week 16 was 14.1% (95% CI, 2.5% to 25.7%; P = 0.018), in favour of tralokinumab every 2 weeks plus TCS over placebo plus TCS. The between-group difference in the secondary end point of EASI-75 at week 26 was 14.1% (95% CI, 2.9% to 25.35%), for which superiority testing was not conducted due to prior failure in the testing hierarchy.

In the ECZTRA 1, 2, and 3 studies, EASI-90, EASI-50, and change from baseline in EASI at week 16 were secondary end points. In the ECZTRA 7 study, EASI-90 scores at weeks 16 and 26 were exploratory end points, and change from baseline in EASI at week 16 and week 26 were secondary end points. Results of these outcomes were in favour of tralokinumab (or tralokinumab plus TCS); however, they were not adjusted for multiplicity and were therefore at an increased risk of type I error (false-positive results).

Scoring Atopic Dermatitis

Adolescents: In the ECZTRA 6 study, the between-group difference in the key secondary end point of adjusted mean change from baseline in SCORAD at week 16 was −19.7 (95% CI, −27.1 to −12.2; P < 0.001), in favour of tralokinumab 300 mg every 2 weeks over placebo. Results of the secondary (treatment policy) and tertiary (composite) estimands were consistent with those of the primary (hypothetical) estimand.

Adults: The between-group differences in the key secondary end point of adjusted mean change from baseline in SCORAD at week 16 were −10.4% (95% CI, −14.4% to −6.5%; P < 0.001) in the ECZTRA 1 study and −14.0% (95% CI, −18.0% to −10.1%; P < 0.001) in the ECZTRA 2 study when comparing tralokinumab every 2 weeks with placebo, and −10.8% (95% CI, −15.2% to −6.5%; P < 0.001) in the ECZTRA 3 study when comparing between tralokinumab every 2 weeks plus TCS with placebo plus TCS; all favouring tralokinumab (or tralokinumab plus TCS). Results of the secondary (treatment policy) and tertiary (composite) estimands were consistent with those of the primary (hypothetical) estimand.

In the ECZTRA 7 study, the between-group differences in the secondary end point of adjusted mean change from baseline in SCORAD were −8.6 (95% CI, −13.0 to −4.2) at week 16 and −8.9 (95% CI, −13.2 to −4.6) at week 26 when comparing tralokinumab every 2 weeks plus TCS with placebo plus TCS. Results of the secondary (treatment policy) and tertiary (COVID-19–modified composite) estimands were consistent with those of the primary estimand at weeks 16 and 26. Neither end point was tested for superiority due to prior failure in the testing hierarchy.

Worst Daily Pruritis Numeric Rating Scale and Adolescent Worst Pruritis Numeric Rating Scale

Adolescents: In the ECZTRA 6 study, the between-group difference in the key secondary end point of the proportion of patients with a reduction of at least 4 points in adolescent worst pruritus NRS at week 16 was 21.7% (95% CI, 12.3% to 31.1%; P < 0.001), favouring tralokinumab 300 mg every 2 weeks over placebo.

Results of the responder analysis based on a 3-point reduction threshold (secondary end point) also favoured tralokinumab. The between-group difference with respect to the secondary end point of adjusted mean change from baseline in adolescent worst pruritus NRS at week 16 was −1.5 (95% CI, −2.4 to −0.6). Neither end point was adjusted for multiplicity and was therefore at increased risk of type I error (false-positive results).

Adults: The between-group differences in the key secondary end point of proportion of patients with a reduction of at least 4 points in worst pruritus NRS at week 16 were 9.7% (95% CI, 4.4% to 15.0%; P = 0.002) in the ECZTRA 1 study and 15.6% (95% CI, 10.3% to 20.9%; P < 0.001) in the ECZTRA 2 study when comparing tralokinumab every 2 weeks with placebo, and 11.3% (95% CI, 0.9% to 21.6%; P = 0.037) in the ECZTRA 3 study when comparing tralokinumab every 2 weeks plus TCS against placebo plus TCS; all were in favour of tralokinumab (or tralokinumab plus TCS). Results of the responder analysis based on a 3-point reduction threshold (a secondary end point) were also in favour of tralokinumab (or tralokinumab plus TCS); however, this end point was not adjusted for multiplicity and was at increased risk of producing false-positive results.

In the ECZTRA 7 study, the proportions of patients with a reduction of at least 4 points in worst pruritus NRS at week 16 and at week 26 were secondary end points. The between-group difference at week 16 was 9.7% (95% CI, −2.0% to 21.4%; P = 0.106) at week 16, which did not indicate a difference between tralokinumab every 2 weeks plus TCS and placebo plus TCS. Results of the secondary (composite) estimand were consistent with those of the primary estimand. The between-group difference at week 26 was 7.3% (95% CI, −4.6% to 19.2%) and was not tested for superiority due to prior failure in the testing hierarchy.

The between-group differences in the secondary end point of adjusted mean change from baseline in worst pruritus NRS at week 16 were −0.9 (95% CI, −1.4 to −0.4) in the ECZTRA 1 study and −1.3 (95% CI, −1.7 to −0.8) in the ECZTRA 2 study when comparing tralokinumab with placebo, and −1.2 (95% CI, −1.7 to −0.7) in the ECZTRA 3 study when comparing tralokinumab every 2 weeks plus TCS against placebo plus TCS. In the ECZTRA 7 study, the between-group differences (exploratory end points) were −0.9 (95% CI, −1.4 to −0.4) at week 16 and −0.9 (95% CI, −1.4 to −0.3) at week 26 when comparing tralokinumab every 2 weeks plus TCS against placebo plus TCS. These end points were not adjusted for multiplicity and were at increased risk of type I error (false-positive results).

Dermatology Life Quality Index and Children’s Dermatology Life Quality Index

Adolescents: In the ECZTRA 6 study, the between-group difference in the key secondary end point of adjusted mean change from baseline in CDLQI at week 16 was −2.6 (95% CI, −4.5 to −0.7; P = 0.007), in favour of tralokinumab 300 mg every 2 weeks over placebo. Results of the secondary (treatment policy) and tertiary estimands (composite) were consistent with those of the primary (hypothetical) estimand.

Results of the responder analysis of proportion of patients with a reduction of at least 6 points in CDLQI from baseline at week 16 (secondary end point) were in favour of tralokinumab; however, this end point was not adjusted for multiplicity and was therefore at increased risk of type I error (false-positive results).

Adults: The between-group differences in the key secondary end point of change from baseline in DLQI at week 16 were −2.1 (95% CI, −3.4 to −0.8; P = 0.002) in the ECZTRA 1 study and −3.9 (95% CI, −5.2 to −2.6; P < 0.001) in the ECZTRA 2 study when comparing tralokinumab every 2 weeks with placebo and −2.9 (95% CI, −4.3 to −1.6; P < 0.001) in the ECZTRA 3 study when comparing tralokinumab every 2 weeks plus TCS against placebo plus TCS; all were in favour of tralokinumab (or tralokinumab plus TCS). Results of the composite estimand were consistent with those of the primary (hypothetical) estimand.

In the ECZTRA 7 study, changes from baseline in DLQI at weeks 16 and 26 were secondary end points. The between-group difference at week 16 was −1.5 (95% CI, −2.6 to −0.4). Results of the secondary (treatment policy) and tertiary (COVID-19–modified composite) estimands were not consistent with the primary (hypothetical) estimand and did not suggest a difference between the treatment groups. At week 26, the between-group difference was −1.6 (95% CI, −2.7 to −0.5). Results of the composite estimand were consistent with those of the primary estimand. Neither end point was tested for superiority due to prior failure of the testing hierarchy.

The proportion of patients with a reduction of at least 4 points in DLQI from baseline was a secondary end point (at week 16) in the ECZTRA 1, 2, and 3 studies, and an exploratory end point in the ECZTRA 7 study. Results favoured tralokinumab (or tralokinumab plus TCS) in the ECZTRA 1, 2, and 3 studies ||| ||| ||| ||||||||||||||| ||||||| |||||||||||| ||||||| ||| ||||||||||| || ||||| || ||| || || |||||| ||. These end points were not adjusted for multiplicity.

Other Efficacy End Points

Adolescents: In the ECZTRA 6 study, results of change from baseline in eczema-related sleep NRS, an exploratory end point; Patient-Oriented Eczema Measure (POEM), a secondary end point; and Hospital Anxiety and Depression Scale (HADS) anxiety scores, an exploratory end point at week 16, were in favour of tralokinumab 300 mg every 2 weeks over placebo. However, these end points were not adjusted for multiplicity and were at increased risk of producing false-positive results.

The results do not suggest a difference between treatment groups in change from baseline in HADS depression scores (an exploratory end point) at week 16. The 95% CI in the between-group difference in proportion of patients with a HADS anxiety or depression score of less than 8 (an exploratory end point) was wide, crossing the null.

Use of TCS and number of days without topical treatment were not assessed in the ECZTRA 6 study.

Adults: Results of change from baseline in eczema-related sleep NRS and POEM (exploratory end points) were in favour of tralokinumab (or tralokinumab plus TCS) across the ECZTRA 1, 2, 3, and 7 trials; however, these end points were not adjusted for multiplicity and were at increased risk of type I error (false-positive results).

Results did not consistently suggest a difference between tralokinumab (or tralokinumab plus TCS) and placebo (or placebo plus TCS) across studies with respect to change from baseline in HADS anxiety and depression scores, proportion of patients with HADS anxiety or depression scores of less than 8 (exploratory end points in the ECZTRA 1, 2, 3, and 7), amount of TCS used, and number of days without topical treatment (secondary end points in the ECZTRA 3 study and exploratory end points in the ECZTRA 7 study). These end points were not adjusted for multiplicity.

Efficacy Results — Maintenance (or Continuous) Treatment Period

IGA Score of 0 or 1 at Week 52 (ECZTRA 1, 2, and 6) or Week 32 (ECZTRA 3) Among Patients With an IGA of 0 or 1 at Week 16

Adolescents: In the ECZTRA 6 study, the proportions of patients receiving tralokinumab 300 mg every 2 weeks with an IGA 0 or 1 at week 16 who maintained their IGA of 0 or 1 response at week 52 were 37.5% (3 out of 8 patients; 95% CI, 13.7% to 69.4%) in the tralokinumab 300 mg every 2 weeks then every 2 weeks group and 87.5% (7 out of 8 patients; 95% CI, 52.9% to 97.8%) in the tralokinumab 300 mg every 2 weeks then every 4 weeks group. No statistical analysis was conducted to assess the between-group difference.

Adults: In the ECZTRA 1 and 2 studies, the proportion of patients with an IGA of 0 or 1 at week 16 (without use of rescue medication) who maintained their IGA of 0 or 1 (without use of rescue medication) at week 52 was included in the statistical hierarchy. In the ECZTRA 1 study, the difference between the tralokinumab every 2 weeks group and the placebo group was 6.0% (95% CI, −21.8% to 33.7%; P = 0.68). Due to failure of this end point, no superiority testing was conducted for the difference between the tralokinumab every 4 weeks group and the placebo group (lower in the testing hierarchy), which was −9.5% (95% CI, −37.1% to 18.0%). In the ECZTRA 2 study, the difference between the tralokinumab every 2 weeks group and the placebo group was 34.1% (95% CI, 13.4% to 54.9%; P = 0.004). The difference between the tralokinumab every 4 weeks group and the placebo group was 19.9% (95% CI, −1.2 to 40.9; P = 0.084). Due to failure of this end point, no superiority testing was conducted for the end point lower in the testing hierarchy (i.e., EASI-75 at week 52 between tralokinumab 300 mg every 4 weeks and placebo).

In the ECZTRA 3 study, the proportions of patients with an IGA of 0 or 1 at week 16 who maintained their IGA of 0 or 1 response at week 32 were 89.6% (95% CI not reported) in the tralokinumab every 2 weeks plus TCS group and 77.6% (95% CI not reported) in the tralokinumab every 4 weeks plus TCS group. No statistical analysis was conducted to assess the between-group difference. This end point was not assessed in the ECZTRA 7 study.

EASI-75 at Week 52 (ECZTRA 1, 2, and 6) or Week 32 (ECZTRA 3) Among Patients With EASI-75 at Week 16

Adolescents: In the ECZTRA 6 study, the proportions of patients with EASI-75 at week 16 (without use of rescue medication) who maintained their EASI-75 response at week 52 (without use of rescue medication) were 44.4% (4 out of 9 patients; 95% CI, 18.9% to 73.3%) in the tralokinumab 300 mg every 2 weeks (week 0 to 16) then every 2 weeks group (week 17 to 52) and 53.8% (7 out of 13 patients; 95% CI, 29.1% to 76.8%) in the tralokinumab 300 mg every 2 weeks (week 0 to 16) then every 4 weeks (week 17 to 52) group. No statistical analysis was conducted to assess the between-group difference on these end points.

Adults: In the ECZTRA 1 study, the proportion of patients with EASI-75 at week 16 (without use of rescue medication) who maintained their EASI-75 response (without use of rescue medication) at week 52 was not tested for superiority due to prior failure in the testing hierarchy (the proportion of patients with an IGA of 0 or 1 at week 16 who maintained their IGA of 0 or 1 at week 52). The difference between the tralokinumab every 2 weeks group and the placebo group was 21.2% (95% CI, −0.2% to 42.6%). The difference between the tralokinumab every 4 weeks group and the placebo group was 11.7% (95% CI, −8.7% to 32.0%).

In the ECZTRA 2 study, the difference in the proportion of patients with EASI-75 at week 16 who maintained their EASI-75 response at week 52 between the tralokinumab 300 mg every 2 weeks and placebo groups was included in the statistical testing hierarchy and was 33.7% (95% CI, 17.3% to 50.0%; P < 0.001). The difference in the proportion of patients with EASI-75 at week 16 who maintained their EASI-75 response at week 52 between the tralokinumab 300 mg every 4 weeks and placebo groups was not tested for superiority due to failure of a prior end point in the statistical testing hierarchy (i.e., IGA of 0 or 1 at week 52 between tralokinumab 300 mg every 4 weeks and placebo).

In the ECZTRA 3 study, the proportions of patients with an IGA of 0 or 1 at week 16 who maintained their IGA of 0 or 1 at week 32 were 92.5% (95% CI not reported) in the tralokinumab every 2 weeks plus TCS group and 90.8% (95% CI not reported) in the tralokinumab every 4 weeks plus TCS group. No statistical analysis was conducted to assess the between-group difference. This end point was not assessed in the ECZTRA 7 study.

Harms Results — Initial Treatment Period

The key harms results in the initial treatment period of ECZTRA 6 (adolescents) and ECZTRA 1, 2, 3, and 7 (adults) are summarized in Table 2 and Table 3, respectively.

Treatment-Emergent Adverse Events

In the initial treatment period of the ECZTRA 1, 2, 3, 6, and 7 studies, the proportion of patients with at least 1 treatment-emergent adverse event (TEAE) ranged between 61.5% and 77.5% in the tralokinumab (or tralokinumab plus TCS) group and between 61.7% and 78.8% in the placebo (or placebo plus TCS) group. No notable between-group difference in the proportion of patients who reported at least 1 TEAE in the initial treatment period was observed across studies. The most common TEAEs reported in the tralokinumab group (in at least 10% of patients) were upper respiratory tract infection (URTI), viral URTI, AD, conjunctivitis, and headache.

Serious Treatment-Emergent Adverse Events

The frequency of serious TEAEs in the initial treatment period ranged between 0.7% and in the tralokinumab (or tralokinumab plus TCS) group, and between 2.5% and 5.3% in the placebo (or placebo plus TCS) group in all pivotal studies.

Withdrawal due to Adverse Events

No treatment withdrawal due to AEs or death was reported in adolescent patients. In the ECZTRA 1, 2, 3, and 7 studies, the proportions of adult patients who withdrew from treatment due to AEs ranged from 0.7% to 3.3% in the tralokinumab (or tralokinumab 300 mg every 2 weeks plus TCS) group and from 0.8% to 4.1% in the placebo (or placebo plus TCS) group.

Mortality

Two deaths (related to an unknown cause and myocardial infarction) were reported in the tralokinumab group in the ECZTRA 1 study, and 1 death (related to metastatic squamous cell carcinoma) was reported in the tralokinumab group in the ECZTRA 2 study. No deaths were reported in all other studies.

Notable Harms

There was no notable difference between the tralokinumab group and the placebo group in the frequency of eczema herpeticum, malignancies, skin infection requiring systemic treatment, and eye disorders reported in adolescents and adults, except that conjunctivitis was consistently reported more frequently in the tralokinumab group (3.0% to 11.1%) than in the placebo group (1.5% to 4.4%) across the studies in adults.

Harms Results — Maintenance (or Continuous) Treatment Period

The overall results in the maintenance (or continuous) treatment period of the ECZTRA 1, 2, and 3 studies were consistent with those of the initial treatment period.

Critical Appraisal

The randomization and allocation concealment methods were adequate; although there were some baseline imbalances in the ECZTRA 3 and 6 studies, these may have been due to chance and did not appear to consistently favour either treatment group. The trials were adequately blinded; however, there is a small potential for bias in measurement of patient-reported outcomes (i.e., adolescent worst daily pruritus NRS, eczema-related sleep NRS, POEM, DLQI or CDLQI, and HADS) leading to inflated efficacy of tralokinumab due to possible unblinding in patients becoming aware of their assignments based on treatment response; however, the presence and extent of such potential bias is unknown. In the initial treatment period, an IGA of 0 or 1, EASI-75, reduction of at least 4 points in adolescent worst daily pruritus NRS from baseline, change from baseline in SCORAD and DLQI outcomes were controlled for multiplicity, while the other end points (secondary and exploratory) were not controlled and were at an increased risk of type I error (false-positive results). Continuous secondary and exploratory end points (change from baseline in EASI, POEM, worst daily pruritus NRS, eczema-related sleep NRS, and HADS scores) were at a high risk of bias due to a large amount of missing data that were not appropriately accounted for in the statistical analysis. No conclusion can be drawn on subgroup analyses due to the lack of sample-size consideration and control for multiplicity. In the maintenance (or continuous) treatment period, the IGA of 0 or 1 and EASI-75 outcomes were adjusted for multiplicity in the ECZTRA 1 and 2 trials; however, results were uncertain due to a sizable reduction in sample sizes, wide CIs for IGAs of 0 or 1 and EASI-75 outcomes, and inconsistent results between the ECZTRA 1 and 2 studies.

