The Mayo Clinic Score (MCS) ranges from 0 to 12 and is a composite of the four following assessments of disease activity: stool frequency subscore, rectal bleeding subscore, endoscopy subscore, and physician's global assessment (PGA) subscore. Each of the four assessments was rated with a score from 0 to 3, with higher scores indicating more severe disease. Remission was defined as MCS less than or equal to (≤)2 with individual subscores ≤1 and a rectal bleeding subscore of 0. Participants were also classified as non-remitters if Week 10 assessments were missing or if they had received permitted/prohibited rescue therapy prior to assessment. Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9/MCS ≥10); the Cochran-Mantel-Haenszel test adjusted the difference in remission rates and associated 95% confidence interval for the stratification factors.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

The Mayo Clinic Score (MCS) ranges from 0 to 12 and is a composite of the four following assessments of disease activity: stool frequency subscore, rectal bleeding subscore, endoscopy subscore, and physician's global assessment (PGA) subscore. Each of the four assessments was rated with a score from 0 to 3, with higher scores indicating more severe disease. Remission was defined as MCS less than or equal to (≤)2 with individual subscores ≤1 and a rectal bleeding subscore of 0. Participants were also classified as non-remitters if Week 10 assessments were missing or if they had received permitted/prohibited rescue therapy prior to assessment. Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9/MCS ≥10). The Cochran-Mantel-Haenszel test adjusted the difference in remission rates and associated 95% confidence interval for the stratification factors.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

The Mayo Clinic Score (MCS) ranges from 0 to 12 and is a composite of the four following assessments of disease activity: stool frequency subscore, rectal bleeding subscore, endoscopy subscore, and physician's global assessment (PGA) subscore. Each of the four assessments was rated with a score from 0 to 3, with higher scores indicating more severe disease. Remission was defined as MCS less than or equal to (≤)2 with individual subscores ≤1 and a rectal bleeding subscore of 0. Participants were also classified as non-remitters if Week 10 assessments were missing or if they had received permitted/prohibited rescue therapy prior to assessment. Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9/MCS ≥10). The Cochran-Mantel-Haenszel test adjusted the difference in remission rates and associated 95% confidence interval for the stratification factors.

GA28948 & GA28949 Pooled Population, Modified Intent-to-Treat: including all randomized participants in studies GA28948 & GA28949 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

The Mayo Clinic Score (MCS) ranges from 0 to 12 and is a composite of the four following assessments of disease activity: stool frequency subscore, rectal bleeding subscore, endoscopy subscore, and physician's global assessment subscore. Each of the four assessments was rated with a score from 0 to 3, with higher scores indicating more severe disease. Clinical Response was defined as: MCS ≥3-point decrease and 30% reduction from baseline as well as ≥1-point decrease in rectal bleeding subscore or an absolute rectal bleeding score of 0 or 1. Non-responders also included participants with missing Week 10 assessments or those who had received permitted/prohibited rescue therapy prior to assessment. Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9 or ≥10); the CMH test adjusted the differences in response rates and associated 95% CIs for the stratification factors.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

The Mayo Clinic Score (MCS) ranges from 0 to 12 and is a composite of the four following assessments of disease activity: stool frequency subscore, rectal bleeding subscore, endoscopy subscore, and physician's global assessment subscore. Each of the four assessments was rated with a score from 0 to 3, with higher scores indicating more severe disease. Clinical Response was defined as: MCS ≥3-point decrease and 30% reduction from baseline as well as ≥1-point decrease in rectal bleeding subscore or an absolute rectal bleeding score of 0 or 1. Non-responders also included participants with missing Week 10 assessments or those who had received permitted/prohibited rescue therapy prior to assessment. Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9 or ≥10); the CMH test adjusted the differences in response rates and associated 95% CIs for the stratification factors.

GA28948 & GA28949 Pooled Population, Modified Intent-to-Treat: including all randomized participants in studies GA28948 & GA28949 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

Improvement in endoscopic appearance of the mucosa was defined as a Mayo Clinic Score (MCS) endoscopy subscore ≤1. Blinded gastroenterologists experienced in inflammatory bowel disease performed central reading of endoscopies at an independent review facility. The rectum, sigmoid, and descending colon segments were assessed and each segment was assigned a score of 0 to 3, with higher scores indicating more severe disease. At baseline all segments were reviewed and the worst score from the three segments was recorded as the endoscopy subscore. Post-baseline the endoscopy score was the worst score of all segments that had been assessed at baseline, if the baseline endoscopy score had a sigmoid colon score ≤1. If at baseline the sigmoid colon score was ≥2, the post-baseline endoscopy score was the sigmoid colon score value. Non-responders also included participants with missing Week 10 assessments or those who had received permitted/prohibited rescue therapy prior to assessment.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

