CADTH Reimbursement Review

Roflumilast (Zoryve)

Sponsor: Arcutis Biotherapeutics Inc.

Therapeutic area: Plaque psoriasis

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Stakeholder Input

Clinical Review

Executive Summary

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

Table 1: Background Information of Application Submitted for Review

NOC = Notice of Compliance; w/w = weight by weight.

Introduction

Psoriasis is a chronic, immune-mediated, inflammatory skin disease that is associated with multiple comorbidities, such as psoriatic arthritis, obesity, and metabolic syndrome.^1,2 Psoriasis can present similarly in children and adults; the peak ages of onset are between 30 and 39 years and between 50 and 69 years.¹ A panel of 11 dermatologists with expertise in psoriasis estimated that the median prevalence of psoriasis is 3% of the general adult population in Canada. The panel estimated that approximately 78% of patients with psoriasis have less than 10% body surface area (BSA) involved (i.e., mild or moderate disease) and 22% of patients have 10% or more BSA involved. The panel further estimated that 50% of patients have less than 3% of BSA involved (i.e., mild disease) and 2% of patients have more than 50% of BSA involved.³ Based on estimates in the US, 50% of patients have facial involvement⁴ and 21% to 30% of patients have intertriginous area involvement.⁵

Chronic plaque psoriasis (also known as psoriasis vulgaris)⁶ is the most common clinical subtype of psoriasis, representing approximately 90% of patient cases in Canada.⁷ Plaque psoriasis is characterized by well-demarcated, erythematous, cutaneous plaques with overlying, coarse, silvery scales. Plaques can be asymptomatic; however, pruritus and pain are often reported by patients. Common areas of involvement include the scalp, elbows, knees, and gluteal cleft. Additionally, intertriginous areas (inverse psoriasis), ear canals, umbilicus, palms, soles, and nails are possible areas of involvement. Intertriginous psoriasis is characterized by well demarcated, smooth, shiny plaques with no to minimal scales.¹

Plaque psoriasis requires lifelong follow-up and treatment. Measures of treatment success in clinical practice may include clearance (absence of signs of disease), control (satisfactory response to therapy as defined by the patient and/or clinician), and remission (maintenance of disease control or suppression of signs and symptoms over time).⁶

Management of mild plaque psoriasis involving the trunk, limbs, and neck includes topical therapies that can be broadly categorized as corticosteroids, vitamin D3 analogues, retinoids, anthralin (commercial formulations of anthralin are not currently available in Canada) and tars, as well as combination therapy.^6,8 The therapeutic options for pediatric patients are generally similar to the options for adult patients.⁶ The management of intertriginous psoriasis, affecting the groin, axillae, inframammary region, abdominal body folds, gluteal cleft, perianal region, and retroarticular fold areas, can be a challenge, as these areas are at an increased risk of adverse reactions to topical therapy because the skin tends to be thinner in these regions.⁶ Moreover, there are currently no available treatments indicated for intertriginous psoriasis. Topical calcineurin inhibitors may be an appropriate treatment option for the management of intertriginous psoriasis; however, they are not currently approved for an indication in psoriasis (hence its off-label use).^6,8 Treatment selection should be individualized to the patient to improve adherence, support patient satisfaction, and achieve treatment success.⁶ Management of moderate to severe plaque psoriasis affecting the trunk and extremities that cannot be adequately controlled by the approaches described previously (adequate control is defined by the patient’s perception of the disease and its burdens) include systemic therapy, phototherapy, combination therapy, and topical therapy as adjunct therapy.⁶

The objective of this report is to review and critically appraise the clinical evidence submitted by the sponsor on the beneficial and harmful effects of roflumilast cream 0.3% for the topical treatment of plaque psoriasis, including treatment of psoriasis in the intertriginous areas, in patients aged 12 years and older.

Stakeholder Perspectives

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

Patient Input

Three patient groups submitted a joint patient input submission. The Canadian Skin Patient Alliance (CSPA), Canadian Psoriasis Network, and Canadian Association of Psoriasis Patients are national, not-for-profit organizations that are dedicated to improving the lives of people with psoriasis across Canada.

According to the input, the impact of psoriasis symptoms is considerable in various areas of patients’ lives, such as mental well-being, daily tasks, intimate relationships, and social lives. Most patients considered effectiveness as the most important factor to be addressed in any new treatment for psoriasis. In addition, patients valued lack of side effects, affordability, treatments that are easy to apply, and treatments that are conducive to their schedules.

More than half of survey respondents reported having moderate psoriasis. The body area with the most impact was the scalp, followed by the legs, arms, genitals, hands, torso, skin folds, and palms. Based on the survey, respondents were using topical corticosteroids, topical combination treatments, and/or biological drugs. Most patients with psoriasis have discontinued their treatment at some point during their disease, with the most frequent reasons being that the treatment stopped working, caused side effects, was unaffordable, or was ineffective. Among respondents, 10 patients reported experience with roflumilast, accessed through a clinical trial. Nine of these patients reported benefits of treatment that included clearing of skin, reduced itch and redness, clearing of skin lesions (plaques), and ease of treatment application. All 10 respondents tolerated roflumilast well, except for 1 who experienced some itching. The patient group input emphasized that psoriasis is a chronic and potentially debilitating disease that poses many challenges and is linked to anxiety, depression, and social isolation. This disease can interfere with social and intimate relationships, productivity, and family life and work life. Furthermore, due to the chronicity of this disease, patients are concerned about recurrence and resistance to treatments in the future.

Clinician Input

Input From the Clinical Expert Consulted by CADTH

Unmet Need

The clinical expert indicated there are currently no available treatment options that can permanently reverse the course of plaque psoriasis, but systemic therapies can suppress psoriasis while they are being used.

The clinical expert also indicated that many patients do not respond to currently available topical treatments; in particular, widespread psoriasis and psoriasis of the scalp, palms, and/or soles are often refractory to topical treatment. The clinical expert noted that although topical therapies are generally well tolerated and that topical steroids are used over the long-term for many patients in clinical practice, clinical guidelines caution against the long-term use of topical steroids for plaque psoriasis. As such, the clinical expert agreed that more effective, safer, and better-tolerated treatment options are an unmet need, particularly for long-term use and in patients with facial, genital, and intertriginous involvement.

The clinical expert further suggested that topical formulations that are less messy and more effective may improve adherence, particularly among patients with intertriginous involvement who are offered topical tacrolimus, as it is only available as an ointment and not widely available because it is not indicated (used off-label) for the treatment of psoriasis.

Place in Therapy

For most patients with plaque psoriasis, the clinical expert suggested they would consider roflumilast as an alternative to other first-line topical treatments for psoriasis within the current stepped approach, as this could prevent having to step up patients to systemic therapy. For patients with facial, genital, and intertriginous involvement, the clinical expert anticipated that roflumilast would be a first-line treatment. The clinical expert further suggested that roflumilast may be preferred for long-term use in some patients and by clinicians, given that it is not a steroid.

For mild disease, the clinical expert suggested that roflumilast would be used as monotherapy. Additionally, depending on patient preference, roflumilast could be used simultaneously with other topical treatments. Roflumilast could also be used in combination with systemic therapy or phototherapy, as topical treatments are typically continued as needed.

Patient Population

The clinical expert identified patients with active psoriasis, regardless of whether it is being actively treated or not, as the most in need of an intervention. The clinical expert noted there is a need for an intervention for the hair-bearing scalp area; however, they indicated that this need would not be addressed by roflumilast. Clinician examination and judgment together with shared decision-making with the patient, would determine whether the drug is best suited for the patient. The clinical expert suggested that patients with psoriasis with limited BSA involvement are the most likely to respond to therapy with roflumilast as monotherapy.

Assessing the Response to Treatment

The clinical expert indicated that clearance or near clearance of psoriasis lesions is commonly used to assess psoriasis severity; this is analogous to the Investigator’s Global Assessment (IGA) scales used in clinical trials. The clinical expert also indicated that patient satisfaction is important and is assessed in a gestalt manner that likely differs between physicians.

According to the clinical expert, extensive measurement scales, such as the Psoriasis Area and Severity Index (PASI), Dermatology Life Quality Index (DLQI), Worst Itch Numeric Rating Scale (WI-NRS), and Psoriasis Symptom Diary (PSD), are not typically used in clinical practice; however, PASI and DLQI may be used if mandated by the insurer to fund a drug.

Discontinuing Treatment

According to the clinical expert, intolerance (e.g., stinging at the application site) and ineffectiveness (i.e., minimal improvement after 2 to 8 weeks of application) are factors to consider when deciding to discontinue treatment with roflumilast.

The clinical expert further indicated that roflumilast would be discontinued when lesions have cleared and can be restarted if and when lesions recur. The clinical expert indicated that phototherapy or systemic therapy may be considered for extensive psoriasis regardless of improvement with roflumilast and, if there is substantial improvement with phototherapy or systemic therapy that obviates the need for topical therapy, then roflumilast would be considered for discontinuation.

Clinician Group Input

Three clinician groups provided input: Fraser Health Dermatology (4 clinicians), the Canadian Dermatology Association (3 clinicians), and the Atlantic Provinces Dermatology Association and Dermatology Association of Ontario (12 clinicians).

The clinician groups noted there is a need for a new treatment to substitute existing therapies for plaque psoriasis. The clinician groups noted that the limitations of current treatments include unfavourable effects, poor compliance, difficult application, high costs, limited efficacy, intolerability due to irritation, and inability to be administered to all areas of the body. The treatment goals noted include reducing the severity of symptoms, minimizing adverse events (AEs), improving tolerability and efficacy, increasing patient quality of life (QoL), and reducing the burden on patients and health care systems.

One clinician group recommended the use of roflumilast as a first-line treatment for the management of mild to severe plaque psoriasis and another group suggested roflumilast should be used after treatment with topical steroids has failed, as topical steroids are an inexpensive and usually well tolerated therapy. The input noted that patients best suited for treatment with the drug under review included those with mild to moderate psoriasis, patients with different disease phenotypes, patients with psoriasis that has failed to respond to topical steroids, and patients with intertriginous psoriasis, noting the importance of steroid sparing in these anatomic sites.

One clinician group suggested that the best outcomes to determine treatment response would be Physician’s Global Assessment (PGA) score and level of BSA involvement. The group noted the PASI score is not commonly used in clinical practice, except when applying for drug plan coverage for systemic therapies for patients with moderate to severe disease. The clinician groups noted that assessment of treatment goals is recommended after 8 weeks and that, typically, patients with psoriasis are initially seen every 3 to 6 months to assess response to treatment, and that treatment should be discontinued due to a lack of efficacy or disease progression. The input received stated that patients with psoriasis are diagnosed and treated as outpatients, typically by both specialists and general practitioners or family doctors.

Drug Program Input

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

• relevant comparators

• consideration for initiation of therapy

• consideration for continuation or renewal of therapy

• consideration for prescribing of therapy

• generalizability

• care provision issues

• system and economic issues.

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

Clinical Evidence

Pivotal Studies and Randomized Controlled Trial Evidence

Description of Studies

The sponsor’s submission included 2 phase III, randomized, double-blind, parallel-group, vehicle-controlled trials, the DERMIS-1⁹ (N = 439) and DERMIS-2¹⁰ (N = 442) trials. The 2 trials assessed success in IGA (defined as an IGA score of clear or almost clear plus an improvement of 2 more grades from baseline at week 8) with roflumilast cream 0.3% compared with matching vehicle in patients with chronic plaque psoriasis involving 2% to 20% of BSA (excluding the scalp, palms, and soles). Patients were excluded if they were unable to discontinue prohibited medications and treatments (defined as systemic, biologic, topical, phototherapy, and investigational treatments) that could affect plaque psoriasis within the prespecified washout period. Patients were randomized in a 2:1 ratio to receive roflumilast or the matching vehicle applied topically once daily for 8 weeks. Secondary outcomes included measures of symptoms and involvement (PASI, IGA, intertriginous IGA [I-IGA], and WI-NRS) and a measure of QoL (PSD). Similarly, exploratory outcomes included measures of involvement (e.g., BSA), measures of health-related quality of life (HRQoL) (e.g., DLQI and Children’s DLQI [CDLQI]), local tolerability, and safety.

At baseline, |||| and |||| of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively, were in the age category of 12 years to 17 years and ||||| and ||||| of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively, were aged 18 years or older. The mean age of all randomized patients was 48.0 years (standard deviation [SD] = 14.69) in the DERMIS-1 trial and 47.0 years (SD = 14.72) in the DERMIS-2 trial. Most patients were male (64.9% and 62.4% of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively), while 33.9% to 39.3% of patients were female. The majority of patients were white (81.5% and 82.8% of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively), while the remainder of the randomized patients (0% to 7.3%) identified as Asian, Black or African American, American Indian or Alaska Native [wording from original source], Native Hawaiian or other Pacific Islander, other, or more than 1 race. A similar proportion of all randomized patients in the DERMIS-1 and DERMIS-2 trials had facial involvement (27.1% and 26.0%, respectively), genital involvement (16.4% and 14.7%, respectively), and a baseline I-IGA score (23.0% and 19.9%, respectively). The majority of patients had a moderate IGA score at baseline (74.7% and 76.5% of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively). The mean baseline BSA affected by psoriasis in all randomized patients was 6.66% (SD = 4.538) in the DERMIS-1 trial and 7.30% (SD = 4.918) in the DERMIS-2 trial. The mean baseline PASI score for all randomized patients was 6.5 (SD = 3.35) in the DERMIS-1 trial and 6.7 (SD = 3.33) in the DERMIS-2 trial.

Efficacy Results

The efficacy end points that were noted to be important to patients and clinicians based on stakeholder input are summarized in Table 2.

Investigator’s Global Assessment

The IGA is an investigator-reported, static evaluation of the overall severity of psoriasis of the whole body. The minimal important difference (MID) in IGA has not been estimated. However, achieving a score of 0 (clear) or 1 (almost clear) on the IGA has generally been accepted as clinically meaningful.^11,12 Alternatively, or in addition to the achievement of a score of 0 or 1, the responder analysis may also consider the proportion of patients with at least a 2-grade improvement from baseline on the static IGA;¹¹ this was consistent with the definition of IGA success that was used in both trials. The primary end point in both trials was an improvement in the overall severity of psoriasis that was measured by the proportion of patients who experienced treatment success based on the IGA, defined as a score of 0 or 1 plus an improvement of 2 or more grades from baseline at week 8.

The primary end point, IGA success at week 8, was met for both trials in the intention-to-treat (ITT) population. In the DERMIS-1 trial, 42.4% of patients in the roflumilast arm versus 6.1% of patients in the vehicle arm experienced treatment success based on the IGA at week 8; the ratio of the odds of IGA success with roflumilast relative to the odds of IGA success with the matching vehicle was ||||| (95% ||| |||| || |||||; | |||||||) at week 8 from baseline in favour of roflumilast. In the DERMIS-2 trial, 37.5% of patients in the roflumilast arm versus 6.9% of patients in the vehicle arm experienced treatment success based on the IGA at week 8; the ratio of the odds of IGA success with roflumilast relative to the odds of IGA success with the matching vehicle was |||| |||| || |||| || |||||| | | ||||||| at week 8 from baseline, also in favour of roflumilast. In both studies, the results of the primary analyses were generally consistent with the sensitivity analyses of the primary end point. IGA success at week 4 was also reported and was also in favour of roflumilast when compared with the matching vehicle in the DERMIS-1 trial (13.44; 95% confidence interval [CI], 3.72 to 48.58; P < 0.0001) and the DERMIS-2 trial (3.91; 95% CI, 1.76 to 8.70; P = 0.0011).

Intertriginous IGA

I-IGA was defined as the IGA scale but was used to evaluate only intertriginous areas in the trials. In both trials, an improvement in the severity of intertriginous psoriasis was measured by the proportion of patients who experienced treatment success based on the I-IGA, defined as a score of 0 (clear) or 1 (almost clear) plus an improvement of 2 or more grades from baseline at week 8, which was a secondary end point tested in a hierarchical manner and adjusted for multiple comparisons. Note that this analysis was based on the prespecified I-IGA ITT population, a subset of patients in the ITT population with intertriginous area involvement and an I-IGA of 2 or more at baseline; 52 patients (82.5%) in the roflumilast arm and 29 patients (90.6%) in the vehicle arm were available in the DERMIS-1 trial for the analysis at week 8, and 47 patients (88.6%) and 27 patients (87.0%), respectively, were available in the DERMIS-2 trial for the analysis at week 8.

In both trials, the proportion of patients who experienced treatment success based on the I-IGA at week 8 was greater with roflumilast treatment compared with the matching vehicle. More specifically, in the DERMIS-1 trial, the ratio of the odds of I-IGA success with roflumilast relative to the odds of I-IGA success with the matching vehicle was 17.94 (95% CI, 2.33 to 138.20; P < 0.0001) at week 8 from baseline, in favour of roflumilast. In the DERMIS-2 trial, the ratio of the odds of I-IGA success with roflumilast relative to the odds of I-IGA success with the matching vehicle was 11.18 (95% CI, 2.33 to 53.68; P = 0.0004) at week 8 from baseline, also in favour of roflumilast.

Psoriasis Area and Severity Index

PASI is an investigator-reported evaluation of the extent and severity of psoriasis. An MID in PASI has not been estimated. In both trials, an improvement in the extent and severity of disease was measured based on the proportion of patients who had a 75% improvement in PASI (PASI 75) from baseline at week 8, which was a secondary end point tested in a hierarchical manner and adjusted for multiple comparisons.

In both trials, the proportion of patients who had a PASI 75 was greater with roflumilast treatment compared with the matching vehicle. More specifically, in the DERMIS-1 trial, the ratio of the odds of a PASI 75 with roflumilast relative to the odds of a PASI 75 with the matching vehicle was 12.00 (95% CI, 5.15 to 27.93; P < 0.0001) at week 8 from baseline, in favour of roflumilast. In the DERMIS-2 trial, the ratio of the odds of a PASI 75 with roflumilast relative to the odds of a PASI 75 with the matching vehicle was 10.42 (95% CI, 4.49 to 24.19; P < 0.0001) at week 8 from baseline, also in favour of roflumilast. Time to PASI 50 was also reported and was also in favour of roflumilast when compared with the matching vehicle in the DERMIS-1 trial; the median Kaplan-Meier (KM) estimate was 31.0 days (95% CI, 29.0 to 41.0) in the roflumilast arm versus 104.0 days (95% CI, 85.0 to not estimable [NE])) in the vehicle arm [P < 0.0001]) and, in the DERMIS-2 trial, the median KM estimate was 30.0 days (95% CI, 29.0 to 42.0) in the roflumilast arm and NE (95% CI, 71.0 to NE] in the vehicle arm [P < 0.0001]).

