ACR20 Responder Index is a composite of clinical, laboratory, and functional measures in rheumatoid arthritis (RA). An ACR20 Responder is a participant who had ≥20% improvement from baseline in both 68 tender and 66 swollen joint counts and ≥20% improvement in at least 3 of 5 criteria: Patient's and Physician's Global Assessment of Disease Activity, Health Assessment Questionnaire-Disability Index (HAQ-DI) (assessment of participant's physical function), pain due to RA, and hsCRP. Participants who discontinue before analysis time point are treated as non-responders. Percentage of participants achieving ACR20 response = (number of ACR20 responders) / (number of participants analyzed) * 100.

All randomized participants who received placebo, 4 mg or 8 mg LY3009104 in Part A. Participants who had missing components of the ACR20 index at Week 12 had these components imputed by last observation carried forward (LOCF).

ACR20 Responder Index is a composite of clinical, laboratory, and functional measures in RA. An ACR20 Responder is a participant who had ≥20% improvement from baseline in both 68 tender and 66 swollen joint counts and ≥20% improvement in at least 3 of 5 criteria: Patient's and Physician's Global Assessment of Disease Activity, HAQ-DI (assessment of participant's physical function), pain due to RA, and hsCRP. Participants who discontinue before analysis time point are treated as non-responders. Percentage of participants achieving ACR20 response = (number of ACR20 responders) / (number of participants analyzed) * 100. Data presented are model-based Bayesian posterior mean response rates with 95% credible interval.

All randomized participants who received study drug in Part A. Participants who had missing components of the ACR20 index at Week 12 had these components imputed by LOCF.

ACR20 Responder Index is a composite of clinical, laboratory, and functional measures in RA. An ACR20 Responder is a participant who had ≥20% improvement from baseline in both 68 tender and 66 swollen joint counts and ≥20% improvement in at least 3 of 5 criteria: Patient's and Physician's Global Assessment of Disease Activity, HAQ-DI (assessment of participant's physical function), pain due to RA, and hsCRP. Participants who discontinue before analysis time points are treated as non-responders. Percentage of participants achieving ACR20 response = (number of ACR20 responders) / (number of participants treated) * 100.

All randomized participants who received study drug in Parts A and B. Participants who had missing components of the ACR20 index at analysis time points had these components imputed by LOCF.

ACR20 Responder Index is a composite of clinical, laboratory, and functional measures in RA. An ACR20 Responder is a participant who had ≥20% improvement from baseline in both 68 tender and 66 swollen joint counts and ≥20% improvement in at least 3 of 5 criteria: Patient's and Physician's Global Assessment of Disease Activity, HAQ-DI (assessment of participant's physical function), pain due to RA, and hsCRP. Participants who discontinue before analysis time points are treated as non-responders. Percentage of participants achieving ACR20 response = (number of ACR20 responders) / (number of participants analyzed) * 100.

All participants who received study drug in Parts C and D. Participants who had missing components of the ACR20 index at analysis time points had these components imputed by LOCF.

ACR50 Responder Index is a composite of clinical, laboratory, and functional measures in RA. An ACR50 Responder is a participant who had ≥50% improvement from baseline in both 68 tender and 66 swollen joint counts and ≥50% improvement in at least 3 of 5 criteria: Patient's and Physician's Global Assessment of Disease Activity, HAQ-DI (assessment of participant's physical function), pain due to RA, and hsCRP. Participants who discontinue before analysis time points are treated as non-responders. Percentage of participants achieving ACR50 response = (number of ACR50 responders) / (number of participants analyzed) * 100.

All randomized participants who received study drug in Parts A and B. Participants who had missing components of the ACR50 index at analysis time points had these components imputed by LOCF.

ACR50 Responder Index is a composite of clinical, laboratory, and functional measures in RA. An ACR50 Responder is a participant who had ≥50% improvement from baseline in both 68 tender and 66 swollen joint counts and ≥50% improvement in at least 3 of 5 criteria: Patient's and Physician's Global Assessment of Disease Activity, HAQ-DI (assessment of participant's physical function), pain due to RA, and hsCRP. Participants who discontinue before analysis time points are treated as non-responders. Percentage of participants achieving ACR50 response = (number of ACR50 responders) / (number of participants analyzed) * 100.

All participants who received study drug in Parts C and D. Participants who had missing components of the ACR50 index at analysis time points had these components imputed by LOCF.

