The secondary endpoint "Change from baseline to Week 16 in HECSI score" was evaluated by determining if there was a dose-response relationship between the change from baseline in HECSI score at Week 16 and the dose administered, using the Multiple Comparison Procedure - Modelling (MCP-Mod) methodology. Several candidate parametric models were assumed and multiple comparison techniques were used to choose the model(s) most likely to represent the true underlying dose-response curve.

P-values were adjusted for multiple comparisons. Models with an adj. p-value <0.025 were statistically sign. different from a flat dose-response model. Model: Change = Treatment + Baseline + Region + Baseline IGA-CHE.

Model selected: Emax model

Least Square (LS) Means were calculated using a mixed model for repeated measurements on the post-baseline responses up to Week 16 with an unstructured covariance matrix, Kenward-Roger approximation to estimate denominator degrees of freedom, and the mean modelled as follows:

Change from baseline in HECSI

= treatment × visit + baseline HECSI × visit + region + baseline IGA-CHE.

The primary comparison between each active delgocitinib dose and vehicle was at Week 16.

The statistical test was not controlled for multiplicity. Data at visits following premature discont. of IMP or initiation of rescue med. before Week 16 were considered missing and imputed with MMRM predicted values (as were any other missing data).

Least Square (LS) Means were calculated using a mixed model for repeated measurements on the post-baseline responses up to Week 16 with an unstructured covariance matrix, Kenward-Roger approximation to estimate denominator degrees of freedom, and the mean modelled as follows:

Change from baseline in HECSI

= treatment × visit + baseline HECSI × visit + region + baseline IGA-CHE.

The primary comparison between each active delgocitinib dose and vehicle was at Week 16.

The statistical test was not controlled for multiplicity. Data at visits following premature discont. of IMP or initiation of rescue med. before Week 16 were considered missing and imputed with MMRM predicted values (as were any other missing data).

Least Square (LS) Means were calculated using a mixed model for repeated measurements on the post-baseline responses up to Week 16 with an unstructured covariance matrix, Kenward-Roger approximation to estimate denominator degrees of freedom, and the mean modelled as follows:

Change from baseline in HECSI

= treatment × visit + baseline HECSI × visit + region + baseline IGA-CHE.

The primary comparison between each active delgocitinib dose and vehicle was at Week 16.

The statistical test was not controlled for multiplicity. Data at visits following premature discont. of IMP or initiation of rescue med. before Week 16 were considered missing and imputed with MMRM predicted values (as were any other missing data).

Least Square (LS) Means were calculated using a mixed model for repeated measurements on the post-baseline responses up to Week 16 with an unstructured covariance matrix, Kenward-Roger approximation to estimate denominator degrees of freedom, and the mean modelled as follows:

Change from baseline in HECSI

= treatment × visit + baseline HECSI × visit + region + baseline IGA-CHE.

The primary comparison between each active delgocitinib dose and vehicle was at Week 16.

The statistical test was not controlled for multiplicity. Data at visits following premature discont. of IMP or initiation of rescue med. before Week 16 were considered missing and imputed with MMRM predicted values (as were any other missing data).

Time to IGA-CHE TS was defined as the time from the date of the first IMP application to first assessment of IGA-CHE TS. Subjects without baseline observation were censored at the date of the first IMP application. Subjects with baseline observation not achieving IGA-CHE TS during the treatment period were censored at the date of the last visit with a valid post-baseline assessment on or prior to the date of discontinuation of IMP or initiation of rescue medication, whichever occurred first.

IGA-CHE TS was reached in less than 50% of participants, therefore the median time was not estimable.

The statistical test was not controlled for multiplicity.

Treatment groups were compared using a 2-sided log-rank test stratified by region and baseline IGA-CHE score. An event was defined as the first time achieving IGA-CHE TS.

Time to IGA-CHE TS was defined as the time from the date of the first IMP application to first assessment of IGA-CHE TS. Subjects without baseline observation were censored at the date of the first IMP application. Subjects with baseline observation not achieving IGA-CHE TS during the treatment period were censored at the date of the last visit with a valid post-baseline assessment on or prior to the date of discontinuation of IMP or initiation of rescue medication, whichever occurred first.

IGA-CHE TS was reached in less than 50% of participants, therefore the median time was not estimable.

The statistical test was not controlled for multiplicity.

Treatment groups were compared using a 2-sided log-rank test stratified by region and baseline IGA-CHE score. An event was defined as the first time achieving IGA-CHE TS.

Time to IGA-CHE TS was defined as the time from the date of the first IMP application to first assessment of IGA-CHE TS. Subjects without baseline observation were censored at the date of the first IMP application. Subjects with baseline observation not achieving IGA-CHE TS during the treatment period were censored at the date of the last visit with a valid post-baseline assessment on or prior to the date of discontinuation of IMP or initiation of rescue medication, whichever occurred first.

The statistical test was not controlled for multiplicity.

Treatment groups were compared using a 2-sided log-rank test stratified by region and baseline IGA-CHE score. An event was defined as the first time achieving IGA-CHE TS.

Time to IGA-CHE TS was defined as the time from the date of the first IMP application to first assessment of IGA-CHE TS. Subjects without baseline observation were censored at the date of the first IMP application. Subjects with baseline observation not achieving IGA-CHE TS during the treatment period were censored at the date of the last visit with a valid post-baseline assessment on or prior to the date of discontinuation of IMP or initiation of rescue medication, whichever occurred first.

The statistical test was not controlled for multiplicity.

Treatment groups were compared using a 2-sided log-rank test stratified by region and baseline IGA-CHE score. An event was defined as the first time achieving IGA-CHE TS.