Adapted ACR30 response was defined as a ≥30% improvement from baseline in at least 3 of the following 5 core response variables, with no more than one of these variables worsening by >30%:

1. Physicians global assessment of disease activity (PhGA) on a 0-100 mm VAS (0=very good and 100=very poor).
2. Participant's assessment of disease activity (PtGA) on a 0-100 mm VAS (0=very good and 100=very poor).
3. Health Assessment Questionnaire- disability index (HAQ-DI): 20 questions across 8 domains assessing the functional abilities. The total score was calculated as the average of the scores for each domain, ranging from 0 (no disability) to 3 (very severe disability).
4. Number of active joints (68 joints evaluated for pain/tenderness and 66 for swelling)
5. Index of inflammation: C-reactive Protein (CRP) levels Additionally, participants were required to have no intermittent fever during the preceding week.

Adapted ACR50 response was defined as 50% reduction between baseline in at least 3 of the 5 response variables, with no more than one of these variables worsening by more than 30%:

1. Physicians global assessment of disease activity on a 0-100 mm VAS (0=very good and 100=very poor).
2. Participant's assessment of disease activity (PtGA) on a 0-100 mm VAS (0=very good and 100=very poor).
3. Health Assessment Questionnaire- disability index (HAQ-DI): 20 questions across 8 domains assessing the functional abilities. The total score was calculated as the average of the scores for each domain, ranging from 0 (no disability) to 3 (very severe disability).
4. Number of active joints (68 joints evaluated for pain/tenderness and 66 for swelling)
5. Index of inflammation: C-reactive Protein (CRP) levels Additionally, participants were required to have no intermittent fever during the preceding week.

Adapted ACR70 response was defined as 70% reduction between baseline in at least 3 of the 5 response variables, with no more than one of these variables worsening by more than 30%:

1. Physicians global assessment of disease activity (PhGA) on a 0-100 mm VAS (0=very good and 100=very poor).
2. Participant's assessment of disease activity (PtGA) on a 0-100 mm VAS (0=very good and 100=very poor).
3. Health Assessment Questionnaire- disability index (HAQ-DI): 20 questions across 8 domains assessing the functional abilities. The total score was calculated as the average of the scores for each domain, ranging from 0 (no disability) to 3 (very severe disability).
4. Number of active joints (68 joints evaluated for pain/tenderness and 66 for swelling)
5. Index of inflammation: C-reactive Protein (CRP) levels Additionally, participants were required to have no intermittent fever during the preceding week.

Adapted ACR90 response was defined as 90% reduction between baseline in at least 3 of the 5 response variables, with no more than one of these variables worsening by more than 30%:

1. Physicians global assessment of disease activity (PhGA) on a 0-100 mm VAS (0=very good and 100=very poor).
2. Participant's assessment of disease activity (PtGA) on a 0-100 mm VAS (0=very good and 100=very poor).
3. Health Assessment Questionnaire- disability index (HAQ-DI): 20 questions across 8 domains assessing the functional abilities. The total score was calculated as the average of the scores for each domain, ranging from 0 (no disability) to 3 (very severe disability).
4. Number of active joints (68 joints evaluated for pain/tenderness and 66 for swelling)
5. Index of inflammation: C-reactive Protein (CRP) levels Additionally, participants were required to have no intermittent fever during the preceding week.

Adapted ACR100 response was defined as 100% reduction between baseline in at least 3 of the 5 response variables, with no more than one of these variables worsening by more than 30%:

1. Physicians global assessment of disease activity (PhGA) on a 0-100 mm VAS (0=very good and 100=very poor).
2. Participant's assessment of disease activity (PtGA) on a 0-100 mm VAS (0=very good and 100=very poor).
3. Health Assessment Questionnaire- disability index (HAQ-DI): 20 questions across 8 domains assessing the functional abilities. The total score was calculated as the average of the scores for each domain, ranging from 0 (no disability) to 3 (very severe disability).
4. Number of active joints (68 joints evaluated for pain/tenderness and 66 for swelling)
5. Index of inflammation: C-reactive Protein (CRP) levels Additionally, participants were required to have no intermittent fever during the preceding week.

The SFS evaluated 10 features, each scored as 1 (present) or 0 (absent). The total score ranged from 0 (none present) to 10 (all present). A negative change from baseline indicated clinical improvement.