Improvement in endoscopic appearance of the mucosa was defined as a Mayo Clinic Score (MCS) endoscopy subscore ≤1. Blinded gastroenterologists experienced in inflammatory bowel disease performed central reading of endoscopies at an independent review facility. The rectum, sigmoid, and descending colon segments were assessed and each segment was assigned a score of 0 to 3, with higher scores indicating more severe disease. At baseline all segments were reviewed and the worst score from the three segments was recorded as the endoscopy subscore. Post-baseline the endoscopy score was the worst score of all segments that had been assessed at baseline, if the baseline endoscopy score had a sigmoid colon score ≤1. If at baseline the sigmoid colon score was ≥2, the post-baseline endoscopy score was the sigmoid colon score value. Non-responders also included participants with missing Week 10 assessments or those who had received permitted/prohibited rescue therapy prior to assessment.

GA28948 & GA28949 Pooled Population, Modified Intent-to-Treat: including all randomized participants in studies GA28948 & GA28949 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

Endoscopic remission was defined as a Mayo Clinic Score (MCS) endoscopy subscore of 0. Blinded gastroenterologists experienced in inflammatory bowel disease performed central reading of endoscopies at an independent review facility. The rectum, sigmoid, and descending colon segments were assessed and each segment was assigned a score of 0 to 3, with higher scores indicating more severe disease. At baseline all segments were reviewed and the worst score from the three segments was recorded as the endoscopy subscore. Post-baseline the endoscopy score was the worst score of all segments that had been assessed at baseline, if the baseline endoscopy score had a sigmoid colon score ≤1. If at baseline the sigmoid colon score was ≥2, the post-baseline endoscopy score was the sigmoid colon score value. Non-responders also included participants with missing Week 10 assessments or those who had received permitted/prohibited rescue therapy prior to assessment.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

Endoscopic remission was defined as a Mayo Clinic Score (MCS) endoscopy subscore of 0. Blinded gastroenterologists experienced in inflammatory bowel disease performed central reading of endoscopies at an independent review facility. The rectum, sigmoid, and descending colon segments were assessed and each segment was assigned a score of 0 to 3, with higher scores indicating more severe disease. At baseline all segments were reviewed and the worst score from the three segments was recorded as the endoscopy subscore. Post-baseline the endoscopy score was the worst score of all segments that had been assessed at baseline, if the baseline endoscopy score had a sigmoid colon score ≤1. If at baseline the sigmoid colon score was ≥2, the post-baseline endoscopy score was the sigmoid colon score value. Non-responders also included participants with missing Week 10 assessments or those who had received permitted/prohibited rescue therapy prior to assessment.

GA28948 & GA28949 Pooled Population, Modified Intent-to-Treat: including all randomized participants in studies GA28948 & GA28949 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

Histologic remission is defined by the resolution of neutrophilic inflammation (e.g., absence of neutrophils in the crypts and lamina propria), defined by a Nancy Histological Index (NHI) score of ≤1. The NHI score ranges from 0 to 4, with the following definitions for each grade: 0 is no histologically significant disease; 1 is chronic inflammatory infiltrate with no acute inflammatory infiltrate; and 2, 3, and 4 are mildly, moderately, and severely active disease, respectively. A small pool of central readers who were blinded to both treatment arm and timepoint performed the histologic scoring. The same reader scored all slides for a given participant based on biopsies from the most inflamed region of the sigmoid colon. Participants were also classified as non-remitters if Week 10 assessments were missing or if they had received rescue therapy prior to assessment. The Cochran-Mantel-Haenszel test adjusted the difference in remission rates and 95% CI for the stratification factors.

GA28948 Histology-Evaluable Population: included all randomized participants in study GA28948 who received at least one dose of study drug and had documented neutrophilic inflammation (i.e., NHI >1) at baseline. This excludes participants who had no baseline histology assessment or had no indication of neutrophilic inflammation at baseline.

