Etanercept Withdrawal And Retreament Study In Subjects With Nr-ax SpA

Participants who experienced ASDAS-Erythrocyte Sedimentation Rate (ESR) level of \>=2.1 were defined as being flared. ASDAS is a score combining the assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on a VAS ranging from 0-10 cm, where 0 = no disease activity and 10= high disease activity. CRP measured in milligram per liter (mg/L) and ESR measured in millimeter per hour (mm/hr). Percentage of participants who flared within 40 weeks after the withdrawal of Etanercept treatment of 24 weeks in Induction period are reported in this outcome measure.

Time frame: Within 40 weeks after Etanercept withdrawal (from Week 24 to Week 64)

Population: Full analysis set for period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed last observation carried forward (LOCF).

ASDAS: score combining assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on VAS ranging 0-10 cm, where 0= no disease activity and 10= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS ranged as inactive disease: 0\<= ASDAS-CRP \<1.3; moderate disease activity: 1.3\<= ASDAS-CRP \<2.1; high disease activity: 2.1\<= ASDAS-CRP \<=3.5; very high disease activity: 3.5\< ASDAS-CRP. The ASDAS-CRP is calculated with the following equation: 0.121\*total back pain+0.110\*participant global+0.073\*peripheral pain/swelling+0.058\*duration of morning stiffness+0.579\*Ln(CRP+1), Ln represents the natural logarithm.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS: score combining assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on VAS ranging 0-10 cm,where 0= no disease activity and 10= high disease activity.CRP measured in mg/L and ESR measured in mm/hr. ASDAS ranged as inactive disease: 0\<= ASDAS-CRP \<1.3; moderate disease activity: 1.3\<= ASDAS-CRP \<2.1; high disease activity: 2.1\<= ASDAS-CRP \<=3.5; very high disease activity: 3.5\< ASDAS-CRP. The ASDAS-CRP is calculated with the following equation: 0.121\*total back pain+0.110\*participant global+0.073\*peripheral pain/swelling+0.058\*duration of morning stiffness+0.579\*Ln(CRP+1), Ln represents the natural logarithm.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS: score combining assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on VAS ranging 0-10 cm, where 0= no disease activity and 10= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS ranged as inactive disease: 0\<= ASDAS-CRP \<1.3; moderate disease activity: 1.3\<= ASDAS-CRP \<2.1; high disease activity: 2.1\<= ASDAS-CRP \<=3.5; very high disease activity: 3.5\< ASDAS-CRP. The ASDAS-CRP is calculated with the following equation: 0.121\*total back pain+0.110\*participant global+0.073\*peripheral pain/swelling+0.058\*duration of morning stiffness+0.579\*Ln(CRP+1), Ln represents the natural logarithm.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3: Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS: score combining assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on VAS ranging 0-10 cm, where 0= no disease activity and 10= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS ranged as inactive disease: 0\<= ASDAS-CRP \<1.3; moderate disease activity: 1.3\<= ASDAS-CRP \<2.1; high disease activity: 2.1\<= ASDAS-CRP \<=3.5; very high disease activity: 3.5\< ASDAS-CRP. The ASDAS-CRP is calculated with the following equation: 0.121\*total back pain+0.110\*participant global+0.073\*peripheral pain/swelling+0.058\*duration of morning stiffness+0.579\*Ln(CRP+1), Ln represents the natural logarithm.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS: score combining assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on VAS ranging 0-10 cm, where 0= no disease activity and 10= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS ranged as inactive disease: 0\<= ASDAS-CRP \<1.3; moderate disease activity: 1.3\<= ASDAS-CRP \<2.1; high disease activity: 2.1\<= ASDAS-CRP \<=3.5; very high disease activity: 3.5\< ASDAS-CRP. The ASDAS-CRP is calculated with the following equation: 0.121\*total back pain+0.110\*participant global+0.073\*peripheral pain/swelling+0.058\*duration of morning stiffness+0.579\*Ln(CRP+1), Ln represents the natural logarithm.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS: score combining assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on VAS ranging 0-10 cm, where 0= no disease activity and 10= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS ranged as inactive disease: 0\<= ASDAS-CRP \<1.3; moderate disease activity: 1.3\<= ASDAS-CRP \<2.1; high disease activity: 2.1\<= ASDAS-CRP \<=3.5; very high disease activity: 3.5\< ASDAS-CRP. The ASDAS-CRP is calculated with the following equation: 0.121\*total back pain+0.110\*participant global+0.073\*peripheral pain/swelling+0.058\*duration of morning stiffness+0.579\*Ln(CRP+1), Ln represents the natural logarithm.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS is a score combining the assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on a VAS ranging from 0-10 cm, where 0 = no disease activity and 100= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS-ESR was calculated with the following equation: 0.8\*total back pain+0.11\*participant global+0.09\*peripheral pain/swelling+0.07\*duration of morning stiffness+ 0.29\*ESR\^1/2. ASDAS ranged as inactive disease: 0 \<= ASDAS-ESR \<1.3; active disease: 1.3 \<= ASDAS-ESR =\<2.1.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS is a score combining the assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on a VAS ranging from 0-10 cm, where 0 = no disease activity and 100= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS-ESR was calculated with the following equation: 0.8\*total back pain+0.11\*participant global+0.09\*peripheral pain/swelling+0.07\*duration of morning stiffness+ 0.29\*ESR\^1/2. ASDAS ranged as inactive disease: 0 \<= ASDAS-ESR \<1.3; active disease: 1.3 \<= ASDAS-ESR =\<2.1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS is a score combining the assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on a VAS ranging from 0-10 cm, where 0 = no disease activity and 100= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS-ESR was calculated with the following equation: 0.8\*total back pain+0.11\*participant global+0.09\*peripheral pain/swelling+0.07\*duration of morning stiffness+ 0.29\*ESR\^1/2. ASDAS ranged as inactive disease: 0 \<= ASDAS-ESR \<1.3; active disease: 1.3 \<= ASDAS-ESR =\<2.1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS is a score combining the assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on a VAS ranging from 0-10 cm, where 0 = no disease activity and 100= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS-ESR was calculated with the following equation: 0.8\*total back pain+0.11\*participant global+0.09\*peripheral pain/swelling+0.07\*duration of morning stiffness+ 0.29\*ESR\^1/2. ASDAS ranged as inactive disease: 0 \<= ASDAS-ESR \<1.3; active disease: 1.3 \<= ASDAS-ESR =\<2.1.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS is a score combining the assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on a VAS ranging from 0-10 cm, where 0 = no disease activity and 100= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS-ESR was calculated with the following equation: 0.8\*total back pain+0.11\*participant global+0.09\*peripheral pain/swelling+0.07\*duration of morning stiffness+ 0.29\*ESR\^1/2. ASDAS ranged as inactive disease: 0 \<= ASDAS-ESR \<1.3; active disease: 1.3 \<= ASDAS-ESR =\<2.1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