DLQI and Children’s DLQI

DLQI and CDLQI are patient-reported tools used to evaluate HRQoL. The estimated within-group MID is 2.2 to 6.9 points in patients with psoriasis and other inflammatory skin disorders.^13,14 In both trials, an improvement in HRQoL was measured by change from baseline in DLQI and CDLQI at week 8, which were exploratory end points that were not included in the statistical hierarchy and not adjusted for multiple comparisons.

A decrease in the DLQI score corresponds to an improvement in HRQoL. In the DERMIS-1 trial, the least squares (LS) mean change from baseline in DLQI at week 8 was |||| || |||||| in the roflumilast arm and |||| || |||||| in the vehicle arm. In the DERMIS-2 trial, the LS mean change from baseline in DLQI at week 8 was |||| || |||||| in the roflumilast arm and |||| || |||||| in the vehicle arm.

Not enough data were collected to carry out an analysis of covariance for change in CDLQI at week 8 in both the DERMIS-1 and DERMIS-2 trials.

Worst Itch Numeric Rating Scale

WI-NRS is a patient-reported outcome measure that is used to assess the severity of itch. The MID in WI-NRS has been estimated to be an improvement of 4 or more points in patients with plaque psoriasis,¹⁵ which is consistent with the definition of WI-NRS success used in the trials. In the DERMIS-1 and DERMIS-2 trials, a reduction in the severity of itch was measured by the proportion of patients who experienced treatment success based on the WI-NRS, defined as a reduction of at least 4 points from baseline at week 8, which was a secondary end point tested in a hierarchical manner and adjusted for multiple comparisons. This analysis was based on a prespecified subset of patients in the ITT population with a WI-NRS pruritus score of 4 or more (the PRU4 ITT population) at baseline; 191 patients (87.6%) in the roflumilast arm and 97 patients (84.3%) in the vehicle arm were available in the DERMIS-1 trial for the analysis at week 8, and 206 patients (89.9%) and 101 patients (87.0%), respectively, were available in the DERMIS-2 trial for the analysis at week 8.

In both the DERMIS-1 and DERMIS-2 trials, the proportion of patients who reported a reduction in the severity of itch based on WI-NRS success was greater with roflumilast treatment compared with the matching vehicle. More specifically, in the DERMIS-1 trial, the ratio of the odds of WI-NRS success with roflumilast relative to the odds of WI-NRS success with the matching vehicle was 7.84 (95% CI, 3.85 to 15.94; P < 0.0001) at week 8 from baseline, in favour of roflumilast. In the DERMIS-2 trial, the ratio of the odds of WI-NRS success with roflumilast relative to the odds of WI-NRS success with the matching vehicle was 3.59 (95% CI, 2.07 to 6.23; P < 0.0001) at week 8 from baseline, also in favour of roflumilast. WI-NRS success at week 4 was also reported and was also in favour of roflumilast when compared with the matching vehicle in the DERMIS-1 trial (4.36; 95% CI, 2.31 to 8.26; P < 0.0001) and in the DERMIS-2 trial (4.93; 95% CI, 2.65 to 9.18; P < 0.0001). However, improvement in the severity of itch as measured by WI-NRS success was not consistently observed at week 2 across trials.

Psoriasis Symptom Diary

PSD is a patient-reported assessment of the impact of plaque psoriasis on overall QoL. An MID in PSD total score has not been estimated. In both trials, an improvement in QoL was measured based on change from baseline in PSD total score at week 8, which was a secondary end point tested in a hierarchical manner and adjusted for multiple comparisons.

In both trials, a greater improvement in QoL was observed in the roflumilast treatment groups compared with the matching vehicle (lower scores indicate less severe or bothersome symptoms). More specifically, in the DERMIS-1 trial, the LS mean difference between roflumilast and vehicle in change from baseline in PSD total score at week 8 was −30.9 (|||||||| ||||| |||| |||||; 95% CI, −37.2 to −24.6; P < 0.0001), in favour of roflumilast. In the DERMIS-2 trial, the LS mean difference between roflumilast and vehicle in change from baseline in PSD total score at week 8 was −26.5 (| |||||; 95% CI, −33.2 to −19.7; P < 0.0001), also in favour of roflumilast. The LS mean difference between roflumilast and vehicle in change from baseline in PSD total score was also reported at week 4 and was also in favour of roflumilast in the DERMIS-1 trial (−25.8; | |||||; 95% CI, −31.7 to −20.0; P < 0.0001) and in the DERMIS-2 trial (−26.0; ||| |||||| 95% CI, −31.9 to −20.0; P < 0.0001).

Body Surface Area

BSA measures the extent of psoriasis as assessed by an investigator. An MID has not been estimated for BSA. In both trials, an improvement in the extent of disease was measured by percent change from baseline in BSA affected by psoriasis at week 8, which was an exploratory end point that was not included in the statistical hierarchy and not adjusted for multiple comparisons.

A decrease in percent BSA corresponds to an improvement in extent of disease. In the DERMIS-1 trial, the LS mean percent change from baseline in BSA affected by psoriasis at week 8 was |||||| || ||||||| in the roflumilast arm and |||||| || ||||||| in the vehicle arm. In the DERMIS-2 trial, the LS mean percent change from baseline in BSA affected by psoriasis at week 8 was |||||| || |||||| in the roflumilast arm and ||||| || ||||||| in the vehicle arm.

Harms Results

The key harms results are summarized in Table 2.

Adverse Events

The proportion of patients in the roflumilast arm with any treatment-emergent adverse event (TEAE) was 25.2% in the DERMIS-1 trial and 25.9% in the DERMIS-2 trial, while the proportion of patients in the vehicle arm with any TEAE was 23.5% in the DERMIS-1 trial and 18.4% in the DERMIS-2 trial. The most common TEAEs (a frequency of 2% or more of patients in either study) reported in the roflumilast arm were diarrhea (3.5% in the DERMIS-1 trial and 2.8% in the DERMIS-2 trial) and headache (1.0% in the DERMIS-1 trial and 3.8% in the DERMIS-2 trial). All remaining TEAEs were reported in less than 2% of patients in the roflumilast arm in either study.

Serious Adverse Events

The proportion of patients in the roflumilast arm with any serious adverse event (SAE) was 0.7% in the DERMIS-1 trial and no patients in the DERMIS-2 trial, while the proportion of patients in the vehicle arm with any SAE was 0.7% in both the DERMIS-1 and DERMIS-2 trials. The SAEs reported in the roflumilast arm were concussion (1 patient in the DERMIS-1 trial) and foot fracture, thorax deformity, and pneumothorax (1 patient in the DERMIS-1 trial).

Withdrawals Due to Adverse Events

The proportion of patients in the roflumilast arm with any withdrawal due to adverse event (WDAE) was |||| in the DERMIS-1 trial and |||| in the DERMIS-2 trial, while the proportion of patients in the vehicle arm with any WDAE was |||| in the DERMIS-1 trial and |||| in the DERMIS-2 trial. All TEAEs leading to discontinuation of treatment and/or study withdrawal were reported in less than || of patients in the roflumilast arm in either study.

Mortality

Based on information the sponsor provided, no deaths occurred during the DERMIS-1 and DERMIS-2 trials.

Notable Harms

The proportion of patients with application site pain in the roflumilast arm was 0.7% in the DERMIS-1 trial and 1.4% in the DERMIS-2 trial. All remaining TEAEs of special interest (application site pruritis, urticaria, dryness, dermatitis, and irritation) were reported in less than 1% of patients in the roflumilast arm in either study.

Table 2: Summary of Key Results From the DERMIS-1 and DERMIS-2 Trials

ANCOVA = analysis of covariance; BSA = body surface area; CI = confidence interval; DLQI = Dermatology Life Quality Index; IGA = Investigator’s Global Assessment; I-IGA = intertriginous Investigator’s Global Assessment; ITT = intention-to-treat; LS = least squares; NA = not applicable; OR = odds ratio; PASI 75 = 75% reduction in the Psoriasis Area and Severity Index score; PRU4 = patients with a WI-NRS pruritus score ≥ 4 at baseline; PRU4 ITT = subset of patients in the ITT population with a WI-NRS pruritus score of 4 or more at baseline; PSD = Psoriasis Symptom Diary; SAE = serious adverse event; SD = standard deviation; SE = standard error; TEAE = treatment-emergent adverse event; WDAE = withdrawal due to adverse event; WI-NRS = Worst Itch Numeric Rating Scale.

Note: The P value was adjusted for multiple testing of the primary and secondary outcomes using a prespecified hierarchical testing strategy.

^aIGA success was defined as a score of 0 (clear) or 1 (almost clear) plus a ≥ 2-grade improvement from baseline.

^bOdds ratio, 95% CI, and P value were obtained from the Cochran-Manel-Haenszel test (stratified by study site, baseline IGA, and baseline intertriginous involvement) comparing roflumilast cream 0.3% with vehicle.

^cI-IGA success was defined as a score of 0 (clear) or 1 (almost clear) plus a ≥ 2-grade improvement from baseline in patients with intertriginous area involvement and I-IGA ≥ 2 at baseline. This analysis was based on the I-IGA ITT population.

^dThe odds ratio, 95% CI, and P value were obtained from the Cochran-Mantel-Haenszel test (stratified by study site and baseline I-IGA) comparing roflumilast cream 0.3% with vehicle.

^eEstimates for LS means (change from baseline and difference from vehicle), 95% CIs, and P values are from an ANCOVA with treatment, site, baseline IGA, baseline intertriginous involvement, and baseline PSD score as independent variables.

^fThe P value is to test for a zero difference between groups (roflumilast cream 0.3% minus vehicle) in change from baseline.

^gWI-NRS success was defined as a ≥ 4-point reduction in WI-NRS pruritus score from baseline in patients with a WI-NRS pruritus score ≥ 4 at baseline. The analysis was based on the PRU4 ITT population.

^hEstimates for LS means (change or percentage change from baseline and difference from vehicle) and accompanying 95% CIs, and P values are from an ANCOVA with treatment, site, baseline IGA, baseline intertriginous involvement, and baseline of the variable being analyzed as independent variables.

ⁱThe P value was not adjusted for multiple testing (i.e., the type I error rate was not controlled).

^jPatients were counted once for each system organ class and once for each preferred term.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Critical Appraisal

Internal Validity

In the DERMIS-1 and DERMIS-2 trials, patients were randomized at baseline according to a computer-generated randomization list and randomization was stratified by study site, baseline IGA (2 versus ≥ 3), and intertriginous involvement at baseline (I-IGA ≥ 2, yes versus no). Based on input from the clinical expert consulted by CADTH for this review, the extent and severity of disease as measured by baseline BSA and PASI are additional effect modifiers. Note that the median and mean BSA and PASI scores were slightly higher in the vehicle arm compared with the roflumilast arm in both trials. IGA assesses severity of lesions, while BSA and PASI account for extent and severity of disease. Therefore, stratification by IGA alone may not result in an optimal comparability in disease severity between treatment arms, and this may have introduced bias in the efficacy results against roflumilast due to the aforementioned imbalance (note that the magnitude of this potential bias is not known). No other baseline demographic and clinical characteristics were identified that could have had a potential impact on the results in either study.

The primary efficacy outcome in both studies was IGA success at week 8. Conclusions about the validity and reliability of the 5-point IGA used in both studies are limited due to the use of the 6- and 7-point IGA in the psychometric validation studies; no evidence for responsiveness of IGA was identified. However, the clinical expert suggested that this difference in scales was unlikely to have introduced bias in the results. Although an MID has not been estimated, having a score of 0 (clear) or 1 (almost clear) on the static IGA has been generally accepted as clinically meaningful (i.e., a responder analysis would consider the proportion of patients with psoriasis who had a score of 0 or 1 in a clinical trial as treatment success).^11,12 Alternatively, or in addition to having a score of 0 or 1, the responder analysis may also consider the proportion of patients with at least a 2-grade improvement from baseline on the static IGA.¹¹ These outcomes were consistent with the definition of IGA success used in the studies.

The primary and secondary outcomes were controlled for multiplicity in both studies using a prespecified hierarchical testing strategy and the Holm procedure to control the familywise type I error. The planned sample size provided sufficient power to analyze up to the first 5 secondary end points; note that the remaining end points demonstrated statistically significant differences between treatment arms that were consistently in favour of roflumilast, with the exception of WI-NRS success at week 2 in the DERMIS-1 trial. The relatively small sample size of patients available for subgroup analysis, particularly patients aged 12 years to 17 years, significantly limited the interpretation of findings and the assessment of treatment benefit in this subgroup of patients. Moreover, as indicated by the clinical expert, the potential heterogeneity of treatment effect by extent and severity of disease as measured by BSA or PASI was not reported. This further compromised the certainty of evidence on the treatment effect of roflumilast among patients by different extent and severity of disease.

External Validity

Based on clinical expert input, the inclusion and exclusion criteria were considered narrow. For example, patients with an IGA score of 1 and a PASI score of 1 would be considered candidates for treatment with roflumilast in clinical practice in Canada; however, they were excluded from the trials, as a score of at least 2 was required for enrolment. Additionally, the clinical expert suggested that patients with plaque psoriasis involving less than 2% and more than 20% of BSA, excluding the scalp, palms, and soles, would potentially be treated with roflumilast but were also excluded from the trials. As such, the effect of roflumilast in the broader patient population is unknown. Note that the Health Canada indication does not restrict the patient population according to percent of BSA involvement.

In both studies, roflumilast was compared with a vehicle that contained only the excipients of the roflumilast cream. However, given the wide range of topical treatment options currently available in clinical practice for plaque psoriasis, the clinical expert agreed that an active comparator would have been more appropriate; in particular, topical steroids and vitamin D analogues would have been appropriate comparators. However, the vehicle cream may be considered an appropriate comparator, as there are limited options for intertriginous areas, which have been identified as an area of unmet need by the clinical expert and clinician groups.

The clinician groups and the clinical expert agreed that the primary, secondary, and selected exploratory outcomes in the trials were clinically relevant (i.e., to capture the extent and severity of disease and determine treatment response in clinical practice). Moreover, the clinical expert indicated that clearance or near clearance of psoriasis lesions is commonly used in clinical practice to assess psoriasis severity, which is analogous to the IGA used in clinical trials. Both the clinician groups and the clinical expert indicated that tools such as the PASI and DLQI are not commonly used in clinical practice unless mandated by the payers for reimbursement.

Long-Term Extension Study

Description of Study

The DERMIS-OLE (ARQ-151-306)¹⁷ trial is an ongoing, phase III, long-term open-label extension (OLE) study conducted to assess the long-term safety of roflumilast in adult and pediatric patients with chronic plaque psoriasis involving up to 25% of BSA. All patients received open-label roflumilast applied once daily to all psoriasis lesions (excluding the scalp) for up to 24 weeks. The study enrolled 267 patients aged 2 years and older, 266 of whom received treatment with roflumilast in the OLE. Patients in cohort 1 (n = 264) successfully completed a prior roflumilast cream study in psoriasis in which they received either roflumilast (n = 171) or vehicle (n = 93). Patients in cohort 2 (n = 2) were naive to treatment with roflumilast and had not yet reached OLE study week 4 at the time of the data cut-off date; therefore, efficacy data were not available for these patients. At the time of data cut-off, a total of 222 patients (83.1%) had completed the OLE study, 32 patients (12.0%) had prematurely discontinued the study, and 12 patients (4.5%) were ongoing in the study.

Efficacy Results

Among patients in cohort 1, at week 24 of the OLE, 50.0% of patients had IGA scores of clear or almost clear. A total of 67.8% of cohort 1 patients had a clear or almost clear IGA status from primary baseline and maintained that status for a median of 93 days. At week 24, a total of 37.1% of cohort 1 patients experienced treatment success based on IGA score, defined by a score of clear or almost clear plus at least a 2-grade improvement from baseline, and this was the case for 54.9% of patients for a median duration of 85 days. For cohort 1 patients with intertriginous area involvement (n = 59), at week 24, 77.8% had I-IGA scores of clear or almost clear and 75.6% experienced treatment success based on I-IGA, defined as a score of clear or almost clear plus at least a 2-grade improvement from baseline. The proportions of cohort 1 patients with PASI 50, PASI 75, and PASI 100 at week 24 of the OLE were 70.5%, 43.8%, and 16.5%, respectively. The proportion of cohort 1 patients who experienced treatment success based on the WI-NRS at week 24, defined as an improvement of 4 or more points, was 62.4%. Overall, the results of the DERMIS-OLE study suggest that efficacy was maintained for up to 24 weeks.

Harms Results

No new safety signals were reported based on the OLE study. No AEs were reported among the patients in cohort 2. Among cohort 1 patients, 26.1% experienced at least 1 AE, the most common of which were sinusitis (2.7%), diarrhea (2.3%), COVID-19 (1.9%), and headache (1.9%). Three patients (1.1%) experienced a total of 5 SAEs, including polycythemia vera, COVID-19 pneumonia, palpitations, dehydration, and syncope, none of which were considered related to the drug. One patient (0.4%) discontinued the study due to an AE of application site irritation. No deaths occurred during the OLE study.

Critical Appraisal

The limitations of the extension study include selection bias, lack of a control group, and a lack of blinding. Reporting of harms and subjective efficacy outcomes such as IGA and I-IGA success may be biased by knowledge of treatment received. As only descriptive statistics were published in this interim report, and without comparator groups, the interpretation of the results is limited. Moreover, there is potential for selection bias, as patients who discontinued the parent studies due to AEs, lack of efficacy, or other reasons were excluded. Furthermore, most patients were white (> 84%), which may limit the generalizability of results to other racial groups. However, the clinical expert consulted by CADTH indicated that the efficacy and safety of roflumilast is not expected to differ by race.

