A Study Comparing the Effect of Albiglutide With Exenatide on Regional Brain Activity Related to Nausea in Healthy Subjects: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study Comparing the Effect of Albiglutide With Exenatide on Regional Brain Activity Related to Nausea in Healthy Subjects (NCT02802514), a randomized phase 4 study. The primary outcome is Blood Oxygen Level Dependent (BOLD) Signal by Functional Magnetic Resonance Imaging (fMRI) Visual Nauseogenic Task.

2. Clinical Question

3. Trial Design

The registry record describes NCT02802514 as a randomized, single-blind, two-part, two-period crossover study. It was classified as Phase 4 and was terminated. The record lists enrollment as 5. Its study description anticipated 24 to 28 participants across Part A and Part B, but the reason for the difference between that expectation and the listed enrollment is not reported in the registry record.

Both parts were described as having screening, dosing and assessment, and follow-up stages. Part A was intended to inform progression, methods and the analysis plan for Part B. The primary-endpoint description says that the intended primary endpoints involved combining data from Part A and Part B. Whether that combined analysis was completed is not reported in the registry record.

The crossover schedule included albiglutide, exenatide and an off-therapy natural-history scan. Treatment sequence varied, and the descriptions include matching placebo dosing. The record describes a 6-9 week washout period between dosing scans. This schedule is relevant to interpretation: the comparison is not simply a parallel-group contrast, and participants could contribute observations under different scan conditions or treatment sequences.

Masking was described as single-blind, with both the subject and the imaging evaluators analyzing MRI data blinded. The stated study dates are 2016-09-20 for the start and 2017-08-10 for primary completion. The trial status is terminated. The reason for termination, the conduct of any interim analyses, and whether the study proceeded as planned from Part A to Part B 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.

Before assignment: A total of 10 participants were screened, of which 6 were screen-failures and 4 were randomized. Participants from United States site were part of the study. This was a 2-part study; however, study was terminated early due to portfolio strategic reason and only Part A was completed.

Session 1 (up to Day 8)

Wash Out Period (up to Week 9)

Session 2 (Up to Day 8)

Analysis population for the primary outcome, as defined in the registry: All randomized Population (included participants who were randomized to receive study treatment, regardless of whether they took randomized study treatment). Imaging data were analyzed in Part A for only one participant who experienced nausea out of four study participants, and that, as a consequence, data cannot be summarized for Part A.

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: Safety Population: Comprised of all participants who received at least one dose of the study treatment.

6. Primary Endpoint

The registry lists the primary outcome as Blood Oxygen Level Dependent (BOLD) Signal by Functional Magnetic Resonance Imaging (fMRI) Visual Nauseogenic Task, measured over Up to Week 11.

The primary endpoint was the blood oxygen level dependent (BOLD) signal measured using functional magnetic resonance imaging during a visual nauseogenic task. The stated timeframe was up to Week 11. The endpoint description focuses on seed-to-voxel analyses, using prespecified brain regions of interest associated with sensory or interoceptive, emotional or affective, cognitive or evaluative, and visual processing. Named regions include the insula, dorsal anterior cingulate cortex, amygdala, pregenual anterior cingulate cortex, dorsolateral prefrontal cortex, occipitofrontal circumference, primary visual cortex and extrastriate cortices.

In general, BOLD fMRI is used to characterize changes in brain activity indirectly through changes in the blood-oxygenation signal. In this study, the visual task was intended to probe brain responses relevant to nausea. Such a signal is not itself a direct measure of how much nausea a person feels, nor does an observed brain response alone establish why it occurred. The endpoint is exploratory and depends on the imaging task, selected regions, preprocessing and analysis choices. The registry record does not report a numerical endpoint estimate, treatment contrast, confidence interval or primary-endpoint p-value.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a crossover design compare?

In general, a crossover design assigns participants to sequences so that observations can be collected under more than one condition. This can make within-participant contrasts possible, which may reduce the influence of stable differences between people. The design described in this record included albiglutide, exenatide and an off-therapy scan, with sequence variations and matching placebos. It was organized into two periods and included a stated 6-9 week washout between dosing scans.

A crossover analysis must account for the fact that measurements are related within a participant. It may also need to address period effects, sequence effects and possible carryover from an earlier treatment period. Those considerations are general methodological issues, not evidence that any particular model or correction was used in this study. The statistical model, its covariates and its handling of period, sequence or carryover are not reported in the registry record.

What does a BOLD signal estimate represent?

In general, BOLD fMRI detects changes in the signal associated with blood oxygenation while a participant performs a task. A task-based contrast can be used to characterize brain responses associated with the task, but it is not a direct assay of nausea itself. The study focused on a visual nauseogenic task and seed-to-voxel analyses from specified brain regions. The registry record does not state a single summary measure or a numerical definition of the treatment contrast, and it does not provide a treatment-effect estimate.