Clinical features:

• Fever (>37.5°C for ≥5 consecutive days)
• Rash (salmon-pink during febrile episodes)
• Serositis
• Lymphadenopathy (>1.5 cm)
• Hepatomegaly/splenomegaly (confirmed by ultrasound or palpation)

Laboratory features:

At baseline, considered present if:

• ESR ≥20 mm/hour
• CRP ≥10 mg/L
• WBC count ≥12×10⁹/L
• Hemoglobin ≤11 g/dL
• Platelets ≥400×10⁹/L

During treatment and follow-up:

• ESR: 0 if <20 mm/hour or decreased ≥30%; 1 if increased or decreased <30%
• CRP: 0 if <10 mg/L or decreased ≥30%; 1 if increased or decreased <30%
• WBC count: 0 if <12×10⁹/L or decreased ≥20%; 1 if increased or decreased <20%
• Hemoglobin: 0 if >11 g/dL or increased ≥20%; 1 if decreased or increased <20%
• Platelets: 0 if <400×10⁹/L or decreased ≥20%; 1 if increased or decreased <20%

The physician rated the participant's disease activity on a 0-100 mm VAS, ranging from no disease activity (0 mm) to very severe disease activity (100 mm). A negative change from baseline indicated improvement

The participant's assessment of disease activity was assessed on the VAS. The VAS scale ranges from 0-100 mm, from no pain/very well (0 mm) to very severe pain/very poor (100 mm). A negative change from baseline indicated improvement

The HAQ was used to assess physical ability and functional status of participants as well as quality of life. The disability dimension consisted of 20 multiple choice items concerning difficulty in performing eight common activities of daily living; dressing, arising, eating, walking, hygiene, reach, grip and usual activities. Each item is scored from 0 (no difficulty) to 3 (unable to do). The score for each domain is the highest item score, and the total score is the mean of the domain scores., ranging from 0 (no disability) to 3 (severe disability). A negative change from baseline indicated improvement.

Active joints were defined as joints with swelling or pain/tenderness. 68 joints were assessed for pain/tenderness and 66 joints for swelling.

The number of active joints ranged from 0 (no active joints) to 68 (all joints tender and /or swollen). A negative change from baseline indicated improvement.

CRP levels in the blood were determined A negative change from baseline indicated improvement in systemic inflammation

The number of participants without intermittent fever associated to AOSD (defined as oral, rectal, or axillary body temperature > 38°C only for several hours during the day) was assessed.

The oral corticosteroid dose was derived from the prednisone equivalent dose per day. The change from baseline was assessed. A negative change from baseline indicated reduced corticosteroid use.

The absence or presence of skin rash was assessed based on physical exam findings including whether it is typical or/and atypical.

Participants with a rash showing both typical and atypical features were counted in both sub-categories.

In the summary table, for each visit participants are classified as: rash present (typical and/or atypical), rash present (typical), rash present (atypical), or rash absent.

For every visit, the denominator used to calculate percentages corresponds to the number of participants with available data at that visit.

DAS28-CRP was a composite index validated for patients with rheumatoid arthritis (RA). It took into account the following components: tender joint count (0-28), swollen joint count (0-28), C-reactive protein (CRP, mg/L), and the Patient's Global Assessment of Disease Activity (Global Health), rated from 0 (best) to 100 (worst). These results were combined to produce the DAS28-CRP score, which ranged from 1.0 (disease remission) to 9.4 (high disease activity). A negative change from baseline in DAS28-CRP indicated an improvement

Serum canakinumab concentrations by visit

Total IL-1β (sum of IL-1β free and bound to canakinumab) in serum was assessed

The ADAs against canakinumab were assessed in serum using a validated immunoassay assay. The number of participants who had a canakinumab ADA positive result was assessed

Number of participants achieving successful corticosteroid tapering at Week 28 was assessed. Successful tapering was defined as meeting any one of the following criteria and maintaining a minimum Adapted ACR30 response:

• Participants with a baseline prednisone equivalent dose > 0.8 mg/kg/day must reduce to ≤ 0.5 mg/kg/day
• Participants with a baseline dose between ≥ 0.5 and ≤ 0.8 mg/kg/day must reduce by ≥ 0.3 mg/kg/day
• Participants with any baseline dose must reduce to ≤ 0.2 mg/kg/day
• Participants with a baseline dose ≤ 0.2 mg/kg/day must achieve any reduction