ACR70 Responder Index is a composite of clinical, laboratory, and functional measures in RA. An ACR70 Responder is a participant who had ≥70% improvement from baseline in both 68 tender and 66 swollen joint counts and ≥70% improvement in at least 3 of 5 criteria: Patient's and Physician's Global Assessment of Disease Activity, HAQ-DI (assessment of participant's physical function), pain due to RA, and hsCRP. Participants who discontinue before analysis time points are treated as non-responders. Percentage of participants achieving ACR70 response = (number of ACR70 responders) / (number of participants analyzed) * 100.

All randomized participants who received study drug in Parts A and B. Participants who had missing components of the ACR70 index at analysis time points had these components imputed by LOCF.

ACR70 Responder Index is a composite of clinical, laboratory, and functional measures in RA. An ACR70 Responder is a participant who had ≥70% improvement from baseline in both 68 tender and 66 swollen joint counts and ≥70% improvement in at least 3 of 5 criteria: Patient's and Physician's Global Assessment of Disease Activity, HAQ-DI (assessment of participant's physical function), pain due to RA, and hsCRP. Participants who discontinue before analysis time points are treated as non-responders. Percentage of participants achieving ACR70 response = (number of ACR70 responders) / (number of participants analyzed) * 100.

All participants who received study drug in Parts C and D. Participants who had missing components of the ACR70 index at analysis time points had these components imputed by LOCF.

ACR50 Responder Index is a composite of clinical, laboratory, and functional measures in RA. An ACR50 Responder is a participant who had ≥50% improvement from baseline in both 68 tender and 66 swollen joint counts and ≥50% improvement in at least 3 of 5 criteria: Patient's and Physician's Global Assessment of Disease Activity, HAQ-DI (assessment of participant's physical function), pain due to RA, and hsCRP. Participants who discontinue before analysis time point are treated as non-responders. Percentage of participants achieving ACR50 response = (number of ACR50 responders) / (number of participants analyzed) * 100. Data presented are model-based Bayesian posterior mean response rates with 95% credible interval.

All randomized participants who received study drug in Part A. Participants who had missing components of the ACR50 index at analysis time point had these components imputed by LOCF.

ACR-N is a continuous measure of clinical, laboratory, and functional outcomes in RA that characterizes percentage (%) of improvement in disease activity from baseline based on ACR core set. This index was calculated as minimum of a) % of improvement in TJC, b) % of improvement in SJC, and c) third highest percentage of improvement of remaining 5 ACR core criteria: If ≥3 components of the 5 ACR core criteria were missing, then c) was set to missing; if any of 3 components a), b), or c) were missing, then ACR-N was set to missing. Percentage of improvement was truncated to a range of -100 to 100 to minimize impact of outliers (greater scores indicate greater % improvement) and negative scores indicate a decline. Data presented are model-based Bayesian posterior mean response rates with 95% credible interval.

All randomized participants who received study drug in Part A. Participants who had missing components of the ACR-N at Week 12 had these components imputed by LOCF.

TJC is the number of tender and painful joints determined for each participant by examination of 68 joints. Joints were assessed by pressure and joint manipulation on physical examination. Participants were asked for pain sensations on these manipulations and watched for spontaneous pain reactions. Any positive response on pressure, movement, or both was translated into a single tender-versus-nontender dichotomy. SJC is the number of swollen joints determined for each participant by examination of 66 joints. Joints were classified as either swollen or not swollen. Swelling was defined as palpable fluctuating synovitis of the joint.

All randomized participants who received study drug in Parts A and B and had TJC and SJC evaluated at analysis time points. LOCF was used to impute missing post-baseline values

TJC is the number of tender and painful joints determined for each participant by examination of 68 joints. Joints were assessed by pressure and joint manipulation on physical examination. Participants were asked for pain sensations on these manipulations and watched for spontaneous pain reactions. Any positive response on pressure, movement, or both was translated into a single tender-versus-nontender dichotomy. SJC is the number of swollen joints determined for each participant by examination of 66 joints. Joints were classified as either swollen or not swollen. Swelling was defined as palpable fluctuating synovitis of the joint.

All participants who received study drug in Parts C and D and had TJC and SJC evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

The HAQ-DI questionnaire assesses the participant's self-perception on the degree of difficulty [0 (without any difficulty), 1 (with some difficulty), 2 (with much difficulty), and 3 (unable to do)] when dressing and grooming, arising, eating, walking, hygiene, reaching, gripping, and performing other daily activities. Scores for each functional area were averaged to calculate HAQ-DI scores, which ranged from 0 (no disability) to 3 (worst disability). A decrease in HAQ-DI score indicated an improvement in the participant's condition.

