Assessment of pain severity using the Brief Pain Inventory (BPI) 4-item pain severity scale at 20 weeks: 0 (Better) - 10 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and post-treatment at 24 weeks (secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 12 and 2 additional Arm B subjects at week 20.

Assessment of pain interference using the Brief Pain Inventory (BPI) 7-item pain intensity scale at 20 weeks: 0 (Better) - 10 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and post-treatment at 24 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 12 and 1 additional Arm B subjects at week 20.

Assessment of pain using average of daily reported pain scores at 20 weeks; average of daily reported scores from days > week 16 and <= week 20. Pain scale: 0 (Better) - 10 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and post-treatment at 24 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 4 Arm A subjects at wk 0, 1 Arm A at wk 4, 1 Arm A at wk 12, and 2 Arm B at wk 0.

Assessment of pain using the % of days taking medication for pain at 20 weeks; percentage of days calculated from days > week 16 and <= week 20.Interim assessments were performed at 4, 8, 12 and 16 weeks and post-treatment at 24 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 4 and 1 additional Arm A subject at week 12.

Assessment of pain improvement using the patient global impression of change in pain scale at 20 weeks: 1 (Better) - 7 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and post-treatment at 24 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 12.

Jaw Function Limitation Scale (JFLS-8) at 20 weeks: 0 (Better) - 10 (Worse).Interim assessments were performed at 4, 8, 12 and 16 weeks and post-treatment at 24 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 4 and 1 additional Arm A subject at week 12.

Depressive and Anxiety symptoms using Patient Health Questionnaire (PHQ-4) at 20 weeks: 0 (Better) - 12 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and post-treatment at 24 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 12, 1 Arm B at week 4, and 1 Arm B at week 8.

Assessment of somatic symptoms using the Somatic Symptoms Scale (SSS-8) at 20 weeks: 0 (Better) - 32 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and post-treatment at 24 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 12.

Assessment of pain severity using the Brief Pain Inventory (BPI) 4-item pain severity scale at 24 weeks: 0 (Better) - 10 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and end of treatment at 20 weeks (primary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 12 and 2 additional Arm B subjects at week 20.

Assessment of pain interference using the Brief Pain Inventory (BPI) 7-item pain intensity scale at 24 weeks: 0 (Better) - 10 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and end of treatment at 20 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 12 and 1 additional Arm B subject at week 20.

Assessment of pain using average of daily reported pain scores at 24 weeks; average of daily reported scores from days > week 20 and <= week 24. Pain scale: 0 (Better) - 10 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and end of treatment at 20 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 4 Arm A subjects at wk 0, 1 Arm A at wk 4, 1 Arm A at wk 12, and 2 Arm B at wk 0.

Assessment of pain using the % of days taking medication for pain at 24 weeks; percentage of days calculated from days > week 20 and <= week 24. Interim assessments were performed at 4, 8, 12 and 16 weeks and end of treatment at 20 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 4 and 1 additional Arm A subject at week 12.

Assessment of pain improvement using the patient global impression of change in pain scale at 24 weeks: 1 (Better) - 7 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and end of treatment at 20 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 12.

Jaw Function Limitation Scale (JFLS-8) at 24 weeks: 0 (Better) - 10 (Worse).Interim assessments were performed at 4, 8, 12 and 16 weeks and end of treatment at 20 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 4 and 1 additional Arm A subject at week 12.

Depressive and Anxiety symptoms using Patient Health Questionnaire (PHQ-4) at 24 weeks: 0 (Better) - 12 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and end of treatment at 20 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 12, 1 Arm B at week 4, and 1 Arm B at week 8.

Assessment of somatic symptoms using the Somatic Symptoms Scale (SSS-8) at 24 weeks: 0 (Better) - 32 (Worse). Interim assessments were performed at 4, 8, 12 and 16 weeks and end of treatment at 20 weeks (additional secondary outcome).

Estimation done using mixed-model ANOVA which includes all available data from all time points for all subjects. This analysis reduces bias compared to an analysis at each time point that excludes subjects with missing data. All study subjects were included, even if they did not have data observed at the given time point. In addition to missing data shown in Participant Flow table, outcome-specific missing data for 1 additional Arm A subject at week 12.