The Adapted ACR30 response requires ≥30% improvement in at least 3 of the following 5 variables, with no more than one worsening by >30%, and no intermittent fever in the preceding week:

1. PhGA [0-100 VAS]
2. PtGA [0-100 VAS]
3. HAQ-DI [0-3 scale]
4. Number of active joints (pain/tenderness: 0-68; swelling: 0-66)
5. CRP levels

FAS: all participants who received at least one dose of study treatment

Adapted ACR30 response was defined as a ≥30% improvement from baseline in at least 3 of the following 5 core response variables, with no more than one of these variables worsening by >30%:

1. Physicians global assessment of disease activity (PhGA) on a 0-100 mm VAS (0=very good and 100=very poor).
2. Participant's assessment of disease activity (PtGA) on a 0-100 mm VAS (0=very good and 100=very poor).
3. Health Assessment Questionnaire- disability index (HAQ-DI): 20 questions across 8 domains assessing the functional abilities. The total score was calculated as the average of the scores for each domain, ranging from 0 (no disability) to 3 (very severe disability).
4. Number of active joints (68 joints evaluated for pain/tenderness and 66 for swelling)
5. Index of inflammation: C-reactive Protein (CRP) levels Additionally, participants were required to have no intermittent fever during the preceding week.

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

Adapted ACR50 response was defined as 50% reduction between baseline in at least 3 of the 5 response variables, with no more than one of these variables worsening by more than 30%:

1. Physicians global assessment of disease activity on a 0-100 mm VAS (0=very good and 100=very poor).
2. Participant's assessment of disease activity (PtGA) on a 0-100 mm VAS (0=very good and 100=very poor).
3. Health Assessment Questionnaire- disability index (HAQ-DI): 20 questions across 8 domains assessing the functional abilities. The total score was calculated as the average of the scores for each domain, ranging from 0 (no disability) to 3 (very severe disability).
4. Number of active joints (68 joints evaluated for pain/tenderness and 66 for swelling)
5. Index of inflammation: C-reactive Protein (CRP) levels Additionally, participants were required to have no intermittent fever during the preceding week.

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

Adapted ACR70 response was defined as 70% reduction between baseline in at least 3 of the 5 response variables, with no more than one of these variables worsening by more than 30%:

1. Physicians global assessment of disease activity (PhGA) on a 0-100 mm VAS (0=very good and 100=very poor).
2. Participant's assessment of disease activity (PtGA) on a 0-100 mm VAS (0=very good and 100=very poor).
3. Health Assessment Questionnaire- disability index (HAQ-DI): 20 questions across 8 domains assessing the functional abilities. The total score was calculated as the average of the scores for each domain, ranging from 0 (no disability) to 3 (very severe disability).
4. Number of active joints (68 joints evaluated for pain/tenderness and 66 for swelling)
5. Index of inflammation: C-reactive Protein (CRP) levels Additionally, participants were required to have no intermittent fever during the preceding week.

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

Adapted ACR90 response was defined as 90% reduction between baseline in at least 3 of the 5 response variables, with no more than one of these variables worsening by more than 30%:

1. Physicians global assessment of disease activity (PhGA) on a 0-100 mm VAS (0=very good and 100=very poor).
2. Participant's assessment of disease activity (PtGA) on a 0-100 mm VAS (0=very good and 100=very poor).
3. Health Assessment Questionnaire- disability index (HAQ-DI): 20 questions across 8 domains assessing the functional abilities. The total score was calculated as the average of the scores for each domain, ranging from 0 (no disability) to 3 (very severe disability).
4. Number of active joints (68 joints evaluated for pain/tenderness and 66 for swelling)
5. Index of inflammation: C-reactive Protein (CRP) levels Additionally, participants were required to have no intermittent fever during the preceding week.

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

Adapted ACR100 response was defined as 100% reduction between baseline in at least 3 of the 5 response variables, with no more than one of these variables worsening by more than 30%:

1. Physicians global assessment of disease activity (PhGA) on a 0-100 mm VAS (0=very good and 100=very poor).
2. Participant's assessment of disease activity (PtGA) on a 0-100 mm VAS (0=very good and 100=very poor).
3. Health Assessment Questionnaire- disability index (HAQ-DI): 20 questions across 8 domains assessing the functional abilities. The total score was calculated as the average of the scores for each domain, ranging from 0 (no disability) to 3 (very severe disability).
4. Number of active joints (68 joints evaluated for pain/tenderness and 66 for swelling)
5. Index of inflammation: C-reactive Protein (CRP) levels Additionally, participants were required to have no intermittent fever during the preceding week.