Indirect Comparisons

Description of Studies

One sponsor-submitted indirect treatment comparison (ITC) was included, consisting of a systematic literature review and a Bayesian network meta-analysis (NMA) comparing roflumilast with the other topical therapies available in Canada in patients with plaque psoriasis.¹⁸ The primary outcome of interest was IGA treatment response at week 8, which was informed by data from 8 studies, including the DERMIS-1 and DERMIS-2 trials. A subgroup analysis of I-IGA treatment response at week 8 among patients with intertriginous psoriasis was informed by data in 4 studies, including the DERMIS trials.

Efficacy Results

The NMA results for IGA treatment response at week 8 favoured roflumilast versus vehicle, vitamin D analogues, tazarotene, or corticosteroids. IGA treatment response data for roflumilast versus corticosteroids plus vitamin D analogues or corticosteroids plus tazarotene did not clearly favour either treatment, as the 95% credible intervals (Crls) included the null value. For the fixed-effect NMA of I-IGA treatment response at week 8, roflumilast was associated with improved odds of treatment response versus vehicle ||| ||||||| || || |||| || |||||||. The results found that roflumilast versus calcineurin inhibitors did not clearly favour either treatment, as the 95% CrIs included the null value.

No harms or HRQoL outcomes were analyzed in the NMA.

Critical Appraisal

Potential sources of heterogeneity could not be fully assessed in the NMA due to the limited reporting of study design characteristics (i.e., inclusion criteria, frequency of treatment withdrawal, and handling of missing data) and patient baseline characteristics (i.e., disease history duration, prior treatment experience, PASI score, and BSA involvement) and, as such, there is uncertainty in the validity of the results of the NMA. The clinical expert consulted for this review noted there were imbalances across treatment groups in effect modifiers of baseline disease severity. As a result, it is possible that the heterogeneity in this baseline characteristic could result in changing the relative treatment effects. The outcomes were limited to the analysis of IGA and I-IGA treatment response; therefore, other relevant efficacy outcomes such as PASI and HRQoL (i.e., DLQI) were not assessed. Long-term efficacy and safety outcomes were not assessed, limiting the external validity of the results. These limitations result in uncertainty in the relative treatment effect estimates between roflumilast and other comparable topical therapies.

Studies Addressing Gaps in the Pivotal and Randomized Controlled Trial Evidence

No additional studies were included in this report for the review of roflumilast.

Conclusion

Two sponsor-submitted, phase III, randomized, double-blind, parallel-group, vehicle-controlled trials (the DERMIS-1 and DERMIS-2 trials) comparing topical treatment with roflumilast versus a matching vehicle in patients aged 2 years or older with chronic plaque psoriasis involving 2% to 20% (inclusive) of BSA, excluding the scalp, palms, and soles, demonstrated that 8 weeks of treatment with roflumilast reduced the overall severity of psoriasis, as measured by the proportion of patients experiencing treatment success based on the IGA compared with the matching vehicle. Roflumilast also demonstrated a reduction in the severity of psoriasis in intertriginous areas (identified as an area of unmet need), extent and severity of psoriasis, and severity of itch compared with the matching vehicle, as measured by the proportion of patients experiencing treatment success based on the I-IGA, a prespecified reduction in the PASI, and treatment success based on the WI-NRS, respectively, as well as improvement in QoL (specifically impacted by psoriasis) compared with the matching vehicle, as measured by PSD. There was supportive evidence on the overall treatment benefit of roflumilast versus vehicle on HRQoL, as measured by DLQI, and extent of disease, as measured by BSA. Of note, the effect of roflumilast in a broader patient population (e.g., those with < 2% and > 20% of BSA affected) and in combination therapy (and as adjunct therapy) in clinical practice is unknown, the line of therapy that would be supported by the evidence from both trials remains unclear, and the comparative efficacy of roflumilast versus other relevant comparators is uncertain in the absence of direct comparative evidence in the treatment of plaque psoriasis of the whole body. Indirect evidence provided by the sponsor-submitted NMA suggested that roflumilast may offer a benefit over vitamin D analogues, tazarotene, or corticosteroids based on IGA treatment response after 8 weeks of treatment, but the results did not clearly favour either treatment when compared with combination treatments (corticosteroids plus vitamin D analogues or corticosteroids plus tazarotene). Further, there was no clearly favoured treatment when compared with calcineurin inhibitors for I-IGA treatment response at week 8. The results of the NMA must be considered alongside potential uncertainty in the validity of the NMA results, as several potential sources of heterogeneity were not assessed, stratified analyses were not conducted to adjust for effect modifiers, and the similarity assumption of the NMA may have been violated; however, the extent of the potential bias is unknown. There were no safety or tolerability concerns associated with the use of topical roflumilast and no new safety signals were reported in the DERMIS-OLE study.

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 roflumilast cream 0.3% for topical use in the treatment of plaque psoriasis, including treatment of psoriasis in the intertriginous areas, in patients 12 years of age and older.

Disease Background

Contents within this section have been informed by materials submitted by the sponsor and clinical expert input. The following have been summarized and validated by the CADTH review team.

Psoriasis is a chronic inflammatory skin disease that is associated with multiple comorbidities, such as psoriatic arthritis, obesity, metabolic syndrome, hypertension, diabetes, and atherosclerotic disease.¹ Psoriasis is a complex, immune-mediated disease in which dysregulation of the immune system (involving interactions between dendritic cells, T-cells, neutrophils, and cytokines) and keratinocyte hyperproliferation contribute to the manifestation of the disease and the typical clinical findings of scaling, induration, and erythema.^1,2 Risk factors for psoriasis include genetics and environmental and behavioural factors. Psoriasis can present similarly in children and adults; the peak ages of onset are between 30 and 39 years and between 50 and 69 years. Of note, children with chronic plaque psoriasis are more likely to present with facial involvement compared with adults.¹

Based on a systematic review of worldwide literature, the estimated prevalence of psoriasis in adults ranged from 0.51% to 11.43% and from 0% to 1.37% in children.¹⁹ However, the prevalence of psoriasis varies geographically²⁰ and the authors acknowledged a geographic gap in knowledge, as the prevalence data were derived from only 20 countries.¹⁹ A panel of 11 dermatologists with expertise in psoriasis estimated that the median prevalence of psoriasis is 3% of the general adult population in Canada. The panel estimated that approximately 78% of patients with psoriasis have less than 10% of BSA involved (i.e., mild or moderate disease) and 22% of patients have 10% or more of BSA involved. The panel further estimated that 50% of patients have less than 3% of BSA involved (i.e., mild disease) and 2% of patients have more than 50% of BSA involved.³ Based on estimates in the US, 50% of patients have facial involvement⁴ and 21% to 30% of patients have intertriginous area involvement.⁵

Chronic plaque psoriasis (also known as psoriasis vulgaris⁶) is the most common clinical subtype of psoriasis, representing approximately 90% of patient cases in Canada.⁷ Plaque psoriasis is characterized by sharply defined, erythematous cutaneous plaques with overlying, coarse, silvery scales.¹ The diameter of plaques can range from less than 1 cm to more than 10 cm. Plaques can occur as single lesions or as generalized disease across wider areas of the body.⁶ Plaques can be asymptomatic; however, pruritus is commonly reported by patients and painful fissures can be present in palm or sole involvement. The scalp, elbows, knees, and gluteal cleft are common areas of involvement. Intertriginous areas (inverse or flexural psoriasis⁶), ear canals, umbilicus, palms, soles, and nails are possible areas of involvement. Intertriginous psoriasis is characterized by well demarcated, smooth, shiny plaques with no to minimal scales. Genital involvement, in particular, may have a negative impact on the patient’s QoL.¹

For most patients, the diagnosis of chronic plaque psoriasis can be made by history and physical examination. A skin biopsy may be performed, if required, such as in cases where the diagnosis remains uncertain following a history and physical examination of the patient or to rule out other conditions.¹

According to the Canadian Guidelines for the Management of Plaque Psoriasis,⁶ the definitions for mild, moderate, and severe plaque psoriasis based on measures of disease severity differ, depending on the setting (clinical trials versus clinical practice). For example, mild plaque psoriasis may be defined as an upper limit of 5% BSA involvement in clinical trials,²¹ while mild plaque psoriasis in clinical practice may be described as disease with a minimal impact on QoL (subjective evaluation of the impact of disease on the patient’s life), and the patient is able to have an acceptable level of symptomatic control with routine skin care and/or topical therapy (objective evaluation of the extent and symptoms of the disease).⁶

Standards of Therapy

Contents within this section have been informed by materials submitted by the sponsor and clinical expert input. The following have been summarized and validated by the CADTH review team.

Plaque psoriasis requires lifelong follow-up and treatment. Measures of treatment success in clinical practice may include clearance (absence of signs of disease), control (satisfactory response to therapy as defined by the patient and/or clinician), and remission (maintenance of disease control or suppression of signs and symptoms over time).⁶ The clinical expert consulted by CADTH for the purpose of this review identified the treatment goals in clinical practice as sufficient clearing of psoriasis lesions resulting in improvement in HRQoL and reduced symptoms (itch and pain) while minimizing drug-related adverse effects. According to the input from patient groups that was received for this submission, patients with psoriasis have identified the following as key outcomes of new treatments: effectiveness (i.e., treatment of publicly visible body areas to improve mental health, treatment of genital psoriasis to improve intimacy with loved ones, and treatment of symptoms, such as itching and burning, to resume normal daily activities), lack of side effects, and a medication that is conducive to their schedule and has the flexibility to be applied to all affected BSAs.

According to the Canadian Guidelines for the Management of Plaque Psoriasis,^6,8 the management of mild plaque psoriasis in clinical practice, specifically disease involving the trunk, limbs, and neck, includes topical therapies that can be broadly categorized as corticosteroids, vitamin D3 analogues, retinoids, anthralin (commercial formulations of anthralin are not currently available in Canada) and tars, as well as combination therapies. According to the clinical expert, anthralin and tars are not commonly used in clinical practice and they were not aware of any that are commercially available in Canada. Corticosteroids are the mainstay of therapy and are widely used for the topical treatment of psoriasis, as they are often effective and well tolerated. However, long-term use of topical corticosteroids and, in particular, high-potency steroids, is limited by their potential to produce side effects, for example, local cutaneous changes and hypothalamic-pituitary-adrenal axis suppression. Corticosteroids are available in a variety of formulations, including creams, ointments, gels, lotions, sprays, foams, and solutions.^6,8 Topical vitamin D3 analogues are most commonly associated with mild irritant contact dermatitis.⁶ The use of topical retinoids as monotherapy is limited by application site irritation (itching, burning, and erythema), a dose-dependent effect.⁶ Tars are often associated with staining and irritation and their use has declined since the introduction of therapeutic alternatives that are typically more acceptable to patients.⁶ Compared with monotherapy, combination therapy is generally more effective and can reduce the incidence of adverse effects. The therapeutic options for pediatric patients are generally similar to the options for adult patients.⁶

Management of intertriginous psoriasis, affecting the groin, axillae, inframammary region, abdominal body folds, gluteal cleft, perianal region, and retroarticular fold areas can be a challenge, as these areas are at an increased risk of adverse reactions to topical treatment, such as corticosteroids, because the skin tends to be thinner in these regions.⁶ Additionally, there are currently no available treatments indicated for intertriginous psoriasis. Although vitamin D analogues have demonstrated efficacy in facial and intertriginous psoriasis, they are associated with irritation and erythema; calcipotriol is not approved for use on the face or intertriginous areas, and calcitriol carries a warning against facial use. Although topical calcineurin inhibitors may be an appropriate treatment option for the management of intertriginous psoriasis, they are not approved for an indication in psoriasis (off-label use). These treatment options may be supplemented as needed for short periods with topical corticosteroids.^6,8

Many of the standard topical therapies for the treatment of mild chronic plaque psoriasis can result in good control of mild psoriatic disease but have distinct profiles of adverse effects, tolerability, and convenience which can have an impact on HRQoL, adherence, and effectiveness of treatment. It should also be noted that psoriatic presentation, personal values, psychosocial health, and treatment expectations can vary between patients.⁶ Consequently, the Canadian Guidelines for the Management of Plaque Psoriasis indicates that the selection of treatments should be individualized to the patient, with the patient participating in the choice of therapy and with consideration of the type of vehicle that would be acceptable to the patient to improve adherence, achieve treatment success, and support overall patient satisfaction.⁶

Moderate to severe chronic plaque psoriasis in clinical practice can be defined as psoriasis affecting the trunk and extremities that cannot be adequately controlled by the aforementioned approaches, with adequate control defined by the patient’s perception of the disease and its burdens.⁶ According to the Canadian Guidelines for the Management of Plaque Psoriasis, management of moderate to severe plaque psoriasis includes systemic therapy with traditional and biologic drugs, phototherapy, combination therapy, and topical therapy as adjunct therapy.⁶ For some patients with moderate to severe psoriasis, amelioration (short-term improvement and limited long-term disease control) may be considered an adequate treatment goal, while full clearance may be considered an appropriate treatment goal for many patients, particularly with the introduction of biologic therapy.⁶ The clinical expert consulted by CADTH for the purpose of this review indicated that phototherapy would not typically be used for intertriginous areas.

Of note, the clinical expert consulted by CADTH for the purpose of this review indicated that if lesions are cleared or almost cleared, but QoL continues to be impacted by the need to apply time-consuming and messy topical therapies, then phototherapy or systemic therapy may be considered.

Drug Under Review

Key characteristics of roflumilast are summarized in Table 3 along with other treatments available for plaque psoriasis.

Roflumilast cream 0.3% is indicated for topical treatment of plaque psoriasis, including treatment of psoriasis in the intertriginous areas, in patients 12 years of age and older; the recommended dosage is to apply to affected areas once daily. The Health Canada indication is aligned with the sponsor’s reimbursement request. Roflumilast cream has not been previously reviewed by CADTH. Roflumilast is a selective phosphodiesterase type 4 inhibitor and a nonsteroidal, anti-inflammatory drug, which is thought to exert its therapeutic action through inhibition of phosphodiesterase type 4 and subsequent inhibition of inflammatory markers associated with plaque psoriasis.²²

Table 3: Key Characteristics of Roflumilast and Other Relevant Topical Therapies for the Treatment of Plaque Psoriasis

ESRD = end-stage renal disease; HPA = hypothalamic-pituitary-adrenal; NA = not applicable; PDE-4 = phosphodiesterase type 4.

Sources: Canadian Guidelines for the Management of Plaque Psoriasis⁶ and the product monographs for roflumilast (Zoryve);²² calcitriol (Silkis);²³ calcipotriol (Dovonex);²⁴ halobetasol propionate and tazarotene (Duobrii);²⁵ and calcipotriol and betamethasone dipropionate ointment, gel, and aerosol foam (Dovobet, Enstilar).^26-28

Stakeholder Perspectives

Patient Group Input

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

Three patient groups submitted a joint patient input submission. CSPA, the Canadian Psoriasis Network, and the Canadian Association of Psoriasis Patients are national, not-for-profit organizations that are dedicated to improving the lives of people with psoriasis across Canada. A survey conducted between January and February 2023, in English and French, was hosted by the CSPA and it invited investigators to share the survey with their clinical trial participants. This survey collected a total of 86 responses from across Canada except for Yukon and Nunavut. Respondents were older than 65 years (26%), 55 to 64 years (21%), 25 to 34 years (19%), 35 to 44 years (16%), 45 to 54 years (16%), and younger than 18 years (3%).

A vast majority (94%) of respondents were living with psoriasis and 6% were caregivers or family members. The impact of psoriasis symptoms was considerable in various areas of patients’ lives, such as mental well-being (14 patients; 24%), daily tasks (10 patients; 17%), intimate relationships (8 patients; 13%), and social lives (7 patients; 12%). Patients also described the challenges of treating and living with their psoriasis as expensive and stressful.

Patients who responded to the survey hosted by the CSPA reported living with psoriasis of varying disease severity, with 26% reporting mild disease, 52% reporting moderate disease, and 22% reporting severe disease. Patients reported that the body areas most impacted by psoriasis include the scalp, followed by the legs, arms, genitals, hands, torso, skin folds, and palms, respectively. Most patients (90%) considered effectiveness as the most important factor to be addressed in any new treatment for psoriasis. In addition, 66% of patients valued lack of side effects, 60% prioritized affordability, 53% preferred treatments that are easy to apply, and 23% preferred treatments that are conducive to their schedules. Based on the survey, a substantial proportion of the respondents were using topical corticosteroids (36%), topical combination treatment (32%), and/or biological drugs (27%), which was the most effective current treatment, with 39% and 30% of patients reporting that biologics work well and very well, respectively. More than half of the respondents (53%; n = 33) in a survey reported facing side effects with the existing treatments available for psoriasis. The majority of patients with psoriasis (76%; n = 47) have discontinued their treatment at some point during their disease, with the most frequent reasons being that the treatment stopped working (61%; n = 37), caused side effects (47%; n = 28), was unaffordable (22%; n = 13), or ineffective (14%; n = 7). Among respondents, 10 patients reported experience with roflumilast, accessed through a clinical trial. Nine of these patients reported benefits of treatment, including clearing of skin, reduced itch and redness, clearing of skin lesions (plaques), and the treatment’s ease of application. All 10 respondents tolerated roflumilast well, except for 1 who experienced some itching.

The patient group input emphasized that psoriasis is a chronic and potentially debilitating disease that poses many challenges and is linked to anxiety, depression, and social isolation. This disease can interfere with social and intimate relationships, productivity, and family life and work life. Furthermore, due to the chronic nature of this disease, patients are concerned about recurrence and resistance to treatments in the future.

Clinician Input

Input From the Clinical Expert Consulted By CADTH

All CADTH review teams include at least 1 clinical specialist with expertise regarding the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., assisting in the critical appraisal of the sponsor’s summary 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 1 clinical specialist with expertise in the diagnosis and management of plaque psoriasis.

Unmet Needs

The clinical expert indicated that there are currently no available treatment options that can permanently reverse the course of plaque psoriasis, but systemic therapy can suppress psoriasis while it is being used.