Histologic remission is defined by the resolution of neutrophilic inflammation (e.g., absence of neutrophils in the crypts and lamina propria), defined by a Nancy Histological Index (NHI) score of ≤1. The NHI score ranges from 0 to 4, with the following definitions for each grade: 0 is no histologically significant disease; 1 is chronic inflammatory infiltrate with no acute inflammatory infiltrate; and 2, 3, and 4 are mildly, moderately, and severely active disease, respectively. A small pool of central readers who were blinded to both treatment arm and timepoint performed the histologic scoring. The same reader scored all slides for a given participant based on biopsies from the most inflamed region of the sigmoid colon. Participants were also classified as non-remitters if Week 10 assessments were missing or if they had received rescue therapy prior to assessment. The Cochran-Mantel-Haenszel test adjusted the difference in remission rates and 95% CI for the stratification factors.

GA28948 & GA28949 Pooled, Histology-Evaluable Population: included all randomized participants in studies GA28948 & GA28949 who received at least one dose of study drug and had documented neutrophilic inflammation (i.e., NHI >1) at baseline. This excludes participants who had no baseline histology assessment or had no indication of neutrophilic inflammation at baseline.

Rectal bleeding data were collected via the participant's diaries and each day a participant provided a score from 0 to 3 according to the following definitions: 0 = no blood in the stool; 1 = streaks of blood with stool less than half the time; 2 = obvious blood with stool most of the time; 3 = blood alone passed. The Mayo Clinic Score (MCS) rectal bleeding subscore was calculated as the worst value of three days of daily diary scores closest to anchor dates at baseline and post-baseline. The data was considered non-parametric and was reported using RANK analysis of covariance (ANCOVA). Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9/MCS ≥10); the model adjusted for these stratification factors along with the baseline rectal bleeding (RB) subscore.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

Rectal bleeding data were collected via the participant's diaries and each day a participant provided a score from 0 to 3 according to the following definitions: 0 = no blood in the stool; 1 = streaks of blood with stool less than half the time; 2 = obvious blood with stool most of the time; 3 = blood alone passed. The Mayo Clinic Score (MCS) rectal bleeding subscore was calculated as the worst value of three days of daily diary scores closest to anchor dates at baseline and post-baseline. The data was considered non-parametric and was reported using RANK analysis of covariance (ANCOVA). Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9/MCS ≥10); the model adjusted for these stratification factors along with the baseline rectal bleeding (RB) subscore.

GA28948 & GA28949 Pooled Population, Modified Intent-to-Treat: including all randomized participants in studies GA28948 & GA28949 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

Stool frequency data were collected via the participant's diaries and each day a participant provided a score from 0 to 3 according to the following definitions: 0 = normal number of stools; 1 = 1 to 2 more stools than normal; 2 = 3 to 4 more stools than normal; 3 = 5 or more stools than normal. The Mayo Clinic Score (MCS) stool frequency subscore was calculated as the average of three days daily diary scores closest to anchor dates at baseline and post-baseline. The data was considered non-parametric and was reported using RANK analysis of covariance (ANCOVA). Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9/MCS ≥10); the model adjusted for these stratification factors along with the baseline stool frequency (SF) subscore.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

Stool frequency data were collected via the participant's diaries and each day a participant provided a score from 0 to 3 according to the following definitions: 0 = normal number of stools; 1 = 1 to 2 more stools than normal; 2 = 3 to 4 more stools than normal; 3 = 5 or more stools than normal. The Mayo Clinic Score (MCS) stool frequency subscore was calculated as the average of three days daily diary scores closest to anchor dates at baseline and post-baseline. The data was considered non-parametric and was reported using RANK analysis of covariance (ANCOVA). Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9/MCS ≥10); the model adjusted for these stratification factors along with the baseline stool frequency (SF) subscore.

GA28948 & GA28949 Pooled Population, Modified Intent-to-Treat: including all randomized participants in studies GA28948 & GA28949 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

The UC-PRO/SS questionnaire was collected in the e-diary and completed by participants for at least 9 to 12 consecutive days prior to a study visit. The bowel movement domain score ranges from 0 to 27, with a higher score indicating a worse disease state. The most recent 7 daily scores available (not including the visit) were selected for the calculation of the visit score. For each item in the questionnaire, a score was calculated for a visit by taking the average of the selected daily scores. The domain score for a visit was calculated as the sum of the averaged items for each question. A Mixed Model for Repeated Measures (MMRM) analysis of the data included the fixed categorical effects of treatment, visit, study stratification factors, and treatment-by-visit interaction, and the continuous covariates of the baseline UC-PRO/SS domain and baseline UC-PRO/SS domain-by-visit interaction. An unstructured covariance matrix was used to model the within-patient errors within the MMRM.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization. The analysis only included participants with non-missing results at baseline and at least one post-baseline timepoint.