All randomized participants who received study drug in Parts A and B and had HAQ-DI evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

The HAQ-DI questionnaire assesses the participant's self-perception on the degree of difficulty [0 (without any difficulty), 1 (with some difficulty), 2 (with much difficulty), and 3 (unable to do)] when dressing and grooming, arising, eating, walking, hygiene, reaching, gripping, and performing other daily activities. Scores for each functional area were averaged to calculate HAQ-DI scores, which ranged from 0 (no disability) to 3 (worst disability). A decrease in HAQ-DI score indicated an improvement in the participant's condition.

All participants who received study drug in Parts C and D and had HAQ-DI evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

hsCRP is a laboratory analyte that is an indicator of inflammation. Decreases in hsCRP represent reductions in inflammation.

All randomized participants who received study drug in Parts A and B and had hsCRP evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

hsCRP is a laboratory analyte that is an indicator of inflammation. Decreases in hsCRP represent reductions in inflammation.

All participants who received study drug in Parts C and D and had hsCRP evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

ESR is a laboratory analyte that is an indicator of inflammation. Decreases represent reductions in inflammation.

All randomized participants who received study drug in Parts A and B and had ESR evaluated at analysis time points.

ESR is a laboratory analyte that is an indicator of inflammation. Decreases represent reductions in inflammation.

All participants who received study drug in Parts C and D and had ESR evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

Physician's and Patient's Assessments of Disease Activity (DA) assessed using a visual analog scale (VAS) that ranged from 0 to 100 millimeters (mm), where 0 indicated no arthritis activity and 100 indicated extremely active arthritis. Patient's assessment of pain due to arthritis was also assessed using a VAS that ranged from 0 (no pain) to 100 mm (worst possible pain).

All randomized participants who received study drug in Parts A and B and had physician's and participant's assessments of disease activity and participant's pain evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

Physician's and Patient's assessments of DA assessed using a VAS that ranged from 0 to 100 mm, where 0 indicated no arthritis activity and 100 indicated extremely active arthritis. Patient's assessment of pain due to arthritis assessed using a VAS that ranged from 0 (no pain) to 100 mm (worst possible pain).

All participants who received study drug in Parts C and D and had physician's and participant's assessments of disease activity and pain evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

Disease Activity Score (DAS) modified to include 28 joint count (DAS28) consisted of composite score of following variables: tender joint count-28 (TJC28), swollen joint count-28 (SJC28), CRP (mg/L), and Patient's Global Assessment of Disease Activity using VAS (patient's global VAS). DAS28-CRP=0.56*square root (sqrt)(TJC28)+0.28*sqrt(SJC28)+0.36*natural log(CRP+1)+0.014*patient's global VAS+0.96. Scores ranged from 1.0-9.4, where lower scores indicated less disease activity, and remission was DAS28-CRP <2.6. A decrease in DAS28-CRP indicated an improvement in participant's condition.

All randomized participants who received study drug in Parts A and B and had DAS28-CRP evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

Disease Activity Score (DAS) modified to include 28 joint count (DAS28) consisted of composite score of following variables: tender joint count (TJC28), swollen joint count (SJC28), CRP (mg/L), and Patient's Global Assessment of Disease Activity using visual analog scale (VAS) (patient's global VAS). DAS28-CRP=0.56*square root (sqrt)(TJC28)+0.28*sqrt(SJC28)+0.36*natural log(CRP+1)+0.014*patient's global VAS+0.96. Scores ranged from 1.0-9.4, where lower scores indicated less disease activity and remission is DAS28-CRP <2.6. A decrease in DAS28-CRP indicated an improvement in participant's condition.

All participants who received study drug in Parts C and D and had DAS28-CRP evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

EULAR28 categorizes clinical response based upon improvement since baseline in DAS modified to include the 28-joint count (DAS28) and post-baseline DAS28. DAS28 consists of a composite score of the following variables: TJC28, SJC28, CRP, and Patient's Global Assessment of their Disease Activity (patient's global VAS). DAS28 scores range from 1.0-9.4. EULAR28 categories include: No Response (improvement in DAS28 of ≤0.6 units or post-baseline DAS28 score >5.1 with improvement by ≤1.2 units), Moderate Response (post-baseline DAS28 ≤5.1 with improvement by >0.6 units but ≤1.2 units or post-baseline DAS28 score >3.2 with improvement by >1.2 units), and Good Response (post-baseline DAS28 score ≤3.2 with improvement by >1.2 units).

All randomized participants who received study drug in Parts A and B and had EULAR28 evaluated at analysis time points.