The clinical expert also indicated that many patients do not respond to currently available topical treatments; in particular, widespread psoriasis and psoriasis of the scalp, palms, and soles are often refractory to topical treatment. In such cases, patients would need to weigh the associated risks and benefits of treatment with phototherapy or systemic therapy. Furthermore, over time, psoriasis is more likely to become refractory to systemic therapy than to topical therapy. The clinical expert noted that although topical therapies are generally well tolerated and that topical steroids are used over the long-term by many patients in clinical practice, clinical guidelines caution against long-term use of topical steroids for plaque psoriasis. Thus, the clinical expert agreed that more effective, safer, and better-tolerated treatment options are an unmet need, particularly for long-term use and in patients with facial, genital, and intertriginous involvement.

The clinical expert suggested that topical formulations that are less messy and more effective may improve adherence; in particular, in patients with intertriginous involvement who are offered topical tacrolimus, as it is only available as an ointment and not widely available because it is not indicated (used off-label) for the treatment of psoriasis.

Place in Therapy

The clinical expert considered roflumilast to be an anti-inflammatory drug, similar to other currently available topical treatments. For most patients with plaque psoriasis, the clinical expert suggested they would consider roflumilast as an alternative to other first-line topical treatments for psoriasis within the current stepped approach, as this could prevent having to step up patients to systemic therapy. For patients with facial, genital, and intertriginous involvement, the clinical expert suggested roflumilast would be a first-line treatment. The clinical expert further indicated that roflumilast would be preferred in scenarios where steroids should be avoided (e.g., intertriginous or facial involvement and due to patient preference). The clinical expert suggested that roflumilast may be preferred for long-term use in some patients and by clinicians, given that it is not a steroid.

For mild disease, the clinical expert suggested that roflumilast would be used as monotherapy. Additionally, depending on patient preference, roflumilast could be used simultaneously with other topical treatments, for example, roflumilast could be used for intertriginous areas and a separate topical treatment in a foam or lotion formulation could be used for the scalp. For extensive psoriasis, potent topical steroids would not be suggested and roflumilast would be considered a suitable topical alternative. However, most patients with extensive psoriasis will need phototherapy or systemic therapy to reach treatment targets. Roflumilast could also be used in combination with systemic therapy or phototherapy, as topical treatments are typically continued as needed.

For most patients with plaque psoriasis, the clinical expert indicated it would be appropriate to suggest treatment with topical steroids with or without vitamin D analogues before suggesting treatment with roflumilast, as they are considered effective and affordable. However, for patients with intertriginous psoriasis, the clinical expert suggested it is not necessary to offer other treatments before roflumilast, given the limited treatment options available specifically for intertriginous areas affected by psoriasis.

Patient Population

The clinical expert identified patients with active psoriasis, regardless of whether it is being actively treated or not, as the most in need of an intervention. The clinical expert noted there is a need for an intervention for the hair-bearing scalp area; however, they indicated that this need would not be addressed by roflumilast. Clinician examination and judgment, together with shared decision-making with the patient, would determine whether the drug is best suited for the patient. The clinical expert suggested that patients with psoriasis with limited BSA involvement are the most likely to respond to therapy with roflumilast as monotherapy.

Although the diagnosis is straightforward and biopsy tests are not needed for most patients, the clinical expert noted that psoriasis of special sites (scalp, intertriginous areas, palms, and soles) can be difficult to differentiate from other conditions, such as eczema; however, this is uncommon in clinical practice.

The clinical expert also noted it is possible for patients to present with only intertriginous or genital involvement but such patients may be underrecognized in clinical practice, as such presentations are often presented in the context of psoriasis on other areas of the body.

Assessing the Response to Treatment

The clinical expert indicated that clearance or near clearance of psoriasis lesions is commonly used to assess psoriasis severity; this is analogous to the IGA scales used in clinical trials. The clinical expert also indicated that patient satisfaction is important and is assessed in a gestalt manner that likely differs between physicians. Components of patient satisfaction can include improvement in symptoms (e.g., itch), the appearance of lesions, QoL, sexual function (relevant in intertriginous psoriasis), and how convenient and pleasant the treatment is from the patient’s perspective.

According to the clinical expert, extensive outcome scales, such as the PASI, DLQI, WI-NRS, and PSD are not typically used in clinical practice; however, the PASI and DLQI may be used if mandated by the insurer to fund a drug.

Discontinuing Treatment

According to the clinical expert, intolerance (e.g., stinging at the application site, odour, or unpleasant sensation when applying the medication) and ineffectiveness (i.e., minimal improvement after 2 to 8 weeks of application) are factors to consider when deciding to discontinue treatment with roflumilast.

The clinical expert further indicated that roflumilast would be discontinued when lesions have cleared and can be restarted if and when lesions have recurred. The clinical expert indicated that phototherapy or systemic therapy may be considered for extensive psoriasis regardless of improvement with roflumilast and, if there is substantial improvement with phototherapy or systemic therapy that obviates the need for topical therapy, then roflumilast would be considered for discontinuation.

Prescribing Considerations

The clinical expert suggested that roflumilast could be initiated and continued by general practitioners and specialists in any treatment setting, including community and hospital practices.

Clinician Group Input

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

Three different clinician groups provided input, Fraser Health Dermatology (4 clinicians), the Canadian Dermatology Association (3 clinicians), and the Atlantic Provinces Dermatology Association and Dermatology Association of Ontario (12 clinicians).

Clinicians have reached a consensus aligned with the clinical expert consulted by CADTH that there is a need for a new treatment as an alternative to existing therapies for plaque psoriasis, which include topical steroids, topical vitamin D analogues and steroid combinations, vitamin A analogues and steroid combinations, and so forth. The clinician groups mentioned different drawbacks for the currently available treatments, including AEs (e.g., topical corticosteroids), poor compliance (e.g., calcipotriol and betamethasone dipropionate ointment or calcipotriene and betamethasone dipropionate foam), difficult application, high cost for patients and the health care system (e.g., phototherapy and biologic therapies), limited efficacy (e.g., vitamin D analogues and steroid combinations), intolerability due to irritation (e.g., vitamin A analogues and steroid combinations) and inability to be administered to all areas of the body (e.g., topical corticosteroids). Lastly, patient compliance decreases when more than 1 topical therapy is prescribed.

The input from the Canadian Dermatology Association was aligned with the opinion of the clinical expert consulted by CADTH that traditional use of long-term steroids can be limiting, with potential local and systemic side effects. The clinician groups concurred that treatment goals include reducing the severity of symptoms, minimizing AEs, improving tolerability and efficacy, improving QoL, and reducing the burden on patients and health care systems.

Input from the Fraser Health Dermatology clinicians indicated topical roflumilast should be used after the failure of topical steroids (i.e., when a patient’s condition does not respond very well to steroids as a first line of treatment) because topical steroids are an affordable and usually well tolerated therapy, which was consistent with the input provided by the clinical expert consulted by CADTH. Of note, the Fraser Health Dermatology clinicians suggested roflumilast is the first choice for intertriginous areas due to the limited number of available treatments. In contrast, the Atlantic Provinces Dermatology Association and Dermatology Association of Ontario suggested prescribing roflumilast before the use of current combination therapies or as a first-line treatment for the management of mild to severe plaques on any area of the body due to the high clinical efficacy and favourable safety profile of roflumilast. The input noted that patients best suited for treatment with the drug under review included patients with mild to moderate psoriasis, patients with different disease phenotypes, patients with psoriasis that has failed to respond to topical steroids, and patients with intertriginous psoriasis, noting the importance of steroid sparing in these anatomic sites.

According to Fraser Health Dermatology and the Atlantic Provinces Dermatology Association, the best outcomes to determine treatment response would be PGA score and BSA involvement. The groups noted that the PASI score is not commonly used in clinical practice, except when applying for coverage for systemic therapies in patients with moderate to severe disease. Clinician groups noted that assessment of treatment goals is recommended after 8 weeks and that, typically, patients with psoriasis are initially seen every 3 to 6 months to assess response to treatment. The input received stated that patients with psoriasis are diagnosed and treated as outpatients, typically by both specialists and general practitioners or family doctors.

Drug Program Input

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

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

CDEC = CADTH Canadian Drug Expert Committee; IGA = Investigator’s Global Assessment; PDE-4 = phosphodiesterase type 4.

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 roflumilast cream 0.3% for topical use in the treatment of plaque psoriasis, including treatment of psoriasis in the intertriginous areas, in patients 12 years of age and older. The focus will be placed on comparing roflumilast with relevant comparators and identifying gaps in the current evidence.

A summary of the clinical evidence included by the sponsor in the review of roflumilast is presented in 3 sections, and CADTH’s critical appraisal of the evidence is included after each section. The first section, the systematic review, includes pivotal studies that were selected according to the sponsor’s systematic review protocol. The second section includes sponsor-submitted long-term extension studies. The third section includes indirect evidence from the sponsor.

Included Studies

Clinical evidence from 2 pivotal studies, 1 long-term extension study, and 1 ITC is included in the CADTH review and appraised in this document.

Pivotal Studies

Contents within this section have been informed by materials submitted by the sponsor. The following has been summarized and validated by the CADTH review team.

Description of Studies

Characteristics of the included studies are summarized in Table 5.

The primary objectives of the DERMIS-1 and DERMIS-2 studies were to assess the safety and efficacy of roflumilast cream 0.3% versus vehicle cream. The studies were identically designed phase III, randomized, parallel-group, double-blind, vehicle-controlled studies in which roflumilast or a matching vehicle was administered topically once daily for 8 weeks to patients with chronic plaque psoriasis involving 2% to 20% (inclusive) of BSA, excluding the scalp, palms, and soles (DERMIS-1, N = 439; DERMIS-2, N = 442). Each study was conducted at 43 different centres in the US and Canada; the DERMIS-1 trial included 8 centres in Canada and the DERMIS-2 trial included 10 centres in Canada.

After meeting all inclusion and none of the exclusion criteria, eligible patients were enrolled and randomized at baseline in a 2:1 ratio to roflumilast or matching vehicle according to a computer-generated randomization list. Randomization was stratified by study site, baseline IGA score (2 versus ≥ 3), and intertriginous involvement at baseline (I-IGA ≥ 2, yes versus no). Each kit, which contained tubes of either roflumilast or vehicle, was assigned to each patient using an internet-based randomization system.

The screening period was up to 5 weeks before the baseline visit followed by an 8-week treatment period and a follow-up visit that was scheduled approximately 1 week after treatment was completed for patients who did not enter the OLE study. Upon completing the treatment phase, patients were eligible to enrol in a separate OLE study, the DERMIS-OLE (ARQ-151-306)¹⁷ study, for up to 6 months. Patients were assessed at approximately 2-week intervals throughout the treatment period. The first patient was screened on December 9, 2019, in both studies and the last patient completed their last study visit on November 16, 2020, in the DERMIS-1 trial and November 23, 2020, in the DERMIS-2 trial.

Table 5: Details of Pivotal Studies Identified by the Sponsor

ACE = angiotensin-converting enzyme; BSA = body surface area; CDI-2 = Children’s Depression Inventory 2; CDLQI = Child Dermatology Life Quality Index; C-SSRS = Columbia-Suicide Severity Rating Scale; DLQI = Dermatology Life Quality Index; IGA = Investigator’s Global Assessment; I-IGA = intertriginous Investigator’s Global Assessment; mPASI 50 = 50% reduction in the modified Psoriasis Area and Severity Index score; mPASI 75 = 75% reduction in the modified Psoriasis Area and Severity Index score; mPASI 90 = 90% reduction in the modified Psoriasis Area and Severity Index score; mPASI 100 = 100% reduction in the modified Psoriasis Area and Severity Index score; OLE = open-label extension; PASI 50 = 50% reduction in the Psoriasis Area and Severity Index score; PASI 75 = 75% reduction in the Psoriasis Area and Severity Index score; PASI 90 = 90% reduction in the Psoriasis Area and Severity Index score; PASI 100 = 100% reduction in the Psoriasis Area and Severity Index score; PDE-4 = phosphodiesterase type 4; PHQ-8 = Patient Health Questionnaire depression scale; PHQ-A = Patient Health Questionnaire-9 modified for adolescents; PSD = Psoriasis Symptom Diary; PUVA = psoralen and UV A; SAE = serious adverse event; TEAE = treatment-emergent adverse event; UVB = UV B; WI-NRS = Worst Itch Numeric Rating Scale.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Populations

Inclusion and Exclusion Criteria

The same inclusion and exclusion criteria were used in the DERMIS-1 and DERMIS-2 studies (Table 5). Patients aged 2 years or older were required to have a clinical diagnosis of psoriasis vulgaris of at least 6 months duration (or 3 months for children) and stable disease for the past 4 weeks. Stable disease was determined by the trained investigator and was intended to indicate no escalation of treatment dose or additional therapy as well as no change in disease severity or extent during the 4 weeks before screening. Patients were required to have psoriasis vulgaris on the face, extremities, trunk, and/or intertriginous areas involving 2% to 20%, inclusive, of BSA (excluding the scalp, palms, and soles) and at least mild disease severity at baseline (IGA score ≥ 2 and PASI score ≥ 2).

Exclusion criteria included the presence of nonplaque or drug-induced forms of psoriasis, known or suspected severe renal insufficiency, and known or suspected moderate to severe liver impairment. Additionally, patients with planned initiation or changes to any concomitant medication that could affect psoriasis vulgaris (e.g., beta-blockers or angiotensin-converting enzyme [ACE] inhibitors) or who were unable to discontinue excluded (prohibited) medications and treatments (defined as systemic, biologic, topical, phototherapy, and investigational treatments) that could affect plaque psoriasis within the prespecified time periods (washout periods) before randomization were excluded from the studies (Table 5).

Interventions

The same interventions were used in the DERMIS-1 and DERMIS-2 studies (Table 5). Patients were randomized to receive roflumilast cream 0.3% or the matched vehicle cream, which were dispensed to the patient or caregiver in identical kits containing 4 45 g tubes. The kits and tubes were labelled in a blinded manner to ensure the patients, investigators, clinical personnel, and the sponsor were not aware of which treatment a patient received until the end of the study. The number of kits dispensed to each patient over the treatment period was based on the BSA involvement of psoriasis.

Patients were directed to apply their assigned treatment once daily topically to psoriatic plaques affecting 2% of BSA up to a maximum application area of 20% of BSA over an 8-week treatment period. Treatment application was in the evening, except on day 1 (baseline) and clinic visit days, when treatment was applied at the study site. Areas of application were all areas affected, including the face, trunk, genitals, intertriginous areas, and limbs, but excluding the scalp. A tar-containing or dandruff shampoo (zinc pyrithione or selenium sulphide) was permitted for treatment of the scalp. Palms and soles were treated but were not counted toward any efficacy measurements. Emollients or moisturizers were not to be applied to treated areas. Treatment of affected areas with roflumilast or vehicle was maintained for the duration of the study, regardless of whether the psoriasis cleared. New lesions that developed during the treatment period were also treated.

The matching vehicle cream contained only the excipients of the roflumilast cream. Of note, a vehicle control arm is distinct from a placebo control arm in that a vehicle control may provide beneficial effects on the signs and symptoms of plaque psoriasis (e.g., reducing scaling, improving skin barrier, and moisturizing the skin).

Treatment was interrupted for up to 1 week if the patient developed an application site reaction with the clinical appearance of an irritation reaction and a severity of a dermal response score of 5 (erythema, edema, and papules) or greater on the Berger and Bowman scale. Treatment was resumed if the reaction had, in the opinion of the investigator, adequately resolved.

Of note, treatment was stopped if, after consultation with the sponsor and medical monitor, a patient required a prohibited treatment or medication during the study period, such as systemic or topical treatments for plaque psoriasis. Any patients with a Columbia-Suicide Severity Rating Scale score indicative of suicidal ideation, a score of 15 or more on the Patient Health Questionnaire depression scale or Patient Health Questionnaire-9 modified for adolescents, a raw total score of 34 on the Children’s Depression Inventory 2, or a weight loss of 5% or more (if not dieting), were considered for treatment discontinuation by a mental health professional (if applicable), the sponsor, and at the discretion of the investigator.

Outcomes

A list of efficacy end points assessed in this Clinical Review Report is provided in Table 6. These end points are further summarized subsequently. The summarized end points comprise those included in the sponsor’s summary of clinical evidence as well as any identified as important to this review according to stakeholders, for example, the clinical expert, clinician groups, or patient groups.

Table 6: Outcomes Summarized From the DERMIS-1 and DERMIS-2 Trials

BSA = body surface area; CDLQI = Children’s Dermatology Life Quality Index; DLQI = Dermatology Life Quality Index; I-IGA = intertriginous Investigator’s Global Assessment; IGA = Investigator’s Global Assessment; NA = not applicable; mPASI 75 = 75% reduction in the modified Psoriasis Area and Severity Index score; mPASI 90 = 90% reduction in the modified Psoriasis Area and Severity Index score; mPASI 100 = 100% reduction in the modified Psoriasis Area and Severity Index score; PASI 50 = 50% reduction in the Psoriasis Area and Severity Index; PASI 75 = 75% reduction in the Psoriasis Area and Severity Index; PASI 90 = 90% reduction in the Psoriasis Area and Severity Index; PASI 100 = 100% reduction in the Psoriasis Area and Severity Index; PSD = Psoriasis Symptom Diary; SAE = serious adverse event; TEAE = treatment-emergent adverse event; WI-NRS = Worst Itch Numeric Rating Scale.

Notes: Upon successful testing of the primary end point, the alpha value was partitioned to test outcomes in partition 1 (alpha = 0.03) and partition 2 (alpha = 0.02).

The definitions of IGA and I-IGA success were identical except that the IGA considered the “whole body” (excluding the scalp, palms, and soles) and the I-IGA considered only the intertriginous regions.

Treatment-free interval and duration of response (remission) were identified as end points of interest by the clinical expert consulted by CADTH for this review, but were not measured in the DERMIS-1 and DERMIS-2 studies.

^aIGA success was defined as a score of 0 (clear) or 1 (almost clear) plus a ≥ 2-grade improvement from baseline.

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

^cI-IGA success was defined as a score of 0 (clear) or 1 (almost clear) plus a ≥ 2-grade improvement from baseline in patients with intertriginous area involvement and I-IGA ≥ 2 at baseline.

^dWI-NRS success was defined as a ≥ 4-point reduction in WI-NRS pruritus score from baseline in patients with a WI-NRS pruritus score ≥ 4 at baseline.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Efficacy Outcomes

A description of the efficacy outcome measures and their measurement properties that were used in both the DERMIS-1 and DERMIS-2 studies and assessed in this Clinical Review Report are presented in Table 8.