The UC-PRO/SS questionnaire was collected in the e-diary and completed by participants for at least 9 to 12 consecutive days prior to a study visit. The bowel movement domain score ranges from 0 to 27, with a higher score indicating a worse disease state. The most recent 7 daily scores available (not including the visit) were selected for the calculation of the visit score. For each item in the questionnaire, a score was calculated for a visit by taking the average of the selected daily scores. The domain score for a visit was calculated as the sum of the averaged items for each question. A Mixed Model for Repeated Measures (MMRM) analysis of the data included the fixed categorical effects of treatment, visit, study stratification factors, and treatment-by-visit interaction, and the continuous covariates of the baseline UC-PRO/SS domain and baseline UC-PRO/SS domain-by-visit interaction. An unstructured covariance matrix was used to model the within-patient errors within the MMRM.

GA28948 & GA28949 Pooled Population, Modified Intent-to-Treat: including all randomized participants in studies GA28948 & GA28949 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization. The analysis only included participants with non-missing results at baseline and at least one post-baseline timepoint.

The UC-PRO/SS questionnaire was collected in the e-diary and completed by participants for at least 9 to 12 consecutive days prior to a study visit. The functional symptoms domain score ranges from 0 to 12, with a higher score indicating a worse disease state. The most recent 7 daily scores available (not including the visit) were selected for the calculation of the visit score. For each item in the questionnaire, a score was calculated for a visit by taking the average of the selected daily scores. The domain score for a visit was calculated as the sum of the averaged items for each question. A Mixed Model for Repeated Measures (MMRM) analysis of the data included the fixed categorical effects of treatment, visit, study stratification factors, and treatment-by-visit interaction, and the continuous covariates of the baseline UC-PRO/SS domain and baseline UC-PRO/SS domain-by-visit interaction. An unstructured covariance matrix was used to model the within-patient errors in the MMRM.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization. The analysis only included participants with non-missing results at baseline and at least one post-baseline timepoint.

The UC-PRO/SS questionnaire was collected in the e-diary and completed by participants for at least 9 to 12 consecutive days prior to a study visit. The functional symptoms domain score ranges from 0 to 12, with a higher score indicating a worse disease state. The most recent 7 daily scores available (not including the visit) were selected for the calculation of the visit score. For each item in the questionnaire, a score was calculated for a visit by taking the average of the selected daily scores. The domain score for a visit was calculated as the sum of the averaged items for each question. A Mixed Model for Repeated Measures (MMRM) analysis of the data included the fixed categorical effects of treatment, visit, study stratification factors, and treatment-by-visit interaction, and the continuous covariates of the baseline UC-PRO/SS domain and baseline UC-PRO/SS domain-by-visit interaction. An unstructured covariance matrix was used to model the within-patient errors in the MMRM.

GA28948 & GA28949 Pooled Population, Modified Intent-to-Treat: including all randomized participants in studies GA28948 & GA28949 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization. The analysis only included participants with non-missing results at baseline and at least one post-baseline timepoint.

The Mayo Clinic Score (MCS) ranges from 0 to 12 and is a composite of the four following assessments of disease activity: stool frequency subscore, rectal bleeding subscore, endoscopy subscore, and physician's global assessment (PGA) subscore. Each of the four assessments was rated with a score from 0 to 3, with higher scores indicating more severe disease. Clinical remission was defined as MCS less than or equal to (≤)2 with individual subscores ≤1. Participants were also classified as non-remitters if Week 10 assessments were missing or if they had received permitted/prohibited rescue therapy prior to assessment. Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9/MCS ≥10). The Cochran-Mantel-Haenszel test adjusted the differences in remission rates and associated 95% confidence intervals for the stratification factors.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

The Mayo Clinic Score (MCS) ranges from 0 to 12 and is a composite of the four following assessments of disease activity: stool frequency subscore, rectal bleeding subscore, endoscopy subscore, and physician's global assessment (PGA) subscore. Each of the four assessments was rated with a score from 0 to 3, with higher scores indicating more severe disease. Remission was defined as MCS less than or equal to (≤)2 with individual subscores ≤1 and a rectal bleeding subscore of 0. Participants were also classified as non-remitters if Week 10 or 14 assessments were missing or the participant received permitted/prohibited rescue therapy prior to assessment. Participants were stratified by concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening (MCS ≤9/MCS ≥10). The Cochran-Mantel-Haenszel test adjusted the differences in remission rates and associated 95% confidence intervals for the stratification factors.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization.