EULAR28 categorizes clinical response based upon improvement since baseline in Disease Activity Score modified to include the 28-joint count (DAS28) and post-baseline DAS28. DAS28 consists of a composite score of the following variables: tender joint count (TJC28), swollen joint count (SJC28), CRP, and Patient's Global Assessment of their Disease Activity (patient's global VAS). DAS28 scores range from 1.0-9.4. EULAR28 categories include: No Response (improvement in DAS28 of ≤0.6 units or post-baseline DAS28 score >5.1 with improvement by ≤1.2 units), Moderate Response (post-baseline DAS28 ≤5.1 with improvement by >0.6 units but ≤1.2 units or post-baseline DAS28 score >3.2 with improvement by >1.2 units), and Good Response (post-baseline DAS28 score ≤3.2 with improvement by >1.2 units).

All participants who received study drug in Parts C and D and had EULAR28 evaluated at analysis time points. LOCF was used to impute missing post-baseline values.

Disease Activity Score (DAS) modified to include 28-joint count (DAS28) consisted of composite score of following variables: tender joint count (TJC28), swollen joint count (SJC28), CRP [milligrams per liter (mg/L)], and Patient's Global Assessment of Disease Activity using VAS (patient's global VAS). Scores ranged from 1.0-9.4, where lower scores indicated less disease activity. DAS28 scores ≤3.2 are considered as low disease activity, and scores <2.6 are considered as remission. Participants who discontinue before analysis time points are treated as non-responders.

All randomized participants who received study drug in Parts A and B and had DAS28-CRP evaluated at analysis time points.

Disease Activity Score (DAS) modified to include 28 joint count (DAS28) consisted of composite score of following variables: tender joint count (TJC28), swollen joint count (SJC28), CRP (mg/L), and Patient's Global Assessment of Disease Activity using VAS (patient's global VAS). Scores ranged from 1.0-9.4, where lower scores indicated less disease activity. DAS28 scores ≤3.2 are considered as low disease activity, and scores <2.6 are considered as remission. Participants who discontinue before analysis time points are treated as non-responders.

All participants who received study drug in Parts C and D and had DAS28-CRP evaluated at analysis time points.

The Investigator asked participants about the duration of their morning stiffness (in minutes) in and around the joints and recorded the duration. The Investigator asked the participants about duration of morning stiffness on the day prior to the study visit to capture actual symptoms. If morning stiffness duration was longer than 12 hours (720 minutes), it was truncated to 720 minutes for statistical presentations and analyses.

All randomized participants who received study drug in Part A and had morning stiffness evaluated at analysis time points. LOCF was used to impute missing post-baseline values for Week 12 analysis.

The SF-36 is a health-related survey that assesses participant's quality of life and consists of 36 questions covering 8 health domains (physical functioning, bodily pain, role limitations due to physical problems and also emotional problems, general health, mental health, social functioning and vitality) and 2 component scores (PCS and MCS). The PCS score consisted of physical functioning, bodily pain, role-physical, and general health scales. The MCS score consisted of social functioning, vitality, mental health, and role-emotional scales. Both PCS and MCS range from 0-100 with higher scores indicating better health or functioning.

All randomized participants who received study drug in Part A and had SF-36 evaluated at analysis time point. LOCF was used to impute missing post-baseline values.

The BPI-sf modified is a self-administered questionnaire developed for the rapid assessment of pain. The BPI-sf modified provides information on the intensity of pain (the sensory dimension) as well as the degree to which pain interferes with function (the reactive dimension). The questionnaire asks questions about pain relief, pain quality, and the participant's perception of the cause of pain. The BPI-sf modified uses a numeric rating scale from 0 ("No pain") to 10 ("Pain as bad as you can imagine"). Since pain can be quite variable over a day, the BPI-sf modified asked participants to rate their pain at the time of responding to the questionnaire (right now), and also at its worst, least and average over the last 24 hours.

All randomized participants who received study drug and had BPI-sf worst-pain-in-the past-24-hours item evaluated at Week 12. LOCF was used to impute missing post-baseline values.

The FACIT-F Scale is a brief 13-item, symptom-specific questionnaire that specifically assesses the participant self-reported severity of fatigue and its impact upon daily activities and functioning. The FACIT-F uses a numeric rating scale of 0 ("Not at all") to 4 ("Very much") for each item to assess fatigue and its impact in the past 7 days. Total scores range from 0 to 52, with higher scores indicating less fatigue.

All randomized participants who received study drug in Part A and had FACIT-F evaluated at Week 12. LOCF was used to impute missing post-baseline values.

All randomized participants who received at least 1 dose of LY3009104 with evaluable LY3009104 PK data.

All randomized participants who received at least 1 dose of LY3009104 with evaluable LY3009104 PK data.

Zero participants were analyzed. Assessment of ENSEMBLE Minimum Data Set 1.0 was not collected at Week 12 and therefore results are not reported for outcome measure.