This post hoc, pooled analysis was designed to identify patient
baseline characteristics that could predict early response to
crisaborole therapy, as defined by ISGA success at day 8 (after
1 week of therapy), ISGA of clear or almost clear at day 8, and
≥1-point improvement in Severity of Pruritus Scale (SPS) score
based on the weekly average SPS score at day 8, using data
from 2 phase 3 studies of patients aged ≥2 years with mild-tomoderate
AD (AD-301 and AD-302).
METHODS
Study Design
Complete study design details and inclusion/exclusion criteria for the 2 phase 3 studies have been previously published.4 Briefly, eligible patients were aged ≥2 years, with a clinical diagnosis of AD, baseline ISGA indicating mild (2) or moderate (3) disease, and percentage of treatable body surface area (%BSA) ≥5. Patients were randomly assigned 2:1 to receive treatment with crisaborole or vehicle applied twice daily to all treatable areas of the body, except the scalp, for 28 days. Patients were ineligible if they had previously used biologic therapy, systemic corticosteroids within 28 days of treatment initiation, or topical calcineurin inhibitors or topical corticosteroids within 14 days of treatment initiation. Patients on stable regimens of inhaled corticosteroids, antihistamines, or topical retinoids for non-AD treatment could continue these medications. Patients could also use acceptable bland emollients throughout the study to manage dry skin areas around, but not overlapping, the treatable AD-involved areas. Use of rescue medication was not permitted.
ISGA, a 5-point rating scale used to measure overall disease severity from clear (0) to severe (4),5 was assessed at baseline and every 7 days (day 8 was the first postbaseline assessment) up to day 29. SPS, a 4-point rating scale of the severity of pruritus from none (0) to severe (3),9 was recorded twice daily up to day 29 via electronic diary by patients or caregivers (for patients aged 2-7 years) before study drug was applied.
Study Outcomes
Early response outcomes
Early response at day 8 was defined using 3 different definitions: ISGA success (defined as ISGA of clear [0] or almost clear [1] with ≥2-grade improvement from baseline), ISGA of clear or almost clear, and SPS response (defined as ≥1-point improvement in SPS based on weekly average scores). The analyses were performed for crisaborole-treated patients, and comparisons were made between subgroups based on baseline characteristics. The subgroups included in the analyses were age (<12 years vs ≥12 years), sex (male vs female), race (white vs nonwhite), ethnicity (Hispanic vs non-Hispanic), baseline ISGA (mild [2] vs moderate [3]), %BSA involvement (mild [5 to <16] vs moderate/severe [≥16]),10 prior AD treatment (prior vs no prior, defined as any prior use of systemic or topical corticosteroids or topical calcineurin inhibitors for the treatment of AD within 90 days before the first dose of study medication), median disease duration (≤6.45 years vs >6.45 years), history of asthma/allergic rhinitis (history vs no history), and concurrent antihistamine use (use vs no use).
Early response as predictor of day 29 response
Correlations between early response at day 8 and response at day 29 were calculated to investigate whether early response was a predictor of response at day 29. The analyses were performed for crisaborole-treated patients using ISGA success, ISGA of clear or almost clear, and SPS responses; comparisons were made between responders and nonresponders at day 8 and day 29.
Statistical Analysis
In the early response outcomes analysis, odds ratios (ORs), 95% CIs, and P values were calculated using a logistic model with response as the dependent variable and subgroup as the independent variable. P values were not adjusted for multiplicity. In the correlation of early response with day-29 response analysis, ORs, CIs, and P values in early response outcomes were calculated using a logistic model with response at day 29 (ISGA success, ISGA clear or almost clear, or SPS response) as the dependent variable and response at day 8 (ISGA success, ISGA clear or almost clear, or SPS response) as the independent variable. P values were not adjusted for multiplicity, with no imputation of missing data.
Complete study design details and inclusion/exclusion criteria for the 2 phase 3 studies have been previously published.4 Briefly, eligible patients were aged ≥2 years, with a clinical diagnosis of AD, baseline ISGA indicating mild (2) or moderate (3) disease, and percentage of treatable body surface area (%BSA) ≥5. Patients were randomly assigned 2:1 to receive treatment with crisaborole or vehicle applied twice daily to all treatable areas of the body, except the scalp, for 28 days. Patients were ineligible if they had previously used biologic therapy, systemic corticosteroids within 28 days of treatment initiation, or topical calcineurin inhibitors or topical corticosteroids within 14 days of treatment initiation. Patients on stable regimens of inhaled corticosteroids, antihistamines, or topical retinoids for non-AD treatment could continue these medications. Patients could also use acceptable bland emollients throughout the study to manage dry skin areas around, but not overlapping, the treatable AD-involved areas. Use of rescue medication was not permitted.
ISGA, a 5-point rating scale used to measure overall disease severity from clear (0) to severe (4),5 was assessed at baseline and every 7 days (day 8 was the first postbaseline assessment) up to day 29. SPS, a 4-point rating scale of the severity of pruritus from none (0) to severe (3),9 was recorded twice daily up to day 29 via electronic diary by patients or caregivers (for patients aged 2-7 years) before study drug was applied.