Investigator’s Global Assessment

In the DERMIS-1 and DERMIS-2 studies, IGA success at weeks 8 and 4 were primary and secondary end points, respectively, and I-IGA success at weeks 8 and 4 were secondary and exploratory end points, respectively. IGA and I-IGA (the latter defined as the IGA scale but used to evaluate the intertriginous areas only) success was defined as a score of clear or almost clear plus an improvement of 2 grades or more from baseline. The IGA was completed at screening, at baseline and every 2 weeks thereafter during the 8-week treatment period, and at follow-up. The I-IGA was completed at baseline and every 2 weeks thereafter during the 8-week treatment period and at follow-up.

The IGA is an investigator-reported, static (i.e., an assessment is made at a static moment in time without comparing it with baseline disease severity),¹² qualitative evaluation of the overall severity of psoriasis of the whole body. The disease is assessed on an ordinal scale with 5 grades of psoriasis severity: clear skin (0), almost clear (1), mild (2), moderate (3), and severe (4). Each grade is defined by distinct morphologic descriptions of plaque thickening, scaling, and erythema (Table 7).

Of note, the PGA has been referred to as the IGA, both of which are used to evaluate the overall severity of psoriasis. Multiple versions of the static IGA and PGA have been described in psoriasis studies, ranging from a 4-point scale (clear to severe) to an 11-point scale (worst imaginable disease to absence of disease activity). The static IGA and PGA evaluate disease severity at a single point in time and do not depend on recalling baseline disease characteristics. It should be noted that different versions of the IGA and PGA use different descriptions for each grade (e.g., a score of 1 may indicate minimal in 1 version of the scale but indicate almost clear in a different version of the scale), and the criteria for assessment for each descriptor may vary. No single IGA or PGA version has been recognized as the standard scale.¹¹

Table 7: Investigator’s Global Assessment of Disease in the DERMIS-1 and DERMIS-2 Trials

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials.¹⁰

Psoriasis Area and Severity Index

In the DERMIS-1 and DERMIS-2 studies, time to PASI 50 (a 50% reduction in PASI score) from baseline was a secondary end point. A 75% reduction in PASI score (PASI 75) from baseline at weeks 8 and 4 were secondary and exploratory end points, respectively. A 75% reduction in modified PASI score (mPASI 75), 90% reduction in PASI score (PASI 90), 90% reduction in mPASI score (mPASI 90), 100% reduction in PASI score (PASI 100), and 100% reduction in mPASI score (mPASI 100) from baseline at week 8 were exploratory end points. The PASI and mPASI were assessed at screening, at baseline and every 2 weeks thereafter during the 8-week treatment period, and at follow-up.

The PASI is an investigator-reported, qualitative evaluation of the severity of psoriasis. Disease severity is evaluated by combining the assessment of the severity of lesions and the area affected into a single score ranging from 0 (no psoriasis) to 72 (maximal psoriasis). To determine the PASI score, the percent of disease in the involved area is estimated and transformed into a grade ranging from 0 (0% of involved area) to 6 (90% to 100% of involved area) for the head (10% of skin), arms (20% of skin), trunk (30% of skin), and legs (40% of skin). Within each anatomic area, the severity is estimated on a scale of 0 (none) to 4 (maximum severity) for each clinical sign, erythema (redness), induration (thickness), and desquamation (scaling), and summed. The final PASI score is calculated by combining the severity rating of the lesions of each anatomic area weighted by the area of involvement and the respective proportion of the patient’s skin. To reflect the status of limited disease, an mPASI score was calculated for patients with an anatomic area with less than 10% involvement using the actual percentage of the involved anatomic area.

Of note, the proportion of patients who had a PASI 75 (at least a 75% reduction in PASI score from baseline) is commonly used as an efficacy end point in clinical trials of biologic therapies for psoriasis; PASI 90 (at least a 90% reduction in PASI score from baseline) and PASI 100 (at least a 100% reduction in PASI score from baseline) are occasionally reported, as well.¹¹

Psoriasis Symptom Diary

In the DERMIS-1 and DERMIS-2 studies, change from baseline in the PSD total score at weeks 8 and 4 were secondary end points. The PSD was completed at screening and at baseline and every 2 weeks thereafter during the 8-week treatment period.

The PSD score is a patient-reported assessment of the impact of plaque psoriasis on overall QoL and was completed by adult patients (aged 18 years or older at screening) in the DERMIS-1 and DERMIS-2 studies.³¹ The patient’s experience of symptom severity, bother, embarrassment, hiding affected skin, and avoiding activities is recorded using a 24-hour recall period. Responses are scored on a scale of 0 (least severe) to 10 (most severe) for each question and the total PSD score is the sum of the scores for all 16 questions, with lower scores indicating less severe or bothersome symptoms.

Worst Itch Numerical Rating Scale

In the DERMIS-1 and DERMIS-2 studies, WI-NRS success at weeks 8, 4, and 2 were secondary end points. WI-NRS success was defined as an improvement of 4 points or more from a baseline score of 4 or more (moderate or worse itch at baseline). The WI-NRS was assessed at screening and baseline and every 2 weeks thereafter during the 8-week treatment period.

The WI-NRS is a patient-reported, single-item tool used to assess the severity of itch and was completed by patients aged 12 years or older in the DERMIS-1 and DERMIS-2 studies. The highest intensity of itch experienced in the previous 24-hour period is rated on a scale of 0 (no itch) to 10 (worst imaginable itch).³²

DLQI and Children’s DLQI

In the DERMIS-1 and DERMIS-2 studies, absolute change in the DLQI or CDLQI at week 8 was an exploratory end point. The DLQI or CDLQI were assessed at screening, baseline, and weeks 2, 4, and 8 during the treatment period.

The DLQI and CDLQI are patient-reported tools used to evaluate HRQoL in patients aged 17 years or older and in patients aged 2 to 16 years, respectively. These 10-item questionnaires cover 6 domains (symptoms and feelings, daily activities, leisure, work and school, personal relationships, and bother with psoriasis treatment) that are scored on a 4-point Likert scale (0 = not at all; 3 = very much). The total DLQI score is calculated by summing the scores of each item and ranges from 0 to 30 points, with lower scores indicating greater HRQoL.

Body Surface Area

In the DERMIS-1 and DERMIS-2 studies, percent change in BSA affected by psoriasis at week 8 was an exploratory end point. BSA was assessed at screening, at baseline and every 2 weeks thereafter during the 8-week treatment period, and at follow-up.

The BSA affected by psoriasis was determined by the patient hand method, where the surface area of the patient’s hand (including fingers) was assumed to equal 1% of BSA, as assessed by the investigator.

Harms Outcomes

In the DERMIS-1 and DERMIS-2 studies, AEs were assessed at screening, at baseline and every 2 weeks thereafter during the 8-week treatment period, and at follow-up.

Local Tolerability Assessment

In the DERMIS-1 and DERMIS-2 studies, local tolerability assessments were completed at baseline, week 4, and week 8.

Local tolerability was graded by investigators before treatment application in the clinic using a scale reported by Berger et al.³³ Reactions at the application site that could have occurred postbaseline were differentiated from the pre-existing inflammation associated with the patient’s psoriasis. Dermal responses were graded on a scale of 0 (no evidence of irritation) to 7 (strong reaction spreading beyond the application site), with other qualitative effects noted (slight glazed appearance, marked glazing, glazing with peeling and cracking, glazing with fissures, film of dried serous exudates, small petechial erosions and/or scabs, or no other effects).

Local tolerability was also graded by patients or their caregivers in the clinic on a scale of 0 (none, i.e., no sensation) to 3 (severe, i.e., hot, tingling, or stinging sensation that has caused definite discomfort) based on the sensation 10 to 15 minutes after treatment application.

Table 8: Summary of Outcome Measures and Their Measurement Properties

BSA = body surface area; DLQI = Dermatology Life Quality Index; HRQoL = health-related quality of life; ICC = interclass correlation; IGA = Investigator’s Global Assessment; I-IGA = intertriginous Investigator’s Global Assessment; LS-IGA = Lattice System Investigator’s Global Assessment; MID = minimal important difference; mPASI = modified version of the PASI score; PASI = Psoriasis Area Severity Index; PGA = Physician’s Global Assessment; PGIC = Patient Global Impression of Change; PSD = Psoriasis Symptoms Diary; SF-36 = Short Form (36) Health Survey; QoL = quality of life; WI-NRS = Worst Itch Numeric Rating Scale.

Statistical Analysis

The statistical analysis of the efficacy end points in the DERMIS-1 and DERMIS-2 studies and assessed in this Clinical Review Report are summarized in Table 9.

Sample Size and Power Calculation

A sample size of approximately 400 patients was planned for each study: approximately 267 patients would receive roflumilast once daily and approximately 133 patients would receive the matching vehicle once daily according to the 2:1 randomization scheme. This sample size would provide more than 99% power to detect a 22.4% difference in the primary end point (IGA success) between treatment groups at an alpha of 0.05 using a 2-sided chi-square test. The intergroup difference was estimated from a phase IIb study (ARQ-151-201, NCT03638258) that compared IGA success rates between roflumilast cream 0.3% and vehicle cream (32.2% versus 9.8%, respectively). The planned sample size would also provide sufficient power to analyze the first 5 secondary end points (Table 5).

Statistical Test or Model

Primary Outcome

The primary efficacy end point (IGA success defined as an IGA score of 0 [clear] or 1 [almost clear] plus an improvement of 2 or more grades from baseline at week 8) was evaluated in the ITT population as the ratio of the odds of IGA success using roflumilast relative to the odds of IGA success using the vehicle after 8 weeks. The primary end point was analyzed using a Cochran-Mantel-Haenszel test with a 2-sided significance level of 5% and stratification by study site, baseline IGA, and baseline intertriginous area involvement. For the analysis of the primary efficacy end point, missing IGA scores were imputed using a regression-based multiple imputation model.

A sensitivity analysis of the primary end point was performed using the original (nonimputed, observed data) dataset, including a repeated-measures logistic regression model with IGA success as the dependent variable and treatment and visit as the independent variables, and a tipping point analysis (post hoc analysis). The primary efficacy analysis and sensitivity analyses were also performed for the primary end point based on the modified ITT population using both observed values and multiple imputation. All other analyses of the primary end point did not adjust for missing data (i.e., used observed data only).

Multiple Testing Procedure

For all primary and secondary efficacy end points, a familywise type I alpha error rate of 0.05 was maintained using a prespecified hierarchical testing strategy. End points were assigned into testing families and testing only proceeded to the next family in accordance with the sequential testing strategy rules.

If the primary end point was met (i.e., a statistically significant improvement in IGA success rate with roflumilast cream 0.3% at week 8), the secondary end points were to be tested inferentially. To control the familywise type I error for multiple comparisons among the secondary end points, the following multiplicity procedure was used.

Upon successful testing of the primary end point (family 1), the alpha was partitioned into 0.03 (partition 1) to sequentially test the secondary end points in families 2 and 3 and into 0.02 (partition 2) to sequentially test WI-NRS time points.

Within the partition 1 alpha of 0.03, the hierarchical testing of family 2 (time to PASI 50, PASI 75 rate at week 8, and I-IGA success at week 8) was performed. Subsequent testing of family 3 (change from baseline in total PSD scores at weeks 8 and 4, and IGA success at week 4) was performed using the alpha available after the testing of family 2, with the Holm procedure used to control for multiple comparisons in family 3.

Within the partition 2 alpha of 0.02, sequential testing was performed for WI-NRS success rates at week 8, then at week 4, and then at week 2.

Subgroup Analyses

Prespecified subgroup analyses were performed for patients with intertriginous area involvement and at least mild severity (i.e., the I-IGA success rate could be assessed) and patients with pruritus who had a WI-NRS score of 4 points or more at baseline (i.e., success in reducing pruritus could be assessed). Both subgroup analyses accounted for multiplicity through alpha partitioning within the hierarchical testing strategy described previously.

Subgroup analyses of the primary end point, which did not account for multiplicity, were also performed using observed data for the following subgroups:

• topical corticosteroid — inadequate response, intolerance, or contraindication (yes or no)

• topical vitamin D derivatives — inadequate response, intolerance, or contraindication (yes or no)

• apremilast — patients with any prior use or no prior use

• psoriasis involvement on the elbow (yes)

• psoriasis involvement on the knee (yes)

• psoriasis involvement on the knee or elbow (either one)

• study sites

• age group — 2 to 11 years, 12 to 17 years, and 18 years and older

  • The permitted age range in the DERMIS-1 and DERMIS-2 studies was 2 years old and above, which extends beyond the Health Canada indication in patients 12 years of age and older. Therefore, results from the subgroup analysis for the age group of 2 to 11 years were not summarized in this report.

• involvement of face or intertriginous area at baseline

• baseline BSA severity — mild (< 5% of BSA), moderate (≥ 5% to < 10% of BSA), and severe (≥ 10% of BSA).

Of the subgroup analyses of the primary end point performed in the studies listed previously, the last 3 subgroups were identified as clinically relevant by the clinical expert consulted by CADTH for this review.

Secondary Outcomes

Binary secondary end points were also analyzed using the Cochran-Mantel-Haenszel test stratified by study site, baseline IGA, and baseline intertriginous area involvement. The binary secondary end points included PASI 75 at week 8, I-IGA success at week 8 (stratified by study site and baseline I-IGA), IGA success at week 4, and WI-NRS success at weeks 8, 4, and 2. Analyses of I-IGA and WI-NRS were performed using the I-IGA ITT and PRU4 ITT populations, respectively; all other analyses were performed on the ITT population.

The continuous secondary end points, change from baseline in total PSD score at week 4 and week 8, were analyzed using an analysis of covariance with the following factors: treatment, study site, baseline IGA, baseline intertriginous area involvement, and baseline of the variable being analyzed. The LS means, standard errors, 95% CIs, and P values were presented. These analyses were performed on the ITT population.

Time to PASI 50 was estimated using the KM method with results calculated as the median and 95% CI for each treatment group, which were compared using the log-rank statistic. A Cox proportional hazards model that included the stratification factors (study site, baseline IGA, and baseline intertriginous area involvement) was used to estimate the hazard ratio and associated 95% CIs. These analyses were performed on the ITT population.

For all secondary efficacy end points other than time to PASI 50, analyses were performed using both multiple imputation for missing data in the primary analysis and observed data. Only observed data were included in the descriptive statistics.

Other Outcomes

Other binary outcomes (e.g., IGA score of clear, mPASI 75, PASI 90, mPASI 90, PASI 100, and mPASI 100) were analyzed using the Cochran-Mantel-Haenszel model and the aforementioned adjustment factors on the ITT population.

Other continuous outcomes (e.g., percent change from baseline in the affected percent of BSA, change in DLQI, and change in CDLQI) were analyzed using an analysis of covariance model and the adjustment factors on the ITT population described previously.

Table 9: Statistical Analysis of Efficacy End Points in the DERMIS-1 and DERMIS-2 Trials

ANCOVA = analysis of covariance; BSA = body surface area; CMH = Cochran-Mantel-Haenszel; CDLQI = Children’s Dermatology Life Quality Index; DLQI = Dermatology Life Quality Index; IGA = Investigator’s Global Assessment; I-IGA = intertriginous Investigator’s Global Assessment; NA = not applicable; mPASI 75 = 75% reduction in the modified Psoriasis Area and Severity Index score; mPASI 90 = 90% reduction in the modified Psoriasis Area and Severity Index score; mPASI 100 = 100% reduction in the modified Psoriasis Area and Severity Index score; PASI 50 = 50% reduction in the Psoriasis Area and Severity Index; PASI 75 = 75% reduction in the Psoriasis Area and Severity Index; PASI 90 = 90% reduction in the Psoriasis Area and Severity Index; PASI 100 = 100% reduction in the Psoriasis Area and Severity Index; PSD = Psoriasis Symptom Diary; WI-NRS = Worst Itch Numeric Rating Scale.

Note: The definitions of IGA and I-IGA success were identical except that the IGA considered the “whole body” (excluding the scalp, palms, and soles) and the I-IGA considered only the intertriginous regions.

^aIGA success was defined as a score of 0 (clear) or 1 (almost clear) plus a ≥ 2-grade improvement from baseline.

^bI-IGA success was defined as a score of 0 (clear) or 1 (almost clear) plus a ≥ 2-grade improvement from baseline in patients with intertriginous area involvement and I-IGA ≥ 2 at baseline.

^cWI-NRS success was defined as a ≥ 4-point reduction in WI-NRS pruritus score from baseline in patients with WI-NRS pruritus score ≥ 4 at baseline.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Analysis Populations

The definition and application of each analysis population in the DERMIS-1 and DERMIS-2 studies are summarized in Table 10.

The ITT analysis set included all randomized patients. The modified ITT analysis set included all randomized patients except for those who missed the week 8 disease assessment specifically due to the COVID-19 disruption.

The I-IGA ITT analysis set was a subset of the ITT population and included patients with intertriginous area involvement and with at least mild severity of the intertriginous lesions (I-IGA ≥ 2) at baseline. The PRU4 ITT analysis set was a subset of the ITT population and included patients with a WI-NRS pruritus score of 4 or more points at baseline.

The safety analysis set included all patients who were enrolled and received at least 1 confirmed dose of roflumilast or the vehicle.

Table 10: Analysis Populations in the DERMIS-1 and DERMIS-2 Trials

IGA = Investigator’s Global Assessment; I-IGA = intertriginous Investigator’s Global Assessment; ITT = intention-to-treat; mITT = modified intention-to-treat; PRU4 = patients with WI-NRS pruritus score ≥ 4 at baseline; WI-NRS = Worst Itch Numeric Rating Scale.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Results

Patient Disposition

Patient disposition was generally consistent between the DERMIS-1 and DERMIS-2 studies (Table 11).

In the DERMIS-1 trial, 539 patients were screened and 439 patients were enrolled; the reasons for screening failure were not reported. A total of 439 patients were randomized in the DERMIS-1 trial, 286 patients to roflumilast cream 0.3% and 153 patients to the vehicle. Similarly, 557 patients were screened and 442 patients were enrolled in the DERMIS-2 trial; the reasons for screening failure were also not reported. A total of 442 patients were randomized in the DERMIS-2 trial, 290 patients to roflumilast cream 0.3% and 152 patients to the vehicle.

