Abatacept for Treating Adults With Giant Cell Arteritis and Takayasu's Arteritis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Abatacept for Treating Adults With Giant Cell Arteritis and Takayasu's Arteritis (NCT00556439), a randomized phase 2 study. The primary outcome is Primary Outcome - Relapse-free Survival (RFS).

2. Clinical Question

3. Trial Design

This was a randomized, double-masked, parallel-group phase 2 trial with concurrent studies in GCA and TAK. The registry identifies the study as a randomized withdrawal design. Its stated enrollment was 97, and the record lists the study status as completed. The recorded start was 2008-12, and primary completion was 2015-08. The sponsor was the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS).

The treatment sequence matters for interpreting the comparator. The registry says all participants received abatacept and prednisone for the first 3 months. It then identifies randomized groups receiving abatacept or placebo within each disease stratum. Thus, the placebo group was not described as receiving placebo throughout the entire study. The available arm descriptions end partway through the prednisone taper instructions: the dose began at 40 to 60mg and was tapered to 20mg by Month 3, but the remainder of the regimen is not reported in the registry record. The timing and details of later treatment and follow-up are also not fully described in the available record.

The parallel comparisons were Group A, abatacept in GCA; Group B, placebo in GCA; Group C, abatacept in TAK; and Group D, placebo in TAK. The primary outcome description says that relapse-free survival curves were constructed separately for the GCA and TAK strata, with treatment arms compared using the logrank test. This stratum-specific approach recognizes that the two diseases are not being treated as one undifferentiated population.

Although the study is marked completed, completion status does not itself provide the outcome findings. The registry record describes the endpoint and proposed analysis but does not report the primary relapse-free survival results. Further protocol details, including a complete account of allocation procedures and the post-initial-period treatment schedule, are not reported in the registry record.

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.

Study Entry to Week 12

Reasons for not completing this period:

After Week 12 (Randomized Study Period)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The primary study endpoint was relapse-free survival (RFS). Kaplan-Meier curves of RFS were constructed for each stratum (giant cell arteritis and Takayasu arteritis), and differences in treatment arms compared using the logrank test. The analysis of the primary outcome was based upon intent to treat.

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: Reflects the population that reached Week 12 blinded randomization. In the giant cell arteritis cohort, 8 of 49 were withdrawn at or before Week 12 with a total of 41 remaining in the active comparator group. In the Takayasu arteritis cohort, 8 of 34 were withdrawn at or before Week 12 with a total of 26 remaining in the active comparator group.

6. Primary Endpoint

The registry lists the primary outcome as Primary Outcome - Relapse-free Survival (RFS), measured over Weeks 0 to 64.

The primary endpoint was relapse-free survival (RFS) over Weeks 0 to 64. It represents the time participants remained free of a disease relapse after remission. In the registry definition, relapse means active disease occurring after a period of remission, while remission means absence of active disease. Active disease could be identified through clinical features, imaging, or both.

The clinical criteria include vascular pain or tenderness, certain persistent headaches, ischemic eye findings or visual loss, jaw or tongue pain or claudication, transient ischemic attack or stroke, extremity claudication, and other symptoms or signs attributed to GCA or TAK. Some symptoms are considered alongside elevated erythrocyte sedimentation rate or C-reactive protein. Imaging evidence includes new vascular stenosis or aneurysm in new vascular territories on MRI/MRA or arteriogram. The full criteria are intended to make the relapse event more explicit than a broad judgment of worsening alone.

In general, a time-to-event endpoint can use information about both whether an event occurred and when it occurred. This is useful when participants may remain event-free for different lengths of time or when follow-up ends before an event is observed. The registry describes Kaplan-Meier curves for RFS, which estimate the event-free experience over time, and a logrank test to compare treatment groups. The registry record does not report the numbers of relapses, the estimated RFS values, or the treatment-comparison result, so the endpoint definition alone cannot establish the size or direction of a treatment difference.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a time-to-event endpoint. Each participant’s outcome is based on the time from the study’s defined starting point until relapse, while participants without a recorded relapse contribute information up to their last observed follow-up or another censoring point. The registry specifies the outcome timeframe as Weeks 0 to 64, but does not fully describe censoring rules or the handling of participants whose follow-up ended early.

What does Kaplan-Meier estimation describe?

In general, the Kaplan-Meier method estimates the proportion of participants who remain event-free over time while accounting for different lengths of follow-up. Participants who have not experienced relapse by the end of their observed time are not treated as if they relapsed; their information is censored at that point. The method depends on assumptions about censoring, including that censoring is not related to the chance of a future event in a way that would distort the estimate. The registry names Kaplan-Meier curves but does not report the curves, estimates at particular times, numbers at risk over follow-up, or censoring details.

What does the logrank test compare?

In general, the logrank test compares event experience across groups over the observed follow-up period. It evaluates whether the groups’ survival patterns differ, rather than comparing only the number of participants with an event at a single time point. The registry says the treatment arms were compared with a logrank test separately within the GCA and TAK strata. The test result, p-value, and any associated effect estimate are not reported in the registry record. A test alone would not communicate the magnitude or precision of a treatment difference; estimates and confidence intervals would also be needed for that interpretation.

