Safety and Efficacy of Exenatide as Monotherapy and Adjunctive Therapy to Oral Antidiabetic Agents in Adolescents With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of Exenatide as Monotherapy and Adjunctive Therapy to Oral Antidiabetic Agents in Adolescents With Type 2 Diabetes (NCT00658021), a randomized phase 3 study. The primary outcome is Adjusted Change From Baseline in Glycated Hemoglobin A1c (HbA1c) at Week 28.

2. Clinical Question

3. Trial Design

This was a Phase 3, randomized, double-masked, parallel-group study recorded as completed. The registry lists enrollment of 122. The intervention arms were exenatide 5 µg, exenatide 10 µg, and placebo; each was administered by subcutaneous injection twice a day. The primary outcome comparison, however, combines the exenatide dose arms as Total Exenatide Twice Daily and compares that group with placebo. The registry does not give separate primary estimates for each exenatide dose.

The trial addressed adolescents with type 2 diabetes, including patients naïve to antidiabetes agents and patients receiving metformin, a sulfonylurea, or both. The primary objective was to compare glycemic control after 28 weeks, using change in HbA1c from baseline to endpoint. Baseline was defined as the last non-missing scheduled or unscheduled assessment on or before the first dose of study medication. The primary outcome timeframe is recorded as baseline (Day 1) and Week 28.

For the primary efficacy analysis, the registry defines an Evaluable Analysis Set. It includes randomized participants who received at least one dose of randomized study medication and had a baseline plus at least one post-baseline HbA1c assessment. The registry’s population description ends mid-sentence after noting that an efficacy-analysis decision was made with agreement by the European Medicines Agency and the Food and Drug Administration. Further detail about that decision is not reported in the registry record.

The registry lists seven secondary outcomes, including HbA1c goal attainment, body weight, fasting serum glucose, self-monitored blood glucose, fasting serum insulin, HOMA-B and HOMA-S, and discontinuation due to failure to maintain glycemic control. Those outcomes address measures beyond the primary HbA1c change comparison. The registry lists nine outcome measures and 16 analyses, but a complete statistical-analysis plan is not reported in the registry record. In particular, the record does not state the rationale for the enrollment target or describe any interim analyses.

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: The study commenced with a 1-week, single-blind, injectable placebo lead-in period before participants were randomized to 1 of 3 treatment groups: exenatide 5 microgram (mcg) twice daily, exenatide 10 mcg twice daily, or placebo twice daily. A total of 122 participants were randomized in this study.

28-Week Treatment Period

Reasons for not completing this period:

Long-term Safety Follow-up Period

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Evaluable Analysis Set included all randomized participants who received at least 1 dose of randomized study medication and had a baseline and at least 1 post-baseline HbA1c assessment. For analysis of efficacy, a decision was made with agreement by the European Medicines Agency and the Food and Drug Administration to pool participants from both exenatide groups (Total EBID) and compare this pooled group with placebo.

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: The Intent-To-Treat analysis set included all randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Adjusted Change From Baseline in Glycated Hemoglobin A1c (HbA1c) at Week 28, measured over Baseline (Day 1) and Week 28.

The primary endpoint is adjusted change from baseline in glycated hemoglobin A1c at Week 28, expressed in %HbA1c. HbA1c is a continuous measure used to assess glycemic control over time. An endpoint based on change compares a participant’s follow-up measurement with that participant’s baseline measurement, while the reported adjusted least-squares means summarize the treatment groups using the specified statistical model.

Change from baseline can be useful because participants may begin the study with different HbA1c values. Comparing changes focuses the analysis on how the measure shifts during follow-up rather than treating the Week 28 value as if all participants started from the same point. The registry defines baseline as the last non-missing assessment on or before the first study-medication dose. The stated primary objective was to test whether the change with exenatide was superior to the change with placebo after 28 weeks.

For this outcome, the registry reports adjusted least-squares mean changes of 0.11 for Total Exenatide Twice Daily and 0.38 for placebo. The reported standard errors are 0.215 and 0.293, respectively. The model-based contrast is reported separately, with its confidence interval and p-value. The contrast is the relevant inferential result for the stated group comparison; the group estimates alone do not convey the uncertainty in that contrast.

HbA1c is one aspect of glycemic control, not a complete account of health or treatment experience. The secondary outcomes extend the registry’s outcome list to other glycemic measures, body weight, and discontinuation related to glycemic control. The endpoint results therefore answer a defined statistical question about HbA1c change, not every question that might matter to patients or clinicians.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry reports the primary outcome as adjusted change from baseline in HbA1c, analyzed with a mixed model with repeated measures (MMRM). It reports least-squares means for the treatment groups and a least-squares mean difference for the primary comparison. These are model-based estimates rather than simple unadjusted summaries. The comparison is Total Exenatide Twice Daily versus placebo, and the endpoint is assessed at Week 28.