Table 2: Key Results From ECZTRA 6 — Initial Treatment Period (Adolescents)

AE = adverse event; CDLQI = Children’s Dermatology Life Quality Index; CI = confidence interval; EASI-75 = reduction of at least 75% in Eczema Area and Severity Index score from baseline; IGA = Investigator’s Global Assessment; NRS = numeric rating scale; q.2.w. = every 2 weeks; SCORAD = Scoring Atopic Dermatitis; SD = standard deviation; SE = standard error; TEAE = treatment-emergent adverse event.

Notes: The key outcomes summarized in this table include the coprimary efficacy end points, the key secondary efficacy end points, and other end points that were noted by the clinical experts consulted by CADTH to be of high importance for clinical decision-making. Unless otherwise specified, AEs are reported based on the Medical Dictionary for Regulatory Activities preferred term.

^aThe analysis was conducted using a Cochran-Mantel-Haenszel test stratified by baseline IGA and region based on the composite estimand.

^bThe analysis was conducted using the repeated measurements model, with baseline IGA, region, and treatment-by-week interaction as factors and interaction between week and baseline value as a covariate, based on the hypothetical estimand.

^cThe end point was not adjusted for multiplicity and was at an increased risk of type I error (false-positive result).

Sources: Clinical Study Report for ECZTRA 6¹³ and the sponsor’s Summary of Clinical Evidence.¹⁴

Table 3: Key Results From ECZTRA 1, 2, 3, and 7 — Initial Treatment Period (Adults; Original Review)

AE = adverse event; CI = confidence interval; DLQI = Dermatology Life Quality Index; EASI-75 = reduction of at least 75% in Eczema Area Severity Index score from baseline; IGA = Investigator’s Global Assessment; NR = not reported; NRS = numeric rating scale; q.2.w. = every 2 weeks; SCORAD = Scoring Atopic Dermatitis; SD = standard deviation; SE = standard error; TCS = topical corticosteroids; TEAE = treatment-emergent adverse event.

Note: Unless otherwise specified, AEs are reported based on the Medical Dictionary for Regulatory Activities preferred term.

^aThe analysis was conducted using a Cochran-Mantel-Haenszel test stratified by baseline IGA and region based on the composite estimand.

^bThe analysis was conducted using a Cochran-Mantel-Haenszel test stratified by prior cyclosporine A use, country, and baseline disease severity based on the COVID-19–modified composite estimand.

^cThis end point was included in the statistical testing hierarchy; however, no superiority testing was conducted for this end point due to prior failure in the statistical testing hierarchy.

^dThe analysis was conducted using the repeated measurements model, with baseline IGA, region, and treatment-by-week interaction as factors and interaction between week and baseline value as a covariate, based on the hypothetical estimand (primary estimand).

^eThe analysis was conducted using the repeated measurements model, with baseline IGA, country, prior cyclosporine A use and treatment-by-week interaction as factors, and interaction between week and baseline value as a covariate, based on the hypothetical estimand (primary estimand).

^fThe end point was not adjusted for multiplicity and was at an increased risk of type I error (false-positive result).

Sources: Clinical Study Reports for ECZTRA 1, 2, 3, and 7^15-18 and the sponsor’s Summary of Clinical Evidence.¹⁴

The study population of the ECZTRA 7 trial (i.e., adults who had uncontrolled disease or were not deemed to be candidates for topical therapy and cyclosporine A) was more reflective of the anticipated place in therapy of tralokinumab compared with other included RCTs in patients who had uncontrolled disease with topical therapy alone. The study interventions of the ECZTRA 3 and 7 studies (i.e., tralokinumab in combination with TCS) were also more reflective of the real-world use of tralokinumab compared with the ECZTRA 1, 2, and 6 studies (i.e., tralokinumab monotherapy) based on clinical expert input that patients typically use biologics in combination with TCS for active lesions. The clinical relevance of SCORAD, POEM, HADS outcomes is unclear given that these instruments are not routinely used in clinical practice. Based on their experience, the clinical experts consulted by CADTH considered the duration of follow-up in the initial treatment period (16 weeks) to be insufficient to adequately assess efficacy, as most patients would require at least 6 months of tralokinumab treatment to achieve an optimal response. Results of the maintenance treatment period (up to 52 weeks) are likely more generalizable but inconclusive due to issues with internal validity. The absence of direct comparative evidence between tralokinumab and relevant comparators (dupilumab, upadacitinib, and abrocitinib) represents a gap in pivotal trial evidence in the treatment of patients with moderate-to-severe AD.

Long-Term Extension Studies

Description of Study

One ongoing, open-label, single-arm, multicentre, long-term extension (LTE) trial, ECZTEND, was submitted by the sponsor. This study involved patients with moderate-to-severe AD who previously participated in clinical trials of tralokinumab (i.e., ECZTRA 1 to 8 and TraSki).¹⁴ Patients were eligible to participate in the ECZTEND trial if they had completed the treatment period(s) in 1 of the parent trials, regardless of the type of previous treatment (i.e., tralokinumab or placebo) or treatment response. All patients received tralokinumab with dosing administered by self-injection as prescribed by the product monograph. Patients were permitted to use concomitant TCS or topical calcineurin inhibitors (TCIs) and were required to apply an emollient at least twice daily for at least 14 days before the ECZTEND trial baseline and continue throughout the trial. The primary outcome was long-term safety, specifically the number of AEs experienced during the study. The secondary outcomes were based on efficacy and included achieving an IGA of 0 or 1 and EASI-75, each measured at weeks 16, 56, 88, 104, 136, 152, 184, 216, and 248. All analyses were descriptive and based on observed cases, with sensitivity analyses using last observation carried forward (LOCF) or modified nonresponder imputation (mNRI) to account for missing data. The 2 major cohorts used for the outcomes analyses were adults and adolescents. The data cut-off dates for the adult cohort for the reported interim analyses were April 30, 2021 (all participants from the ECZTRA 1, 2, 3, 4, 5, and 7 studies enrolled in the ECZTEND study, n = 1,442, with up to 3.5 years of follow-up; 3-year subgroup containing participants from the ECZTRA 1 and 2 studies, n = 347) and April 30, 2022 (4-year subgroup containing participants from the ECZTRA 1 and 2 studies, n = 347). The data cut-off date for the adolescent cohort was April 30, 2022 (participants from the ECZTRA 6 study, up to 3 years of follow-up, n = 127).

Efficacy Results

EASI-75

EASI-75 was assessed relative to the baseline in the parent trials. EASI-75 was achieved in 85.1% of patients (411 of 483, observed data) at week 104 in the ECZTEND study (i.e., up to 3 years of cumulative exposure to tralokinumab in the parent trials and the ECZTEND study) in the all-participants adult cohort; in 84.5% of patients (147 of 174, observed data) at week 152 in the ECZTEND study in the 4-year adult subgroup; and in 84.4% of patients (92 of 109, observed data) at week 56 in the ECZTEND study (i.e., 2 years of cumulative exposure to tralokinumab in the parent trials and the ECZTEND study) in the adolescent cohort. The results of the sensitivity analyses were consistent with those of the primary analysis using observed data.

IGA of 0 or 1

IGA of 0 or 1 was achieved in 50.5% of patients (244 of 483, observed data) at week 104 in the ECZTEND study (i.e., up to 3 years of cumulative exposure to tralokinumab in the parent trials and the ECZTEND study) in the all-participants adult cohort; in 52.6% of patients (92 of 175, observed data) at week 152 in the ECZTEND study in the 4-year adult subgroup; and in 61.5% of patients (67 of 109, observed data) at week 56 in the ECZTEND study (i.e., 2 years of cumulative exposure to tralokinumab in the parent trials and the ECZTEND study) in the adolescent cohort. The results of the sensitivity analyses were consistent with those of the primary analysis using observed data.

Harms Results

In the adult cohort (all participants, n = 1,442), 1,127 patients (78.2%) experienced at least 1 TEAE. In the 3-year adult subgroup (n = 347), 295 patients (85.0%) experienced at least 1 TEAE. In the adolescent cohort (n = 127), 83 patients (65.4%) experienced at least 1 TEAE. In all cohorts, the 3 most common AEs were a viral URTI (13.4% to 28.8%), AD (10.2% to 19.6%), and URTI (7.0% to 10.1%). Between 2.4% and 8.9% of patients reported a serious adverse event (SAE) in these cohorts. Conjunctivitis was reported in 77 patients (5.3%) and 7 patients (3.6%) from the all-participants adult cohort and the adolescent cohort, respectively. Frequency of treatment discontinuation was reported to be between 0.8% to 2.6%. No deaths were reported in the adult cohorts. However, 1 death (0.8%) due to an accident occurred in the adolescent cohort.

Critical Appraisal

Similar to other LTE studies, in the ECZTEND study, it is uncertain if the observed long-term effects can be attributed to tralokinumab treatment due to the lack of a comparison group and no adjustment for potential confounding. A risk of selection bias that favours tralokinumab is also possible, given that patients who perceived the treatment to be benefiting them during the parent trials were more likely to transfer to the extension study. Similarly, long-term safety concerns may be underestimated, as those who had experienced intolerable AEs in the parent trials were excluded from the ECZTEND trial. Given the open-label design of the study, there is also a risk of bias in the measurement of patient-reported outcomes (worst weekly pruritus NRS and DLQI), potentially favouring tralokinumab. The results related to benefits are at risk of being overestimated as they are interim findings.

The ECZTEND trial included patients who completed 1 of the parent trials regardless of treatment response. This is different from clinical practice, in which patients are expected to continue tralokinumab treatment only if they demonstrate objective improvement of disease after an adequate trial of treatment. It is unknown how many patients enrolled in the ECZTEND trial were nonresponders in the parent trial, potentially affecting the generalizability of the study population, because it is unclear what proportion of patients had experienced prior failure of immunosuppressant therapy, which is the likely place in therapy of tralokinumab. Further, the use of concomitant TCS and rescue medications could influence treatment response; however, utilization of such medications was not reported in the study and the impact on generalizability of study findings is therefore unclear.

Indirect Comparisons

In the absence of head-to-head evidence comparing tralokinumab to other relevant therapies used to manage AD, the sponsor submitted 4 indirect treatment comparisons (ITCs) of the effects of tralokinumab and other treatments in patients with moderate-to-severe AD. Of the ITCs submitted, 2 were network meta-analyses (NMAs), including 1 in adults and 1 in adolescents, and 2 were matching adjusted indirect comparisons (MAICs), both in adults.^19-22

Network Meta-Analyses

Description of Studies

The sponsor submitted an NMA conducted by the Institute for Clinical and Economic Review (ICER) that aimed to evaluate the relative efficacy and safety of treatment with tralokinumab versus other therapies in adult patients with moderate-to-severe AD.¹⁹ It is not clear if this NMA was identified by a systematic literature search, and, if so, how it was selected from the available literature. The ICER NMA was used to inform the sponsor-submitted economic model for the treatment effect of tralokinumab up to week 16. A sponsor-commissioned NMA, the LEO Pharma NMA, ||||| || |||||||| ||| |||||||| |||||||| ||| |||||| || ||||||||| |||| |||||||||||| |||||| ||||| ||||||||| || |||||||||| |||||||| ||||||| ||| |||| || || ||| || ||||| |||| |||||||||||||||||| || ||||||| ||||| || ||| |||

Efficacy Results

Efficacy results of the NMA are presented for monotherapy and combination therapy by population (i.e., adults and adolescents). A pairwise comparison against baricitinib is not presented as the treatment is not currently approved for use in Canada.

EASI-50

Adult Population (ICER Network Meta-Analysis): The EASI-50 treatment responses to all included monotherapy interventions in adult patients were greater than those to placebo. Treatments with upadacitinib 30 mg (relative risk [RR] = 1.75; 95% credible interval [CrI], 1.50 to 2.10), abrocitinib 200 mg (RR = 1.59; 95% CrI, 1.31 to 1.95), upadacitinib 15 mg (RR = 1.53; 95% CrI, 1.20 to 1.84), and dupilumab 300 mg (RR = 1.40; 95% CrI, 1.18 to 1.69) were favoured for achievement of EASI-50 compared to tralokinumab 300 mg. The point estimate for EASI-50 favoured abrocitinib 100 mg over tralokinumab 300 mg, but the CrI also included the potential of little-to-no difference between the treatments (RR = 1.21; 95% CrI, 0.95 to 1.53).

The treatment responses to all included combination-therapy interventions on EASI-50 in adult patients exceeded those to placebo. Treatments with upadacitinib 30 mg (RR = 1.45; 95% CrI, 1.27 to 1.71), abrocitinib 200 mg (RR = 1.32; 95% CrI, 1.14 to 1.57), upadacitinib 15 mg (RR = 1.32; 95% CrI, 1.15 to 1.57), dupilumab 300 mg (RR = 1.26; 95% CrI, 1.09 to 1.49), and abrocitinib 100 mg (RR = 1.20; 95% CrI, 1.02 to 1.43) were favoured for achievement of EASI-50 compared to tralokinumab 300 mg.

Adolescent Population (LEO Pharma Network Meta-Analysis): ||| ||||||||| |||||||| || ||| |||||||| ||||||||||| ||||||||||||| || ||||||| || || ||||| || |||||||||| |||||||| |||| |||||||| |||| |||||||| || |||||||||| ||||||||| ||| |||| ||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||||||||| ||||||| ||||||| |||||||||||| ||| ||| ||||| |||||| ||||||||||||

EASI-75

Adult Population (ICER Network Meta-Analysis): The EASI-75 treatment responses to all included monotherapy interventions in adult patients were greater than those to placebo. Treatments with upadacitinib 30 mg (RR = 2.77; 95% CrI, 1.77 to 2.77), abrocitinib 200 mg (RR = 1.89; 95% CrI, 1.45 to 2.49), upadacitinib 15 mg (RR = 1.79; 95% CrI, 1.42 to 2.29), and dupilumab 300 mg (RR = 1.58; 95% CrI, 1.25 to 2.03) were favoured for achievement of EASI-75 compared to tralokinumab 300 mg. The point estimate for EASI-75 favoured abrocitinib 100 mg over tralokinumab 300 mg, but the CrI also included the potential of little-to-no difference between the treatments (RR = 1.29; 95% CrI, 0.93 to 1.76).

The treatment response to all included combination-therapy interventions on EASI-75 in adult patients exceeded those to placebo. Treatments with upadacitinib 30 mg (RR = 1.90; 95% CrI, 1.53 to 2.45), abrocitinib 200 mg (RR = 1.58; 95% CrI, 1.25 to 2.07), upadacitinib 15 mg (RR = 1.48 95% CrI, 1.26 to 2.07), dupilumab 300 mg (RR = 1.46; 95% CrI, 1.15 to 1.90), and abrocitinib 100 mg (RR = 1.34; 9% Crl 1.03 to 1.76) were favoured for achievement of EASI-75 compared to tralokinumab 300 mg.

Adolescent Population (LEO Pharma Network Meta-Analysis): ||| ||||||||| |||||||| || ||| |||||||| ||||||||||| ||||||||||||| || ||||||| || |||||||||| |||||||| |||| |||||||| |||| |||||||| ||||||||| |||| |||||||||||| || || || ||| |||||||| ||| ||||||||||| || ||||||| |||||||| || |||||||||||| ||| || ||||| ||||| ||||| ||||| ||| |||| |||| || |||||| ||| || |||||||||||| ||| || |||| ||||| ||| ||| |||| || ||||||| ||| |||| ||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || ||||||||| ||||||| ||||||| |||||||||||| ||| ||| ||||| |||||| ||||||||||||

EASI-90

Adult Population (ICER Network Meta-Analysis): The EASI-90 treatment responses to all included monotherapy interventions in adult patients were greater than those to placebo. Treatments with upadacitinib 30 mg (RR = 2; 95.89% CrI, 2.19 to 3.95), abrocitinib 200 mg (RR = 2.36; 95% CrI, 1.65 to 3.39), upadacitinib 15 mg (RR = 2.17; 95% CrI, 1.60 to 3.00), and dupilumab 300 mg every 2 weeks (RR = 1.83; 95% CrI, 1.34 to 2.54) were favoured for achievement of EASI-90 compared to tralokinumab 300 mg. The point estimate for EASI-90 favoured abrocitinib 100 mg over tralokinumab 300 mg, but the CrI also included the potential of little-to-no difference between the treatments (RR = 1.39; 95% CrI, 0.91 to 2.09).

The EASI-90 treatment responses to all included combination-therapy interventions in adult patients were superior to those to placebo. Treatments with upadacitinib 30 mg (RR = 2.74; 95% CrI, 1.98 to 3.97), abrocitinib 200 mg (RR = 2.01; 95% CrI, 1.41 to 2.98), upadacitinib 15 mg (RR = 2.01; 95% CrI, 1.43 to 2.96), dupilumab 300 mg (RR = 1.76; 95% CrI, 1.24 to 2.57), and abrocitinib 100 mg (RR = 1.54; 95% CI, 1.05 to 2.31) were favoured for achievement of EASI-90 compared to tralokinumab 300 mg.

Adolescent Population (LEO Pharma Network Meta-Analysis): The EASI-90 treatment responses to all included monotherapy interventions in adolescent patients were superior to those to placebo. Treatment with upadacitinib 15 mg was favoured for achievement of EASI-90 compared to tralokinumab 150 mg (odds ratio [OR] = 15.82; 95% Crl, 1.60 to 734.24) and tralokinumab 300 mg (OR = 17.95; 95% Crl, 1.74, 843.75). Treatment with upadacitinib 30 mg once daily was favoured for achievement of EASI-90 compared to tralokinumab 150 mg (OR = 47.05; 95% CrI, 4.82 to 2,297.05) and to tralokinumab 300 mg (OR = 53.95; 95% Crl, 5.45 to 2,620.94). The Crls for comparisons were too wide to draw any conclusions of certainty in in achieving EASI-90 between tralokinumab and the other active comparators.

Investigation’s Global Assessment

Adult Population (ICER Network Meta-Analysis): The IGA treatment responses to all included monotherapy interventions in adult patients were greater than those to placebo. Treatments with upadacitinib 30 mg (RR = 3.97; 95% CrI, 2.54 to 6.31), upadacitinib 15 mg (RR = 3.07; 95% CrI, 1.88 to 4.99), abrocitinib 200 mg (RR = 2.75; 95% CI, 1.54 to 4.95), and dupilumab 300 mg (RR = 2.15; 95% CrI, 1.31 to 3.60) were favoured for achievement of an IGA of 0 or 1 compared to tralokinumab 300 mg. The Crls for the comparison between tralokinumab and abrocitinib 100 mg were too wide to draw any conclusions of certainty in IGA responses in adult patients receiving monotherapy for AD.