ASDAS is a score combining the assessment of back pain, peripheral pain/swelling, duration of morning stiffness, participant global assessment of disease activity and CRP or ESR. All parameters other than CRP or ESR assessed on a VAS ranging from 0-10 cm, where 0 = no disease activity and 100= high disease activity. CRP measured in mg/L and ESR measured in mm/hr. ASDAS-ESR was calculated with the following equation: 0.8\*total back pain+0.11\*participant global+0.09\*peripheral pain/swelling+0.07\*duration of morning stiffness+ 0.29\*ESR\^1/2. ASDAS ranged as inactive disease: 0 \<= ASDAS-ESR \<1.3; active disease: 1.3 \<= ASDAS-ESR =\<2.1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Here, number analyzed signifies participants evaluable at specific time points.

The BAS-G was a 2-question assessment evaluating the effect of AS on the participants well-being over the last week and last 6 months. Each question scored by the participant on a 100 mm VAS scale ranging from 0 (very Good) to 100 (very Bad), where higher scores indicated worse health condition. The total BAS-G score calculated as the average scores of these two questions and then converted into cm for analysis. Total BAS-G score ranged from 0 to 10 cm, where higher scores indicated worse health condition.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

The BAS-G was a 2-question assessment evaluating the effect of AS on the participants well-being over the last week and last 6 months. Each question scored by the participant on a 100 mm VAS scale ranging from 0 (very Good) to 100 (very Bad), where higher scores indicated worse health condition. The total BAS-G score calculated as the average scores of these two questions and then converted into cm for analysis. Total BAS-G score ranged from 0 to 10 cm, where higher scores indicated worse health condition.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Here, number analyzed signifies participants evaluable at specific time points.

The BAS-G was a 2-question assessment evaluating the effect of AS on the participants well-being over the last week and last 6 months. Each question scored by the participant on a 100 mm VAS scale ranging from 0 (very Good) to 100 (very Bad), where higher scores indicated worse health condition. The total BAS-G score calculated as the average scores of these two questions and then converted into cm for analysis. Total BAS-G score ranged from 0 to 10 cm, where higher scores indicated worse health condition.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

The BAS-G was a 2-question assessment evaluating the effect of AS on the participants well-being over the last week and last 6 months. Each question scored by the participant on a 100 mm VAS scale ranging from 0 (very Good) to 100 (very Bad), where higher scores indicated worse health condition. The total BAS-G score calculated as the average scores of these two questions and then converted into cm for analysis. Total BAS-G score ranged from 0 to 10 cm, where higher scores indicated worse health condition.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

The BAS-G was a 2-question assessment evaluating the effect of AS on the participants well-being over the last week and last 6 months. Each question scored by the participant on a 100 mm VAS scale ranging from 0 (very Good) to 100 (very Bad), where higher scores indicated worse health condition. The total BAS-G score calculated as the average scores of these two questions and then converted into cm for analysis. Total BAS-G score ranged from 0 to 10 cm, where higher scores indicated worse health condition.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: FAS Period 3: all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed = participants evaluable at specific time points.