What is an adjusted least-squares mean?

In general, a least-squares mean is an estimated group mean derived from a statistical model. The model can account for specified factors and repeated observations, so an adjusted mean need not equal the ordinary arithmetic average of the observed values. The adjustment aims to make the group estimates reflect the model’s stated structure. The registry identifies the estimate as adjusted, but the full list of model terms and the complete model specification are not reported in the registry record.

The primary arm estimates are 0.11 for Total Exenatide Twice Daily and 0.38 for placebo, with standard errors of 0.215 and 0.293. A standard error describes the estimated sampling variability of an estimate: smaller standard errors generally correspond to greater precision, all else being equal. It is not a measure of how much individual participants’ outcomes vary. The record reports standard errors for the arm estimates but does not provide confidence limits for each arm estimate.

Why use a mixed model with repeated measures?

In general, repeated-measures models are designed for outcomes measured over time. They can use the pattern of available measurements and account for the fact that measurements from the same participant are related. This can be preferable to treating every observation as independent. The registry says that an MMRM analysis was performed and specifies that measurements after initiation of rescue medication and study-drug discontinuation were excluded. It does not report the covariance structure, the fixed effects, the handling of all other missing observations, or the model’s assumptions in sufficient detail to reproduce the analysis fully.

What does the reported contrast represent?

The reported least-squares mean difference is -0.28 %HbA1c. It is the model’s estimate of the difference between Total Exenatide Twice Daily and placebo for the primary endpoint, using the reported mixed-model analysis. Its 95% confidence interval runs from -1.01 to 0.45, and the associated p-value is 0.444. The interval reflects uncertainty in the contrast, rather than uncertainty in either group estimate considered alone. In general, a confidence interval is built using an estimate and its sampling uncertainty under the analysis method. An interval that contains zero is compatible with no difference as well as with differences in either direction, under the model.

The p-value assesses the compatibility of the observed result with a null hypothesis under the specified analysis assumptions. It is not an effect-size measure, a probability that a treatment works, or a substitute for the estimate and its interval. The registry reports the value but does not report the complete inferential framework, including a stated decision threshold for this comparison. The registry also lists multiple outcomes and analyses, so interpretation of an isolated p-value should take account of the broader analysis context.

What does the analysis population mean?

The Evaluable Analysis Set includes randomized participants who received at least one dose of randomized study medication and had both a baseline and at least one post-baseline HbA1c assessment. This definition restricts the primary analysis to participants meeting those recorded conditions. It differs conceptually from an analysis of everyone randomized, because eligibility for this set depends on receiving medication and having outcome measurements. The registry does not report a full comparison of results across alternative analysis populations.

How should missing or post-rescue measurements be considered?

The description says measurements after starting rescue medication and after study-drug discontinuation were excluded from the MMRM analysis. This defines what data did not contribute after those events, but it does not by itself explain how missing outcomes were handled in every other circumstance. In general, missing-data assumptions matter because observed outcomes may differ systematically from unobserved outcomes. Different assumptions can lead to different estimates, so a reader needs the model specification and sensitivity analyses to assess robustness. Those details are not reported in the registry record.

For general context, over-dispersion occurs when observed variability exceeds what a statistical model’s assumed distribution allows; accounting for it can prevent uncertainty from being understated. The primary endpoint is continuous and is reported using least-squares means and standard errors. The registry does not identify over-dispersion as an issue for this analysis, so it should not be inferred to have affected the result.

The registry also describes a planning illustration based on the observed effect, two-sided significance level 0.05, equal allocation between two arms, and a normal approximation. Such an illustration is a prospective planning calculation, not evidence about the trial’s observed power. It does not reveal the trial’s original design rationale, which is not reported in the registry record. Planning calculations should not be read as changing the uncertainty in the observed estimate or the interpretation of its confidence interval.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Placebo")
n   <- c(42)
m   <- c(0.38)
sp  <- c(0.293)
sd  <- sp * sqrt(n)      # standard errors converted to SDs
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Placebo  42  0.380  1.899  -0.212  0.972

11. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary analysis as a Mixed Models Analysis. Its estimated least-squares mean difference is -0.28 %HbA1c, with a 95% confidence interval from -1.01 to 0.45 and p = 0.444. This adjusted model-based contrast is the principal reported inferential result for Total Exenatide Twice Daily versus placebo.

Recalculated summaries can serve as a check on the posted numbers, but an unadjusted calculation from group summaries would not reproduce the MMRM contrast. The model uses repeated measurements and the registry’s stated exclusions after rescue medication and study-drug discontinuation; the two arm estimates alone do not contain all information needed to reconstruct that analysis. The record does not flag a discrepancy between the recalculated summaries and the sponsor’s result. Accordingly, the sponsor’s reported model analysis remains the relevant result to interpret, rather than replacing it with a different unadjusted calculation.