Because the primary endpoint involves imaging across multiple brain regions and potentially multiple task contrasts or time points, interpretation depends on which tests were prespecified and how the analysis handled multiple comparisons. In general, testing many regions or contrasts increases the chance of at least one apparently unusual result if no multiplicity strategy is applied. The statistical methods and multiplicity procedure are not reported in the registry record. No primary-endpoint analysis is listed.

Why do estimates need uncertainty intervals?

In general, a confidence interval describes the range of effect values compatible with the data and the statistical model at a stated confidence level. It conveys precision as well as the direction and scale of an estimate. A p-value, by contrast, summarizes how unusual data would be under a specified null model; it does not measure the probability that a treatment has no effect, nor does it describe effect size. Neither an estimate, a confidence interval nor a primary-endpoint p-value is reported in the registry record, so those aspects of the imaging comparison cannot be assessed.

The small amount of imaging information explicitly described for Part A makes precision a central concern. In general, sparse data can produce highly uncertain estimates and can make model assumptions influential. A failure to report an analysis is not equivalent to a result showing no difference. It also does not establish that treatments have identical effects. The registry record does not report the results needed to distinguish among these possibilities.

What is over-dispersion, and does it apply?

In general, over-dispersion occurs when count data vary more than expected under a statistical model such as a Poisson model. It can lead to confidence intervals that are too narrow if the extra variability is ignored. That concept is most relevant to count outcomes, not automatically to a continuous fMRI signal. The record does not describe a count-based primary endpoint or report a model that would permit an assessment of over-dispersion. No claim about over-dispersion in this trial is warranted.

How should planning calculations be read?

In general, a planning calculation based on an observed effect uses that estimate as an assumed target for a future design. A calculation for detecting a stated fraction of an effect asks how a hypothetical study might be sized under assumptions such as a significance level, allocation pattern and approximation method. It is not a measure of the power actually achieved in the completed study. The planning assumptions named in the registry record include a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation. The stated calculations are illustrations, not the trial's own design justification. The registry record does not report a sample-size justification for the original study, and no outcome-based sample-size estimate should be interpreted as observed power.

The record's primary outcome is not reduced to one comparison, and no primary-outcome analyses are listed. Consequently, a single treatment contrast cannot be reconstructed from the record. There are no posted primary effect estimates to recalculate from summary statistics. In general, a recalculation from summary data would be an unadjusted estimate and would not replace the sponsor's analysis where that analysis is reported. Here, the sponsor's primary analysis is not reported in the registry record, so no sponsor estimate can be compared with an independent calculation.

10. Multiplicity, Missing Data and Analysis Populations

The primary endpoint is described through a seed-to-voxel approach involving multiple brain regions relevant to nausea-related processing, as well as visual regions. That creates a potential multiplicity issue in general: if many regions, contrasts or time points are examined, the chance of finding a nominally unusual result can increase. The registry record does not report which tests were treated as confirmatory, how many contrasts were analyzed, or whether a family-wise error, false-discovery or other multiplicity procedure was used. No p-values or primary-endpoint analyses are listed.

The record says Part A was intended to inform progression, methods and the analysis plan for Part B, while the primary endpoint description states that Part A and Part B data were to be combined. Whether this process led to a finalized analysis plan or a combined analysis is not reported in the registry record. The reason for termination is also not reported, so it is not possible to determine how that status affected planned analyses.

The primary-outcome population is defined as all randomized participants, regardless of whether they took randomized treatment. The record also states that Part A imaging data were analyzed for only one participant who experienced nausea, out of four study participants, but the accompanying explanation is incomplete. Missing-data methods, the handling of participants without usable imaging, and the number of observations included in any completed analysis are not reported in the registry record. The listed all-randomized population definition alone does not show that every randomized participant contributed evaluable imaging data.

In general, missing outcome data can affect both precision and the interpretation of a treatment comparison, especially when the reasons for missingness differ across treatment or sequence. No assumptions or sensitivity analyses for missing data are reported. The record also does not report how the analysis treated any incomplete crossover sequences.

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 2% reporting threshold.

Most frequent non-serious adverse event terms

The registry record lists serious adverse-event information separately for Albiglutide 50 mg, Exenatide 10 µg, Albiglutide Matching Placebo and Exenatide Matching Placebo. Each listed group has 4 participants at risk for serious adverse events, zero serious adverse events and zero deaths. These are the recorded group-level figures; they should not be added together to infer a total number of distinct participants because the study used a crossover design and group membership can reflect treatment exposure.

For other, non-serious adverse events, the record lists 1 affected participant in the Albiglutide 50 mg group and zero in each of the other listed groups. The record does not provide serious adverse-event terms, and the serious-event term list is empty. These entries describe what was recorded, rather than establishing that no unrecorded events occurred or that the groups have equivalent safety. The listed enrollment is 5 and the trial was terminated, so the safety counts are limited and do not support a precise comparative estimate. The registry record does not report a statistical comparison of adverse events.

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

The study addressed a specific mechanistic question: whether albiglutide and exenatide might be associated with different brain activation during a visual task intended to evoke nausea-related processing. It used MRI and included an off-therapy scan, and its crossover structure was designed to observe participants under more than one condition. The work is therefore best understood as exploratory imaging research, not as a clinical-outcome comparison.

