The American College of Rheumatology (ACR) response is a measurement of improvement in multiple disease assessment criteria. The ACR20 response is defined as: 1) ≥ 20% improvement from baseline in SJC66, and 2) ≥ 20% improvement from baseline in tender TJC68, and 3) ≥ 20% improvement from baseline in at least 3 of the following 5 items: 1. Pain visual analog scale (VAS) (taken from the Health Assessment Questionnaire - Disability Index [HAQ-DI]), 2. Patient's Global Assessment of Disease Activity VAS, 3. Physician's Global Assessment of Disease Activity VAS, 4. Total HAQ-DI score, and 5. CRP. Non-responder imputation was used (ie, to impute a missing response, the participant was assumed to be a non-responder).

Intent-to-treat (ITT) population included all participants in the safety population who had post-randomization data for at least one efficacy parameter.

ACR20 response was defined as: 1) ≥ 20% improvement from baseline in SJC66, and 2) ≥ 20% improvement from baseline in TJC68, and 3) ≥ 20% improvement from baseline in at least 3 of the following 5 items: 1. Pain VAS (taken from the HAQ-DI), 2. Patient's Global Assessment of Disease Activity VAS, 3. Physician's Global Assessment of Disease Activity VAS, 4. Total HAQ-DI score, and 5. CRP. Non-responder imputation was used.

All patients from the ITT population who were randomized to a GLPG0634 arm. The data for participants switching from placebo to GLPG0634 100 mg at Week 12 were not in scope for this analysis and were not reported.

ACR50 response was defined as: 1) ≥ 50% improvement from baseline in SJC66, and 2) ≥ 50% improvement from baseline in TJC68, and 3) ≥ 50% improvement from baseline in at least 3 of the following 5 items: 1. Pain VAS (taken from the HAQ-DI) 2. Patient's Global Assessment of Disease Activity VAS 3. Physician's Global Assessment of Disease Activity VAS 4. Total HAQ-DI score 5. CRP. Non-responder imputation was used. All placebo participants switched to GLPG0634 treatment at Week 12 and were not included in the analysis of Week 24.

ITT population. Participants who switched treatment at Week 12 were handled as if they discontinued at Week 12.

ACR70 response: 1) ≥ 70% improvement from baseline in SJC66, and 2) ≥ 70% improvement from baseline in TJC68, and 3) ≥ 70% improvement from baseline in at least 3 of the following 5 items: 1. Pain VAS (taken from the HAQ-DI), 2. Patient's Global Assessment of Disease Activity VAS, 3. Physician's Global Assessment of Disease Activity VAS, 4. Total HAQ-DI score, and 5. CRP. Non-responder imputation was used. All placebo participants switched to GLPG0634 treatment at Week 12 and were not included in the analysis of Week 24.

ITT population with available data were analyzed. Participants who switched treatment at Week 12 were handled as if they discontinued at Week 12.

The ACR-N is the smallest percentage improvement in swollen and tender joints and the median of the remaining 5 core parameters, and is expected to be more sensitive to change than the ACR20, ACR50 or ACR70. It is a number varying between 0 and 100, with higher numbers indicating less severity of symptoms. Last observation carried forward (LOCF) algorithm was used (ie, to impute a missing value, the last preceding nonmissing value was used). All placebo participants switched to GLPG0634 treatment at Week 12 and were not included in the analysis of Week 24.

ITT population with available data were analyzed. Participants who switched treatment at Week 12 were handled as if they discontinued at Week 12.