The IGA treatment responses to all included combination interventions in adult patients were superior to those to placebo. Treatments with upadacitinib 30 mg (RR = 2.83; 95% CrI, 1.90 to 4.27), abrocitinib 200 mg (RR = 2.24; 95% CI, 1.44 to 3.49), upadacitinib 15 mg (RR = 2.08; 95% CrI, 1.35 to 3.25), dupilumab 300 mg (RR = 1.85; 95% CrI, 1.20 to 2.88), and abrocitinib 100 mg (RR = 1.66; 95% CI, 102 to 2.68) were favoured for achievement of an IGA of 0 or 1 compared to tralokinumab 300 mg.

Adolescent Population (LEO Pharma Network Meta-Analysis): ||| ||||||||| |||||||| || ||| |||||||| ||||||||||| ||||||||||||| || ||| ||| || |||||||||| |||||||| |||| |||||||| |||| |||||||| ||||||||| |||| |||||||||||| || || || ||| |||||||| ||| ||||||||||| || ||| ||| |||||||| || |||||||||||| ||| || |||| ||||| ||| |||| |||| || |||||| ||| || |||||||||||| ||| || |||| ||||| ||| ||| |||| || ||||||| ||| |||| ||| ||||||||||| |||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || || ||||||||| ||| ||| ||||||| |||||||||||| ||| ||| ||||| |||||| ||||||||||||

Peak Puritis NRS Improvement of 4 Points or Greater

Adult Population (ICER Network Meta-Analysis): The treatment responses to all included monotherapy interventions as measured by an improvement of 4 points or greater on the peak pruritis numeric rating scale (PP-NRS) in adult patients were greater that those to placebo. Treatments with upadacitinib 30 mg (RR = 2.16; 95% CrI, 1.14 to 4.58), dupilumab 300 mg (RR = 2.12; 95% CrI, 1.06 to 4.43), and upadacitinib 15 mg (RR = 1.97; 95% CrI, 1.01 to 4.28) were favoured for achievement of an improvement of at least 4 points in PP-NRS compared to tralokinumab 300 mg. The Crls for the remaining comparisons were too wide to draw any conclusions of certainty in improvement of at least 4 points in PP-NRS between tralokinumab and other active comparators among adult patients.

The treatment responses to all included combination-therapy interventions as measured by an improvement of at least 4 points in PP-NRS in adult patients were greater that those to placebo. Treatments with upadacitinib 30 mg (RR = 2.37; 95% CrI, 1.75 to 3.29), abrocitinib 200 mg (RR = 2.04; 95% Crl, 1.47 to 2.89), upadacitinib 15 mg (RR = 1.91; 95% CrI, 1.34 to 2.74), and dupilumab 300 mg (RR = 1.79; 95% CrI, 1.28 to 2.55) were favoured for achievement of improvement of at least 4 points in PP-NRS compared to tralokinumab 300 mg. The point estimate for improvement of at least 4 points in PP-NRS favoured abrocitinib 100 mg over tralokinumab 300 mg, but the CrI also included the potential of little-to-no difference between the treatments (RR = 1.40; 95% Crl, 0.93 to 2.10).

Adolescent Population (LEO Pharma Network Meta-Analysis): ||| ||||||||| |||||||| || ||| |||||||| ||||||||||| ||||||||||||| || |||||||| ||| ||||||||||||| |||| || || |||||||||| |||||||| |||| |||||||| |||| |||||||| || |||||||||| ||||||||| ||| |||| ||| |||||||||||| ||| |||| || |||| ||| ||||||||||| || ||||||||| || |||||||| ||| |||||||||||| |||| || ||||||| |||||||||||| ||| ||| ||||| |||||| ||||||||||||

Children’s Dermatology Life Quality Index

A network meta-analysis of the CDLQI was not reported in the ICER NMA.

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

Patient-Oriented Eczema Measure

A network meta-analysis of POEM was not reported in the ICER NMA.

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

Harms Results

Adverse Events

A network meta-analysis of harms data was not reported in the ICER NMA.

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

Critical Appraisal

ICER Network Meta-Analysis

The ICER NMA was based on studies identified from a systematic literature review of relevant randomized evidence of treatments for adults and adolescent with AD. The systematic literature search was based on a patient, intervention, comparison, and outcome (PICO) model defined a priori, with efficacy and safety outcomes predefined. The systematic literature search was comprehensive. The selection process was not clearly defined, and data extraction was conducted by a single reviewer, increasing the risk of bias and error. While the risk of bias of the comparator trials was assessed, the method used was not reported, and risk of bias was not assessed by outcome. Several sources of clinical and heterogeneity were identified, which challenged the plausibility of the underlying transitivity assumption. These included variations in patient age, duration of disease, disease severity, length of the washout period, time point of follow-up (12 to 16 weeks), and methods of imputation for missing data. To account for differences in corticosteroid use across trials, separate NMAs were conducted for monotherapy and combination therapies. However, the treatment of patients in the control group (placebo plus TCS) were not consistent across the combination-therapy trials. Statistical heterogeneity and consistency were not tested, despite the availability of several closed loops.

The networks were sparse (several comparisons with relatively few studies), and all comparisons to tralokinumab were indirect, which increased the uncertainty in the findings. No sensitivity analysis exploring possible assumptions made by the reviewers were reported. Moreover, there was no indication of model adjustment to account for the correlation in the 3 arm trials. Harms outcomes were not evaluated.

LEO Pharma Network Meta-Analysis

The LEO Pharma NMA was based on studies identified from a systematic review of relevant randomized evidence of treatment for moderate-to-severe AD in adolescent patients. The systematic literature search was based on a PICO model defined a priori, with efficacy and safety outcomes predefined. The systematic literature search was comprehensive. The reasons for study exclusions were reported; and the selection and data-extraction processes were adequate to minimize the risk of bias and error. While the risk of bias of the comparator trials was assessed, the methods used were not reported and the risk of bias was not assessed by outcome. Several sources of heterogeneity were identified across the included studies. These included variation in the time point of follow-up, the predetermined duration of AD for study inclusion, exclusion criterion related to prior use of biologics, and protocol for investigational drug discontinuation for rescue treatment.

No information was given on model fit, and assessment of statistical consistency despite the presence of closed loops. No sensitivity analysis exploring possible assumptions made by the reviewers was reported. All comparisons to tralokinumab were indirect, which introduces increased uncertainty in the findings. Due to the small sample sizes, the CrIs were wide for several comparisons, which precluded drawing conclusions about comparative efficacy and safety for those outcomes.

Matching Adjusted Indirect Comparisons

Description of Studies

The sponsor submitted 2 MAICs conducted on its behalf by a third party comparing the relative efficacy of tralokinumab versus dupilumab in adults with moderate-to-severe AD.^21,22 In both MAICs, evidence for tralokinumab was based on individual patient data, while evidence for dupilumab was based on published aggregate data. ||| |||||||| |||| ||||| || ||| |||||||| ||||||| || |||| |||||| ||||| || ||||||| |||||||||||| ||||||| || |||||||||||| |||| ||||||||| ||||| || ||||||||||| |||| ||||||| ||||||||||||||| |||||| ||||||||| |||||||| || ||||| |||||||||||| ||| || ||||||||||| |||| ||||||| ||||||||||||||| || || ||||| || ||||||||| || |||||| |||||||| |||| || ||| ||| ||||||||| ||||||||| ||||||||| |||||||||||| ||||||||||^|| The unanchored MAIC based on the ECZTRA 3 and LIBERTY AD CHRONOS trials aimed to assess the long-term efficacy outcomes for tralokinumab 300 mg (ECZTRA 3) administered every 2 weeks and 300 mg every 4 weeks against dupilumab (LIBERTY AD CHRONOS) every 2 weeks at 32 to 52 weeks of follow-up in adult patients with moderate-to-severe AD.²¹

Efficacy Results

ECZTRA 7 Versus LIBERTY AD CAFÉ

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

ECZTRA 3 Versus LIBERTY AD CHRONOS

After matching, the reported baseline characteristics of the weighted patient population of the ECZTRA 3 study were matched with those of the LIBERTY AD CHRONOS study. A total of 106 patients were included in the dupilumab treatment group. The effective sample size (ESS) following match adjustment was 123.4 for the tralokinumab treatment arm (49.36% of the original population).

The results of the unanchored efficacy MAIC analysis of the ECZTRA 3 versus LIBERTY AD CHRONOS studies between tralokinumab and dupilumab was in favour of tralokinumab for an IGA of 0 or 1 (risk difference [RD] = 13.9; 95% CI, 0.6 to 27.3) and change in DLQI (mean difference = −1.7; 95% CI, −3.0 to −0.3) at week 52. The CI were too wide to draw any conclusions of certainty about the remaining outcomes between tralokinumab and dupilumab (at week 32: EASI-75, EASI-50, EASI-90, and IGA of 0 or 1; at week 52:% change in EASI, change in worse daily pruritis NRS, percent change in SCORAD, change in POEM; at week 53: EASI-75, EASI-50, EASI-90, worst daily pruritis NRS improvement of at least 4 points, POEM improvement of at least 4 points, DLQI improvement of at least 4 points).

Harms Results

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

No harms end points were evaluated in the ECZTRA 3 versus LIBERTY AD CHRONOS MAIC.

Critical Appraisal

ECZTRA 7 Versus LIBERTY AD CAFÉ

A comparison of the ECZTRA 7 study versus the LIBERTY AD CAFÉ study was chosen after a review of | trials evaluating the treatment of tralokinumab or dupilumab in patients with moderate-to-severe AD. There was no description of a literature search or selection criteria, or any indication of how the 8 trials were located. The sponsor noted that the decision to conduct a MAIC was based on substantial heterogeneity that precluded standard indirect comparisons (e.g., an NMA or Bucher comparison). How the matching variables were selected for the MAIC was not described. Baseline characteristics postmatching were well balanced, with almost perfect matching of the covariates included in the MAIC. However, complete baseline demographic and disease characteristics for patients in both trials were not reported. The application of weights resulted in a reduced ESS of ||||, in which ||||| of enrolled patients in the ECZTRA 7 study were lost. The reduction in sample size in the primary analysis resulted in imprecision, leading to uncertainty of the results. Sensitivity analysis using a larger population by way of an unadjusted indirect comparison was generally consistent with the primary MAIC, but with a narrower CI favouring dupilumab. There was no assessment of residual confounding in the analysis.

ECZTRA 3 Versus CHRONOS

The ECZTRA 3 versus CHRONOS MAIC lacked a description of a literature search or selection criteria, or any indication of how the trials were selected for the MAIC. There was also a lack of transparency in the data-extraction process and quality assessment. Although both the ECZTRA 3 and LIBERTY AD CHRONOS trials included a placebo, an unanchored MAIC was conducted. The decision to conduct an unanchored MAIC was appropriately justified due to differences in trial design (re-randomized versus treat-through) that may have resulted in differences in the treatment of placebo across the ECZTRA 3 and LIBERTY AD CHRONOS trials. Nonetheless, the ECZTRA 3 versus LIBERTY AD CHRONOS MAIC was limited by heterogeneity between the dupilumab target population and the analysis set. First, the dupilumab target population in the LIBERTY AD CHRONOS study was not the same analysis set used in the results reported at week 32 and week 52. Consequently, the matched tralokinumab population may not be completely representative of the dupilumab population results in reports at week 52. Next, the time point at which tralokinumab (week 32) and dupilumab (week 52) were compared were different. The magnitude and direction of bias related to differences in the analysis time point is uncertain. However, because the clinical experts suggested that superior results are expected for tralokinumab at week 52 versus week 32, the analysis may be at risk of a bias in favour of treatment with dupilumab. Unadjusted and match-adjusted baseline covariates were reported. Baseline characteristics postmatching were well balanced, with almost perfect matching of the covariates included in the MAIC. However, complete baseline demographic and disease characteristics for patients in both trials were not reported. The application of weights resulted in a reduced ESS of 123.4, in which 50.64% of enrolled patients in the ECZTRA 3 study were lost. The reduction of sample size in the primary analysis resulted in imprecision, leading to uncertainty of the results. There was no assessment of residual confounding in the analysis.

Studies Addressing Gaps in the Pivotal and RCT Evidence

Description of Studies

Two observational studies were submitted by the sponsor to address gaps in evidence. There was no description of the search or selection methods used to identify these studies. Pezzolo and Naldi was an open-label, retrospective, 12-week case series conducted in Italy (N = 12) and published as a letter to the editor. This study included 12 adult patients whose disease was uncontrolled with dupilumab treatment.²³ Pereyra-Rodriguez et al. (N = 85) was a retrospective, 16-week study conducted in Spain.²⁴ This study assessed 85 adult patients, including those who had previously been treated with either dupilumab (29.4%) or upadacitinib (8.2%). These 2 studies also assessed clinical experience with tralokinumab in real-world settings.

Efficacy Results

In the study by Pezzolo and Naldi, the mean EASI score at baseline before any systemic therapy was 36.58 (range = 21 to 47). All 12 adult patients with AD reached EASI-75 within 8 weeks, with continuing improvement at 12 weeks. The mean EASI scores were 27.58 (range = 20 to 35) at study baseline and 4.67 (range = 0 to 13) at week 12. The mean itch NRS scores were 8.42 (range = 7 to 10) at baseline and 2.92 (range = 0 to 7) at week 12. The mean sleep NRS scores were 7.0 (range = 3 to 10) at baseline and 1.92 (range = 0 to 5) at week 12. In the study by Pereyra-Rodriguez et al., the mean EASI scores at baseline were 25.4 (SD = 8.1) and 7.5 (SD = 6.9) at week 16. The mean SCORAD scores were 55.8 (SD = 13.3) at baseline and 20.0 (SD = 14.78) at week 16. The mean PP-NRS scores were 8.1 (SD = 1.8) at baseline and 3.5 (SD = 2.4) at week 16. At baseline, 47 patients (55.3%) had an IGA of 4. At the end of the follow-up period, 18.8% of patients (the absolute number was not reported) had an IGA of 0 or 1.

Harms Results

In the study by Pezzolo and Naldi, no serious AEs were reported. Also, the conjunctivitis that had been observed in 4 patients during the previous treatment with dupilumab did not recur. In the study by Pereyra-Rodriguez et al., the most frequent AEs were conjunctivitis and red face (5 patients, 5.9% each) with 1 patient having both events at the same time. Of those 5 patients, 2 had experienced conjunctivitis with prior dupilumab treatment, and 3 were naive to advanced therapy with no prior eye-related AEs. Moreover, 3 patients (3.5%) experienced worsening and generalized AD lesions, 2 patients (2.4%) developed reactions at the injection site, and 2 patients (2.4%) reported anxiety-depressive syndrome. One patient discontinued treatment due to severe conjunctivitis.

Critical Appraisal

It is not clear how the studies addressing gaps were selected, creating a potential for study selection bias (i.e., relevant studies may have been left out). There is a high level of uncertainty in the results due to the following study limitations common to both studies: small sample sizes (N = 12 in Pezzolo and Naldi; N = 85 in Pereyra-Rodriguez et al.); potential selection bias in the absence of a clear description of patient selection methods; noncomparative study design with a lack of adjustment for confounding; a lack of clarity on whether the studies were designed a priori; and if retrospective data were collected in a systematic way. As well, no formal hypothesis testing was performed in the study by Pezzolo and Naldi. There was no control for multiple comparisons in Pereyra-Rodriguez et al., which resulted in an increased risk of false-positive results. The durations of follow-up (12 weeks in Pezzolo and Naldi and 16 weeks in Pereyra-Rodriguez et al.) were also inadequate for assessing response to tralokinumab treatment, according to the input of clinical experts consulted by CADTH. As neither of the studies included adolescent patients, the treatment effects in adolescents who had prior dupilumab and/or JAKi treatments were not addressed by these studies.

Conclusions

Evidence from 3 pivotal double-blind RCTs demonstrated that 16 weeks of treatment with tralokinumab resulted in improvements in the severity and extent of AD as measured by EASI, IGA or SCORAD; severity of itching (worst daily pruritus NRS); and health-related quality of life (HRQoL) as measured by DLQI in adults with moderate-to-severe AD who had an inadequate response to topical AD therapy, compared to placebo, when used as monotherapy (in the ECZTRA 1 and 2 studies) and in combination with TCS (the ECZTRA 3 study); however, either the magnitude of improvement in SCORAD and DLQI scores did not meet the literature-reported minimal important difference (MID) estimates or the 95% CI included the potential of a difference falling below conservative MID estimates. Analyses of other clinically important outcomes, including sleep disturbance and symptoms of anxiety and depression, also favoured tralokinumab, although, due to a large amount of missing data and a lack of adjustment for multiplicity, these results are likely to be biased. Similar results were observed in 1 pivotal double-blind RCT (ECZTRA 6) conducted with tralokinumab monotherapy in adolescents who had an inadequate response to topical AD therapy; MID estimates were reached for the improvement in SCORAD score but not for the improvement in CDLQI scores at week 16. The anticipated place of therapy of tralokinumab is in patients with moderate-to-severe AD with inadequate response to topical AD therapy and phototherapy, as well as systemic immunosuppressants. One double-blind RCT (ECZTRA 7) of adults with severe AD who had an inadequate response to topical AD therapy and cyclosporine A provided supporting evidence for the use of tralokinumab in combination with TCS in such patients. The ECZTRA 7 trial found that, compared to placebo, 16 weeks of tralokinumab was associated with a higher proportion of patients with an EASI-75 response but it did not find a difference in reducing the severity of itching (worst daily pruritus NRS). The results of the analysis of other outcomes were inconclusive due to a prior failure in the statistical testing hierarchy. Overall, interpretation of the clinical meaningfulness of findings at week 16 from the RCTs was hindered by the insufficient duration of follow-up given that an optimal response to tralokinumab is usually observed at 6 months in clinical practice, according to clinical expert input. No conclusion can be drawn on the efficacy of tralokinumab beyond week 16 based on the submitted evidence due to important limitations of the included studies, including inconsistent results between trials and evidence of imprecision in the longer-term results in RCTs, and risks of selection bias and confounding due to the noncomparative trial design of the LTE study (ECZTEND), similar to other long-term extension studies.