The registry record does not report the primary imaging results. It gives no treatment effect estimate, uncertainty interval or p-value, and lists no primary-outcome analyses. The limited Part A imaging statement refers to analysis for only one participant who experienced nausea, out of four study participants. The continuation of that description is incomplete. These registry entries mean the posted record cannot establish whether BOLD responses differed between treatments, how large a difference might have been, or how precisely it could be estimated.

Nor does the record establish that the treatments have the same brain effects. An unreported or unavailable comparison is not evidence of equivalence. Likewise, a sparse descriptive safety record with no serious events listed cannot establish that serious events are impossible or that treatments have equal safety. The results should be read as incomplete evidence from a terminated study with very limited reported imaging information.

The study also does not directly answer whether either treatment changes nausea in patients. The study description concerns healthy volunteers susceptible to motion sickness, and its primary endpoint is an imaging signal during a task. Generalization from a task-related brain signal to symptoms, clinical populations or longer-term treatment effects would require evidence beyond the endpoint described in the registry record. The record lists Diabetes Mellitus as a condition, but does not explain the relationship of that listing to the healthy-volunteer study description.

13. Why This Trial Matters Statistically

This trial is statistically notable because it combines a crossover structure with an exploratory imaging endpoint and an intended two-part process. Such a design can be useful for studying within-participant responses, but it makes the analysis dependent on clear handling of treatment sequence, period and repeated measurements. The registry record describes the design but does not report the statistical model used to address those features.

The endpoint also raises a common issue in imaging research: results may be examined across multiple regions or contrasts. In general, the distinction between prespecified tests and exploratory scans of many possible signals matters for interpreting false-positive risk. Here, the registry record describes several brain regions but does not report a multiplicity strategy or primary analysis results. That makes it impossible to determine whether any potential signal was robust to the analysis choices.

The contrast between the anticipated enrollment of 24 to 28 participants and the listed enrollment of 5, alongside terminated status, emphasizes why enrollment and completion context matter when reading an estimate. The record does not give a primary treatment estimate, and it notes only limited Part A imaging analysis. Statistical significance or precision cannot be meaningfully assessed without the results and their uncertainty.

Finally, the record distinguishes an intended analysis from an available analysis: Part A was meant to inform Part B, and combined primary endpoints were described, but completion of that process is not reported. This illustrates why an endpoint definition and planned design are not substitutes for posted outcome data. The record provides context for the research question, but not a quantitative answer to it.

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 largest limitation for interpretation is the absence of reported primary-endpoint analyses. Without BOLD signal estimates, treatment contrasts or uncertainty measures, neither effect direction nor precision can be evaluated. The record says imaging data were analyzed in Part A for only one participant who experienced nausea, out of four study participants. The description stops mid-sentence, so its further context is not reported in the registry record.

Enrollment is listed as 5, although the study description anticipated 24 to 28 participants across the two parts. The trial is marked terminated, but the reason is not reported. The record does not clarify whether Part B was conducted, whether the planned combined analysis occurred, or how the limited enrollment affected the imaging analysis. These omissions prevent a full assessment of study completion and the available analytic population.

The endpoint description spans multiple regions and an imaging task, but the analysis model, multiplicity control, missing-data procedures and treatment of incomplete crossover sequences are not reported. Although the record defines the primary population as all randomized participants, it does not establish that all randomized participants had usable scans or contributed to an analysis. The reason for the condition listing of Diabetes Mellitus alongside a description of healthy volunteers is also not reported.

The safety information is sparse and descriptive, with no comparative analysis reported. The crossover design means group-level exposure counts should not be treated as independent parallel-group samples. Finally, the study's task-based imaging endpoint does not directly establish effects on clinical nausea or outcomes in patients. These limitations constrain interpretation of both efficacy-related mechanisms and safety.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: A Study Comparing the Effect of Albiglutide With Exenatide on Regional Brain Activity Related to Nausea in Healthy Subjects — NCT02802514.
• Registry record last updated 2020-10-30; status last verified 2020-10.
• Results first posted 2020-08-24.
• 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

NCT02802514 was a randomized, single-blind, two-part, two-period crossover study comparing single doses of albiglutide and exenatide, with matching placebos and off-therapy MRI scans. The primary endpoint was BOLD signal during a visual nauseogenic fMRI task, assessed up to Week 11. The study was described as exploratory and intended to examine brain activity related to nausea in healthy volunteers susceptible to motion sickness.

The registry record lists Phase 4, enrollment of 5 and terminated status. It reports no primary-outcome analyses. Its primary-endpoint description notes limited Part A imaging analysis, and the continuation of that statement is incomplete. Serious adverse events and deaths are listed as zero in each treatment group, while one non-serious affected participant is listed in the albiglutide 50 mg group. The reason for termination, the completion of Part B and the primary treatment comparison are not reported in the registry record.