The BAS-G was a 2-question assessment evaluating the effect of AS on the participants well-being over the last week and last 6 months. Each question scored by the participant on a 100 mm VAS scale ranging from 0 (very Good) to 100 (very Bad), where higher scores indicated worse health condition. The total BAS-G score calculated as the average scores of these two questions and then converted into cm for analysis. Total BAS-G score ranged from 0 to 10 cm, where higher scores indicated worse health condition.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Dactylitis is the inflammation of finger and/or toe joints (digits). Dactylitis scores was calculated by evaluating each of the 10 fingers and 10 toes. Each digit was evaluated on a 4-point scale ranging from of 0 to 3 where 0 = none, 1= mild, 2 = moderate, 3 = severe inflammation. The total score was calculated as the sum of scores for the 20 digits, total score ranged from 0 to 60, where higher scores indicated severe inflammation.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Dactylitis is the inflammation of finger and/or toe joints (digits). Dactylitis scores was calculated by evaluating each of the 10 fingers and 10 toes. Each digit was evaluated on a 4-point scale ranging from of 0 to 3 where 0 = none, 1= mild, 2 = moderate, 3 = severe inflammation. The total score was calculated as the sum of scores for the 20 digits, total score ranged from 0 to 60, where higher scores indicated severe inflammation.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Here, number analyzed signifies participants evaluable at specific time points.

Dactylitis is the inflammation of finger and/or toe joints (digits). Dactylitis scores was calculated by evaluating each of the 10 fingers and 10 toes. Each digit was evaluated on a 4-point scale ranging from of 0 to 3 where 0 = none, 1= mild, 2 = moderate, 3 = severe inflammation. The total score was calculated as the sum of scores for the 20 digits, total score ranged from 0 to 60, where higher scores indicated severe inflammation.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Dactylitis is the inflammation of finger and/or toe joints (digits). Dactylitis scores was calculated by evaluating each of the 10 fingers and 10 toes. Each digit was evaluated on a 4-point scale ranging from of 0 to 3 where 0 = none, 1= mild, 2 = moderate, 3 = severe inflammation. The total score was calculated as the sum of scores for the 20 digits, total score ranged from 0 to 60, where higher scores indicated severe inflammation.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

Dactylitis is the inflammation of finger and/or toe joints (digits). Dactylitis scores was calculated by evaluating each of the 10 fingers and 10 toes. Each digit was evaluated on a 4-point scale ranging from of 0 to 3 where 0 = none, 1= mild, 2 = moderate, 3 = severe inflammation. The total score was calculated as the sum of scores for the 20 digits, total score ranged from 0 to 60, where higher scores indicated severe inflammation.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: FAS Period 3: all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed = participants evaluable at specific time points.

Dactylitis is the inflammation of finger and/or toe joints (digits). Dactylitis scores was calculated by evaluating each of the 10 fingers and 10 toes. Each digit was evaluated on a 4-point scale ranging from of 0 to 3 where 0 = none, 1= mild, 2 = moderate, 3 = severe inflammation. The total score was calculated as the sum of scores for the 20 digits, total score ranged from 0 to 60, where higher scores indicated severe inflammation.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The EuroQol-5D is a five dimensional health state classification. Each dimension is assessed on a 3-point ordinal scale (1=no problems, 2=some problems, 3=severe problems). The responses to the five EQ-5D dimensions were scored using a utility-weighted algorithm to derive an EQ-5D health status index score between 0 to 1, with 1.00 indicating no problem and 0 representing ''worst health condition''.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

The EuroQol-5D is a five dimensional health state classification. Each dimension is assessed on a 3-point ordinal scale (1=no problems, 2=some problems, 3=severe problems). The responses to the five EQ-5D dimensions were scored using a utility-weighted algorithm to derive an EQ-5D health status index score between 0 to 1, with 1.00 indicating no problem and 0 representing ''worst health condition''.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Here, number analyzed signifies participants evaluable at specific time points.

The EuroQol-5D is a five dimensional health state classification. Each dimension is assessed on a 3-point ordinal scale (1=no problems, 2=some problems, 3=severe problems). The responses to the five EQ-5D dimensions were scored using a utility-weighted algorithm to derive an EQ-5D health status index score between 0 to 1, with 1.00 indicating no problem and 0 representing ''worst health condition''.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

The EuroQol-5D is a five dimensional health state classification. Each dimension is assessed on a 3-point ordinal scale (1=no problems, 2=some problems, 3=severe problems). The responses to the five EQ-5D dimensions were scored using a utility-weighted algorithm to derive an EQ-5D health status index score between 0 to 1, with 1.00 indicating no problem and 0 representing ''worst health condition''.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

The EuroQol-5D is a five dimensional health state classification. Each dimension is assessed on a 3-point ordinal scale (1=no problems, 2=some problems, 3=severe problems). The responses to the five EQ-5D dimensions were scored using a utility-weighted algorithm to derive an EQ-5D health status index score between 0 to 1, with 1.00 indicating no problem and 0 representing ''worst health condition''.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: FAS Period 3: all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed = participants evaluable at specific time points.

