Open-Label Extension and Safety Study for Participants With Crohn's Disease Previously Enrolled in the Etrolizumab Phase III Study GA29144: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Open-Label Extension and Safety Study for Participants With Crohn's Disease Previously Enrolled in the Etrolizumab Phase III Study GA29144 (NCT02403323), a randomized phase 3 study. The primary outcome is Part 1: Number of Participants With Crohn's Disease Activity Index (CDAI) Remission at 12-week Intervals.

2. Clinical Question

3. Trial Design

This is a Phase III, open-label extension and safety-monitoring study for participants previously enrolled in the etrolizumab Phase III protocol GA29144. Its two parts address different follow-up situations. In Part 1, eligible participants could continue etrolizumab. In Part 2, participants who stopped etrolizumab—either after leaving Part 1 or by entering directly from GA29144—were monitored for progressive multifocal leukoencephalopathy and other safety events.

The registry lists randomized allocation, no masking, and a single-group model. Those descriptors do not establish a randomized comparison for the primary outcome: the posted outcome describes a single etrolizumab extension group, and the record identifies fewer than two arms for that endpoint. There is no concurrent control group specified for interpreting the remission counts. The registry’s allocation label should therefore not be treated as evidence that the extension provides a controlled comparison.

Part 1 treatment was open-label etrolizumab once every 4 weeks, continuing until commercial availability in the participant’s country or the sponsor’s decision to terminate the study, whichever came earlier, up to approximately 10 years after the first participant was enrolled. The registry record lists a start date of 2015-06-08, a primary completion date of 2023-10-09, and a terminated status. These entries describe the study record; they do not state why the study was terminated or how long any individual participant was treated.

Enrollment is listed as 790, while the safety summaries use separate at-risk counts for Part 1 and Part 2. Those figures represent different study groups and should not be substituted for one another. The registry record does not report the reasons for the termination, participant flow between parts, or treatment exposure for each participant.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: This study consists of 2 parts, Part 1: Open-label extension (OLE) period \& Part 2: Progressive multifocal leukoencephalopathy (PML) safety monitoring (SM) period. A total of 790 participants were enrolled in the study, 751 participants in Part 1 and 359 participants in Part 2. Of the 359, 39 participants directly entered into the Part 2 PML SM period from study GA29144.

Part 1: Open Label Extension Period

Reasons for not completing this period:

Part 2: PML Safety Monitoring Period

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: OLE population included all participants who received at least one dose of study drug in Study GA29145 Part 1 (OLE). Number analyzed is the number of participants with data available for analysis at the specified timepoint.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: All Patients Population included all participants who were enrolled in the study.

6. Primary Endpoint

The registry lists the primary outcome as Part 1: Number of Participants With Crohn's Disease Activity Index (CDAI) Remission at 12-week Intervals, measured over Day 1 and Weeks 12, 24, 36, 48, 60, 72, 84, 96, 108, 120, 132, 144, 156, 168, 180, 192, 204, 216, 228, 240 and 252 of OLE.

The primary outcome is the number of participants with Crohn’s Disease Activity Index remission at Day 1 and scheduled 12-week intervals through Week 252 of the OLE. Remission is defined as a CDAI score less than 150. The index combines eight assessments: liquid or soft stools, abdominal pain, general well-being, complications, use of lomotil or other opiates for diarrhea, abdominal mass, hematocrit, and deviation from standard weight. Scores range from 0 to 600, with higher scores indicating worse disease activity; a decrease over time indicates improvement.

A remission count is a binary-outcome summary: each participant either meets the stated threshold at a particular assessment or does not. Counts at repeated visits can describe the observed number meeting the definition at each timepoint, but counts alone do not show how the same individuals’ status changed between visits. The registry notes that the number analyzed at each timepoint is the number with data available then, so the denominator may vary over follow-up.

Repeated measurement can be useful in an extension because the endpoint is assessed over a prolonged period rather than at a single visit. It can describe the observed pattern of remission in the enrolled extension population. However, the registry record does not report remission counts, proportions, confidence intervals, or a modeled comparison over time. It therefore supplies the endpoint definition and schedule, not an estimate of the observed remission pattern.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a binary remission outcome represent?

In general, a binary outcome classifies each participant according to whether a defined condition is met. For this endpoint, the condition is a CDAI score less than 150 at a scheduled assessment. The reported count would describe how many assessed participants met that criterion at that visit. A count is not itself a rate or proportion: interpretation of a proportion requires the relevant denominator, and the registry notes that the analyzed number is based on data available at the specified timepoint.

The definition is threshold-based. A participant just below the cutoff and a participant with a substantially lower score both meet the remission definition, while a participant at or above the cutoff does not. That makes the outcome easy to interpret as a clinical state, but it does not retain all the detail in the underlying CDAI score. The separate score description says higher values indicate worse outcome and that a decrease over time indicates improvement. The registry does not report the distribution of scores or changes in the score itself.