The sponsor’s result should still be read with its uncertainty. The confidence interval includes zero, and the p-value is 0.444. These values describe the statistical evidence and precision for the stated endpoint and analysis; they do not prove equivalence, rule out every meaningful difference, or answer questions about the separate dose arms. Further details about the complete statistical plan and model specification are not reported in the registry record.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists seven secondary outcomes, nine outcome measures, and 16 analyses. The primary result is one component of a broader set of analyses. When several outcomes or comparisons are examined, the chance of observing a seemingly notable result somewhere among them can increase. In general, multiplicity procedures are used to control or describe this issue. Whether a procedure was planned or applied to the primary and secondary analyses is not reported in the registry record. The p-value for the primary endpoint should therefore be understood as the reported value for that analysis, without assuming an unreported adjustment.

The primary analysis population is the Evaluable Analysis Set: randomized participants who received at least one dose of study medication and had a baseline and at least one post-baseline HbA1c assessment. The posted arm estimates are reported for Total Exenatide Twice Daily and placebo, with analysis counts of 78 and 42, respectively. Those analysis counts describe the displayed primary results; they should not be treated as a replacement for the registry’s enrollment figure or as a complete account of every analysis population.

For the MMRM, the registry states that measurements after initiation of rescue medication and study-drug discontinuation were excluded. It does not explain how every other missing value was handled, whether sensitivity analyses were conducted, or how departures from model assumptions were assessed. In general, missing-data methods rely on assumptions about the relationship between observed and unobserved outcomes. Those assumptions are relevant because the analysis may be sensitive to why an outcome is missing. The registry’s description is not sufficient to determine the robustness of the primary result to alternative missing-data assumptions.

The record does not report the sample-size justification, interim-analysis plan, complete statistical-analysis plan, or a full account of multiplicity control. The registry notes that an efficacy-analysis decision was made with agreement by the European Medicines Agency and the Food and Drug Administration, but the available population description is incomplete and does not state the details of that decision. These omissions limit what can be concluded about the prespecification and broader analysis context; they do not alter the numerical primary result reported in the registry.

13. 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: pregnancy (2.7%), cellulitis (2.4%), gastroenteritis (2.4%), staphylococcal abscess (2.4%), intentional self-injury (2.4%).

The registry reports serious adverse events by the exenatide 5 µg, exenatide 10 µg, and placebo groups. The number at risk and the number affected were 41 and 3 for exenatide 5 µg, 37 and 1 for exenatide 10 µg, and 42 and 1 for placebo. No deaths were reported in these groups. These are counts, not a statistical comparison, and they should be read in the context of the group sizes and the trial’s follow-up.

The record also lists other adverse events affecting 26 participants in the exenatide 5 µg group, 24 in the exenatide 10 µg group, and 23 in the placebo group. Serious adverse events and other adverse events are distinct categories in the registry. The serious-event counts should not be confused with the listed terms: the record gives Pregnancy at 2.7%, and Cellulitis, Gastroenteritis, Staphylococcal abscess, Intentional self-injury, and Gastroenteritis viral each at 2.4%. The listed terms are not assigned to specific treatment groups in the available record.

In general, adverse-event counts describe events recorded during a study and do not by themselves establish that a treatment caused an event. The registry information quoted does not provide a formal between-group safety analysis, adjusted comparisons, or detailed event narratives. Small count differences should not be interpreted as evidence of a safety advantage or disadvantage without those additional details. The figures describe the posted safety data, not a clinical recommendation.

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

The primary result is an adjusted estimate of HbA1c change at Week 28 for Total Exenatide Twice Daily compared with placebo. The reported least-squares mean difference is -0.28 %HbA1c, and its 95% confidence interval runs from -1.01 to 0.45. The interval includes zero, while the reported p-value is 0.444. Together, these statistics indicate that the estimate is uncertain and that the registry result does not precisely distinguish a group difference from no difference under the stated analysis.

This does not establish that the groups were equivalent. Equivalence requires a prespecified margin and an analysis designed to assess whether the interval lies within that margin; such a margin and analysis are not reported in the registry record. Nor does the result show that every possible difference is absent. The interval includes a range of values, and its breadth matters when assessing what the data can support.

The primary analysis combines exenatide 5 µg and exenatide 10 µg into Total Exenatide Twice Daily for the reported comparison. It therefore does not provide a separate primary estimate for either dose. The available primary endpoint also concerns HbA1c change, not all secondary outcomes, adverse events, or broader aspects of health. The registry lists secondary outcomes, but their full results are not included in the registry record describing the primary analysis.