The EuroQol-5D is a five dimensional health state classification. Each dimension is assessed on a 3-point ordinal scale (1=no problems, 2=some problems, 3=severe problems). The responses to the five EQ-5D dimensions were scored using a utility-weighted algorithm to derive an EQ-5D health status index score between 0 to 1, with 1.00 indicating no problem and 0 representing ''worst health condition''.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The MASES is an index used to measure the severity of enthesitis. Enthesitis is the inflammation of enthuses (heels). The MASES assesses 13 sites for enthesitis. Each site is scored as 0 or 1 depending on whether enthesitis is present or absent. Sites assessed include 1st costochondral joint (left/right), 7 th costochondral joint (l/r), posterior superior iliac spine (l/r), posterior anterior iliac spine (l/r), iliac crest (l/r), proximal insertion of Achilles tendon (l/r) and 5th lumbar spinous process. The MASES is the sum of all site scores range from 0 (no inflammation) to 13 (worst possible inflammation) where higher scores indicate more severe inflammation of entheses.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The MASES is an index used to measure the severity of enthesitis. Enthesitis is the inflammation of enthuses (heels). The MASES assesses 13 sites for enthesitis. Each site is scored as 0 or 1 depending on whether enthesitis is present or absent. Sites assessed include 1st costochondral joint (left/right), 7 th costochondral joint (l/r), posterior superior iliac spine (l/r), posterior anterior iliac spine (l/r), iliac crest (l/r), proximal insertion of Achilles tendon (l/r) and 5th lumbar spinous process. The MASES is the sum of all site scores range from 0 (no inflammation) to 13 (worst possible inflammation) where higher scores indicate more severe inflammation of entheses.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The MASES is an index used to measure the severity of enthesitis. Enthesitis is the inflammation of enthuses (heels). The MASES assesses 13 sites for enthesitis. Each site is scored as 0 or 1 depending on whether enthesitis is present or absent. Sites assessed include 1st costochondral joint (left/right), 7 th costochondral joint (l/r), posterior superior iliac spine (l/r), posterior anterior iliac spine (l/r), iliac crest (l/r), proximal insertion of Achilles tendon (l/r) and 5th lumbar spinous process. The MASES is the sum of all site scores range from 0 (no inflammation) to 13 (worst possible inflammation) where higher scores indicate more severe inflammation of entheses.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2: Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

The MASES is an index used to measure the severity of enthesitis. Enthesitis is the inflammation of enthuses (heels). The MASES assesses 13 sites for enthesitis. Each site is scored as 0 or 1 depending on whether enthesitis is present or absent. Sites assessed include 1st costochondral joint (left/right), 7 th costochondral joint (l/r), posterior superior iliac spine (l/r), posterior anterior iliac spine (l/r), iliac crest (l/r), proximal insertion of Achilles tendon (l/r) and 5th lumbar spinous process. The MASES is the sum of all site scores range from 0 (no inflammation) to 13 (worst possible inflammation) where higher scores indicate more severe inflammation of entheses.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

The MASES is an index used to measure the severity of enthesitis. Enthesitis is the inflammation of enthuses (heels). The MASES assesses 13 sites for enthesitis. Each site is scored as 0 or 1 depending on whether enthesitis is present or absent. Sites assessed include 1st costochondral joint (left/right), 7 th costochondral joint (l/r), posterior superior iliac spine (l/r), posterior anterior iliac spine (l/r), iliac crest (l/r), proximal insertion of Achilles tendon (l/r) and 5th lumbar spinous process. The MASES is the sum of all site scores range from 0 (no inflammation) to 13 (worst possible inflammation) where higher scores indicate more severe inflammation of entheses.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: FAS Period 3: all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed = participants evaluable at specific time points.

The MASES is an index used to measure the severity of enthesitis. Enthesitis is the inflammation of enthuses (heels). The MASES assesses 13 sites for enthesitis. Each site is scored as 0 or 1 depending on whether enthesitis is present or absent. Sites assessed include 1st costochondral joint (left/right), 7 th costochondral joint (l/r), posterior superior iliac spine (l/r), posterior anterior iliac spine (l/r), iliac crest (l/r), proximal insertion of Achilles tendon (l/r) and 5th lumbar spinous process. The MASES is the sum of all site scores range from 0 (no inflammation) to 13 (worst possible inflammation) where higher scores indicate more severe inflammation of entheses.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Change from baseline in MRI score of spine was assessed using SPARCC method. Scoring was based on 6 consecutive coronal slices from posterior to anterior. Each joint was divided into 4 quadrants. Each quadrant was assigned score of 0=no lesion or 1=increased signal. This part of coring allows for total score ranging from 0-8 for the 2 joints of one coronal slice. For each slice, score was increased by 1 for each joint that exhibits an intense signal in any quadrant (thereby allowing for an increase of up to 2 points in total score for each slice). Also, for each slice, an additional score of 1 was given for each joint that includes a lesion demonstrating continuous increased signal of depth \>=1 cm from articular surface (thereby allowing for an additional increase of up to 2 points for each slice). The total minimum and maximum score for all joints across 6 slices was 0 to 72 where higher scores reflecting worse disease.

Time frame: Baseline (Day 1 Week 1), Week 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Change from baseline in MRI score of spine was assessed using SPARCC method. Scoring was based on 6 consecutive coronal slices from posterior to anterior. Each joint was divided into 4 quadrants. Each quadrant was assigned score of 0=no lesion or 1=increased signal. This part of coring allows for total score ranging from 0-8 for the 2 joints of one coronal slice. For each slice, score was increased by 1 for each joint that exhibits an intense signal in any quadrant (thereby allowing for an increase of up to 2 points in total score for each slice). Also, for each slice, an additional score of 1 was given for each joint that includes a lesion demonstrating continuous increased signal of depth \>=1 cm from articular surface (thereby allowing for an additional increase of up to 2 points for each slice). The total minimum and maximum score for all joints across 6 slices was 0 to 72 where higher scores reflecting worse disease.