Evidence from 4 ITCs comparing tralokinumab to other advanced therapies for the treatment of moderate-to-severe AD suggests that tralokinumab performs worse than or similar to its main comparator, dupilumab, in adults. The lone exception to this trend was noted in the ECZTRA 3 versus LIBERTY AD CHRONOS unanchored MAIC, in which the results favoured tralokinumab versus dupilumab for an IGA of 0 or 1 and change in DLQI at week 52. Comparisons of the efficacy and safety of tralokinumab versus abrocitinib and upadacitinib are considered uncertain. The combined ITC evidence for adults is associated with important uncertainty due to the potential for intransitivity in the NMA, and potential residual confounding and lack of precision in the MAICs. Results of the lone NMA evaluating the efficacy of tralokinumab in adolescents were imprecise and potentially affected by intransitivity that precludes any conclusion of certainty about the comparative efficacies of tralokinumab versus dupilumab, abrocitinib, or upadacitinib. Evidence from 2 observational studies on the use of tralokinumab in patients who had uncontrolled disease with prior dupilumab and/or JAKi treatment was inconclusive given the small sample sizes, and the open-label, noncomparative study designs.

Tralokinumab appeared to be well tolerated in adults and adolescents in the RCTs and remained so beyond 52 weeks based on evidence from the ECZTEND trial. However, based on the indirect evidence, no conclusion about the safety of tralokinumab compared with other advanced therapies for moderate-to-advance AD can be drawn with any certainty, due to imprecision in all relevant analyses (2 MAICs in adults and 1 NMA in adolescents).

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 tralokinumab, 150 mg per 1 mL prefilled syringe and 300 mg per 2 mL prefilled pen, administered by subcutaneous injection for the treatment of moderate-to-severe AD in patients aged 12 years and older whose disease is not adequately controlled with topical prescription therapies or when those therapies are not advisable. Tralokinumab can be used with or without TCS.

Disease Background

Contents within this section were informed by materials submitted by the sponsor and clinical expert input. The following summary was validated by the CADTH review team.

Atopic dermatitis, also known as eczema, is a chronic, heterogeneous, inflammatory, relapsing-remitting skin condition.¹ The pathogenesis of AD involves a complex interplay between genetic predisposition, the environment, skin barrier dysfunction, and immune dysregulation.^4,25-27 The immune dysregulations in AD are predominantly driven by T-helper 2 lymphocytes and their derived cytokines.^4,26 Interleukin (IL)-13 is the primary cytokine responsible for the signs and symptoms of moderate-to-severe AD, and its levels in the serum have shown to be correlated to disease severity.^4,26,28-33

Cases of AD most frequently occur in early childhood, with studies estimating that the prevalence of eczema symptoms across age groups in Canada is around 12% in children aged 6 to 7 years, 8.9% in adolescents aged 13 to 14 years, and 3.5% in adults (aged 18 years and older).^2,3 The majority outgrow the disease later in life, while 25% to 40% of cases will persist into adulthood.³⁴ Although AD is more common in adolescents, adulthood disease tends to be more severe, with 8% to 17% of affected adults experiencing severe disease.^3,35,36 Nonetheless, adolescents who suffer from moderate-to-severe AD are challenged to endure significant burdens that also affect their caregivers during a formative and transitional phase of their life.

Patients with AD often experience acute worsening of their condition, commonly referred to as flares. During flares, patients with AD can experience extremely dry, red, itchy skin that can lead to lesions that can blister, ooze, and crust. Excessive scratching can lead to skin infections, as well as long-term changes to the skin, including skin thickening.³⁷ An intense and debilitating itch (pruritus) and chronically relapsing eczematous lesions are the key clinical hallmarks of moderate-to-severe disease.⁴ In a 2013 Global Burden of Disease study, 41.5% of patients with moderate-to-severe AD had itch for at least 18 hours a day.³⁸ The presence of itch is estimated to be significantly higher in adolescents aged 12 to 15 years than in older patients with AD (98.0% versus 87.3%, respectively; P < 0.05).³⁹ As a result of this itching, patients experience sleep disturbances (e.g., difficulty falling asleep and frequent awakenings), as reported by the Eczema Society of Canada Quality of Life Report for 2016 to 2017, in which 79% of survey respondents reported interrupted or loss of sleep.^37,40

Chronic itching and sleep disturbance are associated with mental health instability.^38,41-43 Individuals with AD reportedly exhibit higher rates of depression, anxiety, suicidal ideation, attention deficit or hyperactivity disorder, and autism spectrum disorder.^4,44-46 Psychosocial impacts such as avoidance of social activities, avoidance of exercise, missing of work and important life events, and the need to change careers or give up certain activities were reported in 48%, 47%, 32% and 30% of adult patients, respectively.³⁷ Similarly, anxiety, difficulty in participating in physical activities, avoidance of social activities, and bullying were reported in 30%, 30%, 29%, and 14% of Canadian children and caregivers, respectively.³⁷ Individuals who suffer from AD and their families or caregivers experience a negative impact on the quality of life due to the psychological distress associated with the disease.

The pan-Canadian (excluding Québec) prevalence of moderate-to-severe AD in adult and adolescent populations who have been treated with off-label immunosuppressants was estimated to be 82,506.^47,14

Standards of Therapy

Contents within this section were informed by materials submitted by the sponsor and clinical expert input. The following summary was validated by the CADTH review team.

The foundation of AD therapy for all patients relies on basic skin care, which involves liberally applying moisturizers and emollients to the skin and avoiding aggravating factors that may be physical, environmental, or chemical.⁴⁸ For patients whose disease is moderate to severe in nature, conventional treatment options may include phototherapy, off-label immunosuppressants, and systemic corticosteroids.⁴⁸ However, these therapies are limited by accessibility and/or safety challenges. Phototherapy is not widely available across Canada and may be impractical for patients and caregivers to access given the frequency of treatments required at specialty clinics (2 to 3 times per week).^48-50 Off-label systemic immunosuppressants such as cyclosporine A and methotrexate are typically used for the shortest duration and at the lowest doses possible due to the lack of long-term efficacy, risk of significant side effects, and burdensome laboratory-monitoring requirements.^49,51-55 The clinical experts consulted by CADTH also mentioned that methotrexate is not recommended in women of childbearing potential given that it is an abortifacient and can be teratogenic. Systemic corticosteroids provide a quick relief of AD, but are associated with short- and long-term safety issues, such that the American Academy of Dermatology has recommended that these therapies be avoided or used in short-term durations only as a bridge to other drugs.^37,52

More recent novel therapies that have been reviewed by CADTH for adults and adolescents with moderate-to-severe AD include biologics (dupilumab) and oral small molecules, including upadacitinib and abrocitinib, both of which are a JAKi.^6,7,56 Despite these important advancements in therapy for AD patients, there are still patients who do not achieve an adequate response to dupilumab and JAKi treatments. While dupilumab is generally well tolerated, drug-induced conjunctivitis has been observed in clinical trials and in clinical practice, necessitating discontinuation for some patients.⁵⁷ Upadacitinib and abrocitinib treatments require baseline and routine laboratory monitoring and have black-box warnings in the product monograph related to infections, malignancies, thrombosis, and major adverse cardiovascular events.^6,7

According to the clinical experts consulted by CADTH for this review, the order of escalation of therapy is as follows: TCS and/or topical nonsteroids (a TCI or phosphodiesterase type 4 [PDE-4] inhibitor), followed by newer systemic treatments, including biologics and a JAKi, and off-label systemic immunosuppressant treatment (typically methotrexate). Most jurisdictions in Canada currently reimburse the cost of dupilumab for patients aged 12 years and older with moderate-to-severe AD. Upadacitinib and abrocitinib are currently undergoing negotiation by the pan-Canadian Pharmaceutical Alliance for patients aged 12 years and older with moderate-to-severe AD. The clinical experts consulted by CADTH noted that physicians prefer to err on the side of a superior safety profile; for this reason, among patients with moderate-to-severe AD who were considered for newer systemic treatments (biologics or JAKi treatment), biologics tend to be trialled first, reserving JAKi options for patients who have an inadequate response to biologics (although there are rare cases in which a JAKi may be trialled before biologics). The clinical experts added that adolescent patients are much more likely to trial a biologic before a JAKi, given the possible longer-term safety implications that come with a JAKi.

The clinical experts consulted by CADTH stated that reducing the severity of symptoms, minimizing adverse effects, improving HRQoL (including improvement in sleep quality anxiety or depressive symptoms), increasing the ability to maintain employment and independence, and reducing caregiver burden are important goals of treatment of patients with moderate-to-severe AD.

Drug Under Review

The key characteristics of tralokinumab are summarized in Table 4, along with other treatments available for patients with moderate-to-severe AD.

Tralokinumab is available in a 150 mg per 1 mL prefilled syringe and 300 mg per 2 mL prefilled pen, both of which can be self-administered as subcutaneous injections.⁵⁸ The 300 mg per 2 mL prefilled pen is not currently marketed in Canada. The recommended dose for adult and adolescent patients 12 years and older is an initial 600 mg (four 150 mg injections with the 150 mg per 1 mL prefilled pen or 2 300 mg injections with the 300 mg per 2 mL prefilled pen) followed by 300 mg (two 150 mg injections with the 150 mg per 1 mL prefilled pen or 1 300 mg injections with the 300 mg per 2 mL prefilled pen) administered every 2 weeks) as a subcutaneous injection. At the prescriber’s discretion, every-fourth-week dosing may be considered for some patients who achieve clear or almost clear skin after 16 weeks of treatment. Some patients with an initial partial response may subsequently improve with continued treatment every 2 weeks beyond 16 weeks.⁵⁸

Tralokinumab is indicated for the treatment of moderate-to-severe AD in adult and adolescent patients 12 years and older whose disease is not adequately controlled with topical prescription therapies or when those therapies are not advisable. Tralokinumab can be used with or without TCS or TCIs.⁵⁸ Tralokinumab was previously reviewed by CADTH for treatment of moderate-to-severe AD in adult patients and received a recommendation not to reimburse.⁸

Tralokinumab is a fully human immunoglobulin G4 monoclonal antibody that binds specifically to type 2 cytokine IL-13 and inhibits its interaction with IL-13 receptors alpha 1 and alpha 2. Tralokinumab neutralizes the activity of IL-13 by blocking its interaction with the IL-13 receptor alpha 1/IL-4 receptor alpha complex. IL-13 is a major driver of type 2 inflammation in AD, with skin showing overexpression of IL-13. IL-13 signals via the IL-13 receptor alpha 1/IL-4 receptor alpha complex and stimulates inflammatory responses, contributes to itch induction, and impairs the expression of proteins necessary for a normal skin barrier.⁵⁸

The sponsor-requested reimbursement criteria differ from the Health Canada–approved indication but align with the February 2023 CADTH Canadian Drug Expert Committee recommendation in response to the Request for Advice for dupilumab,⁵⁹ i.e., for the treatment of patients aged 12 years and older with moderate-to-severe AD whose disease is not adequately controlled with topical prescription therapies or when those therapies are not advisable, and those who had an adequate trial or are ineligible for each of the following therapies: phototherapy (where available) and off-label immunosuppressants.

Table 4: Key Characteristics of Tralokinumab, Dupilumab, Upadacitinib, and Abrocitinib

AD = atopic dermatitis; EASI-75 = reduction of at least 75% in Eczema Area Severity Index score from baseline; IgG4 = immunoglobin G4; IL = interleukin; JAK1 = Janus kinase 1; MACE = major adverse cardiovascular event; q.2.w. = every 2 weeks; q.4.w. = every 4 weeks; STAT = signal transducers and activators of transcriptions; TSLP = thymic stromal lymphopoietin (cytokine family protein).

^aHealth Canada–approved indication.

Sources: Sponsor’s clinical evidence summary¹⁴ and Adtralza,⁵⁸ Dupixent,⁵⁶ Rinvoq,⁶ and Cibinqo⁷ Health Canada product monographs.

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 inputs received by CADTH are included in the stakeholder section at the end of this report.

The Eczema Society of Canada and Eczema Québec in collaboration with the Canadian Skin Patient Alliance submitted 2 separate patient group inputs. The Eczema Society of Canada’s input was based on a survey of 3,000 patients or caregivers and family members; questionnaires; and 1-on-1 interviews (number not reported) with patients and caregivers. Eczema Québec’s input was based on patient testimonials (n = 6), interviews (n = 10), and 2 group discussions (n = 13 in total), as well as insights gleaned from the McGill University Health Centre’s Centre of Excellence for Atopic Dermatitis and a report (The Skin I’m In: 2022 Update) from 2021 to 2023. The groups noted that symptoms of moderate-to-severe AD include inflamed, red, and dry skin that cracks, oozes, bleeds and in some cases involves thickening and/or infections of the skin. Often, patients experience “flare-ups” that are periods of worsening symptoms. Some patients experience remission, but others never experience relief. The input noted that itch is frequently reported as the most burdensome symptom and has been described as “incapacitating,” “debilitating,” and “bugs crawling all over,” leading to disrupted sleep, fatigue, decreased functionality, and significant impacts on daily life, work, and school. Skin rashes were reported to be not only painful but a source of embarrassment and stigmatization affecting self-esteem and social relationships. Family members and/or caregivers reported impacts on intimacy, family dynamics, and relationships, and experience feelings of anxiety, depression, and sleep loss. Patients with moderate-to-severe AD also reported that AD limits their choices of work, clothing, foods, environments, hobbies, regular activities, travel, and hygiene routines. Some patients reported contemplating suicide due to uncontrollable AD. The joint input by Eczema Québec and the Canadian Skin Patient Alliance quoted data from the Canadian Institute for Health Information indicating that patients sometimes end up in the emergency department or become hospitalized when AD is not well controlled. Patients expressed a need for treatments that can result in improvement in symptoms (dryness, flaking, inflammation, blistering, cracking), reduction in itch frequency and/or intensity, long-term improvement in quality of life (sleep, prevention of flares, discomfort, psychological burden), and improved ability to carry out daily activities (work, school, leisure, personal hygiene), and that are safe (reducing infection and minimizing short- and long-term adverse effects), affordable, flexible, and easy to administer.

Clinician Input

Input From Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise in 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 (providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, and providing guidance on the potential place in therapy). The following input was provided by 2 clinical specialists with expertise in the diagnosis and management of AD.

Unmet Needs

The clinical experts noted that none of the available treatment options can reverse the course of the disease. They added that not all patients respond to the newer systemic treatments (dupilumab, upadacitinib, and abrocitinib) for moderate-to-severe AD and that some patients can become refractory to current treatment options. There is therefore an unmet need for more effective and safe treatment options for moderate-to-severe AD. Currently, many patients with AD have access to only 1 biologic drug (dupilumab). In the event of inadequate treatment response or side effects to dupilumab, patients are limited to receiving JAKi treatments, which are associated with more safety concerns compared with biologics. One clinical expert also identified a need for treatment options that can improve adherence to and the convenience of drug administration for patients who are averse to needles (dupilumab is administered as subcutaneous injection) or have trouble adhering to daily administration of oral upadacitinib and abrocitinib.

Place in Therapy

According to the clinical experts, the addition of tralokinumab would likely not result in a paradigm shift. The clinical experts expected tralokinumab to occupy the same place in therapy as dupilumab, serving as an additional biologic option for the treatment of moderate-to-severe AD after failure of off-label immunosuppressants. In the clinical experts’ opinion, the potential toxicities of off-label immunosuppressants and the need for close monitoring make biologics ideal first-line treatments in patients who require a systemic treatment, although they noted that most jurisdictions currently require patients to try an immunosuppressant treatment before receiving dupilumab, and this requirement would likely be applicable to tralokinumab.

Patient Population

In the clinical experts’ opinion, any patient with moderate-to-severe AD could be a candidate for tralokinumab treatment. The clinical experts noted that tralokinumab would most likely be used in patients with AD in the absence of comorbid conditions such as asthma, chronic rhinosinusitis with nasal polyposis, and eosinophilic esophagitis, because these patients could benefit from dupilumab treatment, given that dupilumab is also indicated for the treatment of these conditions. The clinical experts noted that no specific diagnostic test is available to help identify responders to treatment, and clinicians generally use clinical assessments and reviews of medical history to determine if tralokinumab treatment is appropriate.

Assessing the Response Treatment

The clinical experts noted that the effectiveness of therapy is determined by clinical outcomes (e.g., percent BSA involved and number of active areas) and patient-reported outcomes (e.g., itch, sleep, and QoL) that are typically assessed every 3 to 6 months. The clinical experts noted that, in their experience, it takes approximately 6 months to observe the optimal benefits of tralokinumab treatment, and they noted that a lack of response before 6 months should not be considered evidence of treatment failure. The clinical experts described significant improvement in QoL and ability to perform daily activities (e.g., work, school, and household activities) as indicators of a meaningful response to treatment even if the skin is not completely clear of all erythema or lichenification. Measurements of disease improvement such as a PGA (or IGA in clinical trials) and EASI, and patient-reported outcomes, including DLQI (or CDLQI) and worst daily pruritus NRS, are commonly used in clinical practice to assess response to treatment.

Discontinuing Treatment

The clinical experts noted that it would be appropriate to consider a switch of therapy in patients who have no improvement in clinical outcomes or patient-reported outcomes, or who have intolerable side effects (e.g., injection-site reactions, ocular side effects, and recurrent herpetic infections).

Prescribing Considerations

The clinical experts noted that it would be appropriate to limit the authority to prescribe tralokinumab to dermatologists, allergists, immunologists, and pediatricians with expertise in the diagnosis, treatment, and monitoring of patients with AD.

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 inputs received by CADTH are included in the stakeholder section at the end of this report.