The result applies to the defined Evaluable Analysis Set, not automatically to every participant enrolled or to populations with different characteristics. The MMRM excluded measurements after rescue medication and study-drug discontinuation. The record does not provide enough information about other missing data, model assumptions, or sensitivity analyses to judge how dependent the estimate is on those choices.

More generally, a treatment estimate is not a guarantee of an individual response. The reported group-level effect and its uncertainty summarize the trial’s analysis; they do not predict the outcome for any particular person. Nor do the registry results alone establish how the findings should guide care. Their appropriate statistical interpretation is limited to the population, endpoint, timeframe, and analysis reported.

15. Why This Trial Matters Statistically

This trial is statistically informative because it separates a treatment-effect estimate from the uncertainty around that estimate. The primary contrast is -0.28 %HbA1c, but the 95% confidence interval extends from -1.01 to 0.45. Reading only the point estimate would hide the degree of uncertainty. The interval indicates that the data are compatible with a range of possible effects under the analysis model, including no difference.

The design also illustrates why the reported analysis population and model matter. Although the trial was randomized, the primary outcome was analyzed in an Evaluable Analysis Set that required receipt of study medication and at least one post-baseline HbA1c assessment. The registry’s MMRM further excluded measurements after rescue medication and study-drug discontinuation. These choices define the data contributing to the estimate and should be considered when interpreting the result.

The combined exenatide comparison is another important feature. The primary result concerns Total Exenatide Twice Daily versus placebo, while the study had separate 5 µg and 10 µg exenatide arms. A combined estimate cannot be treated as a dose-specific estimate. The registry does not report separate primary contrasts for those doses.

Finally, the registry identifies multiple secondary outcomes and analyses but does not report the complete multiplicity plan or full statistical-analysis plan. This highlights the value of interpreting a result within the full analysis context rather than isolating a single p-value. The trial’s posted primary result offers a defined estimate for a defined endpoint, while the limits of the registry description constrain what can be inferred about analysis choices and the broader evidence.

16. 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 registry’s primary result combines the exenatide 5 µg and 10 µg arms as Total Exenatide Twice Daily. It does not provide a separate primary HbA1c contrast for each dose, so the reported effect should not be attributed to either dose individually. The group summaries also do not provide confidence intervals for each arm estimate.

The Evaluable Analysis Set excludes randomized participants who did not receive at least one dose or did not have both baseline and post-baseline HbA1c measurements. The registry does not provide a complete account of the primary outcome in an alternative population, such as all randomized participants. The reported MMRM excludes measurements after rescue medication and study-drug discontinuation, but the record does not fully describe the treatment of other missing observations or report sensitivity analyses. This limits assessment of robustness to missing-data assumptions.

The registry description of the efficacy population is truncated after referring to a decision made with agreement by the European Medicines Agency and the Food and Drug Administration. The details of that decision are not reported in the registry record. The full model specification, including the model terms and covariance structure, is also not reported, so the analysis cannot be independently reconstructed from the description alone.

Although the registry lists multiple secondary outcomes and analyses, it does not report a complete multiplicity strategy or the full statistical-analysis plan. The rationale for the enrollment target and any interim-analysis plan are not reported in the registry record. These omissions matter when judging prespecification and the wider context of the reported p-value.

Safety information is presented as counts and listed terms, without a formal comparative analysis in the available record. The term percentages are not mapped to treatment arms. As a result, the listed figures do not establish comparative risks or causal relationships. These are limitations of the available registry description, not evidence that an unreported analysis did or did not occur.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Safety and Efficacy of Exenatide as Monotherapy and Adjunctive Therapy to Oral Antidiabetic Agents in Adolescents With Type 2 Diabetes — NCT00658021.
• Registry record last updated 2020-12-01; status last verified 2020-10.
• Results first posted 2020-12-01.
• 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.

20. Record Summary

NCT00658021 is recorded as a completed Phase 3, randomized, double-masked, parallel-group trial with enrollment of 122. It studied subcutaneous exenatide 5 µg, exenatide 10 µg, and placebo, each administered twice a day, in adolescents with type 2 diabetes. The primary objective concerned superiority in change in HbA1c after 28 weeks.

The primary results compare Total Exenatide Twice Daily with placebo. Adjusted least-squares mean changes were 0.11 and 0.38 %HbA1c, respectively. The reported model-based difference was -0.28, with a 95% confidence interval from -1.01 to 0.45 and p = 0.444. The interval includes zero, indicating substantial uncertainty about the direction and magnitude of the group difference.

The efficacy analysis used an Evaluable Analysis Set and an MMRM that excluded measurements after rescue medication and study-drug discontinuation. The registry lists secondary outcomes and safety counts, but complete details of the statistical plan, multiplicity handling, and missing-data sensitivity analyses are not reported in the registry record.