Time frame: Baseline (Day 1 Week 1), Week 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Change from baseline in MRI score of spine was assessed using SPARCC method. Scoring was based on 6 consecutive coronal slices from posterior to anterior. Each joint was divided into 4 quadrants. Each quadrant was assigned score of 0=no lesion or 1=increased signal. This part of coring allows for total score ranging from 0-8 for the 2 joints of one coronal slice. For each slice, score was increased by 1 for each joint that exhibits an intense signal in any quadrant (thereby allowing for an increase of up to 2 points in total score for each slice). Also, for each slice, an additional score of 1 was given for each joint that includes a lesion demonstrating continuous increased signal of depth \>=1 cm from articular surface (thereby allowing for an additional increase of up to 2 points for each slice). The total minimum and maximum score for all joints across 6 slices was 0 to 72 where higher scores reflecting worse disease.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Change from baseline in MRI score of spine was assessed using SPARCC method. Scoring was based on 6 consecutive coronal slices from posterior to anterior. Each joint was divided into 4 quadrants. Each quadrant was assigned score of 0=no lesion or 1=increased signal. This part of coring allows for total score ranging from 0-8 for the 2 joints of one coronal slice. For each slice, score was increased by 1 for each joint that exhibits an intense signal in any quadrant (thereby allowing for an increase of up to 2 points in total score for each slice). Also, for each slice, an additional score of 1 was given for each joint that includes a lesion demonstrating continuous increased signal of depth \>=1 cm from articular surface (thereby allowing for an additional increase of up to 2 points for each slice). The total minimum and maximum score for all joints across 6 slices was 0 to 72 where higher scores reflecting worse disease.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Change from baseline in MRI score of spine was assessed using SPARCC method. Scoring was based on 6 consecutive coronal slices from posterior to anterior. Each joint was divided into 4 quadrants. Each quadrant was assigned score of 0=no lesion or 1=increased signal. This part of coring allows for total score ranging from 0-8 for the 2 joints of one coronal slice. For each slice, score was increased by 1 for each joint that exhibits an intense signal in any quadrant (thereby allowing for an increase of up to 2 points in total score for each slice). Also, for each slice, an additional score of 1 was given for each joint that includes a lesion demonstrating continuous increased signal of depth \>=1 cm from articular surface (thereby allowing for an additional increase of up to 2 points for each slice). The total minimum and maximum score for all joints across 6 slices was 0 to 72 where higher scores reflecting worse disease.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: FAS Period 3: all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed = participants evaluable at specific time points.

Change from baseline in MRI score of spine was assessed using SPARCC method. Scoring was based on 6 consecutive coronal slices from posterior to anterior. Each joint was divided into 4 quadrants. Each quadrant was assigned score of 0=no lesion or 1=increased signal. This part of coring allows for total score ranging from 0-8 for the 2 joints of one coronal slice. For each slice, score was increased by 1 for each joint that exhibits an intense signal in any quadrant (thereby allowing for an increase of up to 2 points in total score for each slice). Also, for each slice, an additional score of 1 was given for each joint that includes a lesion demonstrating continuous increased signal of depth \>=1 cm from articular surface (thereby allowing for an additional increase of up to 2 points for each slice). The total minimum and maximum score for all joints across 6 slices was 0 to 72 where higher scores reflecting worse disease.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Participants assessed their nocturnal back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to centimeter (cm) for analysis.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their nocturnal back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their nocturnal back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their nocturnal back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their nocturnal back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their nocturnal back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Here, number analyzed signifies participants evaluable at specific time points.

Number of swollen joints was determined by examination of 44 joints and identifying when swelling was present. The number of swollen joints was recorded on the joint assessment form at each visit, no swelling =0, swelling =1.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Number of swollen joints was determined by examination of 44 joints and identifying when swelling was present. The number of swollen joints was recorded on the joint assessment form at each visit, no swelling =0, swelling =1.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Number of swollen joints was determined by examination of 44 joints and identifying when swelling was present. The number of swollen joints was recorded on the joint assessment form at each visit, no swelling =0, swelling =1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Number of swollen joints was determined by examination of 44 joints and identifying when swelling was present. The number of swollen joints was recorded on the joint assessment form at each visit, no swelling =0, swelling =1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Number of swollen joints was determined by examination of 44 joints and identifying when swelling was present. The number of swollen joints was recorded on the joint assessment form at each visit, no swelling =0, swelling =1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: FAS Period 3: all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed = participants evaluable at specific time points.

Number of swollen joints was determined by examination of 44 joints and identifying when swelling was present. The number of swollen joints was recorded on the joint assessment form at each visit, no swelling =0, swelling =1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Number of tender joints was determined by examining 44 joints and identified the joints that were painful under pressure or to passive motion. The number of tender joints was recorded on the joint assessment form at each visit, no tenderness = 0, tenderness = 1.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Number of tender joints was determined by examining 44 joints and identified the joints that were painful under pressure or to passive motion. The number of tender joints was recorded on the joint assessment form at each visit, no tenderness = 0, tenderness = 1.