How should repeated assessments be understood?

The endpoint is scheduled repeatedly from Day 1 through Week 252. In general, repeated observations can show the status of the observed group at successive visits. They should not automatically be treated as independent observations: the same participant may contribute data at multiple visits. A method that accounts for within-participant correlation may be appropriate for some longitudinal questions, but the registry record does not report a statistical model or an analysis of change over time.

Separate visit counts also do not reveal individual transitions. A participant could meet the remission threshold at one visit and not at another, but the aggregate counts alone would not identify that pattern. The record gives no participant-level transitions, persistence measure, or time-to-event analysis for remission. Nor does it report whether the scheduled visits were analyzed as separate descriptive summaries or combined in an inferential model.

What can a single-group design estimate?

In general, a single-group extension can describe outcomes among the participants who enter and are followed in that group. It cannot supply a concurrent randomized comparison when no comparator group is part of the endpoint structure. A within-group change is not equivalent to a treatment-versus-control effect: without a comparator, it is not possible to separate treatment-associated changes from other changes over time. Open-label status can also matter because participants and study personnel know the treatment being given, although the registry does not report whether that knowledge affected outcome assessment in this study.

The record labels allocation as randomized but also identifies a single-group model and fewer than two arms for the primary outcome. The record does not explain the relationship between the randomized allocation label and the extension’s single-group outcome structure. It would be inappropriate to infer a randomized comparison from the allocation field alone.

What does the record say about statistical analyses?

The registry lists 14 outcome measures and zero analyses. For the primary remission outcome, it gives the measure, threshold, assessment schedule, and population definition, but no numerical outcome results or inferential analysis. No standard errors, confidence intervals, hypothesis tests, or model estimates are reported for remission. Their absence means precision and statistical compatibility with alternative values cannot be assessed from the registry record.

In general, a confidence interval expresses the uncertainty around an estimated quantity under a specified method and assumptions. It is not a guarantee that the true value lies within the interval, and its interpretation depends on the analysis being used. Since the registry record reports no remission estimate or interval, there is no interval to interpret for this endpoint. A p-value would address compatibility with a specified null hypothesis under a test’s assumptions; no primary-outcome p-value or test is reported.

What are the limits of the safety summaries?

The safety information is descriptive: the record lists participants at risk, participants affected by serious or other adverse events, deaths, and selected serious-event terms with percentages. In general, counts indicate how many participants were affected under the registry’s reporting definitions; they are not event rates unless exposure time and an appropriate denominator are used. No person-time summary or comparative safety analysis is reported. Over long follow-up, differences in observation time can affect the opportunity to record events, but participant-specific follow-up duration is not reported.

Over-dispersion is a general issue for count outcomes when variability exceeds what a simple count model assumes. It matters when selecting or evaluating a statistical model for event counts, but the registry does not report a model for these safety summaries. No model choice, adjustment, or uncertainty interval should be inferred from the counts alone.

10. Multiplicity, Missing Data and Analysis Populations

The primary remission measure is assessed at Day 1 and many scheduled visits through Week 252. In general, repeated testing across visits can raise a multiplicity question: the chance of at least one apparently notable result may increase when many hypotheses are tested. Whether an adjustment is needed depends on the planned inferential aims and how results are interpreted. The registry record does not report hypothesis tests, a multiplicity strategy, or a statistical analysis plan for the repeated remission assessments.

The record defines the analyzed population at each timepoint as participants with data available for analysis at that visit. This specifies how the reported denominator is formed, but does not explain why data could be missing, how missing outcomes were handled, or whether any imputation or sensitivity analysis was used. It also does not report the numbers with available CDAI data at each visit. Consequently, the extent and pattern of missing outcome data cannot be assessed from the registry record.

In general, complete-case summaries describe the participants with observed data, and interpretation may be limited if those participants differ from those without observations. This is a general consideration, not evidence that missingness biased the study’s results. The record does not state a primary analysis population beyond its timepoint-specific availability definition, nor does it describe exclusions or sensitivity analyses for the remission endpoint.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: crohn's disease (9.2%), anal abscess (1.7%), abdominal pain (1.3%), small intestinal obstruction (1.3%), abdominal abscess (0.9%).

For Part 1 (OLE), the registry lists 751 participants at risk, 206 participants affected by serious adverse events, 453 affected by other adverse events, and 3 deaths. For Part 2 (PML SM), it lists 359 participants at risk, 1 participant affected by serious adverse events, 2 affected by other adverse events, and 1 death. These are descriptive counts for different parts of the study, not a comparison between randomized treatment groups.