The IBDQ is a 32-item questionnaire containing four domains: bowel symptoms (10 items), systemic symptoms (5 items), emotional function (12 items), and social function (5 items). An overall total IBDQ score was computed by summing the individual 32-item scores. The range for the IBDQ total score is 32 to 224, with higher scores denoting better health-related quality of life. The unadjusted mean and standard deviation for each study arm are reported. The change from baseline in the IBDQ score was analyzed using an ANCOVA model taking the stratification factors used at randomization into account (concomitant treatment with corticosteroids or immunosuppressants at randomization and disease activity measured during screening [MCS ≤9/MCS ≥10]), and the baseline IBDQ score used as a covariate.

GA28948 Population, Modified Intent-to-Treat: including all randomized participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment assigned at randomization. The analysis only included participants with non-missing results at baseline and at least one post-baseline timepoint.

Serum concentrations of etrolizumab were evaluated at the primary endpoint visit (Week 10) and the secondary endpoint visit (Week 14). Both time points were two weeks after the most recent dose.

GA28948 Pharmacokinetics Evaluable Population: includes participants in study GA28948 who had received at least one dose of study drug and had at least one quantifiable concentration measured post-baseline. Only participants who were treated with etrolizumab were included in this analysis.

An adverse event (AE) is any untoward medical occurrence in a clinical investigation in which a patient is administered a pharmaceutical product, regardless of causal attribution. The investigator independently assessed the severity and seriousness of each recorded AE. The AE severity grading scale for the NCI CTCAE v4.0 was used for assessing severity; any AE not specifically listed was rated according to the following grading scale from 1 to 5: 1 = mild, 2 = moderate, 3 = severe, 4 = life-threatening, 5 = death. AEs of special interest included: elevated AST/ALT in combination with either elevated bilirubin or clinical jaundice; suspected transmission of infectious agent by the study drug; anaphylactic, anaphylactoid and systemic hypersensitivity reactions; and neurological signs, symptoms, and AEs that may suggest possible progressive multifocal leukoencephalopathy (PML).

GA28948 Safety Population: includes all participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment received.

Laboratory tests for hematology parameters were performed and values were compared with the Roche marked reference range. A marked abnormality was defined as a test result that was outside of the Roche marked reference range (labelled as 'High' or 'Low') and represented a clinically significant change from baseline. Not every laboratory abnormality qualified as an adverse event. A laboratory test result must have been reported as an adverse event if it met any of the following criteria: was accompanied by clinical symptoms; resulted in a change in study treatment or a medical intervention; or was clinically significant in the investigator's judgment. The results are presented as a shift from the baseline status to the Week 10 status. Baseline was defined as the last available assessment prior to first receipt of study drug. The 'missing' status only included participants with missing post-baseline values given they had a result at baseline. Abs = absolute count; Ery. = erythrocyte

GA28948 Safety Population: includes all participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment received.

Laboratory tests for chemistry parameters were performed and values were compared with the Roche marked reference range. A marked abnormality was defined as a test result that was outside of the Roche marked reference range (labelled as 'High' or 'Low') and represented a clinically significant change from baseline. Not every laboratory abnormality qualified as an adverse event. A laboratory test result must have been reported as an adverse event if it met any of the following criteria: was accompanied by clinical symptoms; resulted in a change in study treatment or a medical intervention; or was clinically significant in the investigator's judgment. The results are presented as a shift from the baseline status to the Week 10 status. Baseline was defined as the last available assessment prior to first receipt of study drug. The 'missing' status only included participants with missing post-baseline values given they had a result at baseline.

GA28948 Safety Population: includes all participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment received.

Anti-drug antibody (ADA) serum samples were collected from participants and analyzed using validated assays. Participants were considered to be ADA positive post-baseline if they were ADA negative or had missing data at baseline, but developed an ADA response following etrolizumab drug exposure (treatment-induced ADA response), or if they were ADA positive at baseline and the titer of one or more post-baseline samples was at least 0.60 titer units greater than the titer of the baseline sample (treatment-enhanced ADA response). Participants were considered to be ADA negative if they were ADA negative or had missing data at baseline and all post-baseline samples were negative, or if they were ADA positive at baseline but did not have any post-baseline samples with a titer that was at least 0.60 titer unit greater than the titer of the baseline sample (treatment unaffected).

GA28948 Safety Population: includes all participants in study GA28948 who received at least one dose of study drug, with participants grouped according to the treatment received. The analysis of ADAs at baseline and post-baseline only included participants who received treatment with etrolizumab.