Why keep the disease strata distinct?

The trial concerns GCA and TAK, and the primary outcome description explicitly separates these conditions for analysis. Stratified presentation allows each treatment comparison to reflect the relevant disease group rather than treating the two conditions as interchangeable. In general, stratification can also be important when baseline risk or disease course differs between groups. The registry does not report an interaction test or a formal method for combining the stratum-specific results, so no claim about a common treatment effect across the diseases can be made.

What is the role of randomization and masking?

The study is recorded as randomized and double-masked, with a parallel-group model. In general, randomization aims to make treatment groups comparable at allocation, including with respect to factors that may not be measured. Masking can reduce the influence of treatment knowledge on participant reporting, clinical assessment, and study conduct. These design features help protect a comparison from bias, but they do not guarantee balanced groups in a trial or prevent missing data and other limitations. The registry record does not provide details of sequence generation, concealment, or who was masked.

The withdrawal design also affects the meaning of the comparison. The registry states that all participants received abatacept and prednisone in the first 3 months, followed by randomized abatacept and placebo groups. In general, a randomized withdrawal comparison estimates the contrast between continuing active study treatment and withdrawing it under the protocol’s conditions. It is not the same question as starting abatacept versus never receiving it. However, the available arm descriptions are incomplete, and details of the later treatment regimen are not reported in the registry record.

How should the serious-event counts be read?

The adverse-event information gives participants at risk and participants affected within each named group. These counts are descriptive. In general, comparisons of adverse-event counts need context about exposure time, event definitions, repeat events, and uncertainty. A participant count does not indicate how many episodes occurred, how severe each event was beyond its classification, or when the event occurred. The record does not provide those details or a formal inferential comparison of serious adverse events.

The registry includes a sample-size-assumptions statement describing planning illustrations in general terms: two-sided significance level 0.05, equal allocation between two arms, a normal approximation, and effects expressed as a fraction of an observed effect. Those are described as planning illustrations rather than the trial’s own design justification. No calculated planning estimates are included in the registry entries, and they should not be interpreted as observed power or as evidence about the trial’s results. The trial’s own sample-size justification is not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and two analyses, but those counts do not establish how multiplicity was handled. The primary outcome description specifies separate RFS curves and treatment-arm comparisons within the GCA and TAK strata. These are distinct comparisons, and the available record does not report whether any adjustment for multiple testing was planned or used. It also does not identify a hierarchy of hypotheses, a single combined primary test, or how conclusions across the two strata were intended to be drawn.

The endpoint is assessed over Weeks 0 to 64. The record does not report whether analyses at additional time points were prespecified, whether there were secondary endpoints, or whether any subgroup analyses were planned. It also does not report interim analyses or stopping rules. Without that information, the scope of the testing and the potential for selective emphasis cannot be fully assessed.

Missing-data handling is not reported in the registry record. For a time-to-event analysis, the timing and reasons for loss to follow-up, the censoring rules, and the amount of follow-up available can affect interpretation. The registry describes the primary analysis as based on intent-to-treat, but the final wording is truncated and does not explain how participants with incomplete data were retained or treated in the analysis. The number of participants included in each primary comparison is not reported.

The record also does not provide a complete statistical analysis plan, a sample-size justification for the trial, or details of any protocol changes relevant to the outcome analysis. These omissions limit assessment of whether the reported endpoint and analysis were prespecified and how uncertainty was managed. In general, multiplicity control, missing-data procedures, and analysis-population definitions are essential parts of interpreting a randomized trial; their absence from the registry description leaves those aspects unclear.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pain - other (18.2%), infection - other (18.2%), hemorrhage - gi (13.3%), infection with normal anc or grade 1 or 2 neutrophils (13.3%), ocular/visual - other (10.0%).

The serious adverse event data are participant counts within each group. Group A, abatacept in GCA, lists 20 participants at risk and 8 affected. Group B, placebo in GCA, lists 21 at risk and 5 affected. Group C, abatacept in TAK, lists 11 at risk and 5 affected. Group D, placebo in TAK, lists 15 at risk and 7 affected. The registry lists 0 deaths for each group.

These counts should not be read as adjusted comparisons or as evidence of a causal difference. The record does not report confidence intervals or statistical tests for serious adverse events, and it does not describe event duration, timing, recurrence, or participant exposure time. The listed participants at risk provide group denominators for the serious-event counts, but the record does not give a fuller safety-analysis method.

The registry also lists serious event terms with percentages: Pain - other, 18.2%; Infection - other, 18.2%; Hemorrhage - GI, 13.3%; Infection with normal ANC or Grade 1 or 2 neutrophils, 13.3%; Ocular/visual - other, 10.0%; and Renal/genitourinary - other, 10.0%. The denominator for these percentages and whether a participant could contribute to more than one term are not reported in the registry record. They should therefore be treated as listed event summaries, not directly compared across treatment groups.