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

The SFS evaluated 10 features, each scored as 1 (present) or 0 (absent). The total score ranged from 0 (none present) to 10 (all present). A negative change from baseline indicated clinical improvement.

Clinical features:

• Fever (>37.5°C for ≥5 consecutive days)
• Rash (salmon-pink during febrile episodes)
• Serositis
• Lymphadenopathy (>1.5 cm)
• Hepatomegaly/splenomegaly (confirmed by ultrasound or palpation)

Laboratory features:

At baseline, considered present if:

• ESR ≥20 mm/hour
• CRP ≥10 mg/L
• WBC count ≥12×10⁹/L
• Hemoglobin ≤11 g/dL
• Platelets ≥400×10⁹/L

During treatment and follow-up:

• ESR: 0 if <20 mm/hour or decreased ≥30%; 1 if increased or decreased <30%
• CRP: 0 if <10 mg/L or decreased ≥30%; 1 if increased or decreased <30%
• WBC count: 0 if <12×10⁹/L or decreased ≥20%; 1 if increased or decreased <20%
• Hemoglobin: 0 if >11 g/dL or increased ≥20%; 1 if decreased or increased <20%
• Platelets: 0 if <400×10⁹/L or decreased ≥20%; 1 if increased or decreased <20%

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

The physician rated the participant's disease activity on a 0-100 mm VAS, ranging from no disease activity (0 mm) to very severe disease activity (100 mm). A negative change from baseline indicated improvement

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

The participant's assessment of disease activity was assessed on the VAS. The VAS scale ranges from 0-100 mm, from no pain/very well (0 mm) to very severe pain/very poor (100 mm). A negative change from baseline indicated improvement

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

The HAQ was used to assess physical ability and functional status of participants as well as quality of life. The disability dimension consisted of 20 multiple choice items concerning difficulty in performing eight common activities of daily living; dressing, arising, eating, walking, hygiene, reach, grip and usual activities. Each item is scored from 0 (no difficulty) to 3 (unable to do). The score for each domain is the highest item score, and the total score is the mean of the domain scores., ranging from 0 (no disability) to 3 (severe disability). A negative change from baseline indicated improvement.

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

Active joints were defined as joints with swelling or pain/tenderness. 68 joints were assessed for pain/tenderness and 66 joints for swelling.

The number of active joints ranged from 0 (no active joints) to 68 (all joints tender and /or swollen). A negative change from baseline indicated improvement.

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

CRP levels in the blood were determined A negative change from baseline indicated improvement in systemic inflammation

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

The number of participants without intermittent fever associated to AOSD (defined as oral, rectal, or axillary body temperature > 38°C only for several hours during the day) was assessed.

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

The oral corticosteroid dose was derived from the prednisone equivalent dose per day. The change from baseline was assessed. A negative change from baseline indicated reduced corticosteroid use.

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

The absence or presence of skin rash was assessed based on physical exam findings including whether it is typical or/and atypical.

Participants with a rash showing both typical and atypical features were counted in both sub-categories.

In the summary table, for each visit participants are classified as: rash present (typical and/or atypical), rash present (typical), rash present (atypical), or rash absent.

For every visit, the denominator used to calculate percentages corresponds to the number of participants with available data at that visit.

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

DAS28-CRP was a composite index validated for patients with rheumatoid arthritis (RA). It took into account the following components: tender joint count (0-28), swollen joint count (0-28), C-reactive protein (CRP, mg/L), and the Patient's Global Assessment of Disease Activity (Global Health), rated from 0 (best) to 100 (worst). These results were combined to produce the DAS28-CRP score, which ranged from 1.0 (disease remission) to 9.4 (high disease activity). A negative change from baseline in DAS28-CRP indicated an improvement

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with non-missing values were included in the analysis

Serum canakinumab concentrations by visit

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with evaluable values were included in the analysis

Total IL-1β (sum of IL-1β free and bound to canakinumab) in serum was assessed

FAS: all participants who received at least one dose of study treatment. At each time point, only participants with evaluable values were included in the analysis

The ADAs against canakinumab were assessed in serum using a validated immunoassay assay. The number of participants who had a canakinumab ADA positive result was assessed

AS: all participants who received at least one dose of study treatment. At each time point, only participants with evaluable values were included in the analysis