Three clinician groups, the Atlantic Specialist Group Managing AD (7 clinicians), Dermatology Association of Ontario (16 clinicians), Canadian Dermatology Association (an unknown number of clinicians) provided 3 separate inputs. The 3 groups and the clinical experts consulted by CADTH agreed that the goals of therapy are to improve symptoms (long-term and durable relief of chronic itch, minimization of dry and inflamed skin, clear or almost clear skin, less oozing, scaling, cracking, or fissures), and improved QoL (better sleep) and function (ability to focus on work and school). The clinical experts added reduction in anxiety or depressive symptoms and caregiver burnout to this list. As for unmet needs, the clinician groups and the clinical experts consulted by CADTH all agreed that not all patients respond to or tolerate existing systemic treatments. JAKi treatments have safety and contraindication issues (black-box warnings for patients with risk factors for cardiovascular events, cancers, and infections), and dupilumab is associated with conjunctivitis. New treatments are therefore needed to provide more options for patients whose AD is not well controlled with existing systemic therapies. The clinician groups stated that tralokinumab would have the same place in therapy as dupilumab after phototherapy and/or off-label systemic therapies (if required by insurance or public plans) and may be trialled if patients fail to respond to dupilumab and an oral JAKi. The clinician groups said the suitable patient population aligns with the reimbursement request. They also noted that those who did not respond to biologics and/or JAKi treatments or have a history of conjunctivitis, risk factors associated with cardiovascular events, thrombosis, malignancy, serious infections, and/or significant drug-drug interactions or difficulty adhering to stricter dosing schedules, and those over the age of 65 years would be best suited for tralokinumab treatment. The clinical experts added that tralokinumab would most likely be used in patients with “pure” AD without comorbid asthma or eosinophilic esophagitis and those with special site involvement. The 3 clinician groups and the clinical experts consulted by CADTH indicated that they would assess response to treatment based on BSA affected, pruritus NRS, PGA (in clinical practice) and/or EASI, if required by an insurance company or payers, at 6 months after initiation of tralokinumab. According to the clinician groups, a lack of response or efficacy, worsening disease, deterioration of QoL, increased BSA affected, presence of AEs or unacceptable intolerance, and allergies would prompt clinicians to consider discontinuing tralokinumab treatment. Last, the clinician groups and the clinical experts agreed that a dermatologist, allergist, pediatrician, or immunologist well versed in managing moderate-to-severe AD should be allowed to prescribe tralokinumab. The 3 clinician groups raised concerns regarding differential access to tralokinumab, which is currently only funded by private insurance, and the requirement of trying off-label immunosuppressants with lower efficacy and increased risk before accessing newer systemic agents.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may affect their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 5.

Table 5: Summary of Drug Plan Input and Clinical Expert Response

AD = atopic dermatitis; CDEC = CADTH Canadian Drug Expert Committee; JAKi = Janus kinase inhibitor; q.2.w. = every 2 weeks; q.4.w. = every 4 weeks; pCPA = pan-Canadian Pharmaceutical Alliance.

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 tralokinumab 150 mg/mL solution for subcutaneous injection for the treatment of moderate-to-severe AD in adult and adolescent patients aged 12 years and older whose disease is not adequately controlled with topical prescription therapies or when those therapies are not advisable. The focus is on comparing tralokinumab to relevant comparators and identifying gaps in the current evidence.

A summary of the clinical evidence included by the sponsor in the review of tralokinumab is presented in 4 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 a sponsor-submitted LTE study. The third section includes indirect evidence from the sponsor. The fourth section includes additional studies that were considered by the sponsor to address important gaps in the pivotal and RCT evidence.

Included Studies

Clinical evidence addressed in the CADTH review and appraised in this document includes:

• Five pivotal studies or RCTs identified in systematic review (ECZTRA 1, 2, 3, 6, and 7)^9-12

• One LTE study (ECZTEND)^60-64

• Four indirect treatment comparisons (2 NMAs^19,20 and 2 MAICs^21,22)

• Two additional studies addressing gaps in evidence (Pezzolo and Naldi [2023]²³ and Pereyra-Rodriguez et al. [2023]²⁴).

Pivotal Studies and RCT Evidence

Contents within this section were informed by materials submitted by the sponsor. The following summary was validated by the CADTH review team.

Description of Studies

Characteristics of the included studies are summarized in Table 6 (adolescents) and Table 7 (adults).

Adolescents

One randomized, double-blind, placebo-controlled, pivotal phase III, 52-week trial (ECZTRA 6,⁹ N = 301) in the adolescent population met the inclusion criteria of the sponsor-conducted systematic review. The ECZTRA 6 trial evaluated the efficacy of tralokinumab compared with placebo in adolescent patients (aged 12 to less than 18 years) with moderate-to-severe AD. The trial was conducted at 72 study sites across North America, Europe, Asia, and Australia, with 11 sites located in Canada. Patients were enrolled between June 19, 2018, and November 4, 2019. The study is now complete. The study design is shown in Figure 1. It consisted of the following study periods:

• Screening phase (week −6 to week 0): Patients were assessed for study eligibility, and systemic and topical treatments for AD were washed out (TCS, TCIs, and topical PDE-4 inhibitors had a 2-week washout while off-label immunosuppressants and systemic corticosteroids had a 4-week washout). The patient or the patient’s legally acceptable representative signed and dated an informed consent form to participate in the trial.

• Initial treatment period (day 0 to week 16): Enrolled patients were randomized in a 1:1:1 ratio to receive tralokinumab 600 mg for 1 dose then 300 mg every 2 weeks, tralokinumab 300 mg for 1 dose then 150 mg every 2 weeks, or placebo every 2 weeks. Randomization was conducted using a central interactive response system and stratified by region (North America, Europe, Japan, and Australia) and baseline disease severity (IGA of 3 or 4).

• Maintenance treatment period (week 16 to week 52): Patients in the tralokinumab group who achieved either primary end point (an IGA of 0 or 1 or EASI-75) at week 16 without the use of rescue medication were considered responders and re-randomized in a 1:1 ratio to receive tralokinumab every 2 weeks or every 4 weeks at their original dose (150 mg or 300 mg). Patients receiving placebo who met at least 1 of the 2 primary end points at week 16 without the use of rescue medication continued to receive blinded placebo every 2 weeks until week 52. All other patients at week 16, and those who lost response or received rescue medication during maintenance, were transferred to open-label treatment with tralokinumab 300 mg every 2 weeks with optional use of mild-to-moderate-potency TCS or TCIs. Patients were transferred from maintenance to open-label treatment at any visit from week 16 if they met any of the following criteria and transfer was considered appropriate by the investigator:

  • IGA of 0 at week 16: IGA of at least 2 and not achieving EASI-75 over at least a 4-week period

  • IGA of 1 at week 16: IGA of at least 3 and not achieving EASI-75 over at least a 4-week period

  • IGA above 1 at week 16: not achieving EASI-75 over at least a 4-week period

  • Received rescue treatment (from week 16 or later).

• Safety follow-up period (week 52 to week 66): Safety assessments were conducted for those who did not enter the ECZTEND extension trial.

Adults

Three pivotal phase III studies (ECZTRA 1, N = 802; ECZTRA 2,¹⁰ N = 794; ECZTRA 3,¹¹ N = 380) and 1 phase IIIb study (ECZTRA 7;¹² N = 277) in the adult population met the inclusion criteria of the sponsor-conducted systematic review. These double-blind RCTs were included in the original CADTH reimbursement review of tralokinumab in adults. The ECZTRA 1 and 2 studies were designed to compare the efficacy of tralokinumab against placebo in patients with moderate-to-severe AD over 52 weeks of treatment. The ECZTRA 3 and 7 studies were designed to demonstrate that tralokinumab in combination with TCS is superior to placebo in combination with TCS; the ECZTRA 3 study assessed patients with moderate-to-severe AD over 32 weeks of treatment and the ECZTRA 7 study assessed patients with severe AD who were not adequately controlled with or had contraindications to oral cyclosporine A over 26 weeks of treatment. They were multicentre studies that included study sites in North America, Europe, Asia, and Australia; specifically, the ECZTRA 2 and 3 studies included 19 and 10 study sites, respectively, in Canada. All studies are now complete. The study designs are shown in Figure 2 (ECZTRA 1 and 2), Figure 3 (ECZTRA 3), and Figure 4 (ECZTRA 3). The studies consisted of the following study periods:

Figure 1: ECZTRA 6 Trial Design (Adolescents)

AD = atopic dermatitis; EASI75 = reduction of at least 75% in Eczema Area Severity Index score from baseline; IGA = Investigator’s Global Assessment; Q2W = every 2 weeks; Q4W = every 4 weeks; TCS = topical corticosteroids.

Note: Transfer criteria were patients with an IGA of 0 at week 16 who over 3 consecutive visits had an IGA of 2 or greater and did not achieve EASI-75; patients with an IGA of 1 at week 16 who over 3 consecutive visits had an IGA of 3 or higher and did not achieve EASI-75; patients with an IGA above 1 at week 16 who over 3 consecutive visits did not achieve EASI-75; and patients who received rescue treatment during the maintenance phase.

Source: Clinical Study Report for ECZTRA 6.¹³

ECZTRA 1 and 2

• Screening phase (2 to 6 weeks): Patients were assessed for study eligibility and systemic and topical treatments for AD were washed out (TCS, TCIs, and topical PDE-4 inhibitors had a 2-week washout while off-label immunosuppressants and systemic corticosteroids had a 4-week washout).

• Initial treatment period (day 0 to week 16): Enrolled patients were randomized in a 3:1 ratio to either tralokinumab (600 mg for 1 dose then 300 mg every 2 weeks) or placebo every 2 weeks. Randomization was conducted using a computer-generated randomization schedule and stratified by region (North America, Japan, and Europe) and baseline disease severity (an IGA of 3 or 4).

• Maintenance treatment period (week 16 to week 52): Patients in the tralokinumab group who achieved a clinical response at week 16 were re-randomized at a ratio of 2:2:1 to receive either tralokinumab every 2 weeks, tralokinumab every 4 weeks (alternating dose administrations of placebo and tralokinumab every 2 weeks), or placebo every 2 weeks. Patients who did not achieve a clinical response at week 16, as well as patients who did not maintain an adequate clinical response during the maintenance treatment period were transferred to open-label tralokinumab every 2 weeks with optional use of TCS. Patients randomized to the placebo group in the initial treatment period who achieved a clinical response at week 16 continued to receive placebo every 2 weeks in the maintenance treatment period while maintaining blinding.

• Safety follow-up period (week 52 to week 66): Safety assessments were conducted.

Figure 2: ECZTRA 1 and 2 Trial Design (Adults, Original Review)

EASI 75 = reduction of at least 75% in Eczema Area Severity Index score from baseline; IGA = Investigator’s Global Assessment; Q2W = every 2 weeks; TCS = topical corticosteroids.

Sources: Clinical Study Reports for ECZTRA 1¹⁵ and 2.¹⁶

ECZTRA 3

• Screening phase (2 to 6 weeks): Patients were assessed for study eligibility and systemic and topical treatments for AD were washed out (TCS, TCIs, and topical PDE-4 inhibitors had a 2-week washout while off-label immunosuppressants and systemic corticosteroids had a 4-week washout).

• Initial treatment period (day 0 to week 16): Enrolled patients were randomized 2:1 to receive either tralokinumab (600 mg for 1 dose then 300 mg every 2 weeks) or placebo every 2 weeks. Randomization was conducted using a central interactive voice-response system and stratified by region (Europe and North America) and baseline disease severity (an IGA of 3 or 4). All patients were instructed to use supplied TCS once daily as needed.

• Continuation treatment period (week 16 to week 32): Patients who were assigned to the tralokinumab group in the initial treatment period and had a clinical response at week 16 were re-randomized in 1:1 ratio to either tralokinumab every 2 weeks, or tralokinumab every 4 weeks (alternating dose administrations of placebo and tralokinumab every 2 weeks). Randomization was stratified by region (Europe and North America) and IGA response at week 16 (an IGA of 0 or 1, or above 1). Patients who received placebo in the initial treatment period and had a clinical response at week 16 continued to receive placebo every 2 weeks. Patients randomized to tralokinumab or placebo in the initial treatment period who did not have a clinical response at week 16 were given tralokinumab every 2 weeks in the continuation treatment period. All patients continued to stay on the TCS regimen as needed.

• Follow-up period (week 32 to week 46): All patients who did not enter the ECZTEND trial after completion of the continuation treatment period were followed for safety assessments until week 46.

ECZTRA 7

• Screening period (2 to 6 weeks): Patients were assessed for study eligibility and systemic and topical (except for TCS and TCIs) treatments for AD were washed out (topical PDE-4 inhibitors had a 2-week washout while off-label immunosuppressants and systemic corticosteroids had a 4-week washout).

• Treatment period (day 0 to week 26): Enrolled patients were randomized 1:1 to receive either tralokinumab (600 mg for 1 dose then 300 mg every 2 weeks) or placebo. Randomization was conducted using interactive response technology and stratified by prior cyclosporine A use (yes or no), country (Germany: yes or no), and baseline disease severity (an IGA of 3 or 4). All patients were instructed to use a supplied TCS once daily as needed.

• Safety follow-up period (week 27 to week 40): All patients who did not enter the ECZTEND trial after completion of the treatment period were followed for safety assessments until week 40.

Figure 3: ECZTRA 3 Trial Design (Adults, Original Review)

AD = atopic dermatitis; Q2W = every 2 weeks; Q4W = every 4 weeks; TCS = topical corticosteroids.

Note: Clinical response is defined as patients achieving an Investigator’s Global Assessment of 0 or 1 or at least 75% reduction in the Eczema Area and Severity Index.

Source: Clinical Study Report for ECZTRA 3.¹⁷

Figure 4: ECZTRA 7 Trial Design (Adults, Original Review)

AD = atopic dermatitis; q2w = every 2 weeks; TCI = topical calcineurin inhibitor; TCS = topical corticosteroids.

Source: Clinical Study Report for ECZTRA 7.¹⁸

Table 6: Study Design of ECZTRA 6 (Adolescents)

AD = atopic dermatitis; AE = adverse event; BSA = body surface area; CDLQI = Children’s Dermatology Life Quality Index; EASI = Eczema Area and Severity Index; EASI-50 = reduction of at least 50% in Eczema Area and Severity Index score from baseline; EASI-75 = reduction of at least 75% in Eczema Area and Severity Index score from baseline; EASI-90 = reduction of at least 90% in Eczema Area and Severity Index score from baseline; HADS = Hospital Anxiety and Depression Scale; IGA = Investigator’s Global Assessment; NRS = numeric rating scale; PDE-4 = phosphodiesterase type 4; POEM = Patient-Oriented Eczema Measure; q.2.w. = every 2 weeks; q.4.w. = every 4 weeks; RCT = randomized controlled trial; SCORAD = Scoring Atopic Dermatitis; SCORAD-50 = 50% decrease in Scoring Atopic Dermatitis; SCORAD-75 = 75% decrease in Scoring Atopic Dermatitis; TCI = topical calcineurin inhibitor; TCS = topical corticosteroids.

^aTralokinumab 150 mg q.2.w. and q.4.w. regimens are not of interest to this review as these regimens are not approved by Health Canada.

Sources: Clinical Study Report for ECTZRA 6¹³ and sponsor’s Summary of Clinical Evidence.¹⁴

Table 7: Study Designs of ECZTRA 1, 2, 3, and 7 (Adults, Original Review)

AD = atopic dermatitis; AE = adverse event; BSA = body surface area; DLQI = Dermatology Life Quality Index; EASI = Eczema Area and Severity Index; EASI-50 = reduction of at least 50% in Eczema Area and Severity Index score from baseline; EASI-75 = reduction of at least 75% in Eczema Area and Severity Index score from baseline; EASI-90 = reduction of at least 90% in Eczema Area and Severity Index score from baseline; HADS = Hospital Anxiety and Depression Scale; IGA = Investigator’s Global Assessment; NA = not applicable; NRS = numeric rating scale; PDE-4 = phosphodiesterase type 4; POEM = Patient-Oriented Eczema Measure; PGI-B = Patient Global Impression of Bother; PGI-S = Patient Global Impression of Severity; q.2.w. = every 2 weeks; q.4.w. = every 4 weeks; RCT = randomized controlled trial; SC = subcutaneous; SAE = serious adverse event; SCORAD = Scoring Atopic Dermatitis; SCORAD-50 = 50% decrease in Scoring Atopic Dermatitis; SCORAD-75 = 75% decrease in Scoring Atopic Dermatitis; SF-36 = Short Form (36) Health Survey; TCI = topical calcineurin inhibitor; TCS = topical corticosteroids; TSQM = Treatment Satisfaction Questionnaire for Medication; VAS = visual analogue scale; WPAI-GH = Work Productivity and Activity Impairment – General Health.

^aAt screening and baseline.

^bDuring the week before baseline.

^cPatients were permitted to continue using stable doses of such emollients if initiated before the screening visit.

^dPatients were instructed to apply a thin layer of supplied TCS once daily to areas with active lesions as needed in the treatment period.

Sources: Clinical Study Reports for ECZTRA 1,¹⁵ 2,¹⁶ 3,¹⁷ and 7,¹⁸ CADTH Reimbursement Review for Adtralza (2022),⁶⁵ and sponsor’s Summary of Clinical Evidence.¹⁴

Populations

Key inclusion and exclusion criteria of included studies are summarized in Table 6 (adolescents) and Table 7 (adults).

Inclusion and Exclusion Criteria

Adolescents

The ECZTRA 6 study included adolescents aged 12 to 17 years who were diagnosed with AD for at least 1 year and had an inadequate response or intolerance to TCS and/or TCIs, or for whom topical medications were medically inadvisable. Patients were required to have an EASI of at least 12 at screening and at least 16 at baseline, an IGA score of at least 3 at screening and at baseline, AD involvement of at least 10% of BSA, and an adolescent pruritus NRS average score of at least 4 during the week before baseline. Patients who had been treated with a topical PDE-4 inhibitor, TCS, or TCI within 2 weeks prior, or used systemic immunosuppressive or immunomodulating drugs, systemic corticosteroids, or 3 or more bleach baths within the past 4 weeks were excluded, as were those who had used tanning beds or phototherapy within the preceding 6 weeks.

Adults

The inclusion and exclusion criteria of the ECZTRA 1, 2, 3, and 7 studies were the same as for the ECZTRA 6 study, except that they were conducted in adult patients. Inadequate response was explicitly defined as failure to achieve and maintain remission or a low disease-activity state (comparable to an IGA of 0 [clear] to 2 [mild]) despite treatment with a daily regimen of TCS of medium to high potency. As well, the pruritus NRS average score was used to assess the severity of pruritus (as opposed to adolescent pruritis NRS in the ECZTRA 6 study). Patients enrolled in the ECZTRA 7 study had a history of inadequate response, intolerance, or unacceptable toxicity to cyclosporine A, or were not deemed to be candidates for cyclosporine A treatment and an EASI of at least 20 at screening and baseline (as opposed to an EASI of at least 12 at screening and at least 16 at baseline in other studies), and were allowed to use TCS and TCIs within 2 weeks before enrolment (which was not permitted in other studies).

Interventions

Adolescents

Initial Treatment Period

In the initial treatment period of the ECZTRA 6 study, patients were randomly assigned to receive 1 of the following until week 16:

• Tralokinumab 600 mg (i.e., 4 injections of 150 mg, 1 mL each) subcutaneously on day 0, followed by tralokinumab 300 mg (i.e., 2 injections of 150 mg, 1 mL each) subcutaneously every 2 weeks

• Tralokinumab 300 mg (i.e., 2 mL) and placebo 2 mL subcutaneously on day 0, followed by tralokinumab 150 mg (1 mL) and placebo (1 mL) subcutaneously every 2 weeks

• Placebo 4 mL on day 0 then 2 mL subcutaneously every 2 weeks.