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

The record establishes that this was a randomized, double-masked phase 2 trial of abatacept in GCA and TAK, using a randomized withdrawal design and RFS as the primary endpoint. It also identifies how the investigators intended to compare relapse-free survival within each disease stratum. This information explains the clinical question and planned analytical framework, but it is not a report of the primary treatment effect.

The primary RFS results are not reported in the registry record. No estimate of time without relapse, comparison statistic, p-value, or confidence interval is available in the information. Consequently, neither the direction nor the magnitude of an efficacy difference can be determined. The endpoint’s definition and the existence of a logrank analysis do not show whether the groups differed, how precisely any difference was estimated, or whether it would be clinically important.

The serious adverse event counts and listed terms provide descriptive safety information, including 0 deaths recorded in each group. They do not by themselves establish that the groups had the same or different safety profiles. A robust interpretation would also require information about event timing, duration, exposure, uncertainty, and the methods used for comparison. Those details are not reported in the registry record.

In general, a randomized trial can reduce confounding through allocation, while masking can limit some forms of assessment or reporting bias. Those design features strengthen the basis for a treatment comparison when results are available, but they do not replace complete outcome reporting. The record’s completion status should not be mistaken for evidence of benefit or lack of benefit. The appropriate statistical reading is that the design and endpoint are described, while the primary comparison and its precision remain unreported.

13. Why This Trial Matters Statistically

This trial is statistically notable because its clinical question spans two named forms of large-vessel vasculitis, while the primary analysis treats them as separate strata. That distinction matters: an overall average could obscure different patterns between GCA and TAK, whereas stratum-specific comparisons retain the disease-specific framing. At the same time, separate analyses raise questions about how evidence across strata was to be interpreted and whether multiplicity was addressed. Those details are not reported in the registry record.

The randomized withdrawal design also defines a particular estimand. Participants initially received abatacept and prednisone, followed by randomized abatacept or placebo groups. In general, this design evaluates continued treatment against withdrawal after an initial treatment period; it does not answer the same question as an initial-treatment comparison from the outset. Clear reporting of the transition, follow-up, and analysis population is therefore important. The registry’s arm descriptions are incomplete, limiting assessment of the full treatment sequence.

RFS is a time-to-event endpoint, and the registry names Kaplan-Meier estimation and logrank testing. These methods can use event timing and accommodate different observed follow-up durations, subject to assumptions and transparent censoring rules. The available description identifies those methods but omits the estimates, uncertainty, and event data needed to judge the treatment contrast.

Finally, the record illustrates why trial status and design information should be distinguished from results. A completed randomized trial can still have an incomplete public outcome record. Without the RFS estimates and their precision, the central statistical question remains unanswered in the registry information. This limitation is particularly consequential when the record also lacks details on missing data, multiplicity, and the trial’s own sample-size rationale.

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 limitation is that the registry record does not report the primary RFS findings. It gives the endpoint definition and analysis approach but not relapse counts, Kaplan-Meier estimates, comparison statistics, p-values, or confidence intervals. The treatment effect and its precision therefore cannot be assessed.

The available protocol descriptions are incomplete. They state that all participants received abatacept and prednisone for the first 3 months and identify randomized abatacept and placebo groups afterward, but the arm text is truncated. The prednisone taper is described through Month 3, with the remaining instructions absent. Details of subsequent treatment, follow-up, and the full allocation procedure are not reported in the registry record.

The record also does not explain censoring or missing-data procedures, the number analyzed for the primary endpoint, or handling of participants who did not complete follow-up. It identifies two disease strata and two analyses but does not report multiplicity control or an analysis hierarchy. Interim analyses, stopping rules, and a complete statistical analysis plan are not reported in the registry record.

Safety reporting is limited to group counts, death counts, and a list of serious event terms with percentages. The record does not specify the denominator for those term percentages, whether terms can overlap, event timing, exposure-adjusted summaries, or formal safety comparisons. Finally, the registry describes planning assumptions in general terms but does not provide the trial’s sample-size justification. These gaps limit a full assessment of design adequacy, analysis conduct, and the precision of both efficacy and safety findings.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Abatacept for Treating Adults With Giant Cell Arteritis and Takayasu's Arteritis — NCT00556439.
• Registry record last updated 2018-02-26; status last verified 2018-01.
• Results first posted 2018-02-26.
• 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

NCT00556439 is a completed phase 2 randomized, double-masked, parallel-group study sponsored by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS). It enrolled 97 adults in concurrent GCA and TAK studies. The registry describes a randomized withdrawal design, with abatacept and prednisone during the first 3 months, followed by named abatacept and placebo groups within each disease stratum.

The primary endpoint was relapse-free survival over Weeks 0 to 64. The registry describes Kaplan-Meier curves and logrank comparisons within the GCA and TAK strata and says the primary analysis was based on intent-to-treat, though that sentence is truncated. Primary outcome estimates and comparisons are not reported in the registry record. Serious adverse event counts and 0 deaths per group are listed, but detailed comparative safety analyses are not included.