The proportion of patients who discontinued from the DERMIS-1 trial was 10.8% and 13.1% of the patients randomized to roflumilast and vehicle, respectively. Similarly, 9.0% and 13.8% of patients randomized to roflumilast and vehicle, respectively, discontinued from the DERMIS-2 trial. Withdrawal by patient in the DERMIS-1 trial was reported in 3.8% versus 7.2% of patients randomized to roflumilast and vehicle, respectively. Similarly, withdrawal by patient in the DERMIS-2 trial was reported in 3.4% versus 7.2% of patients randomized to roflumilast and vehicle, respectively.

Lost to follow-up in the DERMIS-1 trial was reported in 4.2% and 2.6% of patients randomized to roflumilast and vehicle, respectively. Similarly, 5.2% and 4.6% of patients randomized to roflumilast and vehicle, respectively, were lost to follow-up in the DERMIS-2 trial. Study discontinuation as a result of any AEs in the DERMIS-1 trial was reported in 1.7% and 1.3% of patients randomized to roflumilast and vehicle, respectively. Similarly, 0.3% and 1.3% of patients randomized to roflumilast and vehicle, respectively, discontinued the study as a result of any AEs in the DERMIS-2 trial.

Table 11: Summary of Patient Disposition From the DERMIS-1 and DERMIS-2 Trials (All Randomized Patients)

ITT = intention-to-treat; I-IGA = intertriginous Investigator’s Global Assessment; NR = not reported; PRU4 = patients with a WI-NRS pruritus score of ≥ 4 at baseline; WI-NRS = Worst Itch Numeric Rating Scale.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Protocol Deviations

A summary of the protocol deviations in the DERMIS-1 and DERMIS-2 trials is presented in Table 12.

Overall, the proportion of randomized patients with any protocol deviation was 54.2% in the roflumilast arm versus 62.7% in the vehicle arm in the DERMIS-1 trial, and 59.0% in the roflumilast arm and 60.5% in the vehicle arm in the DERMIS-2 trial. The proportion of patients with any protocol violation related to COVID-19 disruption was numerically greater in the DERMIS-1 trial than in the DERMIS-2 trial (10.3% of all randomized patients versus 4.3% of all randomized patients, respectively), but was generally well balanced between treatment arms within each study.

In the DERMIS-1 trial, major protocol deviations were reported in 26.9% of patients in the roflumilast arm versus 33.3% in the vehicle arm. In the DERMIS-2 trial, major protocol deviations were reported in 22.4% of patients in the roflumilast arm versus 26.3% in the vehicle arm. The most common major protocol deviation reported in both studies was efficacy assessment (14.4% of all randomized patients in the DERMIS-1 trial and 14.9% of all randomized patients in the DERMIS-2 trial); the proportion of randomized patients with a major protocol deviation related to efficacy assessment was generally well balanced between the treatment arms within each study and across studies. This was due to a workflow issue with the electronic clinical outcomes assessment tablet that allowed the study coordinator to remain logged into the tablet while the investigator was performing efficacy assessments. A reconciliation was performed to identify trained and approved investigators who had conducted end point assessments while the study coordinator was logged into the system. At the study sites where this issue was identified, the principal investigators signed affidavits to confirm that the investigator performed the assessment at the time of entry.

Table 12: Summary of Protocol Deviations in the DERMIS-1 and DERMIS-2 Trials (Safety Population)

^aDeviations related to the study drug due to intervention noncompliance (i.e., less than 80% of doses achieved over the duration of the study or more than 3 consecutive missed doses).⁴⁷

^bOther deviations included application of the long-term extension intervention at week 8 (DERMIS-1) and intervention tubes or kits lost during the study or not returned at the study completion visit.⁴⁷

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials.¹⁰

Baseline Characteristics

Note, all patients randomized in the DERMIS-1 and DERMIS-2 trials received at least 1 dose of the intervention; thus, the ITT population and the safety population were identical in each study.

Patients in the DERMIS-1 and DERMIS-2 studies had generally consistent demographic characteristics (Table 13). The proportion of patients in the age category of 12 to 17 years was similar across both studies (|||| || |||| of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively). The proportion of patients in the age category of 18 years and older was also similar across both studies (||||| || ||||| of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively). The mean age of all randomized patients was 48.0 years (SD = 14.69) in the DERMIS-1 trial and 47.0 years (SD = 14.72) in the DERMIS-2 trial. Most patients were male (64.9% and 62.4% of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively), while 33.9% to 39.3% of patients were female. The majority of patients were white (81.5% and 82.8% of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively), while the remainder of the randomized patients (0% to 7.3%) identified as Asian, Black or African American, American Indian or Alaska Native [wording from original source], Native Hawaiian or other Pacific Islander, other, or more than 1 race.

Patients in the DERMIS-1 and DERMIS-2 studies had generally consistent clinical characteristics (Table 14). The mean baseline of BSA covered (involved) in all randomized patients was 6.66% (SD = 4.538) in the DERMIS-1 trial and 7.30% (SD = 4.918) in the DERMIS-2 trial. Across both studies, the majority of patients had moderate IGA at baseline (74.7% and 76.5% of randomized patients in the DERMIS-1 and DERMIS-2 trial, respectively). A similar proportion of all randomized patients in the DERMIS-1 and DERMIS-2 trials had facial involvement (27.1% and 26.0%, respectively) and genital involvement (16.4% and 14.7%, respectively). Across both studies, the majority of patients had mild to moderate I-IGA at baseline (mild, 11.2% and 8.6% of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively; moderate, 9.8% and 10.0% of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively). The mean PASI at baseline for all randomized patients was 6.5 (SD = 3.35) in the DERMIS-1 trial and 6.7 (SD = 3.33) in the DERMIS-2 trial.

Across both studies, the majority of patients had a baseline WI-NRS score of 4 or more points (75.9% and 78.1% of randomized patients in the DERMIS-1 and DERMIS-2 trial, respectively) (Table 14). Note, the mean baseline WI-NRS score was 5.7 (SD = 2.78) in the DERMIS-1 trial and 5.9 (SD = 2.66) in the DERMIS-2 trial across treatment arms. The mean PSD total score at baseline was 72.6 (SD = 42.21) in the DERMIS-1 trial, while in the DERMIS-2 trial, the mean baseline PSD total score was 69.3 (40.66) in the roflumilast arm and 77.4 (41.24) in the vehicle arm. The mean DLQI score at baseline for all randomized patients was 7.3 (SD = 5.47) in the DERMIS-1 trial and 7.2 (SD = 5.61) in the DERMIS-2 trial.

Data on treatment history of psoriasis and data on the number of prior treatments for psoriasis were limited. The majority of patients (85.6% and 84.8% of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively) reported at least 1 prior medication (Table 14.) Therefore, approximately 15% of randomized patients in each study did not receive any prior medication, irrespective of plaque psoriasis indication (i.e., treatment-naive patients). Most patients across studies, 56.3% and 66.3% of randomized patients in the DERMIS-1 and DERMIS-2 trials, respectively, reported inadequate response, intolerance, or contraindication to topical corticosteroids for the treatment of psoriasis.

In both the DERMIS-1 and DERMIS-2 trials, the demographic and clinical characteristics of patients in the I-IGA ITT and PRU4 ITT populations were generally balanced between treatment arms and similar to the ITT population.

The baseline characteristics outlined in the following tables are limited to those which are most relevant to this review or were felt to impact the outcomes or interpretation of the study results.

Table 13: Summary of Baseline Demographic Characteristics of the DERMIS-1 and DERMIS-2 Trials (Safety Population)

SD = standard deviation.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Table 14: Summary of Baseline Clinical Characteristics of the DERMIS-1 and DERMIS-2 Trials (Safety Population)

BMI = body mass index; BSA = body surface area; CDLQI = Children’s Dermatology Life Quality Index; DLQI = Dermatology Life Quality Index; I-IGA = intertriginous Investigator’s Global Assessment; IGA = Investigator’s Global Assessment; mPASI = modified Psoriasis Area and Severity Index; PASI = Psoriasis Area and Severity Index; PSD = Psoriasis Symptom Diary; SD = standard deviation; WI-NRS = Worst Itch Numeric Rating Scale.

^aWI-NRS was determined by asking the patient to assess their worst itch over the past 24 hours. The scale was from 0 to 10, which ranged from “no itch” to “worst imaginable itch.”

^bPrior medications are all medications that were started before the application of the assigned treatment.

^cPatients could be counted more than once if they had received multiple treatment types.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Exposure to Study Treatments

The extent of exposure to treatment was based on the total number of applications and total weight of treatment; patients were considered compliant if they (or their caregiver) applied 80% or more of the expected applications and they did not miss more than 3 consecutive doses. Treatment compliance rates were generally consistent across studies and were greater than 90% for each treatment arm (Table 15).

The mean number of applications per patient during the treatment period was also generally consistent between the DERMIS-1 (55.1 applications and 54.4 applications in the roflumilast and vehicle arms, respectively) trial and DERMIS-2 trial (54.3 applications and 53.0 applications in the roflumilast and vehicle arms, respectively). Of note, the mean total weight of roflumilast and vehicle applied was 122.73 g (SD = 113.114) and 179.89 g (SD = 183.030), respectively, in the DERMIS-1 trial. Similarly, 122.02 g (SD = 122.465) of roflumilast versus 187.78 g (SD = 180.495) of vehicle was applied in the DERMIS-2 trial (Table 15).

Concomitant Medications

Note, topical and systemic treatments as well as phototherapy for plaque psoriasis were considered prohibited medications and treatments and were not permitted in the DERMIS-1 and DERMIS-2 studies. Eye and ear drops as well as nasal corticosteroid preparations were permitted. Inhaled corticosteroid preparations were permitted if used for a stable condition and at a stable dose for more than 28 days before screening and were continued at the same dose for the duration of the study. Nonmedicated emollients, moisturizers, and sunscreens were permitted as used normally by patients. These could be applied to nontreated areas as needed but should not have been used within 12 hours of a study visit. A tar-containing or dandruff shampoo (zinc pyrithione or selenium sulphide) was permitted for treatment of the scalp.

The addition of new medications, including nonprescription medications, during the study was generally discouraged; however, short-term use of a medication could be authorized by the investigator and chronic medication use was permitted during the study (except as prohibited according to the prespecified list of excluded medications and treatments). At least 1 concomitant medication was used by ||||| of randomized patients in the DERMIS-1 trial (||||| in the roflumilast arm and ||||| in the vehicle arm) and by 67.0% of randomized patients in the DERMIS-2 trial (||||| in the roflumilast arm and ||||| in the vehicle arm). The most common (frequency of 10% or greater) classes of concomitant medications in the DERMIS-1 trial were 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (|||||), propionic acid derivatives (|||||), and ACE inhibitors (|||||); note ||||| of all randomized patients used ACE inhibitors as a prior medication. Similarly, the most common (frequency of 10% or greater) classes of concomitant medications in the DERMIS-2 trial were propionic acid derivatives (|||||) followed by ACE inhibitors (||||); note, |||| of all randomized patients used ACE inhibitors as a prior medication (Table 16).

Table 15: Summary of Patient Exposure From the DERMIS-1 and DERMIS-2 Trials (Safety Population)

SD = standard deviation.

^aTotal weight was determined by subtracting returned tube weight from the dispensed tube weight for each tube that was dispensed and summing the weights.

^bCompliance was calculated based on number of applications divided by the expected number of applications for each patient multiplied by 100.

^cA patient was considered compliant if they (or their caregiver) applied ≥ 80% of the expected applications during the application period and did not miss > 3 consecutive doses.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Table 16: Summary of Concomitant Medications From the DERMIS-1 and DERMIS-2 Trials (Safety Population)

ACE = angiotensin-converting enzyme; HMG-CoA = 3-hydroxy-3-methylglutaryl coenzyme A.

Note: Summary of concomitant medications was according to anatomic therapeutic chemical class level 4 and preferred term. Concomitant medications are all medications that were continued or started after the first application of the intervention. Patients were counted only once for each class and preferred term.

^aFrequency ≥ 10% of all randomized patients in the study.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials.¹⁰

Efficacy

The summarized end points comprise those included in the sponsor’s summary of clinical evidence as well as any end points identified as important to this review according to stakeholders, for example, the clinical expert, clinician groups, or patient groups.

Investigator’s Global Assessment

The primary end point, IGA success at week 8, was met for the DERMIS-1 and DERMIS-2 trials in the ITT population (Table 17). In the DERMIS-1 trial, 42.4% of patients in the roflumilast arm versus 6.1% of patients in the vehicle arm experienced treatment success based on the IGA at week 8; the ratio of the odds of IGA success with roflumilast relative to the odds of IGA success with the matching vehicle was ||||| |||| || |||| || |||||| | | ||||||| at week 8 from baseline in favour of roflumilast. In the DERMIS-2 trial, 37.5% of patients in the roflumilast arm versus 6.9% of patients in the vehicle arm experienced treatment success based on the IGA at week 8; the ratio of the odds of IGA success with roflumilast relative to the odds of IGA success with the matching vehicle was |||| |||| || |||| || |||||| | | ||||||| at week 8 from baseline, also in favour of roflumilast.

In both studies, the results of the primary analyses were generally consistent with the sensitivity analyses of the primary end point (modified ITT dataset, only observed data in the ITT dataset, and fitted point estimates from a generalized estimating equation) (Table 31 in Appendix 1).

The secondary end point, IGA success at week 4, was tested in a hierarchical manner and adjusted for multiple comparisons (Table 17). In the DERMIS-1 trial, the ratio of the odds of IGA success with roflumilast relative to the odds of IGA success with the matching vehicle was 13.44 (95% CI, 3.72 to 48.58; P < 0.0001) at week 4 from baseline in favour of roflumilast. In the DERMIS-2 trial, the ratio of the odds of IGA success with roflumilast relative to the odds of IGA success with the matching vehicle was 3.91 (95% CI, 1.76 to 8.70; P = 0.0011) at week 4 from baseline, also in favour of roflumilast.

The exploratory end points, IGA score of clear at weeks 4 and 8, were not included in the statistical hierarchy and not adjusted for multiple comparisons (Table 19). In the DERMIS-1 trial, 3.4% of patients in the roflumilast arm and no patients in the vehicle arm had an IGA score of clear at week 4; the results were consistent with the results in the DERMIS-2 trial. In the DERMIS-1 trial, 14.1% of patients in the roflumilast arm and 1.5% of patients in the vehicle arm had an IGA score of clear at week 8. Similarly, in the DERMIS-2 trial, 11.0% of patients in the roflumilast arm and no patients in the vehicle arm had an IGA score of clear at week 8.

Subgroup Analyses of the Primary Efficacy End Point

The subgroup analyses were not included in the statistical hierarchy or adjusted for multiple comparisons (Table 18).

Age Group

A prespecified subgroup analysis of the primary efficacy end point by age group (2 to 11 years, 12 to 17 years, and 18 years or older) was performed. In alignment with the Health Canada indication in patients aged 12 years or older, only the results for the 12 to 17 years and 18 years or older age groups are summarized subsequently.

In the DERMIS-1 trial, of the patients who were aged 12 to 17 years, ||||| in the roflumilast arm and ||||| in the vehicle arm, experienced treatment success based on the IGA at week 8. The results among patients who were aged 18 years or older were consistent with the results in the overall ITT population. In the DERMIS-2 trial, of the patients who were aged 12 to 17 years, no patients in either treatment arm experienced treatment success based on the IGA at week 8. The results among patients aged 18 years and older were consistent with the results for the overall ITT population.

Involvement of Intertriginous Area or Face

A prespecified subgroup analysis of the primary efficacy end point by intertriginous or facial involvement was performed.

In the DERMIS-1 trial, among patients with intertriginous or facial involvement, ||||| in the roflumilast arm and no patients in the vehicle arm experienced treatment success based on the IGA at week 8. In the DERMIS-2 trial, among patients with intertriginous or facial involvement, ||||| in the roflumilast arm and 3.9% in the vehicle arm experienced treatment success based on the IGA at week 8.

A post hoc pooled (DERMIS-1 and DERMIS-2) subgroup analysis of patients with facial and/or intertriginous and/or genital involvement was conducted. A numerically greater proportion of patients with plaque psoriasis involving the face and/or genital and/or intertriginous areas in the roflumilast arm experienced treatment success based on the IGA at week 8 compared with the vehicle arm; this was consistent with the results of the DERMIS-1 and DERMIS-2 studies.⁴⁸

BSA Affected by Psoriasis at Baseline

A prespecified subgroup analysis of the primary efficacy end point by amount of BSA affected by psoriasis at baseline was performed.

In the DERMIS-1 trial, among patients with mild BSA affected (less than 5% of BSA), ||||| in the roflumilast arm and |||| in the vehicle arm experienced treatment success based on the IGA at week 8. In the DERMIS-2 trial, ||||| in the roflumilast arm and |||| in the vehicle arm experienced treatment success based on the IGA at week 8.

In the DERMIS-1 trial, among patients with moderate BSA affected (5% to less than 10% of BSA), ||||| in the roflumilast arm and |||| in the vehicle arm experienced treatment success based on the IGA at week 8. In the DERMIS-2 trial, ||||| in the roflumilast arm and ||||| in the vehicle arm experienced treatment success based on the IGA at week 8.

In the DERMIS-1 trial, among patients with severe BSA (10% and greater BSA), ||||| in the roflumilast arm and no patients in the vehicle arm experienced treatment success based on the IGA at week 8. In the DERMIS-2 trial, ||||| in the roflumilast arm and |||| in the vehicle arm experienced treatment success based on the IGA at week 8.

Intertriginous IGA

Note, the following analyses were based on the prespecified I-IGA ITT population, a subset of patients in the ITT population with intertriginous area involvement and an I-IGA of 2 or more at baseline.

The secondary end point, I-IGA success at week 8, was tested in a hierarchical manner and adjusted for multiple comparisons (Table 17). In the DERMIS-1 trial, the ratio of the odds of I-IGA success with roflumilast relative to the odds of I-IGA success with the matching vehicle was 17.94 (95% CI, 2.33 to 138.20; P < 0.0001) at week 8 from baseline in favour of roflumilast. In the DERMIS-2 trial, the ratio of the odds of I-IGA success with roflumilast relative to the odds of I-IGA success with the matching vehicle was 11.18 (95% CI, 2.33 to 53.68; P = 0.0004) at week 8 from baseline, also in favour of roflumilast.