Maintenance Treatment Period

In the maintenance treatment period of the ECZTRA 6 study, patients who achieved a clinical response at week 16 (defined as an IGA of 0 or 1 or EASI-75 without the use of rescue treatment) while receiving tralokinumab in the initial treatment period were re-randomized to receive tralokinumab every 2 weeks or every 4 weeks at the same dose as in the initial treatment period until week 52.

Patients initially randomized to tralokinumab 300 mg every 2 weeks were re-randomized to either tralokinumab 300 mg every 2 weeks in a 2 mL injection or tralokinumab 300 mg every 4 weeks in alternating 2 mL injections of 300 mg of tralokinumab and placebo.

Among patients initially randomized to tralokinumab 150 mg, patients re-randomized to tralokinumab 150 mg every 2 weeks received tralokinumab 150 mg (1 mL) + placebo (1 mL). Patients re-randomized to tralokinumab 150 mg every 4 weeks received alternating 1 mL doses of tralokinumab 150 mg and placebo or a 2 mL dose of placebo.

Patients receiving placebo every 2 weeks who met the primary end points at week 16 without the use of rescue medication continued to receive placebo every 2 weeks until week 52.

Patients who transferred to open-label treatment received tralokinumab 300 mg (2 mL) every 2 weeks with optional use of TCS or TCIs of mild-to-moderate potency.

Blinding

Tralokinumab and placebo were visually distinct. All patients, study investigators, and study personnel who were involved in the treatment, clinical evaluation, and monitoring of patients were blinded from treatment assignment in the treatment period of all studies. All treatment groups received the same number of injections at each visit. The packaging and labelling of tralokinumab and placebo contained no evidence of their identity. The interventions were administered by a health care provider, the patient, or the patient’s caregiver (open-label treatment only).

Rescue Medications and Concomitant Medications

All patients were to use an emollient twice daily (or more often, as needed) for at least 14 days before randomization and continue this treatment throughout the trial. Rescue treatment for AD could be provided to patients at the discretion of the investigator. Investigators were encouraged to try rescue topical treatments first for at least 14 days before escalating to systemic medications. TCS of any WHO class could be used as topical rescue treatment (it was unclear if any restrictions were imposed on frequency or dose of topical rescue treatment). Systemic rescue treatment with corticosteroids or nonsteroidal immunosuppressive drugs (e.g., cyclosporine A or methotrexate) required immediate discontinuation of tralokinumab. After the treatment with these drugs was completed, tralokinumab could be resumed if deemed appropriate by the investigator and sponsor’s medical expert. Use of biological rescue treatment was not allowed for the entire trial duration. The following concomitant medications related to AD treatment were permitted from screening through safety follow-up: oral antibiotics, antiviral, or antifungal therapy for skin infections, and oral antihistamines.

Adults

Initial Treatment Period

In the initial treatment period of the ECZTRA 1 and 2 trials, patients were randomized to receive either of the following until week 16:

• tralokinumab 600 mg (i.e., 4 injections of 150 mg, 1 mL each) subcutaneously on day 0, followed by tralokinumab 300 mg (i.e., 2 injections of 150 mg, 1 mL each) subcutaneously every 2 weeks until week 16

• placebo 4 mL on day 0, and then 2 mL every 2 weeks.

The same interventions were also used in the initial treatment period of the ECZTRA 3 study and the treatment period of the ECZTRA 7 study. In addition, all patients in the ECZTRA 3 and 7 studies were instructed to initiate treatment once daily with a supplied TCS (mometasone furoate 0.1% cream) on active lesions as needed throughout the treatment period. In the ECZTRA 7 study, patients received these interventions until week 26.

Maintenance (or Continuous) Treatment Period (ECZTRA 1, 2, and 3)

In the maintenance treatment period of ECZTRA 1 and ECZTRA 2, patients randomized to tralokinumab in the initial treatment period who had a clinical response (i.e., those who were in tralokinumab group during the initiation treatment period and had a clinical response, defined as an IGA of 0 or 1 or EASI-75 with rescue medication use) at week 16 were re-randomized to receive 1 of the following until week 52:

• tralokinumab 300 mg (2 mL) every 2 weeks

• tralokinumab 300 mg every 4 weeks (alternating dose administrations of 300 mg tralokinumab and 2 mL placebo every 2 weeks to maintain blinding)

• placebo (2 mL) every 2 weeks.

In the continuous treatment period of the ECZTRA 3 study, patients who had clinical response to tralokinumab at week 16 were re-randomized to receive either of the following until week 32:

• tralokinumab 300 mg (2 mL) every 2 weeks in combination with an TCS on an as-needed basis

• tralokinumab 300 mg (2 mL) every 4 weeks, in combination with an TCS on an as-needed basis.

In the ECZTRA 1, 2, and 3 studies, patients randomized to the placebo group in the initial treatment period who achieved a clinical response at week 16 continued to receive placebo every 2 weeks until the end of the maintenance treatment period. Patients who did not achieve a clinical response to tralokinumab at week 16 (as well as patients who did not maintain an adequate clinical response during the maintenance treatment period of ECZTRA 1 and 2) were transferred to tralokinumab every 2 weeks with optional use of TCS.

The blinding procedures in these trials were the same as those in the ECZTRA 6 study. The protocol for rescue and concomitant medication use was the same as in the ECZTRA 6 study, except that in the ECZTRA 3 and 7 studies, only higher-potency TCS (Europe Class above 3 or below 4) could be used as topical rescue treatment.

Outcomes

A list of efficacy end points assessed in this clinical review report is provided in Table 8 and summarized in the text that follows. Summarized end points are based on those included in the sponsor’s Summary of Clinical Evidence as well as any identified as important to this review according to stakeholders (e.g., clinical experts, clinician groups, or patient groups).

Unless otherwise specified, the baseline value was defined as the latest predose assessment.

Investigator’s Global Assessment

The IGA score is an investigator-reported assessment used to rate the severity of AD.¹⁵ It is based on a 5-point scale ranging from 0 to 4, in which “0” indicates clear, and “4” indicates severe AD. Evidence for validity and reliability of this instrument is summarized in Table 9. No MID was identified in adult or adolescent patients with AD.¹⁵

Table 8: Outcomes Summarized from ECZTRA 1, 2, 3, and 7

AD = atopic dermatitis; AE = adverse event; CDLQI = Children’s Dermatology Life Quality Index; DLQI = Dermatology Life Quality Index; EASI = Eczema Area and Severity Index; EASI-50 = reduction of at least 50% in Eczema Area and Severity Index score from baseline; EASI-75 = reduction of at least 75% in Eczema Area and Severity Index score from baseline; EASI-90 = reduction of at least 90% in Eczema Area and Severity Index score from baseline; HADS = Hospital Anxiety and Depression Scale; IGA = Investigator’s Global Assessment; NA = not applicable; NRS = numeric rating scale; POEM = Patient-Oriented Eczema Measure; SAE = serious adverse event; SCORAD = Scoring Atopic Dermatitis; TCS = topical corticosteroids; WDAE = withdrawal due to adverse event.

^aStatistical testing for these end points was adjusted for multiple comparisons (e.g., hierarchal testing).

^bAdolescent worst pruritis NRS, an age-appropriate version of the worst daily pruritus NRS, was used to assess itch in adolescents in the ECZTRA 6 study.

^cCDLQI was used in the adolescent population in the ECZTRA 6 study.

Sources: Clinical Study Reports for ECZTRA 1,¹⁵ 2,¹⁶ 3,¹⁷ 6,¹³ and 7¹⁸ and sponsor’s Summary of Clinical Evidence.¹⁴

Eczema Area and Severity Index

The EASI is an investigator-reported scale that is used to assess the severity and extent of AD.¹⁵ The severity of 4 disease characteristics of AD (erythema, infiltration and/or papulation, excoriations, and lichenification) are assessed on a 4-point scale ranging from “0” (none or absent) to “3” (severe). The total EASI score ranges from 0 to 72 points, with higher values indicating more severe and/or more extensive condition. EASI-50, EASI-75, and EASI-90 indicate improvements of at least 50%, 75%, and 90% from baseline in the EASI score, respectively.¹⁵ Evidence for validity, reliability and responsiveness of this instrument is summarized in Table 9. The MID was estimated to be 6.6 in a study that included adults with AD.⁶⁶ No MID for adolescents was identified.

Scoring Atopic Dermatitis

Scoring Atopic Dermatitis (SCORAD) is an investigator-administered tool to measure the extent and severity of AD lesions, along with subjective symptoms. The instrument assesses 3 components of AD: the extent of affected BSA, severity, and symptoms. The total SCORAD is calculated based on the 3 components, with a maximum possible total score of 103, in which a higher score indicates more severe disease. Evidence for validity and reliability, and responsiveness of this instrument are summarized in Table 9. The MID was estimated to be 8.7 points in a study that included adults and children with AD.⁶⁶

Worst Pruritus Numeric Rating Score and Adolescent Worst Pruritus Numeric Rating Score

The pruritus NRS is a tool used for patients to self-report the worst itch over the past 24 hours using an 11-point scale, with 0 indicating “no itch” and 10 indicating “worst itch possible.” Evidence for reliability, validity, and responsiveness in patients with AD is summarized in Table 9. A study by Simpson et al. (2017), estimated that improvement on the pruritus NRS of at least 3 to 4 points from baseline is a clinically meaningful change, which was calculated using anchor- and distribution-based methods based on data from a study in adults with moderate-to-severe AD.⁶⁷ Another study, by Yosipovitch et al. (2019), estimated the MID to be between 2 and 4 points based on anchors (EASI, IGA, and Pruritus Categorical Scale), and 1.0 based on distribution methods.⁶⁸

The adolescent worst pruritis NRS tool is an age-appropriate version of the worst daily pruritus NRS tool developed for adults. Evidence for reliability, validity, responsiveness, and MID of adolescent worst pruritus NRS in patients with AD is not available.

Eczema-Related Sleep Numeric Rating Scale

The eczema-related sleep NRS is used by the patient to rate how much their eczema had interfered with their sleep the last night using an 11-point NRS, with 0 indicating that it “did not interfere” and 10 indicating that it “completely interfered.”⁶⁹ Evidence for validity is summarized in Table 9. No MID estimate was identified in adult or adolescent patients with AD.

Patient-Oriented Eczema Measure

The POEM is a 7-item, AD-specific, symptom questionnaire.⁷⁰ Based on the self-reported frequency of occurrence during the past week, the 7 items (itching, sleep, bleeding, weeping, cracking, flaking, and dryness) are assessed using a 5-point scale (0 = “no days,” 1 = “1 to 2 days,” 2 = “3 to 4 days,” 3 = “5 to 6 days,” and 4 = “every day”). The maximum total score is 28; with a higher score indicative of worse disease severity.⁷⁰ Evidence for validity, reliability, and responsiveness of the POEM is summarized in Table 9. The MID has been established in AD as 3.4 points in adults⁶⁶ and from 3.0 to 3.9 points in children.^71,72 Another study established 5 points as the MID for adults using global severity of AD as anchor.⁷⁰

Dermatology Life Quality Index and Children Dermatology Life Quality Index

The DLQI is a widely used dermatology-specific HRQoL instrument for use in adults. It is a 10-item questionnaire in which patients self-report 6 different aspects that may affect QoL (symptoms and feelings, daily activities, leisure, work and school performance, personal relationships, and treatment).¹⁵ The DLQI score ranges between 0 and 30. The higher the score, the greater QoL is impaired.¹⁵ Evidence for reliability, validity, and responsiveness of the DLQI instrument in patients with moderate-to-severe AD was not identified. Estimates of the MID for a variety of skin conditions range from 2.2 to 6.9 points, but no information about MID had been found for adults with AD.^73-75

The CDLQI questionnaire is based on the adult version (DLQI) and is designed and validated in patients with dermatological conditions aged 3 to 16.¹³ The CDLQI is available in text and cartoon versions; the text version was used in the ECZTRA 6 study. The questionnaire consists of 10 items addressing the patient’s perception of the impact of their skin disease on various aspects of their QoL over the last week, including dermatology-related symptoms and feelings, leisure, school, friendships, sleep, and the impact of treatment. The total score ranges from 0 to 30, with a higher score indicative of a poor QoL.¹³ Evidence for validity, reliability, and responsiveness of this instrument is summarized in Table 9. In adolescents, a reduction in the CDLQI of 6 to 8 points has been suggested as the clinically relevant threshold for within-person change in CDLQI in moderate-to-severe AD.⁷⁶

Hospital Anxiety and Depression Scale

The HADS is a widely used patient-reported questionnaire designed to identify anxiety disorders and depression.⁸⁸ The HADS questionnaire contains 14 items that are used to assess symptoms experienced in the previous week, among which 7 items are related to anxiety and 7 items are related to depression. The score ranges between 0 and 21 for each subscale (anxiety and depression); a high score is indicative of a poor state.⁸⁸ Evidence for reliability, validity, responsiveness, and MID of HADS in adult or adolescent patients with AD is not available.

Harms Outcomes

Harms outcomes were assessed as secondary end points. This includes AEs, SAEs, withdrawal due to AEs, and death, as well as notable harms of interest to this review, including eye disorder (e.g., conjunctivitis), eczema herpeticum, malignancies, skin infections requiring systemic treatment, injection-site reactions, oral herpes, upper respiratory infection, and acne.

Table 9: Summary of Outcome Measures and Their Measurement Properties

AD = atopic dermatitis; AUC = area under the curve; CDLQI = Children’s Dermatology Life Quality Index; CI = confidence interval; DLQI = Dermatology Life Quality Index; EASI = Eczema Area and Severity Index; EASI-50 = reduction of at least 50% in Eczema Area and Severity Index score from baseline; EASI-75 = reduction of at least 75% in Eczema Area and Severity Index score from baseline; EASI-90 = reduction of at least 90% in Eczema Area and Severity Index score from baseline; HADS = Hospital Anxiety and Depression Scale; ICC = intraclass correlation coefficient; IGA = Investigator’s Global Assessment; MID = minimal important difference; NRS = numeric rating scale; OSAAD = Objective Severity Assessment of Atopic Dermatitis; PCS = Pruritus Categorical Scale; PGADS = Patient Global Assessment of Disease Status; POEM = Patient-Oriented Eczema Measure; QoL = quality of life; SCORAD = Scoring Atopic Dermatitis; SD = standard deviation; VAS = visual analogue scale.

^aIGA categories:

0 (clear): No inflammatory signs of AD; No erythema and no elevation (papulation and/or infiltration).

1 (almost clear): Just perceptible erythema, and just perceptible papulation and/or infiltration; barely perceptible erythema and/or minimal lesion elevation (papulation and/or infiltration) that is not widespread.

2 (mild disease): Mild erythema and mild papulation and/or infiltration; visibly detectable, light pink erythema and very slight elevation (papulation and/or infiltration).

3 (moderate disease): Moderate erythema and moderate papulation and/or infiltration; dull red, clearly distinguishable erythema and clearly perceptible but not extensive elevation (papulation and/or infiltration).

4 (severe disease): Severe erythema and severe papulation and/or infiltration; deep or dark-red erythema, marked and extensive elevation (papulation and/or infiltration).

^bThe EASI is a composite index, including an assessment of disease extent and percent of BSA involved, converted to a proportional factor (scale of 0 to 6), in 4 body regions (head and neck, lower limbs, upper limbs, and trunk). The proportion allocated to each body region depends on the patient’s age. In patients aged 8 years or older, the proportions are 10% for head and neck, 20% for upper extremities, 30% for trunk, and 40% for lower extremities; in patients aged 7 years or younger, the proportions are 20% for head and neck, 20% for upper extremities, 30% for trunk, and 30% for lower extremities. The EASI also includes an assessment of erythema (E), infiltration and/or papulation (I), excoriation (Ex) and lichenification (L), each on a scale of 0 to 3. The algorithm for calculating the EASI uses, for each body region, the sum of the clinical sign scores (E + I + Ex + L) multiplied by the area, multiplied by the proportional factor. The total EASI score is the sum of the 4 body-region scores.

^cThe SCORAD is calculated as: (A)/5 + 7 × (B)/2 + (C):

Extent (A) is assessed as a percentage of each defined body area and reported as the sum of all areas, with a maximum score of 100%.

Intensity (B) of 6 specific symptoms of AD (erythema, edema and/or papulation, oozing and/or crusting, excoriation, lichenification, and dryness) is assessed on an average representative area using a 4-point scale ranging from 0 (none or absent) to 3 (severe). The sum of the intensity score of the 6 symptoms is reported, with a maximum score of 18.

Subjective symptoms (C) involve a subjective assessment of the average itch and sleeplessness over the last 3 days or nights, recorded for each symptom by the patient on a VAS ranging from 0 (no itch or no sleeplessness) to 10 (worst imaginable itch or sleeplessness), with a maximum combined possible score of 20.

^dThe text version was used in the ECZTRA 6 study.

Sources: Sponsor’s clinical evidence summary, ECZTRA 6 Clinical Study Report,¹³ Bozek and Reich (2017),⁷⁸ Barbier et al. (2004),⁷⁸ Hanifin et al. (2001),⁷⁹ Schram et al. (2012),⁶⁶ Schmitt et al. (2013),⁸⁰ Rehal and Armstrong (2011),⁸¹ Lewis-Jones and Finlay (1995),⁸⁶ Salek et al. (2013),⁸⁷ Simpson et al. (2019),⁷⁶ Yosipovitch et al. (2019),⁶⁸ Simpson et al. (2017),⁵¹ Dias-Barbosa et al. (2020),⁶⁹ Silverberg et al. (2020),⁷⁰ Howells et al. (2018),⁷² Zigmond and Snaith (1983),⁸⁸ Basra et al. (2008),⁸³ Basra et al. (2015),⁷⁴ Shikiar et al. (2005),⁷³ Heinl et al. (2016),⁷⁵ Lewis and Finlay (2004),⁸⁴ and Badia et al. (1999).⁸⁵

Statistical Analysis

Statistical analysis methods of the primary estimand of efficacy end points are summarized in Table 10.

Sample-Size Considerations

ECZTRA 1 and 2: A sample-size calculation determined that approximately 780 randomized patients were required (3:1 ratio, i.e., 585 patients to tralokinumab and 195 patients to placebo) to demonstrate a statistically significant difference between tralokinumab and placebo with respect to both primary end points (an IGA of 0 or 1 at week 16 and EASI-75 at week 16) at a 2-sided significance level of 0.05 with over 99% power, assuming the IGA of 0 or 1 response rates in tralokinumab and placebo groups were 30% and 10%, respectively, and EASI-75 response rates in tralokinumab and placebo groups were an IGA 0 or 1, and 40% and 15%, respectively.