The listed serious-event terms and percentages are Crohn’s disease, 9.2%; anal abscess, 1.7%; abdominal pain, 1.3%; small intestinal obstruction, 1.3%; abdominal abscess, 0.9%; and anaemia, 0.8%. The registry record does not state the denominator for these percentages or give event-level timing. The term list should not be read as a causal assessment. The record also does not report the causes, timing, or relationship to etrolizumab of the deaths. Differences in counts across the two parts should not be interpreted as differences in risk without comparable follow-up and an appropriate analysis, neither of which is reported.

12. What the Result Does, and Does Not, Mean

The record describes an extension intended to evaluate long-term safety and efficacy of continued etrolizumab in eligible participants previously enrolled in GA29144, alongside safety monitoring for participants who stopped treatment. It supplies a primary remission definition and a schedule of assessments, as well as descriptive safety counts for Part 1 and Part 2. Those details clarify what the study planned to observe and how the reported safety summaries are organized.

The registry record does not report the number of participants in remission at any scheduled visit. It therefore does not show the observed remission frequency, a change in remission over time, or the precision of any such estimate. It also does not report an inferential analysis of the primary outcome. The remission endpoint description should not be mistaken for a finding.

The study’s single-group primary-outcome structure has no concurrent comparator. Even a reported remission pattern would describe the extension participants rather than establish what would have happened under a different treatment or without continued treatment. The safety counts likewise describe recorded events in the two study parts; they do not establish that etrolizumab caused or prevented any event. The registry does not report the causes or treatment relationship of the deaths.

These limitations concern what can be concluded from this registry record, not whether the study may have generated other results. The posted record also does not explain the reason for termination, the handling of missing remission data, or the statistical plan. Conclusions about those topics are therefore not reported in the registry record.

13. Why This Trial Matters Statistically

Extension studies address follow-up beyond an initial trial, and this record illustrates why their design and population definitions matter when reading outcomes. Part 1 follows eligible participants who continue open-label etrolizumab, while Part 2 monitors participants after treatment stops. Those distinct purposes lead to different populations and different interpretations of the safety counts.

The primary outcome is measured repeatedly, with remission defined using a specific CDAI threshold. The scheduled assessments provide a framework for describing disease activity over time, but the record’s timepoint-specific availability rule also makes the denominator important. Without reported counts and denominators at each visit, the outcome schedule cannot be translated into an observed remission pattern.

The study also demonstrates a basic distinction between descriptive follow-up and comparative evidence. A single-group extension can record outcomes among participants who entered it, but it does not provide a concurrent control for attributing changes to treatment. The registry’s randomized allocation field does not override the single-group primary-outcome structure.

Finally, registry completeness shapes statistical interpretation. The safety summaries include affected-participant counts and selected serious-event terms, while the primary outcome has no posted results or analyses. The absence of a reported statistical plan, missing-data approach, and reason for termination leaves important questions unanswered. Recognizing those limits helps keep conclusions proportional to what the record documents.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary-outcome record gives a definition and schedule but no remission counts, analyzed numbers by visit, effect estimates, confidence intervals, or tests. The outcome’s precision and longitudinal pattern therefore cannot be evaluated. The record does not report a statistical analysis plan, a model for repeated measures, or a multiplicity approach.

The timepoint-specific population includes participants with data available at that visit. Reasons for missing data, methods for handling missing values, and sensitivity analyses are not reported in the registry record. The record also does not give participant flow between Part 1 and Part 2 or individual follow-up duration, limiting interpretation of the safety counts over time.

The study is open-label and the primary outcome is structured as a single group, without a concurrent comparator. This restricts causal and comparative interpretation. Although allocation is recorded as randomized, the record does not explain how that designation relates to the single-group model. The reason for the terminated status is not reported in the registry record.

For safety, the record gives participant counts and selected serious-event terms with percentages, but does not specify the percentage denominator, event timing, exposure-adjusted rates, or an analysis of between-part differences. It does not report the causes or treatment relationship of deaths. These omissions prevent stronger interpretation of the safety summaries.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Open-Label Extension and Safety Study for Participants With Crohn's Disease Previously Enrolled in the Etrolizumab Phase III Study GA29144 — NCT02403323.
• Registry record last updated 2024-12-02; status last verified 2024-11.
• Results first posted 2024-12-02.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

NCT02403323 is a Phase III, open-label extension and safety-monitoring study sponsored by Hoffmann-La Roche for participants with Crohn’s disease previously enrolled in GA29144. The registry lists randomized allocation and a single-group model. Part 1 describes etrolizumab treatment once every 4 weeks for eligible participants; Part 2 describes 92 weeks of monitoring after treatment cessation for PML and other safety events.

The primary outcome is the number of Part 1 participants with CDAI remission, defined as a score less than 150, at scheduled assessments through Week 252. The registry record lists 14 outcome measures and zero analyses, but reports no primary remission counts or statistical estimates. It lists safety counts separately for Part 1 and Part 2. The study has terminated status, a start date of 2015-06-08, and a primary completion date of 2023-10-09; the reason for termination is not reported in the registry record.