The exploratory end point, I-IGA success at week 4, was not included in the statistical hierarchy and not adjusted for multiple comparisons (Table 19). In the DERMIS-1 trial, 42.3% of patients in the roflumilast arm and 27.6% of patients in the vehicle arm experienced treatment success based on the I-IGA at week 4. In the DERMIS-2 trial, 52.1% of patients in the roflumilast arm and 17.9% of patients in the vehicle arm experienced treatment success based on the I-IGA at week 4.

Psoriasis Area and Severity Index

The secondary end point, time to PASI 50, was tested in a hierarchical manner and adjusted for multiple comparisons (Table 17). In the DERMIS-1 trial, the median KM estimate of time to PASI 50 was 31.0 days (95% CI, 29.0 to 41.0) in the roflumilast arm versus 104.0 days (95% CI, 85.0 to NE) in the vehicle arm in favour of roflumilast (P < 0.0001). The hazard ratio was 3.867 (95% Wald CI, 2.795 to 5.351). In the DERMIS-2 trial, the median KM estimate of time to PASI 50 was 30.0 days (95% CI, 29.0 to 42.0) in the roflumilast arm and NE (95% CI, 71.0 to NE) in the vehicle arm in favour of roflumilast (P < 0.0001). The hazard ratio was 4.207 (95% Wald CI, 3.029 to 5.844).

The secondary end point, PASI 75 at week 8, was tested in a hierarchical manner and adjusted for multiple comparisons (Table 17). In the DERMIS-1 trial, the ratio of the odds of PASI 75 with roflumilast relative to the odds of PASI 75 with the matching vehicle was 12.00 (95% CI, 5.15 to 27.93; P < 0.0001) at week 8 from baseline in favour of roflumilast. In the DERMIS-2 trial, the ratio of the odds of PASI 75 with roflumilast relative to the odds of PASI 75 with the matching vehicle was 10.42 (95% CI, 4.49 to 24.19; P < 0.0001) at week 8 from baseline, also in favour of roflumilast.

The exploratory end point, PASI 75 at week 4, was not included in the statistical hierarchy and not adjusted for multiple comparisons (Table 19). In the DERMIS-1 trial, 21.8% of patients in the roflumilast arm and 3.0% of patients in the vehicle arm had a PASI 75 at week 4. In the DERMIS-2 trial, 16.2% of patients in the roflumilast arm and 3.6% of patients in the vehicle arm had a PASI 75 at week 4.

The exploratory end point, mPASI 75 at week 8, was not included in the statistical hierarchy and not adjusted for multiple comparisons (Table 32). In the DERMIS-1 trial, 63.1% of patients in the roflumilast arm and 22.0% of patients in the vehicle arm had an mPASI 75 at week 8. In the DERMIS-2 trial, 56.8% of patients in the roflumilast arm and 13.7% of patients in the vehicle arm had an mPASI 75 at week 8.

The exploratory end points, PASI 90 and mPASI 90 at week 8, were not included in the statistical hierarchy and not adjusted for multiple comparisons (Table 19 and Table 32, respectively). In the DERMIS-1 trial, 22.4% of patients in the roflumilast arm and 2.3% of patients in the vehicle arm had a PASI 90 at week 8, and 44.7% of patients in the roflumilast arm and 9.8% of patients in the vehicle arm had an mPASI 90 at week 8. In the DERMIS-2 trial, 17.0% of patients in the roflumilast arm and 2.3% of patients in the vehicle arm had a PASI 90 at week 8, and 35.2% of patients in the roflumilast arm and 8.4% of patients in the vehicle arm had an mPASI 90 at week 8.

The exploratory end points, PASI 100 and mPASI 100 at week 8, were not included in the statistical hierarchy and not adjusted for multiple comparisons (Table 19 and Table 32, respectively). In the DERMIS-1 trial, 13.7% of patients in the roflumilast arm and 1.5% of patients in the vehicle arm had a PASI 100 at week 8; the proportion of patients in each treatment arm who had an mPASI 100 at week 8 was consistent with the results for PASI 100 at week 8. In the DERMIS-2 trial, 11.0% of patients in the roflumilast arm and no patients in the vehicle arm had a PASI 100 at week 8; the proportion of patients in each treatment arm who had an mPASI 100 at week 8 was consistent with the results for PASI 100 at week 8.

DLQI and Children’s DLQI

The exploratory end point, absolute change in DLQI at week 8, was not included in the statistical hierarchy and not adjusted for multiple comparisons (Table 19). In the DERMIS-1 trial, the LS mean change from baseline in DLQI at week 8 was |||| || |||||| in the roflumilast arm and |||| || |||||| in the vehicle arm. In the DERMIS-2 trial, the LS mean change from baseline in DLQI at week 8 was ||| || |||||| in the roflumilast arm and |||| || |||||| in the vehicle arm.

Not enough data were collected in the DERMIS-1 and DERMIS-2 trials to carry out an analysis of covariance for change in CDLQI at week 8 (Table 19).

Worst Itch Numeric Rating Scale

The secondary end points, WI-NRS success at weeks 8, 4, and 2, were tested in a hierarchical manner and adjusted for multiple comparisons (Table 17). Note, the following analyses were based on the prespecified PRU4 ITT population, a subset of patients in the ITT population with a WI-NRS pruritus score of 4 or more at baseline.

In the DERMIS-1 trial, the ratio of the odds of WI-NRS success with roflumilast relative to the odds of WI-NRS success with the matching vehicle was 7.84 (95% CI, 3.85 to 15.94; P < 0.0001) at week 8 from baseline in favour of roflumilast. In the DERMIS-2 trial, the ratio of the odds of WI-NRS success with roflumilast relative to the odds of WI-NRS success with the matching vehicle was 3.59 (95% CI, 2.07 to 6.23; P < 0.0001) at week 8 from baseline, also in favour of roflumilast.

In the DERMIS-1 trial, the ratio of the odds of WI-NRS success with roflumilast relative to the odds of WI-NRS success with the matching vehicle was 4.36 (95% CI, 2.31 to 8.26; P < 0.0001) at week 4 from baseline in favour of roflumilast. In the DERMIS-2 trial, the ratio of the odds of WI-NRS success with roflumilast relative to the odds of WI-NRS success with the matching vehicle was 4.93 (95% CI, 2.65 to 9.18; P < 0.0001) at week 4 from baseline, also in favour of roflumilast.

In the DERMIS-1 trial, the ratio of the odds of WI-NRS success with roflumilast relative to the odds of WI-NRS success with the matching vehicle was 1.76 (95% CI, 0.98 to 3.19; P = 0.1197) at week 2 from baseline (note, this end point failed to reach statistical significance and was potentially underpowered by the planned sample size). In the DERMIS-2 trial, the ratio of the odds of WI-NRS success with roflumilast relative to the odds of WI-NRS success with the matching vehicle was 2.56 (95% CI, 1.43 to 4.58; P = 0.0026) at week 2 from baseline in favour of roflumilast.

Psoriasis Symptom Diary

The secondary end points, change from baseline in PSD total score at weeks 8 and 4, were tested in a hierarchical manner and adjusted for multiple comparisons (Table 17).

Table 17: Summary of Key Efficacy Results From the DERMIS-1 and DERMIS-2 Trials (ITT Population)

ANCOVA = analysis of covariance; CI = confidence interval; HR = hazard ratio; IGA = Investigator’s Global Assessment; I-IGA = intertriginous Investigator’s Global Assessment; I-IGA ITT = a subset of patients in the ITT population with intertriginous area involvement and an I-IGA of 2 or more at baseline; ITT = intention-to-treat; KM = Kaplan-Meier; LS = least squares; NE = not estimable; OR = odds ratio; PASI 50 = 50% reduction in the Psoriasis Area and Severity Index; PASI 75 = 75% reduction in the Psoriasis Area and Severity Index; PRU4 ITT = patients in the ITT population with a WI-NRS pruritus score of ≥ 4 at baseline; PSD = Psoriasis Symptom Diary; SD = standard deviation; SE = standard error; WI-NRS = Worst Itch Numeric Rating Scale.

Note: The P value was adjusted for multiple testing using a prespecified hierarchical testing strategy.

^aIGA success was defined as a score of 0 (clear) or 1 (almost clear) plus a ≥ 2-grade improvement from baseline.

^bOdds ratio, 95% CI, and P value were obtained from the Cochran-Manel-Haenszel test (stratified by study site, baseline IGA, and baseline intertriginous involvement) comparing roflumilast cream 0.3% with vehicle.

^cTime to PASI 50 (50% reduction in PASI from baseline) was calculated as date of PASI 50 minus the day 1 date plus 1. Patients who did not have PASI 50 were censored at day of discontinuation or date lost to follow-up, whichever was earlier. Censored values were not included in the descriptive statistics; only patients who had PASI 50 were included in the descriptive statistics.

^dP value was obtained from the log-rank test where roflumilast cream 0.3% was compared with the matching vehicle.

^eThe HR and 95% Wald CI were obtained from a Cox proportional hazards model comparing roflumilast cream 0.3% with the vehicle with the site, baseline IGA score category, and baseline intertriginous involvement score category as factors in the model.

^fI-IGA success was defined as a score of 0 (clear) or 1 (almost clear) plus a ≥ 2-grade improvement from baseline in patients with intertriginous area involvement and I-IGA ≥ 2 at baseline. This analysis was based on the I-IGA ITT population.

^gThe odds ratio, 95% CI, and P value were obtained from the Cochran-Mantel-Haenszel test (stratified by study site and baseline I-IGA) comparing roflumilast cream 0.3% with vehicle.

^hEstimate for LS means (change from baseline and difference from vehicle), 95% CIs and P values are from an ANCOVA with treatment, site, baseline IGA, baseline intertriginous involvement, and baseline PSD score as independent variables.

ⁱThe P value is to test for a zero difference between groups (roflumilast cream 0.3% minus vehicle) in change from baseline.

^jWI-NRS success was defined as a ≥ 4-point reduction in WI-NRS pruritus score from baseline in patients with a WI-NRS pruritus score of ≥ 4 at baseline. The analysis was based on the PRU4 ITT population.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Table 18: Subgroup Analyses of the Primary Efficacy End Point in the DERMIS-1 and DERMIS-2 Trials (ITT Population)

BSA = body surface area; CI = confidence interval; IGA = Investigator’s Global Assessment; ITT = intention-to-treat; NE = not estimable; OR = odds ratio.

Note: IGA success was defined as a score of 0 (clear) or 1 (almost clear) plus a ≥ 2-grade improvement from baseline.

^aThe odds ratio, 95% CI, and P value were obtained from a Cochran-Mantel-Haenszel test (stratified by site, baseline IGA, and baseline intertriginous involvement) comparing roflumilast cream 0.3% with vehicle.

^bThe P value was not adjusted for multiple testing (i.e., the type I error rate was not controlled).

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials.¹⁰

Table 19: Exploratory Outcomes in the DERMIS-1 and DERMIS-2 Trials (ITT Population)

ANCOVA = analysis of covariance; BSA = body surface area; CDLQI = Children’s Dermatology Life Quality Index; CI = confidence interval; DLQI = Dermatology Life Quality Index; IGA = Investigator’s Global Assessment; I-IGA = intertriginous Investigator’s Global Assessment; I-IGA ITT = a subset of patients in the ITT population with intertriginous area involvement and an I-IGA of 2 or more at baseline; ITT = intention-to-treat; LS = least squares; NE = not estimable; OR = odds ratio; PASI 75 = 75% reduction in the Psoriasis Area and Severity Index score; PASI 90 = 90% reduction in the Psoriasis Area and Severity Index score; PASI 100 = 100% reduction in the Psoriasis Area and Severity Index score; SD = standard deviation; SE = standard error.

^aThe odds ratio, 95% CI, and P value were obtained from a Cochran-Mantel-Haenszel test (stratified by site, baseline IGA, and baseline intertriginous involvement) comparing roflumilast cream 0.3% with vehicle.

^bThe P value was not adjusted for multiple testing (i.e., the type I error rate was not controlled).

^cI-IGA success was defined as a score of 0 (clear) or 1 (almost clear) plus a ≥ 2-grade improvement from baseline in patients with intertriginous area involvement and I-IGA ≥ 2 at baseline. This analysis was based on the I-IGA ITT population.

^dThe odds ratio, 95% CI, and P value were obtained from a Cochran-Mantel-Haenszel test (stratified by study site and baseline I-IGA) comparing roflumilast cream 0.3% with vehicle.

^eEstimates for LS means (change divided by percentage change from baseline and difference from vehicle) and accompanying 95% CIs, and P values are from an analysis of covariance with treatment, site, baseline IGA, baseline intertriginous involvement, and baseline of the variable being analyzed as independent variables.

^fNot enough data were collected to carry out an ANCOVA for this parameter.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials.¹⁰

In the DERMIS-1 trial, the LS mean difference between roflumilast and vehicle in change from baseline in PSD total score at week 8 was −30.9 (standard error [SE] = 3.22; 95% CI, −37.2 to −24.6; P < 0.0001), in favour of roflumilast. In the DERMIS-2 trial, the LS mean difference between roflumilast and vehicle in change from baseline in PSD total score at week 8 was −26.5 (SE = 3.44; 95% CI, −33.2 to −19.7; P < 0.0001), also in favour of roflumilast.

In the DERMIS-1 trial, the LS mean difference between roflumilast and vehicle in change from baseline in PSD total score at week 4 was −25.8 (SE = 3.00; 95% CI, −31.7 to −20.0; P < 0.0001) in favour of roflumilast. In the DERMIS-2 trial, the LS mean difference between roflumilast and vehicle in change from baseline in PSD total score at week 4 was −26.0 (SE = 3.03; 95% CI, −31.9 to −20.0; P < 0.0001), also in favour of roflumilast.

Body Surface Area

The exploratory end point, percent change in BSA affected by psoriasis at week 8, was not included in the statistical hierarchy and not adjusted for multiple comparisons (Table 19). In the DERMIS-1 trial, the LS mean percent change from baseline in BSA affected by psoriasis at week 8 was |||||| || ||||||| in the roflumilast arm ||| |||||| || ||||||| in the vehicle arm. In the DERMIS-2 trial, the LS mean percent change from baseline in BSA affected by psoriasis at week 8 was |||||| || |||||| in the roflumilast arm and |||| || ||||||| in the vehicle arm.

Harms

A summary of harms from the DERMIS-1 and DERMIS-2 trials is presented in Table 20.

Adverse Events

The proportion of patients in the roflumilast arm with any TEAE was 25.2% in the DERMIS-1 trial and 25.9% in the DERMIS-2 trial, while the proportion of patients in the vehicle arm with any TEAE was 23.5% in the DERMIS-1 trial and 18.4% in the DERMIS-2 trial. The most common TEAEs reported in the roflumilast arm (a frequency of 2% or more of patients in either study) were diarrhea (3.5% in the DERMIS-1 trial and 2.8% in the DERMIS-2 trial) and headache (1.0% in the DERMIS-1 trial and 3.8% in the DERMIS-2 trial). All remaining TEAEs were reported in less than 2% of patients in the roflumilast arm in either study.

Serious Adverse Events

The proportion of patients in the roflumilast arm with any SAE was 0.7% in the DERMIS-1 trial and no patients in the DERMIS-2 trial, while the proportion of patients in the vehicle arm with any SAE was 0.7% in both trials. The SAEs reported in the roflumilast arm were concussion (1 patient in the DERMIS-1 trial) and foot fracture, thorax deformity, and pneumothorax (1 patient in the DERMIS-1 trial).

Withdrawals Due to Adverse Events

The proportion of patients in the roflumilast arm with any WDAE was |||| in the DERMIS-1 trial and |||| in the DERMIS-2 trial, while the proportion of patients in the vehicle arm with any WDAE was |||| in the DERMIS-1 trial and |||| in the DERMIS-2 trial. TEAEs leading to discontinuation of treatment and/or study withdrawal were reported in less than || of patients in the roflumilast arm in either study.

Mortality

Based on information provided by the sponsor, no deaths occurred during the DERMIS-1 and DERMIS-2 trials.

Notable Harms

The proportion of patients with application site pain in the roflumilast arm was 0.7% in the DERMIS-1 trial and 1.4% in the DERMIS-2 trial. All remaining TEAEs of special interest (application site pruritus, urticaria, dryness, dermatitis, and irritation) were reported in less than 1% of patients in the roflumilast arm in either study.

Local Tolerability Assessments

In the DERMIS-1 trial, the majority of patients in the roflumilast arm (||||| investigator-rated; ||||| patient-rated) and in the vehicle arm (97.6% investigator-rated; 90.6% patient-rated) did not have tolerability concerns based on local tolerability assessments at week 8.

In the DERMIS-2 trial, the majority of patients in the roflumilast arm (|||| investigator-rated; ||||| patient-rated) and in the vehicle arm (||||| investigator-rated; ||||| patient-rated) did not have tolerability concerns based on local tolerability assessments at week 8.

Table 20: Summary of Harms From the DERMIS-1 and DERMIS-2 Trials (Safety Population)

NA = not applicable; NR = not reported; SAE = serious adverse event; TEAE = treatment-emergent adverse event; WDAE = withdrawal due to adverse event.

^aTEAEs in > 1% of roflumilast-treated patients in the DERMIS-1 or DERMIS-2 trials.

^bPatients were counted once for each system organ class and once for each preferred term.

Sources: Clinical Study Reports for the DERMIS-1⁹ and DERMIS-2 trials;¹⁰ details from the table were taken from the sponsor’s summary of clinical evidence.¹⁶

Critical Appraisal

Internal Validity

The DERMIS-1 and DERMIS-2 studies were randomized, double-blind, and vehicle-controlled. Patients were randomized at baseline according to a computer-generated randomization list and randomization was stratified by study site, baseline IGA score (2 versus ≥ 3), and intertriginous involvement at baseline (I-IGA ≥ 2, yes versus no). Based on input from the clinical expert consulted by CADTH for this review, the extent and severity of disease as measured by baseline BSA and PASI are additional effect modifiers. Note that the median and mean BSA and PASI scores were slightly higher in the vehicle arm compared with the roflumilast arm in both trials. IGA assesses severity of lesions, while BSA and PASI account for extent and severity of disease. Therefore, stratification by IGA alone may not result in an optimal comparability in disease severity between treatment arms, and this may have introduced bias in the efficacy results against roflumilast due to the aforementioned imbalance (note that the magnitude of this potential bias is not known). No other baseline demographic and clinical characteristics were identified that could have had a potential impact on the results in either study.