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Number of tender joints was determined by examining 44 joints and identified the joints that were painful under pressure or to passive motion. The number of tender joints was recorded on the joint assessment form at each visit, no tenderness = 0, tenderness = 1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Number of tender joints was determined by examining 44 joints and identified the joints that were painful under pressure or to passive motion. The number of tender joints was recorded on the joint assessment form at each visit, no tenderness = 0, tenderness = 1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Number of tender joints was determined by examining 44 joints and identified the joints that were painful under pressure or to passive motion. The number of tender joints was recorded on the joint assessment form at each visit, no tenderness = 0, tenderness = 1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: FAS Period 3: all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed = participants evaluable at specific time points.

Number of tender joints was determined by examining 44 joints and identified the joints that were painful under pressure or to passive motion. The number of tender joints was recorded on the joint assessment form at each visit, no tenderness = 0, tenderness = 1.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The physician assessed the overall disease activity of participants over the last 48 hours by using a 100 mm VAS pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The physician assessed the overall disease activity of participants over the last 48 hours by using a 100 mm VAS pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

The physician assessed the overall disease activity of participants over the last 48 hours by using a 100 mm VAS pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

The physician assessed the overall disease activity of participants over the last 48 hours by using a 100 mm VAS pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The physician assessed the overall disease activity of participants over the last 48 hours by using a 100 mm VAS pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

The physician assessed the overall disease activity of participants over the last 48 hours by using a 100 mm VAS pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Here, number analyzed signifies participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: FAS Period 3: all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed = participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores (physical component scores \[PCS\]; mental component scores \[MCS\]). Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 32, 48, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: FAS Period 3: all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed = participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

SF-36 is widely used generic quality of life instrument that assesses the participant's general health and functional status. SF-36 consists of 36 questions that grouped into 8 domains (physical functioning, vitality, social functioning, mental health, role physical, bodily pain, role emotional and general health). Domain scores range from 0 (worst value) to 100 (best value), with greater scores reflecting better health status. Scores of 8 health aspects were summarized to derive the 2 component scores: PCS, MCS. Four domains of the SF-36 comprises the PCS score (physical functioning, role-physical, bodily pain, and general health) and remaining 4 domains comprises of the MCS score (vitality, social functioning, role-emotional, and mental health). Each PCS and MCS are scored from 0 to 100 with higher scores indicating better health (0= worst value and 100= best value).

Time frame: Baseline (Day 1 Week 1), Week 12, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

Participants assessed their overall disease activity over the last 48 hours by using a 100 mm pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their overall disease activity over the last 48 hours by using a 100 mm pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their overall disease activity over the last 48 hours by using a 100 mm pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their overall disease activity over the last 48 hours by using a 100 mm pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their overall disease activity over the last 48 hours by using a 100 mm pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their overall disease activity over the last 48 hours by using a 100 mm pain scale that ranges from 0 mm (none) to 100 mm (severe pain), where higher scores indicated more pain. The reported values then converted to cm for analysis purposes.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their total back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their total back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their total back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their total back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their total back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

Participants assessed their total back pain over the last 48 hours on a 100 mm VAS scale ranged from 0 mm (none) to 100 mm (severe), where higher scores indicated more pain. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Here, number analyzed signifies participants evaluable at specific time points.

The WPAI assesses work productivity and impairment. It is a 6-item questionnaire used to assess the degree to which a specified health problem here AS affected work attendance, work productivity and productivity in non-work regular activities. Participants were asked to consider the past 7 days prior to each questionnaire day. The questionnaire asks: current employment status, hours worked, hours missed from work for any reason other than AS, hours missed from work due to AS, degree to which AS affected work productivity, and degree to a which AS affected non-work regular activities. Four component scores were then calculated: percent work time missed due to AS; percent impairment while working due to AS, percent overall work impairment due to AS, and percent non-work activity impairment due to AS. The computed percentage range for each sub-scale was from 0-100, with higher numbers indicating greater impairment and less productivity.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: FAS for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF .Here, Overall Number of Participants Analyzed included participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The WPAI assesses work productivity and impairment. It is a 6-item questionnaire used to assess the degree to which a specified health problem here AS affected work attendance, work productivity and productivity in non-work regular activities. Participants were asked to consider the past 7 days prior to each questionnaire day. The questionnaire asks: current employment status, hours worked, hours missed from work for any reason other than AS, hours missed from work due to AS, degree to which AS affected work productivity, and degree to a which AS affected non-work regular activities. Four component scores were then calculated: percent work time missed due to AS; percent impairment while working due to AS, percent overall work impairment due to AS, and percent non-work activity impairment due to AS. The computed percentage range for each sub-scale was from 0-100, with higher numbers indicating greater impairment and less productivity.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The WPAI assesses work productivity and impairment. It is a 6-item questionnaire used to assess the degree to which a specified health problem here AS affected work attendance, work productivity and productivity in non-work regular activities. Participants were asked to consider the past 7 days prior to each questionnaire day. The questionnaire asks: current employment status, hours worked, hours missed from work for any reason other than AS, hours missed from work due to AS, degree to which AS affected work productivity, and degree to a which AS affected non-work regular activities. Four component scores were then calculated: percent work time missed due to AS; percent impairment while working due to AS, percent overall work impairment due to AS, and percent non-work activity impairment due to AS. The computed percentage range for each sub-scale was from 0-100, with higher numbers indicating greater impairment and less productivity.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: FAS Period 3: all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Overall Number of Participants Analyzed = participants who were evaluable for this outcome measure and number analyzed = participants evaluable at specific time points.