Study Outcomes
Early response outcomes
Early response at day 8 was defined using 3 different definitions: ISGA success (defined as ISGA of clear [0] or almost clear [1] with ≥2-grade improvement from baseline), ISGA of clear or almost clear, and SPS response (defined as ≥1-point improvement in SPS based on weekly average scores). The analyses were performed for crisaborole-treated patients, and comparisons were made between subgroups based on baseline characteristics. The subgroups included in the analyses were age (<12 years vs ≥12 years), sex (male vs female), race (white vs nonwhite), ethnicity (Hispanic vs non-Hispanic), baseline ISGA (mild [2] vs moderate [3]), %BSA involvement (mild [5 to <16] vs moderate/severe [≥16]),10 prior AD treatment (prior vs no prior, defined as any prior use of systemic or topical corticosteroids or topical calcineurin inhibitors for the treatment of AD within 90 days before the first dose of study medication), median disease duration (≤6.45 years vs >6.45 years), history of asthma/allergic rhinitis (history vs no history), and concurrent antihistamine use (use vs no use).
Early response as predictor of day 29 response
Correlations between early response at day 8 and response at day 29 were calculated to investigate whether early response was a predictor of response at day 29. The analyses were performed for crisaborole-treated patients using ISGA success, ISGA of clear or almost clear, and SPS responses; comparisons were made between responders and nonresponders at day 8 and day 29.
Statistical Analysis
In the early response outcomes analysis, odds ratios (ORs), 95% CIs, and P values were calculated using a logistic model with response as the dependent variable and subgroup as the independent variable. P values were not adjusted for multiplicity. In the correlation of early response with day-29 response analysis, ORs, CIs, and P values in early response outcomes were calculated using a logistic model with response at day 29 (ISGA success, ISGA clear or almost clear, or SPS response) as the dependent variable and response at day 8 (ISGA success, ISGA clear or almost clear, or SPS response) as the independent variable. P values were not adjusted for multiplicity, with no imputation of missing data.
RESULTS
Patient Disposition
A total of 1522 patients were included in the combined phase 3 studies, and 1016 and 506 patients received crisaborole and vehicle, respectively (Table 1). The mean age in the crisaborole group was 12.3 years; the mean %BSA was 18.3. Greater proportions of patients were aged <12 years (61.7%), female (55.7%), white (60.7%), and non-Hispanic (80.3%); had moderate disease per ISGA (61.3%); had mild %BSA involvement (64.8%); and had not received prior AD treatment within 90 days before starting the study (59.0%). Most patients did not have a history of asthma/allergic rhinitis (65.7%) and/or did not concurrently use antihistamines (75.9%). The median disease duration was 6.45 years.
Early Responder Outcomes
Patients were more likely to be early (day 8) ISGA success responders if they were aged <12 years (<12 years vs ≥12 years, OR [95% CI], 1.83 [1.24-2.69]; P=0.0023), were white (white vs non-white, 1.51 [1.04-2.19]; P=0.0316), had moderate baseline disease by ISGA (mild vs moderate, 0.48 [0.32-0.72]; P=0.0003), had not received prior AD treatment (prior vs no prior, 0.65 [0.45-0.94]; P=0.0213), had disease duration ≤6.45 years (≤6.45 years vs >6.45 years, 1.46 [1.03-2.08]; P=0.0349), or did not concurrently use antihistamines (use vs no use, 0.56 [0.35-0.89]; P=0.0148; Figure 1).
A total of 1522 patients were included in the combined phase 3 studies, and 1016 and 506 patients received crisaborole and vehicle, respectively (Table 1). The mean age in the crisaborole group was 12.3 years; the mean %BSA was 18.3. Greater proportions of patients were aged <12 years (61.7%), female (55.7%), white (60.7%), and non-Hispanic (80.3%); had moderate disease per ISGA (61.3%); had mild %BSA involvement (64.8%); and had not received prior AD treatment within 90 days before starting the study (59.0%). Most patients did not have a history of asthma/allergic rhinitis (65.7%) and/or did not concurrently use antihistamines (75.9%). The median disease duration was 6.45 years.
Early Responder Outcomes
Patients were more likely to be early (day 8) ISGA success responders if they were aged <12 years (<12 years vs ≥12 years, OR [95% CI], 1.83 [1.24-2.69]; P=0.0023), were white (white vs non-white, 1.51 [1.04-2.19]; P=0.0316), had moderate baseline disease by ISGA (mild vs moderate, 0.48 [0.32-0.72]; P=0.0003), had not received prior AD treatment (prior vs no prior, 0.65 [0.45-0.94]; P=0.0213), had disease duration ≤6.45 years (≤6.45 years vs >6.45 years, 1.46 [1.03-2.08]; P=0.0349), or did not concurrently use antihistamines (use vs no use, 0.56 [0.35-0.89]; P=0.0148; Figure 1).