DAS28 (CRP) was categorized into EULAR response categories (none, moderate, good) as follows: None = Actual DAS28 (CRP) ≤ 3.2, > 3.2 to ≤ 5.1, or > 5.1 AND Improvement in DAS28 (CRP) from baseline ≤ 6.0 or > 0.6 to ≤ 1.2; Moderate = Actual DAS28 (CRP) ≤ 3.2 AND Improvement in DAS28 (CRP) from baseline > 0.6 to ≤ 1.2, Actual DAS28 (CRP) > 3.2 to ≤ 5.1 or > 5.1 AND Improvement in DAS28 (CRP) from baseline > 1.2, or Actual DAS28 (CRP) > 3.2 to ≤ 5.1 AND Improvement in DAS28 (CRP) from baseline > 0.6 to ≤ 1.2; Good = Actual DAS28 (CRP) ≤ 3.2 AND Improvement in DAS28 (CRP) from baseline > 1.2. LOCF algorithm was used. All placebo participants switched to GLPG0634 treatment at Week 12 and were not included in the analysis of Week 24.

ITT population with available data were analyzed. Participants who switched treatment at Week 12 were handled as if they discontinued at Week 12.

A participant's disease activity status can be defined as being in remission when scores on the TJC28, SJC28, CRP (actual value in mg/dL) and Patient Global Assessment of Disease Activity (cm) are all ≤ 1. Non-responder imputation was used. All placebo participants switched to GLPG0634 treatment at Week 12 and were not included in the analysis of Week 24.

ITT population with available data were analyzed. Participants who switched treatment at Week 12 were handled as if they discontinued at Week 12.

The SDAI is the numerical sum of 5 outcome parameters: TJC28, SJC28, Patient Global Assessment of Disease Activity (in cm), Physician's Global Assessment of Disease Activity (in cm), and CRP (mg/dL). The SDAI was categorized as follows: • High disease activity: SDAI > 26 • Moderate disease activity: 11 to 26 • Low disease activity: 3.3 to 11 • Remission: ≤ 3.3. LOCF algorithm was used. The SDAI total score ranges from 0 to approximately 86. All placebo participants switched to GLPG0634 treatment at Week 12 and were not included in the analysis of Week 24.

ITT population with available data were analyzed. Participants who switched treatment at Week 12 were handled as if they discontinued at Week 12.

The CDAI is the SDAI modified to exclude CRP and is the sum of the 4 outcome parameters: TJC28, SJC28, Patient Global Assessment of Disease Activity (in cm), and Physician's Global Assessment of Disease Activity (in cm). The CDAI was be categorized as follows: • High disease activity: > 22 • Moderate disease activity: 10 to 22 • Mild disease activity: 2.8 to 10 • Remission: ≤ 2.8. LOCF algorithm was used. The CDAI total score ranges from 0 to approximately 76. All placebo participants switched to GLPG0634 treatment at Week 12 and were not included in the analysis of Week 24.

ITT population with available data were analyzed. Participants who switched treatment at Week 12 were handled as if they discontinued at Week 12.

FACIT-Fatigue scale is a 13-item questionnaire, each scored on a 5-point scale: 0 (Not at all) to 4 (Very much). The larger the participant's response to the questions (with the exception of 2 negatively stated that are scored reversely), the greater the fatigue. The sum of all responses resulted in the FACIT-Fatigue score for a total possible score of 0 (worse score) to 52 (better score), with a higher score indicating a better quality of life. LOCF algorithm was used. All placebo participants switched to GLPG0634 treatment at Week 12 and were not included in the analysis of Week 24.

ITT population with available data were analyzed. Participants who switched treatment at Week 12 were handled as if they discontinued at Week 12. All placebo participants switched to GLPG0634 treatment at Week 12 and were not included in the analysis of Week 24.

The SF-36 is a 36-item questionnaire measuring 8 domains (physical functioning, role physical, bodily pain, general health, vitality, social functioning, role emotional, and mental health). Each domain score ranges from 0 (worst) to 100 (best), with higher scores reflecting better health-related functional status. Two summary scale scores were computed based on weighted combinations of the 8 domain scores: the Physical Component Summary (PCS) and the Mental Component Summary (MCS). LOCF algorithm was used. All placebo participants switched to GLPG0634 treatment at Week 12 and were not included in the analysis of Week 24.

ITT population with available data were analyzed. Participants who switched treatment at Week 12 were handled as if they discontinued at Week 12.