In both the DERMIS-1 and DERMIS-2 studies, the duration of the drug washout period specific to each class of excluded medication or treatment, if inappropriately selected, was expected to have a possible carryover effect on psoriasis vulgaris. The duration of the washout period was considered reasonable for each treatment in plaque psoriasis by the clinical expert. Therefore, any continued effect from these treatments on plaque psoriasis was not expected after the washout period and not likely to have an impact on the efficacy assessment of roflumilast.

More than 90% of patients in each treatment arm across studies were compliant with the assigned intervention, defined as a patient (or caregiver) who applied 80% or more of the expected applications during the application period and did not miss more than 3 consecutive doses. Topical and systemic treatments as well as phototherapy for plaque psoriasis were prohibited during the treatment period in both studies. Short-term and chronic medications were permitted during the treatment period; similar proportions of patients who reported use of the most common concomitant medications were observed between treatment arms in both studies. The most common major protocol deviation reported in both studies was efficacy assessment (14.4% of all randomized patients in the DERMIS-1 trial and 14.9% of all randomized patients in the DERMIS-2 trial), which was similar between treatment arms across studies. In both studies, the high rate of treatment compliance, concomitant medication use, and deviation of trial performance were considered unlikely to have introduced bias in the efficacy results.

Sensitivity analyses of the primary efficacy end point were performed using the observed data, modified ITT population, and tabulation of fitted point estimates from generalized estimating equations for binary response on observed data; the results of these sensitivity analyses were generally consistent with the results of the primary analysis. Missing data were imputed for the primary, secondary (with the exception of time to PASI 50), and select exploratory outcomes (I-IGA success, PASI 75, PASI 90, and PASI 100) that were included in this report using a regression-based multiple imputation model. According to the investigator, this approach to imputation was considered superior to other strategies, such as the last observation carried forward approach. However, no sensitivity analysis using the last observation carried forward approach was reported; that method is considered a relatively more conservative approach for a chronic disease such as plaque psoriasis, when missingness is likely at random and nondifferential between treatment arms.

The primary efficacy outcome in both studies was IGA success, defined as an IGA score of 0 (clear) or 1 (almost clear), plus an improvement of 2 or more grades from baseline at week 8. Evidence for the validity^34-36 and reliability^36,37 of the 6- and 7-point IGA scales was identified in the literature; however, no evidence for the responsiveness of the IGA was identified. Note that conclusions about the validity and reliability of the 5-point IGA used in the DERMIS-1 and DERMIS-2 studies are limited due to the use of the 6- and 7-point IGA in the psychometric validation studies. However, the clinical expert suggested that this difference in scales was unlikely to have introduced bias in the results. Although an MID has not been estimated, achieving a score of 0 (clear) or 1 (almost clear) on the static IGA has been generally accepted as clinically meaningful (i.e., a responder analysis would consider the proportion of patients with psoriasis who experienced treatment success based on the IGA, defined as a score of 0 or 1 plus at least a 2-grade improvement from baseline).^11,12 Alternatively, or in addition to a score of 0 or 1, the responder analysis may also consider the proportion of patients with at least a 2-grade improvement from baseline on the static IGA.¹¹These were consistent with the definition of IGA success used in both studies.

The secondary efficacy outcomes in both studies included the PASI, I-IGA, PSD, IGA, and WI-NRS. Evidence for the validity^36,38 and reliability³⁹ of the PASI was identified in the literature. The responsiveness of the PASI was found to be low when the affected BSA was less than 10%;^34,40 an mPASI score was calculated for patients with an anatomic area with less than 10% involvement to reflect the status of limited disease in the studies. An MID in the PASI has not been estimated. The measurement properties of the mPASI and I-IGA were not identified in the literature. Evidence for the validity, reliability, and responsiveness of the PSD was identified in the literature.⁴¹ An MID in PSD total score has not been estimated. Evidence for the validity, reliability, and responsiveness of WI-NRS in patients with plaque psoriasis was identified in the literature.¹⁵ The MID in WI-NRS has been estimated to be an improvement of 4 or more points in patients with plaque psoriasis;¹⁵ this was used in the definition of WI-NRS success in both studies.

Of note, IGA and PASI are investigator-reported outcome measures, while the PSD, WI-NRS, DLQI, and CDLQI are patient-reported outcome measures. As such, there was a potential risk of bias, likely in favour of roflumilast, due to subjective reporting of patient-reported outcomes and the potential for the unblinding of treatment assignment as a result of patients observing an improvement in the affected area with roflumilast compared with the matching vehicle, which contained only the excipients of the roflumilast cream. Note that this potential for the unblinding of the treatment assignment may not affect investigator-assessed end points, such as IGA, in the same manner. The clinical expert identified treatment-free interval and duration of response (remission) as additional outcomes important to patients that were not captured in the sponsor’s systematic review protocol.

The primary and secondary outcomes were controlled for multiplicity in both studies using a prespecified hierarchical testing strategy and the Holm procedure to control the familywise type I error. The planned sample size provided sufficient power to analyze up to the first 5 secondary end points; note that the remaining end points demonstrated statistically significant differences between treatment arms that were consistently in favour of roflumilast, with the exception of WI-NRS success at week 2 in DERMIS-1. The relatively small sample size of patients available for subgroup analysis, in particular, the number of patients aged 12 to 17 years, significantly limited the interpretation of findings and the assessment of treatment benefit in patient subgroups. Moreover, as indicated by the clinical expert, the potential heterogeneity of treatment effect by extent and severity of disease as measured by BSA or PASI was not reported. This further compromised the certainty of the evidence on the treatment effect of roflumilast among patients by different extent and severity of disease.

External Validity

Based on input from the clinical expert consulted by CADTH for this review, the inclusion and exclusion criteria were considered narrow. For example, patients with an IGA score of 1 and a PASI score of 1 would be considered candidates for treatment with roflumilast in clinical practice in Canada; however, they were excluded from the trials, as a score of at least 2 was required for enrolment. Additionally, the clinical expert suggested that patients with plaque psoriasis involving less than 2% and more than 20% of BSA, excluding the scalp, palms, and soles, would potentially be treated with roflumilast but were also excluded from the trials. As such, the effect of roflumilast in the broader patient population is unknown; however, based on clinical expert input, it is anticipated that roflumilast will be used in the broader patient population in clinical practice despite the lack of evidence. Note that the Health Canada indication also does not restrict the patient population according to percent BSA involvement.

According to the Canadian Guidelines for the Management of Plaque Psoriasis, the definitions for mild, moderate, and severe plaque psoriasis based on disease severity measures differ, depending on the setting (clinical trials versus clinical practice).⁶ Measures of disease severity in clinical practice include the impact of disease on the patient’s QoL and level of symptomatic control by routine skin care measures and/or topical therapy.⁶ Similarly, the clinical expert indicated that defining the severity of disease in clinical practice is often a gestalt assessment and would require consideration of clearance or near clearance of psoriasis lesions, treatment history, HRQoL, and indication for topical versus systemic therapy. Measures of disease severity (i.e., various cut-offs and scales) in clinical trials are not consistently defined and can differ, depending on the class of drug being evaluated.⁶ Note that the IGA assesses the severity of lesions, while the PASI accounts for extent and severity of disease and, as such, the clinical expert agreed the PASI is more clinically meaningful compared with the IGA. Nonetheless, the PASI and IGA are correlated, as both tools assess the severity of lesions.^35,36 The lower limit of moderate to severe disease may be set at a PASI score of 8 and the lower limit of severe disease may be set at a PASI score of 10⁶; similarly, the clinical expert estimated a cut-off of 10 on the PASI scale to indicate moderate disease. According to these thresholds for moderate to severe disease, the majority of patients in both studies had mild disease based on the mean baseline PASI score, but based on the baseline PASI range, both studies also included patients with moderate to severe disease (i.e., candidates for systemic therapy).

The intervention used in both studies, roflumilast cream 0.3% administered topically to lesions of plaque psoriasis once daily, was consistent with the dosage recommended in the product monograph.²² The clinical expert indicated that in clinical practice, roflumilast would probably be stopped by most patients when lesions are clear; however, the treatment of affected areas with roflumilast or the vehicle was maintained for the duration of the studies, regardless of whether the psoriasis cleared.

In both studies, roflumilast was compared with a matching vehicle cream that contained only the excipients of the roflumilast cream. However, given the wide range of topical treatment options currently available in clinical practice for plaque psoriasis, the clinical expert agreed that an active comparator would have been more appropriate; in particular, topical steroids and vitamin D analogues would have been appropriate comparators. However, the vehicle cream may be considered an appropriate comparator, as there are limited options for intertriginous areas, which have been identified as an area of unmet need by the clinical expert and clinician groups.

The clinician groups and the clinical expert consulted by CADTH for this review agreed the IGA (otherwise referred to as the PGA in practice), PASI, DLQI or CDLQI, and percent of BSA involvement were clinically relevant and meaningful (i.e., capture the extent and severity of disease and determine treatment response in clinical practice). The clinical expert also indicated that the I-IGA, PSD, and WI-NRS were clinically relevant. Moreover, the clinical expert indicated that clearance or near clearance of psoriasis lesions is commonly used in clinical practice to assess psoriasis severity, which is analogous to the use of the IGA in clinical trials. Both the clinician groups and the clinical expert indicated that tools such as the PASI and DLQI are not commonly used in clinical practice unless mandated by the payers for reimbursement. Additionally, the clinician groups and the clinical expert indicated that patient satisfaction and/or feedback is an important outcome used to determine treatment response in clinical practice. The clinical expert indicated that components of patient satisfaction can include improvement in symptoms (itch), appearance of lesions, QoL and sexual function (relevant in intertriginous psoriasis), and convenience of treatment and pleasantness of treatment formulation from the patient’s perspective. These components are generally consistent with the improved outcomes identified in the input submitted by the patient groups for this review.

The treatment period in each study was 8 weeks, which was considered by the clinical expert to be an appropriate amount of time to assess the efficacy and safety of roflumilast. According to the clinical expert, in the context of topical therapy for plaque psoriasis, 4 weeks is often sufficient to determine whether an adequate response will be achieved; although extensive psoriasis may require more time to achieve an adequate response, it should be well on its way after 4 weeks. Moreover, the Canadian Dermatology Association indicated that treatment goals are assessed after 8 weeks of treatment. The clinical expert also suggested that roflumilast may be preferred for long-term use, as it is a nonsteroidal drug; however, the collection of evidence on the durability of effect and long-lasting remission (based on a longer follow-up period) was limited to the DERMIS-OLE study.

Long-Term Extension Study

The contents within this section have been informed by materials submitted by the sponsor. The following have been summarized and validated by the CADTH review team.

Description of Study

This section includes a summary of 1 OLE study, the DERMIS-OLE (ARQ-151-306) study,¹⁷ which was included in the sponsor’s submission to CADTH and was considered to provide further information on the long-term safety and efficacy of roflumilast cream 0.3% in patients with chronic plaque psoriasis.

This is a multicentre, open-label, phase III, long-term safety and efficacy study in which patients with chronic plaque psoriasis involving up to 25% of BSA received roflumilast cream 0.3% applied once a day for up to 24 weeks. Patients included those who successfully completed a prior roflumilast cream study in psoriasis (cohort 1) or were treatment-naive with roflumilast cream (cohort 2). A total of 267 patients were enrolled (n = 264 in cohort 1 and n = 3 in cohort 2), of which 266 patients received treatment. The study period was between February 12, 2020, and December 14, 2020.

Of note, 2 out of 3 patients in cohort 2 received treatment but they had not reached study week 4 at the time of the data cut-off. Therefore, efficacy data are not available for those patients. No TEAEs occurred in the 2 treatment-naive patients in cohort 2 as of the data cut-off. As a result, the presentation of efficacy and harms results in this interim report focused on the 264 patients in cohort 1.¹⁷

Populations

The eligibility criteria at screening are briefly summarized subsequently.

The DERMIS-OLE study included patients aged 2 years and older with a diagnosis of chronic plaque psoriasis. Patients were enrolled into 1 of 2 cohorts. To be eligible for cohort 1, patients were required to have successfully completed 1 of the following roflumilast cream clinical trials: the phase I trial (ARQ-151-107), a phase II trial (ARQ-151-215 or ARQ-151-216), the DERMIS-1 (ARQ-151-301) trial, or the DERMIS-2 (ARQ-151-302) trial. Patients eligible for cohort 2 were naive to treatment with roflumilast, with a clinical diagnosis of psoriasis vulgaris of at least 3 months duration on the face, extremities, trunk, and/or intertriginous areas involving 2% to 25% of BSA (excluding the scalp, palms, and soles), and with an IGA score of at least mild (score of 2) at day 1. Of note, at the time of the data cut-off, no patient had rolled over from studies ARQ-151-107 (which is an open-label, phase I study in adolescents and adults with plaque psoriasis) or ARQ-151-215 or ARQ-151-216, which are open-label, phase II, 4-week trials in children with plaque psoriasis aged between 2 and 5 years; thus, the cohort 1 data reflect only patients enrolled from the ARQ-151-301 and ARQ-151-302 studies.

Patients were excluded from the study if they experienced a roflumilast treatment–related AE or SAE or were currently taking lithium or an antimalaria drug or medication that can affect psoriasis vulgaris (e.g., beta-blockers, ACE inhibitors). Patients suspected of having severe renal insufficiency or hepatic disorders, HIV infection, or hypersensitivity to the investigational product, were not included in the study. Additionally, patients with a history of severe depression and suicidal ideation or a history of chronic alcohol or drug abuse within 6 months of the initiation of the study medication were excluded.

In cohort 1 (N = 264), the mean age of patients was 47.9 years (SD = 15.79), with ||||| adult, |||| adolescent, and |||| pediatric patients. The proportion of patients who were male was larger than the proportion who were female (||||| versus |||||, respectively). Most patients were white (84.1%), and the mean weight was 92.88 kg (SD = 24.720). In cohort 1, most patients (66.3%) had a moderate IGA score at baseline, the mean baseline BSA affected by psoriasis was 6.49% (SD = 4.48), and the mean baseline PASI score was 6.1 (SD = 3.19).

Interventions

Roflumilast cream 0.3% was applied topically by patients or caregivers once daily to active psoriatic plaques (up to a maximum application area of 25% of BSA) for up to 24 weeks at home.¹⁷

Treatment application with roflumilast was to occur in the evening, at least 15 minutes after the patient had bathed and at least 20 minutes before the patient went to bed, and the area was not to be washed until the following morning. Patients applied the roflumilast to all active psoriasis lesions, including any new plaques that developed during the study unless otherwise instructed by the investigator. Application was to all areas affected, including intertriginous and/or genital regions and the face, but excluding the scalp. Patients were not required to continue treatment if the psoriasis plaques completely resolved.

Concomitant Medications

The use of concomitant medication was prohibited for biological therapies, oral corticosteroids, retinoids, apremilast, methotrexate, cyclosporine, and other systemic immunosuppressants. Topical antipsoriasis medication (e.g., corticosteroids), vitamin D analogues, prescription shampoos, psoralen and UV A phototherapy, UV B, and antihistamines were also not allowed during treatment with roflumilast 0.3%.¹⁷

Outcomes

The primary outcomes of the DERMIS-OLE study were safety, including TEAEs and SAEs. Secondary efficacy outcomes included the following:

• an IGA score of clear or almost clear

• IGA success (defined as an IGA score of clear or almost clear plus a 2-grade improvement from baseline)

• duration of response, including duration of clear or almost clear status and duration of IGA success (defined as the period from the date of the first observation of an IGA score of clear or almost clear or IGA success until the date of restarting roflumilast)

• a 50%, 75%, 90%, or 100% reduction in PASI over time

• an I-IGA score of clear or almost clear over time

• I-IGA success over time (defined as an I-IGA score of clear or almost clear plus a 2-grade improvement from baseline)

• change in WI-NRS score over time.¹⁷

Other relevant outcomes assessed included change in BSA score over time, the proportion of patients who met the criteria for disease clearance (i.e., having clear scores on the IGA, I-IGA, PASI, and mPASI and stopping treatment at some time during the study after primary baseline), and the time to restarting study drug (i.e., treatment-free interval), defined as the time when patients experienced disease clearance and stopped treatment to all lesions.

Statistical Analysis

No formal statistical test was performed. Descriptive statistics were used to give a summary of the efficacy and safety results. The analysis was based on the number of patients with measurements available and no imputation was performed. A December 14, 2020, data cut-off date was set for interim reporting because this investigation was ongoing. Of note, the efficacy data for 2 out of the 3 treatment-naive patients (cohort 2) were not available because they had not yet reached trial week 4 at the time of the data cut-off.¹⁷

Results

Patient Disposition

At the time of the data cut-off, 264 patients (98.9%) who had participated in parent studies and 3 treatment-naive patients (1.1%), 2 of whom had not reached the week 4 visit, were enrolled into this OLE study. Therefore, 267 patients started the OLE study and 266 (99.5%) of them had received roflumilast. Of note, the patients in the OLE study who had received roflumilast 0.3% in the parent studies received treatment for a total of 32 weeks (8 weeks of treatment in the parent studies plus an additional 24 weeks), whereas patients who received vehicle cream received up to 24 weeks of treatment with roflumilast in the OLE study. At the time of the data cut-off, 222 patients (83.1%) had completed the study, 32 (12%) had discontinued the study, and 12 (4.5%) were ongoing in the study. The most common reasons for discontinuation were lost to follow-up (5.7%) and withdrawal (4.5%). Two patients discontinued due to noncompliance and 1 patient discontinued because of an AE.¹⁷ A detailed summary of the patient disposition in the DERMIS-OLE study is available in Table 21.

Table 21: Summary of Patient Disposition From the DERMIS-OLE Study (All Enrolled Patients)