With an IGA response rate of 30% at week 16, 175 IGA responders initially treated with tralokinumab were expected to be re-randomized into the maintenance treatment period (2:2:1 ratio, i.e., 70 patients on tralokinumab every 2 weeks, 70 patients on tralokinumab every 4 weeks, and 35 patients on placebo). Assuming IGA response rates at week 52 of 80%, 50%, and 5%, respectively, for tralokinumab every 2 weeks, tralokinumab every 4 weeks, and placebo, the nominal power to show a difference at the 4% significance level would be greater than 99% between tralokinumab every 2 weeks and placebo, and greater than 99% between tralokinumab every 4 weeks and placebo.

With an EASI-75 response rate of 40% at week 16, 235 EASI-75 responders initially treated with tralokinumab were expected to enter the maintenance treatment period (94 patients on tralokinumab every 2 weeks, 94 patients on tralokinumab every 4 weeks, and 47 patients on placebo). Assuming EASI-75 response rates at week 52 of 90%, 55%, and 5%, respectively, for tralokinumab every 2 weeks, tralokinumab every 4 weeks, and placebo, the nominal power to show a difference at the 4% significance level would be greater than 99% between tralokinumab every 2 weeks and placebo and greater than 99% between tralokinumab every 4 weeks and placebo.

ECZTRA 3: Approximately 369 randomized patients were required in the initial treatment period (2:1 ratio, 246 patients to tralokinumab every 2 weeks plus TCS and 123 patients to placebo plus TCS) to demonstrate a statistically significant difference between tralokinumab plus TCS and placebo plus TCS with respect to an IGA of 0 or 1 at week 16 (primary end point) at a 2-sided significance level of 0.05 with 90% power, assuming the IGA of 0 or 1 response rates in tralokinumab plus TCS and placebo plus TCS groups were 30% and 15%, respectively. For the primary end point of EASI-75 at week 16, a sample size of 369 patients provided a power exceeding 99.9% to detect a difference between the 2 groups, assuming response rates of 40% and 15%, respectively.

ECZTRA 6: Approximately 294 randomized patients were required in the initial treatment period (1:1:1 ratio, i.e., 98 patients to tralokinumab 300 mg, 98 patients to tralokinumab 150 mg, and 98 patients to placebo) to demonstrate a statistically significant difference between tralokinumab 300 mg and placebo with respect to an IGA of 0 or 1 at week 16 (primary end point) at a 2-sided significance level of 0.05 with approximately 94% power, assuming response rates for tralokinumab 300 mg and placebo groups of 30% and 10%, respectively. For the primary end point of EASI-75 at week 16, a sample size of 294 patients would provide a power of approximately 98% to detect a difference between tralokinumab 300 mg and placebo, assuming response rates of 40% and 15%, respectively.

ECZTRA 7: A sample size of 250 patients would provide 99% power to detect a treatment difference for the primary end point at a significance level of 0.05, assuming EASI-75 response rates at week 16 of 40% and 15% for tralokinumab plus TCS and placebo plus TCS, respectively. Assuming a response rate of 30% versus 15% in reduction of worst daily pruritus NRS (weekly average) score of at least 4 from baseline to week 16 for tralokinumab plus TCS and placebo plus TCS, respectively, the sample size would provide at least 80% power to reject the hypotheses related to the primary and secondary end point evaluating pruritus at week 16.

Multiplicity Adjustment

ECZTRA 1 and 2: To control the overall type I error rate, the primary analyses for the primary and secondary end points for the initial and maintenance treatment followed the testing procedures shown in Figure 5. IGAs of 0 or 1 at week 16, followed by EASI-75 at week 16, were evaluated at a 5% significance level. If both tests were statistically significant, the significance level (alpha) was split between the analyses of the 3 secondary end points at week 16 (alpha = 1%) and the analyses of the 2 maintenance end points at week 52 (alpha = 4%). A different testing procedure was used for the US regulatory submission but will not be summarized in this report.

Figure 5: Testing Procedures for Primary, Secondary, and Maintenance End Points in ECZTRA 1 and 2 (Global, Non-US Submission)

DLQI = Dermatology Life Quality Index; EASI75 = reduction of at least 75% in Eczema Area and Severity Index score from baseline; IGA = Investigator’s Global Assessment; SCORAD = Scoring Atopic Dermatitis; Q2W = every 2 weeks; Q4W = every 4 weeks.

Source: Clinical Study Reports for ECZTRA 1¹⁵ and 2.¹⁶

ECZTRA 3: The overall type I error rate for the primary analysis of the primary estimands for the primary and confirmatory secondary end points was controlled by a combination of hierarchical testing and Holm-Bonferroni multiplicity adjustment as outlined in Figure 6. The testing procedure was similar to that in the ECZTRA 1 and 2 studies, except that the maintenance treatment end points were not included in the testing hierarchy in the ECZTRA 3 study. A different testing procedure was used for the US regulatory submission but will not be summarized in this report.

Figure 6: Testing Procedures for Primary and Confirmatory Secondary End Points in ECZTRA 3 (Global, Non-US Submission)

DLQI = Dermatology Life Quality Index; EASI75 = reduction of at least 75% in Eczema Area and Severity Index score from baseline; IGA = Investigator’s Global Assessment; SCORAD = Scoring Atopic Dermatitis.

Note: Arrows indicate order of testing when the hypothesis is rejected for an end point within a box.

Source: Clinical Study Report for ECZTRA 3.¹⁷

ECZTRA 6: To control the overall type I error rate, the primary analyses of the primary estimands for the primary and confirmatory secondary end points followed the testing procedures outlined in Figure 7. An IGA of 0 or 1 at week 16, followed by EASI-75 at week 16, between tralokinumab 300 mg every 2 weeks and placebo were evaluated at a 5% significance level. If both tests were significant, the significance level (alpha) was split between the analyses of the 3 secondary end points between tralokinumab 300 mg every 2 weeks and placebo at week 16 (alpha = 2.5%), and the analyses of the 2 primary end points and 3 secondary end points between tralokinumab 150 mg every 2 weeks and placebo at week 16 (alpha = 2.5%). A different testing procedure was used for the US regulatory submission but it will not be summarized in this report.

Figure 7: Testing Procedure for Primary and Secondary End Points in ECZTRA 6 (Global, Non-US Submission)

CDLQI = Children’s Dermatology Life Quality Index; EASI75 = reduction of at least 75% in Eczema Area and Severity Index score from baseline; IGA = Investigator’s Global Assessment; SCORAD = Scoring Atopic Dermatitis.

Notes: The numbers in parentheses indicate significance levels that have been passed on from rejected hypotheses for the other tralokinumab dose levels.

Source: Clinical Study Report for ECZTRA 6.¹³

ECZTRA 7: The primary and secondary end points were evaluated hierarchically as shown in Figure 8. The hypothesis relating to a specific end point was rejected only if all hypotheses relating to end points earlier in the hierarchy were also rejected at the 5% significance level. Hypothesis testing was based on the primary analysis of the primary estimand for each associated end point.

Statistical Analysis for Primary Efficacy End Points

ECZTRA 1, 2, 3, and 6

Statistical model: The difference in response rates between treatment groups with respect to the 2 primary outcomes was analyzed at a 2-sided significance level of 5% using the Cochran-Mantel-Haenszel test (single imputation analyses) or using combined inference from multiple Mantel-Haenszel risk differences and associated standard errors using Rubin’s rule (multiple imputation analyses). The stratification factors included region and baseline disease severity. The primary estimand (composite strategy) assessed treatment differences in response rates of IGA 0 or 1 and EASI-75 after 16 weeks achieved without rescue medication, regardless of treatment discontinuation. Patients who had received rescue medication before the week 16 visit were considered nonresponders. Patients with missing data at week 16 and where rescue medication had not been used before week 16 were imputed as nonresponders.

Figure 8: Testing Hierarchy for Primary and Secondary End Points in ECZTRA 7

Alpha = statistical significance level; DLQI = dermatology life quality index; EASI75 = reduction of at least 75% in Eczema Area and Severity Index score from baseline; IGA = investigator’s global assessment; NRS = numeric rating scale; SCORAD = scoring atopic dermatitis.

Source: Clinical Study Report for ECZTRA 7.¹⁸

Sensitivity analyses: Three sensitivity analyses were conducted:

• All patients who permanently discontinued the investigational medicinal product before week 16 were considered nonresponders, even if no rescue medication had been used.

• Missing data at week 16 was imputed using the LOCF rather than nonresponder imputation for patients who did not receive rescue medication and did not withdraw due to an AE or lack of efficacy.

• Tipping-point analysis using multiple imputations with patients who, before the week 16 visit, had received rescue medication and were considered nonresponders. Missing week 16 responses were imputed from a Bernoulli distribution with a varying parameter, P, for patients in the placebo group who did not use rescue medication. Patients in the tralokinumab group with missing week 16 data were imputed as nonresponders. Different percentages of placebo patients were considered responders for the different P values. The tipping point is the P value that changed the conclusion from significant to nonsignificant.

Subgroup analyses: According to the clinical experts, this review is interested in 1 subgroup analysis: baseline IGA.

Two additional estimands were defined, including:

• Secondary hypothetical estimand: All values were censored after permanent discontinuation of the investigational product or initiation of rescue medication, and multiple imputations of missing values were applied within each treatment group, assuming data were missing at random.

• Tertiary treatment policy estimand: All data were used as observed, and multiple imputation of missing values was applied, assuming data were missing at random within the 4 groups defined according to the randomized treatment group and whether patients permanently discontinued the investigational product before week 16.

ECZTRA 7

Statistical model: The difference in EASI-75 between treatment groups was analyzed at a 2-sided significance level of 5% using the Cochran-Mantel-Haenszel test (single imputation analyses) or multiple Mantel-Haenszel risk differences and associated standard errors to allow for a combined inference using Rubin’s rule (multiple imputation analyses). The stratification factors included prior cyclosporine A use (yes or no) and baseline disease severity (IGA of 3 or 4). The primary estimand (COVID-19–modified composite) assessed treatment difference in response rates of EASI-75 after 16 weeks achieved without rescue medication or treatment discontinuation, as if the COVID-19 pandemic had not happened. Patients who received rescue treatment or permanently discontinued the investigational product before the week 16 visit, without prior patient-onset of COVID-19, were considered nonresponders at all visits after the relevant event occurred. Any missing or collected data from patients who had patient-onset of COVID-19 as their first prior intercurrent event were handled differently and instead were imputed as missing at random following the start of patient-onset of COVID-19. Data missing before any intercurrent event were handled as nonresponses, unless data were missing due to the COVID-19 pandemic, in which case they were imputed as missing at random.

Sensitivity analysis: To examine the deviations from the missing-at-random assumption in the primary analysis, a tipping-point analysis, which assumed data were not missing at random, was performed.

Subgroup analyses: The following subgroup analyses were of interest to this review: prior cyclosporine A use (yes or no) and disease severity at baseline (IGA of 3 or 4). Interaction between subgroups and treatment effect were tested using a conditional logistic regression model.

Three additional estimands were defined for the primary end point, including:

• secondary estimand (“composite”), which assessed the treatment difference in response rates of EASI-75 after 16 weeks achieved without either rescue treatment or treatment discontinuation, regardless of the occurrence of the COVID-19 pandemic

• tertiary estimand (“treatment policy”), which assessed the treatment difference in response rate of EASI-75 after 16 weeks between tralokinumab plus TCS and placebo plus TCS regardless of rescue treatment and IMP discontinuation, as if the COVID-19 pandemic did not happen

• quaternary estimand (“hypothetical”), which assessed the treatment difference in response rates of EASI-75 after 16 weeks if all patients adhered to the treatment regimen in the sense that they did not discontinue the investigational product permanently, no rescue treatment was prescribed, and as if the COVID-19 pandemic did not happen before week 16.

Statistical Analysis for Secondary Efficacy End Points

ECZTRA 1, 2, 3, and 6

Statistical model: Binary secondary outcomes were conducted as described for the primary end points. In the primary estimand (hypothetical strategy) of continuous secondary outcomes, data collected after permanent discontinuation of the investigational product or after initiation of rescue medication were excluded. Analyses were conducted using a repeated measurements model as follows: change in measurement value = (treatment × week) + (baseline value × week) + region + baseline IGA. For patients for whom no postbaseline data were collected before initiation of rescue medication, the week 2 change was imputed as 0. Unless otherwise stated, all significance tests were 2-sided using a 5% significance level.

Sensitivity analysis: Multiple imputations of missing values were applied based on regression models fitted on observed data from the placebo group. The following analysis of covariance (ANCOVA) model at week 16 was used: change in measurement value = treatment + baseline value + region + baseline IGA. Estimates and standard errors from analyses of multiple imputed datasets were combined using Rubin’s rule.

Two additional estimands were defined for the key continuous secondary end points, including the treatment policy estimand and the composite policy estimand. The treatment policy estimand was as described in the preceding section for the primary end points. In the composite policy estimand, patients who had received rescue medication before the week 16 visit were considered nonresponders by using the worst observation carried forward and multiple imputation of missing values was applied using the ANCOVA model.

ECZTRA 7

Statistical model: Binary secondary outcomes were analyzed as described for the primary end points. In the primary estimand (hypothetical strategy) of continuous secondary outcomes, data collected after permanent discontinuation of the investigational product, after initiation of rescue treatment, or after patient-onset of COVID-19 were excluded. Analyses were conducted using a repeated measurements model as follows: change in measurement value = (treatment × week) + (baseline value × week) + prior cyclosporine A use + country + baseline IGA, assuming the data were missing at random.

Sensitivity analysis: Imputation of missing data at week 16 and week 26 was performed using a pattern-mixture model in which missing data in the tralokinumab plus TCS group as well as the placebo plus TCS group were imputed using data from the placebo plus TCS group and the so-called copy-reference approach. In the ANCOVA model, change from baseline in measurement value = treatment + baseline value + prior cyclosporine A use + country + baseline IGA. The estimates and standard errors from analyses of multiple imputed datasets were combined using Rubin’s rule.

Subgroup analysis: Subgroup analyses for the key secondary end points were the same as described for the primary end point.

Two additional estimands were defined for the key continuous secondary end points using the treatment policy strategy and the COVID-19–modified composite strategy.

Statistical Analysis for Maintenance Efficacy End Points

ECZTRA 1, 2, and 6

Statistical model: Two binary maintenance end points were assessed: an IGA of 0 or 1 at week 52 among patients with an IGA of 0 or 1 at week 16 achieved without rescue medication after initial randomization to tralokinumab, and EASI-75 at week 52 among patients with EASI-75 at week 16 achieved without rescue medication after initial randomization to tralokinumab. The difference in response rates between treatment groups was analyzed using the Cochran-Mantel-Haenszel test stratified by region in the ECZTRA 1 and 2 studies, a binomial model (providing response rates), and corresponding 95% CIs based on the Wilson score method in the ECZTRA 6 study. All patients who had received rescue medication before the week 52 visit, including TCS, and/or those who had been transferred to open-label treatment with tralokinumab, were considered nonresponders. Patients with missing data at week 52 were imputed as nonresponders.

Sensitivity analysis: Data missing at week 52 for patients who did not receive rescue medication did not transfer to open-label treatment, and did not withdraw from the trial due to AE or lack of efficacy were imputed using LOCF instead of nonresponder imputation.

ECZTRA 3

Descriptive statistics were used for end points in the continuation treatment period (an IGA of 0 or 1 and EASI-75 at week 32).

Analysis Populations

Analysis Populations of the ECZTRA 1, 2, 3, 6, and 7 studies are summarized in Table 11.

Results

This section summarizes the results of the pivotal and RCT evidence identified by the sponsor. The results of the tralokinumab 150 mg groups (every 2 weeks and every 4 weeks) in the ECZTRA 6 study are not of interest to this review and are not presented in this report as the dosing does not align with the recommended dosing stated in the product monograph.

Patient Disposition

Patient disposition in the ECZTRA 6 study (adolescents) is presented in Table 12 (initial treatment period) and Table 13 (maintenance treatment period). Patient disposition in the ECZTRA 1, 2, 3, and 7 (adults) studies is presented in Table 14 (initial treatment period), Table 15 (maintenance treatment period in the ECZTRA 1 and 2 studies), and Table 16 (continuous treatment period in the ECZTRA 3 study).

Table 10: Statistical Analysis of Efficacy End Points (Primary Estimand)

AE = adverse event; ANCOVA = analysis of covariance; CDLQI = Children's Dermatology Life Quality Index; CMH = Cochran-Mantel-Haenszel; CsA = cyclosporine A; DLQI = Dermatology Life Quality Index; EASI = Eczema Area and Severity Index; EASI-50 = reduction of at least 50% in Eczema Area and Severity Index score from baseline; EASI-75 = reduction of at least 75% in Eczema Area and Severity Index score from baseline; HADS = Hospital Anxiety and Depression Scale; IGA = Investigator's Global Assessment; IMP = investigational medicinal product; LOCF = last observation carried forward; MAR = missing at random; NA = not applicable; NRS = numeric rating scale; POEM = Patient-Oriented Eczema Measure; SCORAD = Scoring Atopic Dermatitis; SCORAD-50 = 50% reduction in Scoring Atopic Dermatitis; SCORAD-75 = 75% reduction in Scoring Atopic Dermatitis; TCS = topical corticosteroids.

^aDLQI in the ECZTRA 1, 2, and 3 studies; CDLQI in the ECZTRA 6 study.

^bModel used: change in (SCORAD or DLQI, or CDLQI) = treatment × week + baseline (SCORAD or DLQI, or CDLQI) × week + region + baseline IGA.

^cModel used: change from baseline in (SCORAD or DLQI, or CDLQI) = treatment + baseline value + region + baseline IGA.

^dModel used: change from baseline in SCORAD or DLQI = (treatment × visit) + (baseline SCORAD or DLQI × visit) + prior CSA use + country (Germany: yes or no) + baseline IGA.

^eANCOVA model: change from baseline in SCORAD or DLQI = treatment + baseline SCORAD or DLQI + prior CSA use + country (Germany: yes or no) + baseline IGA.

^fExcluding EASI-50, SCORAD-75, SCORAD-50, and reduction in worst daily pruritus NRS of 3 or more points from baseline (not assessed in the study).