The WPAI assesses work productivity and impairment. It is a 6-item questionnaire used to assess the degree to which a specified health problem here ankylosing spondylitis (AS) affected work attendance, work productivity and productivity in non-work regular activities. Participants were asked to consider the past 7 days prior to each questionnaire day. The questionnaire asks: current employment status, hours worked, hours missed from work for any reason other than AS, hours missed from work due to AS, degree to which AS affected work productivity, and degree to a which AS affected non-work regular activities. Four component scores were then calculated: percent work time missed due to AS; percent impairment while working due to AS, percent overall work impairment due to AS, and percent non-work activity impairment due to AS. The computed percentage range for each sub-scale was from 0-100, with higher numbers indicating greater impairment and less productivity.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: FAS for Period 1 included all participants who took study medication and had one evaluation after baseline. Here, Overall Number of Participants Analyzed included participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The WPAI assesses work productivity and impairment. It is a 6-item questionnaire used to assess the degree to which a specified health problem here AS affected work attendance, work productivity and productivity in non-work regular activities. Participants were asked to consider the past 7 days prior to each questionnaire day. The questionnaire asks: current employment status, hours worked, hours missed from work for any reason other than AS, hours missed from work due to AS, degree to which AS affected work productivity, and degree to a which AS affected non-work regular activities. Four component scores were then calculated: percent work time missed due to AS; percent impairment while working due to AS, percent overall work impairment due to AS, and percent non-work activity impairment due to AS. The computed percentage range for each sub-scale was from 0-100, with higher numbers indicating greater impairment and less productivity.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

The WPAI assesses work productivity and impairment. It is a 6-item questionnaire used to assess the degree to which a specified health problem here AS affected work attendance, work productivity and productivity in non-work regular activities. Participants were asked to consider the past 7 days prior to each questionnaire day. The questionnaire asks: current employment status, hours worked, hours missed from work for any reason other than AS, hours missed from work due to AS, degree to which AS affected work productivity, and degree to a which AS affected non-work regular activities. Four component scores were then calculated: percent work time missed due to AS; percent impairment while working due to AS, percent overall work impairment due to AS, and percent non-work activity impairment due to AS. The computed percentage range for each sub-scale was from 0-100, with higher numbers indicating greater impairment and less productivity.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Overall Number of Participants Analyzed signifies participants who were evaluable for this outcome measure and number analyzed signifies participants evaluable at specific time points.

BASDAI consisted of 6 questions related to disease activity. Each of the first 5 questions was scored by the participant on a 100 mm scale ranging from 0 (none) to 100 (very severe), where higher scores indicated more severe disease activity. The sixth question, related to duration of morning stiffness measured on a scale for 0 (0 hours) to 100 (2 hours), where higher scores indicated larger duration of morning stiffness. The BASDAI score was obtained by computing the mean score for the 2 questions related to morning stiffness (questions 5 \[severity of morning stiffness\] and 6 \[duration of morning stiffness\]) and then adding that value to the sum of the scores for the first 4 questions and then dividing the total by 5. This can be written as BASDAI=(Q1+Q2+Q3+Q4+(Q5+Q6)/2)/5. The BASDAI total score ranged from 0 to 10, where higher scores indicated more severe disease activity. The reported values were converted to cm for analysis.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

BASDAI consisted of 6 questions related to disease activity. Each of the first 5 questions was scored by the participant on a 100 mm scale ranging from 0 (none) to 100 (very severe), where higher scores indicated more severe disease activity. The sixth question, related to duration of morning stiffness measured on a scale for 0 (0 hours) to 100 (2 hours), where higher scores indicated larger duration of morning stiffness. The BASDAI score was obtained by computing the mean score for the 2 questions related to morning stiffness (questions 5 \[severity of morning stiffness\] and 6 \[duration of morning stiffness\]) and then adding that value to the sum of the scores for the first 4 questions and then dividing the total by 5. This can be written as BASDAI=(Q1+Q2+Q3+Q4+(Q5+Q6)/2)/5. The BASDAI total score ranged from 0 to 10, where higher scores indicated more severe disease activity. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

BASDAI consisted of 6 questions related to disease activity. Each of the first 5 questions was scored by the participant on a 100 mm scale ranging from 0 (none) to 100 (very severe), where higher scores indicated more severe disease activity. The sixth question, related to duration of morning stiffness measured on a scale for 0 (0 hours) to 100 (2 hours), where higher scores indicated larger duration of morning stiffness. The BASDAI score was obtained by computing the mean score for the 2 questions related to morning stiffness (questions 5 \[severity of morning stiffness\] and 6 \[duration of morning stiffness\]) and then adding that value to the sum of the scores for the first 4 questions and then dividing the total by 5. This can be written as BASDAI=(Q1+Q2+Q3+Q4+(Q5+Q6)/2)/5. The BASDAI total score ranged from 0 to 10, where higher scores indicated more severe disease activity. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

BASDAI consisted of 6 questions related to disease activity. Each of the first 5 questions was scored by the participant on a 100 mm scale ranging from 0 (none) to 100 (very severe), where higher scores indicated more severe disease activity. The sixth question, related to duration of morning stiffness measured on a scale for 0 (0 hours) to 100 (2 hours), where higher scores indicated larger duration of morning stiffness. The BASDAI score was obtained by computing the mean score for the 2 questions related to morning stiffness (questions 5 \[severity of morning stiffness\] and 6 \[duration of morning stiffness\]) and then adding that value to the sum of the scores for the first 4 questions and then dividing the total by 5. This can be written as BASDAI=(Q1+Q2+Q3+Q4+(Q5+Q6)/2)/5. The BASDAI total score ranged from 0 to 10, where higher scores indicated more severe disease activity. The reported values were converted to cm for analysis.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Here, number analyzed signifies participants evaluable at specific time points.

BASDAI consisted of 6 questions related to disease activity. Each of the first 5 questions was scored by the participant on a 100 mm scale ranging from 0 (none) to 100 (very severe), where higher scores indicated more severe disease activity. The sixth question, related to duration of morning stiffness measured on a scale for 0 (0 hours) to 100 (2 hours), where higher scores indicated larger duration of morning stiffness. The BASDAI score was obtained by computing the mean score for the 2 questions related to morning stiffness (questions 5 \[severity of morning stiffness\] and 6 \[duration of morning stiffness\]) and then adding that value to the sum of the scores for the first 4 questions and then dividing the total by 5. This can be written as BASDAI=(Q1+Q2+Q3+Q4+(Q5+Q6)/2)/5. The BASDAI total score ranged from 0 to 10, where higher scores indicated more severe disease activity. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2. Here, number analyzed signifies participants evaluable at specific time points.

BASDAI consisted of 6 questions related to disease activity. Each of the first 5 questions was scored by the participant on a 100 mm scale ranging from 0 (none) to 100 (very severe), where higher scores indicated more severe disease activity. The sixth question, related to duration of morning stiffness measured on a scale for 0 (0 hours) to 100 (2 hours), where higher scores indicated larger duration of morning stiffness. The BASDAI score was obtained by computing the mean score for the 2 questions related to morning stiffness (questions 5 \[severity of morning stiffness\] and 6 \[duration of morning stiffness\]) and then adding that value to the sum of the scores for the first 4 questions and then dividing the total by 5. This can be written as BASDAI=(Q1+Q2+Q3+Q4+(Q5+Q6)/2)/5. The BASDAI total score ranged from 0 to 10, where higher scores indicated more severe disease activity. The reported values were converted to cm for analysis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Here, number analyzed signifies participants evaluable at specific time points.

BASFI is composed of 10 questions related to the participant's ability to function. Each question scored by the participant on a 100 mm scale ranging from 0 (easy) to 100 (impossible), where higher scores indicated more difficulty in participant's ability to function. The BASFI total score calculated as mean of the scores for these 10 questions and converted to cm for analysis. BASFI total score was ranged from 0 (easy) to 10 (impossible), where higher scores indicated more difficulty in participant's ability to function due to ankylosing spondylitis.

Time frame: Baseline (Day 1 Week 1), Week 4, 8, 12, 16, 24

Population: Full analysis set for Period 1 included all participants who took study medication and had one evaluation after baseline. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

BASFI is composed of 10 questions related to the participant's ability to function. Each question scored by the participant on a 100 mm scale ranging from 0 (easy) to 100 (impossible), where higher scores indicated more difficulty in participant's ability to function. The BASFI total score calculated as mean of the scores for these 10 questions and converted to cm for analysis. BASFI total score was ranged from 0 (easy) to 10 (impossible), where higher scores indicated more difficulty in participant's ability to function due to ankylosing spondylitis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Week 28, 32, 40, 48, 56, 64

Population: Full analysis set for Period 2 included all participants who had at least one evaluation during period 2.Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.

BASFI is composed of 10 questions related to the participant's ability to function. Each question scored by the participant on a 100 mm scale ranging from 0 (easy) to 100 (impossible), where higher scores indicated more difficulty in participant's ability to function. The BASFI total score calculated as mean of the scores for these 10 questions and converted to cm for analysis. BASFI total score was ranged from 0 (easy) to 10 (impossible), where higher scores indicated more difficulty in participant's ability to function due to ankylosing spondylitis.

Time frame: Period 1 Baseline (Day 1 Week 1), Period 2 Baseline (last visit before treatment withdrawal), Period 3 Baseline (last visit before retreatment), Week 64, 68, 72, 76

Population: Full analysis set for Period 3 included all participants who took study retreatment medication and had at least one evaluation after restarting active therapy. Missing data was imputed using mixed LOCF. Here, number analyzed signifies participants evaluable at specific time points.