^gModel used: the amount of TCS used (or a number of days without topic treatment use) = (treatment × week) + region + baseline IGA.

Sources: Clinical Study Reports for ECZTRA 1,¹⁵ 2,¹⁶ 3,¹⁷ 6,¹³ and 7¹⁸ and the sponsor’s Summary of Clinical Evidence.¹⁴

Table 11: Analysis Populations of ECZTRA 1, 2, 3, 6, and 7

IMP = investigational medicinal product.

Note: In addition to the analysis populations, a safety follow-up analysis set (included in all studies), a per-protocol analysis set (included in the ECZTRA 1, 2, 3, 6), and an open-label safety analysis set (included in the ECZTRA 1, 2, and 6 studies) were defined in the studies but they were not summarized as they are not of interest to this review.

^aPatients from site 340 (n = 2) and site 341 (n = 7) were excluded due to several issues that involved noncompliance with good clinical practice.

Source: Clinical Study Reports for ECZTRA 1,¹⁵ 2,¹⁶ 3,¹⁷ 6,¹³ and 7¹⁸ and the sponsor’s Summary of Clinical Evidence.¹⁴

Adolescents

Initial Treatment Period

Of 347 screened patients in the ECZTRA 6 study, 46 (13.3%) were screening failures, mostly due to failure to meet eligibility criteria (10.1%). A total of 301 patients were randomized to the tralokinumab 300 mg every 2 weeks group (n = 101), tralokinumab 150 mg every 2 weeks group (n = 100) (results not presented in this report), and the placebo group (n = 100) in the initial treatment period. Study treatment discontinuation was reported in 3% and 8% of patients receiving tralokinumab and placebo, respectively, with the most common reason being withdrawal by parents or guardians in both groups. The full analysis set and the safety analysis set included 97 patients (96.0%) in the tralokinumab group and 94 patients (94.0%) in the placebo group.

Maintenance Treatment Period

In the ECZTRA 6 study, 27 patients were week 16 tralokinumab responders assigned to maintenance treatment with tralokinumab 300 mg every 2 weeks (n = 13) or tralokinumab 300 mg every 4 weeks (n = 14); 6 patients were week 16 placebo responders who continued to receive placebo as maintenance treatment. One patient discontinued study treatment in each of the tralokinumab groups, and no patient discontinued study treatment in the placebo group.

Adults

Initial Treatment Period

Screening failure rates ranged between 12.9% to 25.0% across the ECZTRA 1, 2, 3, and 7 studies, most due to failure to meet eligibility criteria. The respective numbers of patients randomized to tralokinumab (or tralokinumab plus TCS) and placebo (or placebo plus TCS) were 603 and 199 in the ECZTRA 1 study, 593 and 201 in the ECZTRA 2 study, 253 and 127 in the ECZTRA 3 study, and 140 and 137 in the ECZTRA 7 study. No notable between-group imbalance in study treatment discontinuation was noted across studies (ranging between 4.4% and 10.9%), except in the ECZTRA 2 study, in which more patients in the placebo group (10.9%) discontinued from study treatment than from the tralokinumab group (5.6%). The full analysis set and safety analysis set included all or close to all randomized patients (at least 98.6%) in these studies.

Maintenance (or Continuous) Treatment Period

ECZTRA 1: 185 patients were week 16 tralokinumab responders assigned to maintenance treatment (71 in the tralokinumab every 2 weeks group, 78 in the tralokinumab every 4 weeks group, and 36 in the placebo group) and 29 patients were week 16 placebo responders who continued to receive placebo as maintenance treatment.

ECZTRA 2: 227 patients were week 16 tralokinumab responders assigned to maintenance treatment (91 in the tralokinumab every 2 weeks group, 90 in the tralokinumab every 4 weeks group, and 46 in the placebo group) and 31 patients were week 16 placebo responders who continued to receive placebo as maintenance treatment.

ECZTRA 3: Of the 353 patients assigned to continuation treatment, 138 were week 16 tralokinumab responders (evenly split between tralokinumab every 2 weeks plus TCS and tralokinumab every 4 weeks plus TCS), 95 were week 16 tralokinumab nonresponders who received tralokinumab every 2 weeks plus TCS maintenance treatment, 79 were week 16 placebo nonresponders who received tralokinumab every 2 weeks plus TCS maintenance treatment, and 41 were week 16 placebo responders who continued to receive placebo as maintenance treatment.

Table 12: Patient Disposition for the Initial Treatment Period in ECZTRA 6 (Adolescents)

q.2.w. = every 2 weeks.

^aRefers to all screened patients, including those who were eventually randomized to tralokinumab 150 mg q.2.w. (not summarized in this table).

^bRefers to all randomized patients, including those who were randomized to tralokinumab 150 mg q.2.w. (not summarized in this table).

^cFour patients were excluded from the full analysis set due to good clinical practice noncompliance issues at the study site (n = 3) and not receiving study intervention (n = 1).

^dSix patients were excluded from the full analysis set due to good clinical practice noncompliance issues at the study site (n = 5) and not receiving study intervention (n = 1).

Sources: Clinical Study Reports for ECZTRA 6¹³ and the sponsor’s Summary of Clinical Evidence.¹⁴

Table 13: Patient Disposition for Maintenance Treatment Period in ECZTRA 6 (Adolescents)

q.2.w. = every 2 weeks; q.4.w. = every 4 weeks.

Source: Clinical Study Reports for ECZTRA 6¹³ and the sponsor’s Summary of Clinical Evidence.¹⁴

Table 14: Patient Disposition for Initial Treatment Period in ECZTRA 1, 2, 3, and 7 (Adults, Original Review)

NA = not applicable; q.2.w. = every 2 weeks; TCS = topical corticosteroids.

Sources: Clinical Study Reports for ECZTRA 1,¹⁵ 2,¹⁶ 3,¹⁷ and 7¹⁸ and the sponsor’s Summary of Clinical Evidence.¹⁴

Table 15: Patient Disposition for Maintenance Treatment Period in ECZTRA 1 and 2 (Adults, Original Review)

NR = not reported; q.2.w. = every 2 weeks; q.4.w. = every 4 weeks.

Source: Clinical Study Reports for ECZTRA 1¹⁵ and ECZTRA 2¹⁶ and the sponsor’s Summary of Clinical Evidence.¹⁴

Table 16: Patient Disposition for Continuation Treatment Period in ECZTRA 3 (Adults, Original Review)

q.2.w. = every 2 weeks; q.4.w. = every 4 weeks; TCS = topical corticosteroids.

Source: Clinical Study Report for ECZTRA 3¹⁷ and the sponsor’s Summary of Clinical Evidence.¹⁴

Baseline Characteristics

Baseline patient characteristics of the included studies are summarized in Table 17 (adolescents) and Table 18 (adults). The baseline characteristics outlined in the Table 17 and Table 18 are limited to those most relevant to this review or considered to affect the outcomes or interpretation of the study results.

Adolescents

The study population of ECZTRA 6 (n = 191) had a mean age of 14.6 (SD = 1.7) years. There was about an equal proportion of males (51.5%) and females (48.5%). The majority of patients were white (57.5%). Patients had a mean duration of AD diagnosis of 12.3 years (SD = 3.6), a mean BSA involvement of 50.9% (SD = 23.0%), and a mean EASI score of 31.68 (SD = 13.60). Moderate disease and severe disease (as measured by an IGA) were present in 54.0% and 46.0% of patients, respectively. The proportions of patients who had prior TCS, systemic immunosuppressant, monoclonal antibody, and phototherapy treatment for AD were 100%, 21.1%, 2.4%, and 25.6%, respectively.

The baseline patient characteristics were generally balanced between the tralokinumab 300 mg every 2 weeks group and the placebo group, except that a notably higher proportion of patients in the placebo group had received prior systemic corticosteroid treatment (52.1%), wet wraps (30.9%), and phototherapy (30.9%) compared with the tralokinumab 300 mg every 2 weeks group (34.0%, 21.6%, and 16.5%, respectively).

Adults

In the ECZTRA 1 study (n = 802), 2 (n = 794), 3 (n = 380), and 7 (n = 277), the mean age of the overall study population ranged between 36.5 years (SD = 14.1) and 39.1 years (SD = 15.2). More than half of patients were male (55.0% to 59.6%). At baseline, the mean duration of AD diagnosis ranged from 26.2 years (SD = 13.9) to 28.3 years (SD = 14.7). The mean BSA involvement ranged from 48.1% (SD = 24.2%) to 54.7% (SD = 22.2%). The mean EASI score ranged between 29.35 (SD = 12.25) to 32.95 (SD = 12.54). The proportion of patients with severe disease (as assessed by IGA) was between 46.3% and 50.7%. The proportion of patients who received prior TCS was between 98.0% to 99.6%. In the ECZTRA 3 and 7 studies, in which prior use of monoclonal antibodies was reported, the proportions were 6.3% and 7.6%, respectively. There was generally no notable difference in the baseline patient characteristics across the studies, except that the ECZTRA 7 study enrolled a higher proportion of patients who are white (98.2%), received prior cyclosporine A treatment (74.7%), and received prior phototherapy (58.8%), compared with the ECZTRA 1, 2, and 3 studies (with respective ranges of 55.0% to 59.6%, 31.1% to 36.4%, and 43.7% to 48.1%).

No notable between-group difference in baseline patient characteristics was present in the ECZTRA 1 study. In the ECZTRA 2 study, a higher proportion of patients received prior systemic corticosteroids in the tralokinumab group (69.1%) compared with the placebo group (62.2%). In the ECZTRA 3 study, several imbalances between treatment groups were noted, including a lower proportion of patients who were male (49.4%) and who had received prior systemic corticosteroids (58.5%) and prior methotrexate treatment (11.5%) in the tralokinumab every 2 weeks plus TCS group compared with the placebo plus TCS group (66.1%, 67.7%, and 23.6% respectively). As well, a higher proportion of patients were white (80.2%) and had received prior phototherapy (48.2%) in the tralokinumab every 2 weeks plus TCS group compared with the placebo plus TCS group (66.9% and 41.7%, respectively). In the ECZTRA 7 study, a lower proportion of patients received prior phototherapy in the tralokinumab every 2 weeks plus TCS group (56.4%) compared with the placebo plus TCS group (61.3%)

Table 17: Summary of Baseline Characteristics of ECZTRA 6 — Full Analysis Set (Adolescents)

AD = atopic dermatitis; BSA = body surface area; CDLQI = Children’s Dermatology Life Quality Index; EASI = Eczema Area and Severity Index; IGA = Investigator’s Global Assessment; q.2.w. = every 2 weeks; NRS = numeric rating scale; SCORAD = Scoring Atopic Dermatitis; SD = standard deviation.

Source: Clinical Study Report for ECTZRA 6¹³ and the sponsor’s Summary of Clinical Evidence.¹⁴

Table 18: Baseline Characteristics in ECZTRA 1, 2, 3, and 7 — Full Analysis Set (Adult Population, Original Review)

BSA = body surface area; DLQI = Dermatology Life Quality Index; EASI = Eczema Area and Severity Index; IGA = investigator’s global assessment; NR = not reported; NRS = numeric rating scale; SCORAD = Scoring Atopic Dermatitis; q.2.w. = every 2 weeks; SD = standard deviation; TCS = topical corticosteroids.

Sources: Clinical Study Reports for ECZTRA 1,¹⁵ 2,¹⁶ 3,¹⁷ and 7,¹⁸ CADTH Reimbursement Review for Adtralza (2022),⁶⁵ and the sponsor’s Summary of Clinical Evidence.¹⁴

Exposure to Study Treatments

Study treatment exposure in the ECZTRA 6 study (adolescents) is presented in Table 19 (initial treatment period) and Table 20 (maintenance treatment period). Patient disposition in the ECZTRA 1, 2, 3, and 7 (adults) is presented in Table 21 (initial treatment period) and Table 22 (maintenance [or continuous] treatment period).

Adolescents

In the initial treatment period of ECZTRA 6, the majority of patients in both treatment groups received all doses of the study treatment (93.8% in the tralokinumab 300 mg every 2 weeks group and 90.4% in the placebo group). The mean durations of exposure were 0.304 patient-years of exposure (PYE) (SD = 0.030) in the tralokinumab group, and 0.297 PYE (SD = 0.054) in the placebo group.

In the maintenance treatment period of ECZTRA 6, the proportions of patients who received all doses of study treatment were 63.6% in the tralokinumab 300 mg every 2 weeks group, 84.6% in the tralokinumab 300 mg every 4 weeks group, and 50.0% in the placebo group. The mean durations of exposure were 0.510 PYE (SD = 0.232) in the tralokinumab 300 mg every 2 weeks group, 0.522 PYE (SD = 0.244) in the tralokinumab 300 mg every 4 weeks group, and 0.497 PYE (SD = 0.316) in the placebo group.

Adults

More than 70% of patients received all doses of the study treatment in the initial treatment period across the ECZTRA 1, 2, 3, and 7 studies. No notable between-group difference in the mean duration of exposure was noted in the initial treatment period across these studies. In the ECZTRA 1, 2, and 3 studies, the mean duration of exposure in both treatment groups was approximately 0.30 PYE. || |||||| || ||| |||| |||||||| || |||||||| ||| |||| ||||||| ||| || ||| |||||||||||||||| ||| |||| |||||| ||| || ||| ||||||||||| ||||||

In the maintenance (or continuous) treatment period of the ECZTRA 1, 2, and 3 studies, between 76.9% and 94.2% of patients who were week 16 tralokinumab responders received all doses of study treatment. No notable between-group difference in the mean duration of exposure was noted in the ECZTRA 3 study. In the ECZTRA 1 study, the mean duration of exposure was lower in patients who transitioned from tralokinumab to placebo (0.52 PYE; SD = 0.24) compared with the tralokinumab every 2 weeks group (0.56 PYE; SD = 0.22) and tralokinumab every 4 weeks group (0.57 PYE; SD = 0.21). In the ECZTRA 2 study, the mean duration of exposure was also lower in in the patients who transitioned from tralokinumab to placebo (0.44 PYE; SD = 0.22) compared with other treatment groups (tralokinumab every 2 weeks group [0.52 PYE; SD = 0.24]; tralokinumab every 4 weeks group [0.50 PYE; SD = 0.24]).

Table 19: Patient Exposure for ECZTRA 6 in Initial Treatment Period — Safety Analysis Set (Adolescents)

IQR = interquartile range; PYE = patient-years of exposure; q.2.w. = every 2 weeks; SD = standard deviation.

Sources: Clinical Study Report for ECZTRA 6¹³ and the sponsor’s Summary of Clinical Evidence.¹⁴

Table 20: Patient Exposure for ECZTRA 6 in Maintenance Treatment Period — Safety Analysis Set (Adolescents)

IQR = interquartile range; PYE = patient-years of exposure; q.2.w. = every 2 weeks; q.4.w. = every 4 weeks; SD = standard deviation.

Source: Clinical Study Report for ECZTRA 6¹³ and the sponsor’s Summary of Clinical Evidence.¹⁴

Table 21: Patient Exposure from ECZTRA 1, 2, 3, and 7 in Initial Treatment Period — Safety Analysis Set (Adults, Original Review)

IQR = interquartile range; PYE = patient-years of exposure; q.2.w. = every 2 weeks; SD = standard deviation; TCS = topical corticosteroids.

^aRefers to the patient exposure in the 26-week treatment period.

Sources: Clinical Study Reports for ECZTRA 1,¹⁵ 2,¹⁶ 3,¹⁷ and 7¹⁸ and the sponsor’s Summary of Clinical Evidence.¹⁴

Table 22: Patient Exposure from ECZTRA 1, 2, and 3 in Maintenance (Continuous Treatment) Period — Maintenance (or Continuous) Safety Analysis Set (Adults, Original Review)

IQR = interquartile range; PYE = patient-years of exposure; q.2.w. = every 2 weeks; q.4.w. = every 4 weeks; SD = standard deviation; TCS = topical corticosteroids.

Sources: Clinical Study Reports for ECZTRA 1,¹⁵ 2,¹⁶ and 3¹⁷ and the sponsor’s Summary of Clinical Evidence.¹⁴

Rescue Treatment Exposure

Rescue treatment exposure in the ECZTRA 6 study (adolescents) is presented in Table 23. Patient disposition in the ECZTRA 1, 2, 3, and 7 studies (adults) is presented in Table 24 (initial treatment period) and Table 25 (maintenance [or continuous] treatment period).

Adolescents

In the initial treatment period of the ECZTRA 6 study, 29.9% of patients in the tralokinumab 300 mg every 2 weeks group and 56.4% of patients in the placebo group received rescue medication. The most frequently used rescue medication was TCS in both groups (29.9% in the tralokinumab 300 mg every 2 weeks group and 54.3% in the placebo group).

In the maintenance treatment period of the ECZTRA 6 study, 2 patients (18.2%) in the tralokinumab 300 mg every 2 weeks group and 1 patient (7.7%) in the tralokinumab 300 mg every 4 weeks group received rescue medication.

Adults

In the initial treatment period of the ECZTRA 1, 2, 3, and 7 studies, the frequency of rescue medication use was consistently lower in the tralokinumab (or tralokinumab plus TCS) group compared with the placebo (or placebo plus TCS) group (ECZTRA 1: 35.8% versus 46.2%; ECZTRA 2: 22.8% versus 44.3%; ECZTRA 3: 2.8% versus 10.2%; ECZTRA 7: 5.7% versus 13.9%).

In the ECZTRA 1 and 2 studies, among week 16 tralokinumab responders who were re-randomized to receive maintenance treatment, the proportion of patients who received rescue medication was the highest in the placebo group (ECZTRA 1: 41.7%; ECZTRA 2: 19.6%), followed by the tralokinumab 300 mg every 4 weeks group (ECZTRA 1: 33.3%; ECZTRA 2: 17.8%), and the tralokinumab 300 mg every 2 weeks group (ECZTRA 1: 28.2%; ECZTRA 2: 16.5%). In the ECZTRA 3 study, among week 16 tralokinumab responders who were re-randomized to receive continuous treatment, no patients in the tralokinumab 300 mg every 2 weeks plus TCS group and 1 patient (1.4%) in the tralokinumab 300 mg every 4 weeks plus TCS group received rescue medication.

In both the initial and maintenance (or continuous) treatment periods, the most frequently used rescue medication was TCS in all treatment groups across ECZTRA 1, 2, and 3.

Table 23: Rescue Medication Exposure from ECZTRA 6 (Adolescents)

q.2.w. = every 2 weeks; q.4.w. = every 4 weeks.

Source: Clinical Study Report for ECZTRA 6.¹³